Discovery of bactericides as an acute mitochondrial membrane damage inducer

Abstract

Mitochondria evolved from endosymbiotic bacteria to become essential organelles of eukaryotic cells. The unique lipid composition and structure of mitochondrial membranes are critical for the proper functioning of mitochondria. However, stress responses that help maintain the mitochondrial membrane integrity are not well understood. One reason for this lack of insight is the absence of efficient tools to specifically damage mitochondrial membranes. Here, through a compound screen, we found that two bis-biguanide compounds, chlorhexidine and alexidine, modified the activity of the inner mitochondrial membrane (IMM)-resident protease OMA1 by altering the integrity of the IMM. These compounds are well-known bactericides whose mechanism of action has centered on their damage-inducing activity on bacterial membranes. We found alexidine binds to the IMM likely through the electrostatic interaction driven by the membrane potential as well as an affinity for anionic phospholipids. Electron microscopic analysis revealed that alexidine severely perturbated the cristae structure. Notably, alexidine evoked a specific transcriptional/proteostasis signature that was not induced by other typical mitochondrial stressors, highlighting the unique property of alexidine as a novel mitochondrial membrane stressor. Our findings provide a chemical-biological tool that should enable the delineation of mitochondrial stress-signaling pathways required to maintain the mitochondrial membrane homeostasis.

FIGURE 1:

Identification of bactericides as a stabilizer of PD-related PINK1 (C125G) mutant. (A) PINK1 KO HeLa cells stably expressed PINK1 (C125G)-EYFP were transfected with control or OMA1 siRNA. After 72 h, cells were treated with 20 μM CCCP or 10 μM MG132 for 4 h. The lysate was analyzed by SDS–PAGE. Note that OMA1 is known to be degraded on CCCP through autocatalytic activation (see also Supplemental Figure S1B). (B) PINK1 KO HeLa cells stably expressed PINK1 (C125G)-EYFP were seeded on 384 well plate and treated with each compound (5 μM, FDA-approved compounds). After 18 h, cells were treated with 20 μM CCCP for 4 h. EYFP fluorescence intensity of each well was measured by the high content image analyzer. (C, D) The chemical structure of a hit compound, chlorhexidine (C), and a similar compound, alexidine (D). Guanidium groups that have delocalized positive charges are highlighted in blue. (E, F) PINK1 KO HeLa cells stably expressed PINK1 (C125G)-EYFP were treated with the indicated drugs for the indicate time period, and the lysate was analyzed by SDS–PAGE. (G, H) HeLa cells stably expressed PINK1 (C125G)-EYFP were transfected with control or OMA1 siRNA. After 72 h, cells were treated with 20 μM CCCP or 5 μM alexidine for 8 h and subjected to ICC. High-magnification images in the indicated regions (G) are shown in H. TOM20 was utilized as a mitochondrial marker. Scale bars; 25 μm (G) and 5 μm (H).

FIGURE 2:

Alexidine shows the substrate-dependent inhibition on OMA1-mediated proteolysis. (A, B) PARL KO HeLa cells were pretreated with alexidine for the indicated time period (A), or at the indicated concentration for 30 min (B), and after, treated with 20 μM CCCP for 2 h. Note that in PARL KO cells, the CCCP-dependent PGAM5 cleavage is mediated by OMA1. The lysate was analyzed by SDS–PAGE. (C) WT HeLa cells were pretreated with 5 μM alexidine for 1 h, and after, treated with 150 μM actinonin for 2 h. The lysate was analyzed by SDS–PAGE. (D) HeLa cells transiently expressed with DELE1-HA was pretreated with 5 μM alexidine for 1 h, and after, treated with 20 μM CCCP for 4 h. The lysate was analyzed by SDS–PAGE. (E) Alexidine showed the inhibitory effect on the OMA1-mediated proteolysis in a substrate-dependent manner. (F–H) The indicated HeLa cells were transfected with control or PHB2 siRNA. After 72 h, cells were harvested, and the lysate was analyzed by SDS–PAGE; 10 μM doxycycline was added for last 24 h to induce DELE1-HA in H. *Nonspecific bands. (I) PHB complex differentially regulates the OMA1-mediated proteolysis in a substrate-dependent manner.

FIGURE 3:

Alexidine has an affinity for the IMM. (A and B) WT HeLa cells or HeLa cells stably expressed Su9-mCherry (a matrix marker) were treated with 5 μM alexidine or 20 μM CCCP for the indicated time period. After the drug-treatment, cells were washed with PBS for twice and stained with NAO. NAO staining was analyzed by live-cell imaging. Scale bars; 25 μm (A) and 10 μm (B). (C) NAO fluorescence intensity (Ex 485 nm /Em 535 nm) in individual wells of a microtiter plate which were coated with or without the indicated phospholipid species was measured by microtiter plate reader. Alexidine or CCCP was added at the indicated concentration for 30 min before the NAO staining. Data are shown as mean ± SD (n = 3 or n = 6 per condition). **P < 0.01, ***P < 0.001, and ****P < 0.0001 (one-way ANOVA followed by Turkey’s multiple comparison). (D) WT HeLa cells were treated with 5 μM alexidine, 20 μM CCCP, or the combination of these drugs for the indicated time period. After the drug-treatment, cells were washed with PBS twice and stained with NAO. NAO staining was analyzed by live-cell imaging. Scale bars; 50 μm. (E) FACS analysis of NAO fluorescence intensity in D.

FIGURE 4:

Alexidine induces an acute perturbation of IMM integrity. (A) The EM images of the WT HeLa cells that were treated with DMSO, 20 μM CCCP, or 5 μM alexidine for 4 h. Scale bars; 800 nm (upper panels), 100 nm (lower panels). (B, C) OMA1 KO cells stably expressed OMA1 (E328Q)-EYFP were treated with 5 μM alexidine or 20 μM CCCP for 4 h and subjected to ICC. Cox IV was used as a typical IMM marker protein (C). *Nonspecific signal. Scale bars; 5 μm. (D) WT HeLa cells were treated with 5 μM alexidine or 20 μM CCCP for 2 h and subjected to ICC. TIM50 was used as a typical IMM marker protein. Scale bars; 5 μm. (E) Pearson’s R values between MIC60 and TIM50 fluorescent signals from five images for each condition in D were analyzed using Image J. Data are shown as mean ± SD (n = 5 per condition). ****P < 0.0001 (one-way ANOVA followed by Turkey’s multiple comparison).

FIGURE 5:

Alexidine evokes a unique transcriptional/proteostasis signature. (A, B) WT HeLa cells were treated with 5 μM alexidine or 20 μM CCCP for 8 h and subjected to TMT-based quantitative proteomics. Scatter plot for Log₂ FC of protein amount in CCCP treated/untreated (x-axis) and Log₂ FC of protein amount in alexidine treated/untreated (y-axis) (A). Values are from Supplemental Table SS2. The enrichment analysis of proteins whose amount was significantly changed (t test q value < 0.05) in the alexidine-treated cells compared with the mock-treated cells (B). The enriched Gene Ontology Cellular Component (GOCC) classes and each enrichment value was shown. (C) WT HeLa cells were treated with 5 μM alexidine or 20 μM CCCP for 4 h and subjected to RNA-seq analysis. Scatter plot for Log₂ FC of mRNA amount in CCCP treated/untreated (x-axis) and Log₂ FC of mRNA amount in alexidine treated/untreated (y-axis). Values are from Supplemental Table SS3. (D) Venn diagram of down-regulated proteins (fold change < 0.8, t test q value < 0.05) after the alexidine or CCCP treatment in A. (E) List of the alexidine-specific down-regulated mitochondrial proteins in A. Fold change of mRNA level of each protein was obtained from RNA-seq results in C. IMS; inner membrane space. (F) Venn diagram of up-regulated proteins (fold change > 1.5, t test q value < 0.05) after the alexidine or CCCP treatment in A. (G) List of the alexidine-specific up-regulated proteins in A. Fold change of mRNA level of each protein was obtained from RNA-seq results in C. (H) Validation of TMT-based proteomics results by SDS–PAGE. WT HeLa cells were treated with the indicated drugs for 8 h (5 μM alexidine, 20 μM CCCP, 10 μM rotenone, 150 μM actinonin, and 10 μM CDDO) and subjected to further analyses. (I) Graphical summary of this study. See text for detail.