Caspase-9 (CASP9)

The canonical initiator caspase of the intrinsic (mitochondrial) apoptosis pathway. Caspase-9 sits at the convergence point where mitochondrial outer membrane permeabilization (MOMP) — mediated by bak and bax — translates into irreversible cell-death commitment. It is not simply a protease: it functions as an allosteric signaling platform whose activity is inseparable from the apaf-1 apoptosome scaffold on which it assembles.

Identity

UniProt: P55211 (CASP9_HUMAN)
NCBI Gene: 842
HGNC symbol: CASP9
Mouse ortholog: Casp9 (functional conservation confirmed by rescue experiments)
Aliases: ICE-LAP6 (Duan et al. 1996 original cloning), Mch-6, APAF-3 (Srinivasula et al. 1998 identity assignment)
Canonical length: 416 amino acids (procaspase-9; isoform 9α, full-length)

Isoforms

Isoform	Length	Activity	Notes
9α (9L)	416 aa	Pro-apoptotic	Full-length; canonical
9β (9S)	~115 aa	Dominant-negative	Lacks residues 140–289 (p20 catalytic region); competes for CARD-CARD interaction with APAF-1, inhibiting apoptosome assembly
9γ	CARD only	Endogenous inhibitor	Generated by alternative splicing; further characterization limited needs-replication

The 9α/9β ratio is regulated by splicing factors and shifts in tumors toward the dominant-negative 9β isoform, providing an alternative resistance mechanism to classical anti-apoptotic protein overexpression. needs-human-replication

Domain structure and active site

Procaspase-9 is organised as a single polypeptide (416 aa) that is cleaved post-translationally but does not require cleavage for activity:

CARD (caspase activation and recruitment domain): Residues 1–92. Six-helix-bundle fold. Mediates homotypic CARD–CARD interaction with apaf-1 for apoptosome recruitment. Not directly catalytic.
p20 (large subunit): Residues ~140–315. Contains the catalytic cysteine at Cys287 (UniProt P55211 numbering; active site; nucleophilic attack on LEHD↓X substrate). Note: Renatus et al. 2001 use the caspase-1 numbering convention and refer to this residue as Cys-285 — the two designations refer to the same position ¹.
Linker / auto-cleavage site: Asp315 — the primary autocleavage site (between p20 and p10). Cleavage at Asp315 by caspase-9 itself (or by granzyme B) separates the large and small subunits but is not required for catalytic activity — apoptosome-bound procaspase-9 adopts an active conformation without proteolytic processing ¹.
p10 (small subunit): Residues ~330–416. Contributes to the substrate-binding groove together with p20; contains the second catalytic His (histidine in the catalytic dyad).

Substrate specificity: Canonical recognition sequence LEHD↓X (Leu-Glu-His-Asp; cleavage after the Asp). Primary physiological substrates are procaspase-3 (at IETD↓Asp175) and procaspase-7 (similar site), generating the active executioner caspases that commit the cell to death.

Apoptosome assembly and activation

The sequence of events from cytochrome-c release to caspase-9 activation is one of the best-characterised multi-step apoptotic signalling modules:

MOMP — bax/bak permeabilise the outer mitochondrial membrane, releasing cytochrome-c (and smac-diablo) from the intermembrane space into the cytosol.
APAF-1 activation — Cytochrome c binds the WD40 propeller domain of apaf-1, triggering ATP/dATP exchange at the NBD (nucleotide-binding domain) of the NB-ARC region. The resulting dATP-loaded APAF-1 monomer undergoes conformational change that exposes the CARD domain ².
Oligomerisation — Seven activated APAF-1 monomers oligomerise into the characteristic 7-spoke wheel (“apoptosome”) ³. needs-replication — the 7-spoke, ~280 Å cryo-EM architecture was established by subsequent structural work (Acehan et al. 2002 Mol Cell 9:423-432), not by Li 1997.
Procaspase-9 recruitment — Seven procaspase-9 zymogens bind via CARD–CARD interactions to the central hub of the apoptosome. The local concentration of procaspase-9 on the platform drives activation by the induced proximity model ⁴: forced dimerisation of two caspase-9 molecules — regardless of processing state — generates a catalytically competent active site. This is mechanistically distinct from the “zymogenicity” model in which cleavage is the activating event.
Apoptosome-bound activity — Unlike downstream executioner caspases-3/-7, active caspase-9 remains physically bound to the apoptosome and does not undergo substantial cytosolic release. The apoptosome-caspase-9 holoenzyme is the operative species in vivo ¹. This tethering limits the spatial spread of caspase-9 activity and confines the amplification signal to an organised platform.

Dimension	Status
Apoptosome mechanism conserved in humans?	yes — human APAF-1 and caspase-9 reconstitute functional apoptosome in vitro
Executioner caspase activation conserved?	yes
Replicated in humans (loss-of-function)?	yes — Casp9-null embryos phenocopy Apaf1-null; equivalent pathway in human cancers

Genetics and knockout phenotype

Mice homozygous null for Casp9 were independently described by two groups in 1998:

Hakem et al. 1998 — Casp9-null embryos die perinatally (collected up to E16.5; majority fail to survive) due to massive brain malformation (protrusion of brain mass, stenosis of ventricles, heterotopias, and interruption of the telencephalic wall — arising from failure to execute developmental apoptosis in the neuroepithelium) ⁵. ES cells and MEFs also showed broad resistance to apoptotic stimuli (UV, gamma-irradiation, etoposide, adriamycin). The neural phenotype parallels that of Apaf1-null and Casp3-null mice, consistent with the linear order: Apaf-1 → Caspase-9 → Caspase-3 in the intrinsic pathway.
Kuida et al. 1998 — Parallel independently-derived Casp9-null line confirmed perinatal lethality with brain overgrowth (onset of brain malformation visible at ~E12.5; most homozygotes die before P3); also demonstrated that cytochrome c–activated caspase-9 processing is absent and downstream caspase-3 activation is severely impaired ⁶. Critically, apoptosis in response to etoposide, dexamethasone, and gamma-irradiation (intrinsic triggers) was abolished in thymocytes, while Fas-mediated (extrinsic) apoptosis was intact, establishing pathway selectivity.

Both knockouts confirmed that caspase-9 is non-redundant for intrinsic apoptosis during development and is not substituted by other initiator caspases. needs-human-replication — equivalent human loss-of-function is not observed as a germline variant; only somatic downregulation in tumours.

Regulation

XIAP — direct inhibitor

xiap (X-linked inhibitor of apoptosis) directly inhibits caspase-9 via its BIR3 domain, which docks onto the neo-N-terminus exposed by Asp315 cleavage. XIAP BIR3 sterically occludes the caspase-9 dimerisation interface, preventing homodimerisation and thus suppressing activity. This is a post-processing regulation mechanism distinct from XIAP’s separate BIR2-mediated inhibition of caspase-3.

Relevance to aging: XIAP levels are reported to be elevated in senescent cells, contributing to the apoptosis resistance that defines the senescent state. XIAP accumulation is one mechanism by which senescent cells tolerate high levels of pro-apoptotic signalling [^gap_xiap_senescence]. unsourced — specific primary citation needed for XIAP elevation in human senescent cells.

SMAC/DIABLO — antagonist of XIAP

smac-diablo is co-released with cytochrome-c from the intermembrane space during MOMP. The N-terminal AVPI tetrapeptide of mature SMAC/DIABLO binds XIAP BIR3 with higher affinity than caspase-9, displacing XIAP and de-repressing caspase-9 activity. SMAC/DIABLO thereby functions as a rheostat that sets the threshold for apoptosome-dependent cell death.

AKT phosphorylation — inhibitory

akt (PKB) directly phosphorylates procaspase-9 at Ser196 (human), impairing cleavage at Asp315 and attenuating catalytic activity ⁷. This phosphorylation represents a direct survival signal from the pi3k-akt-pathway into the core apoptosis machinery, coupling growth-factor availability to intrinsic cell death threshold. Relevant in aging contexts where chronic AKT activation (e.g., in response to nutrient excess or oncogenic signalling) contributes to apoptosis resistance and senescent cell accumulation.

Additional regulatory phosphorylations documented by large-scale proteomics: Thr125 (by MAPK1/ERK2; inhibitory; links MAPK survival signalling to caspase-9 threshold), Ser302, Ser307, Ser310 (functional significance less well characterised). needs-replication for these sites.

Summary of regulatory inputs

Regulator	Site/mechanism	Effect on caspase-9	Pathway connection
APAF-1 apoptosome	CARD–CARD; proximity-induced dimerisation	Activating	Intrinsic apoptosis
XIAP BIR3	Binds post-Asp315 neo-N-term	Inhibitory	IAP survival signalling
SMAC/DIABLO	Displaces XIAP from BIR3	De-repressing	IMS release upon MOMP
AKT	Ser196 phosphorylation	Inhibitory	PI3K/AKT survival
ERK2/MAPK1	Thr125 phosphorylation	Inhibitory	MAPK survival

Role in aging and cellular senescence

Caspase-9 is mechanistically upstream of the apoptotic arm of the senescence machinery. Its relevance to aging operates at two levels:

1. Apoptosis resistance in senescent cells

Senescent cells are characterised by paradoxical resistance to apoptosis despite high expression of bax, bim, and other pro-apoptotic factors [^apoptosis_pathway_resistance]. This resistance is at least partly attributable to elevation of anti-apoptotic proteins (BCL-2, BCL-xL, MCL-1, XIAP) that suppress MOMP upstream of or coincident with caspase-9 activation. The net result is that the apoptosome is primed but not triggered — cytochrome-c release is the gated bottleneck, not caspase-9 activity itself.

From a senolytic strategy perspective, caspase-9 is therefore a secondary target: eliminating senescent cells by restoring pro-apoptotic signalling (via BH3 mimetics targeting BCL-xL or BCL-2) ultimately depends on caspase-9 to execute the death signal once MOMP is restored. This makes caspase-9 competence in aged cells a prerequisite for senolytic efficacy. unsourced — direct measurement of apoptosome function in primary human senescent cells is lacking; most evidence is cell-line based.

2. Developmental apoptosis and tissue homeostasis

The embryonic lethality of Casp9-null mice establishes that developmental apoptosis — which sculpts the nervous system, removes interdigital webbing, and eliminates autoreactive lymphocytes — is entirely caspase-9-dependent. Age-related deficits in tissue clearance (excess cells, protein aggregates, mis-specified cells) may involve declining apoptosome competence, though the causal direction is not established. no-mechanism

Pathway membership

apoptosis-pathway — intrinsic branch; downstream of MOMP, upstream of executioner caspases
cellular-senescence — apoptosis resistance in senescent cells; senolytic target dependency

Key interactors

apaf-1 — apoptosome scaffold; CARD–CARD recruitment
cytochrome-c — APAF-1 activator; indirect co-activator
bax / bak — upstream MOMP executors; set cytochrome-c release threshold
caspase-3 / caspase-7 — direct substrates; executioner phase
xiap — BIR3 inhibitor (direct binding post-Asp315 cleavage)
smac-diablo — XIAP antagonist (indirect de-repressor of caspase-9)
akt — Ser196 kinase; survival signalling link

Pharmacology and therapeutic context

No clinical-stage agent directly targets caspase-9. Indirect modulation occurs via:

BH3 mimetics (navitoclax, venetoclax, A1331852) — restore MOMP → liberate caspase-9 from sequestration → senolytic or anti-tumour effect. Caspase-9 competence is required for these agents to work.
SMAC mimetics (e.g., birinapant, LCL161) — antagonise XIAP → de-repress caspase-9/caspase-3 axis. In clinical trials for cancer; not yet aging-focused.
AKT inhibitors — remove Ser196 inhibitory phosphorylation; lowers apoptosis threshold in cells with hyperactive PI3K/AKT.

For senolytic applications, the critical question is whether aged and senescent cells retain functional apoptosome competence (cytochrome-c → APAF-1 oligomerisation → caspase-9 activation) downstream of the MOMP bottleneck. Evidence is incomplete. no-mechanism

Limitations and open questions

Apoptosome competence in aged cells: No systematic study has measured apoptosome assembly or caspase-9 auto-processing kinetics in primary aged (non-senescent) human cells. Whether aging per se impairs apoptosome function is unknown. unsourced
XIAP in human senescent cells: The claim that XIAP accumulates in senescent cells to resist caspase-9-dependent killing is widely cited but lacks well-powered primary human data. needs-replication
9β isoform in aging: Dominant-negative 9β isoform ratio changes with age have not been characterised in aged tissues. unsourced
Cross-talk with necroptosis / pyroptosis: Caspase-9 suppresses caspase-1 activation in some contexts; the interaction between the apoptosome and the nlrp3-inflammasome inflammasome in aged macrophages is not mechanistically resolved. no-mechanism

Footnotes

doi:10.1073/pnas.231465798 · Renatus M et al. 2001 (PNAS 98:14250-14255) · in-vitro (structural/biochemical) · model: recombinant full-length and ΔCARD caspase-9 (E. coli) + gel filtration + crystal structure at 2.8 Å · dimer-formation drives activation; proteolysis is neither sufficient nor necessary; at physiological concentrations caspase-9 is a monomer (Kd in micromolar range); apoptosome raises local concentration above Kd to drive dimerisation; uses caspase-1 numbering convention (active-site Cys = Cys-285 in their numbering = Cys-287 in UniProt P55211 numbering) · local PDF available ↩ ↩² ↩³
doi:10.1016/s0092-8674(00)80434-1 · Li P et al. 1997 (Cell 91:479-489) · in-vitro (reconstituted system) · model: HeLa S-100 cytosolic extract + purified recombinant APAF-1, caspase-9 (APAF-3), caspase-3 + dATP · APAF-3 identity as caspase-9 established; cytochrome c + dATP + APAF-1 + procaspase-9 → caspase-3 activation reconstituted · local PDF available ↩
The 7-spoke wheel apoptosome architecture (~280 Å) was established by cryo-EM structural work subsequent to Li 1997 — see Acehan D et al. 2002 (Mol Cell 9:423-432) and Riedl SJ et al. 2005 (Nature 434:926-933). Li 1997 demonstrated functional reconstitution of the complex but did not resolve its architecture. needs-replication — apoptosome structure reference not yet extracted to a study page. ↩
doi:10.1073/pnas.96.20.10964 · Salvesen GS & Dixit VM 1999 (PNAS 96:10964) · review/conceptual · induced-proximity model of initiator caspase activation · 928 citations · not_oa (no local PDF) ↩
doi:10.1016/s0092-8674(00)81477-4 · Hakem R et al. 1998 (Cell 94:339-352) · in-vivo · model: Casp9-null mice (129/C57BL/6 background) + ES cells + MEFs + thymocytes · perinatal lethal; brain malformation (protrusion, ventricular stenosis, heterotopias); ES cells/MEFs resistant to UV, gamma-irradiation, etoposide, adriamycin; dexamethasone/gamma-irradiation resistance in thymocytes; Fas-apoptosis intact · local PDF available ↩
doi:10.1016/s0092-8674(00)81476-2 · Kuida K et al. 1998 (Cell 94:325-337) · in-vivo · model: Casp9-null mice (129/C57BL/6 and 129/CD1 backgrounds; phenotype similar in both) + thymocytes · perinatal lethal; brain malformation onset ~E12.5; <3% of pups survive to weaning; cytochrome c–mediated caspase-3 activation absent in embryonic lysates; restored by in-vitro-translated CASP9; thymocytes resistant to etoposide, dexamethasone, gamma-irradiation; Fas (extrinsic) apoptosis intact · local PDF available ↩
doi:10.1126/science.282.5392.1318 · Cardone MH et al. 1998 (Science 282:1318-1321) · in-vitro + in-vivo · model: primary thymocytes + recombinant proteins · AKT phosphorylates Ser196; inhibits cleavage/activity · 3,110 citations · not_oa (no local PDF) ↩

🧬 Aging Wiki

Explorer

caspase-9