Streptococcus agalactiae Infects Glial Cells and Invades the Central Nervous System via the Olfactory and Trigeminal Nerves

Streptococcus agalactiae causes neonatal meningitis and can also infect the adult central nervous system (CNS). S. agalactiae can cross the blood-brain barrier but may also reach the CNS via other paths. Several species of bacteria can directly invade the CNS via the olfactory and trigeminal nerves, which extend between the nasal cavity and brain and injury to the nasal epithelium can increase the risk/severity of infection. Preterm birth is associated with increased risk of S. agalactiae infection and with nasogastric tube feeding. The tubes, also used in adults, can cause nasal injuries and may be contaminated with bacteria, including S. agalactiae. We here investigated whether S. agalactiae could invade the CNS after intranasal inoculation in mice. S. agalactiae rapidly infected the olfactory nerve and brain. Methimazole-mediated model of nasal epithelial injury led to increased bacterial load in these tissues, as well as trigeminal nerve infection. S. agalactiae infected and survived intracellularly in cultured olfactory/trigeminal nerve- and brain-derived glia, resulting in cytokine production, with some differences between glial types. Furthermore, a non-capsulated S. agalactiae was used to understand the role of capsule on glial cells interaction. Interestingly, we found that the S. agalactiae capsule significantly altered cytokine and chemokine responses and affected intracellular survival in trigeminal glia. In summary, this study shows that S. agalactiae can infect the CNS via the nose-to-brain path with increased load after epithelial injury, and that the bacteria can survive in glia.


INTRODUCTION
Cranial nerves extending between the nasal cavity and the brain constitute a route by which certain microbes can invade the central nervous system (CNS). These nerves are the olfactory nerve and the intranasal branches of the trigeminal nerve. The cell bodies of sensory (primary) olfactory neurons are localized in the olfactory neuroepithelium of the nasal cavity. Their dendrites extend into the nasal epithelium, where odorant molecules are detected, and their axons reach all the way into the olfactory bulb in the brain, where they connect with second order neurons. The axons of sensory olfactory neurons together constitute the highly fasciculated olfactory nerve. The olfactory nerve is thus a direct connection between the periphery and the brain (Graziadei and Monti Graziadei, 1985). The anatomy of the trigeminal nerve is distinctly different; the cell bodies of its sensory neurons are found in the trigeminal ganglia, and there are only trigeminal axons near the nasal epithelium. The distal ends of these axons, however, are localized very close to the apical surface of the nasal epithelium (Schaefer et al., 2002). Thus, both the olfactory and the trigeminal nerves constitute potential invasion paths for infectious agents  ( Figure 1A).
The nasal epithelium constitutes a physical barrier against microbial infection of underlying nerves, and exhibits powerful innate and adaptive immune system components, supported by the nasal-associated lymphoid tissue (NALT) (Jochems et al., 2019). Should microbes penetrate this layer, they encounter the glia of the olfactory and trigeminal nerves, olfactory ensheathing cells (OECs) and trigeminal Schwann cells (TgSCs), respectively. OECs and TgSCs have innate immune functions and are considered to be the key phagocytes in these nerves (Panni et al., 2013). The astrocytes of the glia limitans layer (the A B D E C FIGURE 1 | S. agalactiae can invade the CNS via the olfactory nerve but infects the trigeminal nerve only after epithelial injury. Mice were treated with either methimazole or vehicle (PBS), followed by intranasal inoculation with S. agalactiae (1 x 10 6 CFU) or vehicle (also PBS). (A) Schematic of sagittal view of mouse brain, with olfactory mucosa (OM) and olfactory nerve (green), trigeminal nerve (Tg; blue), olfactory bulb (OB; orange), and brain. Tissue of Tg was collected from region shown by blue arrow. Graphs show the amount of S. agalactiae isolated from tissues 4 h and 24 h after intranasal inoculation for the four groups, (1) PBS ! PBS (data not shown as no bacteria was isolated, Supplementary Data Table 1), (2) methimazole ! PBS (data not shown as no bacteria was isolated, Supplementary Data Table 1), (3) PBS ! S. agalactiae and (4) methimazole ! S. agalactiae. (B) olfactory mucosa, (C) olfactory bulb, (D) trigeminal nerve and (E) brain. Data are pooled results from two independent biological experiments and are shown as median and interquartile range (4 h timepoint = 9 mice per group, 24 h timepoint = 12 mice). The S. agalactiae load from the different groups were compared using two-way ANOVA with significance determined using Tukey's multiple comparison test, *p ≤ 0.05, ***p ≤ 0. 001. For mice with zero bacterial counts, these were assigned a value of 1 in order to plot on log 10 y-axis.
physical and immunological barrier between the peripheral nervous system and the CNS), where the nerves connect with the olfactory bulb/brainstem, are also innate immune cells and constitute a "third layer of defence" against invasion of the CNS by infectious agents (Nazareth et al., 2019). Thus, the cranial nerves and the CNS are well protected from microbial invasion, and only a small number of pathogens are thought to infect the CNS via these paths. Of pathogens that can use these paths, the ability to escape glial phagocytosis, and instead infect and survive inside glia, has been suggested to be a key mechanism for bacterial infection of the CNS via cranial nerves (Macedo-Ramos et al., 2011;Dando et al., 2014;Walkden et al., 2020). Injuries to the nasal epithelium, for example caused by allergies, localized viral infections and mechanical trauma, are relatively common (Upadhyay and Holbrook, 2004) and may expose the underlying cranial nerves to infection. Experimental injuries to the nasal epithelium of mice have been shown to increase the risk of bacterial invasion of the olfactory nerve and bulb (Walkden et al., 2020). Premature infants are frequently fed with nasogastric tubes or intubated with nasal continuous positive airway pressure (nCPAP), which are associated with nasal injuries (Imbulana et al., 2018). Nasogastric feeding tubes are frequently contaminated with potentially pathogenic bacteria at the prongmucosal interface (Petersen et al., 2016); thus, preterm infants may be at risk for contracting CNS infection via the nose-to-brain path.
Streptococcus agalactiae (group B streptococcus, GBS) is the leading cause of neonatal meningitis, and sometimes meningoencephalitis, affecting 0.5-3 infants per 1000 live births. S. agalactiae infection has high mortality and leads to permanent neurological deficits in approximately half of surviving infants (Chin and Fitzhardinge, 1985;Horvath-Puhoé t al., 2021). Transmission is typically vertical (from the genouritary tract of the mother to the naso-and oropharynx of the child during delivery) (Yadeta et al., 2018), but horizontal transmission can also occur (Morinis et al., 2011). S. agalactiae infection can cause two clinical syndromes: early-onset disease (EOD) in the first week of life, or late-onset disease (LOD), occurring in infants aged one week to three months. Infants born prematurely also have three-to 30-fold increased risk of developing both EOD and LOD than full-term infants (Melin, 2011;Nanduri et al., 2019;Mynarek et al., 2021). It has been suggested that contaminated feeding tubes may be a source of transmission for S. agalactiae (Le Doare and Kampmann, 2014;Jauneikaite et al., 2018). After delivery of breast milk via nasogastric feeding tubes, preterm infants have been shown to develop LOD, sometimes with meningitis (Olver et al., 2000).
Although less commonly, S. agalactiae can also cause meningitis in adults, with increasing incidence in recent years. Nosocomial S. agalactiae disease in adults can arise from new acquisitions or from pre-existing colonization of skin and mucosal tissues, and droplet transmission has been described. S. agalactiae meningitis particularly affects people with severe underlying conditions, can be fatal or result in serious sequelae (Gupta et al., 2018). Gastric tube feeding has also led to increased amounts of S. agalactiae in the oral flora of elderly patients, suggesting tube contamination (Takeshita et al., 2011).
S. agalactiae can cross the blood-brain barrier (Maisey et al., 2008), but the mechanisms for CNS invasion have not been fully elucidated and other paths may also be involved . S. agalactiae can colonize the nasal epithelium in infants (Foster-Nyarko et al., 2016) and, whilst specific data regarding colonization of the nasal mucosa in adults is lacking, S. agalactiae can be cultured from the oropharynx and/or respiratory tract of~20-25% of adults (Roloff et al., 2018). Because other commensal bacteria in the nasopharynx that cause CNS disease can reach the brain via the nose-to-brain cranial nerve path, we hypothesized that this path constitutes a potential mechanism by which S. agalactiae can infect the CNS. S. agalactiae also exhibits capacity for intracellular survival in a variety of cell types, including macrophages, neutrophils and microvascular endothelial cells (Cumley et al., 2012); thus, perhaps S. agalactiae may also survive in glia.
In the current study, we investigated whether S. agalactiae sequence type 17 (ST17); serotype III, which is epidemiologically the most relevant in neonatal (Edmond et al., 2012) and adult (Paveenkittiporn et al., 2020;Vuillemin et al., 2020) Group B streptococcal meningitis, can infect the brain via the olfactory and/or trigeminal nerves in mice. Because epithelial injury is associated with increased infection of the olfactory nerve, we also determined the effects of prior experimental injury to the nasal epithelium on S. agalactiae infection via this path. As the ability to infect glia is thought to be important for bacterial invasion of both cranial nerves and the brain, we also determined how the key glial types in the olfactory/trigeminal nerves and glia limitans layer responded to S. agalactiae.

Bacterial Strains and Culture Conditions
Wild-type (WT) S. agalactiae strain 874391, which is a serotype III isolate and a member of the hypervirulent ST17 lineage (Sullivan et al., 2017), and an isogenic DcpsE strain  deficient in capsule synthesis, were routinely grown at 37°C in Todd-Hewitt broth (Thermo Fisher Scientific) with shaking at 180 rpm for 16 h. Retrospective colony counts for cultures were performed by serial dilution in phosphate-buffered saline (PBS) and plating onto tryptone soya agar plates or selective media as described below.
intranasally inoculated with 10 mL containing 1x10 6 colony-forming units (CFUs) of WT S. agalactiae suspended in PBS, or vehicle (PBS only) as described previously for other bacteria (Nazareth et al., 2021). Mice were sacrificed 4 h and 24 h post inoculation by rising CO 2 asphyxiation and tissues (olfactory mucosa, olfactory bulb, brain, trigeminal nerve (from within the cranial cavity), lungs, liver and blood) were collected. Tissues were homogenized using a TissueLyzer (QIAGEN) using 5 mm tungsten carbide beads and homogenates were suspended in 1 mL PBS for colony count assays by plating serial dilutions on CHROMID Strepto B agar (Biomerieux Cat number: 43461) or fixed for use in immunohistochemistry assays according to our established protocols (Nazareth et al., 2021).
For S. agalactiae infection post methimazole-induced epithelial injury, 7-10 weeks old female BALB/c mice (Animal Resource Centre) under anaesthesia with isofluorane (1.5-2%) were injected with methimazole (Sigma-Aldrich, 50 mg/kg, 10 mg/ml in PBS) or vehicle (PBS) using intraperitoneal injection according to our published protocol (Walkden et al., 2020). The administration of methimazole causes tissue-specific death of olfactory neurons and supporting cells secondary to degeneration of the olfactory epithelium in rodents which constitutes a more robust model of olfactory nerve injury than that caused by other chemicals (Herbert et al., 2012). Three days later when neuroepithelial damage peaks, animals were intranasally inoculated with WT S. agalactiae or vehicle as described above. Mice were sacrificed 4 h and 24 h post inoculation and used for colony count or immunohistochemistry assays as described above.

Culture of Primary Glia
Primary glia cultures were obtained from S100b-DsRed transgenic mice (Windus et al., 2007), in which all glia express the bright red protein DsRed, as previously described . Briefly, the olfactory bulb and trigeminal ganglia were dissected out for preparations of OECs and TgSCs, respectively, from postnatal day 7 pups. Tissue explants were plated into 24-well plates, with wells pre-coated with Matrigel basement membrane matrix (Corning Matrigel Basement Membrane Matrix, FAL354234). The explants were maintained in glial medium, constituting of Dulbecco's Modifed Eagle Medium (DMEM) (Gibco) with 10% fetal bovine serum (FBS), 50 mg/mL gentamycin (Gibco), 200 mM L-glutamine (Gibco) and G5 supplement diluted as per manufacturer's instructions (Gibco). This established method yields 80% DsRed-positive cells which are also positive for glial markers (the p75 neurotrophin receptor, s100) .
Primary astrocytes with approximately 70-80% purity were obtained from the cortices of postnatal day 3 pups following a previous published protocol (Schildge et al., 2013). The cortices were dissected out and cut into small pieces in glial medium. The cortical cell suspension was then plated in a poly-D-lysine hydrobromide (Sigma-Aldrich P6407)-coated T75 flask. Following 7 to 8 days (90% confluency), astrocytes were separated from microglia and oligodendrocyte precursor cells by shaking the flask on an orbital shaker as previously described (Schildge et al., 2013). All the glia were cultured to 80% confluency after which they were trypsinized (using Gibco TrypLE Express, 1X) and used for experiments.

In Vitro Infection of Primary Glia
To analyse the interactions of WT S. agalactiae and the capsuledeficient DcpsE mutant with glia, we exposed cultured primary OECs, TgSCs and astrocytes to the bacteria (n = 3 biological and 3 technical replicates for each condition/cell type). Methods for the routine culture and cell infection of cells were performed as previously described (Nazareth et al., 2021) with the following modifications: glia were seeded in 96-well plates at 6000 cells per well and incubated at 37°C in 5% CO 2 until approximately 80% confluence was achieved. Glial cell monolayers were then infected with S. agalactiae at a multiplicity of infection (MOI) of 100 bacteria per cell (or, for control, medium alone was added to the cultures) for 1 h. The plate was then incubated at 37°C in 5% CO 2 . Bacteria were initially allowed to attach to glia for 1 h, before the infecting inoculum was aspirated and unattached bacteria was removed by washing the well three times with PBS. Then, adherent bacteria were enumerated by colony counts on Todd-Hewitt agar.
To determine the concentration of intracellular bacteria, antibiotic protection assays were utilised (Dando et al., 2016). Adherent extracellular bacteria were removed by the addition of glial medium containing antibiotics (penicillin 250 U/mL and streptomycin 250 U/mL with gentamycin 50 µg/mL, Gibco) and the number of intracellular bacteria was determined at 2 h and 24 h post infection.

Immunocytochemistry
Glial cells were seeded onto 96-well plates at 6000 cells per well and the following day were infected with WT or DcpsE S. agalactiae, or medium alone (control). Infected monolayers at various times were fixed with 4% PFA for 10 min, rinsed with PBS, followed by blocking/permeabilising solution (3% BSA in PBS with 0.3% Triton X-100) for 30 min at room temperature on a shaker.

Imaging
All high magnification images were collected using an Olympus FV3000 confocal microscope and panels show confocal images of maximum projection of z-stacks. Images were colour-balanced using Adobe Photoshop CS5 (Adobe Systems Incorporated) with the figures being compiled in Adobe Illustrator CS5 (Adobe Systems Incorporated). 3D reconstruction of z-stacks were generated and videos were compiled using Imaris 9.5.1 software.

Cytokine and Chemokine Assays
The levels of cytokines were determined in the cell culture supernatants at 24 h post inoculation with Bio-Plex Pro Mouse Cytokine 23-plex assays (Bio-Rad, Gladesville, NSW, Australia). Cell culture supernatants were collected and centrifuged at 1,000 × g for 15 min at 4°C. Samples were then stored at −80°C until analysis. Thawed samples were then assayed according to the manufacturer's instructions. The plates were read on a BioPlex 200 Luminex bead array reader (Bio-Rad) and data were acquired with Bio-Plex Manager Software (version 5.0; Bio-Rad).

Data Analysis
Statistical testing and graphical analysis were conducted using GraphPad Prism version 8.0. Colony count data from organ lysates and from cell lysates were examined by Kruskal-Wallis Test followed by Dunn's multiple comparison post-tests. Cytokine and chemokine expression data were compared using two-way analysis of variance (ANOVA) followed by Tukey's multiple comparison post-hoc tests.

S. agalactiae Can Invade the CNS via the Olfactory Nerve
To determine whether S. agalactiae could invade the CNS via the olfactory nerve and/or the intranasal branches of the trigeminal nerve, we intranasally inoculated mice with S. agalactiae (10 6 bacteria in 10 mL) and determined whether live S. agalactiae bacteria could be isolated from mucosal, nerve and CNS tissues. Control inoculations contained PBS only (vehicle). At 4 h and 24 h post inoculation, the mice were sacrificed and homogenates of the olfactory mucosa, olfactory bulb, trigeminal nerve, brain (beyond the olfactory bulb; Figure 1A), lungs, liver and blood (Supplementary Data, Table 1) were analysed for S. agalactiae load. For the control group (PBS ! PBS), no bacteria were isolated from any of the tissues (olfactory mucosa, olfactory bulb, trigeminal nerve, brain, lungs, liver and blood; Supplementary Data, Table 1 and Figures 1B-E). For the S. agalactiae infected group at 4 h, median tissue load was 4 log 10 CFU/0.1g tissue (interquartile range (IQR) 3.3-4.4) in olfactory mucosa and no bacteria were present in all the other organs tested ( Figures 1B-E). For the S. agalactiae infected group at 24 h post infection, median tissue loads were 3.6 log 10 CFU/0.1g tissue (IQR 0.8-4.3) in olfactory mucosa, 3 log 10 CFU/0.1g tissue (IQR 0-3.8) in the olfactory bulb, and 1.6 log 10 CFU/0.1g tissue (IQR 0.2-2.2) in the brain ( Figures 1B-E). Interestingly, however, no S. agalactiae bacteria were isolated from the trigeminal nerve, lungs, liver, and blood 4 h and 24 h post infection ( Figure 1D; Supplementary Data Table 1).

Injury to the Nasal Epithelium Exacerbates Infection via the Olfactory Nerve Route and Results in Infection of the Trigeminal Nerve
As injury to the olfactory epithelium has been shown to increase the amount of bacteria infecting the olfactory nerve/bulb (Walkden et al., 2020), in parallel assays to those described above we pre-injured the olfactory neuroepithelium with our well-established method using methimazole (Walkden et al., 2020), which causes death of primary olfactory neurons and damage to the olfactory epithelium. Three days later, we inoculated the mice intranasally with S. agalactiae. At this time, death of olfactory neurons is at its peak (Nazareth et al., 2015) and methimazole has been cleared, limiting potential unknown side-effects of methimazole (Xie et al., 2011). Epithelial injury significantly increased the bacterial load isolated from the olfactory mucosa, and brain ( Figures 1B, C, E) at 4 h and 24 h, compared to parallel infection without injury. This also resulted in recovery of S. agalactiae from the trigeminal nerve with median 3.4 log 10 CFU/0.1g tissue (IQR 2.9-3.7) at 4 h and median 3.4 log 10 CFU/0.1g tissue (IQR 0.7-4) at 24 h post infection ( Figure 1D). Also, no S. agalactiae bacteria were isolated from the lungs, liver, and blood 4 h and 24 h post infection. Thus, S. agalactiae can infect the CNS via the olfactory nerve, and epithelial injury resulted in significantly elevated numbers of invading bacteria in the brain and olfactory bulb. Epithelial injury also exposes the trigeminal nerve for S. agalactiae infection.
To further examine the dynamics of S. agalactiae invasion via the CNS, we immunolabelled tissue sections from the olfactory epithelium, bulb and brain for S. agalactiae. In the control uninfected animals (PBS ! PBS), the olfactory epithelium was an intact layer of dense cells (Figure 2A). After methimazole treatment alone (methimazole ! PBS), the epithelial layer sloughed off into the nasal cavity ( Figure 2B). When S. agalactiae bacteria were intranasally inoculated after methimazole injury (methimazole ! S. agalactiae), bacteria were readily detected within the remaining tissue of the olfactory epithelium as well as within the exudate in the nasal cavity ( Figures 2D-F). We could not detect S. agalactiae from the olfactory mucosa of uninjured mice ( Figure 2C), but, consistent with a~100-fold reduction in the number of culturable bacteria from the uninjured vs the injured group (3.6 vs 5.3 log 10 CFU/0.1g tissue, respectively), this is likely due to sparse distribution of bacteria within the tissue. In the other tissues analysed, we could not clearly distinguish S. agalactiae cocci from background immunolabelling and thus, we only show immunolabelling data for the olfactory epithelium.

S. agalactiae Can Invade and Survive Inside Glia
Glia (OECs and TgSCs) are the key phagocytes in the olfactory and trigeminal nerves, respectively, whilst astrocytes (CNS glia) in the glia limitans layer between the nerves and the CNS constitute another layer of defence against pathogens. As the capability to infect these glia is thought to constitute a key mechanism for CNS invasion, we therefore determined the susceptibility of these glia to S. agalactiae infection. Using antibiotic protection assays, we compared the adhesion and invasion capacity of S. agalactiae between primary OECs, TgSCs and astrocytes. To investigate bacterial adhesion to the cells, we exposed the cells to S. agalactiae at a multiplicity of infection of 100 bacteria per cell (MOI 100) for 1 h and determined the number of bacteria that had attached to the cells. To determine whether S. agalactiae could invade and survive within OECs, TgSCs and astrocytes, an antibiotic cocktail was added to the medium to kill extracellular bacteria. Cells were then incubated for a further 2 h or 24 h and the number of intracellular bacteria was determined for the different cell types by lysing monolayers at these timepoints. We found that S. agalactiae adhered to the cells within 1 h ( Figure 3A), and bacteria could be recovered from all glia at 2 h post antibiotic addition, representing invasive S. agalactiae ( Figure 3B). The frequency of invasion (% of adhered bacteria recovered at 2 h post antibiotics) was similar for the three glial types (Supplementary Figures 1A, B). After 24 h, however, bacteria were only recovered from OECs and astrocytes, and not in TgSCs ( Figure 3C and  Supplementary Figure 1C) suggesting that S. agalactiae was eliminated from TgSCs at 24h post inoculation. To assess the cytotoxicity of S. agalactiae on glial cells, percentage of dead cell count and total cell count was performed at 1 h and 24 h post infection. No significant cytotoxicity was observed in glial cells (OECs, TgSCs and astrocytes) post infection (Supplementary Figure 2). Thus, the trends observed in the results are due to glial responses to bacteria and not due to cell death occurring post infection. To resolve the interactions of S. agalactiae with OECs, TgSCs and astrocytes at the cellular level, we fixed and imaged infected cells at 1 h, 2 h and 24 h time-points. The glia was isolated from S100b-DsRed mice, in which the S100 promoter drives the expression of the fluorescent protein DsRed in glia, allowing easy visualisation of the cells. The expression of DsRed is lower in astrocytes than in the other glia; therefore, the identity of these cells was verified using immunolabelling for glial fibrillary associated protein (GFAP), a well-established astrocyte marker ( Although bacterial counts were below the detection limit from TgSCs at 24 h post inoculation ( Figure 3C), we could occasionally find S. agalactiae cocci in some TgSCs using imaging ( Figure 4K). We observed that some of the cells were bi-or multinucleated, or had atypical nuclei ( Figures 4I, K, L).

The S. agalactiae Capsule Contributes to Invasion and Survival Within Trigeminal Schwann Cells
Lack of S. agalactiae polysaccharide capsule may alter colonization and cellular responses to the bacteria (Alkuwaity et al., 2012;Sullivan et al., 2016). Therefore, in parallel assays, we compared glial responses to an isogenic S. agalactiae mutant in cpsE (DcpsE) that is devoid of capsule . We found that colony counts of adhesion, invasion and intracellular survival for the DcpsE strain were similar to WT for OECs, TgSCs and astrocytes ( Figures 3A-C); while the two-way ANOVA analysis did not reveal significant differences between any of the treatment groups, for the 24 h intracellular survival in TgSCs there was a p value of 0.08 for the comparison between DcpsE strain and WT. Interestingly, with a comparison of percentage of bacteria load to initial inoculum, there was significantly more DcpsE S. agalactiae recovered from TgSCs at 24 h compared to WT (p ≤ 0.001, Supplementary Figure 1C). In addition, DcpsE bacteria also had significantly higher capacity for intracellular survival in TgSCs than in astrocytes (both WT and DcpsE, p≤ 0.01) and also compared to OECs (WT only, ≤ 0.01) (Supplementary Figure 1C). Imaging confirmed the DcpsE strain could adhere to, invade and survive inside all three glial types ( Figure 5).

Phagocytosis of S. agalactiae in Phagolysosomes in Glial Cells
To determine the fate of S. agalactiae after internalisation within the glia, we immunolabelled the cells at 24 h post infection for lysosome-associated membrane protein 2 (LAMP-2). LAMP-2 is the major protein component of the lysosome membrane which is required for the fusion between the late phagosome and lysosomes as well as acidification of the lysosomal lumen that helps in degradation of the internalized foreign particles such as bacteria (Eskelinen et al., 2002;Huynh et al., 2007). We observed that all glial cells (OECs, TgSCs and astrocytes) showed welldefined LAMP2-positive lysosomes around both WT and DcpsE S. agalactiae (Figures 6, 7) demonstrating that the bacteria were internalized or colocalized inside lysosomes after infection. To support the above statement 3D reconstruction and videos were generated which showed co-localisation of bacteria and LAMP within the cells (Supplementary Video 1 (OECs), Video 2 (TgSCs) and Video 3 (astrocytes).

Glia Respond to S. agalactiae by Secreting Cytokines and Chemokines
To gain an understanding of the innate immune responses of the different glia to S. agalactiae, we also analysed secretion of cytokines and chemokines. As the S. agalactiae capsule has A B C

FIGURE 3 | Comparison of adhesion/invasion of glia by S. agalactiae. (A)
Cell monolayers were inoculated with WT or DcpsE bacteria for 1 h (MOI 100), followed by colony counts to determine the number of adherent bacteria. (B, C) After 1 h, penicillin, streptomycin and gentamicin were added to the medium to kill extracellular bacteria, and at 2 h (B) or 24 h (C) cells were lysed to recover intracellular bacteria and colony counts were performed. Data shows mean ± SEM of three independent experiments with n = 3 technical replicates. Data were compared between glial types using two-way ANOVA with Tukey's multiple comparison test; no significant differences were detected. CFU/mL values of zero were assigned a value of 1 to enable display on log10 axes (in graph (C), TgSCs WT value is zero). previously been found to alter cytokine production by other cell types (Lemire et al., 2012b), we also assessed whether the capsule affected secretion of cyto-/chemokines. We inoculated the cells with both S. agalactiae WT (white bars) and DcpsE strain (grey bars) and cyto/chemokine levels were analysed at 24 h post inoculation.

Cytokines
Infection of all the glial types with S. agalactiae WT and DcpsE strain resulted in the production of numerous pro-inflammatory cytokines (Figure 8), anti-inflammatory cytokines (Figure 9), and chemokines ( Figure 10) at 24 h post-inoculation (all at significantly higher levels than non-infected control glia).  The glial cytokine response included pro-inflammatory cytokines previously reported to be associated with S. agalactiae infection (Patras and Nizet, 2018), such as interleukin 1b (IL-1b), interleukin 17 (IL-17), interferon g (IFN-g) and tumour necrosis factor a (TNF-a). The glia also responded to both the S. agalactiae WT and DcpsE strain with secretion of the anti-inflammatory (regulatory) cytokines interleukin 4 (IL-4), interleukin 5 (IL-5), interleukin 6 (IL-6), interleukin 10 (IL-10) and interleukin 13 (IL-13) at 24 h post inoculation ( Figure 9). Overall, S. agalactiae-infected OECs also produced higher levels of both pro-and anti-inflammatory cytokines than TgSCs and astrocytes (Figures 8, 9). The glia exhibited differential cytokine responses to the wildtype and capsule-deficient S. agalactiae strains. All three glia  produced significantly higher levels of IL-1a ( Figure 8A), while OECs and TgSCs produced significantly higher levels of IL-10 and only TgSCs produced significantly higher levels of IL-12p70 in response to the DcpsE strain than to the WT strain (Figures 8,9). For TgSCs in particular, the DcpsE strain was generally associated with secretion of higher levels of cytokines than the WT strain (with individual variations depending on cytokine and/or cell type; Figures 8, 9). In response to DcpsE strain, TgSCs secreted significantly more IL-2, IL-12p40, IL-12p70, IL-17 than in response to WT. OECs, in contrast, secreted significantly lower levels of GM-CSF after exposure to the capsule-deficient strain than to the wild-type strain ( Figure 8K).

Chemokines
The glia also responded to S. agalactiae with the secretion of the chemokines C-X-C motif ligand 1 (CXCL-1, also known as Gro/ KC or Groa), eotaxin, monocyte chemoattractant protein 1 (MCP-1), macrophage inflammatory proteins (MIP-1a and MIP-1b, also known as CCL3 and CCL4) and regulated upon activation-normal T cell expressed and presumably secreted (RANTES) (Figure 10). At 24 h post inoculation, OECs secreted higher levels of CXCL-1, eotaxin, MCP-1 and RANTES than the other cell types ( Figure 10). Some differences in secretion of chemokines were found between cells that had been inoculated with the WT strain versus the DcpsE strain. At 24 h post inoculation, all three glial types secreted higher levels of MCP-1 in response to the capsuledeficient strain than to the wild-type strain, although only OECs had significantly higher levels. TgSCs also secreted more MIP-1a in response to the S. agalactiae DcpsE strain than to the wild-type strain ( Figure 10).

DISCUSSION
This study demonstrates that S. agalactiae can rapidly (within 24 h) infect the CNS via the olfactory nerve in mice, and that injury to the nasal epithelium exacerbates infection via this route. We also found that injury, often associated with nasal infections or mechanical trauma in humans, exposes the trigeminal nerve to infection. In cultured cells, we showed (1) that S. agalactiae can infect and survive inside glia (primary mouse OECs, TgSCs and astrocytes), (2) resulting in glial-mediated secretion of cytokines and chemokines, and (3) that the S. agalactiae capsule deficiency plays a major role in cellular interactions with TgSCs and cytokine/chemokine secretion. Together, these findings show that cranial nerves extending between the nasal cavity and the brain constitute an alternative path by which S. agalactiae can infect the CNS, which may explain why meningitis can occur in patients despite negative blood cultures (Garges et al., 2006). The outcomes also suggest that intracellular survival in glia constitutes a mechanism behind the capacity for invading the brain via these routes. The olfactory nerve has long been hypothesized to be a potential invasion route for pathogens causing bacterial meningitis (Filippidis and Fountas, 2009). Rapid infection of the meninges/CNS via this route has been demonstrated for Neisseria meningitidis (Sjolinder and Jonsson, 2010) and Streptococcus pneumoniae (van Ginkel et al., 2003), even in the absence of blood infection. Chlamydia pneumoniae, which is predominantly a respiratory pathogen, is also considered to infect the CNS via this path (Balin et al., 2018) and Treponema species (Riviere et al., 2002) appear to primarily reach the CNS via the trigeminal nerve. Chlamydia muridarum (Nazareth et al., 2021) (Chlamydia commonly used to model rodent chlamydia lung infection), Burkholderia pseudomallei Dando et al., 2016) and Listeria monocytogenes have also been shown to infect the CNS via both the olfactory and the trigeminal nerve routes in mice, with circumstantial evidence of trigeminal nerve/brainstem involvement in humans for the latter two (Drevets and Bronze, 2008;Pagelow et al., 2018). Certain viruses such as herpes simplex type 1 (HSV1) (Shivkumar et al., 2013;Doll et al., 2019) and SARS-CoV2 (Brouwer et al., 2020) are also considered capable of invading the brain via the olfactory and/or trigeminal nerve routes. The majority of these infectious agents can invade the CNS via the nerves rapidly (within 24-48 h) .
Infection of the olfactory and/or trigeminal nerve by B. pseudomallei  and N. meningitidis (Sjolinder and Jonsson, 2010) are associated with patchy injury to the nasal epithelium. In the current study, we did not observe epithelial damage when mice were infected with S. agalactiae alone; despite this, the bacteria were able to enter the olfactory nerve. This is in alignment with previous studies of olfactory nerve infection by S. pneumoniae (van Ginkel et al., 2003), but in contrast to studies on S. aureus, which requires epithelial injury to enter the nerve. From the trigeminal nerve, however, no bacteria were isolated without prior epithelial injury. In contrast, when the olfactory epithelium was This finding suggests that perhaps other cell types restricted to the olfactory neuroepithelium are involved in the mechanism by which S. agalactiae enters the olfactory nerve in the absence of injury. The capacity for intracellular survival in glia has been suggested as a key mechanism by which bacteria can infect the olfactory and trigeminal nerves. Intracellular survival in OECs and/or TgSCs has previously been shown for S. pneumoniae (Macedo-Ramos et al., 2014), C. muridarum (Nazareth et al., 2021), B. pseudomallei (Walkden et al., 2020), and N. meningitidis . The current study demonstrates that S. agalactiae can survive inside cultured OECs and astrocytes for at least 24 h. The ability to resist killing and to survive in OECs is likely important for olfactory nerve infection, and in astrocytes, for traversing the glia limitans layer and for entry into the CNS (Ruiz-Mendoza et al., 2016). Whilst intracellular survival in OECs has not been previously investigated for S. agalactiae, one previous study has shown that S. agalactiae (including serotype III) can survive in astrocytes for at least 12 h (Stoner et al., 2015), which is in accordance with our findings. We found that S. agalactiae is attenuated for intracellular survival in TgSCs compared to other glial types. This finding was surprising, given the fact that OECs have been shown to mount a much more powerful immune response to Escherichia coli and pathogenassociated molecular patterns (PAMPs) (Vincent et al., 2007), as well as exhibit higher levels of immune-and inflammatory-related factors on the transcriptional level (Vincent et al., 2005) than Schwann cells. It is important to note that these previous investigations used Schwann cells from the sciatic nerve, and not the trigeminal nerve; TgSCs may exhibit differential immune responses than other Schwann cells due to their anatomical location near the nasal cavity (Choudhury et al., 2021). A recent study, however, has shown that OECs are better protected against intracellular survival of C. muridarum (Nazareth et al., 2021), and are more efficient phagocytes of necrotic cellular debris, than TgSCs . It is thus possible that OECs and TgSCs may respond differently to distinct phagocytic targets and while FIGURE 8 | Pro-inflammatory cytokine responses in glia after inoculation with wild-type and capsule-deficient S. agalactiae. OECs, TgSCs and astrocytes were inoculated with either WT or DcpsE (MOI 100:1) after which the cytokine levels in the medium were analysed at 24 h post inoculation. Data show mean ± SEM, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 (two-way ANOVA with Tukey's multiple comparison test, n = 3 technical replicates x 100,000 cells/well).

Chacko et al.
S. agalactiae CNS Infection via Nerves colocalisation of LAMP2 with the internalised bacteria appeared similar across the different glial types, further investigations may reveal mechanisms that modulate intracellular survival or destruction of the bacteria in the different glia. Whilst the intact nasal epithelium is likely to be the key to protecting the trigeminal nerve from infection, the low capacity for survival in TgSCs may also contribute to the lack of trigeminal nerve S. agalactiae infection. Glial activation is one of the first hallmarks of nervous system infection and neuroinflammation, characterized by cytokine and chemokine release (Refolo and Stefanova, 2019). These factors are protective in that they stimulate clearance of infectious agents and cell debris, with some factors also directly promoting tissue repair. However, if inflammation is excessive and long-lasting, it can lead to neurotoxicity and tissue damage. The host inflammatory response to S. agalactiae contributes significantly to the pathogenesis of meningitis and CNS injury (Patras and Nizet, 2018). We here showed that S. agalactiae infection resulted in the production of many pro-and anti-inflammatory cytokines and chemokines by glia.
All three glial types responded with secretion of the proinflammatory cytokines IL-12, TNF-a, IL-1b, IL-17 and IFN-g, which are also secreted at high levels in mouse models of S. agalactiae sepsis (Patras and Nizet, 2018). IL-12, TNF-a and FIGURE 9 | Anti-inflammatory cytokine responses in glia after inoculation with wild-type and capsule-deficient S. agalactiae. OECs, TgSCs and astrocytes were inoculated with either WT or DcpsE (MOI 100:1) after which the cytokine levels in the medium were analysed at 24 h post inoculation. Data show mean ± SEM, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 (two-way ANOVA with Tukey's multiple comparison test, n = 3 technical replicates x 100,000 cells/well).
FIGURE 10 | Chemokine responses in glia after inoculation with wild-type and capsule-deficient S. agalactiae. OECs, TgSCs and astrocytes were inoculated with either WT or DcpsE (MOI 100:1) after which the chemokine levels in the medium were analysed at 24 h post inoculation. Data show mean ± SEM, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 (two-way ANOVA with Tukey's multiple comparison test, n = 3 technical replicates x 100,000 cells/well). IL-1b are key mediators of for the initial inflammation cascade in response to bacteria (Wennekamp and Henneke, 2008), including S. agalactiae (Patras and Nizet, 2018). IL-12, in particular, has key roles in combatting S. agalactiae infections (Patras and Nizet, 2018), including urinary tract infection  and sepsis (Landwehr-Kenzel and Henneke, 2014). High levels of TNF-a can cause damage to the brain tissue, but peripheral glia produce pituitary adenylate cyclase activating peptide (PACAP) (Hegg et al., 2003), which may protect against TNF-a-mediated damage (Kanekar et al., 2010). Production of pro-inflammatory cytokines was overall highest in OECs, followed by TgSCs and then astrocytes. Secretion of these factors may be important for OECs and TgSCs (peripheral glia) in preventing microbial invasion of the brain via peripheral nerves.
The pro-inflammatory glial responses were counter-balanced by the generation of the anti-inflammatory cytokines IL-6 and IL-10. IL-6 is important for regeneration/cell survival, whereas IL-10 overall limits inflammation (Landwehr-Kenzel and Henneke, 2014). Astrocytes induce transcription of IL-6 in response to S. agalactiae infection (Stoner et al., 2015), which is considered important for prevention of S. agalactiae-induced damage to the CNS tissue (Landwehr-Kenzel and Henneke, 2014). OECs and TgSCs have previously been shown to produce IL-12 and IL-6 in response to both Gram-negative and Gram-positive bacteria (Dando et al., 2016;Choudhury et al., 2021;Nazareth et al., 2021). Thus, secretion of IL-12 and IL-6 may be key mediators of the peripheral glial responses to a wide range of infectious agents. In the current study, we found that the glia also responded to S. agalactiae with secretion of several chemokines (usually secreted secondary to pro-inflammatory cytokines) that mediate recruitment of leukocytes (Graves and Jiang, 1995) and have important roles in neuroinflammation and neural repair (Miller et al., 2008).
S. agalactiae produces a polysaccharide capsule, which protects against opsonisation and, thus, opsonophagocytosis by several cell types, as well as suppresses oxidative burst and subsequent release of proteases by neutrophils (Carlin et al., 2009). S. agalactiae are classified into 10 different serotypes based on capsule composition (Afshar et al., 2011), however, 0.5% of clinical S. agalactiae isolates do not react with any diagnostic anti-capsular serum (Ippolito et al., 2010), and are considered non-typeable (NT). NT strains may either completely or in part lack the capsule. Human clinical isolates completely lacking the entire capsular locus have been identified using PCR (Creti et al., 2012). In a murine model of urinary tract infection, lack of the S. agalactiae type III capsule was shown to enhance bladder colonization .
In the current study, we showed that the capsule did not affect adhesion and intracellular survival by S. agalactiae in OECs and astrocytes. These findings suggest that capsule-independent mechanisms mediate attachment and internalization of S. agalactiae in these cells, as has previously been shown for dendritic cells in which both capsule-dependent and independent internalization mechanisms have been described (Lemire et al., 2012a). Regarding astrocytes, this finding contradicts a previous study, which showed that presence of the capsule impaired intracellular survival of S. agalactiae, however the study used S. agalactiae serotype I and not III (Alkuwaity et al., 2012). In TgSCs, however, intracellular survival of wild-type S. agalactiae was negligible, whereas the capsuledeficient mutant exhibited significant intracellular survival (in fact, higher than in astrocytes); thus, the capsule appeared to repress intracellular survival in TgSCs, however the mechanism is yet to be determined. It is worth noting that our assays used heat-inactivated bovine serum (inactivated complement system) in the medium; this would reduce S. agalactiae capsule interaction with the opsonizing effects of the complement system (Maisey et al., 2008;Patras and Nizet, 2018).
We found that the capsule also moderately affected the production of cyto/chemokines. Overall, the presence of the capsule repressed the production of multiple cyto/chemokines by the glia, but enhanced secretion of some specific ones by OECs. Overall, the capsule promoted secretion of some chemokines by OECs, whilst having the opposite effects on TgSCs and astrocytes. Thus, the interrelationship between cytokine production and intracellular survival remains unclear.
The study has several limitations, which are important to address in future investigations of how S. agalactiae invade the CNS via cranial nerves and infect glia. Long-term in vivo studies are required to determine S. agalactiae survival in the nerves and CNS, and whether S. agalactiae causes associated neuropathologies as has been shown for C. pneumoniae and HSV-1 (Balin et al., 2018) for example. The cell cultures had purities of 70-80% and thus the cytokine responses reflect the total cell population and not just the target glial cells; organoid type 3D cultures in future may reveal more in vivo like responses. Further investigations focussing on the potential for S. agalactiae to cause morphological changes, such as nuclear abnormalities, are also needed.

CONCLUSION
In the current study, S. agalactiae was shown to invade the CNS via the olfactory nerve in mice after intranasal inoculation. We also showed that epithelial injury was associated with increased infection of the olfactory nerve and also led to S. agalactiae infection of the trigeminal nerve. Finally, we showed that S. agalactiae also infected and survived intracellularly in the glia of both the olfactory and the trigeminal nerves and CNS. Importantly, this ability to infect glia may also constitute a key reason for the bacterial invasion of these cranial nerves and the brain.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
All procedures were approved by the Griffith University Biosafety Committee (NLRD/09/15_var7) and the Griffith University Animal Ethics Committee (MSC/08/18/AEC) in accordance with guidelines of the Australian Commonwealth Office of Gene Technology Regulator and the National Health and Medical Research Council of Australia.

AUTHOR CONTRIBUTIONS
Experiments were conducted by AC, AD, IC, TE, MS, and CG. Figures were prepared by AC, AD, JS, and JE. All authors analysed the data. AC, AD, and JE wrote the main manuscript and all authors reviewed and edited the manuscript. JE and GU provided the overall supervision of the project. All authors contributed to the article and approved the submitted version.