BFL1 modulates apoptosis at the membrane level through a bifunctional and multimodal mechanism showing key differences with BCLXL

Abstract

BFL1 is a relatively understudied member of the BCL2 protein family which has been implicated in the pathogenesis and chemoresistance of a variety of human cancers, including hematological malignancies and solid tumours. BFL1 is generally considered to have an antiapoptotic function, although its precise mode of action remains unclear. By quantitatively analyzing BFL1 action in synthetic membrane models and in cells, we found that BFL1 inhibits apoptosis through three distinct mechanisms which are similar but not identical to those of BCLXL, the paradigmatic antiapoptotic BCL2 family protein. Strikingly, alterations in lipid composition during apoptosis activate a prodeath function of BFL1 that is based on noncanonical oligomerization of the protein and breaching of the permeability barrier of the outer mitochondrial membrane (OMM). This lipid-triggered prodeath function of BFL1 is absent in BCLXL and also differs from that of the apoptotic effector BAX, which sets it apart from other BCL2 family members. Our findings support a new model in which BFL1 modulates apoptosis through a bifunctional and multimodal mode of action that is distinctly regulated by OMM lipids compared to BCLXL.

The Incomplete Puzzle of the BCL2 Proteins

“…These models propose that antiapoptotic proteins repress apoptosis neutralizing either BH3-only activators (direct model, MODE 1) or BAX-type proteins (indirect model, MODE 2) [6,7,8,9,10,11,12,13,14]. In addition, retrotranslocation or inhibition MODE 0 postulates that BCL2-type proteins inhibit apoptosis by keeping BAX-type proteins inactive through continuous retrotranslocation from the mitochondrial surface into the cytosol [15,16,17,18,19]. These models, however, do not consider an enigmatic property shared by all BCL2-type proteins, which is their ability to promote, rather than inhibit, apoptosis under specific conditions (PRODEATH MODE) [15,20,21,22].…”

“…In addition, retrotranslocation or inhibition MODE 0 postulates that BCL2-type proteins inhibit apoptosis by keeping BAX-type proteins inactive through continuous retrotranslocation from the mitochondrial surface into the cytosol [15,16,17,18,19]. These models, however, do not consider an enigmatic property shared by all BCL2-type proteins, which is their ability to promote, rather than inhibit, apoptosis under specific conditions (PRODEATH MODE) [15,20,21,22]. The complex interaction network that orchestrates these proteins’ actions is commonly termed the BCL2 interactome, which constitutes an intricate puzzle yet unresolved (Figure 1).…”

“…First, BCL2 proteins perform their function, that is, regulating mitochondrial outer membrane (MOM) permeabilization (MOMP) to release apoptogenic factors into the cytosol and therefore induce apoptosis, when targeted to the membrane. The membrane and its constituting lipids affect the pieces of the BCL2 puzzle by modulating the affinities between the different family members [23,24] or by altering their canonical phenotype or function, for example switching their antiapoptotic nature to proapoptotic activity [15]. In addition to the membrane environment, posttranslational modifications have also been shown to modify the affinity and function of the BCL2 family members [13,25,26].…”

Bax mitochondrial residency is more critical than Bax oligomerization for apoptosis

“…Lipid mixtures containing egg phosphatidylcholine and cardiolipin in a 80:20 ratio were used. To obtain proteoliposomes, LUVs (100 nm) were prepared as described elsewhere (Flores-Romero et al, 2018, Subburaj et al, 2015. LUVs were incubated with 5 nM cBid and 2.5 nM Bax (S4C C62S C126S)-488 or Bax G179P (S4C C62S C126S)-488 mutants for 1h at room temperature.…”

“…Samples were illuminated for 35 ms with a delay time between frames of 25 ms (number of frames 1200) with an intensity of ~ 0.1 kW/cm 2 . Stoichiometry analysis was performed as described elsewhere (Flores-Romero et al, 2018, Subburaj et al, 2015. Data are provided as raw values and no correction for partial labelling was applied, because this maneuver would introduce uncertainty, particularly, with BaxG179P, which had a relatively low labeling efficiency.…”