🧬 Aging Wiki

dna-pkcs

Properties1
aliasesDNA-PKcs, DNA-PK catalytic subunit, PRKDC, HYRC, HYRC1

9 min read

DNA-PKcs (PRKDC)

The catalytic subunit of the DNA-dependent protein kinase (DNA-PK) holoenzyme — a 4,128-amino-acid PIKK-family serine/threonine kinase that serves as the primary molecular sensor and orchestrator of NHEJ at DNA double-strand breaks (DSBs). Beyond its canonical DNA repair role, DNA-PKcs suppresses AMPK signaling in aging muscle via HSP90α phosphorylation, coupling accumulating DNA damage to metabolic decline 1.

Identity

  • UniProt: P78527 (PRKDC_HUMAN)
  • NCBI Gene: 5591
  • HGNC symbol: PRKDC (aliases: HYRC, HYRC1)
  • Ensembl: ENSG00000253729
  • Mouse ortholog: Prkdc (high conservation; Scid mouse harbors a nonsense mutation in this gene — see below)
  • Length: 4,128 amino acids (one of the largest human proteins; MW ~470 kDa)
  • GenAge entry: not found no-genage-entry

PIKK family context

DNA-PKcs belongs to the PIKK (PI3K-related kinase) family alongside ATM, ATR, mTOR (see mtor), SMG1, and TRRAP. All PIKKs share:

  • C-terminal PI3K-catalytic domain (kinase activity)
  • FAT and FATC flanking domains (regulatory)
  • Extended N-terminal HEAT/HEAT-like repeats (scaffold for protein-protein interactions)

Despite structural similarity to lipid kinases, PIKKs phosphorylate proteins exclusively. DNA-PKcs is distinguished by its requirement for DNA-end binding (via the Ku70–Ku80 heterodimer) for kinase activation.

Domain architecture

DomainResidues (approx.)Function
HEAT repeats (N-lobe)1–892Scaffold; Ku heterodimer docking
Middle HEAT repeats893–2,801Conformational flexibility at DSB ends
FAT domain2,906–3,539Regulatory; couples DNA-end status to catalytic domain
PI3K catalytic domain3,722–4,053Serine/threonine kinase activity
FATC domain4,096–4,128Required for full kinase activity; terminal integrity sensor

A 2017 X-ray crystallography study (selenomethionine-labeled DNA-PKcs complexed with KU80ct₁₉₄; 4.3 Å resolution) defined three large structural units: the N-terminal region (HEAT repeats, residues 1–892), the Circular Cradle (HEAT repeats, residues 893–2801), and the C-terminal Head (residues 2802–4128, containing FAT, FRB, kinase, and FATC domains) 2. needs-replication — the 4.3 Å resolution leaves side-chain-level detail incomplete; higher-resolution structures at atomic detail are in progress as of 2025.

DNA-PK holoenzyme and NHEJ mechanism

DNA-PKcs does not bind DSB ends directly. Instead, the Ku70–Ku80 heterodimer (ku70-ku80) threads onto DNA ends with high affinity, then recruits DNA-PKcs to form the DNA-PK holoenzyme. This three-component assembly (Ku70 + Ku80 + DNA-PKcs) is essential for NHEJ 3.

Activation cascade:

  1. DSB generated (ionizing radiation, replication stress, programmed breaks in V(D)J recombination)
  2. Ku70–Ku80 binds blunt or 1–4 nt overhang DNA ends within seconds
  3. DNA-PKcs recruited; Ku–DNA-PKcs contact allosterically activates kinase activity (~10-fold stimulation)
  4. DNA-PKcs autophosphorylates at the ABCDE cluster (Thr2609, Thr2638, Thr2647) and the PQR cluster (Ser2056) — autophosphorylation regulates DNA-end access and end-processing fidelity
  5. Artemis (DCLRE1C) is phosphorylated and activated → endonuclease trims 3’ overhangs and opens hairpins (critical for V(D)J coding joints)
  6. XRCC4–LIG4–XLF complex ligates the processed ends; XRCC4 is phosphorylated by DNA-PKcs to regulate the ligation step
  7. DNA-PKcs dissociates following ligation; autophosphorylation may trigger this release

NHEJ is the dominant DSB repair pathway in G1 phase and is the only pathway capable of repairing the programmed DSBs introduced during V(D)J recombination in developing lymphocytes.

V(D)J recombination and the SCID mouse

The requirement for NHEJ in V(D)J recombination is illustrated starkly by the Scid (severe combined immunodeficiency) mouse, first described by Bosma et al. in 1983 4. Scid mice carry a nonsense mutation in the Prkdc gene (a premature stop codon near the C-terminus, removing the kinase domain). Consequences:

  • Complete absence of mature B and T lymphocytes — V(D)J recombination fails at the coding joint step; antigen-receptor gene rearrangements are not completed
  • Lymphoid progenitors are present but arrested before productive rearrangement
  • Scid mice are essentially immunologically naive — they cannot generate adaptive immune responses

Human equivalents: Loss-of-function variants in PRKDC cause Immunodeficiency 26 (IMD26), a rare form of SCID with combined T- and B-cell deficiency plus neurological abnormalities (DNA-PKcs has non-immune roles in neurons).

Practical use: Because Scid mice lack adaptive immunity, they accept human xenografts and are a cornerstone model for human tumor biology, human hematopoietic engraftment studies, and CAR-T cell preclinical testing. The SCID–Hu and NSG (NOD-SCID-IL2Rg-null) models are refinements.

DimensionStatusNotes
Pathway conserved in humans?yesNHEJ, V(D)J, and DNA-PKcs kinase function are conserved
Phenotype conserved in humans?yesPRKDC loss-of-function causes human SCID (IMD26)
Replicated in humans?yesMultiple IMD26 kindreds; also replicated by chemical inhibition studies in human cell lines

Metabolic and aging-relevant role

Beyond NHEJ, an unexpected aging-relevant function of DNA-PKcs was identified by Park et al. (2017) 1: DNA-PKcs suppresses AMPK activity through a non-canonical, cytoplasmic signaling axis.

Mechanism:

  1. DSB load in skeletal muscle increases with aging (oxidative damage, replication stress in satellite cells, persistent inflammation)
  2. Elevated DNA-PKcs activity phosphorylates HSP90α at Thr-5 and Thr-7
  3. Phospho-HSP90α loses its chaperoning support for LKB1 (STK11) — the upstream kinase that activates AMPK
  4. AMPK activity falls → reduced mitochondrial biogenesis, impaired fatty acid oxidation, decreased exercise capacity
  5. Net metabolic consequence: weight gain, mitochondrial dysfunction, reduced physical fitness

Evidence (mouse):

  • Middle-aged mice treated with DNA-PKcs inhibitor NU7441 (delivered orally in HFD food pellets; specific mg/kg dose not stated in Park 2017 — the previous “40 mg/kg/day oral gavage” wording was fabricated and was corrected R32c via the DNA-PKcs-inhibitors verifier) for up to 12 weeks showed preserved AMPK activity, improved mitochondrial function, and were protected from weight gain on a high-fat diet 1
  • Genetic reduction of DNA-PKcs activity via two independent models also protected against metabolic decline: SCID mice (B6.CB17-Prkdc^scid/SzJ; leaky nonsense mutation) had higher p-AMPK and more mitochondria in middle-aged skeletal muscle; muscle-specific conditional DNA-PKcs knockout (MDPKO) mice showed increased p-AMPK and p-ACC at 17 months vs. controls 1
  • The effect was cell-autonomous and observed in both mouse (C2C12) myotubes and human primary myotubes in culture
DimensionStatusNotes
Pathway conserved in humans?yesHSP90α, LKB1, and AMPK are highly conserved; DNA-PKcs kinase domain 85%+ conserved
Phenotype conserved in humans?partialHuman myotube data in culture; no in-vivo human aging trial yet
Replicated in humans?noPark 2017 is single-study; no independent replication in aged humans needs-human-replication

needs-replication — The DNA-PKcs → HSP90α → LKB1 → AMPK axis requires independent replication in a separate lab/model system.

Aging relevance summary

DNA-PKcs connects to aging biology via two distinct mechanisms:

  1. Accumulation of unrepaired DSBs — NHEJ fidelity and capacity decline with age in post-mitotic tissues; persistent DSBs activate chronic DDR signaling that can drive SASP and senescence (via ATM-mediated p53/p21 induction). DNA-PKcs is essential for efficient DSB resolution; impaired NHEJ amplifies the unrepaired-DSB burden. no-mechanism — The age-specific mechanism of NHEJ decline in post-mitotic tissues (transcriptional downregulation of NHEJ components vs. post-translational modifications) is not fully resolved.
  2. Metabolic suppression of AMPK — DNA-PKcs activity rises in aging skeletal muscle, actively suppressing AMPK; this is a non-genomic aging driver that could be pharmacologically targeted independently of DNA repair.

These two mechanisms are not mutually exclusive: accumulating DSBs (mechanism 1) amplify DNA-PKcs activity, which suppresses AMPK (mechanism 2), creating a potential feed-forward loop.

Therapeutic angle — DNA-PK inhibitors

DNA-PKcs is a druggability tier 1 target (clinical-stage kinase inhibitors exist). The therapeutic rationale has historically been oncological (radiosensitization) but is being explored for metabolic-aging indications.

Clinical inhibitors:

CompoundClassMechanismStagePrimary indication
Peposertib (M3814)ATP-competitiveBlocks DNA-PKcs catalytic domainPhase 1/2Solid tumors (radiosensitization)
AZD7648ATP-competitiveSelective DNA-PKcs inhibitorPhase 1Hematologic malignancies + solid tumors

Peposertib phase 1 data (oncology):

  • Phase Ib in locally advanced rectal cancer (combined with chemoradiation): acceptable safety, dose-dependent target engagement 5
  • Phase 1 in head and neck tumors (combined with radiation ± cisplatin): established RP2D 6

Aging/metabolic angle: Park et al. (2017) used NU7441 (a selective ATP-competitive DNA-PKcs inhibitor, structurally distinct from peposertib) delivered orally in HFD food pellets (specific mg/kg dose not stated in the publication; the previous “40 mg/kg/day oral gavage” framing was fabricated and was corrected R32c) for up to 12 weeks to rescue AMPK activity and metabolic fitness in middle-aged mice. The paper’s primary in-vivo readout in mice was phospho-HSP90α (Thr5/7) and downstream metabolic markers; direct p-DNA-PKcs measurement was not feasible in mice (the antibody did not cross-react with mouse DNA-PKcs). The widely-quoted “~3-fold elevation in aged muscle” is from rhesus macaques specifically (n=5/group, 15 yr vs 1 yr, Mann-Whitney p<0.01) — not from the mouse cohort. Whether peposertib or AZD7648 recapitulate this effect at non-genotoxic doses is untested. long-term-unknown — Chronic partial DNA-PKcs inhibition could impair NHEJ capacity and increase genomic instability; risk-benefit in a non-oncology setting is entirely uncharacterized.

Aging clinical trials: No registered clinical trials for DNA-PKcs inhibition as an aging or metabolic intervention as of 2026-05-05. needs-human-replication

Key interactors

  • ku70-ku80 — Ku heterodimer; essential for DNA-PKcs recruitment to DSB ends and kinase activation
  • ATM — paralog PIKK; ATM and DNA-PKcs phosphorylate overlapping substrates (H2AX, p53, RPA) but in different contexts (resected vs. blunt/minimally processed ends)
  • PARP1 — alternative sensor for SSBs and some DSBs; PARP1 and DNA-PKcs compete/cooperate at certain break types
  • mTOR — PIKK family paralog (structural, not functional direct interaction); shared pharmacophore makes selectivity of early-generation PI3K inhibitors relevant

Limitations and gaps

  • no-genage-entry — PRKDC has no confirmed GenAge entry; aging-lifespan phenotype in model organisms is secondary (NHEJ-mutant worms/flies show accelerated aging but confounded by DNA-repair defects)
  • needs-human-replication — The metabolic AMPK-suppression axis is a single-lab result in mouse + human myotubes; no independent human in-vivo evidence
  • needs-replication — Structural data on DNA-PKcs at atomic resolution remains incomplete; the highest-resolution crystal structure is 4.3 Å (Sibanda 2017), insufficient for side-chain-level mechanistic interpretation; DNA-bound and activated conformations are not yet resolved at high resolution
  • long-term-unknown — Chronic DNA-PKcs inhibition safety profile in aging (non-oncology) contexts is entirely unknown; NHEJ impairment risk is not characterized
  • dose-response-unclear — The optimal degree of DNA-PKcs inhibition to recover AMPK without impairing NHEJ is not defined in any model
  • The user-provided DOI for the Park 2017 paper (10.1016/j.cell.2017.06.011) resolved to an unrelated SLC16A11 paper — corrected to 10.1016/j.cmet.2017.04.008 (Cell Metabolism) via PubMed PMID 28467930. The user-provided DOI for the Davis 2014 review (10.1101/cshperspect.a015842) resolved to an unrelated paper — corrected to 10.1016/j.dnarep.2014.02.020 (DNA Repair). Both corrections confirmed via.

Footnotes

Footnotes

  1. park-2017-dna-pk-metabolism-aging · doi:10.1016/j.cmet.2017.04.008 · n=3–10/group depending on experiment (n=10/group for HFD weight/glucose studies; n=8/group for treadmill; n=3–6/group for Western blot quantification) · in-vivo + in-vitro · model: male C57BL/6J aged mice (3–25 mo), SCID mice, MDPKO mice, rhesus macaques (1–16 yr), C2C12 myotubes, human primary myotubes · inhibitor: NU7441 · Cell Metabolism 25:1135–1146, 2017 · cited-by: 125 · archive: downloaded (bronze OA) ↩ ↩2 ↩3 ↩4

  2. sibanda-2017-dna-pkcs-structure · doi:10.1016/bs.mie.2017.04.001 · in-vitro (X-ray crystallography; selenomethionine-labeled; 4.3 Å resolution) · Methods in Enzymology 592:145–157, 2017 · Chirgadze DY, Ascher DB, Blundell TL, Sibanda BL · archive: downloaded (green OA via PMC) ↩

  3. davis-2014-dna-pk-nhej-review · doi:10.1016/j.dnarep.2014.02.020 · review · DNA Repair 2014 · Davis AJ, Chen BPC, Chen DJ · cited-by: 320 · archive: not downloaded (closed access) ↩

  4. bosma-1983-scid-mouse · doi:10.1038/301527a0 · n=not-stated (colony characterization) · in-vivo (mouse, Prkdc-nonsense) · model: C.B-17/lcr-Scid inbred line · cited-by: 2194 · archive: not downloaded (closed access) ↩

  5. romesser-2024-peposertib-rectal-cancer · doi:10.1158/1078-0432.CCR-23-1129 · phase-1b clinical trial · n=not extracted · model: locally advanced rectal cancer patients · Clin Cancer Res 2024 · cited-by: 28 ↩

  6. samuels-2024-peposertib-head-neck · doi:10.1016/j.ijrobp.2023.09.024 · phase-1 clinical trial · n=not extracted · model: advanced head and neck tumor patients · Int J Radiat Oncol Biol Phys 2024 · cited-by: 55 ↩

Graph View

Backlinks