Glucose and Oxygen Levels Modulate the Pore-Forming Effects of Cholesterol-Dependent Cytolysin Pneumolysin from Streptococcus pneumoniae

Abstract

A major Streptococcus pneumoniae pathogenic factor is the cholesterol-dependent cytolysin pneumolysin, binding membrane cholesterol and producing permanent lytic or transient pores. During brain infections, vascular damage with variable ischemia occurs. The role of ischemia on pneumolysin's pore-forming capacity remains unknown. In acute brain slice cultures and primary cultured glia, we studied acute toxin lysis (via propidium iodide staining and LDH release) and transient pore formation (by analyzing increases in the intracellular calcium). We analyzed normal peripheral tissue glucose conditions (80 mg%), normal brain glucose levels (20 mg%), and brain hypoglycemic conditions (3 mg%), in combinations either with normoxia (8% oxygen) or hypoxia (2% oxygen). At 80 mg% glucose, hypoxia enhanced cytolysis via pneumolysin. At 20 mg% glucose, hypoxia did not affect cell lysis, but impaired calcium restoration after non-lytic pore formation. Only at 3 mg% glucose, during normoxia, did pneumolysin produce stronger lysis. In hypoglycemic (3 mg% glucose) conditions, pneumolysin caused a milder calcium increase, but restoration was missing. Microglia bound more pneumolysin than astrocytes and demonstrated generally stronger calcium elevation. Thus, our work demonstrated that the toxin pore-forming capacity in cells continuously diminishes when oxygen is reduced, overlapping with a continuously reduced ability of cells to maintain homeostasis of the calcium influx once oxygen and glucose are reduced.

Keywords: Streptococcus pneumoniae; astrocytes; brain; hypoglycemia; hypoxia; ischemia; microglia; pneumolysin; pore formation; transient pores.

Figure 1

Schematic representation of different tissue layers surrounding an ischemic core.

Figure 2

Lytic effects in acute brain slices in normoxic and hypoxic environments. (A) Propidium iodide (PI) permeabilization of mixed glia following 2 HU/mL PLY in normoxic and hypoxic conditions, showing increased permeabilization during hypoxia. Glucose level at 80 mg%, comparable with a normal peripheral (non-brain) tissue glucose concentration. (B) Exposure to 2 HU/mL PLY in acute slices with active perfusion with carbogen (95%O₂/5% CO₂) (normoxia) and with reduced oxygen levels (on a shaker in a 5% CO₂ incubator with normal atmospheric air). Lysis was measured 8 h following each PLY challenge according to the release of LDH (lactate dehydrogenase) in the medium. (C) Effect of 10 µM MK-801 on the lytic properties of 2 HU/mL PLY in the acute slices incubated in hypoxic conditions. The elevated lysis in the presence of PLY was not reverted. Brain slices were incubated at a glucose level of 80 mg%. All values represent the mean ± SEM; groups are compared using a one-way ANOVA; values above the bars indicate p when significant; n indicates independent experiments.

Figure 3

Lytic effects of PLY in different conditions of hypoxia/hypoglycemia in mixed glia. Dynamic propidium iodide-based permeabilization of mixed primary mouse glial cells following exposure to 2 HU/mL PLY. Only in conditions of isolated hypoglycemia was permeabilization elevated. In dual hypoxic/hypoglycemic conditions, permeabilization was diminished and the half-time was elevated, indicative of slower toxin binding and/or pore formation. For all groups, n = 6 independent experiments.

Figure 4

Calcium elevation in mixed glia caused by PLY. (A) Cal520 fluorescence increase as a marker of Ca elevation in 3 individual microglial cells from primary glial cultures (2 HU/mL PLY was applied at timepoint 0) for demonstrative purposes. Schematic cell images from BioRender. (B) Cal520 fluorescence increase following PLY exposure in 4 astrocyte cells. Note the difference in the pattern of Ca fluctuation due to the syncytium-like interconnections between the astrocytes, allowing very flexible Ca buffering and wave distribution. (C) Cumulative curves of Ca elevation in microglia following exposure to 2 HU/mL PLY; n indicates the number of cells, pooled together from 4 independent experiments. Below the curves, a statistical comparison of the areas under the curves is presented. (D) Cumulative curves of Ca elevation in astrocytes following exposure to 2 HU/mL PLY; n indicates the number of cells, pooled together from 4 independent experiments. Below the curves, a statistical comparison of the areas under the curves is presented. (E) Comparison of the area under the curve (AUC) between the cumulative elevation of Ca induced by PLY in microglia and in astrocytes in conditions of normal oxygen and normal glucose, indicating a stronger effect in microglia. All values represent the mean ± SEM; groups are compared using the Mann–Whitney U-test; values above the bars indicate p when significant; n indicates individual cells; pooled together from 3 experiments. (F) Tabular presentation of the Ca changes in microglia and in astrocytes under different conditions ((~) indicates normoxic/normoglycemic conditions and their equivalents).

Figure 5

Cholesterol content in microglia and astrocytes. (A) Relative fluorescent intensity of filipin staining as a marker of cholesterol content in microglia and astrocytes, demonstrating a higher content in the microglia. (B) False-color fluorescent filipin image of an astrocyte (large cell, As) and microglia (Mi). (C) Relative fluorescent intensity of the bound PLY-EGFP after 10 min of exposure (4 HU/mL), demonstrating increased PLY binding in the microglia. (D) Images of microglia/astrocyte co-culture with the microglia outlined (left, transmission image). Microglial cells with exceptionally strong PLY-EGFP binding are indicated with arrows (fluorescent image of PLY-EGFP, right). All values represent the mean ± SEM; groups are compared using the Mann–Whitney U-test; values above the bars indicate p when significant; n indicates independent experiments. All scale bars: 20 µm.

Figure 6

Diminished endocytosis rate during hypoglycemia. Following 2 h of cell conditioning at 20 mg% (normoglycemia) and 3 mg% (hypoglycemia), the basic FM 4-64 turnover was recorded and compared between normoglycemic brain conditions (20 mg%) and hypoglycemic brain conditions (3 mg%). All values represent the mean ± SEM; groups are compared using the Mann–Whitney U-test; values above the bars indicate p when significant; n indicates independent experiments.

Figure 7

Effect of 0.1 and 1 HU/mL PLY on red blood cell hemolysis under various glucose conditions, and in normoxic and hypoxic environments (37 °C, 30 min). While in normoxia, glucose variation did not produce any change in hemolysis, in hypoxia, diminishing glucose concentrations demonstrated a mild reduction in the toxin’s effects. All values represent the mean ± SEM; groups are compared using the Wilcoxon paired test; values above the bars indicate p when significant; n indicates independent replicates.

Figure 8

Schematic presentation of the major changes from normoxia/normoglycemia to hypoxia/hypoglycemia (ischemia).