µ-Calpain conversion of antiapoptotic Bfl-1 (BCL2A1) into a prodeath factor reveals two distinct alpha-helices inducing mitochondria-mediated apoptosis

Abstract

Anti-apoptotic Bfl-1 and pro-apoptotic Bax, two members of the Bcl-2 family sharing a similar structural fold, are classically viewed as antagonist regulators of apoptosis. However, both proteins were reported to be death inducers following cleavage by the cysteine protease µ-calpain. Here we demonstrate that calpain-mediated cleavage of full-length Bfl-1 induces the release of C-terminal membrane active α-helices that are responsible for its conversion into a pro-apoptotic factor. A careful comparison of the different membrane-active regions present in the Bfl-1 truncated fragments with homologous domains of Bax show that helix α5, but not α6, of Bfl-1 induces cell death and cytochrome c release from purified mitochondria through a Bax/Bak-dependent mechanism. In contrast, both helices α5 and α6 of Bax permeabilize mitochondria regardless of the presence of Bax or Bak. Moreover, we provide evidence that the α9 helix of Bfl-1 promotes cytochrome c release and apoptosis through a unique membrane-destabilizing action whereas Bax-α9 does not display such activities. Hence, despite a common 3D-structure, C-terminal toxic domains present on Bfl-1 and Bax function in a dissimilar manner to permeabilize mitochondria and induce apoptosis. These findings provide insights for designing therapeutic approaches that could exploit the cleavage of endogenous Bcl-2 family proteins or the use of Bfl-1/Bax-derived peptides to promote tumor cell clearance.

Figure 1. μ-calpain cleaves Bfl-1 at two major sites in its N-terminus and releases a large C-terminal fragment with cytotoxic activity.

(A) GST-Bfl-1(1–151) was digested with recombinant µ-calpain in vitro and the fragments were separated by SDS-PAGE (left panel). Bands corresponding to cleaved products (arrows) were analyzed by mass spectrometry (MS). A higher concentration of recombinant GST-Bfl-1(1–151) was treated with µ-calpain, products were separated by SDS-PAGE and blotted to PVDF (right panel). A sub-band (asterisk) was excised and subjected to Edman degradation, which indicated that Bfl-1 had been N-terminally cleaved between residues F71 and N72. (B) Schematic representation of the wild type Bfl-1 protein showing the location of the identified μ-calpain cleavage sites (asterisks, upper sequence), of mutant Bfl-1 protein with a 6 aminoacids deletion surrounding the two cleaved residues (Bfl-1DD, middle sequence), and of mutant Bfl-1 in which the region overlapping the first cleavage site was swapped with a structurally homologous region in Bcl2L10 (Bfl-1SD, bottom sequence) (C) Confirmation of the two calpain cleavage sites identified in Bfl-1 using noncleavable mutants. 293T cell lysates expressing GFP-tagged Bfl-1 constructs were exogenously treated with μ-calpain. Lysates containing equal amount of GFP-tagged Bfl-1 proteins were separated by SDS-PAGE and analyzed by western blot with an anti-GFP antibody to detect the full length protein and N-terminal truncated fragments and with a polyclonal anti-Bfl-1 antibody to detect C-terminal truncated fragments. Upper and lower panels represent two independent experiments with different time of exposure. (D) BJAB cells were cultured with or without treatment with TNF/CHX in the presence or absence of the calpain inhibitor ALLN. Lysates were separated by SDS-PAGE and the presence of a cleaved fragment was assayed by western blot using an anti-Bfl-1 antibody. (E) Secondary structure of Bfl-1 in which the nine helices of the protein are represented by boxes along with the different BH domains (left panel).The two cleavage sites mapped in (A) are indicated. A comparison with the previously published µ-calpain cleavage site in Bax is also shown (bottom sequence). 3D structure of Bfl-1 (2VM6) and position of the different cleaved sites (red circles) are indicated. The different Bcl-2 Homology domains are colored in yellow (putative BH4), red (BH3), green (BH1) and blue (BH2). (F) FACS assays of Annexin V staining in HT1080 cells. Chimeric GFP constructs encoding GFP alone, or fusions of GFP with full-length Bfl-1 or Bax or with the various membrane-active α-helices corresponding to the C-terminal part of Bfl-1 or Bax, i.e. α5, α6 (PFD, pore forming domain) and α9 (FE, final exon) are represented. The α-helical topology of Bax and Bfl-1 corresponds to the structures solved in aqueous environment , . Transfected cells were stained for phosphatidylserine exposure using Cy3-conjugated Annexin V and the percentage of apoptotic GFP-expressing cells was determined by FACS 24 hours post transfection (right panel). Death of GFP-expressing and staurosporine (STS)-treated cells were also monitored as controls. Graphs shown are representative of three independent experiments.

Figure 2. The ectopic overexpression of GFP-tagged Bfl-1-α5/α9 and Bax-α5/α6 fragments induces cell death but with distinct Bax/Bak requirements.

(A) Chimeric GFP proteins used in this study. GFP-tagged constructs encoding GFP alone, or fusions of GFP with full-length Bfl-1 or Bax or with the various membrane-active α-helices of Bfl-1 or Bax, i.e. α1, α5, α6 and α9, are represented. The α-helical topology of Bax and Bfl-1 correspond to the structures solved in aqueous environment , . Because the structures of the membrane-bound forms of these proteins are unknown, we designed sequence boundaries that extend a few residues beyond the α-helical regions in the structures of the water-soluble forms. (B) FACS assays of Annexin V staining in HT1080 cells. Transfected cells were stained for phosphatidylserine exposure using Cy3-conjugated Annexin V and the percentage of apoptotic GFP-expressing cells was determined by FACS. Histograms represent the percentage of GFP-expressing cells binding Annexin V (upper panel). Assays were performed in triplicate (error bars correspond to standard deviations). Staurosporine (STS) treatment was included for comparison. (C) FACS histogram showing Annexin-V staining of MEF (left panel) and MEF-DKO cells (right panel) expressing the different GFP-tagged constructs described in (A). GFP-transfected cells treated with staurosporine (STS) or left untreated were used as controls. Assays were performed in triplicate (error bars correspond to standard deviations). (D) Expression and analysis of the various GFP-tagged proteins in mammalian cells. Western Blot analyses on transiently transfected MEF (top panel) or MEF DKO cells (bottom panel) at 24 h post-transfection. Proteins were separated by SDS-PAGE and the presence of fusion proteins with the correct size was tested by immunoblot with anti-GFP antibody. Analysis of caspase-3 activation below each panel shows the cleaved 17 kDa product indicative of activated caspase-3.

Figure 3. Subcellular localization of the GFP-tagged, Bfl-1/Bax-derived (poly)peptides.

MEF (left panels) and MEF DKO (right panels) cells were co-transfected with mitoDsRed plasmid (encoding DsRed2 fused to the mitochondrial targeting sequence from subunit VIII of human cytochrome c oxidase) and the GFP-tagged constructs. Subcellular distribution was analyzed by confocal microscopy 24 h after transfection. Confocal images showing GFP (green) and MitoDsRed (red) fluorescence. The DNA staining dye Topro-3 (blue) was used to visualize the nuclei. In merged images, the yellow color shows the co-localization of GFP and MitoDsRed at mitochondria. Scale bar, 10 µm.

Figure 4. Bfl-1-derived peptides have different abilities to permeabilize the MOM of mitochondria isolated from cultured cells.

Peptides were incubated at different concentrations (10 µM and 25 µM) with isolated mitochondria for the indicated times (5, 15, 30 and 60 min) and the release of cytochrome c was monitored by immunoblot (IB). MitoHsp70 (mHsp70) was used as control indicative of equal-loading and proper isolation of the pellet fraction containing mitochondria (Mito) in comparison to the supernatant fraction (SN). Cytochrome c release assays were performed using iBMKW2 (wild type) and iBMKD3 (double KO Bax/Bak) for all tested peptides. For Bfl-1-α5, wild type MEF and MEF DKO (Bax/Bak −/− double KO) cells were used in parallel.

Figure 5. The Bfl-1-α9 peptide induces mitochondrial permeabilization through a membrane-destabilizing mechanism.

Bfl-1-α9 peptide was incubated for the indicated times with mitochondria at lower concentrations (0.5 µM and 2.5 µM, top panels) and previously used in figure 4 (10 µM and 25 µM, bottom panels). The release of cytochrome c and the expression of MitoHsp70 were monitored as in figure 4, combined with the detection of the external membrane associated protein hexokinase 1 (HK1) and the matrix-contained protein MnSOD.