BAK (BCL2 Antagonist/Killer 1)

The constitutively mitochondria-resident pro-apoptotic effector of the intrinsic pathway. BAK — together with its obligate paralogue bax — forms the pores that permeabilize the mitochondrial outer membrane (MOMP), releasing cytochrome c and committing a cell irreversibly to death. Unlike BAX (which is cytosolic in healthy cells), BAK is anchored to the mitochondrial outer membrane at steady state, making it the immediate-response effector when anti-apoptotic restraint is lifted. In aging biology, BAK is a convergence target for senolytics: anti-apoptotic Bcl-2 family proteins (MCL-1 dominant; BCL-xL secondary) hold BAK in check in senescent cells, and BH3-mimetic drugs release this brake to trigger selective senescent-cell clearance.

Identity

UniProt: Q16611 (BAK_HUMAN)
NCBI Gene: 578
HGNC symbol: BAK1 (aliases: BAK, BCL2L7, CDN1)
Mouse ortholog: Bak1 (one-to-one; function conserved)
Length: 211 amino acids; ~23.4 kDa

Discovery

BAK was identified in 1995 by two independent groups. Farrow et al. cloned Bak as a Bcl-2 homologue through a yeast two-hybrid screen using adenovirus E1B 19K as bait; overexpression accelerated apoptosis in nerve-growth-factor-deprived neurons; the Farrow clone encoded a 216-amino-acid ORF ¹. Simultaneously, Chittenden et al. independently cloned BAK from a human B-cell library and characterized it as an inducer of apoptosis whose enforced expression was sufficient to kill FL5.12 cells and serum-deprived fibroblasts, and whose BH3 domain was required for both cell killing and binding to BCL-xL; the Chittenden clone encoded 211 amino acids (Mr 23.4 kDa), which is the canonical length now in UniProt Q16611 ² ³. A third concurrent paper (Kiefer et al. 1995, Nature 374:736–739) also described BAK independently and likewise reported 211 amino acids. The 216 aa Farrow clone likely reflects a longer N-terminal isoform or cloning artifact; the 211 aa sequence is the canonical reference. Yeast two-hybrid data from Farrow showed BAK does not homodimerize with itself in that assay, unlike Bax. The parallel discovery of two functionally redundant effectors (BAK and BAX) foreshadowed the central importance of their combined function in developmental biology.

Domain structure

Domain	Residues (approx.)	Function
BH3 motif	74–88	Essential for pro-apoptotic activity; mediates binding to anti-apoptotic Bcl-2 proteins
BH1 motif	117–136	Contributes to hydrophobic groove structure
BH2 motif	169–184	Contributes to hydrophobic groove structure
Transmembrane helix (alpha-9)	188–205	C-terminal membrane anchor; constitutively inserts into mitochondrial OMM
Zinc-binding site	Asp160 (D160), His164 (H164)	Mediates homodimerization; sub-millimolar Zn inhibits MOMP (Moldoveanu 2006)

The BH3-BH1-BH2 region forms the hydrophobic groove that is the docking site for activating BH3-only proteins (BIM, BID, PUMA) and the competitive binding surface for anti-apoptotic proteins (BCL-xL, MCL-1, BCL-2). The hydrophobic groove is also the structural target of small-molecule BH3 mimetics.

Key PTMs

Met1 removal + Ala2 N-acetylation — constitutive; confirmed by UniProt annotation (Q16611). No known regulatory role reported.
No phosphorylation or ubiquitination sites are as well-characterized on BAK as on bax. unsourced — BAK’s PTM repertoire is less studied than BAX’s; see GenAge / PhosphoSitePlus for updates.

Mechanism of action

Resting state (constitutively OMM-anchored)

In healthy cells, BAK is already inserted into the mitochondrial outer membrane (OMM) via its C-terminal transmembrane helix — in contrast to BAX, which resides in the cytosol. BAK is held as an inactive monomer through sequestration by anti-apoptotic Bcl-2 family members (predominantly MCL-1 and BCL-xL) that bind its BH3 domain via their hydrophobic grooves. This tonic sequestration is the primary brake on BAK activity.

Activation: direct and indirect models

1. Direct activation — BH3-only “activator” proteins (BIM, tBID) bind the hydrophobic groove of BAK, triggering conformational changes: exposure of the N-terminus (alpha-1 helix), disruption of a key disulfide bond between helices, and release from anti-apoptotic partners. This model is supported by structural and biochemical reconstitution data. PUMA is classified in Chipuk 2010 primarily as a sensitizer/derepressor (binding the anti-apoptotic repertoire) rather than a direct activator of BAK/BAX, though some data suggest it may also act directly; its role remains contested ⁴.

2. De-repression (indirect) model — BH3-only “sensitizers” (NOXA, BAD, BIK) bind and neutralize anti-apoptotic proteins that were sequestering BAK, freeing it to oligomerize. NOXA is particularly relevant for BAK because it selectively targets MCL-1 (the dominant BAK guard) while having low affinity for BCL-xL/BCL-2.

Current consensus: both mechanisms likely operate simultaneously; relative contributions are cell-type-dependent ⁴. A critical distinction from BAX: because BAK is already at the OMM, there is no translocation step — activation leads directly to conformational change and oligomerization.

Mitochondrial execution sequence

Conformational change — N-terminal alpha-1 helix exposed; BH3 domain disengaged from anti-apoptotic partner
Oligomerization — BAK (and/or BAX) molecules form symmetric dimers → higher-order oligomeric pore assemblies; the Zn²⁺-bridged homodimer seen in the cBAK crystal structure (Moldoveanu 2006), coordinated via D160–H164 pairs from each monomer ⁵, represents an inhibitory resting conformation — the BH3 binding pocket is occluded in this state and is not the active pore state
MOMP — cytochrome c, Smac/DIABLO, and HtrA2/Omi released into cytosol
Apoptosome — cytochrome c + Apaf-1 + dATP → procaspase-9 recruitment
Caspase cascade — caspase-9 → caspase-3/7 → substrate cleavage → cell death

MOMP is the point-of-no-return: once achieved, cell death proceeds even if caspases are inhibited.

BAK vs BAX: key differences

Feature	BAK	BAX
Resting location	Mitochondrial OMM (constitutive)	Cytosol (translocates upon activation)
Primary anti-apoptotic guard	MCL-1 (dominant); BCL-xL secondary	BCL-xL (dominant); BCL-2, MCL-1 secondary
Activation mechanism	Conformational change in situ	Translocation + conformational change
Single-KO lethality	Not lethal (BAX compensates)	Not lethal (BAK compensates)
Double-KO lethality	Yes (Bak-/-Bax-/- perinatally lethal)	(same; see Lindsten 2000)

The differential dependence on MCL-1 vs BCL-xL has direct pharmacological implications: NOXA-like sensitizers or MCL-1 inhibitors preferentially disrupt BAK, whereas BIM/BAD-like sensitizers or BCL-xL inhibitors preferentially disrupt BAX. A cell’s survival may depend on inhibiting both nodes simultaneously.

Regulation by anti-apoptotic Bcl-2 family members

Protein	Affinity for BAK	Senolytic releasing it
mcl-1	High (dominant BAK guard)	MCL-1 inhibitors (S63845, AMG-176); also NOXA induction
bcl-xl	High (secondary guard)	Navitoclax; BCL-xL-selective A1331852; BCL-xL PROTACs (DT2216)
bcl-2	Low-moderate	Venetoclax; navitoclax

The dominance of MCL-1 over BAK (vs BCL-xL over BAX) creates a therapeutic divergence: navitoclax (BCL-2/BCL-xL dual inhibitor) triggers apoptosis largely via the BAX node, while MCL-1 inhibitors access the BAK node more directly.

Genetic evidence: BAK/BAX double knockout

Lindsten et al. 2000 generated Bak/Bax double-knockout (DKO) mice ⁶:

Perinatal lethality — the majority of DKO mice die perinatally (within ~48 hours of birth; fewer than 10% survive to adulthood); interdigital web tissue persists (no developmental apoptosis); imperforate vagina in all adult females; neurological abnormalities including deafness and circling behavior; accumulation of excess hematopoietic progenitors and mature lymphocytes
Single-KO phenotype — Bak-/- alone (with WT Bax) is essentially normal, confirming functional redundancy between the two effectors. Bax-/- alone produces lymphoid hyperplasia and male infertility (Knudson 1995)
Conclusion: BAK and BAX together constitute the minimal required set of OMM pore-forming effectors for developmental apoptosis; neither alone is sufficient for normal development

Degenhardt et al. 2002 demonstrated using baby mouse kidney cells from bax-/-, bak-/-, and bax-/-bak-/- mice that loss of either Bax or Bak alone did not abrogate TNF-α-mediated apoptosis, but loss of both conferred strong resistance — confirming functional redundancy in the extrinsic pathway as well as the intrinsic ⁷. (Note: a claim about BAK and BAX independently promoting cytochrome c release from isolated mitochondria derives from a different study — Suzuki et al. 2000 by Suzuki M, Youle RJ, Tjandra N — which is not locally archived; the original footnote had an incorrect DOI-author assignment.) needs-replication

Dimension	Status	Notes
Pathway conserved in humans?	yes	BH3-domain mechanism, OMM topology, and BCL-2 family interactions conserved; BAK structure (Moldoveanu 2006) is human protein
Phenotype conserved in humans?	yes	BCL-2 family dysregulation drives human lymphomas; MCL-1 and BCL-xL overexpression found in many human cancers and senescent cells
Replicated in humans?	partial	Pharmacological BCL-2/BCL-xL inhibition (BH3 mimetics) validated in human oncology; aging-specific human replication sparse

Aging relevance

Senescent cell apoptosis resistance and BAK suppression

Senescent cells are characteristically resistant to apoptosis despite chronic genotoxic and proteotoxic stress. The prevailing mechanistic model is “survival priming”: senescent cells upregulate anti-apoptotic Bcl-2 family members — particularly MCL-1 and BCL-xL — which maintain BAK (and BAX) in a sub-threshold activated state, poised but blocked from oligomerizing.

Zhu et al. 2016 identified navitoclax (ABT-263) as a senolytic: it reduced viability of senescent human endothelial cells, lung fibroblasts, and murine embryonic fibroblasts, with sensitivity correlating with the specific pattern of anti-apoptotic protein expression in each cell type ⁸. Because navitoclax inhibits both BCL-2 and BCL-xL, it addresses both BAX and BAK guard nodes. The study did not independently dissect BAK’s contribution from BAX’s.

The multi-target requirement seen in some senescent cell types is demonstrated directly in Zhu 2016: in IMR90 human lung fibroblasts, siRNA knockdown of BCL-xL alone or BCL-xL + BCL-2 was not senolytic, but the triple BCL-2 + BCL-xL + BCL-w siRNA combination (reflecting navitoclax’s target set) was senolytic (Fig. 4A). This implies that BAK is protected by overlapping redundant anti-apoptotic coverage in some lineages, requiring simultaneous disruption of multiple guards ⁸. Navitoclax was NOT senolytic in human primary preadipocytes, demonstrating that senolytic activity is cell-type-restricted. needs-replication

MCL-1 as a BAK-specific senolytic target

Because MCL-1 is the dominant anti-apoptotic regulator of BAK specifically, MCL-1 inhibitors represent a BAK-selective senolytic strategy complementary to BCL-xL inhibitors (which are more BAX-relevant). First-generation MCL-1 inhibitors (S63845, AMG-176) have shown senolytic activity in preclinical models, though cardiac toxicity from MCL-1 inhibition in cardiomyocytes remains a concern. needs-human-replication dose-response-unclear

Stem cell pool maintenance

Insufficient BAK/BAX-mediated apoptosis leads to accumulation of damaged or aberrant cells, contributing to tissue dysfunction and cancer risk. Conversely, chronic BAK activation under oxidative or genotoxic stress may deplete stem cell pools, a separate driver of aging phenotypes. The BAK/BAX apoptosis threshold in stem cell compartments is likely regulated tissue-specifically, and dysregulation in either direction (too much or too little) is pathological. no-mechanism

GenAge status

BAK1 is not listed in the GenAge human database (Build 21, ~307 genes as of 2026). BAK’s aging relevance derives from its role as the primary effector in senolytic apoptosis pathways, not from direct lifespan-modification evidence in model organisms.

Pharmacology: BAK-relevant agents

BH3 mimetics (release BAK from anti-apoptotic guards)

Drug	Primary BAK-relevant target	Clinical stage	Notes
Navitoclax (ABT-263)	BCL-xL (+ BCL-2)	Phase 2 (oncology + senolytics)	Dose-limiting thrombocytopenia (BCL-xL in platelets); senolytic in vivo mouse data
Venetoclax (ABT-199)	BCL-2 selective	FDA-approved (CLL, AML)	Spares BCL-xL → lower platelet toxicity; less BAK-relevant than navitoclax
S63845	MCL-1 selective	Preclinical / Phase 1 (oncology)	Directly targets dominant BAK guard; cardiac toxicity concern
AMG-176	MCL-1 selective	Phase 1 (oncology)	As above
DT2216	BCL-xL PROTAC (platelet-sparing)	Preclinical / Phase 1	Platelet-sparing via absence of CRBN E3 ligase in platelets

No BAK-direct small-molecule activators (analogous to BTSA1 for BAX) are in clinical development as of 2026. unsourced — confirm at ClinicalTrials.gov.

Interaction network (key nodes)

p53 — transcriptional inducer of PUMA and NOXA, which then activate/sensitize BAK; p53 also directly induces Bax (less direct role via BAK)
puma — BH3-only protein; p53 transcriptional target; classified primarily as a sensitizer/derepressor in Chipuk 2010 (binds anti-apoptotic repertoire) rather than a direct activator of BAK, though this classification remains debated ⁴
bim — BH3-only activator; binds both BAK and BAX; released from BCL-xL by navitoclax
bid — BH3-only activator; cleaved by caspase-8 (tBID) at OMM to activate BAK
bax — obligate paralogue; partially compensates when BAK is absent; cooperates in pore formation
mcl-1 — primary anti-apoptotic guard of BAK; targeted by NOXA and MCL-1 inhibitors
bcl-xl — secondary anti-apoptotic guard; targeted by navitoclax and BCL-xL PROTACs
bcl-2 — tertiary anti-apoptotic guard; targeted by venetoclax
cytochrome-c — released downstream of MOMP; initiates apoptosome
apaf-1 — forms apoptosome with cytochrome c; activates caspase-9

Pathway membership

apoptosis-pathway — intrinsic mitochondrial apoptosis; BAK is the constitutively-OMM-resident pore-forming effector
p53-pathway — downstream effector target via PUMA/NOXA/BIM induction
bcl-2-family-signaling — central node of the Bcl-2 rheostat; MCL-1/BCL-xL regulate BAK directly

Limitations and gaps

#gap/needs-human-replication — Senolytic efficacy of BAK-targeted strategies (navitoclax, MCL-1 inhibitors) in human aging contexts is preclinical-only; no completed human aging trials as of 2026.
#gap/no-mechanism — Whether a cell activates BAK (apoptosis) versus enters senescence under the same stress conditions is incompletely understood; likely involves BAK expression level, MCL-1 abundance, NOXA induction kinetics, and cell-type-specific thresholds.
#gap/needs-replication — The specific MCL-1-dominance model for BAK in senescent cells is mechanistically inferred from pharmacological studies; direct molecular demonstration in aged human tissues is absent.
#gap/dose-response-unclear — Therapeutic windows for MCL-1 inhibitors in senolytic applications (senescent cell killing vs normal-tissue toxicity, especially cardiac) have not been established.
BAK’s PTM regulatory landscape is less characterized than BAX’s. PhosphoSitePlus as of 2026 lists fewer regulatory modifications on BAK than on BAX.
BAK1 is not listed in GenAge human database (Build 21). Its aging relevance is mechanistic (effector role in senolytic pathways), not genetic-longevity-association based.

Footnotes

doi:10.1038/374731a0 · Farrow SN, White JH, Martinou I, Raven T, Pun KT, Grinham CJ, Martinou JC, Brown R · 1995 · Nature 374:731–733 · in-vitro + in-vivo · model: yeast two-hybrid + neurons (NGF withdrawal); first cloning of BAK; showed BAK overexpression accelerates apoptosis; locally available in archive ↩
doi:10.1038/374733a0 · Chittenden T, Harrington EA, O’Connor R, Flemington C, Lutz RJ, Evan GI, Guild BC · 1995 · Nature 374:733–736 · in-vitro · model: cell lines; demonstrated enforced BAK expression induces rapid apoptosis antagonized by BCL-2/BCL-xL; locally available in archive ↩
doi:10.1002/j.1460-2075.1995.tb00246.x · Chittenden T, Flemington C, Houghton AB, Ebb RG, Gallo GJ, Elangovan B, Chinnadurai G, Lutz RJ · 1995 · EMBO J 14(22):5589–5596 · in-vitro · model: cell lines; characterized BH3 domain (residues 67–94) as the conserved death-mediating domain necessary and sufficient for cytotoxic activity and BCL-xL binding; identified same domain as critical in Bax and Bip1; locally available in archive ↩
doi:10.1016/j.molcel.2010.01.025 · Chipuk JE, Moldoveanu T, Llambi F, Parsons MJ, Green DR · 2010 · Mol Cell 37:299–310 · review · comprehensive unified model of BCL-2 family activation mechanisms (direct activator vs. de-repression); 1,398 citations; locally available in archive ↩ ↩² ↩³
doi:10.1016/j.molcel.2006.10.014 · Moldoveanu T, Liu Q, Tocilj A, Watson M, Shore G, Gehring K · 2006 · Mol Cell 24:677–688 · in-vitro (crystallography) · 1.47 Å crystal structure of calpain-proteolysed BAK (cBAK) homodimer; identified zinc-binding site coordinated by D160 (Asp) and H164 (His) as mediating inhibitory homodimerization; demonstrated sub-millimolar Zn inhibits MOMP in bak/bax DKO MEF mitochondria; inhibition alleviated by H164A mutation ↩
doi:10.1016/s1097-2765(00)00136-2 · Lindsten T, Ross AJ, King A, Zong WX, Rathmell JC, Shiels HA, et al. (Thompson lab) · 2000 · Mol Cell 6(6):1389–1399 · in-vivo · model: Bak/Bax double-knockout mouse (mixed 129S3×129X1×C57BL/6 background); 1,466 citations; demonstrated combined loss of BAK + BAX causes perinatal lethality (~48h), persistent interdigital webs, imperforate vagina, neurological defects, lymphoid accumulation; locally available in archive ↩
doi:10.1074/jbc.m109939200 · Degenhardt K, Sundararajan R, Lindsten T, Thompson C, White E · 2002 · J Biol Chem 277(16):14127–14134 · in-vitro + in-vivo · model: primary baby mouse kidney (BMK) cells from bax-/-, bak-/-, bax-/-bak-/- mice; demonstrated loss of both Bax and Bak required for resistance to TNF-α-mediated apoptosis; either alone was sufficient for cell death signaling; also showed Bax and Bak are dispensable for p53-dependent transformation suppression. NOTE: This DOI was previously mislabeled in this wiki as “Suzuki M, Youle RJ, Tjandra N” — that is a different paper (Suzuki et al. 2000, Cell 103:645–654) not available in archive. ↩
doi:10.1111/acel.12445 · Zhu Y, Tchkonia T, Fuhrmann-Stroissnigg H, Dai HM, Ling YY, Stout MB et al. · 2016 · Aging Cell 15(3):428–435 · in-vitro + in-vivo · model: senescent human HUVECs, lung fibroblasts, MEFs; first identification of navitoclax as a senolytic; sensitivity correlated with BCL-2 family protein expression pattern; locally available in archive ↩ ↩²

🧬 Aging Wiki

Explorer

bak