Bax exists in a dynamic equilibrium between the cytosol and mitochondria to control apoptotic priming

Abstract

The proapoptotic Bcl-2 protein Bax is predominantly found in the cytosol of nonapoptotic cells and is commonly thought to translocate to mitochondria following an apoptotic stimulus. The current model for Bax activation is that BH3 proteins bind to cytosolic Bax, initiating mitochondrial targeting and outer-membrane permeabilization. Here, we challenge this and show that Bax is constitutively targeted to mitochondria but in nonapoptotic cells is constantly translocated back to the cytosol. Using live-cell spinning-disk confocal imaging with a combination of FLIP, FRAP, and photoactivatable GFP-Bax, we demonstrate that disrupting adhesion-dependent survival signals slows the rate of Bax's dissociation from mitochondria, leading to its accumulation on the outer mitochondrial membrane. The overall accumulation of mitochondrial Bax following loss of survival signaling sensitizes cells to proapoptotic BH3 proteins. Our findings show that Bax is normally in a dynamic equilibrium between cytosol and mitochondria, enabling fluctuations in survival signals to finely adjust apoptotic sensitivity.

Graphical abstract

Figure 1

Bax Constantly Dissociates from Mitochondria in Nonapoptotic Cells (A) FLIP analysis of GFP-Bax and GFP-BaxS184V stably expressed in MECs. The cytosolic fractions of GFP-Bax and GFP-BaxS184V were photobleached in the yellow ROI. The red line indicates the outline of the imaged cell. Dissociation of the mitochondrial GFP was followed within the green ROIs over 10 min. The signals obtained post bleaching were analyzed as explained in the Experimental Procedures and normalized to 100% fluorescence. For all FLIP data, the first image post bleaching is set at 0 s, n = number of cells analyzed per condition, and error bars represent SEM. Data were analyzed by ANOVA. ^∗ = p < 0.05; ^∗∗ = p < 0.01; ^∗∗∗ = p < 0.0001. Data indicate that GFP-Bax has a dissociation rate with t_1/2 ≈ 70 s, which is unaffected by ABT-737. GFP-BaxS184V has a significantly reduced dissociation rate compared with GFP-Bax. See Movie S1. (B) FLIP analysis carried out as in (A) on MECs transiently expressing GFP-Bax or GFP-BaxS184V. Error bars represent SEM. (C) FLIP analysis of Bax^−/−Bak^−/− DKO MEFs transiently expressing GFP-Bax or GFP-BaxS184V. See Movie S2. Error bars represent SEM. (D) FLIP analysis of GFP-Bax expressed in either Bax^−/−Bak^−/− or Bim^−/−PUMA^−/− DKO MEFs. In the lower panel, Bax^−/−Bak^−/− DKO MEFs expressed both GFP-Bax and mCherry-tBid. See Movie S3. Error bars represent SEM.

Figure 2

Bax Exists in a Dynamic Equilibrium between Mitochondria and the Cytosol (A) MECs transiently expressing paGFP-BaxS184V were photoactivated at 405 nm in the blue ROI. Association of paGFP-BaxS184V on mitochondria was monitored over 10 min within the green ROI. mRFP-H2B was coexpressed from the same plasmid as paGFPBaxS184V to allow identification of cells for photoactivation. See Movie S4. Error bars represent SEM. (B) MECs transiently expressing paGFP-Bax were photoactivated as in (A). Fifteen minutes post activation, the same ROI was photobleached at 488 nm. Cells were imaged for 10 min. See Movie S5. (C) paGFP-BaxS184V was photoactivated at 405 nm within the blue ROI, and its accumulation on mitochondria was imaged as in (A). FLIP was then performed by photobleaching within the yellow ROI, and redistribution to the bleached area followed. mRFP-H2B is shown in red.

Figure 3

Bax and Bcl-XL Can Stabilize Each Other on Mitochondria (A) MECs stably expressing GFP-Bax were transfected with expression vectors for either mRFP-Bcl-XL or mRFP-Mcl-1. Eighteen hours post transfection, FLIP analysis was carried out as in Figure 1A in the presence or absence of ABT-737. The three graphs shown are from the same experiment, and the data for GFP-Bax (green line) are the same in all three plots. The bar graph on the left indicates that mRFP-Bcl-XL and mRFP-Mcl-1 inhibited detachment-induced apoptosis. Error bars represent SEM. See Figure S2. (B) FRAP analysis of Bcl-XL and DsRed-Mito. MECs transiently expressing GFP-BclXL and DsRed-Mito were photobleached in the yellow ROI and imaged every 5 s. Fluorescence intensity was analyzed within the ROI and normalized to 100%. Error bars represent SEM. See Movie S6. (C) Bax^−/−Bak^−/− DKO MEFs transiently expressing mRFP-Bcl-XL, either alone or with GFP-Bax, were analyzed by FRAP as in (B). Error bars represent SEM. See Movie S7.

Figure 4

Detachment of MECs Followed by Reattachment Rapidly Alters the Subcellular Equilibrium of Bax (A) MECs transiently expressing GFP-Bax were left adherent, detached for 30 min, or detached for 30 min before reattaching to ECM for the indicated times. After fixation, cells were immunostained for anti-GFP and anti-mtHsp70. The number of cells displaying predominantly mitochondrial GFP-Bax was quantified. Error bars represent SEM. Data was analyzed by ANOVA; ^∗∗∗ = p < 0.0001. Representative images and line profiles of each image indicating coincidence of GFP and mtHsp70 are shown on the right. Scale bar represents 10 μm. See also Figure S3A. (B) Detachment of MECs leads to a rapid loss of FAK and Akt activity. Adherent cells and cells detached for various times were immunoblotted (IB) for anti-FAK pY397, pY577, pY925, and anti-total FAK (left panel) or for anti-Akt pS473, anti-Akt pT308, anti-total Akt, and anti-active caspase 3 (right panel). (C) MECs were transfected with pEGFP-Bax either alone or in combination with pCDNA6-myrFAK or the inactive variant myrFAKY397F. Eighteen hours post transfection, cells were detached for 15 min, cytospun on polysine slides, and immunostained for anti-GFP and anti-V5 (for detection of coexpressed myrFAK). The number of V5-positive cells with predominantly mitochondrial GFP-Bax was determined. Data represent the mean of three independent experiments. Error bars represent SEM. Data was analyzed by ANOVA; ^∗∗ = p < 0.01. ^∗∗∗; = p < 0.0001. The panel on the right shows representative images. (D) MECs were transfected with either pCDH-GFPBax or pCDH-myrAkt-GFPBax. Eighteen hours post transfection, cells were left adherent or detached for 1 hr. Cells were immunostained for anti-GFP and anti-mtHsp70. The number of cells with predominantly mitochondrial GFP-Bax was determined. The mean of three independent experiments is shown. Error bars represent SEM. Data were analyzed by ANOVA; ^∗∗∗ = p < 0.0001. The panel on the right shows representative images. See also Figure S3B.

Figure 5

Activation of FAK and Akt Leads to the Dissociation of Mitochondrial Bax in Detached Cells (A) MECs transiently expressing myrFAKER (upper panel) or myrAktER (lower panel) were treated with 4-OHT for the indicated times. Lysates were immunoblotted with the indicated antibodies. IP, immunoprecipitation. (B) Reattachment of MECs leads to the rapid activation of FAK and Akt. MECs were analyzed by immunoblotting for phosphorylation of FAK on tyrosine 397 and Akt on serine 473. (C) MECs transiently expressing GFP-Bax alone or in combination with myrAktER or myrAkt (expressed from the same plasmid) were detached for 30 min. Cells were left untreated for an additional 30 min or treated with 4-OHT for 30 min, then immunostained for anti-GFP and anti-mtHsp70. The proportion of cells with predominantly mitochondrial GFP-Bax was quantified. The mean of three independent experiments is shown. Error bars represent SEM. Data were analyzed by ANOVA; ^∗∗∗ = p < 0.0001. The panel on the right shows representative images. (D) Adherent MECs coexpressing GFP-Bax and either myrAktK179M or myrAktK179MER were untreated or treated with 4-OHT. The distribution of GFP-Bax was determined as in (C). The mean of three independent experiments is shown. Error bars represent SEM. Data were analyzed by ANOVA; ^∗∗ = p < 0.01. Representative images are shown on the right. (E) MECs expressing GFP-Bax and myrAktK179MER were left untreated or treated with 4-OHT, followed by 8 hr either with or without washing out the 4-OHT. The subcellular distribution of GFP-Bax was then quantified as in (C). Error bars represent SEM.

Figure 6

FAK and Akt Signaling Control Bax Mitochondrial Dissociation (A) Effect of the small-molecule inhibitors on the FAK-Akt signaling axis. Adherent MECs were treated with either ABT-737, FAK inhibitor 14, SH-6, or Wortmannin and immunoblotted for Akt pS473, total AKT, FAK pY397, total FAK, and β actin. (B) MECs expressing GFP-Bax or GFP-BaxS184V were treated with ABT-737, FAK inhibitor 14, SH-6, or Wortmannin. Cells were immunostained with anti-GFP and anti-mtHsp70, and the number of cells with predominantly mitochondrial Bax was quantified. Data represent the mean of three independent experiments. Error bars represent SEM. Data were analyzed by ANOVA; ^∗∗ = p < 0.01; ^∗∗∗ = p < 0.0001. (C) MECs stably expressing GFP-Bax were treated with DMSO or FAK inhibitor 14. FLIP was performed as in Figure 1, and the dissociation rate of mitochondrial Bax was quantified. See Movie S8. (D) MECs expressing paGFP-Bax were treated with FAK inhibitor 14, photoactivated (blue ROI), and imaged over 15 min. Untreated MECs expressing paGFP-BaxS184V are shown in the lower panel for comparison, before and after photoactivation. (E) MECs transiently expressing GFP-Bax alone or with myrAkt or myrAktK179M were left adherent or detached for 1 hr. The proportion of cells with mitochondrial Bax was quantified as before. Error bars represent SEM. (F) FLIP analysis on MECs expressing GFP-Bax alone or with either myrAkt or myrAktK179M. The t_1/2 for dissociation of GFP-Bax was determined. ABT-737 (5 μM) had no effect on the dissociation of GFP-Bax in MECs coexpressing myrAkt. Error bars represent SEM. See Movie S9.

Figure 7

Inhibiting Mitochondrial Bax Dissociation Sensitizes Cells to BH3 Mimetics (A) MECs expressing GFP-Bax alone or with myrAkt or myrAktK179M were detached from the ECM for 8 hr, and apoptosis was quantified. Data represent the mean of three independent experiments. Error bars represent SEM. Data were analyzed by ANOVA; ^∗∗ = p < 0.01. (B) Flow-cytometery analysis of GFP in MECs stably expressing GFP-Bax or GFP-BaxS184V. (C) MECs from (B) were detached from the ECM for 8 hr. Apoptosis was then quantified. Error bars represent SEM. Data were analyzed by ANOVA; ^∗∗ = p < 0.01. (D) MECs from (B) were treated with 5 μM ABT-737 as indicated. Apoptosis was quantified. Data represent the mean of three independent experiments. Error bars represent SEM. Data were analyzed by ANOVA; ^∗ = p < 0.05; ^∗∗ = p < 0.01. (E) MECs from (B) were treated with 125 μM etopisode as indicated. Apoptosis was quantified. Data represent the mean of three independent experiments. Error bars represent SEM. Data were analyzed by ANOVA; ^∗∗∗ = p < 0.0001. (F) Adherent MECs expressing GFP-BaxS184V were either left untreated or treated with 5 μM ABT-737. Cells were immunostained with anti-Bax monoclonal 6A7 and anti-GFP. Exposure and postcapture manipulation were identical for all images. (G) Adherent MECs stably expressing either GFP-Bax or GFP-BaxS184V were treated with FAK inhibitor either alone or in combination with ABT-737. Cells were immunostained as in (F).