Introduction:Streptococcus pneumoniae is the most common cause of bacterial meningitis and meningoencephalitis in humans. The bacterium produces numerous virulence determinants; among them, hydrogen peroxide (H2O2) and pneumolysin (Ply) contribute to bacterial cytotoxicity. Microglia, the resident phagocytes in the brain, are distinct from other macrophages, and we thus compared their susceptibility to pneumococcal toxicity and their ability to phagocytose pneumococci with those of bone-marrow-derived macrophages (BMDMs). Methods: Microglia and BMDMs were co-incubated with S. pneumoniae D39 to analyze the survival of phagocytes by fluorescence microscopy, bacterial growth by quantitative plating, and phagocytosis by an antibiotic protection assay. Ply was detected by hemolysis assay and Western blot analysis. Results: We found that microglia were killed during pneumococcal infection with a wild-type and an isogenic ply-deficient mutant, whereas the viability of BMDMs was not affected by pneumococci. Treatment with recombinant Ply showed a dose-dependent cytotoxic effect on microglia and BMDMs. However, high concentrations of recombinant Ply were required, and under the chosen experimental conditions, Ply was not detectable in the supernatant during infection of microglia. Inactivation of H2O2 by exogenously added catalase abolished its cytotoxic effect. Consequently, infection of microglia with pneumococci deficient for the pyruvate oxidase SpxB, primarily producing H2O2, resulted in reduced killing of microglia. Conclusion: Taken together, in the absence of Ply, H2O2 caused cell death in primary phagocytes in concentrations produced by pneumococci.

Streptococcus pneumoniae (pneumococcus) is a commensal bacterium on the mucosal surfaces of humans, which, however, can become an invasive pathogen causing severe infections such as pneumonia, septicemia, and meningitis. Carriage of pneumococci is essential for subsequent infection [1]. Pneumococcal meningitis has case fatality rates of 20–37% in high-income countries and up to 51% in low-income countries [2]. A variety of factors is responsible for invasion, dissemination, and translocation into the central nervous system (CNS), yet the mechanisms of the pathogenesis of CNS infections are incompletely understood [3]. Two factors known to be involved in pneumococcal disease are hydrogen peroxide (H2O2) and pneumolysin (Ply) [4‒10].

H2O2 is a metabolic by-product during aerobic growth, produced up to 90% by the enzyme pyruvate oxidase (SpxB) [11], which was first described by Spellerberg et al. in 1996 [12]. Notably, the amount of released H2O2 is high enough to cause bacteriostatic and bactericidal effects on other bacteria in the upper respiratory tract [4]. Moreover, Ply release correlates with the amount and pyruvate oxidase enzymatic activity, i.e., H2O2 production [13]. In mice infection experiments, the production of H2O2 by SpxB led to prolonged colonization of the nasopharynx. During lung infection, the concerted action of Ply and SpxB contributes to bacterial replication in the lungs and translocation into the bloodstream [5]. Although powerful destructive effects were shown for H2O2 in mice in vivo, the knowledge about its role during brain infections remains limited. H2O2 caused ciliary stasis in a ciliated ependymal cell line [9] and played an auxiliary role in Ply toxicity on a microglial cell line [10]. With respect to primary brain cells, a cytotoxic effect of H2O2 has been shown for neurons and microglia [10].

Ply belongs to the group of cholesterol-dependent cytolysins (CDCs). CDC toxins are able to form pores in cholesterol-containing cytoplasmic membranes and thereby lead to the lysis of host cells [14]. Ply production is necessary for the colonization of the nasopharynx, the first step in pneumococcal infection [6]. Furthermore, Ply induces the formation of nitric oxide, a key feature of septic shock [15], and in mouse infections, it leads to lung damage [7]. Ply is a well-established major virulence factor on the path toward and during the progression of pneumonia. In a rat meningitis model, pneumococci deficient in Ply were reported to cause very mild diseases/symptoms in rats [8]. In a mouse model, animals infected with Ply-deficient S. pneumoniae appeared healthier, had lower bacterial concentrations in blood, and lived longer, whereas bacterial titers in the cerebellum, meningeal inflammation, and neuronal damage scores were comparable to mice infected with wild-type S. pneumoniae [16]. It is known that Ply and fragments of the pneumococcal capsule can inhibit the motility of microglia [17] and that Ply is able to cause cell death in primary rat cortical neurons [18]. Furthermore, synaptic damage during meningitis is caused by Ply at disease-relevant non-lytic concentrations probably because it leads to the release of glutamate via pore formation [19].

Microglia are the resident macrophages of the parenchyma of the CNS and make up 10–15% of all glial cells, the most abundant cells in the CNS [20]. They are of different origins than other macrophages in the body. Microglia are derived from the yolk sac and are later replenished by resident progenitor cells, whereas almost all other macrophages are derived from bone marrow progenitors [21‒24]. They have a key role during infections of the CNS as they display the first line of defense against invading pathogens. They express a broad range of receptors to recognize pathogens; they are able to phagocytose them and produce cytokines which can recruit immune cells from the periphery and enable microglia to communicate with cells within the CNS [25].

In this study, we found that primary microglia, but not bone-marrow-derived macrophages (BMDMs), were killed during pneumococcal infection. This prompted us to find out which factor is responsible for the rapid death of microglia. Our results indicate that cell death was mainly dependent on H2O2 under experimental conditions in which Ply was not released by pneumococci in sufficient concentrations to harm microglia. Both factors can be toxic for host cells; however, in relation to the concentrations produced by pneumococci during infection, very high concentrations of recombinant Ply (rPly) and low concentrations of H2O2 were necessary. BMDMs were more effective than microglia in killing pneumococci and, therefore, prevented the accumulation of high concentrations of pneumococcal H2O2.

Bacterial Strains and Culture Conditions

All mutant strains used in this study were generated in S. pneumoniae D39 (NCTC 7466, serotype 2) in previous studies [26‒29]. We used cryopreserved bacteria for all experiments. For the preparation of cryostocks, S. pneumoniae was grown on Columbia blood agar plates (Oxoid, Thermo Fisher Scientific) at 37°C and 5% CO2 overnight and subsequently cultivated in Todd Hewitt broth (Bacto, Becton Dickinson) containing 0.5% yeast extract (Carl Roth) (THY) until the optical density at 600 nm reached 0.6. Pneumococci were washed with phosphate-buffered saline (PBS; Gibco, Thermo Fisher Scientific), diluted in a medium with a final concentration of 15% glycerol, and snap-frozen in liquid nitrogen. Cryopreserved bacteria were stored at −80°C, until they were thawed and diluted to the appropriate concentrations of colony-forming units (CFU) needed in the experiments.

Eukaryotic Cell Culture

L929 cells (ATTC®: CCL-1) were cultured in 50 mL Dulbecco’s modified Eagle medium (DMEM; low glucose [1 g/L], l-glutamine, and pyruvate [Gibco, Thermo Fisher Scientific]) containing 5% fetal calf serum (FCS; Biochrom) and 2 mml-glutamine (Gibco, Thermo Fisher Scientific) in 175 cm2 cell culture flasks at 37°C and 8% CO2. Culture supernatants were harvested weekly, centrifuged twice, and frozen for later use to stimulate or differentiate microglia or BMDMs. The human laryngeal epithelial cell line HEp-2 (ATCC® CCL-23™) and the murine macrophage cell line J774A.1 (ATCC® TIB-67) were cultured in DMEM (low glucose [1 g/L], l-glutamine, and pyruvate [Gibco, Thermo Fisher Scientific]) containing 10% FCS (Biochrom) and 2 mml-glutamine (Gibco, Thermo Fisher Scientific) in 75 cm2 cell culture flasks at 37°C and 8% CO2.

Primary microglia were prepared from neonatal (0–2 days) mice as previously described [30, 31]. Briefly, after preparation, mixed glial cultures were cultured in poly-l-lysine (Merck)-coated 75 cm2 cell culture flasks with DMEM (high glucose (4.5 g/L) and l-glutamine, without pyruvate [Gibco, Thermo Fisher Scientific]) containing 10% FCS (Biochrom), 2 mml-glutamine (Gibco, Thermo Fisher Scientific), and penicillin (50 units/mL)-streptomycin (50 μg/mL) (Gibco, Thermo Fisher Scientific) for 10 days or until confluency of the cell layer was reached at 37°C and 8% CO2. To increase microglial cell division, one-third of the supernatants of the cell line L929 were added to the medium without antibiotics for 3–4 days at 37°C and 8% CO2. For detachment of microglia, mixed glial cell culture flasks were shaken for 30 min at 37°C at 130 rounds per minute on an orbital shaker. Cells from supernatants were counted and microglia seeded as described for individual assays. Depending on the experiment, a medium with or without phenol red was used for the cultivation and seeding of microglia.

For obtaining primary BMDMs, bone marrow cells were prepared from mice and differentiated as previously described [32]. Bone marrow cells were cultivated in DMEM (low glucose [1 g/L], l-glutamine, and pyruvate [Gibco, Thermo Fisher Scientific]), supplemented with 10% FCS (Biochrom), 2 mml-glutamine (Gibco, Thermo Fisher Scientific), 1/5 M-CSF/L929 supernatant, and penicillin (50 units/mL)-streptomycin (50 μg/mL) (Gibco, Thermo Fisher Scientific). Cells were stimulated with 1/5 L929 culture supernatants for 3 days at 37°C and 8% CO2. Then, the medium including stimulating supernatants was replaced, and cells were cultivated likewise for additional 6–7 days. Cells were detached via cold shock for 3 min on ice and afterward resuspended in warm medium without antibiotics and seeded according to the individual assays.

Phagocytosis and Intracellular Survival Assay

Primary microglia or BMDMs were seeded at 125,000 cells/mL in DMEM without antibiotics in a 6-well plate (2 mL/well; Fig. 1a, b) or 96-well plate (200 μL/well, in triplicate; Fig. 1c) and incubated at 37°C and 5% CO2. The following day, cells were infected in vitro at a multiplicity of infection (MOI) of 10 with S. pneumoniae D39, the capsule- (D39Δcps), Ply- (D39Δply), or double-knockout mutant (D39ΔcpsΔply) for 1 h at 37°C and 5% CO2 in the absence or presence of rPly (1–1000 hemolytic units per mL [HU/mL]). Subsequently, eukaryotic cells were washed with PBS containing MgCl2 and CaCl2 (Merck) and then incubated in medium containing final concentrations of 100 μg/mL gentamicin (Carl Roth) and 10 μg/mL penicillin G (Merck) for 1 h at 37°C and 5% CO2. Afterward, eukaryotic cells were washed with PBS containing MgCl2 and CaCl2, lysed with distilled water, and viable intracellular bacteria were plated at serial dilutions in triplicate. By using this plating method, a minimum of 100 CFU/well is detectable (indicated by the dotted line in the respective graph).

Fig. 1.

Bacterial growth is impaired in the presence of BMDMs due to phagocytosis. Microglia (a) or BMDMs (b) were infected with S. pneumoniae D39 or the capsule-deficient mutant D39Δcps at a MOI of 10 in a 6-well plate (2 mL/well). Growth of pneumococci was determined in medium only or in the presence of primary macrophages (“+ primary cells”). Colony-forming units (CFU) are shown as mean ± SD of three independent experiments (one technical replicate each). Significance is indicated compared to bacteria grown in medium by **p < 0.01 and ****p < 0.0001 (one-way ANOVA followed by Sidak’s multiple comparison test). c In a penicillin-gentamicin protection assay, microglia and BMDMs were infected with D39 or isogenic mutants, respectively, and the number of intracellular bacteria was determined via plating. Host cells were infected for 1 h at a MOI of 10 in a 96-well plate (200 μL/well) and afterward treated with antibiotics for 1 h. The number of recovered bacteria (CFU/well, 100 μL/well) is shown as mean ± SD of at least three independent experiments (technical triplicate each). Significant differences between microglia and BMDMs are indicated by **p < 0.01 (unpaired t test). The dotted line represents the limit of detection for the applied plating method. SD, standard deviation.

Fig. 1.

Bacterial growth is impaired in the presence of BMDMs due to phagocytosis. Microglia (a) or BMDMs (b) were infected with S. pneumoniae D39 or the capsule-deficient mutant D39Δcps at a MOI of 10 in a 6-well plate (2 mL/well). Growth of pneumococci was determined in medium only or in the presence of primary macrophages (“+ primary cells”). Colony-forming units (CFU) are shown as mean ± SD of three independent experiments (one technical replicate each). Significance is indicated compared to bacteria grown in medium by **p < 0.01 and ****p < 0.0001 (one-way ANOVA followed by Sidak’s multiple comparison test). c In a penicillin-gentamicin protection assay, microglia and BMDMs were infected with D39 or isogenic mutants, respectively, and the number of intracellular bacteria was determined via plating. Host cells were infected for 1 h at a MOI of 10 in a 96-well plate (200 μL/well) and afterward treated with antibiotics for 1 h. The number of recovered bacteria (CFU/well, 100 μL/well) is shown as mean ± SD of at least three independent experiments (technical triplicate each). Significant differences between microglia and BMDMs are indicated by **p < 0.01 (unpaired t test). The dotted line represents the limit of detection for the applied plating method. SD, standard deviation.

Close modal

Assessment of Cell Viability by Hoechst/Propidium Iodide Double Staining

Primary microglia or BMDMs were seeded at 100,000 cells/mL in DMEM without antibiotics in a 24-well plate (1 mL/well) and incubated at 37°C and 5% CO2. The following day, primary macrophages were infected with either S. pneumoniae D39, D39Δcps, D39Δply, and D39ΔcpsΔply or the pyruvate oxidase-knockout strain (D39ΔspxB), respectively, at a MOI of 3 or 10 or were treated with different concentrations of H2O2 (Merck) for up to 6 h at 37°C and 5% CO2. For experiments including catalase pretreatment, the enzyme (Merck) was used in a concentration of 2,000 U/mL in cell culture medium and added 30 min prior to infection with pneumococci or treatment with H2O2 [33].

After the respective incubation time, the nuclei of these cells were stained with 15 nmol/mL propidium iodide (Merck) to label the nuclei of dead cells and 1.78 nmol/mL Hoechst 33342 (Merck) to label the nuclei of all cells for 10 min at 37°C. Three images from different areas per time point and condition (on average resulting in 1,200 cells) were recorded at 200-fold magnification with a Nikon Eclipse Ti-S microscope (Nikon) using NIS-Elements BR 4.51.01 (Nikon) software. Cells were later analyzed using CellProfiler 3.0 (Broad Institute) [34] and ImageJ/Fiji (NIH) [35].

H2O2 Assay

Primary microglia were seeded at 100,000 cells/mL in DMEM without antibiotics and without phenol red in 24-well plates (1 mL/well) and incubated at 37°C and 5% CO2. The following day, microglia were infected with S. pneumoniae D39, D39Δcps, D39Δply, D39ΔspxB, or the double knockout strains D39ΔcpsΔply and D39ΔcpsΔspxB at a MOI of 3 for up to 6 h at 37°C and 5% CO2.

After centrifugation, supernatants were transferred to a 48-well plate (900 μL/well) and H2O2 levels were determined by the method described by Pick and Keisari [36] with the following modifications: Phenol red (Merck) was used at a final concentration of 0.072 mm, horseradish peroxidase (Merck) at a final concentration of 9 U/mL, and the reaction was stopped by a final concentration of 10 mm NaOH (Carl Roth). Moreover, the experiment was performed in DMEM (high glucose [4.5 g/L] without l-glutamine, without phenol red, and without pyruvate [Gibco, Thermo Fisher Scientific]) supplemented with 10% FCS (Biochrom) and 2 mml-glutamine (Gibco, Thermo Fisher Scientific). Changes in absorption due to the oxidation of phenol red were detected at 605 nm in 96-well plates (200 μL/well) using a microplate reader (SpectraMax i3x, Molecular Devices). H2O2 concentrations in the supernatants were calculated by interpolation of absorption values to an H2O2 standard. If necessary, samples were diluted 1:4 or 1:6 with medium before performing H2O2 measurements.

In addition, the amount of H2O2 produced by pneumococci in pooled human cerebrospinal fluid (CSF) and in DMEM in the absence of microglia was compared. CSF was supplemented with the following amino acids: alanine, arginine, asparagine, aspartic acid, glutamic acid, glycine, histidine, hydroxyl proline, isoleucine, leucine, methionine, phenylalanine, proline, serine, tryptophan, tyrosine, valine (final concentration 50 mg/L), glutamine, and threonine (final concentration 100 mg/L), as well as adenine and uracil (final concentration 100 mg/L and 200 mg/L) (modified from [37]). Due to the limited amount of human CSF available, the sample volumes had to be adjusted for this assay. Briefly, S. pneumoniae D39 was incubated at a concentration of 106 CFU/mL (corresponding to a MOI of 10 in the presence of cells) in CSF or DMEM in a 48-well plate (250 μL/well) for up to 6 h at 37°C and 5% CO2. The supernatants were centrifuged, and 100 μL of the supernatants was transferred to 96-well plates to perform the H2O2 assay as described above. If necessary, samples were diluted 1:2, 1:4, or 1:8 with medium before performing the H2O2 measurement.

Cytotoxicity Assay

A lactate dehydrogenase release assay was performed to test the cytotoxicity of Ply, using rPly heterologously produced in Escherichia coli [38]. Primary cells were seeded at concentrations of 300,000 cells/mL in 96-well plates (200 μL/well), whereas the cell lines were seeded at 150,000 cells/mL (J774A.1) and 100,000 cells/ml (HEp-2), respectively, resulting in similar amounts of cells on the next day. On the following day, cells were exposed for 2 h to 128–8,200 HU/mL of rPly. The CytoTox 96® Non-Radioactive Cytotoxicity Assay (Promega) was used according to the instructions of the manufacturer.

Hemolysis Assay

The assay was performed to first determine the HU/mg of rPly [38] and later to determine the concentration of active Ply produced by pneumococci. Briefly, 100 μL samples with known concentrations of rPly were serially diluted 1:2 in 0.9% NaCl solution or DMEM (high glucose [4.5 g/L] and l-glutamine, without pyruvate and phenol red [Gibco, Thermo Fisher Scientific]) supplemented with 10% FCS (Biochrom) and 2 mml-glutamine (Gibco, Thermo Fisher Scientific), respectively, in conical-shaped 96-well plates. Afterward, 100 μL of a 2% erythrocyte suspension, prepared from sheep blood (Fiebig Nährstoffe, Germany) in a 0.9% NaCl solution or DMEM, was added. Completely lysed or untreated cells served as controls. The plates were incubated on an orbital shaker at 100 rounds per minute for 2 h at 37°C. Plates were centrifuged, supernatants were transferred into flat-bottom 96-well plates, and absorption was measured at 550 nm in a microplate reader (SpectraMax i3x, Molecular Devices). One HU/mL was defined as the concentration of toxin causing 50% hemolysis of a 2% sheep erythrocyte suspension. Depending on the lot and activity of rPly, one HU corresponded to 2.2–5 ng rPly.

To determine the amount of active Ply produced by pneumococci, S. pneumoniae D39 was incubated at a concentration of 106 CFU/mL (corresponding to an MOI of 10 in the presence of cells) in THY or DMEM in 24-well plates (1 mL/well) for up to 6 h at 37°C and 5% CO2. The supernatants of five wells were pooled and centrifuged, and 4 mL of the supernatant was concentrated using the Amicon® Ultra-4 centrifugal filter unit (Merck). Hemolysis assay was conducted with 100 μL of the concentrated supernatant (concentration factor 10) as described above, and the HU/mL supernatant was calculated based on the volume of concentrated supernatant that induced 50% hemolysis of the 2% sheep erythrocyte suspension. A minimum of 1 HU/mL supernatant could be detected by this method.

Western Blot Analysis

Concentrated supernatants obtained for the hemolysis assay were separated electrophoretically using a 5% stacking and a 10% running sodium dodecyl sulfate-polyacrylamide gel and were transferred to a polyvinylidene fluoride membrane (Merck). The membranes were blocked for 1 h at room temperature (RT) with 5% skimmed milk powder in Tris-buffered saline (TBS) with 1% Tween®20 (Carl Roth). Afterward, they were incubated with a polyclonal rabbit antiserum raised against Ply (Davids Biotechnologie GmbH, Germany; dilution 1:500 in 1% skimmed milk powder in TBS with 1% Tween®20) [38] overnight at 4°C, followed by incubation with HRP-linked goat anti-rabbit IgG (Cell Signaling; dilution 1:5,000 in 1% skimmed milk powder in TBS with 1% Tween®20) for 1 h at RT. The membranes were developed with SuperSignal™ West Pico Chemiluminescent Substrate (Thermo Fisher Scientific), and the chemiluminescent signal was detected with ChemoCam Imager 3.2 (Intas, Germany).

Flow Cytometric Analysis

For the detection of rPly binding to microglia and J774A.1 macrophages and the detection of viable cells, 5 × 105 cells were co-incubated with rPly (1–1,000 HU/mL) in the respective cell culture medium for 1 h at 37°C. Afterward, cells were centrifuged (300 g, 5 min, RT), the supernatant was removed, and cells were washed in PBS, followed by live/dead staining using the Zombie NIR Fixable Viability Kit (BioLegend) for 10 min at RT. Cells were washed and incubated with a primary polyclonal serum raised against rPly in rabbits (Davids Biotechnologie GmbH; dilution 1:250 in PBS + 2% FCS) for 30 min at 37°C. After washing, cells were incubated with a secondary anti-rabbit Alexa-Fluor 488-conjugated antibody (Dianova) for 10 min on ice. Then, cells were washed and resuspended in 2% paraformaldehyde. Samples were acquired on a BD LSRII cytometer (BD Biosciences). All data evaluation was performed using FlowJo software (FlowJo LLC).

The following controls were included in the analysis: cells were incubated with 0 HU/mL rPly + primary polyclonal serum raised against rPly in rabbits + secondary anti-rabbit Alexa-Fluor 488-conjugated antibody (refers to “-”) and cells were preincubated with 1,000 HU/mL rPly + a naïve rabbit serum + secondary anti-rabbit Alexa-Fluor 488-conjugated antibody (refers to “antibody control”).

Statistical Analysis

All experiments were performed at least three times unless otherwise indicated, and results are shown as mean ± standard deviation. All statistical analyses were carried out using GraphPad Prism (version 9.0.0; GraphPad Software, USA). The normal (Gaussian) distribution of the data was tested with the Shapiro-Wilk test and statistically significant differences were identified by the unpaired t test or one-way ANOVA followed by Tukey’s, Sidak’s, or Dunnett’s multiple comparison test. Statistically significant differences are indicated by *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

S. pneumoniae Grows in the Presence of Microglia but Is Killed by BMDMs

Primary microglia and primary BMDMs were infected in vitro with wild-type or nonencapsulated S. pneumoniae D39 (D39Δcps) at a MOI of 10. Nonencapsulated pneumococci were included, because they occasionally show different growth kinetics, depending on the medium used, and are usually phagocytosed more efficiently by macrophages than encapsulated pneumococci. Both effects could result in differences in bacterial numbers during growth in cell culture medium or in the presence of macrophages. Remarkably, equal concentrations (CFU/well) of pneumococci were recovered when bacteria were grown in the presence of microglia or in medium (DMEM 4.5 g/L glucose) without eukaryotic cells (Fig. 1a). Similar bacterial densities were reached, when pneumococci were grown in DMEM supplemented with 4.5 g/L glucose (medium used for cultivation of microglia) or 1 g/L glucose (medium used for cultivation of BMDMs) (Fig. 1a, b). In the presence of BMDMs, the number of recovered pneumococci from the supernatant was significantly lower after 4 h and 8 h of incubation compared to growth in the medium alone (Fig. 1b). Taken together, this indicates that pneumococci were killed by BMDMs, but not by microglia, and pneumococcal growth was independent of the glucose concentration.

To evaluate how pneumococcal survival in the presence of microglia or BMDMs was influenced by phagocytic uptake and killing, a penicillin-gentamicin protection assay was performed to determine the number of intracellular pneumococci (Fig. 1c). Phagocytic uptake was observed in BMDMs and microglia after 1 h of infection followed by 1 h of antibiotic treatment and was similar for nonencapsulated pneumococci, but the number of intracellular encapsulated bacteria was lower in BMDMs when compared to microglia, suggesting a faster intracellular killing of the pneumococci by BMDMs. Additionally, we observed killing of pneumococci in the cell-free supernatant of BMDMs (online suppl. Fig. S1; for all online suppl. material, see https://doi.org/10.1159/000536514), indicating that also extracellularly secreted factors of BMDMs are involved in the killing of pneumococci. Thus, killing of S. pneumoniae was delayed in microglia compared to BMDMs and pneumococci were killed extra- and intracellularly by BMDMs.

As described in previous studies with S. pneumoniae and other streptococci, nonencapsulated pneumococci were phagocytosed in higher numbers by microglia as well as BMDMs compared to encapsulated bacteria (Fig. 1c). Conversely, no differences could be observed between the wild-type strain and the Ply-deficient isogenic mutant. To answer the question whether Ply affects phagocytosis under the chosen experimental conditions, rPly was added in increasing concentrations (up to 1,000 HU/mL) to microglia infected with S. pneumoniae D39ΔcpsΔply. However, addition of rPly in the chosen concentrations did not alter bacterial uptake by microglia (online suppl. Fig. S2).

Microglia Are Killed during in vitro Pneumococcal Infection, whereas BMDMs Survive

Next, we analyzed whether the observed differences in pneumococcal growth in the presence of microglia or BMDMs (Fig. 1) were caused by differences in macrophage numbers due to cell death induced by pneumococci. Vital staining and microscopic analysis were performed to assess the cell death of primary phagocytes during infection with pneumococci (Fig. 2). Remarkably, a significant decrease in the percentage of viable microglia but not of BMDMs was detectable over time. At 6 h post infection, less than 20% of microglia were still vital (Fig. 2a), whereas BMDMs retained almost full viability over time in the presence of pneumococci. Only nonencapsulated pneumococci were able to induce approximately 10% cell death in BMDMs (Fig. 2b). The lethal effect of pneumococcal infection on microglia was independent of the capsule and Ply as the reduction of microglial cells during infection was similar for all strains (Fig. 2a). The percentage of microglial cell death depended on the relation of bacteria and phagocytes for infection of microglia with D39Δcps. After 4 h of infection, at a MOI of 3 approximately 40% of viable microglia could be detected, but only 20% when infected at a MOI of 10 (Fig. 2a; online suppl. Fig. S3a). Notably, the differences were not a consequence of different growth behaviors, because comparable amounts of S. pneumoniae were recovered from the supernatants throughout the experiments (online suppl. Fig. S3b; Fig. 1a). Taken together, S. pneumoniae infection did not impair BMDM viability but led to dose- and time-dependent cell death in microglial cells, which was independent of the presence of the pneumococcal capsule or Ply.

Fig. 2.

BMDMs survive in vitro pneumococcal infection, while microglia die in a time-dependent manner. After infection of microglia (a) or BMDMs (b) at a MOI of 10 with S. pneumoniae D39 or different mutant strains, cytotoxicity was visualized by propidium iodide-Hoechst staining of the nuclei. On average, 1,200 cells were analyzed per condition by fluorescence microscopy to determine the number of dead cells. Results are shown as mean ± SD of at least three independent experiments (one technical replicate each). Significant differences between uninfected cells (control) and cells infected with S. pneumoniae are indicated by *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (one-way ANOVA followed by Dunnett’s multiple comparison test). SD, standard deviation.

Fig. 2.

BMDMs survive in vitro pneumococcal infection, while microglia die in a time-dependent manner. After infection of microglia (a) or BMDMs (b) at a MOI of 10 with S. pneumoniae D39 or different mutant strains, cytotoxicity was visualized by propidium iodide-Hoechst staining of the nuclei. On average, 1,200 cells were analyzed per condition by fluorescence microscopy to determine the number of dead cells. Results are shown as mean ± SD of at least three independent experiments (one technical replicate each). Significant differences between uninfected cells (control) and cells infected with S. pneumoniae are indicated by *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (one-way ANOVA followed by Dunnett’s multiple comparison test). SD, standard deviation.

Close modal

Cytotoxic Effect of Ply on Phagocytes Is Only Apparent at High Concentrations

To analyze the possible cytotoxic effects of Ply on phagocytes in more detail, we added different concentrations of rPly to primary microglia and BMDMs, as well as the murine macrophage cell line J774.A1 and the human laryngeal epithelial cell line HEp-2 for reasons of comparison. Dose-dependent cytotoxic effects on macrophages were only detected at very high concentrations (1,025–8,200 HU/mL) of rPly (Fig. 3a). Cytotoxic effects that caused more than 20% cell damage in macrophages were only observed at concentrations of 4,100 and 8,200 HU/mL rPly. Moreover, at these high rPly concentrations, BMDMs and J774.A1 cells were significantly more susceptible toward the toxin than microglia. Notably, the epithelial cell line HEp-2 was more susceptible toward rPly than BMDMs and J774A.1 cells, because 2,050 HU/mL rPly caused 45% cytotoxicity in HEp-2 cells and only 10–15% cytotoxicity in all other cell types tested (Fig. 3a).

Fig. 3.

Cytotoxic effect of Ply on macrophages is only apparent at high toxin concentrations, which are not produced by pneumococci. a Primary microglia and BMDMs as well as the J774A.1 macrophage cell line and the epithelial cell line HEp-2 were stimulated with rPly (HU) for 2 h. Cytotoxicity was determined by the lactate dehydrogenase release assay. Results are shown as mean ± SD of at least three independent experiments (technical duplicates each). Significant differences between the cells are indicated by *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (one-way ANOVA followed by Tukey’s multiple comparison test). b HU/mL of Ply produced by S. pneumoniae D39 after 6 h of incubation in DMEM or in the presence of microglia were determined by hemolysis assay with 2% sheep erythrocytes. Results are shown as mean ± SD of three independent experiments (one technical replicate each). The dotted line represents the limit of detection. SD, standard deviation.

Fig. 3.

Cytotoxic effect of Ply on macrophages is only apparent at high toxin concentrations, which are not produced by pneumococci. a Primary microglia and BMDMs as well as the J774A.1 macrophage cell line and the epithelial cell line HEp-2 were stimulated with rPly (HU) for 2 h. Cytotoxicity was determined by the lactate dehydrogenase release assay. Results are shown as mean ± SD of at least three independent experiments (technical duplicates each). Significant differences between the cells are indicated by *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (one-way ANOVA followed by Tukey’s multiple comparison test). b HU/mL of Ply produced by S. pneumoniae D39 after 6 h of incubation in DMEM or in the presence of microglia were determined by hemolysis assay with 2% sheep erythrocytes. Results are shown as mean ± SD of three independent experiments (one technical replicate each). The dotted line represents the limit of detection. SD, standard deviation.

Close modal

To prove that rPly can bind to microglia under the chosen experimental conditions, a flow cytometric analysis with microglia, and in comparison with J774.A1 macrophages, was conducted (online suppl. Fig. S4). Using concentrations that are sublytic for microglia, a dose-dependent binding of rPly was detected to J774.A1 and microglia. In contrast to the cytotoxicity assay (Fig. 3a), the highest concentration (1,000 HU/mL rPly) induced cell death in J774.A1 macrophages (online suppl. Fig. S4d) but not in microglia (online suppl. Fig. S4c).

As the activity of rPly might be influenced by the culture conditions, the hemolytic activity typically tested in 0.9% sodium chloride (NaCl) solution was determined in cell culture medium (DMEM) as well as in the bacterial growth medium (THY). Indeed, the activity of rPly was significantly higher in 0.9% NaCl compared to DMEM and THY, respectively (online suppl. Fig. S5).

To correlate the concentrations of rPly used to induce cytotoxic effects in microglia and BMDMs to the concentration of active Ply released by pneumococci in vitro, HU/mL of Ply released by S. pneumoniae D39 were determined after 6 h of incubation in DMEM only and in the presence of microglia, respectively (Fig. 3b). No active toxin was detectable in the supernatant of infected microglia. In addition, we determined the HU/mL of Ply produced by S. pneumoniae D39 in THY medium in comparison to DMEM (online suppl. Fig. S6c) and found that the protein itself as well as its hemolytic activity was only detectable in THY (2–8 HU/mL) but not in DMEM. Plating of the supernatants revealed that bacteria did not grow beyond 108 CFU/mL during incubation in DMEM, whereas in the THY medium we counted more than 108 CFU/mL after 6 h of incubation (online suppl. Fig. S6d, e). Even though the bacterial number in the THY medium is high enough to induce autolysis of the pneumococci and release Ply, the measured hemolytic activity was far below the concentration of active toxin necessary to damage microglia. Taken together, Ply is not present under the given experimental conditions (DMEM) and therefore cannot be responsible for microglial cell death.

H2O2 Produced by S. pneumoniae Kills Microglia

Because Ply seems not to be responsible for pneumococcal cytotoxicity in microglial infection, we tested the possible effects of H2O2 in the next step. The ability of pneumococci to produce high concentrations of H2O2 is well known, and H2O2 is a frequently mentioned virulence determinant of pneumococci. To analyze the concentration of H2O2 produced during pneumococcal infection of microglia, the supernatants of infected cells were tested in a phenol red-horseradish peroxidase assay (Fig. 4a). The wild-type and nonencapsulated mutant strains produced up to 2.5 mm H2O2 in the presence of microglia after 6 h of incubation, whereas the Ply-deficient mutant strain produced only up to 1.5 mm. As expected, D39ΔspxB released only a limited amount of H2O2 (0.2 mm), with SpxB (pyruvate oxidase) being the enzyme that produces the vast amount of H2O2 during pneumococcal metabolism. In order to prove that measurable H2O2 was mainly produced by the pneumococci and not by the microglia themselves, H2O2 production was also analyzed in the absence of microglia and comparable amounts of H2O2 were detected in medium plus pneumococci only (online suppl. Fig. S7).

Fig. 4.

Hydrogen peroxide (H2O2) kills microglia in concentrations produced by S. pneumoniae. a Microglia were infected with S. pneumoniae D39 or different mutant strains at a MOI of 3 for up to 6 h. H2O2 concentration in the supernatant was determined via phenol red-horseradish peroxidase assay. Results are shown as mean ± SD of at least three independent experiments (one technical replicate each). Significant differences compared to D39 are indicated by ****p < 0.0001 (one-way ANOVA followed by Dunnett’s multiple comparison test). b Microglia were exposed to different concentrations of H2O2 in DMEM for up to 6 h. Cytotoxicity was determined by propidium iodide-Hoechst staining of the nuclei. Results are shown as mean ± SD of three independent experiments (one technical replicate each). Significant differences between untreated cells (control) and cells treated with H2O2 are indicated by **p < 0.01, ***p < 0.001, and ****p < 0.0001 (one-way ANOVA followed by Dunnett’s multiple comparison test). c Microglia were infected with S. pneumoniae D39 or D39ΔspxB at a MOI of 3 for 6 h after pretreatment with catalase (2,000 units/mL). Cytotoxicity was determined by propidium iodide-Hoechst staining of the nuclei. Results are shown as mean ± SD of three independent experiments (with the exception of H2O2 + catalase, which was only performed twice; one technical replicate each). Significant differences between cells pretreated with catalase and untreated cells are indicated by ***p < 0.001 and ****p < 0.0001 (unpaired t test). SD, standard deviation.

Fig. 4.

Hydrogen peroxide (H2O2) kills microglia in concentrations produced by S. pneumoniae. a Microglia were infected with S. pneumoniae D39 or different mutant strains at a MOI of 3 for up to 6 h. H2O2 concentration in the supernatant was determined via phenol red-horseradish peroxidase assay. Results are shown as mean ± SD of at least three independent experiments (one technical replicate each). Significant differences compared to D39 are indicated by ****p < 0.0001 (one-way ANOVA followed by Dunnett’s multiple comparison test). b Microglia were exposed to different concentrations of H2O2 in DMEM for up to 6 h. Cytotoxicity was determined by propidium iodide-Hoechst staining of the nuclei. Results are shown as mean ± SD of three independent experiments (one technical replicate each). Significant differences between untreated cells (control) and cells treated with H2O2 are indicated by **p < 0.01, ***p < 0.001, and ****p < 0.0001 (one-way ANOVA followed by Dunnett’s multiple comparison test). c Microglia were infected with S. pneumoniae D39 or D39ΔspxB at a MOI of 3 for 6 h after pretreatment with catalase (2,000 units/mL). Cytotoxicity was determined by propidium iodide-Hoechst staining of the nuclei. Results are shown as mean ± SD of three independent experiments (with the exception of H2O2 + catalase, which was only performed twice; one technical replicate each). Significant differences between cells pretreated with catalase and untreated cells are indicated by ***p < 0.001 and ****p < 0.0001 (unpaired t test). SD, standard deviation.

Close modal

To test whether H2O2 itself is sufficient to kill primary macrophages, a concentration-effect curve of H2O2 with microglia was performed (Fig. 4b). The tested concentrations were chosen according to the concentrations of H2O2 released by pneumococci during infection (Fig. 4a). After 4 h, less than 40% of the microglia were still alive, and after 6 h, the cells were efficiently killed even by the lowest concentration of H2O2 (0.25 mm). Therefore, concentrations of H2O2 produced by pneumococci are sufficient to kill the microglia. Likewise, H2O2 was able to kill BMDMs, yet the cells were not as susceptible for H2O2 as microglia (online suppl. Fig. S8).

Finally, we treated microglia with catalase prior to pneumococcal infection and tested cell viability at 6 h post infection (Fig. 4c). After infection with D39, less than 5% viable cells were detected; conversely, when pretreated with catalase, significantly more viable microglia (approximately 80%) were observed after 6 h. As a control, microglia were incubated with 2.5 mm H2O2. Prior treatment with catalase retained the cell viability of H2O2-treated microglia at the level of uninfected control cells (Fig. 4c). Consequently, we also tested the pneumococcal spxB mutant, which produced smaller amounts of H2O2 during the infection of microglia (Fig. 4a). After 6 h of infection with D39ΔspxB, approximately 70% of microglial cells were still viable. Moreover, when pretreated with catalase, microglial viability could be restored to the level of the uninfected control (Fig. 4c). These results show that pneumococcal H2O2 is responsible for microglial cell death under the chosen experimental conditions and sufficient concentrations of H2O2 are released by pneumococci during microglial infection.

S. pneumoniae Produces H2O2 in Medium and in Human CSF

To assess if the observed effects of strong H2O2 production by S. pneumoniae also occur under more physiological conditions, pneumococci were incubated in pooled human CSF and for comparison in DMEM for up to 6 h (Fig. 5). Pneumococci grew to higher concentrations (CFU/mL) in CSF compared to DMEM (Fig. 5a), and comparable amounts of H2O2 were produced by the bacteria under both conditions (Fig. 5b). This indicates that H2O2 is released in cytotoxic amounts under physiological conditions.

Fig. 5.

S. pneumoniae produces comparable amounts of H2O2 in DMEM and CSF. S. pneumoniae D39 was grown in DMEM or human CSF in 48-well plates (250 μL/well) for 6 h, with a starting CFU of 1 × 106 CFU/mL. a Growth of pneumococci (CFU/mL) in DMEM and CSF was determined by plating. b Concentration of H2O2 (mM/well) produced by pneumococci during growth in DMEM and CSF was measured via the phenol red-horseradish peroxidase assay. Results are shown as mean ± SD of three independent experiments (a: four technical replicates were pooled for plating, b: technical duplicates each). Significant differences between DMEM and CSF are indicated by *p < 0.05, **p < 0.01, and ****p < 0.0001 (unpaired t test). SD, standard deviation.

Fig. 5.

S. pneumoniae produces comparable amounts of H2O2 in DMEM and CSF. S. pneumoniae D39 was grown in DMEM or human CSF in 48-well plates (250 μL/well) for 6 h, with a starting CFU of 1 × 106 CFU/mL. a Growth of pneumococci (CFU/mL) in DMEM and CSF was determined by plating. b Concentration of H2O2 (mM/well) produced by pneumococci during growth in DMEM and CSF was measured via the phenol red-horseradish peroxidase assay. Results are shown as mean ± SD of three independent experiments (a: four technical replicates were pooled for plating, b: technical duplicates each). Significant differences between DMEM and CSF are indicated by *p < 0.05, **p < 0.01, and ****p < 0.0001 (unpaired t test). SD, standard deviation.

Close modal

Here, we demonstrate that pneumococcal infection was cytotoxic for primary microglia, while it was not for BMDMs. In contrast to microglia, BMDMs efficiently killed S. pneumoniae intra- and extracellularly. The observed cell death was independent of Ply, as under our experimental conditions Ply was not detectable in the supernatants of infected cells, whereas catalase (an H2O2-degrading enzyme) treatment of cells diminished the cytotoxic effect of pneumococci on microglia. H2O2 was cytotoxic to microglia in concentrations released by pneumococci under the given experimental conditions. We detected up to 2 mm H2O2 produced by the wild-type pneumococcal strain D39 in our experimental setup (Fig. 4a). These concentrations correspond to an earlier finding, reporting that 1.8 mm H2O2 was formed by 5 × 107 CFU S. pneumoniae within 30 min [11]. Moreover, mutant strains lacking the major producer of pneumococcal H2O2, the pyruvate oxidase, were not cytotoxic to microglial cells in our experimental setting. After H2O2 titration on primary microglia, the resulting concentration-dependent cell death curve reflected the curve of microglial cell death caused by pneumococci (Fig. 2a). H2O2 concentrations as low as 0.25 mm were able to cause microglial cell death in our experiments. A slightly lower concentration of H2O2 (approximately 0.2 mm after 6 h of incubation) was produced by D39ΔspxB during growth in the presence of microglia (Fig. 4a), and a reduction of approximately 30% in cell viability was observed after co-cultivation of microglia with D39ΔspxB for 6 h in comparison with microglia treated with catalase (Fig. 4c). Under long-term exposure of cells to H2O2, concentrations of 10 nm were sufficient to induce eukaryotic cell death [39, 40]. Microglia are, compared to other CNS cell types such as oligodendrocytes, astrocytes, and neurons, the slowest in degrading H2O2 [41]. It needs to be determined in vivo whether microglia have a similar sensitivity to H2O2 as in vitro since microglial survival is regulated by CSFR1 and CSFR1 antagonists which make microglia more susceptible to cell death induced by H2O2 [42].

In contrast to microglia, BMDMs efficiently killed pneumococci extra- and intracellularly. Exposing BMDMs to increasing concentrations of H2O2 revealed that they also can be harmed by it. Nevertheless, higher concentrations are needed, because BMDMs are more resistant to H2O2 toxicity than microglia and are able to survive up to 0.25 mm H2O2 for 6 h. Even after treatment with 0.5 mm H2O2 for 6 h, survival is only reduced by approximately 25% (online suppl. Fig. S8). In addition to a higher H2O2 tolerance, BMDMs are obviously able to prevent the accumulation of harmful pneumococcal H2O2 by efficiently killing the bacteria.

Microglia play a key role during infections of the CNS as they represent the first line of defense against invading pathogens. They express a broad range of receptors for the recognition of pathogens, and they are able to phagocytose them and produce cytokines which can recruit immune cells from the periphery [25]. On the other hand, the activation of microglia, the release of pro-inflammatory cytokines and chemokines, and the recruitment of peripheral leukocytes can result in an excessive inflammatory response and the breakdown of the blood-brain barrier [43]. In addition, microglial activation by bacterial products contributes directly to neuronal death during bacterial meningitis [44], leading to neurological sequelae in up to 50% of survivors of pneumococcal meningitis [45‒47]. Thus, activation of microglia by bacterial pathogens might facilitate invasion of the CNS and manifestation of bacterial meningitis. However, an increased inflammatory response poses a risk for the pathogen, and therefore induction of microglial cell death might represent a strategy to ensure bacterial survival. Studies by Braun et al. [10] demonstrated that Ply and H2O2 both can induce neuronal and microglial apoptosis in vitro and mediate brain cell apoptosis during bacterial meningitis in vivo. Mutants lacking Ply and H2O2 significantly reduced neuronal damage during pneumococcal meningitis in rabbits, and in the absence of Ply, catalase treatment reduced neuronal damage in vivo. Moreover, H2O2 acts as an important vasodilator in early cerebral hyperfusion during pneumococcal meningitis [47]. Catalase treatment in rats reduced the regional cerebral blood flow and brain water content in the early phase of pneumococcal meningitis [48], but did not reduce the intracranial hypertension which appears to depend on the inflammatory host response such as the production of nitric oxide [49]. For that reason, catalase treatment might be, in combination with modulation of the host response, an effective treatment to reduce neuronal damage and regional cerebral blood flow.

Previous studies revealed that the production of H2O2 by pneumococci is directly correlated with the release of Ply, as strains lacking pyruvate oxidase (and lactate oxidase) showed decreased cytotoxicity and hemolytic activity [13, 50]. Notably, in our study, we have measured up to 2 mm H2O2 in the supernatant of infected microglia, but we were not able to detect any Ply in the cell culture medium with a hemolysis assay (limit of detection 1 HU/mL; Fig. 3b). Additionally, we analyzed whether Ply is released in the absence of microglial cells in the eukaryotic cell culture medium (DMEM) compared to a classical bacterial growth medium (THY). Indeed, after 4 and 6 h of growth, Ply was detectable in THY inoculated with S. pneumoniae D39 and D39Δcps by a hemolysis assay and Western blot analysis (online suppl. Fig. S6), but not when bacteria were cultured in the medium used for microglial cultivation and infection (DMEM). One reason for the lack of Ply in the eukaryotic cell culture medium might be the density of bacteria reached under these experimental conditions. In THY, approximately 5 × 108 CFU/mL were detected after 6 h of incubation in contrast to 5 × 107 CFU/mL in DMEM. It is known that pneumococci undergo autolysis when the stationary growth phase (>108 CFU/mL) is reached and that Ply is mainly released during autolysis [51]. We speculate that in DMEM the stationary growth phase is reached less rapidly, and therefore Ply is not released. Furthermore, we observed that the primary phagocytes used in our study are more resistant to the actions of rPly compared to the cell lines J774A.1 and HEp-2 (Fig. 3a). It is well known that cell lines often react differently compared to primary cells. It has been shown for Anthrolysin O, another CDC, that, in contrast to concentrations required for toxicity seen on the monocyte cell line THP-1, 10-fold higher doses of the toxin were required in primary human monocyte-derived macrophages [52].

A concentration of at least 1,025 HU/mL rPly (reflecting 2.2–5 μg/mL rPly) was required to measure cytotoxic effects of approximately 10% in primary microglia (Fig. 3a). Our results are in line with other studies in which Ply was cytotoxic to primary microglia in concentrations of 60 and 200 nm (3 and 10 μg/mL) [53]. In humans, 0.85–180 ng/mL Ply was found in CSF during meningitis [54]. Another study in humans suffering from S. pneumoniae meningitis revealed that Ply concentrations in CSF were higher (median 1 mg/mL) than concentrations inducing microglial cell injury in vitro. Moreover, in survivors, Ply levels significantly declined between the first and second lumbar punctures (interval approximately 48 h), while in non-survivors Ply levels remained high [55]. This suggests that Ply is produced in vivo and it is very likely that it exerts important functions during CNS infections. However, under our in vitro conditions, it was not possible to reproduce the release of these high concentrations of Ply. Ply has been shown to be toxic to brain cells, in particular neurons [10, 18, 19, 56, 57], and is able to inhibit microglial taxis [17]. An effect of Ply on microglial cell death has been attributed to different cell death mechanisms in microglial cell lines, namely, apoptosis, autophagy, and pyroptosis [10, 58]. At sublytic concentrations, Ply is able to induce changes in the cytoskeleton via RhoA and Rac1 [59] as well as a direct transmembrane actin interaction [60] and enhance dynamin endocytosis in primary microglia [61]. However, the effects of Ply on microglia at sublytic concentrations were not addressed in the present study.

In order to show that Ply can bind under the chosen experimental conditions to microglia and is active in the used cell culture medium, we performed a flow cytometric analysis (online suppl. Fig. S4) and determined the HU of rPly in DMEM (online suppl. Fig. S5), respectively. Indeed, the activity of rPly was decreased in the cell culture medium compared to a sodium chloride solution, but binding of Ply to microglial cells, in comparison with the macrophage cell line J774.A1, was detectable.

In conclusion, rPly is able to cause dose-dependent cytotoxic effects on microglia; however, the required concentrations were very high, and in the absence of Ply, H2O2 strongly affected eukaryotic cell viability. As the observed effect of high H2O2 production by pneumococci in medium was also seen in human CSF ex vivo, this effect is likely of physiological relevance in the brain. We show that primary microglia, brain phagocytes that should clear a pneumococcal infection, are rapidly and effectively killed by pneumococcal H2O2, presumably affecting the ability of the brain to prevent the entrance of pneumococci into nervous tissue.

We would like to thank Johanna Weißhaupt and Anna Richter, who actively supported the performance of some experiments, as well as Gerhard Burchardt and his team for providing the amino acid solutions for the H2O2 experiments. This project was performed by FJ in partial fulfillment of the requirements for the Ph.D. degree from the University of Veterinary Medicine Hannover.

C57BL/6 WT mice were cared for in accordance with the principles outlined in the “European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes” at the Institute for Biochemistry, University of Veterinary Medicine Hannover. The killing of neonatal mice for the extraction of primary cells was approved by the Lower Saxony State Office for Consumer and Food Safety, Germany.

The anonymized leftovers of human CSF samples from patients receiving diagnostic lumbar punctures were pooled. The patients did not suffer from any infectious disease and had no CSF pleocytosis. Written informed consent was obtained from the patients. This study protocol was reviewed and approved by the Ethics Committee of the University Medical Center Göttingen, Georg-August-University Göttingen, Germany (approval number: 34/1/05).

The authors have no conflict of interest to declare.

This project was partially supported by the Niedersachsen-Research Network on Neuroinfectiology (N-RENNT) of the Ministry of Science and Culture of Lower Saxony (PVW, FJ), the German Research Foundation (DFG, 374031971 – TRR 240 to SH), and BMBF (FZK 01DP19007) (SH). The funding bodies did not influence the design of the study, collection, analysis, and interpretation of data, and writing of the manuscript.

F.J. conducted experiments, took microscopic pictures, analyzed data, and drafted the manuscript. D.S. conducted experiments, analyzed data, and drafted the manuscript. R.N. supervised the study. T.P.K. and S.H. produced rPly, constructed mutant and wild-type S. pneumoniae strains, and supported the study design. D.H. contributed to the design of the flow cytometric experiments and analyzed data. P.V.-W. conceived the study and drafted the manuscript. J.S. contributed to the study design, performed experiments, analyzed data, and drafted the manuscript. All authors read and approved the final version of the manuscript.

Additional Information

Franziska Jennert and Désirée Schaaf contributed equally to this work.

All relevant data are contained within the manuscript and supporting information files.

1.
Bogaert
D
,
De Groot
R
,
Hermans
PW
.
Streptococcus pneumoniae colonisation: the key to pneumococcal disease
.
Lancet Infect Dis
.
2004
;
4
(
3
):
144
54
.
2.
Brouwer
MC
,
Tunkel
AR
,
van de Beek
D
.
Epidemiology, diagnosis, and antimicrobial treatment of acute bacterial meningitis
.
Clin Microbiol Rev
.
2010
;
23
(
3
):
467
92
.
3.
Doran
KS
,
Fulde
M
,
Gratz
N
,
Kim
BJ
,
Nau
R
,
Prasadarao
N
, et al
.
Host-pathogen interactions in bacterial meningitis
.
Acta Neuropathol
.
2016
;
131
(
2
):
185
209
.
4.
Pericone
CD
,
Overweg
K
,
Hermans
PW
,
Weiser
JN
.
Inhibitory and bactericidal effects of hydrogen peroxide production by Streptococcus pneumoniae on other inhabitants of the upper respiratory tract
.
Infect Immun
.
2000
;
68
(
7
):
3990
7
.
5.
Orihuela
CJ
,
Gao
G
,
Francis
KP
,
Yu
J
,
Tuomanen
EI
.
Tissue-specific contributions of pneumococcal virulence factors to pathogenesis
.
J Infect Dis
.
2004
;
190
(
9
):
1661
9
.
6.
Ogunniyi
AD
,
LeMessurier
KS
,
Graham
RM
,
Watt
JM
,
Briles
DE
,
Stroeher
UH
, et al
.
Contributions of pneumolysin, Pneumococcal Surface Protein A (PspA), and PspC to pathogenicity of Streptococcus pneumoniae D39 in a mouse model
.
Infect Immun
.
2007
;
75
(
4
):
1843
51
.
7.
García-Suárez
MM
,
Flórez
N
,
Astudillo
A
,
Vázquez
F
,
Villaverde
R
,
Fabrizio
K
, et al
.
The role of pneumolysin in mediating lung damage in a lethal pneumococcal pneumonia murine model
.
Respir Res
.
2007
;
8
(
1
):
3
.
8.
Hirst
RA
,
Gosai
B
,
Rutman
A
,
Guerin
CJ
,
Nicotera
P
,
Andrew
PW
, et al
.
Streptococcus pneumoniae deficient in pneumolysin or autolysin has reduced virulence in meningitis
.
J Infect Dis
.
2008
;
197
(
5
):
744
51
.
9.
Hirst
RA
,
Sikand
KS
,
Rutman
A
,
Mitchell
TJ
,
Andrew
PW
,
O'Callaghan
C
.
Relative roles of pneumolysin and hydrogen peroxide from Streptococcus pneumoniae in inhibition of ependymal ciliary beat frequency
.
Infect Immun
.
2000
;
68
(
3
):
1557
62
.
10.
Braun
JS
,
Sublett
JE
,
Freyer
D
,
Mitchell
TJ
,
Cleveland
JL
,
Tuomanen
EI
, et al
.
Pneumococcal pneumolysin and H(2)O(2) mediate brain cell apoptosis during meningitis
.
J Clin Invest
.
2002
;
109
(
1
):
19
27
.
11.
Pericone
CD
,
Park
S
,
Imlay
JA
,
Weiser
JN
.
Factors contributing to hydrogen peroxide resistance in Streptococcus pneumoniae include pyruvate oxidase (SpxB) and avoidance of the toxic effects of the fenton reaction
.
J Bacteriol
.
2003
;
185
(
23
):
6815
25
.
12.
Spellerberg
B
,
Cundell
DR
,
Sandros
J
,
Pearce
BJ
,
Idanpaan-Heikkila
I
,
Rosenow
C
, et al
.
Pyruvate oxidase, as a determinant of virulence in Streptococcus pneumoniae
.
Mol Microbiol
.
1996
;
19
(
4
):
803
13
.
13.
Bryant
JC
,
Dabbs
RC
,
Oswalt
KL
,
Brown
LR
,
Rosch
JW
,
Seo
KS
, et al
.
Pyruvate oxidase of Streptococcus pneumoniae contributes to pneumolysin release
.
BMC Microbiol
.
2016
;
16
(
1
):
271
.
14.
Barnett
TC
,
Cole
JN
,
Rivera-Hernandez
T
,
Henningham
A
,
Paton
JC
,
Nizet
V
, et al
.
Streptococcal toxins: role in pathogenesis and disease
.
Cell Microbiol
.
2015
;
17
(
12
):
1721
41
.
15.
Braun
JS
,
Novak
R
,
Gao
G
,
Murray
PJ
,
Shenep
JL
.
Pneumolysin, a protein toxin of Streptococcus pneumoniae, induces nitric oxide production from macrophages
.
Infect Immun
.
1999
;
67
(
8
):
3750
6
.
16.
Wellmer
A
,
Zysk
G
,
Gerber
J
,
Kunst
T
,
Von Mering
M
,
Bunkowski
S
, et al
.
Decreased virulence of a pneumolysin-deficient strain of Streptococcus pneumoniae in murine meningitis
.
Infect Immun
.
2002
;
70
(
11
):
6504
8
.
17.
Hupp
S
,
Grandgirard
D
,
Mitchell
TJ
,
Leib
SL
,
Hathaway
LJ
,
Iliev
AI
.
Pneumolysin and the bacterial capsule of Streptococcus pneumoniae cooperatively inhibit taxis and motility of microglia
.
J Neuroinflammation
.
2019
;
16
(
1
):
105
.
18.
Braun
JS
,
Hoffmann
O
,
Schickhaus
M
,
Freyer
D
,
Dagand
E
,
Bermpohl
D
, et al
.
Pneumolysin causes neuronal cell death through mitochondrial damage
.
Infect Immun
.
2007
;
75
(
9
):
4245
54
.
19.
Wippel
C
,
Maurer
J
,
Fortsch
C
,
Hupp
S
,
Bohl
A
,
Ma
J
, et al
.
Bacterial cytolysin during meningitis disrupts the regulation of glutamate in the brain, leading to synaptic damage
.
PLoS Pathog
.
2013
;
9
(
6
):
e1003380
.
20.
Nayak
D
,
Roth
TL
,
McGavern
DB
.
Microglia development and function
.
Annu Rev Immunol
.
2014
;
32
:
367
402
.
21.
Ginhoux
F
,
Greter
M
,
Leboeuf
M
,
Nandi
S
,
See
P
,
Gokhan
S
, et al
.
Fate mapping analysis reveals that adult microglia derive from primitive macrophages
.
Science
.
2010
;
330
(
6005
):
841
5
.
22.
Sheng
J
,
Ruedl
C
,
Karjalainen
K
.
Most tissue-resident macrophages except microglia are derived from fetal hematopoietic stem cells
.
Immunity
.
2015
;
43
(
2
):
382
93
.
23.
McGrath
KE
,
Frame
JM
,
Palis
J
.
Early hematopoiesis and macrophage development
.
Semin Immunol
.
2015
;
27
(
6
):
379
87
.
24.
Gordon
S
,
Taylor
PR
.
Monocyte and macrophage heterogeneity
.
Nat Rev Immunol
.
2005
;
5
(
12
):
953
64
.
25.
Mariani
MM
,
Kielian
T
.
Microglia in infectious diseases of the central nervous system
.
J Neuroimmune Pharmacol
.
2009
;
4
(
4
):
448
61
.
26.
Zysk
G
,
Schneider-Wald
BK
,
Hwang
JH
,
Bejo
L
,
Kim
KS
,
Mitchell
TJ
, et al
.
Pneumolysin is the main inducer of cytotoxicity to brain microvascular endothelial cells caused by Streptococcus pneumoniae
.
Infect Immun
.
2001
;
69
(
2
):
845
52
.
27.
Rennemeier
C
,
Hammerschmidt
S
,
Niemann
S
,
Inamura
S
,
Zähringer
U
,
Kehrel
BE
.
Thrombospondin-1 promotes cellular adherence of gram-positive pathogens via recognition of peptidoglycan
.
FASEB J
.
2007
;
21
(
12
):
3118
32
.
28.
Battig
P
,
Muhlemann
K
.
Influence of the spxB gene on competence in Streptococcus pneumoniae
.
J Bacteriol
.
2008
;
190
(
4
):
1184
9
.
29.
Surabhi
S
,
Jachmann
LH
,
Shumba
P
,
Burchhardt
G
,
Hammerschmidt
S
,
Siemens
N
.
Hydrogen peroxide is crucial for NLRP3 inflammasome-mediated IL-1β production and cell death in pneumococcal infections of bronchial epithelial cells
.
J Innate Immun
.
2022
;
14
(
3
):
192
206
.
30.
Seele
J
,
Nau
R
,
Prajeeth
CK
,
Stangel
M
,
Valentin-Weigand
P
,
Seitz
M
.
Astrocytes enhance Streptococcus suis-glial cell interaction in primary astrocyte-microglial cell Co-cultures
.
Pathogens
.
2016
;
5
(
2
):
43
.
31.
Redlich
S
,
Ribes
S
,
Schutze
S
,
Eiffert
H
,
Nau
R
.
Toll-like receptor stimulation increases phagocytosis of Cryptococcus neoformans by microglial cells
.
J Neuroinflammation
.
2013
;
10
:
71
.
32.
Abdissa
K
,
Ruangkiattikul
N
,
Ahrend
W
,
Nerlich
A
,
Beineke
A
,
Laarmann
K
, et al
.
Relevance of inducible nitric oxide synthase for immune control of Mycobacterium avium subspecies paratuberculosis infection in mice
.
Virulence
.
2020
;
11
(
1
):
465
81
.
33.
Duane
PG
,
Rubins
JB
,
Weisel
HR
,
Janoff
EN
.
Identification of hydrogen peroxide as a Streptococcus pneumoniae toxin for rat alveolar epithelial cells
.
Infect Immun
.
1993
;
61
(
10
):
4392
7
.
34.
McQuin
C
,
Goodman
A
,
Chernyshev
V
,
Kamentsky
L
,
Cimini
BA
,
Karhohs
KW
, et al
.
CellProfiler 3.0: next-generation image processing for biology
.
PLoS Biol
.
2018
;
16
(
7
):
e2005970
.
35.
Schneider
CA
,
Rasband
WS
,
Eliceiri
KW
.
NIH Image to ImageJ: 25 years of image analysis
.
Nat Methods
.
2012
;
9
(
7
):
671
5
.
36.
Pick
E
,
Keisari
Y
.
A simple colorimetric method for the measurement of hydrogen peroxide produced by cells in culture
.
J Immunol Methods
.
1980
;
38
(
1–2
):
161
70
.
37.
Hartel
T
,
Eylert
E
,
Schulz
C
,
Petruschka
L
,
Gierok
P
,
Grubmuller
S
, et al
.
Characterization of central carbon metabolism of Streptococcus pneumoniae by isotopologue profiling
.
J Biol Chem
.
2012
;
287
(
6
):
4260
74
.
38.
Jahn
K
,
Handtke
S
,
Palankar
R
,
Weißmüller
S
,
Nouailles
G
,
Kohler
TP
, et al
.
Pneumolysin induces platelet destruction, not platelet activation, which can be prevented by immunoglobulin preparations in vitro
.
Blood Adv
.
2020
;
4
(
24
):
6315
26
.
39.
Takeda
H
,
Tomita
M
,
Tanahashi
N
,
Kobari
M
,
Yokoyama
M
,
Takao
M
, et al
.
Hydrogen peroxide enhances phagocytic activity of ameboid microglia
.
Neurosci Lett
.
1998
;
240
(
1
):
5
8
.
40.
Pathipati
P
,
Muller
S
,
Jiang
X
,
Ferriero
D
.
Phenotype and secretory responses to oxidative stress in microglia
.
Dev Neurosci
.
2013
;
35
(
2–3
):
241
54
.
41.
Hirrlinger
J
,
Resch
A
,
Gutterer
JM
,
Dringen
R
.
Oligodendroglial cells in culture effectively dispose of exogenous hydrogen peroxide: comparison with cultured neurones, astroglial and microglial cells
.
J Neurochem
.
2002
;
82
(
3
):
635
44
.
42.
Soto-Diaz
K
,
Vailati-Riboni
M
,
Louie
AY
,
McKim
DB
,
Gaskins
HR
,
Johnson
RW
, et al
.
Treatment with the CSF1R antagonist GW2580, sensitizes microglia to reactive oxygen species
.
Front Immunol
.
2021
;
12
:
734349
.
43.
Thorsdottir
S
,
Henriques-Normark
B
,
Iovino
F
.
The role of microglia in bacterial meningitis: inflammatory response, experimental models and new neuroprotective therapeutic strategies
.
Front Microbiol
.
2019
;
10
:
576
.
44.
Nau
R
,
Bruck
W
.
Neuronal injury in bacterial meningitis: mechanisms and implications for therapy
.
Trends Neurosci
.
2002
;
25
(
1
):
38
45
.
45.
Bohr
V
,
Paulson
OB
,
Rasmussen
N
.
Pneumococcal meningitis. Late neurologic sequelae and features of prognostic impact
.
Arch Neurol
.
1984
;
41
(
10
):
1045
9
.
46.
Durand
ML
,
Calderwood
SB
,
Weber
DJ
,
Miller
SI
,
Southwick
FS
,
Caviness
VS
Jr
, et al
.
Acute bacterial meningitis in adults. A review of 493 episodes
.
N Engl J Med
.
1993
;
328
(
1
):
21
8
.
47.
Pfister
HW
,
Feiden
W
,
Einhaupl
KM
.
Spectrum of complications during bacterial meningitis in adults. Results of a prospective clinical study
.
Arch Neurol
.
1993
;
50
(
6
):
575
81
.
48.
Hoffmann
OM
,
Becker
D
,
Weber
JR
.
Bacterial hydrogen peroxide contributes to cerebral hyperemia during early stages of experimental pneumococcal meningitis
.
J Cereb Blood Flow Metab
.
2007
;
27
(
11
):
1792
7
.
49.
Kim
YS
,
Tauber
MG
.
Neurotoxicity of glia activated by gram-positive bacterial products depends on nitric oxide production
.
Infect Immun
.
1996
;
64
(
8
):
3148
53
.
50.
Bazant
J
,
Ott
B
,
Hudel
M
,
Hain
T
,
Lucas
R
,
Mraheil
MA
.
Impact of endogenous pneumococcal hydrogen peroxide on the activity and release of pneumolysin
.
Toxins
.
2023
;
15
(
10
):
593
.
51.
Boulnois
GJ
,
Paton
JC
,
Mitchell
TJ
,
Andrew
PW
.
Structure and function of pneumolysin, the multifunctional, thiol-activated toxin of Streptococcus pneumoniae
.
Mol Microbiol
.
1991
;
5
(
11
):
2611
6
.
52.
Mosser
EM
,
Rest
RF
.
The Bacillus anthracis cholesterol-dependent cytolysin, Anthrolysin O, kills human neutrophils, monocytes and macrophages
.
BMC Microbiol
.
2006
;
6
:
56
.
53.
Ebert
S
,
Gerber
J
,
Bader
S
,
Mühlhauser
F
,
Brechtel
K
,
Mitchell
TJ
, et al
.
Dose-dependent activation of microglial cells by Toll-like receptor agonists alone and in combination
.
J Neuroimmunol
.
2005
;
159
(
1–2
):
87
96
.
54.
Spreer
A
,
Kerstan
H
,
Bottcher
T
,
Gerber
J
,
Siemer
A
,
Zysk
G
, et al
.
Reduced release of pneumolysin by Streptococcus pneumoniae in vitro and in vivo after treatment with nonbacteriolytic antibiotics in comparison to ceftriaxone
.
Antimicrob Agents Chemother
.
2003
;
47
(
8
):
2649
54
.
55.
Wall
EC
,
Gordon
SB
,
Hussain
S
,
Goonetilleke
URS
,
Gritzfeld
J
,
Scarborough
M
, et al
.
Persistence of pneumolysin in the cerebrospinal fluid of patients with pneumococcal meningitis is associated with mortality
.
Clin Infect Dis
.
2012
;
54
(
5
):
701
5
.
56.
Reiss
A
,
Braun
JS
,
Jager
K
,
Freyer
D
,
Laube
G
,
Buhrer
C
, et al
.
Bacterial pore-forming cytolysins induce neuronal damage in a rat model of neonatal meningitis
.
J Infect Dis
.
2011
;
203
(
3
):
393
400
.
57.
Stringaris
AK
,
Geisenhainer
J
,
Bergmann
F
,
Balshüsemann
C
,
Lee
U
,
Zysk
G
, et al
.
Neurotoxicity of pneumolysin, a major pneumococcal virulence factor, involves calcium influx and depends on activation of p38 mitogen-activated protein kinase
.
Neurobiol Dis
.
2002
;
11
(
3
):
355
68
.
58.
Kim
JY
,
Paton
JC
,
Briles
DE
,
Rhee
DK
,
Pyo
S
.
Streptococcus pneumoniae induces pyroptosis through the regulation of autophagy in murine microglia
.
Oncotarget
.
2015
;
6
(
42
):
44161
78
.
59.
Iliev
AI
,
Djannatian
JR
,
Nau
R
,
Mitchell
TJ
,
Wouters
FS
.
Cholesterol-dependent actin remodeling via RhoA and Rac1 activation by the Streptococcus pneumoniae toxin pneumolysin
.
Proc Natl Acad Sci U S A
.
2007
;
104
(
8
):
2897
902
.
60.
Hupp
S
,
Förtsch
C
,
Wippel
C
,
Ma
J
,
Mitchell
TJ
,
Iliev
AI
.
Direct transmembrane interaction between actin and the pore-competent, cholesterol-dependent cytolysin pneumolysin
.
J Mol Biol
.
2013
;
425
(
3
):
636
46
.
61.
Hupp
S
,
Förtsch
C
,
Graber
F
,
Mitchell
TJ
,
Iliev
AI
.
Pneumolysin boosts the neuroinflammatory response to Streptococcus pneumoniae through enhanced endocytosis
.
Nat Commun
.
2022
;
13
(
1
):
5032
.