The Bcl‐2 proteins Bax and Bak can permeabilize the outer mitochondrial membrane and commit cells to apoptosis. Pro‐survival Bcl‐2 proteins control Bax by constant retrotranslocation into the cytosol of healthy cells. The stabilization of cytosolic Bax raises the question whether the functionally redundant but largely mitochondrial Bak shares this level of regulation. Here we report that Bak is retrotranslocated from the mitochondria by pro‐survival Bcl‐2 proteins. Bak is present in the cytosol of human cells and tissues, but low shuttling rates cause predominant mitochondrial Bak localization. Interchanging the membrane anchors of Bax and Bak reverses their subcellular localization compared to the wild‐type proteins. Strikingly, the reduction of Bax shuttling to the level of Bak retrotranslocation results in full Bax toxicity even in absence of apoptosis induction. Thus, fast Bax retrotranslocation is required to protect cells from commitment to programmed death.
Pro‐apoptotic proteins Bax and Bak kill cells by permeabilizing the outer mitochondrial membrane. Mitochondrial localization and thus apoptosis induction by both proteins is controlled by their retrotranslocation dynamics governed by the hydrophobicity of the C‐terminal membrane anchor.
The pro‐apoptotic Bcl‐2 protein Bak is retrotranslocated from the mitochondria into the cytosol dependent on pro‐survival Bcl‐2 proteins.
Bax and Bak retrotranslocate at different rates by the same retrotranslocation process.
Rapid Bax shuttling protects cells from apoptosis in the presence or absence of apoptotic stimuli.
The hydrophobicity of the membrane anchor determines shuttling and localization of Bax and Bak.
In response to stress, cells can initiate mitochondrial apoptosis signaling in multicellular animals. The intrinsic cell suicide program converges at the activation of the Bcl‐2 proteins Bax and Bak that permeabilize the outer mitochondrial membrane (OMM). The release of cytochrome c (cyt c) and other intermembrane space proteins into the cytosol after OMM permeabilization (mitochondrial outer membrane permeabilization, MOMP) results in mitochondrial dysfunction and initiates the caspase cascade that efficiently dismantles the cell (Bratton & Cohen, 2001; Green & Kroemer, 2004). The activation of Bax or Bak commits the cell to apoptosis necessitating a tight control of pro‐apoptotic Bcl‐2 proteins (Lindsten et al, 2000). Thus, the Bcl‐2 protein family contains two groups of structurally similar proteins involved in arbitration to apoptosis: pro‐survival Bcl‐2 proteins harboring four Bcl‐2 homology domains (BH1‐4, e.g. Bcl‐2, Bcl‐xL or Mcl‐1) and pro‐apoptotic Bcl‐2 proteins with three BH domains (BH1‐3, e.g. Bax or Bak). Proteins of both groups are regulated by a diverse group of proteins sharing only the BH3 domain with Bcl‐2 proteins (BH3‐only proteins). Pro‐survival Bcl‐2 proteins inhibit Bax and Bak via direct interactions or by sequestering ‘activator’ BH3‐only proteins, thereby preventing their interaction with Bax and Bak (Letai et al, 2002; Kuwana et al, 2005; Willis et al, 2005, 2007; Kim et al, 2006; Llambi et al, 2011). Regulatory interactions between pro‐ and anti‐apoptotic Bcl‐2 proteins and Bax and Bak can only be observed in the presence of the OMM or liposomes and could result in membrane‐integral protein complexes (Roucou et al, 2002), suggesting also mitochondrial apoptosis signaling via membrane‐embedded proteins (Leber et al, 2007; Lovell et al, 2008; García‐Sáez et al, 2009). Upon apoptosis induction, Bax and Bak oligomerize and at least partially insert into the OMM (Eskes et al, 1998; Antonsson et al, 2000; Wei et al, 2001).
The C‐terminal Bax and Bak segments are sufficient for targeting protein fusions to the OMM, whereas deletion of the C‐terminal segments abolishes mitochondrial localization and pro‐apoptotic activity of Bax and Bak (Nechushtan et al, 1999; Schinzel et al, 2004; Setoguchi et al, 2006). C‐terminal membrane anchors (MAs) direct proteins post‐translationally to the target organelle and may insert into the membrane (Habib et al, 2003). Single amino acid substitutions in the C‐terminal helix can either localize Bax almost completely to the mitochondria or prevent OMM binding at all (Nechushtan et al, 1999). The potential binding site of Bax for pro‐apoptotic BH3 motifs interacts with its MA in cytosolic Bax inhibiting MA exposure and binding to BH3 motifs (Suzuki et al, 2000). The exposure of the C‐terminal Bax helix may involve prolyl isomerization in the loop preceding the MA or BH3‐only protein binding to a low affinity site in the N‐terminal part of Bax (Schinzel et al, 2004; Gavathiotis et al, 2008).
Bak could be controlled by mechanisms similar to Bax regulation, considering the dependence of activation and oligomerization of both proteins on major conformational changes, exposing the BH3 motifs (Wang et al, 1998; Dewson et al, 2008; Edlich et al, 2011; Moldoveanu et al, 2013). The conversion of inactive into active Bak seems to depend exclusively on transient BH3‐only protein interactions with the hydrophobic cleft of Bak (Dai et al, 2011). In healthy cells, Bax primarily resides in the cytosol (Wolter et al, 1997), contrasting with the predominant mitochondrial Bak localization. Although Bax is constantly translocating to the OMM, it is stabilized in the cytosol by interactions with pro‐survival Bcl‐2 activities on the OMM, establishing an equilibrium between cytosolic and mitochondrial Bax (Edlich et al, 2011; Schellenberg et al, 2013). Bax retrotranslocation from the mitochondria requires recognition of its exposed BH3 motif by the hydrophobic groove of Bcl‐xL and interaction between the C‐terminal Bcl‐xL helix and Bax (Edlich et al, 2011; Todt et al, 2013). Bax shuttling could also involve Bcl‐xL‐independent mechanisms (Schellenberg et al, 2013). When Bax retrotranslocation is compromised, Bax accumulates on the OMM, but requires further stimulation to become active dependent on the size of the mitochondrial Bax pool prior to apoptosis stimulation (Todt et al, 2013). Reversible mitochondrial Bax accumulation can be observed during anoikis (Valentijn et al, 2003).
The functional redundancy between Bax and Bak raises the question, why Bak is predominantly found on the mitochondria. Are Bax and Bak controlled by different mechanisms or does differential regulation of the same mechanism cause different localizations of these pro‐apoptotic Bcl‐2 proteins? Considering that resistance to apoptosis is one hallmark of cancer (Hanahan & Weinberg, 2011), understanding the regulation of Bax and Bak is imperative for targeting the mitochondrial apoptosis pathway in cancer therapy.
Bak is present in the cytosol
The importance of the subcellular localization for the regulation of Bax raises the question whether Bak is also regulated by shuttling between mitochondria and cytosol in spite of its predominant mitochondrial localization. Thus, the presence of Bak in the cytosol (C) and the mitochondria‐containing heavy membrane fraction (HM) from human tissues was analyzed with specific antibodies (Fig 1A and B). Different levels of Bak are present on the mitochondria of all tissue samples (Fig 1C). Strikingly, in heart, kidney and lung tissue, Bak is also present in the cytosolic fraction (Fig 1D; Supplementary Fig S1). Interestingly, human lung tissue not only shows cytosolic Bak but the HM fraction also contains membrane‐associated Bak in addition to membrane‐integral protein (Fig 1E). The presence of Bak in the cytosol of human tissues raises the question, whether Bak is shuttled into the cytosol by a mechanism similar to Bax retrotranslocation (Edlich et al, 2011; Todt et al, 2013).
Bcl‐xL retrotranslocates Bak
We tested the possibility of Bak shuttling by Fluorescence Loss in Photobleaching (FLIP) measurements of mitochondrial Bak with and without ectopically expressed Bcl‐xL. The experiments were performed with transient protein expression in HCT116 Bax/Bak DKO cells to avoid high protein levels resulting from stable protein expression (Supplementary Fig S2A–D). FLIP experiments target cytosolic fluorescence by repeated cycles of bleaching within a defined region, while changes in mitochondrial fluorescence are monitored by confocal imaging between bleaching events (Ishikawa‐Ankerhold et al, 2012).
GFP‐Bak localizes to the mitochondria but cytosolic Bak is apparent when Bcl‐xL is overexpressed (Fig 2A). After the cytosolic fluorescence was bleached within the initial cycles, mitochondrial GFP‐Bak fluorescence diminished during the FLIP measurement indicating Bak shuttling (Fig 2B and C). Retrotranslocation of mitochondrial GFP‐Bak is increased by Bcl‐xL, accelerating the loss of mitochondrial fluorescence during the measurements (Fig 2B and C, arrows). Therefore, Bcl‐xL retrotranslocates mitochondrial Bak into the cytosol. Low Bak shuttling rates could explain the differential localization of Bax and Bak in cells (Table 1).
The same retrotranslocation process shuttles Bax and Bak
Bcl‐xL increases Bak shuttling. Therefore, the effect of wild‐type Bcl‐xL or Bcl‐xL G138A on Bak localization was analyzed in HCT116 Bax/Bak DKO cells. The G138A substitution prevents Bax BH3 binding to the hydrophobic groove of Bcl‐xL and Bax retrotranslocation (Sedlak et al, 1995; Desagher et al, 1999; Edlich et al, 2011). While Bak is robustly found in the HM fraction, an additional Bak pool is also present in the cytosol (Fig 3A and B; Supplementary Fig S3A), corroborating the Bak localization in human tissues (Fig 1). Elevated levels of wild‐type Bcl‐xL, but not Bcl‐xL G138A, increase the cytosolic Bak pool (Fig 3A and C; Supplementary Fig S3A and B). Increased levels of Mcl‐1 accelerate Bak retrotranslocation similar to Bax shuttling (Fig 3C; Edlich et al, 2011). On the other hand, the BH3 mimetic ABT‐737 decreases Bak retrotranslocation in the presence of ectopically expressed Bcl‐xL (Supplementary Fig S3C). In contrast to previously observed Bax shuttling (Edlich et al, 2011), ectopic Bcl‐2 expression does not accelerate Bak retrotranslocation.
The D83R substitution prevents Bak BH3 interactions with pro‐survival Bcl‐2 proteins (Kvansakul et al, 2007). Consistent with the effects of ABT‐737 and Bcl‐xL G138A, the Bak D83R substitution inhibits pro‐survival Bcl‐2 protein‐mediated Bak shuttling (Fig 3C). Therefore, interactions between the BH3 motif and hydrophobic groove of Bcl‐xL are required for Bak/Bax retrotranslocation. In parallel to ectopically expressed Bak, endogenous Bak is significantly increased in the cytosol of HeLa cells when Bcl‐xL levels are elevated (Fig 3D; Supplementary Fig S3D and E). Cytosolic Bak results from constant shuttling and is not diminished when protein synthesis is blocked (Supplementary Fig S3F). Thus, different Bak shuttling rates can account for the differential localization of Bak in human tissues (Fig 1).
Bax and Bak shuttling is determined by the membrane anchor
Single amino acid substitutions in the C‐terminal MA of Bax can shift the protein to the mitochondria and decelerate its retrotranslocation (Nechushtan et al, 1999; Edlich et al, 2011; Schellenberg et al, 2013). Therefore, the impact of MA substitutions on the subcellular localization and retrotranslocation of Bax and Bak was analyzed (Fig 4A). Both chimeras (BaxTBak and BakTBax) were expressed to levels comparable to the wild‐type proteins in the presence or the absence of ectopically expressed Bcl‐xL (Supplementary Fig S4A and B). While Bax resides in most cells primarily in the cytosol, BaxTBak is largely localized to the mitochondria, similar to the localization of wild‐type Bak (Fig 4B and C). Interestingly, HCT116 Bax/Bak DKO cells expressing BaxTBak often show a punctate fluorescence pattern (Fig 4C), as is characteristic for active Bax (Karbowski et al, 2002). BakTBax, on the other hand, exhibits a large pool of cells with significant amounts of cytosolic protein, corroborating the central role of the Bak MA in protein localization (Ferrer et al, 2012). However, the different localization observed for BakTBax chimeras (Fig 4C and Ferrer et al, 2012) could be caused by differences in MA composition, protein expression and the resulting cell stress. Despite increased mitochondrial BakTBax levels compared to wild‐type Bax, cell fractionations substantiated greater similarity between BakTBax and wild‐type Bax localization compared to Bak (Fig 4B–D).
In parallel to localization, the shuttling of BaxTBak and BakTBax is altered compared to their wild‐type counterparts. Bax retrotranslocation is slowed down to the level of Bak shuttling by the Bak MA, while Bak shuttling is significantly accelerated by the Bax MA (Supplementary Fig S4C and D; Table 1). These measurements show that the C‐terminal membrane anchor determines retrotranslocation rates, and thus, localization of Bax and Bak.
Increased retrotranslocation is required to protect cells from Bax activation
The size of the mitochondrial Bax pool determines cellular commitment to apoptosis (Todt et al, 2013). The pronounced influence of shuttling rates on the subcellular localization of Bax and Bak suggests that both proteins are regulated by the rate of mitochondrial‐cytosolic shuttling. Thus, the apoptotic activities of Bax, BaxTBak, Bak, or BakTBax ectopically expressed in HCT116 Bax/Bak DKO cells were analyzed by measuring caspase‐3/7 activity and the cleavage of the caspase substrate poly (ADP‐ribose) polymerase (PARP). In the absence of staurosporine (STS), Bax, Bak and BakTBax expression results in low caspase activity that can be inhibited by Bcl‐xL overexpression (Fig 5A; Supplementary Fig S5A). Strikingly, BaxTBak gains full activity in the absence of apoptotic stimuli that is not completely inhibited by Bcl‐xL overexpression, contrasting the regulation of both wild‐type proteins.
In response to STS‐induced apoptosis, BaxTBak activity in the presence of Bcl‐xL overexpression is even higher than Bax, Bak, or BakTBax activities in the absence of overexpressed Bcl‐xL (Fig 5B; Supplementary Fig S5B). Bcl‐xL fails to prevent the transition towards the active conformation of BaxTBak and cytochrome c release despite robust binding to the Bax variant (Fig 5C and D; Supplementary Fig S5C). Both wild‐type proteins and BakTBax can be inhibited to basal apoptotic activity by Bcl‐xL overexpression. The Bak activity is reduced when containing the Bax MA. These results establish the link between increased retrotranslocation rates and reduced pro‐apoptotic activity of Bax and Bak but show also the requirement for accelerated Bax shuttling.
Ectopic Bcl‐xL expression reduces the pool of Annexin V‐positive cells expressing Bax, Bak or BakTBax but not that of cells expressing BaxTBak (Fig 5E). Colony formation after apoptosis stimulation in the presence of Bax, BaxTBak, Bak or BakTBax reflects the results of caspase‐3/7 activity measurements and PARP cleavage (Fig 5F and G). In the presence of BaxTBak the predominant fate is cell death even when Bcl‐xL was overexpressed. Bcl‐xL also fails to protect cells from BaxTBak activity in the absence of apoptotic stimuli (Fig 5H). Thus, slow Bax retrotranslocation commits cells to apoptosis.
Bax and Bak are shuttled depending on the MA hydrophobicity
Apoptosis activity measurements revealed a striking difference between BaxTBak and Bak despite similar localization and shuttling rates. Bax is regulated by conformational changes (Hsu & Youle, 1997; Edlich et al, 2011). Therefore, the presence of the Bak MA might induce conformational changes and pre‐activate Bax. We tested whether BaxTBak activity was likely based on pre‐activation by conformational changes or increased MA hydrophobicity by comparing BaxTBak and its V197/198S variant (BaxTBakSS, Supplementary Fig S6A). Similar hydrophobicity of the MA of BaxTBakSS and Bax S184V suggests similar shuttling rates of both Bax variants, if MA hydrophobicity determines Bax/Bak retrotranslocation. Bax S184V is shifted to the mitochondria compared to wild‐type Bax due to the exposure and the hydrophobicity of the MA (Nechushtan et al, 1999; Suzuki et al, 2000). In fact, BaxTBakSS shuttles at similar rates as Bax S184V and significantly faster than BaxTBak, suggesting that Bax/Bak shuttling depends on the hydrophobicity but not the sequence of the MA (Fig 6A, Table 1). In contrast to BaxTBak, BaxTBakSS is not activated in the absence of apoptosis stimuli, suggesting BaxTBak activation based on MA hydrophobicity rather than pre‐activation (Fig 6B–D).
The importance of the MA hydrophobicity for Bax/Bak shuttling demands an assessment of Bax and Bak with the same MA. To this end we made use of the Bax MA to avoid Bax activation at low shuttling rates. In addition, we introduced the S184V substitution to prevent MA binding to the hydrophobic Bax pocket, because substantial differences in the amino acid composition of Bak render similar interactions unlikely (Supplementary Fig S6A). The comparison of Bax, BakTBax and both variants containing the S184V substitution: (I) reveals the MA‐independent influence of Bax and Bak on shuttling and commitment to apoptosis (Fig 6E), (II) indicates the effect of increased MA hydrophobicity, and (III) can indicate the contribution of MA binding to the hydrophobic pocket in Bax inhibition. The S184V substitution decreases the rate of Bax and BakTBax retrotranslocation (Supplementary Fig S6B; Table 1). The apoptotic activity of Bax S184V and BakTBax S184V is increased after apoptosis induction compared to the variants without S184V substitution (Fig 6F and G). Although both variants share only the C‐terminal helix, Bax S184V and BakTBax S184V induce similar apoptosis signaling. Thus, the commitment of cells to apoptosis is determined by Bax/Bak shuttling rates. The retrotranslocation rates of both Bcl‐2 proteins exhibit a linear dependence on the hydrophobicity, but not the sequence, of the MA (Fig 6H). Thus, the shuttling rates, localization and activity of Bax and Bak are determined by the hydrophobicity of the MA.
Membrane‐integral Bcl‐2 proteins are shuttled by retrotranslocation
Bak retrotranslocation, like Bax shuttling, is accelerated by the pro‐survival Bcl‐2 proteins, depending on the recognition of Bak BH3 by the pro‐survival Bcl‐2 protein. Bax shuttling involves the co‐retrotranslocation of Bcl‐xL. Therefore we tested, whether the shuttling of Bax and Bak can shift the localization of Bcl‐xL accordingly. Indeed, wild‐type Bcl‐xL shifts in the presence of Bax or Bak from the membrane‐integral to the OMM‐associated and cytosolic forms (Fig 7A; Supplementary Fig S7A). This Bcl‐xL migration from the OMM‐integral form is not observed for the Bax/Bak shuttling‐incompetent Bcl‐xLTBax (Fig 7B). This Bcl‐xL variant containing the Bax MA does not co‐retrotranslocate Bax or Bak, but interacts with Bax in contrast to Bcl‐xL G138A (Fig 7B and C; Supplementary Fig S7B). Therefore, the Bcl‐xL localization is retrotranslocation‐dependent. Although only a small fraction of wild‐type Bax is present in the OMM‐integrated protein pool of proliferating cells, larger pools of Bax variants (e.g. Bax S184V) are membrane integral and shuttle completely into the cytosol (Fig 7D; Supplementary Fig S7C and D; Edlich et al, 2011). Noteworthy, OMM‐integral Bax is not in the active conformation (Supplementary Fig S7E).
These observations raise the question whether an increased mitochondrial Bax pool initiates apoptosis, although mitochondrial Bax accumulation is insufficient to commit cells to apoptosis (Valentijn et al, 2003; Todt et al, 2013). Perhaps extremely slow shuttling could raise levels above a required threshold. Thus, wild‐type Bax expression was adapted to match the level of BaxTBak on the mitochondria (Fig 7E). Despite similar levels of wild‐type Bax and BaxTBak on the OMM, in the absence of apoptotic stimuli, the release of the IMS protein Smac into the cytosol as well as PARP cleavage were only observed in cells expressing BaxTBak (Fig 7E and F). These results confirm that even high mitochondrial Bax levels do not per se commit cells to apoptosis. Bax activation seems to be determined by low rates of Bax shuttling, suggesting the importance of the dwell time of Bax molecules on the OMM. The co‐retrotranslocation of Bcl‐xL and Bax or Bak involves the shuttling of the OMM‐integral forms of Bax, Bak and Bcl‐xL into the cytosol (Fig 7G). Although Bax and Bak are shuttled by the same process from the OMM, cells must accelerate Bax retrotranslocation to prevent commitment to apoptosis at low Bax shuttling rates in the absence of apoptosis signaling.
Bax and Bak share a common regulation of their localization and activity despite differences in their distribution between cytosol and mitochondria. Both pro‐apoptotic Bcl‐2 proteins are retrotrans‐located by pro‐survival Bcl‐2 proteins from the mitochondria into the cytosol of healthy cells, dependent on the interaction with their BH3 motif. While Bax is retrotranslocated by Bcl‐xL, Bcl‐2 and Mcl‐1, Bcl‐2 fails to accelerate Bak shuttling. This difference between Bax and Bak emphasizes the central role of retrotranslocation in Bax/Bak regulation, as Bcl‐2 seems to be not involved in Bak regulation (Oltersdorf et al, 2005). The hydrophobicity of the C‐terminal MAs of Bax or Bak determines different localization pattern and differential shuttling of both Bcl‐2 proteins by the same retrotranslocation process (Figs 6H and 7G). Bax and Bak share similar apoptotic activity and subcellular localization at high shuttling rates, when the same MA is exposed, emphasizing the major role of the C‐terminal MA in the regulation of Bax and Bak. Interestingly, retrotranslocation shuttles OMM‐integral forms of Bax, Bak and Bcl‐xL that either are in equilibrium with OMM‐associated protein or are directly shuttled into the cytosol by the retrotranslocation machinery. However, active Bax is not retrotranslocated, as Bax activation blocks shuttling into the cytosol (Edlich et al, 2011).
The analysis of Bax/Bak chimeras and variants revealed a potential requirement for increased Bax retrotranslocation in the absence of apoptosis signaling. Strikingly, the reduction of Bax retrotranslocation to the level of Bak shuttling initiates full Bax toxicity in the absence of apoptotic stimuli (Figs 5 and 7G). If the Bax shuttling rate is reduced sufficiently, mitochondrial Bax commits the cell to apoptosis in the absence of an apoptotic stress. Bcl‐xL overexpression does not prevent Bax activation at low shuttling rates. Therefore, the survival of the cell requires fast Bax retrotranslocation from the mitochondria. Bak, however, commits the cell to apoptosis only in the presence of apoptotic stimuli despite its predominant mitochondrial localization. Mitochondrial Bax activation adds to the differential regulation of both redundant proteins (Sarosiek et al, 2013). Bax activation probably only occurs with wild‐type proteins when Bax shuttling is decreased, for instance, by BH3‐only protein signaling (Edlich et al, 2011). However, the underlying mechanism of Bax activation at low shuttling rates remains to be solved. Mitochondrial Bax activation might explain cellular ‘priming’ to death by BH3‐only proteins regardless of the presence of Bid, Bim and Puma (Ni Chonghaile et al, 2011; Vo et al, 2012). Depending on individual retrotranslocation rates, Bax activation could differ among different mitochondria, resulting in different organelle fates under stress conditions (Tait et al, 2010). On the other hand, mitochondrial Bax accumulation does not per se lead to Bax activation (Todt et al, 2013). Accordingly, similar levels of wild‐type Bax and BaxTBak on the OMM result in BaxTBak but not Bax activity (Fig 7E). Under these conditions, both Bax variants most likely differ only in their translocation and retrotranslocation rates, and thus the residence time individual molecules spend on the mitochondria. Therefore, healthy cells inhibit Bax by accelerated shuttling into the cytosol and minimizing the time Bax molecules spend on the OMM in order to prevent Bax activation in the absence of apoptosis signaling.
Materials and Methods
Bax and Bak constructs were cloned in pEGFP‐C1 or pEYFP‐C1 expression vectors. The chimeras BaxTBak and BakTBax were obtained by overlap‐extension‐PCR using the following primers: Bak–BaxTail rev (5′‐CAAAGATGGTCACGGTGGGACCATTGCCC‐3′); Bak–BaxTail for (5′‐GGGCAATGGTCCCACCGTGACCATCTTTG‐3′); XhoI–Bak for (5′‐GCTACTCGAGCTATGGCTTCGGGGCAAG‐3′); Bax–EcoRI rev (5′‐GCGCGAATTCTCAGCCCATCTTCTTCCAGATG‐3′); Bax–BakTail rev (5′‐GCACGTTCAGGATCTGCCACGTGGGC‐3′); Bax–BakTail for (5′‐GCCCACGTGGCAGATCCTGAACGTGC‐3′); XhoI–Bax for (5′‐ATATCTCGAGCTATGGACGGGTCCGGGG‐3′) and Bak–EcoRI rev (5′‐GCGGGAATTCTCATGATTTGAAGAATCTTCGTACCACAAAC‐3′). BaxTBak contains Bax1–171 and Bak188–211 and BakTBax consists of Bak1–187 and Bax172–192.
Cell culture and transfection
HCT116 cells and HCT116 Bax/Bak DKO cells were cultured in McCoy's 5A medium supplemented with 10% heat‐inactivated fetal bovine serum and 10 mM Hepes in 5% CO2 at 37°C. Cells were transfected with TurboFect (Fermentas) or Lipofectamine LTX (Invitrogen), typically with 100 ng of the wild‐type or mutant constructs of Bax and Bak in pcDNA3.1 or pEGFP vector according to the manufacturer's instructions. In every comparative experiment the same amounts of all tested constructs were transfected in parallel using the same protocol resulting in similar transfection rates of approx. 90% (Supplementary Fig S2). Co‐transfections with pcDNA3.1 Bcl‐xL were performed with a ratio of 1:6 between pro‐apoptotic and pro‐survival Bcl‐2 protein. Cells were incubated for 6–8 h for confocal imaging.
LZRS‐DD‐Bcl‐xL was generated by inserting human Bcl‐xL into LZRS zeocin at EcoRI/XhoI sites and FKBP L106P in frame upstream via the EcoRI site using PCR/restriction digest based cloning (Banaszynski et al, 2006). HeLa cells expressing degradation‐prone DD‐Bcl‐xL were generated by retroviral transduction as follows. Amphotropic Phoenix cells (0.5 × 106 in a 10‐cm dish) were transfected with LZRS DD Bcl‐xL using Lipofectamine 2000 (Invitrogen). Two days later virus‐containing supernatant was harvested, filtered and used to infect target HeLa cells in the presence of polybrene (1 μg/ml). Two days post‐infection, stably expressing cells were selected by growth in zeocin (200 μg/ml, Invitrogen).
Confocal microscopy and FLIP
HCT116 Bax/Bak DKO cells were seeded on a chambered cover glass (Thermo Scientific) in McCoy's 5A medium, grown for 36 h and transfected for 6–8 h. Cells were then incubated with MitoTracker far red for 10 min and imaged using a Zeiss 510 META confocal LSM microscope equipped with argon (458/488/514 nm lines) and HeNe (543/633 nm) lasers.
Fluorescence loss in photobleaching (FLIP) experiments were performed as described previously (Edlich et al, 2011). In short, cells were imaged prior to bleaching. Then a single region (diameter of 1 μm) within the nucleus was repeatedly bleached with two iterations of a 488 nm laser line (100% output) using a Zeiss LSM 510 META with a 63× PlanFluor lens. Two images were collected after each bleach pulse, with 30 s between pulses. After 15 cycles of bleaching and collecting 30 images, separate measurements on the mitochondria were taken to analyze loss in fluorescence. Unbleached cells neighboring analyzed cells served as controls for photobleaching during image acquisition of each measurement. Fluorescence intensities were normalized by setting the pre‐bleach fluorescence to 100% signal.
Apoptosis activity assays
For caspase‐3/7 measurements, HCT116 Bax/Bak DKO cells were seeded in 6‐well plates and transfected with the same amount of plasmids containing different Bax or Bak variants in the presence or absence of Bcl‐xL. Then, cells were treated with 1 μM STS or left untreated. After 12 h incubation, whole‐cell lysates were isolated and fluorogenic caspase‐3 substrate N‐acetyl‐DEVD‐AMC (BD Pharmigen) was added to each sample according to manufacturer's protocol. Samples were incubated for 1 h at 37°C and then fluorescence signal was measured 30 times at 2‐min intervals with an excitation wavelength of 355 nm and an emission wavelength of 460 nm in a plate reader. The same samples were also subjected to analysis of PARP cleavage by Western blot.
Whole‐cell lysis and subcellular fractionation
Cells were harvested and incubated in cell lysis buffer (20 mM Tris, 100 mM NaCl, 1 mM EDTA, 0,5% Triton X‐100, protease inhibitor cocktail) for 15 min on ice. Whole‐cell extracts were obtained by centrifugation at 15,000 × g for 10 min at 4°C. Protein concentrations were determined using a Bradford Assay (RotiQuant, Roth). Samples were boiled in SDS sample buffer for 10 min at 95°C and subsequently subjected to SDS–PAGE and Western Blot analysis. Subcellular fractionations were performed as previously described (Todt et al, 2013).
Mitochondrial pellets were prepared as described above and resuspended in 100 mM Na2CO3 at pH 11.5. The samples were incubated on ice for 20 min to disrupt protein–protein interactions of peripheral proteins while interactions protected by the lipid bilayer like lipid–protein interactions remain intact. Then, the membranes were pelleted at 150,000 × g for 30 min at 4°C. The supernatant, containing membrane‐associated proteins, was subjected to protein precipitation by acetone. The pellet was resuspended once more in 100 mM Na2CO3 at pH 11.5 and incubated on ice for 20 min. Finally, the resuspended samples were centrifuged at 155,000 × g for 30 min at 4°C to obtain carbonate inextractable proteins. Both fractions were assayed by Western blot.
Clonogenic survival assay
For the evaluation of clonogenic survival, HCT116 Bax/Bak DKO cells were seeded in 6‐well plates. After the transfection with different Bax or Bak variants in the presence or the absence of Bcl‐xL, 1 μM STS was applied to each sample for 24 h or cells were left untreated. After 12–16 days, surviving colonies were fixed and stained with 1% methylene blue.
Cells were seeded in 6‐well plates and transfected with equal DNA amounts of Bax or Bak and pcDNA or Bcl‐xL. Cells were harvested and washed with PBS. Cells were incubated in Annexin V binding buffer substituted with Annexin V–PE (eBioscience) for 10 min at room temperature. Finally, cells were fixed with 1% paraformaldehyde (neoLab) and measured using a BD FACSCalibur flow cytometer (BD Biosciences). Data were analyzed using FlowJo Version 10.0.7.
FT, ZC and FR designed and performed experiments, analyzed data and wrote the paper. FEm, JL, GI and AK performed experiments and analyzed data. SWGT performed experiments, analyzed data and edited the paper. SF and HFL analyzed data and helped to conceptualize the project. FEd conceptualized the project, designed experiments, analyzed data and wrote the paper.
Conflict of interest
The authors declare that they have no conflict of interest.
Supplementary Figures S1–S7
Legends for Supplementary Figures
We thank Dr. J.‐C. Martinou and Dr. R.J. Youle for reading and fruitful discussion of the manuscript. We are also grateful to S. Liebscher for superb technical assistance. We thank Dr. T. Wandless for the provision of Shield‐1 reagent. This work is supported by the Emmy Noether program of the German Research Council (Deutsche Forschungsgemeinschaft, DFG), the Else Kröner Fresenius Foundation, the Spemann Graduate School of Biology and Medicine (SGBM, GSC‐4) and the Centre for Biological Signalling Studies (BIOSS, EXC‐294) funded by the Excellence Initiative of the German Federal and State Governments.
FundingEmmy Noether program of the German Research Council (Deutsche Forschungsgemeinschaft, DFG)
- © 2014 The Authors