🧬 Aging Wiki

caspase-3

Properties1
aliasesCASP3, apopain, CPP32, Yama, SCA-1, caspase-3

11 min read

Caspase-3 (CASP3)

The canonical executioner caspase and primary effector protease of programmed cell death. Once activated by initiator caspases (caspase-9 from the intrinsic pathway; caspase-8 from the extrinsic pathway), caspase-3 cleaves >100 cellular substrates to execute the apoptotic program: DNA fragmentation, membrane blebbing, organelle dismantling, and cytoskeletal collapse. Central to the aging field because apoptosis resistance in senescent cells — and conversely, non-apoptotic caspase-3 activity in proliferating and differentiating cells — complicate simple “more apoptosis = slower aging” framings.

Identity

  • UniProt: P42574 (CASP3_HUMAN; Swiss-Prot, manually reviewed)
  • NCBI Gene: 836
  • HGNC: 1504 (symbol: CASP3)
  • Ensembl: ENSG00000164305
  • Mouse ortholog: Casp3 (one-to-one; ~81% sequence identity)
  • Canonical length: 277 aa (pro-form)
  • Historical synonyms: CPP32 (Fernandes-Alnemri 1994) 1; apopain (Nicholson 1995) 2; Yama (Tewari 1995) 3; SCA-1 (separate cDNA clone)

Structure and zymogen processing

Caspase-3 is synthesized as an inactive zymogen, procaspase-3 (~32 kDa, 277 aa). The pro-form consists of:

RegionResiduesNotes
N-terminal prodomain1–28Two propeptide segments (1–9 and 10–28); short compared with initiator caspases
Large subunit (p17)29–175Contains QACRG active-site motif; Cys163 catalytic residue
Small subunit (p12)176–277His121 completes the catalytic dyad

Activation requires proteolytic cleavage at Asp175 (the large/small subunit junction) and removal of the prodomain. The mature enzyme is an obligate heterotetramer composed of two large (p17) and two small (p12) subunits arranged in a (p17–p12)₂ structure. The active sites are located at the two p17/p12 interfaces 2.

Active site

The caspase fold presents an acyl-enzyme mechanism using a catalytic dyad:

  • Cys163 — nucleophilic cysteine; the QACRG pentapeptide motif surrounding it is conserved across all caspases
  • His121 — general base; positioned by the small subunit

S-nitrosylation of Cys163 (producing S-nitrosocysteine) inhibits caspase-3 activity; this PTM has been proposed as an anti-apoptotic regulatory mechanism in some cell types (#gap/no-mechanism for physiological relevance of S-nitrosylation in aging contexts).

Substrate specificity

Caspase-3 requires Asp at P1 (the residue immediately N-terminal to the scissile bond) and strongly prefers a hydrophobic or acidic residue at P4. The canonical tetrapeptide recognition motif is:

DEVD (Asp–Glu–Val–Asp) — embedded in the PARP-1 cleavage site

More broadly: DEXD with Asp at P1 and P4; contrast with caspase-7 (also DEVD-preferring) and caspase-8/9 (IETD/LEHD). Walsh et al. 2008 demonstrated that caspase-3 and caspase-7 — despite near-identical substrate consensus sequences — cleave substantially different subsets of cellular proteins using immunodepletion and 2D-gel analysis in human Jurkat cell-free extracts: of 20 substrates examined, 12 were preferentially cleaved by caspase-3, while only 1 (cochaperone p23) was more susceptible to caspase-7, implying structural determinants beyond the tetrapeptide 4.

Activation pathways

Intrinsic (mitochondrial) pathway

cytochrome-c released from the mitochondrial intermembrane space via bax/bak-dependent MOMP binds apaf-1 → triggers apoptosome assembly (heptameric APAF-1 wheel) → recruits and autoactivates caspase-9 → activated caspase-9 cleaves procaspase-3 at Asp175 → executioner caspase-3 is released and amplifies the apoptotic signal.

Extrinsic (death receptor) pathway

FasL/TRAIL/TNF → death receptor DISC assembly → caspase-8 autoactivation → caspase-8 cleaves procaspase-3 directly (type I cells: sufficient for apoptosis) or cleaves bid to form tBID → tBID activates bax/bak → MOMP → intrinsic pathway amplification (type II cells) → caspase-9 → caspase-3.

Other activating proteases

  • Caspase-10 — death-receptor pathway initiator; can cleave procaspase-3 in parallel with caspase-8
  • Granzyme B — cytotoxic T-lymphocyte/NK cell serine protease; cleaves at IETD sites and can directly activate procaspase-3, bypassing upstream caspases

Key apoptotic substrates

Once active, caspase-3 cleaves >100 substrates to execute cell death. The highest-confidence canonical substrates:

SubstrateCleavage siteConsequence
PARP-1DEVD↓G between Asp216 and Gly217Inactivates DNA repair; generates 89/24 kDa fragments (marker of apoptosis) 2
ICAD (DFF45)DETD (aa 117) and DAVD (aa 224)Releases CAD/DFF40 nuclease → internucleosomal DNA fragmentation 5 6
Lamin A/B—Nuclear envelope disassembly; nuclear collapse
ROCK1—Constitutive kinase activity → membrane blebbing
Gelsolin—Actin filament severing; cytoskeletal collapse
Procaspase-6, -7—Caspase cascade amplification
PKCδ—Activated truncated kinase amplifies apoptosis

DFF/CAD — the DNA fragmentation effector

Liu et al. 1997 identified the DNA Fragmentation Factor (DFF) as a heterodimer of a 45 kDa and a 40 kDa subunit purified from HeLa cell S-100 cytosol 5. Caspase-3 cleaves the 45 kDa subunit (DFF-45) at two sites — DETD (aa 117) and DAVD (aa 224) — generating 30 kDa and 11 kDa fragments; this releases the active 40 kDa nuclease (DFF-40), which executes internucleosomal DNA cleavage — the ~180 bp ladder seen on apoptotic gel electrophoresis. The cDNA cloning of DFF-40 was incomplete at the time of publication. Enari et al. 1998 independently characterized the same system, naming the components CAD (caspase-activated DNase) and ICAD (inhibitor of CAD); the DFF45/DFF40 and ICAD/CAD nomenclatures refer to the same proteins 6.

DimensionStatusNotes
Pathway conserved in humans?yesDFF45/DFF40 heterodimer conserved; C. elegans CPS-6 is functional homolog
Phenotype conserved in humans?yesDNA laddering observed in human apoptosis
Replicated in humans?yesFundamental cell biology, not model-organism-specific

Regulation

XIAP — the primary endogenous inhibitor

XIAP (X-linked inhibitor of apoptosis) inhibits the mature (processed) caspase-3 heterotetramer via its BIR2 domain, which sterically occludes the active site. XIAP also inhibits caspase-7 (BIR2) and caspase-9 (BIR3). This is a direct stoichiometric interaction (7 independent experimental observations; IntAct EBI-524064/EBI-517127). unsourced for quantitative Ki values against caspase-3 in vivo context.

SMAC/DIABLO (released from mitochondria during MOMP) antagonizes XIAP by competing for the BIR2/BIR3 binding grooves via its IAP-binding tetrapeptide motif (AVPI), thereby de-repressing caspase-3 and caspase-9.

HSP90 / CHIP

Procaspase-3 is a client of HSP90; inhibition of HSP90 promotes procaspase-3 degradation via the E3 ligase CHIP. This constitutes a second anti-apoptotic checkpoint in some cancer cell lines. needs-replication for relevance in normal aging tissues.

Phosphorylation

Ser26 phosphorylation (UniProt annotated) has been proposed to modulate activity, but the physiological kinase and functional consequence are not well characterized. no-mechanism

Knockout phenotype

Kuida et al. 1996 generated Casp3-null mice on the C57BL/6 background 7:

  • ~50% perinatal lethality
  • Surviving mice showed brain malformation — supernumerary cells, ectopic masses in the cortex and striatum, reflecting impaired developmental apoptosis
  • Brain weight ~30–40% greater than controls due to excess cells
  • No gross abnormalities outside the CNS
  • The phenotype is notably strain-dependent: on mixed or other backgrounds, Casp3-null mice survive normally and are fertile, with more subtle phenotypes
DimensionStatusNotes
Pathway conserved in humans?yesHuman CASP3 fully functional; identical executioner role
Phenotype conserved in humans?partialHumans with caspase-3 loss-of-function not well-characterized; single-gene CNS malformation parallels exist
Replicated in humans?noNo germline CASP3 KO humans characterized needs-human-replication

Discovery

Three independent groups converged on caspase-3 in 1994–1995:

  1. Fernandes-Alnemri et al. 1994 cloned CPP32 from human Jurkat cells as a novel ICE/CED-3 homolog; showed it is activated during apoptosis 1.
  2. Nicholson et al. 1995 identified apopain as the ICE/CED-3 protease required for mammalian apoptosis, using biochemical purification from apoptotic THP-1 human monocytic leukemia cells; showed it cleaves PARP between Asp216 and Gly217; showed potent inhibition by Ac-DEVD-CHO (K_i < 1 nM) and pan-caspase inhibitor zVAD-fmk 2.
  3. Tewari et al. 1995 independently cloned CPP32β (Yama); demonstrated PARP cleavage and CrmA inhibition, linking the protease to known apoptosis regulators 3.

The convergence of three independent molecular identifications with prior biochemical activity (the ~32 kDa PARP-cleaving protease in apoptotic cells) established caspase-3 as the canonical executioner caspase.

Role in aging and senescence

Apoptosis resistance of senescent cells

Senescent cells (cellular-senescence) are characteristically apoptosis-resistant despite elevated pro-apoptotic signaling. This is explained primarily by upregulation of anti-apoptotic bcl-xl, bcl-2, and mcl-1 upstream of caspase-3, which suppress bax/bak activation and prevent cytochrome-c release. Caspase-3 itself is not generally upregulated in senescent cells — the block is upstream at the MOMP level. The selective killing of senescent cells by senolytics (e.g., navitoclax, ABT-263) ultimately depends on de-suppressing bax/bak → MOMP → caspase-9 → caspase-3 → apoptotic execution.

Non-apoptotic caspase-3 functions

At sub-apoptotic activation levels, caspase-3 has documented non-apoptotic functions:

  • Cell proliferation — partial caspase-3 activation may promote mitogenic signaling in some contexts
  • Differentiation — muscle differentiation, platelet biogenesis, and lens fiber maturation have caspase-3 roles
  • Synaptic plasticity — LTD in neurons involves local, limited caspase-3 activity without cell death

The mechanisms preventing sub-apoptotic caspase-3 activity from triggering full apoptotic commitment are incompletely understood. no-mechanism

The “caspase-3 paradox” in cancer and aging

High caspase-3 expression has been reported in some tumors without commensurate apoptosis, possibly reflecting either substrate mis-targeting or concurrent caspase-3 non-apoptotic roles in proliferation. This “paradox” is relevant to aging biology because the relationship between apoptotic competence and tissue homeostasis is not monotonic — insufficient apoptosis favors senescent cell accumulation; excessive apoptosis depletes stem cell pools. The net effect on aging rate likely depends on tissue, life stage, and stress context. no-mechanism needs-replication

Pathway membership

Key interactors

  • apaf-1 — apoptosome scaffold that activates caspase-9, which then activates caspase-3
  • caspase-9 — initiator (intrinsic pathway); direct activator of procaspase-3
  • caspase-8 — initiator (extrinsic pathway); direct activator of procaspase-3 in type I cells
  • xiap — BIR2-domain inhibitor of processed caspase-3
  • smac-diablo — XIAP antagonist; de-represses caspase-3
  • bid — tBID links extrinsic to intrinsic pathway upstream of caspase-3

Pharmacology

No clinically approved caspase-3 inhibitors or activators exist specifically for aging. Research tools:

  • zVAD-fmk — irreversible pan-caspase inhibitor; standard in vitro probe; not clinically useful
  • zDEVD-fmk — caspase-3/-7 selective inhibitor; research use
  • DEVD-AFC / DEVD-AMC — fluorogenic substrates for measuring caspase-3 activity; K_m (Ac-DEVD-AMC) = 9.7 ± 1.0 ÎĽM for purified CPP32/apopain 2
  • Senolytics (navitoclax, dasatinib+quercetin) target upstream BCL-2 family proteins, ultimately causing caspase-3 activation in senescent cells as a desired downstream consequence — caspase-3 is the effector, not the therapeutic target

Limitations and gaps

  • needs-human-replication — Casp3-null developmental phenotype (Kuida 1996) is in mouse; human CASP3 loss-of-function not characterized clinically
  • no-mechanism — sub-apoptotic caspase-3 activation thresholds and the signals that prevent commitment to full apoptosis are incompletely understood
  • no-mechanism — S-nitrosylation of Cys163 as a physiological aging-context regulator needs characterization
  • no-mechanism — Ser26 phosphorylation: kinase and functional consequence unknown
  • needs-replication — “caspase-3 paradox” in tumors: whether non-apoptotic caspase-3 activity has a causal role in tumorigenesis vs aging accumulation needs systematic investigation
  • unsourced — quantitative Ki values for XIAP BIR2 inhibition of active caspase-3 in cellular context
  • Caspase-3 vs caspase-7 redundancy in apoptotic execution in vivo: Walsh 2008 shows distinct substrate sets in cell-free systems (caspase-3 more promiscuous; 12/20 substrates preferentially cleaved by casp-3 vs 1/20 by casp-7) but physiological consequences of each in aging contexts are not delineated 4 needs-replication

Footnotes

Footnotes

  1. doi:10.1016/s0021-9258(18)47344-9 · Fernandes-Alnemri, Litwack, Alnemri · J Biol Chem 1994 · in-vitro (Jurkat cell lysates) · model: human Jurkat cells · CPP32 original cloning; ICE/CED-3 homology; activation during apoptosis. Archive: pending download. ↩ ↩2

  2. doi:10.1038/376037a0 · Nicholson et al. · Nature 1995 · in-vitro / biochemical purification · model: THP-1 human monocytic leukemia cells · Identifies apopain (= CPP32) as the ICE/CED-3 protease required for mammalian apoptosis; purified from THP-1 cytosol as two subunits (M_r 17K and 12K); cleaves PARP between Asp216–Gly217; K_m (Ac-DEVD-AMC) = 9.7 ± 1.0 μM; K_i (Ac-DEVD-CHO) < 1 nM. Archive: local PDF verified. ↩ ↩2 ↩3 ↩4 ↩5

  3. doi:10.1016/0092-8674(95)90541-3 · Tewari et al. · Cell 1995 · in-vitro · model: human cells · Yama/CPP32β; PARP cleavage; CrmA inhibition. Archive: pending download. Note: the DOI 10.1016/0092-8674(95)90426-3 given in the seeding brief is a BUG-2 mismatch — that DOI resolves to a retinotectal projection paper (Mek4/ELF-1). Correct DOI confirmed via PubMed PMID 7774019. ↩ ↩2

  4. doi:10.1073/pnas.0707715105 · Walsh, Cullen, Sheridan, Lüthi, Gerner, Martin · PNAS 105:12815–12819, September 2, 2008 · in-vitro/biochemical · model: human Jurkat cell-free extracts + recombinant purified human caspase-3 and caspase-7 · Method: immunodepletion of endogenous caspases + 2D-gel analysis + immunoblotting (not SILAC) · Of 20 substrates examined, 12 were preferentially cleaved by caspase-3 vs only 1 (cochaperone p23) by caspase-7; caspase-3 is the principal apoptosis-associated effector caspase. Archive: local PDF verified (downloaded 2026-05-04). Note: DOI 10.1073/pnas.0805089105 in seeding brief was BUG-2 mismatch (plant fatty acid dehydratase paper); correct DOI confirmed via PMID 18723680. ↩ ↩2

  5. doi:10.1016/s0092-8674(00)80197-x · Liu, Zou, Slaughter, Wang · Cell 89:175–184, April 18, 1997 · in-vitro · model: HeLa cell S-100 cytosol; hamster liver nuclei (fragmentation assay) · DFF purified as 45 kDa + 40 kDa heterodimer from HeLa S-100; caspase-3 cleaves DFF-45 at DETD (aa 117) and DAVD (aa 224) (not DEVD); releases active DFF-40 nuclease; DFF-40 cDNA cloning incomplete at time of publication. Year confirmed as 1997 against PDF. Archive: local PDF verified. ↩ ↩2

  6. doi:10.1038/34112 · Enari et al. · Nature 1998 · in-vitro · model: FL5.12 murine cells + cell-free systems · CAD/ICAD nuclease cascade; independent identification of same system as Liu 1997; ICAD cloning. Archive: not_oa (no local copy). no-fulltext-access ↩ ↩2

  7. doi:10.1038/384368a0 · Kuida et al. · Nature 1996 · in-vivo (mouse, targeted KO) · model: C57BL/6 Casp3-null mice · ~50% perinatal lethality; brain malformation with excess supernumerary cells; strain-dependent phenotype. Archive: not_oa (no local copy). no-fulltext-access ↩

Graph View

Backlinks