Missing elimination via membrane vesicle shedding contributes to the diminished calcium sensitivity of listeriolysin O

Abstract

The lytic capacity of cholesterol-dependent cytolysins is enhanced in the extracellular calcium-free environment through a combination of limited membrane repair and diminished membrane toxin removal. For a typical neurotoxin of the group, pneumolysin, this effect has already been observed at reduced (1 mM) calcium conditions, which are pathophysiologically relevant. Here, we tested another neurotoxin of the group, listeriolysin O from L. monocytogenes, active in the primary vacuole after bacterium phagocytosis in host cells. Reduced calcium did not increase the lytic capacity of listeriolysin (in contrast to pneumolysin), while calcium-free conditions elevated it 2.5 times compared to 10 times for pneumolysin (at equivalent hemolytic capacities). To clarify these differences, we analyzed membrane vesicle shedding, known to be a calcium-dependent process for toxin removal from eukaryotic cell membranes. Both pneumolysin and listeriolysin initiated vesicle shedding, which was completely blocked by the lack of extracellular calcium. Lack of calcium, however, elevated the toxin load per a cell only for pneumolysin and not for listeriolysin. This result indicates that vesicle shedding does not play a role in the membrane removal of listeriolysin and outlines a major difference between it and other members of the CDC group. Furthermore, it provides new tools for studying membrane vesicle shedding.

Figure 1

Cytolytic capacity of LLO and PLY in 2 mM Ca²⁺ and calcium-free conditions. (A) The lytic capacity (as judged by the fraction of propidium iodide-positive cells) of 2 and 5 HU/ml LLO increases with a reduction of calcium from 2 mM (left) to calcium-free conditions (right). Significance is calculated with the Wilcoxon matched pair test with individual time points paired to compare the lytic capacity of equivalent LLO concentrations at 0 and at 2 mM Ca²⁺. (B) Comparison of equivalent lytic amounts of LLO and PLY reveal much stronger increase after calcium removal for PLY (approx. 10 times) vs. LLO (approx. 2.5 times). All values represent the mean ± SEM of 5 independent experiments (n = 5).

Figure 2

Cytolytic capacity of LLO in reduced calcium conditions. The reduction of calcium from 2 to 1 mM does not increase the lytic capacity of LLO when cells were challenged with 2 HU/ml (A) or 5 HU/ml (B). All values represent the mean ± SEM of 5 independent experiments (n = 5).

Figure 3

Extracellular vesicle shedding in primary mixed glial cells. Confocal imaging of fluorescent CellMask stained membranes (z-plane reconstruction in panels A–D) at 0 and at 5 min after exposure to toxin. (A) 2 HU/ml pneumolysin (PLY) in 2 mM Ca²⁺ conditions. Red arrows indicate shed extracellular vesicles. Scale bar: 8 µm. (B) Lack of membrane vesicle shedding after cell challenge with 2 HU/ml PLY in calcium-free conditions. (C) Membrane vesicle shedding (red arrows at 5 min) after challenge with 2 HU/ml listeriolysin (LLO). (D) Inhibited vesicle shedding after challenge with 2 HU/ml LLO. E. Diagrams of the number of vesicles/field (z-plane reconstruction field with the size 60 × 10 µm) before and after toxin challenge at 2 mM Ca²⁺ (left) and calcium-free conditions (right), clearly indicating the inhibition of vesicle shedding for both toxins in calcium-free conditions. All values represent the mean ± SEM of 5 independent experiments (n = 5). Significance calculated with the Wilcoxon matched pair test with individual time point pairing.

Figure 4

Western blot analysis of toxin load per membrane and per a cell. (A) Membrane-bound LLO after exposure to 2 HU/ml LLO for 20 min in 2 mM Ca²⁺ and in calcium-free conditions, normalized to the membrane marker Na/K-ATPase. The drop of LLO in calcium-free conditions indicates more Na/K-ATPase and less LLO is present in shed vesicles. #1 to #5 indicate independent experiments‘ samples. (B) Membrane-bound PLY after exposure to 2 HU/ml PLY for 20 min remains unchanged in 2 mM Ca2+ and in calcium-free conditions, normalized to the membrane marker Na/K-ATPase. This indicates shed vesicles contained proportional amounts of toxin and membrane marker. #1 to #5 indicate independent experiments‘ samples. (C) Membrane toxin load normalized to cytosolic actin (i.e. the number of cells) demonstrating increased load per a cell after calcium depletion for PLY, but not for LLO. (D) Schematic explanation of the results when normalizing to a membrane and to a cytosolic protein. All values represent mean ± SEM of 5 or 6 independent experiments (n = 5–6). All bands of interest are cropped from the source blots, as lines of cropping are indicated with dotted line. All loading controls with actin and Na/K-ATPase correspond to the identical samples with LLO or PLY staining in the same panel of the figure. Source blots in unprocessed form are included in Supplementary Figures S2–S4. Analysis is performed using the Wilcoxon matched pair test.