🧬 Aging Wiki

puma

Properties1
aliasesBBC3, p53-upregulated modulator of apoptosis, BCL2-binding component 3

12 min read

PUMA (BBC3)

p53-Upregulated Modulator of Apoptosis — a 193-amino-acid BH3-only member of the Bcl-2 family and the most potent inducer of apoptosis among the BH3-only proteins. Transcriptionally activated by p53 in response to DNA damage, hypoxia, and oncogenic stress, PUMA bridges nuclear stress sensing to mitochondrial apoptosis execution by binding and neutralizing all five major anti-apoptotic Bcl-2 family members (BCL-2, BCL-xL, BCL-w, MCL-1, A1). Its “promiscuous” binding profile, combined with a weak direct-activator function on BAX, explains its exceptional potency. Discovered simultaneously in 2001 by Nakano and Vousden 1 and by Yu, Zhang, Kinzler, and Vogelstein 2.

Naming note: The gene is officially BBC3 (BCL2 Binding Component 3); the protein is universally referred to as PUMA in the literature. This page uses PUMA as the primary name, consistent with p53-pathway.

Identity

FieldValue
UniProtQ9BXH1 (BBC3_HUMAN)
NCBI Gene27113
HGNC ID17868
Chromosomal location19q13.32
Length193 amino acids
MW~20.5 kDa
Isoforms4 (alternative splicing)
Mouse orthologBbc3 / Puma (one-to-one)

Functional domains

  • BH3 motif (residues ~137–151) — the sole structured interaction domain; intrinsically disordered in isolation but adopts an alpha-helical conformation upon binding the hydrophobic groove of anti-apoptotic Bcl-2 family proteins. The BH3 helix is essential for binding to BCL-2 and BCL-xL (demonstrated by co-IP in discovery papers 12); binding to BCL-w, MCL-1, and A1 attributed to later work 3 no-fulltext-access.
  • Intrinsically disordered N-terminus (~residues 1–28) and central disordered region (~71–138) — most of the PUMA protein is unstructured, which is characteristic of BH3-only proteins. The disorder may facilitate binding kinetics and multiple partner engagement.
  • Low-complexity region (71–82) — function not fully characterized. no-mechanism

PUMA lacks other BH domains (BH1, BH2, BH4) that characterize multi-domain Bcl-2 family members, consistent with its role as a pure activator/sensitizer rather than a pore-forming executioner.

Post-translational modifications

PTMSiteFunctional consequence
PhosphorylationSer10Annotated in UniProt; functional consequence not fully characterized no-mechanism

PUMA is primarily regulated at the transcriptional level rather than by PTM; its short half-life (rapidly degraded in non-stressed cells) also contributes to tight regulation. needs-replication — mechanisms of PUMA protein turnover (proteasomal vs lysosomal; relevant E3 ligases) need primary citations.

Core mechanism: how PUMA kills cells

PUMA operates via two partially overlapping mechanisms that together explain its exceptional potency 3:

1. Universal anti-apoptotic neutralization (sensitizer function)

PUMA’s BH3 helix binds the hydrophobic BH3-binding groove of all five canonical anti-apoptotic Bcl-2 family proteins 3. Note: the discovery papers (Nakano 2001, Yu 2001) directly demonstrated binding only to BCL-2 and BCL-xL by co-immunoprecipitation; binding to BCL-w, MCL-1, and A1 is attributed to Chipuk 2005 (closed-access; not independently verifiable here) no-fulltext-access:

Anti-apoptotic targetSignificance
bcl-2Displaced BIM, BID, tBID from sequestration; allows BAX/BAK activation
bcl-xlSame; BCL-xL is major survival factor in many cell types
BCL-wContributes to survival of neurons and certain epithelial cells
MCL-1Key survival factor in hematopoietic cells; PUMA binding overcomes MCL-1-dependent resistance
A1/BFL-1Rounds out coverage; no single anti-apoptotic escape route

By neutralizing all five targets, PUMA leaves no anti-apoptotic “escape valve” for the cell — a critical distinction from more selective BH3-only proteins such as BAD (prefers BCL-2/BCL-xL/BCL-w but not MCL-1 or A1) or NOXA (prefers MCL-1 and A1 only). The combination of these two selective BH3-only proteins essentially phenocopies PUMA’s activity, supporting the mechanistic interpretation 3.

2. Direct activator vs sensitizer — contested classification contradictory-evidence

PUMA’s status as a direct BAX/BAK activator is contested across the literature. Two competing classifications:

  • Direct activator (Chipuk 2005, Vogler 2025): PUMA can directly engage and conformationally activate BAX in a manner analogous to BIM and BID — the basis for the original framing on this page.
  • Sensitizer/derepressor only (Chipuk 2010 “BCL-2 Family Reunion” review, Mol Cell 37:299–310, doi:10.1016/j.molcel.2010.01.025, Fig 1B): PUMA’s pan-binding profile across all five anti-apoptotic family members is necessary but not sufficient to qualify as a direct activator under the Chipuk 2010 framework. Chipuk 2010 explicitly classifies PUMA as a sensitizer/derepressor; only BIM and tBID retain direct activator status. The bcl-2-family-signaling.md verifier (2026-05-04) confirmed this against Chipuk 2010 Fig 1B.

Both readings have substantial support; the field has not converged. This page retains the Chipuk 2005 dual-function framing but flags the tension; downstream pages (bcl-2-family-signaling, bak, bim, apoptosis-pathway) should refer to this section for the contested classification.

The downstream execution pathway

PUMA activity converges on bax (and BAK) activation → mitochondrial outer membrane permeabilization (MOMP) → cytochrome c release → apoptosome formation → caspase-9/caspase-3 cascade → cell death. The PUMA → MOMP axis is the commitment point of intrinsic apoptosis; once MOMP occurs, cell death is essentially irreversible.

Transcriptional regulation by p53

PUMA is among the most robustly p53-induced genes 12. The BBC3 promoter contains two tandem p53-response elements (p53REs) that bind p53 with high affinity. Induction kinetics are rapid (detectable mRNA within 1–2 h of DNA damage) and robust across cell types.

Stimuli that induce PUMA via p53:

  • DNA double-strand breaks (ionizing radiation, chemotherapy)
  • Replication stress (UV, hydroxyurea)
  • Oncogene activation (e.g., Myc, Ras)
  • Hypoxia (partial; HIF-1 independent contribution via p53 stabilization)
  • Oxidative stress

p53-independent PUMA induction also occurs via:

  • FOXO3a (nutrient deprivation, PI3K/Akt pathway suppression) unsourced — FOXO3a-mediated PUMA induction is established in the literature (Sunters et al. 2003 and others) but not from Yu 2001; citation needs to be replaced with appropriate primary source
  • ER stress / unfolded protein response (CHOP-dependent pathway)
  • Glucocorticoids (lymphocyte-specific; relevant to immune cell apoptosis during stress)

The p53-independence of some induction pathways means PUMA can execute apoptosis even when p53 is mutated or inactivated, relevant to some cancer therapy contexts but less central to normal aging biology.

Role in aging

Central effector of p53-dependent apoptosis under chronic stress

In normal aging, persistent low-level DNA damage (from telomere erosion, replication stress, ROS) chronically activates p53. PUMA is a primary transcriptional target through which this chronic p53 activation can drive apoptosis — particularly in rapidly dividing stem and progenitor cell compartments. The aging-relevant question is whether p53 → PUMA signaling eliminates damaged cells appropriately (tumor suppression) or excessively depletes tissue stem cells (stem-cell exhaustion).

Contribution to accelerated aging in hyperactive-p53 models

The Tyner 2002 “p53+/m” mouse model demonstrated that chronic p53 hyperactivation drives an accelerated-aging phenotype with early loss of tissue regenerative capacity 4. PUMA upregulation is implicated as one effector of this phenotype, though the quantitative contribution of PUMA vs other p53 transcriptional targets (p21, BAX, NOXA) in the p53+/m accelerated-aging phenotype has not been resolved. no-mechanism

DimensionStatusNotes
Pathway conserved in humans?yesp53 → BBC3 transcription; BH3-only mechanism; BCL-2 family binding — all conserved
Phenotype conserved in humans?partialExcess p53 activity depletes stem cells in mouse models; human stem cell exhaustion involves similar pathways, but direct PUMA measurements in aged human tissues are limited unsourced
Replicated in humans?noNo PUMA-targeted intervention data in humans needs-human-replication

PUMA in hematopoietic stem cell biology

PUMA mediates p53-dependent apoptosis in hematopoietic progenitor cells. The transcription factor Slug (SNAI2) represses PUMA expression in hematopoietic progenitors, providing protection against ionizing radiation-induced apoptosis 5. Loss of this protection (e.g., in aged bone marrow with declining Slug expression, or under chronic DNA damage) could contribute to the reduced self-renewal capacity of aged hematopoietic stem cells. needs-replication — direct evidence for PUMA as the executioner of aged-HSC depletion (vs other BH3-only proteins) is limited to indirect inference.

PUMA in the senescent cell / senolytic context

A key insight connecting PUMA to aging biology and senolytics:

  1. Senescent cells induced by DNA damage upregulate pro-apoptotic BH3-only proteins, including PUMA — a consequence of chronic p53/p21 activation. unsourced — primary citation for PUMA protein upregulation specifically in naturally senescent cells needed; Baker 2016 does not address this claim (that paper uses p16Ink4a-driven apoptosis, not PUMA).
  2. However, senescent cells simultaneously upregulate anti-apoptotic Bcl-2 family members (bcl-xl, BCL-2, BCL-w) as part of a survival program — these sequester BH3-only proteins in inactive complexes, allowing the senescent cell to persist rather than die. unsourced — the anti-apoptotic upregulation model in senescent cells is established in the senolytic literature (Zhu et al. 2015; Chang et al. 2016 and related work) but Baker 2016 does not discuss this mechanism; citation 6 removed from these specific claims.
  3. BH3-mimetic compounds (navitoclax/ABT-263; ABT-737; the dasatinib + quercetin combination by an indirect mechanism) displace PUMA and other BH3-only proteins from anti-apoptotic sequestration, reactivating the latent apoptotic program and selectively killing senescent cells.

This model explains why senescent cells are paradoxically sensitized to BH3 mimetics despite having a loaded apoptotic machinery — the threshold has already been primed by PUMA/BIM accumulation; only the anti-apoptotic blockade needs to be relieved.

needs-replication — the direct demonstration that PUMA specifically (vs BIM or BID) is the sequestered effector responsible for senescent-cell apoptotic priming has not been resolved at a protein-level in primary aged tissues.

Pathway membership

  • p53-pathway — PUMA is the primary pro-apoptotic transcriptional target of p53; induced by both the canonical stress-response arm and the senescence arm of p53 signaling
  • apoptosis-pathway — PUMA operates at the MOMP commitment step; upstream of cytochrome c release, downstream of BH3-only protein induction
  • dna-damage-response — ATM/ATR → CHK2/CHK1 → p53 stabilization → PUMA transcription; PUMA is the distal effector linking DNA damage to cell death

Key interactors

InteractorInteraction typeFunctional consequence
p53transcriptional inducerBinds BBC3 promoter p53RE; induces PUMA expression under stress
bcl-2BH3-domain binding (PUMA BH3 → BCL-2 groove)Displaces BIM/BID from BCL-2 sequestration; enables BAX activation
bcl-xlBH3-domain binding (high affinity)Same; particularly important in lymphocytes and neurons
MCL-1BH3-domain binding (high affinity)Overcomes MCL-1-mediated apoptotic resistance
baxdirect activator (weak; contested per Chipuk 2010 Fig 1B)Low-level direct BAX conformational activation per Chipuk 2005, independent of BH3-groove binding; reclassified as sensitizer-only by Chipuk 2010 — see “contested classification” section
SLUG/SNAI2transcriptional repressor of BBC3Represses BBC3 in hematopoietic progenitors; loss of repression sensitizes to p53-dependent killing

Therapeutic implications in aging

  • Anti-senolytic context: BH3 mimetics act partly by relieving PUMA sequestration in senescent cells, enabling their removal and delaying age-associated pathologies. See dasatinib, quercetin, and fisetin pages and the senolytics category page.
  • Stem cell protection: Transient PUMA inhibition during periods of acute stress (radiation, chemotherapy) has been explored as a strategy to protect the hematopoietic stem cell pool 5. Whether this approach could be applied to protect against age-related stem cell depletion is a conceptual extension not yet tested in aging contexts. needs-human-replication
  • Cancer risk tradeoff: PUMA suppression to protect stem cells must be weighed against reduced p53-dependent tumor suppression — the same antagonistic pleiotropy tension that runs through p53-centered aging biology. No clinical framework for this tradeoff exists. dose-response-unclear

Discovery

Two independent groups published back-to-back in Molecular Cell in 2001:

  • Nakano and Vousden 2001 — used microarray screening of p53-inducible SAOS-2 (p53-null osteosarcoma) and H1299-p53 cells to identify PUMA as a novel BH3-only protein robustly induced by p53 but not by p53 mutants lacking transactivation activity; showed that PUMA-α and PUMA-β overexpression alone is sufficient to induce rapid apoptosis (sub-G1 fraction ~26–28% at 24 hr in H1299 cells; ~82% at 24 hr in SAOS-2 cells); antisense oligonucleotide knockdown of PUMA partially reduced p53-mediated apoptosis, supporting a specific but not exclusive role 1. Note: PUMA genetic knockout data is from later work (Jeffers et al. 2003; Villunger et al. 2003), not from this discovery paper.
  • Yu, Zhang, Hwang, Kinzler, and Vogelstein 2001 — used SAGE (serial analysis of gene expression) of DLD1 colorectal cancer cells with doxycycline-inducible p53 to identify PUMA; showed PUMA is induced >10-fold within 9 hr of p53 activation and is rapidly induced by 5-FU and adriamycin in p53-wild-type HCT116 and SW48 cells but not in p53-null HCT116 cells; demonstrated PUMA co-immunoprecipitates with BCL-2 and BCL-xL (BH3-domain dependent); PUMA overexpression reduced colony formation by >1000-fold across four cancer cell lines regardless of p53 genotype. Originally named the gene JFY-1; agreed with Nakano/Vousden to adopt “PUMA” 2. Note: the paper did not test binding to BCL-w, MCL-1, or A1.

Both groups simultaneously defined the protein’s identity, p53 dependence, and pro-apoptotic function. The Nakano/Vousden paper introduced the “PUMA” name.

Limitations and open questions

GapTagNotes
PUMA vs other BH3-only proteins in aged tissuesneeds-replicationWhether PUMA specifically (vs BIM or NOXA) drives stem-cell depletion in normal aging is not established
Direct quantification in aged human tissuesunsourcedPUMA protein/mRNA levels in aged human bone marrow, muscle, or gut not well-documented
Ser10 phosphorylation functional significanceno-mechanismThe only known PTM on PUMA; kinase and functional consequence not fully characterized
Protein turnover mechanismsneeds-replicationE3 ubiquitin ligases targeting PUMA for proteasomal degradation not fully defined
Quantitative contribution in Tyner 2002 phenotypeno-mechanismRelative share of PUMA vs p21, BAX, NOXA in hyperactive-p53 accelerated aging not resolved
Therapeutic window for PUMA modulationdose-response-unclearNo clinical framework for transient PUMA suppression to protect stem cells vs cancer risk increase
Human trial dataneeds-human-replicationAll aging-relevant PUMA intervention data is preclinical; no direct PUMA-targeting agents in human trials for aging indications

Footnotes

Footnotes

  1. nakano-2001-puma-p53-proapoptotic · doi:10.1016/s1097-2765(01)00214-3 · n=N/A · in-vitro · model: SAOS-2 (p53-null osteosarcoma), H1299-p53 inducible, RKO, MCF-7, U2OS · archive: locally downloaded ↩ ↩2 ↩3 ↩4

  2. yu-2001-puma-bbc3-colorectal · doi:10.1016/s1097-2765(01)00213-1 · n=N/A · in-vitro · model: DLD1 colorectal (p53-inducible via tet-off), HCT116, SW48, SW480, H1299 · SAGE discovery method · archive: locally downloaded ↩ ↩2 ↩3 ↩4

  3. doi:10.1126/science.1114297 · in-vitro (biochemical + cell-based) · model: cell-free systems; HCT116 colorectal cells · key finding: PUMA displaces direct activators (BIM, BID) from BCL-xL sequestration AND has weak direct BAX activation; explains superior potency vs selective BH3-only proteins · 541 citations · archive: not OA; not locally downloaded no-fulltext-access ↩ ↩2 ↩3 ↩4

  4. tyner-2002-p53-mutant-aging · doi:10.1038/415045a · n=35 (p53+/m) + 56 (p53+/+) · in-vivo (transgenic mouse) · p<0.0001 survival · model: p53+/m C57BL/6×129/Sv · 1,434 citations · locally downloaded ↩

  5. doi:10.1016/j.cell.2005.09.029 · Wu et al. 2005 · in-vivo (mouse) · n=8–10 per genotype group · model: slug−/− and slug+/− mice (C57BL6.SJL background); 7.0 Gy TBI; slug−/−puma−/− double-KO rescue experiment (n=8–10/group, p<0.0001) · key finding: Slug directly represses puma transcription at intron-1 SBS3 Slug binding site; slug−/− mice radiosensitive due to excess puma-dependent apoptosis in myeloid progenitors; puma loss rescues slug−/− radiosensitivity · archive: locally downloaded ↩ ↩2

  6. doi:10.1038/nature16932 · Baker et al. 2016 · in-vivo (naturally aged INK-ATTAC transgenic mice) · n=multiple cohorts (mixed C57BL/6-129Sv-FVB and congenic C57BL/6) · model: p16Ink4a-promoter-driven FKBP-Casp8 apoptosis induction from 12 months; AP20187 biweekly · median lifespan extended 27% (mixed background) and 24% (C57BL/6) · paper uses p16Ink4a as the senescent-cell driver, not PUMA; does NOT characterize BCL-2-family anti-apoptotic upregulation in senescent cells; citation retained here for lifespan-extension context only · locally downloaded ↩

Graph View

Backlinks