Role of Bax and Bak in mitochondrial morphogenesis


Bcl-2 family proteins are potent regulators of programmed cell death. Although their intracellular localization to mitochondria and the endoplasmic reticulum has focused research on these organelles, how they function remains unknown. Two members of the Bcl-2 family, Bax and Bak, change intracellular location early in the promotion of apoptosis to concentrate in focal clusters at sites of mitochondrial division. Here we report that in healthy cells Bax or Bak is required for normal fusion of mitochondria into elongated tubules. Bax seems to induce mitochondrial fusion by activating assembly of the large GTPase Mfn2 and changing its submitochondrial distribution and membrane mobility—properties that correlate with different GTP-bound states of Mfn2. Our results show that Bax and Bak regulate mitochondrial dynamics in healthy cells and indicate that Bcl-2 family members may also regulate apoptosis through organelle morphogenesis machineries.


Programmed cell death is essential in multicellular organisms for development, tissue turnover and host defence. In mammalian cells, mitochondria have a central role in apoptosis that is regulated by members of the Bcl-2 family1. Activation of either Bax or Bak, two homologous pro-apoptotic members of the Bcl-2 family, is associated with changes in their conformation that are postulated to induce permeabilization of the outer mitochondrial membrane (OMM)2. Bax and Bak double knockout (Bax/Bak DKO) cells show resistance to apoptosis induced by various treatments, demonstrating the crucial role of these proteins in the regulation of cell death3.

During apoptosis, mitochondria fragment into smaller units. Simultaneous with this mitochondrial fragmentation, Bax and Bak change their intracellular locations to concentrate into foci at sites of mitochondrial scission, where they colocalize with mitofusin 2 (Mfn2) and dynamin-related protein 1 (Drp1)4, two proteins that mediate mitochondrial fusion and fission, respectively. We therefore examined the possibility that Bax and Bak could regulate mitochondrial morphology in healthy cells.

Mitochondrial morphology depends on Bax or Bak

Mitochondrial shape and size were examined in six independent cell lines or explants from mice lacking expression of both Bax and Bak. We compared primary and transformed mouse embryonic fibroblasts (Bax/Bak DKO MEFs)3 and baby mouse kidney cell lines (Bax/Bak DKO BMKs)5 from Bax/Bak DKO mice with lineage-matched cell lines from wild-type mouse strains. Cells were labelled with the mitochondria-specific probe Mitotracker Red (data not shown) or by immunostaining the mitochondrial marker cytochrome c (Fig. 1a, b). In 68.1 ± 6.7% (mean ± s.d.) of transformed Bax/Bak DKO MEFs, more than 90% of mitochondria in a cell showed a short, fragmented morphology and only occasionally contained elongated tubules (Fig. 1b, c). By contrast, most single Bax or single Bak knockout cells (94.8 ± 1.5% and 80.0 ± 4.7% of Bax KO and Bak KO, respectively) had elongated tubular mitochondria resembling those in 80.9 ± 8.8% of wild-type MEFs from the same mouse mixed background 129/CD1 (ref. 3 and Fig. 1c), in addition to those in MEFs previously described6 from a 129/SvEv background (98.6 ± 0.7% had long branching mitochondria). Similar results were seen in primary MEF explants from two Bax/Bak DKO mice as compared with cells from one wild-type mouse and one Bax-/-Bak+/+ mouse (Supplementary Fig. S1). In addition, only 11 ± 4.2% and 8 ± 2.8% of the cells had tubular, elongated mitochondria in two independent Bax/Bak DKO BMK cell lines, in marked contrast to two independent wild-type BMK cell lines, which had elongated tubular mitochondria (83 ± 4.2% and 40 ± 2.8%; Supplementary Fig. S1).

Figure 1: Mitochondrial morphology defects in Bax/Bak DKO cells.

a, b, Wild-type (a) and Bax/Bak DKO (b) MEFs stained with antibodies to cytochrome c (ref. 19) were analysed by confocal microscopy. c, Mitochondrial morphology was scored in the cell types indicated. Data represent the mean ± s.d. of three independent experiments, each with >100 cells counted per cell type. df, IL-3-dependent bone-marrow-derived cells from DKO mice (d) and DKO mice permanently re-expressing Bak (e; clone Bak-8) stained with Mitotracker were analysed by confocal microscopy and compared with GFP-transfected IL-3-dependent bone-marrow-derived cells from Bax/Bak DKO mice (f). Cells with predominantly elongated mitochondria were scored; data represent the mean ± s.d. of three experiments, each with 50 cells per condition. g, Mitochondrial length was measured in the indicated cell types (see Supplementary Methods). Data represent the mean ± s.d. h, Mitochondrial network complexity in wild-type, Bax/Bak DKO, Bak KO and Mfn2 KO MEFs was measured by FRAP as described8. Differences in the mean fluorescence recovery between wild-type and Bax/Bak DKO MEFs or between wild-type and Mfn2 KO MEFs measured at 40?s were statistically significant at P < 0.001 (two-tailed t-test).

The average length of a mitochondrion was substantially shorter in MEFs from three different Bax/Bak DKO mice (3.3–4.4?µm), as compared with MEFs from three different control mice (8.4–11.1?µm); and in Bax/Bak DKO BMK cells (2.8 ± 1.7?µm), as compared with wild-type BMKs (8.0 ± 2.1?µm; Fig. 1g). The extent of mitochondrial fragmentation in the Bax/Bak DKO cell lines was comparable to that induced by the knockout of Mfn1 or Mfn2 (2.4 ± 0.6?µm in Mfn2 KO versus 11.5 ± 5.5?µm in 129/SvEv wild-type MEFs, Fig. 1g; ref. 6). Bone-marrow-derived cells from a Bax/Bak DKO mouse also showed markedly fragmented mitochondria (Fig. 1d, f), and permanent re-expression of Bak into this cell line7 induced considerable mitochondrial elongation in healthy cells (Fig. 1e, f). Moreover, knock down of Bax and Bak by RNA interference (RNAi) in HeLa cells led to an overall decrease in mitochondrial size and network complexity (Supplementary Fig. S2), further confirming the requirement of Bax and Bak for mitochondrial morphogenesis.

Mitochondrial network complexity in cells lacking Bax and Bak was also evaluated by fluorescence recovery after photobleaching (FRAP; Fig. 1h), a technique that, by assessing the depth of bleach and rate of redistribution of mitochondrial matrix-localized yellow fluorescence protein (mito-YFP) into regions of irreversible fluorophore photobleach, provides a semiquantitative measure of the volume of single units of the mitochondrial network8. Fast but minor refilling of bleached regions occurred in Bax/Bak DKO cells (to 40.15 ± 6.3% of initial fluorescence; mean ± s.d.), comparable to that in Mfn2 KO cells (to 44.9 ± 7.4% of initial fluorescence). By contrast, refilling of bleached areas in wild-type MEFs or single Bak KO MEFs was more prominent (to 59.41 ± 11.16% and 58.2 ± 6.2% of initial fluorescence, respectively), indicating that diffusion of the mito-YFP occurred from a greater distance and volume in wild-type MEFs than in Bax/Bak DKO or Mfn2 KO cells (Fig. 1h). Thus, the shorter mitochondria in Bax/Bak DKO cells are discontinuous, separate units.

Bax overexpression elongates mitochondria

The cytomegalovirus protein, viral mitochondria-associated inhibitor of apoptosis (vMIA), prevents apoptosis by sequestering and inactivating Bax9,10 and Bak (Supplementary Fig. S3). As reported in primary human fibroblasts11, vMIA expression in DU145 cells (Fig. 2b, compared with control cells in Fig. 2a) and HCT116 Bax-/- cells (data not shown) caused the mitochondria to fragment into shorter units. We reasoned that binding of Bax and Bak and inhibition of their mitochondrial elongation activity by vMIA may cause this mitochondrial fragmentation. Ectopic expression of Bax reversed the vMIA-induced mitochondrial fragmentation, as assessed visually and by FRAP in DU145 cells and in HCT116 cells (Fig. 2c, e, f and data not shown). Bax lacking the BH3 domain (Fig. 2d–f), an anti-apoptotic and a BH3-only member of the Bcl-2 family (Supplementary Fig. S4), did not reverse vMIA-induced mitochondrial fragmentation. These data indicate that specific activities of Bax and Bak, which might be linked to their apoptosis-related functions, are required for the formation of balanced mitochondrial networks.

Figure 2: vMIA-induced mitochondrial fragmentation is reversed by Bax overexpression.

ae, Mito-YFP labelled mitochondria (green) in DU145 cells (ae) and HCT116 Bax KO cells (e) expressing the indicated constructs were imaged after anti-Myc immunostaining of vMIA (red) and scored for elongation. Data represent the mean ± s.d. of three experiments, each with 150 cells scored per condition. f, DU145 cells were analysed for mitochondrial contiguity by FRAP. Data represent an average of >30 measurements obtained in at least two experiments.

Mitochondrial fission and fusion machinery

On inhibition of mitochondrial fission by loss of Drp1 activity, mitochondria become unusually interconnected owing to unbalanced fusion12,13. Inactivation of Drp1 with the dominant-negative inhibitor Drp1-K38A12, however, did not cause mitochondrial elongation in Bax/Bak DKO MEFs (Supplementary Fig. S5H) or in vMIA-expressing DU145 cells (Supplementary Figs S5F–H and S6D), HCT116 Bax-/- cells (data not shown) or Cos-7 cells (Supplementary Fig. S5A–E), indicating that the fusion step of mitochondrial morphogenesis requires Bax or Bak activity. Use of Drp1 RNAi14 confirmed that inhibition of mitochondrial fission caused less elongation of mitochondria in Bax/Bak DKO cells than in Mfn2 KO cells (Supplementary Fig. S7).

Further indicating a role of Bax/Bak in mitochondrial fusion, ectopic overexpression of Mfn2 overcame the fusion deficit in Bax/Bak DKO cells (Fig. 3), and this was dependent on the Mfn2 carboxy-terminal coiled-coil domain (Supplementary Fig. S11D). In addition, ectopic expression of both Bax and Mfn2 in Bax/Bak DKO cells induced greater mitochondrial interconnectivity than Mfn2 alone (Fig. 3c), suggesting that Bax functions in mitochondrial morphogenesis by regulating Mfn2.

Figure 3: Ectopic Mfn2 but not Mfn2-G12V reverses fragmentation owing to loss of Bax/Bak.

a, b, Images of Bax/Bak DKO cells transfected with YFP–OMP25 (a; green) or with Mfn2–YFP (b; yellow) and stained for cytochrome c to visualize mitochondria (red). c, FRAP analysis of Bax/Bak DKO cells expressing mito-YFP alone or mito-YFP with Mfn2 or with Mfn2 plus Bax. df, Images of Mfn2 KO (d, e) and Bax/Bak DKO (f) MEFs transfected with Mfn2–YFP (d; green) or Mfn2-G12V–YFP (e, f; green) and immunostained for cytochrome c (df; red, top; white, bottom). g, Effects of Mfn2–YFP and Mfn2-G12V–YFP on mitochondrial morphology in Mfn2 KO cells and Bax/Bak DKO cells assessed by blind counting. Data represent the mean ± s.d. of two experiments, each with triplicates of 50 cells counted per condition.

Mfn2 GTPase activity and Bax/Bak expression

The subcellular location, expression level and mobility of fission proteins Drp1 and Fis1 were not detectably different with and without Bax/Bak expression (Supplementary Fig. S8 and data not shown). Although the expression level of endogenous Mfn2 was also the same (Fig. 4f and Supplementary Fig. S8), the distribution of Mfn2 in the OMM in wild-type MEFs differed from that in Bax/Bak DKO MEFs. In wild-type MEFs, Mfn2 localized partially around the OMM, but primarily concentrated into foci at sites of mitochondrial fusion and fission4,15 (Fig. 4a and Supplementary Fig. S9), and the relative amounts of Mfn2 in the different locations could be quantified by analysis of confocal images (Fig. 4d and Table 1). In Bax/Bak DKO MEFs, however, Mfn2 coated the OMM more evenly and had diminished foci (Fig. 4b, d, and Supplementary Fig. S9). Re-expression of Bax in Bax/Bak DKO MEFs significantly (P = 0.01) reconstituted focal localization of Mfn2 (Fig. 4c, d, and Supplementary Fig. S9). The ratio of Mfn2 in foci relative to that circumscribing the mitochondria increased more than 12-fold in the presence of either endogenous Bax/Bak or ectopic Bax (Fig. 4d).

Figure 4: Mfn2 complex formation, subcellular distribution and membrane mobility are altered by Bax expression.

ac, Wild-type (a) and Bax/Bak DKO (b, c) MEFs transiently transfected with Mfn2–YFP alone, and Bax/Bak DKO MEFs coexpressing Mfn2–YFP and Bax (c) were stained for cytochrome c and analysed by confocal microscopy for Mfn2 distribution (green or white) and mitochondria (red). d, Wild-type MEFs, Bax/Bak DKO MEFs transfected with Mfn2–YFP, and Bax/Bak DKO MEFs cotransfected with Mfn2–YFP and Bax were analysed for YFP distribution along the mitochondria (see Supplementary Methods). e, OMM mobility of Mfn2–YFP, Mfn2(1–703)–YFP, Mfn2-K109T–YFP and YFP–Fis1 (inset) expressed in Bax/Bak DKO MEFs and in the same cells cotransfected with Bax was measured by FRAP. Each line represents the mean of ≥30 measurements. f, Mfn2 complexes in wild-type, Bax/Bak DKO and Mfn2 KO MEFs were analysed by BN–PAGE19. Isolated mitochondria were incubated at 35?°C for 20?min in the absence (lane 1) or presence of 100?µM GTP-γS (lane 2) or 1?mM GDP (lane 3). Arrow indicates the 440-kDa band. SDS gels of whole-cell lysates show similar expression of Mfn2 with and without Bax/Bak expression.

Table 1 Bax-altered distribution of Mfn2 and mutant Mfn2 along mitochondria

We used FRAP to quantify Mfn2 membrane mobility in the presence or absence of Bax expression. Mfn2–YFP was significantly less mobile than three other proteins localizing to the OMM—OMP25 (P < 0.005), Fis1 (P < 0.005), or YFP fused to the membrane anchor of Bcl-xL (P < 0.005) (Supplementary Fig. S10)—consistent with its more focal distribution. Deletion of the Mfn2 C-terminal coiled-coil domain (Mfn2(1–703)–YFP), which mediates Mfn2 complex formation16, led to increased membrane mobility (Fig. 4e) and shifted Mfn2 localization to circumscribe the mitochondria completely (Supplementary Fig. S11A–C). By contrast, the GTPase-inactive mutant Mfn2-K109T showed less mobility in the OMM than did wild-type Mfn2 (Fig. 4e), consistent with its increased focal mitochondrial distribution15 (Supplementary Fig. S11E, F).

Thus, various mutants of Mfn2 show a correlation between their degree of membrane mobility and location on the mitochondrial surface. The focal form seems to be considerably less mobile than the circumscribing population. Consistent with the shift of Mfn2 into foci along the OMM due to Bax/Bak expression, introducing Bax into Bax/Bak DKO MEFs caused a significant decrease in Mfn2 mobility (P < 0.05; Fig. 4e). However, Bax expression did not affect the mobility or distribution of Mfn2(1–703)–YFP (P > 0.1), Mfn2-K109T (P > 0.1) or a control protein in the OMM—the regulator of mitochondrial fission, Fis1 (Fig. 4e, Table 1 and Supplementary Fig. S11C). We therefore examined the mobility of endogenous Mfn2 using blue native gel electrophoresis (BN–PAGE) in the absence and presence of Bax/Bak (Fig. 4f). Mfn2 migrated in a series of different size complexes, varying in molecular weight between 240 and 669?kDa, in 129/CD1 Bax/Bak DKO cells, but assembled into a single 440-kDa complex with Bax/Bak expression in wild-type MEFs. Ectopic expression of Bax in Bax/Bak DKO cells also shifted Mfn2 into 440-kDa complexes (data not shown). However, Opa1, another dynamin family protein required for mitochondrial fusion, did not show any change in complex formation, as assessed by BN–PAGE, between wild-type and Bax/Bak DKO MEFs (Supplementary Fig. S12), showing the specificity of Bax modulation of Mfn2 complex formation. Neither incubation of GDP nor incubation of GTP-γS with the cell lysates affected the 440-kDa complex formation (Fig. 4f). Gel filtration of haemagglutinin-tagged Mfn2 also showed a change in Mfn2 complex size owing to the expression of ectopic Bax (Supplementary Fig. S13). These three changes in Mfn2—localization, membrane mobility and complex formation (Fig. 4)—that are regulated by Bax expression may reflect a mechanism of Bax/Bak-induced mitochondrial tubule maintenance. In contrast to the GTPase-inactive mutant Mfn2-K109T, a mutant of Mfn2 with reduced GTPase activity and increased GDP–GTP exchange to favour the GTP bound-form, Mfn2-G12V, has been shown to circumscribe the mitochondria with reduced focal distribution15. We confirmed that Mfn2-G12V primarily circumscribes the OMM and found that, unlike wild-type Mfn2, it is not shifted into foci by Bax (Table 1 and Supplementary Fig. S11G, H). Notably, Mfn2-G12V fully reconstituted mitochondrial fusion in Mfn2 KO MEFs but not in Bax/Bak DKO MEFs (Fig. 3d–g).

These data indicate that Bax or Bak is required for Mfn2-activity-dependent mitochondrial fusion. We therefore directly examined mitochondrial fusion in Bax/Bak DKO cells by using a quantitative assay based on the redistribution and fluorescence dilution of a mitochondrial matrix-targeted photoactivable green fluorescent protein (mito-PAGFP)17. Small regions of interest (ROIs) in wild-type and Bax/Bak DKO cells transfected with mito-PAGFP were briefly illuminated by a 413-nm laser, resulting in photoactivation of the mitochondrial matrix GFP fluorophore in the ROI. In control 129/CD1 MEFs, dilution and redistribution of mito-PAGFP fluorescence were detected 30 and 60?min after activation (Fig. 5a, c), as previously described in 129/SvEv MEFs17. Distinct fusion was observed in 42 of 45 cells analysed 30?min after activation and further fusion occurred in 34 of 38 cells analysed 60?min after activation, indicating efficient mitochondrial fusion. By contrast, no dilution or redistribution of mito-PAGFP fluorescence was detectable in 32 of 43 Bax/Bak DKO cells analysed (Fig. 5b, c). Quantifying the extent of fluorescence dilution of GFP after photoactivation in two wild-type MEF lines, Bak/Bak DKO MEFs and Mfn1 KO MEFs, showed that the rate of fusion decreased several fold in the Bax/Bak DKO cells relative to the rate in wild-type MEFs, comparable to the decrease seen in Mfn1 KO cells (Fig. 5c). Similarly, Cos-7 and HeLa cells expressing vMIA showed a complete absence of mitochondrial fusion activity (Supplementary Fig. S6A–C).

Figure 5: Mitochondrial fusion is inhibited in the absence of Bax and Bak.

ac, Wild-type 129/CD1 (a, c), Bax/Bak DKO (b, c), wild-type 129/SvEv (c) and Mfn1 KO (c) MEFs were analysed for mitochondrial fusion as described17. Images to the right of a and b show magnified areas (from red squares) located in close vicinity to the activated ROIs (yellow circles). c, Quantification of fusion in wild-type 129/CD1, Bax/Bak DKO, wild-type 129/SvEv and Mfn1 KO MEFs as described17. Data represent the mean ± s.d. of >30 single-cell measurements per cell type obtained on three separate occasions.

In contrast to the defect in mitochondrial morphology, we saw no defect in mitochondrial membrane potential or ATP concentrations in Bax/Bak DKO cells (Supplementary Fig. S14) suggesting a normal metabolic status of Bax/Bak DKO cells. Consistent with this, the amount of hexokinase binding to mitochondria, which seems to be linked to both the rate of glycolysis and adenine nucleotide exchange across the OMM, is unchanged in Bax/Bak DKO MEFs relative to wild-type MEFs18.


Our results show that Bax and Bak are required for normal morphogenesis of mitochondria. The mitochondria in cells lacking Bax and Bak are shorter, have less network continuity and lower fusion rates. Bax also regulates Mfn2 complex assembly, membrane mobility and distribution on the OMM, regulating formation of Mfn2 foci that occur at sites of mitochondrial fusion. In yeast two-hybrid analysis, the 392–602 region of Mfn2, which contains a coiled-coil domain, interacts with Bax (Y.-J. Lee-Wickner and R.Y., unpublished data), suggesting that Bax may directly activate Mfn2. In addition, endophilin B1, a protein normally involved in mitochondrial morphogenesis14 that binds to Bax20, may coordinate Bax/Bak activity with the mitochondrial morphogenesis machinery.

Our results also yield insights into the role of Bcl-2 family members during apoptosis. On induction of apoptosis, the interconnected mitochondrial network converts into a punctiform morphology21. Overexpression of Mfn2 inhibits apoptosis15,22, Mfn2 colocalizes with Bax and Bak in foci during apoptosis4,15, and apoptotic fragmentation of mitochondria occurs in the same time frame as Bax translocation to mitochondria and cytochrome c release across the OMM17, suggesting that mitochondrial structural remodelling participates in transduction of apoptosis. Bax and Bak change conformation during this step of apoptosis, perhaps reflecting inactivation of their mitochondrial elongation activity to allow unbalanced mitochondrial fission; alternatively, because Bax and Bak colocalize with Mfn2 during the fragmentation process, Bax and Bak in their apoptotic conformation may actively inhibit Mfn2 activity to cause Drp1-dependent fragmentation. Similarly, Ced-9, the Bcl-2 family member in Caenorhabditis elegans, is thought to convert between anti- and pro-apoptotic conformations and, in concert with Drp1, may regulate both mitochondrial morphology and apoptosis23. Notably, Ced-9 can induce mitochondrial fusion and co-immunoprecipitates with Mfn2 in mammalian cells24.


Measurements of mitochondrial connectivity

FRAP was done as described8 to measure mitochondrial connectivity. In brief, bleaching of mitochondrial matrix-targeted YFP fluorescence was applied at randomly chosen perinuclear regions where indicated (bleach), and fluorescence intensity was normalized to the intensity of the ROI from the first image in the series. Differences in the fluorescence recovery between different cell types were measured at 40?s.

Measurements of membrane motility

Circular ROIs, 2.5?µm in diameter, were imaged (α-Plan-FLUAR 100 × /1.45 oil objective; Carl Zeiss) before and after photobleach (20 iterations of 514-nm laser set to 100%) of 0.5-µm circular ROIs located in the centre of imaging area. We collected 150 images (1 scan every 33?ms) and digitalized the fluorescence intensity in imaged ROIs with LSM 510 software (Zeiss MicroImaging). Curves were corrected for both nonspecific photobleaching that occurred during imaging and background, and normalized to the first image in the series (taken as 100%). Each line represents the average of ≥30 measurements obtained on 2–3 separate occasions.


Mitochondrial complexes of Mfn2 in transformed wild-type, Bax/Bak DKO and Mfn2 KO MEFs were analysed by BN–PAGE. Mitochondria were isolated by differential centrifugation and incubated at 35?°C for 20?min in the absence or presence of 100?µM GTP-γS (Upstate) and 1?mM GDP (Upstate). After the incubation, samples were centrifuged at 100,000g for 30?min and the resulting pellets were resuspended in membrane lysis buffer (1% CHAPS, 10% glycerol and 0.5?M aminocaproic acid in 50?mM Bis/Tris; pH 7.0), incubated for 20?min on ice, and then centrifuged at 100,000g for 30?min. Supernatants were applied to BN–PAGE at 40?µg of protein per lane. Proteins were resolved on 4–12% gradient polyacrylamide Tris-Glycine gels (Invitrogen) at 4?°C and a current of 3?mA. Proteins were transferred into polyvinylidene fluoride membranes and subjected to immunoblotting with polyclonal antibodies to Mfn2 (a gift from P. Hajek and G. Attardi), monoclonal antibodies to Opa1 (BD Biosciences) and monoclonal antibodies to Hsp60 (Stressgen) as loading controls. The sizes of the complexes in BN–PAGE experiments were established on the basis of the mobility of marker proteins (Amersham): thyroglobulin (669?kDa), apoferritin (440?kDa) and catalase (232?kDa).


  1. 1

    Cory, S., Huang, D. C. & Adams, J. M. The Bcl-2 family: roles in cell survival and oncogenesis. Oncogene 22, 8590–8607 (2003)

    CAS  Article  Google Scholar 

  2. 2

    Martinou, J. C. & Green, D. R. Breaking the mitochondrial barrier. Nature Rev. Mol. Cell Biol. 2, 63–67 (2001)

    CAS  Article  Google Scholar 

  3. 3

    Wei, M. C. et al. Proapoptotic BAX and BAK: a requisite gateway to mitochondrial dysfunction and death. Science 292, 727–730 (2001)

    ADS  CAS  Article  Google Scholar 

  4. 4

    Karbowski, M. et al. Spatial and temporal association of Bax with mitochondrial fission sites, Drp1, and Mfn2 during apoptosis. J. Cell Biol. 159, 931–938 (2002)

    CAS  Article  Google Scholar 

  5. 5

    Degenhardt, K., Sundararajan, R., Lindsten, T., Thompson, C. & White, E. Bax and Bak independently promote cytochrome c release from mitochondria. J. Biol. Chem. 277, 14127–14134 (2002)

    CAS  Article  Google Scholar 

  6. 6

    Chen, H. et al. Mitofusins Mfn1 and Mfn2 coordinately regulate mitochondrial fusion and are essential for embryonic development. J. Cell Biol. 160, 189–200 (2003)

    CAS  Article  Google Scholar 

  7. 7

    Lum, J. J. et al. Growth factor regulation of autophagy and cell survival in the absence of apoptosis. Cell 120, 237–248 (2005)

    CAS  Article  Google Scholar 

  8. 8

    Szabadkai, G. et al. Drp-1-dependent division of the mitochondrial network blocks intraorganellar Ca2+ waves and protects against Ca2+-mediated apoptosis. Mol. Cell 16, 59–68 (2004)

    CAS  Article  Google Scholar 

  9. 9

    Arnoult, D. et al. Cytomegalovirus cell death suppressor vMIA blocks Bax- but not Bak-mediated apoptosis by binding and sequestering Bax at mitochondria. Proc. Natl Acad. Sci. USA 101, 7988–7993 (2004)

    ADS  CAS  Article  Google Scholar 

  10. 10

    Poncet, D. et al. An anti-apoptotic viral protein that recruits Bax to mitochondria. J. Biol. Chem. 279, 22605–22614 (2004)

    CAS  Article  Google Scholar 

  11. 11

    McCormick, A. L., Smith, V. L., Chow, D. & Mocarski, E. S. Disruption of mitochondrial networks by the human cytomegalovirus UL37 gene product viral mitochondrion-localized inhibitor of apoptosis. J. Virol. 77, 631–641 (2003)

    CAS  Article  Google Scholar 

  12. 12

    van der Bliek, A. M. A mitochondrial division apparatus takes shape. J. Cell Biol. 151, F1–F4 (2000)

    CAS  Article  Google Scholar 

  13. 13

    Shaw, J. M. & Nunnari, J. Mitochondrial dynamics and division in budding yeast. Trends Cell Biol. 12, 178–184 (2002)

    CAS  Article  Google Scholar 

  14. 14

    Karbowski, M., Jeong, S. Y. & Youle, R. J. Endophilin B1 is required for the maintenance of mitochondrial morphology. J. Cell Biol. 166, 1027–1039 (2004)

    CAS  Article  Google Scholar 

  15. 15

    Neuspiel, M., Zunino, R., Gangaraju, S., Rippstein, P. & McBride, H. M. Activated Mfn2 signals mitochondrial fusion, interferes with Bax activation and reduces susceptibility to radical induced depolarization. J. Biol. Chem. 280, 25060–25070 (2005)

    CAS  Article  Google Scholar 

  16. 16

    Koshiba, T. et al. Structural basis of mitochondrial tethering by mitofusin complexes. Science 305, 858–862 (2004)

    ADS  CAS  Article  Google Scholar 

  17. 17

    Karbowski, M. et al. Quantitation of mitochondrial dynamics by photolabeling of individual organelles shows that mitochondrial fusion is blocked during the Bax activation phase of apoptosis. J. Cell Biol. 164, 493–499 (2004)

    CAS  Article  Google Scholar 

  18. 18

    Majewski, N. et al. Hexokinase-mitochondria interaction mediated by Akt is required to inhibit apoptosis in the presence or absence of Bax and Bak. Mol. Cell 16, 819–830 (2004)

    CAS  Article  Google Scholar 

  19. 19

    Ishihara, N., Eura, Y. & Mihara, K. Mitofusion 1 and 2 play distinct roles in mitochondria fusion reactions via GTPase activity. J. Cell Sci. 117, 6535–6546 (2004)

    CAS  Article  Google Scholar 

  20. 20

    Takahashi, Y. et al. Loss of Bif-1 suppresses Bax/Bak conformational change and mitochondrial apoptosis. Mol. Cell. Biol. 25, 9369–9382 (2005)

    CAS  Article  Google Scholar 

  21. 21

    Youle, R. J. & Karbowski, M. Mitochondrial fission in apoptosis. Nature Rev. Mol. Cell Biol. 6, 657–663 (2005)

    CAS  Article  Google Scholar 

  22. 22

    Sugioka, R., Shimizu, S. & Tsujimoto, Y. Fzo1, a protein involved in mitochondrial fusion, inhibits apoptosis. J. Biol. Chem. 279, 52726–52734 (2004)

    CAS  Article  Google Scholar 

  23. 23

    Jagasia, R., Grote, P., Westermann, B. & Conradt, B. DRP-1-mediated mitochondrial fragmentation during EGL-1-induced cell death in C. elegans. Nature 433, 754–760 (2005)

    ADS  CAS  Article  Google Scholar 

  24. 24

    Delivani, P., Adrain, C., Taylor, R. C., Duriez, P. J. & Martin, S. J. Role for CED-9 and Egl-1 as regulators of mitochondrial fission and fusion dynamics. Mol. Cell 21, 761–773 (2006)

    CAS  Article  Google Scholar 

Download references


We thank N. Swarup for help in the early stages of this work; C. Thompson, J. Lum and T. Lindsten for Bax/Bak DKO bone marrow cells, Bak-transfected DKO cells and DKO primary MEFs; E. White, S. Korsmeyer and D. Chan for Bax/Bak DKO BMK cells, Bax/Bak DKO MEFs and Mfn KO MEFs, respectively; B. Vogelstein for HCT116 Bax-/- cells, J. Norris for DU145 cells, V. Goldmacher for vMIA HeLa cells, H. McBride for antibodies to Mfn2 and mutant Mfn2 constructs; P. Hajek and G. Attardi for antibodies to Mfn2; A. Antignani, A. Neutzner, B. Fell, S.-W. Ryu and S. Smith for technical assistance; and C. Smith for assistance with confocal microscopy and data analysis and for reading the manuscript. This work was supported by the Intramural Research Program of the NIH, National Institute of Neurological Disorders and Stroke. Author Contributions M.K., K.N. and R.Y. contributed to the conception, interpretation, execution and presentation of the experiments; S-Y.J. contributed to the RNAi experiments; M.C. participated in FRAP experiments and data analysis.

Author information



Corresponding authors

Correspondence to Mariusz Karbowski or Richard J. Youle.

Ethics declarations

Competing interests

Reprints and permissions information is available at The authors declare no competing financial interests.

Supplementary information

Supplementary Notes

This file contains Supplementary Methods and Supplementary Figure Legends (DOC 107 kb)

Supplementary Figure 1

Quantification of mitochondrial morphology with and without Bax/Bak expression in baby mouse kidney cells (BMK) and primary mouse embryonic fibroblasts (MEFs). (JPG 16 kb)

Supplementary Figure 2

Development and mitochondrial phenotype of Bax and Bak double RNAi cells (JPG 22 kb)

Supplementary Figure 3

Viral mitochondria-associated inhibitor of apoptosis (vMIA) interacts with both Bax and Bak and inhibits apoptosis induced by Bak-overexpression (JPG 17 kb)

Supplementary Figure 4

Certain Bcl-2 family members other than Bax do not reverse vMIA-induced mitochondrial fragmentation (JPG 14 kb)

Supplementary Figure 5

Blocking fission does not reverse fragmentation induced by vMIA expression or loss of Bax/Bak (JPG 63 kb)

Supplementary Figure 6

vMIA blocks mitochondrial fusion (JPG 26 kb)

Supplementary Figure 7

Drp1 RNAi effect on mitochondria in Bax/Bak DKO MEFs (JPG 12 kb)

Supplementary Figure 8

Expression and sub-cellular distribution of mitochondrial fusion and fission proteins in 129/CD1 WT and 129/CD1 Bax/Bak DKO MEFs (JPG 10 kb)

Supplementary Figure 9

Quantification of Mfn2-YFP sub-mitochondrial localization in WT and Bax/Bak DKO MEFs (JPG 10 kb)

Supplementary Figure 10

FRAP of mitochondrial outer membrane-associated proteins in Bax/Bak DKO MEFs (JPG 17 kb)

Supplementary Figure 11a-d

Sub-mitochondrial distribution of YFP chimeras of several Mfn2 mutants. (JPG 27 kb)

Supplementary Figure 11e-h

Sub-mitochondrial distribution of YFP chimeras of several Mfn2 mutants. (JPG 54 kb)

Supplementary Figure 12

Opa1 complex formation is not altered in WT and Bax/Bak DKO MEFs (JPG 9 kb)

Supplementary Figure 13

Gel Filtration analysis of Mfn2 complex formation (JPG 10 kb)

Supplementary Figure 14

Mitochondrial membrane potential (Δψm) and ATP levels in WT and Bax/Bak DKO MEFs (JPG 15 kb)

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Karbowski, M., Norris, K., Cleland, M. et al. Role of Bax and Bak in mitochondrial morphogenesis. Nature 443, 658–662 (2006).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing