Anti-apoptotic BH3-only proteins inhibit Bak-dependent apoptosis

This article has been Reviewed by the following groups

Read the full article See related articles

Listed in

Log in to save this article


Bcl-2 family proteins regulate induction of intrinsic apoptosis through initiating mitochondrial outer membrane permeabilization (MOMP). Activation of the MOMP effectors Bax and Bak is controlled by interplay levels of anti-apoptotic Bcl-2 proteins (e.g. Mcl-1) and pro-apoptotic BH3-only proteins (e.g. BIM). Using a genome-wide CRISPR-dCas9 transactivation screen we identified two Bcl-2 family proteins, BNIP5 and Bcl-G, as inhibitors of Bak, but not Bax induced apoptosis. BNIP5 was able to block Bak activation in different cell types and in response to various cytotoxic therapies. The BH3 domain of BNIP5 was both necessary and sufficient to block Bak activation. Mechanistically, the BH3 domains of BNIP5 and Bcl-G act as a selective Bak activators, while not inhibiting anti-apoptotic proteins. This led to increased binding of activated Bak to Mcl-1, which prevented apoptosis engagement, identifying BNIP5 and Bcl-G as anti-apoptotic BH3-only proteins.

Article activity feed

  1. Review coordinated via ASAPbio’s crowd preprint review

    This review reflects comments and contributions by Joe Biggane, Luciana Gallo, Rachel Lau, Sam Lord, Dipika Mishra, Claudia Molina. The comments were synthesized by Iratxe Puebla.

    The study reports two two Bcl-2 family proteins, BNIP5 and Bcl-G, which inhibit Bak-dependent apoptosis through engagement of MODE 2 inhibition. The BH3 domains of these proteins act as selective Bak activators, while not inhibiting anti-apoptotic proteins, leading to increased binding of activated Bak to Mcl-1, which prevents apoptosis.

    The reviewers raised a couple of questions about the methodology and several other suggestions for the paper, outlined below:


    Throughout the study various BH3 mimetics are used, but the combinations in which they are used and/or the doses employed could be more clearly reported. For example, in Figure 1E and 1F ABT-737 and S63845 are used at 1 μM. Then, in Figure 1H, A-331852 is substituted for ABT-737 in combination with S63845 and the concentration is not reported. In Figure 1H, ABT-737 and S63845 are used again, but this time at a concentration of 2 μM each. Other concentrations are used in Figures 2, 3, and 5. There seems to be a dose-response assay in Figure 3B, but it is used for a specific use case. It would be beneficial to report all combinations and doses employed, and the rationale for them in the main text, to allow readers to fully interpret the data presented.

    In various figures, there is a concern about the statistical approach to calculate p-values based on multiple measurements or cells within each sample. The t-test and ANOVA assume that each measurement is independent, and multiple nuclei within the same sample are not independent. Recommend either not reporting p-values or averaging together the values from each biological replicate to calculate the p-value using those sample-level means. For more information, see and

    Specific comments

    • Introduction ‘The two MOMP effectors Bcl-2 associated x (Bax) and Bcl-2 antagonist killer (Bak) are inactive in resting cells as these cells exhibit low levels of proapoptotic BH3-only proteins (e.g. BIM)....and some are additionally able to activate Bax and Bak (sensitizers and direct activators, e.g. BIM)’ - Recommend revising the fragment for clarity, adding references to support the statements and possibly an introductory figure to help visualize the proteins involved.
    • Introduction, last paragraph ‘We found that two Bcl-2 proteins, Bcl-2 interacting protein 5 (BNIP5) and Bcl-G, act as selective inhibitors of Bak-dependent but not Bax-dependent apoptosis…’- The fragment is unclear, BNIP5 and Bcl-G are first reported as Bak-inhibitors, then activators and back to inhibitors. Does this mean to describe protein-protein interaction and changes in conformation?

    Figure 1

    -Recommend using a different color scheme for Figure 1E to assist visual interpretation of the results, in particular consider using a color-blind friendly color palette.

    -‘colony formation (F)’ - The text later on refers to ‘clonogenic survival’, would it be possible to clarify in the legend or text what is being assessed, i.e. recovery assay, clonogenic survival or colony formation?

    -Figure 1G - Please clarify whether 2 uM of each are used in this experiment.

    -‘We transduced PC9 lung cancer or A375 melanoma cells…’ - It is nice to see that different cell lines were assessed to address any cell line-specific effects. Would be interesting to see if this effect occurs in normal cell lines and not just cancer cell lines.

    • Figure 2 - The inline color-coded legends are useful when bars are displayed but in the figure several bars are close to zero, consider an alternative method to label the bars.
    • Results ‘...suggesting posttranscriptional regulation of Bak levels by BNIP5’ - Maybe large proteome databases of multiple cell lines (e.g CCLE) can be datamined to determine the correlation between BNIP and Bak expression?
    • Figure 3B, 3D and 3E - Please clarify the concentrations used for each treatment in the figure legend.
    • Methods, Cell viability and cell death measurements - The study assessed cell death or cell viability with either live cell imaging, or in fixed cells, can the methodology for this be elaborated upon? Also, propidium iodide staining is used in several sections of the results, recommend adding information about this under the Methods section.
    • Methods - There are several missing references in the Methods section.
    • Suggest adding Supplemental Figure 6 as a graphical abstract.