Group B Streptococcus interactions with human meningeal cells and astrocytes in vitro

Abstract

Background: Streptococcus agalactiae (Group B Streptococcus, GBS) is a leading cause of life-threatening neonatal meningitis and survivors often suffer permanent neurological damage. How this organism interacts with the meninges and subsequently with astrocytes that constitute the underlying cortical glia limitans superficialis is not known.

Methodology/principal findings: In this paper, we demonstrate dose-dependent adherence of GBS over time to human meningioma cells and fetal astrocytes in vitro, which was not influenced by expression of either β-haemolysin/cytolysin (β-h/c) toxin, different capsule serotypes or by absence of capsule (p>0.05). Internalization of GBS by both cell types was, however, a slow and an infrequent event (only 0.02-0.4% of associated bacteria were internalised by 9 h). Expression of β-h/c toxin did not play a role in invasion (p>0.05), whereas capsule expression lead to a reduction (p<0.05) in the numbers of intracellular bacteria recovered. GBS strains induced cytotoxicity as demonstrated by the measurement of lactate dehydrogenase (LDH) enzyme release by 9 h and by viable staining. Increasing levels of meningioma cell death correlated with bacterial growth and the phenotype of β-h/c toxin production, i.e. from weakly, to normo- to hyper-haemolytic. However, cytotoxicity was significantly greater (p<0.05) towards astrocytes, and infection with initial MOI≥0.003 induced 70-100% LDH release. By comparing wild-type (β-h/c(+)) and mutant (ΔcylE β-h/c(-)) strains and β-h/c toxin extracts and by using the surfactant dipalmitoylphosphatidylcholine in cytotoxicity inhibition experiments, β-h/c toxin was demonstrated as principally responsible for cell death.

Conclusions/significance: This study has described key events in the interactions of GBS with meningeal cells and astrocytes in vitro and a major virulence role for β-h/c toxin. Understanding the mechanisms involved will help to identify potential therapies for improving patient survival and for reducing the incidence and severity of neurological sequelae.

Figure 1. Association of GBS strains with human meningioma cell lines.

Human meningioma cell lines were infected with various MOI (0.0003–3000) of GBS strains of different serotypes and also with strains deficient in expression of β-h/c toxin (ΔcylE) or capsule. Bacterial association was quantified over time with the symbols representing the mean and the error bars the standard error of the mean from three independent experiments. Similar data were obtained using both cell lines.

Figure 2. Meningioma cell injury induced by wild type GBS strains.

A) Release of LDH: cells were infected for 9 h with different initial MOI of GBS strains and LDH release was measured. Results shown are the mean values of LDH release compared with maximum release by lysed cells and the error bars are the SEM from at least two independent experiments, each performed in triplicate. B) Confocal microscopy: cell cultures were infected with a range of initial MOI of different serotypes of wild type GBS for 9 h. Cell death was examined using the LIVE/DEAD fluorescent dye assay, where uptake of the red dye (ethidium homodimer) identifies dead cells and uptake of the green dye (calcein AM) identifies live cells. The scale bar shows 75 µm (×40 lens magnification). Images are of monolayers infected for 9 h, except for the COH-1 strain infected for 24 h, and they are representative of experiments carried out in triplicate. Similar data were obtained using both cell lines.

Figure 3. Comparison of β-hemolysin/cytolysin (β-h/c) toxin-expressing and toxin-deficient strains on meningioma cell death.

Cell cultures were infected with various doses of wild-type A909 and NCTC10/84 (β-h/c⁺) strains and their isogenic ΔcylE (β-h/c⁻) mutants and meningioma cell death quantified after 9 h. (A) Bacterial dose-dependent LDH release from meningioma cells. The symbols represent the mean values of LDH release compared with maximum release by lysed cells and the error bars are the SEM from at least two independent experiments. (B) Confocal microscopy images of meningioma cells challenged with different concentrations of GBS A909 and NCTC10/84 ΔcylE mutant strains for 9 h and 24 h. The red colour identifies dead cells, whereas the green colour corresponds to viable cells. The scale bar at bottom right of each image measures 75 µm (×40 magnification). Similar data were obtained using both cell lines.

Figure 4. The surfactant DPPC protects meningioma cells from GBS-induced death.

(A) Effect of DPPC on cell death induced by live bacterial infection. Meningioma cells were infected with doses of GBS strains that induced the highest levels of LDH release (Figure 2A) and cell death quantified at 9 h. The columns represent the mean levels of LDH released from monolayers and the arrows indicate the percentage of reduction of LDH release in the presence of DPPC in comparison with LDH in the absence of DPPC. The error bars show the SEM of at least two independent experiments. The two confocal microscopy images show monolayers infected with a cytotoxic dose (initial MOI, 30) of the hyper-haemolytic NCTC10/84 strain for 9 h in the absence and presence of DPPC. (B) The protective effect of DPPC is dependent on time of addition. Meningioma cell monolayers were infected with GBS strain NCTC10/84 and DPPC (3 mg/ml) added at 0, 3 and 6 h. LDH release was measured at 9 h. The columns represent the mean levels of LDH release and the error bars the SEM of 3 independent infection experiments. (C) Cell death induced by β-h/c toxin extracts and inhibition by DPPC. The graph shows LDH release from monolayers treated with β-h/c toxin extracts prepared from wild type NCTC and A909 strains (containing 100HU and 50HU, respectively). Equal volumes of extracts prepared from their β-h/c deficient mutant strains were also used. Results shown are mean values of LDH release and error bars are the SEM from four independent experiments using two independent batches of β-h/c toxin extracts. Arrows indicate the percentage inhibition of LDH release in the presence and absence of DPPC. The confocal microscopy images show monolayers treated with β-h/c extracts (250HU) prepared from wild type GBS strains for 9 h in the absence and presence of DPPC. An equivalent volume of extract prepared from the ΔcylE (β-h/c deficient) mutant strain was also used. Using the LIVE/DEAD assay, the red colour identifies dead cells, whereas the green colour corresponds to viable cells. The scale bar at bottom right of each image measures 75 µm (×40 magnification). Similar data were obtained using both cell lines.

Figure 5. GBS interactions with human fetal astrocytes.

a&b) Association of wild-type A909 and NCTC10/84 (β-h/c⁺) strains and their isogenic ΔcylE (β-h/c⁻) mutants. SVGmm fetal astrocyte cell monolayers were infected with various initial MOI of GBS strains and association measured over time. Data are from representative experiments (n = 3 for each bacterium) and the symbols show the mean cfu values and the error bars the standard deviations of triplicate wells. c) Astrocyte cell death measured by LDH release. Astrocytes were infected with various initial MOI of wild-type A909 and NCTC10/84 (β-h/c⁺) strains and their isogenic ΔcylE (β-h/c⁻) mutants. LDH release was measured after 9 h and the symbols represent the mean levels of LDH release and the error bars the SEM from 3 independent experiments. Open circles denote NCTC10/84 (β-h/c⁺) and open triangles NCTC10/84 ΔcylE (β-h/c⁻); closed circles denote A909 (β-h/c⁺) and closed triangles A909 ΔcylE (β-h/c⁻); closed diamonds denote background release of LDH. d) Effect of DPPC on astrocyte cell death induced by normo-haemolytic GBS infection. Astrocyte cell monolayers were infected with wild-type strains A909 (initial MOI 3000) in the presence (3 mg/ml) or absence of DPPC, and LDH release measured after 9 h. Control (Con) is spontaneous LDH release from cells. The columns represent mean levels of LDH release and the error bars the SEM from 3 independent experiments. e) Effect of DPPC on astrocyte cell death induced by hyper-haemolytic GBS infection. Cell monolayers were infected with different initial MOI of NCTC10/84 (0.0003–3000) for 9 h in the presence (3 mg/ml) or absence of DPPC. Data are from a representative experiment (n = 3) and the symbols show the mean levels of LDH release and the error bars the standard deviation of triplicate wells. Open and closed circles denote GBS without and with DPPC respectively; open triangles denote background release of LDH. f) Effect of DPPC on β-h/c toxin induced astrocyte cell death. Fetal astrocyte cell monolayers were treated for 9 h with β-h/c⁺ and β-h/c⁻ extract (100 µl volume per well and prepared as described in Materials and Methods) in the presence or absence of DPPC (3 mg/ml). LDH release was measured after 9 h. The columns represent mean levels of LDH release and the error bars the SEM from experiments carried out with 3 independent preparations of toxin extracts.