Connexin 43 hemichannels regulate mitochondrial ATP generation, mobilization, and mitochondrial homeostasis against oxidative stress

Abstract

Oxidative stress is a major risk factor that causes osteocyte cell death and bone loss. Prior studies primarily focus on the function of cell surface expressed Cx43 channels. Here, we reported a new role of mitochondrial Cx43 (mtCx43) and hemichannels (HCs) in modulating mitochondria homeostasis and function in bone osteocytes under oxidative stress. In murine long bone osteocyte-Y4 cells, the translocation of Cx43 to mitochondria was increased under H₂O₂-induced oxidative stress. H₂O₂ increased the mtCx43 level accompanied by elevated mtCx43 HC activity, determined by dye uptake assay. Cx43 knockdown (KD) by the CRISPR-Cas9 lentivirus system resulted in impairment of mitochondrial function, primarily manifested as decreased ATP production. Cx43 KD had reduced intracellular reactive oxidative species levels and mitochondrial membrane potential. Additionally, live-cell imaging results demonstrated that the proton flux was dependent on mtCx43 HCs because its activity was specifically inhibited by an antibody targeting Cx43 C-terminus. The co-localization and interaction of mtCx43 and ATP synthase subunit F (ATP5J2) were confirmed by Förster resonance energy transfer and a protein pull-down assay. Together, our study suggests that mtCx43 HCs regulate mitochondrial ATP generation by mediating K⁺, H⁺, and ATP transfer across the mitochondrial inner membrane and the interaction with mitochondrial ATP synthase, contributing to the maintenance of mitochondrial redox levels in response to oxidative stress.

Keywords: ATP synthase; cell biology; connexin 43; mouse; osteocytes; oxidative stress; proton.

Figure 1.. Connexin 43 (Cx43) translocated to mitochondrial and formed functional hemichannels (HCs) induced by oxidative stress in osteocytes.

(A) H₂O₂ treatment induced the translocation of Cx43 to mitochondria in the murine long bone osteocyte-Y4 (MLO-Y4) cells. MLO-Y4 cells were treated with 0.3 mM H₂O₂ for 2 hr. After being stained with MitoTracker, cells were fixed and immunostained with Cx43(E2) antibody. Scale bar: 20 μm. (B) Manders’ overlap coefficient co-localization analysis of MitoTracker and Cx43 based on fluorescence signals. Data collected from four independent experiments. Two-tailed t-test, **p<0.01. (C) Co-localization of succinate dehydrogenase (SDHA) and Cx43 in isolated mitochondria. Mitochondria were isolated from MLO-Y4 cells, fixed and subsequently immunostained with anti-Cx43 C-terminal (CT) and anti-SDHA antibodies. Scale bar: 1 μm. (D) Western blot and statistical data showed a significant increase of Cx43 in isolated mitochondria after 0.3 mM H₂O₂ treatment. One-way ANOVA analysis was conducted, **p<0.01. Data collected from N≥4 individual experiments. (E) Mitochondria dye uptake increased after H₂O₂ and were inhibited by Cx43 CT antibody and carbenoxolone (CBX). MLO-Y4 cells were treated with 0.3 mM H₂O₂ for 2 hr for mitochondria isolation. Mitochondrial dye uptake assay was performed with Lucifer yellow (LY)/TRITC-dextran in the absence and presence of 0.5 μg/mL Cx43CT, Cx43(E2) antibody or 1 μM CBX. N≥3, two-way ANOVA analysis was conducted, *p<0.05. TRITC-dexdran was used for calibration.

Figure 1—figure supplement 1.. The purity of mitochondria fractions isolated from murine long bone osteocyte-Y4 (MLO-Y4) cells.

(A) Mitochondria marker protein succinate dehydrogenase (SDHA) was accumulated in mitochondria fraction. Western blot showed enriched SDHA expression in mitochondria fraction compared to the whole cell lysate. There were contaminations from ER ( protein disulphide isomerase [PDI] as a marker protein) and Golgi (STX6 as a marker protein). (B) Fluorescence confocal images of isolated mitochondria stained with organelles-specific dyes. Mitochondria (MitoTracker) were indicated as red, and ER (ER-Tracker) and nuclei (Hochest) were indicated as blue. Scale bar: 20 μm. (C) The number of fluorescent dots per area was calculated. Most dots in mitochondria fractions were stained with MitoTracker. Data is expressed as mean ± SEM and compared by one-way ANOVA. N=4; ****p<0.0001.

Figure 1—figure supplement 2.. Connexin 43 (Cx43) translocated to mitochondria in MC3T3-E1 osteoblast cell line after H₂O₂ treatment.

(A) H₂O₂ treatment induced the translocation of Cx43 to mitochondria in MC3T3-E1 cells. MC3T3-E1 cells were treated with 0.3 mM H₂O₂ for 2 hr and incubated with MitoTracker Deepred before fixation. Cx43(E2) antibody was used. Scale bar: 20 μm. (B) Manders’ overlap coefficient co-localization analysis of MitoTracker and Cx43 based on fluorescence signals in MC3T3-E1 cells. Data were collected from four independent experiments. Two-tailed t-test, **p<0.01.

Figure 2.. Connexin 43 (Cx43) knockdown (KD) reduced hemichannel (HC) function in mitochondria.

(A) Cx43 was significantly knocked down in murine long bone osteocyte-Y4 (MLO-Y4) cells by the CRISPR-Cas9 system. Control and Cx43 KD (CxKD) MLO-Y4 cells were collected, and crude membrane extracts were analyzed using an anti-Cx43 antibody. Band intensity was quantified and compared by a two-tailed t-test (right panel). N=3; **p<0.01. (B) Cx43 was significantly knocked down in isolated mitochondria. Mitochondria were isolated from control (Ctrl) and CxKD MLO-Y4 cells and mitochondrial lysates immunoblotted with anti-Cx43 antibody. The KD efficiency in isolated mitochondria was quantified by the band intensity and compared by two-tailed t-test. N=3; *p<0.05 and **p<0.01. (C) Cx43 signal was absent in the mitochondria of Cx43KD MLO-Y4 cells. MLO-Y4 cells with or without Cx43 knockdown were immunolabeled with anti-Cx43 antibody and SDHA. Scale bar: 1 μm. (D) Dye uptake in isolated mitochondrial Cx43 KD MLO-Y4 cells was abolished after H₂O₂ treatment. Mitochondria were isolated from MLO-Y4 cells with 2 hr 0.3 mM H₂O₂ incubation, and Lucifer yellow (LY)/RD dye uptake assay was conducted and quantified. (E) Mitochondrial membrane potential (∆ψm) was significantly decreased in the Cx43 KD MLO-Y4 cells. MLO-Y4 cells with or without Cx43 KD were incubated with tetramethyl rhodamine and ethyl ester (TMRE) which is used as an indicator for ∆Ψm. TMRE fluorescence was shown in red. (F) TMRE fluorescence signals were detected and quantified (right panel). N=5 of independent experiments. Two-tailed t-test was used for statistical analysis. Each experiment was done in triplicates with three images from each repeat. ** p<0.01. Scale bar: 100 μm. (G) Cx43 KD significantly attenuated reactive oxygen species (ROS) production. MLO-Y4 cells with or without Cx43 KD were stained with MitoSOX, and MitoSOX fluorescence signals were quantified by NIH Image J and compared using two-way ANOVA, N=4, * p<0.05, and ** p<0.001. The cells were treated in the presence of complex inhibitors for 1 hr. AA: antimycin A and Rot: rotenone. (G and H) Mitochondrial ROS stained with fluorescent 2',7'-dichlorofluorescein (DCF) decreased in Cx43 impaired mitochondria. Mitochondria were isolated from either MLO-Y4 control or Cx43KD cells and treated with the inhibitors for 20 min. The fluorescence signals were measured using a microplate reader and quantified by NIH Image J. Two-way ANOVA analysis and multiple comparisons were conducted in each group, N=5. * p<0.05, ** p<0.001.

Figure 3.. Mitochondrial functions were changed in the connexin 43 (Cx43) knockdown (KD) Y4 cell line.

(A) The mitochondrial oxygen consumption was impaired in Cx43 KD Y4 cells. Seahorse XF Cell Mito Stress Test was used to determine mitochondrial function in cells. Oxygen consumption rate (OCR) was measured in each state in which ADP, oligomycin (OLIGO), p-trifluoromethoxy carbonyl cyanide phenylhydrazine (FCCP) and antimycin A (AA) were used to treat cells. OCR level calibrated based on 10,000 cells. (B) The OCR was significantly reduced after ADP and FCCP injections in Cx43KD Y4 cells. ***p<0.001. (C) The proton leak and ADP-induced ATP production, calculated by OCR, was decreased in Cx43KD experimental groups. N≥3 repeats in each experiment, Two-tailed t-test, **p<0.01, and ***p<0.001. (D) ATP was dramatically decreased in Cx43KD Y4 cells. ATP determination was performed in isolated whole-cell lysis, isolated mitochondria, and cytosolic components. N=3. Two-tailed t-test, **p<0.01, and ***p<0.001. (E) The mRNA expression level of related genes in Y4 cell line. Mitochondrial complex II component protein succinate dehydrogenase (Sdha) and F(1)F(0) ATP synthase associated ATP synthase membrane subunit f (Atp5j2) were decreased in Cx43KD group. Mitochondrial permeable pore component ADP/ATP translocase 2 (Slc25a5) expression level had no significant changes between control and Cx43KD Y4 cells. N=6, Two-tailed t-test, *p<0.05, and **p<0.01.

Figure 4.. pHluorin signal indicated an attenuated sensitivity to oxidative stress in connexin 43 (Cx43) knockdown (KD) Y4 cells.

(A) pHluorin fluorescent protein localized in the cytoplasmic area of transfected murine long bone osteocyte-Y4 (MLO-Y4) cells. BFP was used as calibrations. (B) Cytosolic pHluorin signal change due to H₂O₂ stimuli was attenuated in Cx43KD cells. Cytosolic pH change detected in Cx43 KD and control Y4 cells by pHluorin after 0.6 mM H₂O₂ stimuli at 25 s time point. Statistical analysis showed the difference in the 40 s time point. ***p<0.001. (C) ATP5J2 conjugated pHluorin fluorescent protein localized on mitochondria. Green: pHluorin signal (GFP channel); Blue: BFP. Scale bar: 10 μm. (D) Mitochondrial pHluorin signal change due to H₂O₂ stimuli was attenuated in Cx43KD cells. ATP5J2-pHluorin signal indicated the mitochondrial pH level from Cx43 KD and control Y4 cells after 0.6 mM stimuli at 25 s time point. Statistical analysis showed the difference in the 40 s time point in ATP5J2-pHluorin transfected Y4 cells. Data were collected from n≥7 cells in three independent experiments. *p<0.05. (E) Mitochondrial pHluorin signal oscillation was inhibited in Cx43KD Y4 cells. Representative ATP5J2-pHluorin signal oscillation in control and Cx43KD Y4 cells. Two-tailed t-test analysis showed significantly decreased oscillation amplitude in the CxKD group. ****p<0.0001.

Figure 4—figure supplement 1.. ATP5J2-pHluorin localized on mitochondria.

ATP5J2-pHluorin signal was merged with MitoTracker on mitochondria in Atp5j2-mTagBFP-pHluorin transfected murine long bone osteocyte-Y4 (MLO-Y4) cells. Scale bar: 20 µm.

Figure 5.. Connexin 43 C-terminal (Cx43CT) Ab treatment decreased mitochondrial respiratory capacity in murine long bone osteocyte-Y4 (MLO-Y4) cells.

(A) Oxygen consumption rate (OCR) was measured on permeabilized MLO-Y4 cells, using an XF96e analyzer. The sequential injection of mitochondrial inhibitors as indicated by arrows. OLIGO, oligomycin; AA, antimycin A. Rabbit IgG, Cx43CT antibody, and carbenoxolone (CBX) were pre-added in the assay medium. Basal respiration (state 2), after the injections of ADP (state 3), OLIGO (state 4o), p-trifluoromethoxy carbonyl cyanide phenylhydrazine (FCCP; state 3u). (B) The respiratory control ratio RCR (RCR = state 3/state 4o) and uncoupling control ratio UCR (UCR = state 3u/state 4o), which reflects the mitochondrial respiratory capacity, had no difference between groups. One-way ANOVA test. (C) OCR was detected significantly different in state 3 and state 3u. Values corresponding to the different respiratory states are represented as mean ± SEM (n=15–18 replicate of three independent experiments/group). Two-way ANOVA analysis was performed. *p<0.05, ***p<0.001, and ****p<0.001. (D) Both C-terminal (CT) antibody and CBX treated groups showed an inhibitory role on complex I activity. Complex I activity was measured in XFe96 Seahorse bioanalyzed by sequential administration of a combination of pyruvate (10 mM)/malate (0.5 mM), FCCP (0.5 μM), and antimycin A (4 μM). The activity was normalized to the control group. *p<0.05, **p<0.01. (E). CT antibody and CBX treated group showed a similar inhibitory role on complex II activity. Complex II enzymatic activity was analyzed similarly to Complex I. After basal OCR measurements cells were sequentially treated with a combination of rotenone (2 μM)/succinate (10 mM), FCCP (0.5 μM), and antimycin A (4 μM). All data were reported as the mean ± SEM. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.001.

Figure 5—figure supplement 1.. Mitochondrial function was affected under connexin 43 C-terminal (Cx43CT) Ab treatment in Cx43 knockdown (KD) murine long bone osteocyte-Y4 (MLO-Y4) cells.

(A) Oxygen consumption rate (OCR) was measured on permeabilized Cx43 KD MLO-Y4 cells, using an XF96e analyzer. The sequential injection of mitochondrial inhibitors is indicated by arrows. OLIGO, oligomycin; AA, antimycin A. rabbit IgG, Cx43CT antibody, and carbenoxolone (CBX) were pre-added in the assay medium. Basal respiration (state 2), after the injections of ADP (state 3), OLIGO (state 4o), p-trifluoromethoxy carbonyl cyanide phenylhydrazine (FCCP) (state 3u). (B) The respiratory control ratio RCR (RCR = state 3/state 4o) and uncoupling control ratio UCR (UCR = state 3u/state 4o), which reflects the mitochondrial respiratory capacity, had no difference between groups. One-way ANOVA test. (C) OCR was detected significantly different in state 3 and state 3u. Values corresponding to the different respiratory states are represented as mean ± SEM (n=15–18 replicates from three independent experiments/group). Two-way ANOVA analysis was performed. *p<0.05, ***p<0.001, and ****p<0.001. (D and E) CT antibody and CBX treated group showed the inhibitory effect on complexes I and II activity in Cx43KD MLO-Y4 cells. Activities of mitochondrial complexes I and II were normalized to the control MLO-Y4 cell line group. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.001.

Figure 5—figure supplement 2.. Connexin 43 (Cx43) C-terminal (CT) antibody localized to mitochondria with plasma membrane permeabilizer (PMP).

(A) Schematic illustration of the experiments. Murine long bone osteocyte-Y4 (MLO-Y4) cells were treated with 0.3 mM H₂O₂. The media was then changed to the mitochondria isotonic solution, and PMP was added to a final concentration of 2 nM. After the incubation with IgG and Cx43(CT) or Cx43(E2) antibody in the presence of PMP, immunofluorescence was conducted only with the secondary antibody. (B) Immunofluorescent staining with anti-IgG showed signal on the plasma membrane in Cx43(E2) group and on the mitochondria in Cx43(CT) group. MitoTracker was used for mitochondria staining. Scale bar: 20 µm.

Figure 6.. Mitochondrial connexin 43 (mtCx43) hemichannels (HCs) were permeable to protons indicated by Cx43-pHluorin signal.

(A) Typical images of Cx43-pHluorin in murine long bone osteocyte-Y4 (MLO-Y4) cells. Cx43-pHluorin fluorescent protein localization in transfected MLO-Y4 cells. Green: pHluorin signal (GFP channel); Blue: BFP; Red: MitoTracker Red. Scale bar: 10 μm. (B) Cx43-pHluorin signal was reduced in whole-cell area after 0.6 mM H₂O₂ treatment. BFP signal was used for calibrations. (C) Representative mitochondrial Cx43-pHluorin oscillation in MLO-Y4 cells after 0.6 mM H₂O₂ treatment. BFP signal was used for calibrations. (D and E) C-terminal (CT) antibody reduced Cx43-pHluorin signal oscillation. Mitochondrial pHluorin signal oscillation was recorded in Cx43-pHluorin transfected MLO-Y4 cells. The maximal amplitude of the oscillation was analyzed. **p<0.01.

Figure 6—figure supplement 1.. ATP5J2-pHluorin spontaneous oscillation on mitochondria.

(A) C-terminal (CT) antibody showed no effect on ATP5J2-pHluorin signal. Mitochondrial pHluorin signal oscillation in ATP5mf-pHluorin was detected as the same as Cx43-pHluorin in transfected murine long bone osteocyte-Y4 (MLO-Y4) cells. (B) The maximal amplitude of the oscillation was analyzed with a two-tailed t-test.

Figure 7.. Connexin 43 (Cx43) interacted with ATP5J2 on mitochondria.

(A) Förster resonance energy transfer (FRET) assay showed the interaction between Cx43 and ATP5J2 in murine long bone osteocyte-Y4 (MLO-Y4) cells. MLO-Y4 cells co-expressing Cx43-cyan fluorescent protein (CFP) and ATP5J2-EYFP were presented. YFP fluorescence (EYFP was excited, and its fluorescence recorded; the first column), CFP fluorescence (CFP was excited and its fluorescence recorded; the second column), and ‘FRET index’ (images processed with the ‘FRET and co-localization analyzer,’ fourth column). Scale bars: 20 μm. (B) Immunoprecipitation showed physically binding of Cx43 with endogenous ATP5J2. Cx43 band was detected after immunoprecipitation using ATP5J2-GFP as the bait. After immunoprecipitation, samples were collected and ran through sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Anti-Cx43 (1:300) was used to incubate the membrane. (C) GST pull-down assay proved the physical binding between ATP5J2 and Cx43-CT. Purified GST-CT was conjugated to glutathione agarose beads with GST as control.

Figure 7—figure supplement 1.. Overexpression of endogenous ATP5J2-GFP in murine long bone osteocyte-Y4 (MLO-Y4) cells.

(A) MLO-Y4 cells were transfected with GFP-N1 or ATP5j2-GFP plasmids. Images were taken in BX-Z100 microscopy with a GFP channel and bright field. Scale bar: 100 µm. (B) Schematic diagram of GST-pull down assay using transfected MLO-Y4 cells with ATP5J2-GFP and GFP-N1 plasmids as a control.

Figure 8.. Mitochondrial connexin 43 (mtCx43) response to oxidative stress in primary osteocytes.

(A) Osteocytes from Csf1^± have an increased reactive oxygen species (ROS) level by DCF staining. Data collected from three animals in each group, each point indicated the ROS level from each individual well. Scale bar: 50 µm. Two-tailed t-test was conducted, ****p<0.0001. (B) Osteocytes from Csf1^± mice have an increased co-localization of succinate dehydrogenase (SDHA, green) and Cx43 (red). Primary osteocytes were stained with SDHA and Cx43 antibodies. Scale bar: 20 µm. Coefficient co-localization analysis of SDHA and Cx43 based on fluorescence signals. Data collected from three mice in each group. Two-tailed t-test, *p<0.05. (C) Schematic diagram on mtCx43 function in osteocytes. Organelles (mitochondria and nucleus) are not drawn to scale. In oxidative stress, Cx43 is translocated onto mitochondria. With a special position close to ATP synthase, mtCx43 hemichannels (HCs) enabled the H⁺ influx between the mitochondrial inner membrane, accelerating ATP production. Thus, mtCx43 HCs play a crucial role in osteocytes under oxidation.