Ku70/Ku80 (XRCC6/XRCC5 heterodimer)

The Ku heterodimer is the foundational sensor and gatekeeper of non-homologous-end-joining (NHEJ) — the dominant double-strand break (DSB) repair pathway in mammalian cells. As a ring-shaped complex, Ku threads onto DNA ends at DSBs, shields them from nucleolytic resection, and recruits DNA-PKcs to form the DNA-PK holoenzyme. Its functions extend to telomere-end protection, where it prevents illegitimate telomere-to-telomere fusions. Ku-deficient mice display progeroid features, immunodeficiency, and cancer predisposition, establishing the complex as aging-relevant.

Subunit identity

Field	Ku70 (XRCC6)	Ku80 (XRCC5)
UniProt	P12956	P13010
Gene symbol	XRCC6	XRCC5
NCBI Gene	2547	7494
Length	609 aa	732 aa
Mouse ortholog	Xrcc6	Xrcc5

Naming note: This page covers the obligate Ku70/Ku80 heterodimer as a functional unit. Per complex-page convention (CLAUDE.md), the primary catalytic anchor subunit’s UniProt (P12956, Ku70/XRCC6) is used in the top-level uniprot: field; the complex-subunits: list carries both IDs.

Complex architecture

The Ku70/Ku80 heterodimer adopts a crescent-shaped architecture with a central channel that encircles the DNA double helix ¹. Key structural features:

Central channel (pre-formed ring): Fits dsDNA within a cylindrical Van der Waals exclusion volume with a radius of ~11.5 Å (diameter ~23 Å) ¹; the heterodimer loads onto DNA ends by threading, not by wrapping around. This geometry means Ku cannot bind circular DNA — it requires a free DNA terminus.
Central polypeptide rings: Ku70 residues 277–341 and Ku80 residues 267–336 form the two rings that encircle DNA ¹. The broader β-barrel domains span larger regions of each subunit.
N-terminal α/β domains: Ku70 residues 34–250; Ku80 residues 6–238 (13% sequence identity); adopt a Rossman fold-type topology and lie at the periphery of the heterodimer ¹. The VWFA classification of the Ku80 N-terminal domain is from other structural analyses, not Walker 2001 directly. unsourced — VWFA attribution needs citation.
Ku80 C-terminal domain: The ~19 kDa C-terminal domain of Ku80 (absent from crystallographic structure) is required for DNA-PKcs recruitment ¹. The specific EEXXXDDL motif and residue range (720–728) cited in the literature are not stated in Walker 2001. unsourced — EEXXXDDL motif residue range needs independent citation.
Ku70 SAP domain: Located in the C-terminal region of Ku70; proposed DNA-tethering function and contribution to telomere association ¹. Walker 2001 places this domain beginning at residue ~536 with a disordered linker (~539–558); the specific 573–607 range is not directly stated in Walker 2001. unsourced — precise SAP domain boundary needs citation.
Ku70 C-terminal region: Anti-apoptotic; sequesters Bax at mitochondria through direct interaction.

The crystal structure at 2.5 Å resolution (Walker et al. 2001) revealed that DNA-induced conformational change is minimal — Ku binds pre-organized, not induced-fit ¹.

NHEJ initiation mechanism

NHEJ is the primary DSB repair pathway in mammalian cells, operating throughout the cell cycle (dominant in G1) ¹.

Step 1 — DSB sensing: Within seconds of a DSB, Ku70/Ku80 binds both broken ends via its preformed channel. Ku has high affinity for DNA ends (Kd 1.5–4.0 × 10⁻¹⁰ M, i.e., ~0.15–0.4 nM, per ref. 22 in Walker 2001 ¹); the ~400,000 molecules/nucleus abundance figure cited elsewhere is not stated in Walker 2001. unsourced — nuclear abundance figure needs independent citation from a biochemical quantification study.

Step 2 — DNA-PK holoenzyme assembly: Ku80’s C-terminal EEXXXDDL motif recruits DNA-PKcs (PRKDC), forming the ~600 kDa DNA-PK holoenzyme. DNA-PKcs autophosphorylation (at the ABCDE and PQR clusters) regulates end processing and synapsis.

Step 3 — End protection and tethering: Ku physically shields DNA ends from exonucleolytic degradation, thereby suppressing homologous recombination by limiting resection. This commits the cell to NHEJ over HR; the balance between pathways is partly regulated by cell-cycle phase and resection factor competition.

Step 4 — Ligation complex recruitment: Ku recruits the XRCC4-LIG4 ligation complex (and co-factors PAXX, XLF/NHEJ1, APLF) for end ligation. XRCC4 directly contacts Ku80’s VWFA domain.

Step 5 — Ku eviction: After ligation, Ku must be removed from the ligated product (Ku cannot slide off a circularized DNA). This is mediated by ubiquitin-dependent TRIP12 / UBR5 and SPRTN proteases. unsourced — cite the relevant eviction-mechanism paper.

The pathway-level article is non-homologous-end-joining (stub; not yet created). stub

Telomere-end protection

Ku associates with telomeres independently of its canonical NHEJ function and contributes to telomere-end protection in parallel with shelterin ²:

Ku at telomeres prevents fusions: In Ku70-null and Ku80-null mice, chromosome ends undergo end-to-end fusions, indicating that Ku is required to mask telomeric DSB-like structures from aberrant NHEJ-mediated joining. This is paradoxical — Ku normally promotes ligation, yet at telomeres it inhibits it, likely by occluding DNA-PKcs recruitment or by cooperating with shelterin to adopt a non-ligatable end configuration.
Ku80 interaction with telomerase: In human cells, Ku80 has been reported to associate with TERC (the telomerase RNA component), though the mechanistic significance remains contested. contradictory-evidence
Species note: Ku’s role in telomere protection is stronger in mice than in humans; murine telomeres are ~10-fold longer and more Ku-dependent for structural integrity. The relevance to human telomere biology should be interpreted with this caveat.

Dimension	Status
Ku telomere-protection conserved in humans?	partial — telomere fusions in Ku-null mice; human Ku localizes to telomeres but phenotype at human telomere lengths is less severe
Phenotype (telomere fusions) conserved in humans?	partial — observed in human cell lines, severity unclear in vivo
Replicated in humans?	no (in vivo) / in-progress (cell lines)

Mouse genetics: progeroid phenotypes

Loss of either Ku subunit in mice produces a characteristic syndrome that combines immunodeficiency (SCID) with accelerated aging ³ ² ⁴:

Growth retardation: Ku70-null mice maintained body weights at 50–60% of littermate controls at all examined ages ². Ku80-null mice were similarly 40–60% the size of littermate controls ³.
Immunodeficiency: Both Ku70-null and Ku80-null mice fail to complete V(D)J recombination, lacking mature B cells ² ³. Ku70-null mice show a “leaky” T cell defect — DP thymocytes are present but reduced; peripheral T cells are markedly fewer than controls ². This is distinct from the complete T cell block in Ku80-null mice ³.
Lymphomagenesis: Ku70-null mice develop thymic lymphomas at significant frequency — 6 cases in approximately 40 Ku70⁻/⁻ mice, ranging from 2 to 7 months of age; tumors had a CD4⁺CD8⁺ TCRβ⁻ phenotype ². Ku80-null mice, by contrast, were not observed to develop early thymic tumors — Nussenzweig 1996 explicitly noted the absence of such predisposition in Ku80-null animals ³.
Progeroid tissue changes: Ku-null MEFs show premature replicative senescence — Ku70-null MEFs accumulated non-dividing cells faster than controls and stopped growing earlier ². Specific histological progeroid changes (skin atrophy, osteopenia, liver dysfunction) are not reported in either Gu 1997 or Nussenzweig 1996. unsourced — these tissue-level progeroid phenotypes require separate primary citations.
Shortened lifespan: Ku70-null mice die primarily from thymic lymphoma; the premature senescence of MEFs suggests additional aging-related deterioration, but lifespan data are not formally reported in Gu 1997 ².
Emphysema: Bax-induced apoptosis in Ku70-deficient lung leads to emphysema-like pathology, suggesting Ku70’s anti-apoptotic Bax-sequestration function is tissue-protective in lung independently of DNA repair ⁵.

Dimension	Status
Pathway (NHEJ) conserved in humans?	yes — identical molecular machinery
Progeroid phenotype conserved in humans?	partial — Ku mutations are not reported as germline survivable in humans; phenotype inferred from somatic contexts
Replicated in humans?	no — no human germline Ku-null kindred identified

needs-human-replication — the progeroid interpretation of Ku70/Ku80 deficiency rests entirely on mouse knockout data.

Aging relevance: Ku protein decline with age

Ku70 and Ku80 protein levels drop approximately two-fold in replicatively senescent human fibroblasts relative to early-passage cells ⁶. Key observations:

⚠️ The claims in this section derive from Seluanov 2007 (doi:10.1016/j.dnarep.2007.06.010), which could not be retrieved via a local paper archive (2026-05-05). Quantitative claims (two-fold decline, cytoplasmic redistribution) are AI-extracted and unverified against the primary PDF. See Limitations. no-fulltext-access

In senescent cells, the nuclear Ku pool remains stable, but the cytoplasmic Ku fraction (present in young cells) disappears, leaving Ku incapable of responding dynamically to new DSBs ⁶.
Ku in senescent cells is unable to relocate to sites of induced DNA damage, consistent with impaired NHEJ efficiency as a contributor to the accumulation of unrepaired DSBs in aged cells.
Reduced NHEJ efficiency has also been measured functionally in tissues from aged rodents, though the relative contribution of Ku level vs. accessibility vs. post-translational modification is not established. no-mechanism

This decline fits the broader pattern of genomic-instability accumulation with age: impaired Ku function → slower or error-prone DSB repair → increased mutational burden and chromosomal rearrangements.

Dimension	Status
Ku decline with age in humans?	partial — shown in replicatively senescent human cells in vitro; organismal data limited
Replicated in humans (in vivo)?	no — needs-human-replication

Key interactions

DNA-PKcs (PRKDC): Recruited by Ku80’s C-terminal EEXXXDDL motif; forms the DNA-PK holoenzyme; required for NHEJ completion and H2AX phosphorylation (as part of the kinase complex).
XRCC4-LIG4: Ligase complex docked to the Ku80 VWFA domain via XRCC4.
XLF (NHEJ1) / PAXX: Co-factors that stabilize the XRCC4-LIG4-Ku complex at DNA ends.
shelterin: Ku co-occupies telomeres with shelterin components; functional interplay, especially with TRF1/TRF2, regulates the anti-fusion activity at telomeres.
Bax: Ku70 directly binds and sequesters Bax in the cytoplasm via its C-terminal domain, providing a non-DNA-repair anti-apoptotic function.
cgas-sting: Cytoplasmic Ku (present in young, non-senescent cells) binds STING and suppresses cGAS-STING innate immune signaling; loss of cytoplasmic Ku in senescent cells may contribute to chronic cGAS-STING activation and sasp. needs-replication — this interaction is reported but mechanistic evidence in primary aging contexts is limited.
atm: Ku-bound DSBs are recognized by ATM independently; the pathways converge at H2AX and 53BP1 recruitment.

Pharmacology and druggability

Ku70/Ku80 is classified as druggability tier 2 (high-quality probe compounds exist but no approved drug). Rationale:

Cancer context: Ku inhibitors are being explored as radiosensitizers and chemosensitizers in cancer therapy, where disabling NHEJ in tumor cells increases DSB-induced lethality. Several small-molecule Ku inhibitors (e.g., NU7441 targets DNA-PKcs rather than Ku directly; STL127705 binds Ku80 directly) are at probe/preclinical stage.
Aging context: Ku inhibition is not a viable aging intervention given that Ku deficiency accelerates genomic instability and cancer. Therapeutic interest is the inverse: restoring Ku function or activity in aged tissues where levels decline.
No approved drug targets Ku directly. unsourced — specific probe compounds and their on-target validation need primary citations.

Limitations and gaps

#gap/needs-human-replication — Progeroid phenotypes of Ku deficiency are established only in mice; no human germline Ku-null individuals are known.
#gap/needs-human-replication — Ku protein decline with age is shown in replicatively senescent human fibroblasts; in vivo human tissue aging data are sparse.
#gap/no-fulltext-access — Seluanov 2007 (doi:10.1016/j.dnarep.2007.06.010, PMC:PMC2699370) could not be downloaded via a local paper archive (2026-05-05). The two-fold Ku protein decline figure and nuclear/cytoplasmic redistribution claims cannot be verified against the primary source. Recommend manual download and re-verification.
#gap/no-mechanism — Mechanism by which Ku at telomeres prevents fusions while Ku at internal DSBs promotes ligation is not fully resolved.
#gap/no-mechanism — How cytoplasmic Ku loss in senescent cells suppresses or fails to suppress cGAS-STING needs further mechanistic elucidation.
#gap/unsourced — Specific Ku abundance in nucleus (~400,000 molecules/nucleus) needs a primary biochemical quantification citation. Walker 2001 reports Kd 1.5–4.0 × 10⁻¹⁰ M (citing Yoo & Dynan 1999 as ref. 22) — this value is from a secondary citation in Walker, not directly measured in Walker 2001; verify against Yoo & Dynan 1999.
#gap/unsourced — VWFA domain classification for Ku80 N-terminal domain, EEXXXDDL motif residue range (720–728), and precise Ku70 SAP domain boundary (573–607) are not stated in Walker 2001 and require independent primary citations (likely Singleton et al. 1999 or Gell & Jackson 1999 for EEXXXDDL; Aravind & Koonin 2000 for VWFA/SAP).
#gap/unsourced — Ku eviction mechanism after ligation needs its own primary citation.
#gap/needs-canonical-id — GenAge ID for XRCC6/Ku70 was not confirmed; field left null pending lookup at HAGR GenAge database.

Footnotes

doi:10.1038/35088000 · Walker JR, Corpina RA, Goldberg J · Nature 2001 · in-vitro (crystal structure, X-ray 2.5 Å) · model: human Ku70/Ku80 heterodimer · locally available: ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹
doi:10.1016/s1074-7613(00)80386-6 · Gu Y, Seidl KJ, Rathbun GA, Zhu C, Manis JP, van der Stoep N, Davidson L, Cheng HL, Sekiguchi JM, Frank K, Stanhope-Baker P, Schlissel MS, Roth DB, Alt FW · Immunity 1997 · in-vivo (Ku70-null mice) · n not reported (knockout colony; ~40 Ku70⁻/⁻ mice used across tumor study) · model: Mus musculus (mixed 129/Sv × C57BL/6J background) · locally available: ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸
doi:10.1038/382551a0 · Nussenzweig A, Chen C, da Costa Soares V, Sanchez M, Sokol K, Nussenzweig MC, Li GC · Nature 1996 · in-vivo (Ku80-null mice) · n not reported (knockout colony) · model: Mus musculus (mixed 129/Sv × C57BL/6J background, injected into C57BL/6 blastocysts) · locally available: ↩ ↩² ↩³ ↩⁴ ↩⁵
doi:10.1073/pnas.94.15.8076 · Gu Y, Jin S, Gao Y, Weaver DT, Alt FW · PNAS 1997 · in-vitro (Ku70-deficient ES cells) · model: Mus musculus embryonic stem cells · not locally downloaded (pending, OA available) ↩
doi:10.1177/1535370216654587 · Matsuyama S et al. · Experimental Biology and Medicine 2016 · in-vivo (Ku70-null mice) · model: Mus musculus · not locally downloaded (pending) ↩
doi:10.1016/j.dnarep.2007.06.010 · Seluanov A, Danek J, Hause N, Gorbunova V · DNA Repair 2007 · 6(12):1740–1748 · in-vitro (replicatively senescent human fibroblasts) · n not reported · model: human fibroblasts (WI-38 or similar; cell line not independently confirmed — unsourced pending full-text access) · not locally downloadable (green OA, PMC:PMC2699370; a local paper DOI lookup failed 2026-05-05 — cannot verify quantitative claims in Seluanov 2007-derived sections) ↩ ↩²

🧬 Aging Wiki

Explorer

ku70-ku80