Streptolysin S induces mitochondrial damage and macrophage death through inhibiting degradation of glycogen synthase kinase-3β in Streptococcus pyogenes infection

Abstract

Group A Streptococcus (GAS) infection is associated with a variety of human diseases. Previous studies indicate GAS infection leads to RAW264.7 cell death, but the mechanism is unclear. Here, analyzing the timing of reactive oxygen species (ROS) production and using mitochondrial ROS scavenger, we found the wild type GAS-induced RAW264.7 cell death was associated with mitochondrial ROS. The wild type GAS infection could activate glycogen synthase kinase-3β (GSK-3β). Inhibition of GSK-3β activity by lithium chloride or decreasing GSK-3β expression by lentivirus-mediated short hairpin RNA for GSK-3β could not only decrease the wild type GAS-induced mitochondrial ROS generation, mitochondria damage and cell death, but also reduced GAS intracellular replication. Streptolysin S (SLS), a GAS toxin, played the important role on GAS-induced macrophage death. Compared to the wild type GAS with its isogenic sagB mutant (SLS mutant)-infected macrophages, we found sagB mutant infection caused less mitochondrial ROS generation and cell death than those of the wild type GAS-infected ones. Furthermore, the sagB mutant, but not the wild type or the sagB-complementary mutant, could induce GSK-3β degradation via a proteasome-dependent pathway. These results suggest that a new mechanism of SLS-induced macrophage death was through inhibiting GSK-3β degradation and further enhancing mitochondrial damage.

Figure 1

ROS-mediated cytotoxicity in the GAS-infected RAW264.7 cells. RAW264.7 cells were infected with the wild type GAS at MOI 10 or 25. The levels of ROS were measured by carboxy-H₂DCFDA (a) at different times post-infection or by MitoSOX (b) staining at 3 h post-infection. Fluorescence intensity % was shown and expressed as described in Materials and Methods. Results are represented as mean ± standard deviation (SD). **P < 0.01, ***P < 0.001 compared with medium only group (One-way ANOVA test followed by Tukey’s test; n = 4). The culture supernatants of RAW264.7 cells infected with the GAS at MOI 10 or 25 (c), and the culture supernatants of RAW264.7 cells infected with GAS at a MOI of 25 in the presence of different concentrations of NAC (d) were collected at 18 h post-infection and then measured by LDH detection kit. In (c), LDH release % was shown and expressed as the mean ± SD, as described in Materials and Methods. ***P < 0.001 compared with medium only group (One-way ANOVA test followed by Tukey’s test; n = 4). In (d), LDH release % was shown and expressed as the mean ± SD, as described in Materials and Methods. ***P < 0.001 compared with GAS-infected group versus medium only group. (One-way ANOVA test followed by Tukey’s test; n = 4).***P < 0.001 compared with 20 mM NAC- treated group versus GAS-infected group (One-way ANOVA test followed by Tukey’s test; n = 4). *P < 0.05 compared with 10 mM NAC- treated group versus GAS-infected group (One-way ANOVA test followed by Tukey’s test; n = 4).

Figure 2

LiCl inhibited the mitochondrial ROS production and prevented the loss of Δψm and cell death in the GAS-infected RAW264.7 cells. RAW264.7 cells were infected with the wild type GAS for 1 h at a MOI of 25 in the presence of different concentrations of LiCl. Mitochondrial ROS and Δψm of the GAS-infected cells were measured by MitoSOX and JC-1 respectively at 3 h post-infection. (a) Mitochondrial ROS production in the GAS-infected RAW264.7 cells was inhibited by LiCl in a dose-dependent manner. Fluorescence intensity % was shown and expressed as described in Materials and Methods. Results are represented as mean ± SD. In LiCl-treated groups, **P < 0.01, ***P < 0.001 compared with the GAS-infected group. In the GAS-infected group, ***P < 0.001 compared with medium only (without GAS) group (One-way ANOVA test followed by Tukey’s test; n = 4). (b) The change of Δψm in GAS-infected cells was inhibited by LiCl. The level of Δψm in RAW264.7 cells was detected by JC-1 as described in Materials and Methods. In LiCl-treated groups, **P < 0.01, ***P < 0.001 compared with the GAS-infected group. In the GAS-infected group, ***P < 0.001 compared with medium only (without GAS) group (One-way ANOVA test followed by Tukey’s test; n = 4). (c) The LDH release of GAS-infected cells was dose-dependently inhibited by LiCl. The culture supernatants of GAS-infected RAW264.7 cells were collected at 18 h post-infection and then measured by LDH detection kit. LDH release % was shown and expressed as the mean ± SD, as described in Materials and Methods. In LiCl-treated groups, **P < 0.01, ***P < 0.001 compared with GAS-infected group. In the GAS-infected group, ***P < 0.001 compared with medium only (without GAS) group (One-way ANOVA test followed by Tukey’s test; n = 4).

Figure 3

Knockdown of GSK-3β attenuated mitochondrial ROS production and cell death of the GAS-infected RAW264.7 cells. The expression of GSK-3β was silenced in RAW264.7 cells using lentiviral-based shRNA (shGSK-3β) construct and the luciferase shRNA construct was using as negative control. The transfected cells and RAW264.7 cells (without transfection) were infected with GAS at a MOI of 25 as described in Materials and Methods. (a) Mitochondrial ROS levels were measured by MitoSOX staining at 3 h post GAS infection. Fluorescence intensity % was shown and expressed as described in Materials and Methods. Results are represented as mean ± SD. In GAS-infected groups, ***P < 0.001 compared with shGSK-3β knockdown group versus negative control group (One-way ANOVA test followed by Tukey’s test; n = 4) or versus the group of RAW264.7 cells (One-way ANOVA test followed by Tukey’s test; n = 4). (b) The GAS-mediated cell deaths were measured by LDH release at 18 h post GAS infection, as described in Materials and Methods. **P < 0.01 compared with shGSK-3β knockdown group versus negative control group (One-way ANOVA test followed by Tukey’s test; n = 4) or versus the group of RAW264.7 cells (One-way ANOVA test followed by Tukey’s test; n = 4).

Figure 4

Inhibition of GSK-3β by LiCl or knockdown of GSK-3β reduced the bacterial loads of the GAS-infected RAW264.7 cells. RAW264.7 cells that pretreated with 1,000 μM of LiCl (a) or transfected with the shLuc construct or shGSK-3β construct (b) were infected with GAS at MOI of 25. The number of remnant bacteria in each group was determined by plate count at 1.5 h or 18 h post GAS infection and shown as described in Materials and Methods. Results are represented as mean ± SD. In (a), **P < 0.01 compared with non-LiCl group (Unpaired Student’s t-test; n = 4). In (b), *P < 0.05 compared with shGSK-3β knockdown group versus negative control group (One-way ANOVA test followed by Tukey’s test; n = 4) or versus the group of RAW264.7 cells (One-way ANOVA test followed by Tukey’s test; n = 4).

Figure 5

The sagB mutation decreased mitochondrial ROS production and prevented mitochondrial dysfunction and cell death in the GAS-infected RAW264.7 cells. RAW264.7 cells were infected with the wild type GAS, isogenic sagB mutant or sagB- complementary mutant at a MOI of 25 as described in Materials and Methods. (a) The levels of ROS were measured by carboxy-H₂DCFDA at 1 h post GAS infection, as described in Materials and Methods. Results are represented as mean ± SD. **P < 0.01 compared with medium only group. *P < 0.05 compared with sagB mutant group versus GAS group or versus sagB-complementary group (One-way ANOVA test followed by Tukey’s test; n = 4). Mitochondrial ROS levels and Δψm were measured by MitoSOX (b) and JC-1 (c) at 3 h post-infection respectively, as described in Materials and Methods. Results are represented as mean ± SD. In (b), ***P < 0.001 compared with GAS-infected group versus medium only group. ***P < 0.001 compared with sagB mutant group versus GAS group or versus sagB-complementary group (One-way ANOVA test followed by Tukey’s test; n = 4). In (c), ***P < 0.001 compared with GAS-infected group versus medium only group. **P < 0.01 compared with sagB mutant group versus GAS group. *P < 0.05 compared with sagB mutant group versus sagB-complementary group (One-way ANOVA test followed by Tukey’s test; n = 4). (d) The cell death of the wild type, the isogenic sagB mutant or sagB-complementary mutant-infected cells were determined by LDH release at 18 h post-infection. ***P < 0.001 compared with GAS-infected group versus medium only group. **P < 0.01 compared with sagB mutant group versus GAS group. *P < 0.05 compared with sagB mutant group versus sagB-complementary group (One-way ANOVA test followed by Tukey’s test; n = 4).

Figure 6

Loss of contact between GAS and RAW264.7 cells reduced mitochondrial ROS production and cell death in the GAS-infected RAW264.7 cells. RAW264.7 cells were infected with the wild type GAS or its isogenic sagB mutant, which was added in culture medium or within the hanging insert with 0.4 μm membrane, at a MOI of 25 as described in Materials and Methods. (a) Mitochondrial ROS levels were measured by MitoSOX at 3 h post-infection, as described in Materials and Methods. Results are represented as mean ± SD. ***P < 0.001 compared with GAS-infected group versus medium only group or versus sagB mutant-infected group (Two-way ANOVA test followed by Bonferroni test; n = 4). (b) The cell death of the wild type or the isogenic sagB mutant-infected cells were determined by LDH release at 18 h post-infection. Results are represented as mean ± SD. ***P < 0.001 compared with GAS-infected group versus medium only group. **P < 0.01 compared with sagB mutant-infected group versus GAS-infected group (Two-way ANOVA test followed by Bonferroni test; n = 4).

Figure 7

The sagB mutation enhanced the proteasome-mediated GSK-3 β degradation in GAS-infected RAW264.7 cells. RAW264.7 cells were infected with the wild type GAS or isogenic sagB mutants at a MOI of 25 as described in Materials Methods. (a) The phosphorylation of GSK-3β at serine 9 (pGSK-3β) and the total GSK-3β protein in GAS-infected cells were detected at 3 h post-infection by Western blotting, as described in Materials and Methods. Representative immunoblots from 3 independent experiments were shown. The black arrow indicates low MW of GSK-3β. The quantitative ratios of full-length GSK-3β relative to GAPDH were determined by ImageJ and shown in (b). Displayed blots of GSK-3β and GAPDH are cropped from the same gel, and then immunoblotted with anti-GSK-3β antibody and anti-GAPDH antibody respectively. Full-length blots are presented in Supplementary Fig. S9a,b. The quantitative ratios of pGSK-3β relative to GAPDH were determined by ImageJ and shown in (c). Displayed blots of pGSK-3β are cropped from the independent gel. Full-length blots are presented in Supplementary Fig. S9c,d. (d) RAW264.7 cells were infected with the wild type, sagB mutant, or sagB-complementary mutant at a MOI of 25. In some groups, cells were pretreated with 10 μM of MG132 (proteasome inhibitor) or 10 μM of ALLN (calpain inhibitor I) before the sagB mutant infection, as described in Materials and Methods. Total GSK-3β protein in GAS-infected cells were detected at 3 h post-infection by Western blotting. GAPDH was used as an internal control. The black arrow indicates low MW of GSK-3β. The quantitative ratios of full-length GSK-3β relative to GAPDH were determined by ImageJ and shown in (e). Displayed blots are not cropped from the different gels. Full-length blots are presented in Supplementary Fig. S9e,f.