Oncofusion

start / graph / gene-prkdc

Geneestablished confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.Q25-partner-phospho

PRKDC / DNA-PKcs - the kinase that phosphorylates AF9 on the one regulatory residue the fusion keeps

also called PRKDC · DNA-PKcs · DNA-dependent protein kinase catalytic subunit · p350

DNA-PK is a second alarm kinase for broken DNA. It puts a phosphate on one exact spot of AF9 - and that spot is inside the piece of AF9 that the leukaemia fusion protein keeps.

The same thing, for a scientist

PRKDC (P78527, 4128 aa) is the catalytic subunit of DNA-dependent protein kinase, a PIKK-family kinase recruited to double-strand breaks by the Ku70/Ku80 heterodimer and central to non-homologous end joining. Like ATM and ATR it phosphorylates S/T-Q motifs. In this beat it is the kinase that phosphorylates MLLT3/AF9 at S395 - the single S/T-Q motif in the entire AF9 sequence - releasing AF9 from the super elongation complex. S395 lies C-terminal of residue 376 and is therefore on retained sequence in roughly 92% of KMT2A::MLLT3 patients.

Go deeper
The specificity argument, made from sequence

A scan of the P42568 sequence performed in this session finds exactly one S/T-Q motif in AF9, at S395. That is the residue the primary paper assigns to DNA-PKcs 2. A one-motif protein is the rare case where a motif match is not the weak evidence this project usually treats it as: there is no second candidate for a PIKK to choose.

The paralogue check is informative too. ENL (Q03111) carries two S/T-Q motifs, at S394 and S398, and the AF9 S395 site aligns with ENL S394 - both sit in the shared P-S-Q ... G-P-L-R-S context. So the single PIKK site of AF9 and the first PIKK site of ENL are the same conserved position 3.

Facts and figures
uniprotP78527
length AA4128
accession resolutionDirect accession fetch of P78527; primary geneName read back as PRKDC and length as 4128.
substrate motifS/T-Q, shared with ATM and ATR
inhibitor used in primary paperNU7441, 2 uM
gene symbolPRKDC
substrates in this beat
[
  "gene-mllt3"
]
How this could bite us back
  • A DNA-PK inhibitor used as a chemosensitiser in a KMT2A::MLLT3 leukaemia would, on this mechanism, prevent the phosphorylation that releases the fusion from the SEC. The sensitising intent and the fusion-specific effect point in opposite directions, and nobody has measured which wins.
  • DNA-PKcs is required for V(D)J recombination and its loss causes severe combined immunodeficiency. In a paediatric patient that is not an abstract toxicity.
Still open
  1. Is AF9 S395 phosphorylated in a KMT2A::MLLT3 leukaemia cell after cytarabine or an anthracycline, and does the fusion protein lose SEC binding when it is?
  2. DNA-PK inhibitors are in clinical trials as chemo- and radio-sensitisers. What do they do to a KMT2A::MLLT3 blast, where blocking S395 phosphorylation would preserve rather than disrupt the fusion's elongation complex?
Every source behind this entry (3)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    PRKDC is UniProt P78527, a 4128-residue DNA-dependent protein kinase catalytic subunit; the primary gene name and length were read back from the accession.

    What this evidence cannot show

    A database identity record only.

    Evidence that argues against this (1)
    • Resolving an accession correctly says nothing about which PIKK acts on a given substrate in a given cell, since ATM, ATR and DNA-PKcs share the S/T-Q motif and are distinguished only pharmacologically or genetically. Ui A et al. 2016
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    DNA-PKcs phosphorylates AF9 at Serine 395 upon ionising radiation; treatment with the DNA-PK inhibitor NU7441 at 2 micromolar abolishes the IR-induced phosphorylation, and an AF9 S395A mutant shows defective phosphorylation with unaltered SEC interaction after IR, while the phosphomimic S395D reduces SEC binding.

    PMID 38750015DOI
    What this evidence cannot show

    Full text read via Europe PMC fullTextXML (PMC11096357). HEK293T cells throughout; several key panels are n=1 or n=2 replicates as the authors state in the figure legends. No haematopoietic cell, no KMT2A fusion, and no patient material.

    Evidence that argues against this (1)
    • The same group's earlier work places ATM, not DNA-PKcs, on the paralogous elongation factor ELL at its S/T-Q sites during genotoxic stress, so the PIKK assignment within this protein family is context-dependent and a single inhibitor experiment in one cell line does not fix it. Pal S et al. 2022
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    AF9 (P42568) contains exactly one S/T-Q motif in its 568 residues, at S395, and ENL (Q03111) contains exactly two, at S394 and S398; AF9 S395 and ENL S394 occupy the same aligned position, both within a shared P-S-Q...GPLRS sequence context.

    What this evidence cannot show

    This is a sequence computation performed in this session, not a published alignment. It is a motif scan, and this project's standing rule is that a short linear motif match means almost nothing on its own; it is reported here only because the AF9 motif count is one, and because an independently published functional site coincides with it.

    Evidence that argues against this (1)
    • The standing counter-argument against motif reasoning in this project is its own worked failure: a 14-3-3 motif scan of KMT2A produced a single candidate that the phosphosite atlas then showed is not phosphorylated at all. A motif is a hypothesis for an experiment, never a result - and neither ENL S394 nor ENL S398 appears in the PeptideAtlas set for Q03111. EMBL-EBI 2026
Connected to