ATR phosphorylates KMT2A at S516 to break the SCF(Skp2) contact - a phospho-gated degradation switch sitting inside the retained fragment
also called MLL S516 · ATR-MLL S-phase checkpoint · Skp2 phosphoswitch
When DNA is damaged during copying, a sensor enzyme called ATR puts a phosphate on KMT2A at one specific spot. That phosphate stops a disposal machine from grabbing KMT2A, so the protein survives and slows DNA copying down. The spot is on the piece that leukaemia fusions keep.
The same thing, for a scientist
MLL is phosphorylated at **serine 516 by ATR** in response to genotoxic stress in S phase. The phosphate **disrupts MLL's interaction with SCF(Skp2)** and so blocks its degradation, letting MLL accumulate on chromatin, methylate H3K4 at late replication origins and inhibit CDC45 loading to delay replication. Mll-null MEFs are rescued by wild-type but not **S516A** MLL. S516 lies at residue 516 of 3,969, **inside the fragment every KMT2A fusion retains**, and is independently observed as a phosphosite in the PRIDE/PTMeXchange re-analysis. The trap in reading this as a second handle is the paper's own conclusion: **MLL fusions behave as dominant negatives that abrogate ATR-mediated phosphorylation of wild-type MLL**, so the fusion is described as disabling this axis in trans rather than using it. The residue is retained; whether the switch operates on the fusion is untested.
What the switch does
Order of events as reported 1:
- Replication stress activates ATR.
- ATR phosphorylates MLL S516.
- Phospho-S516 MLL no longer binds SCF(Skp2), so it is not degraded at S phase.
- MLL accumulates on chromatin, deposits H3K4me at late origins, blocks CDC45 loading.
- Replication slows - the S-phase checkpoint.
MLL-deficient cells show radioresistant DNA synthesis and chromatid-type abnormalities; myeloid progenitors carrying an Mll-CBP knock-in allele show a severe radioresistant DNA synthesis phenotype.
Retained, but is it used?
Retained, on sequence. S516 is 884 residues N-terminal to the conventional 1400 boundary and sits in the same region as the CXXC domain and the menin/LEDGF clamp. Every prevalent fusion carries it.
Used, unknown. The paper frames fusions as dominant negatives against the wild-type axis. Two readings are compatible with that:
- the fusion carries S516, ATR still phosphorylates it, and the consequence is different because the Skp2 degron's other determinants are missing or masked by the partner;
- the fusion carries S516 and nothing writes or reads it.
The companion 2007 paper is the reason to take the first reading seriously: it reports that MLL degradation initiates at the N-terminal ~1400 amino acids and that tested prevalent MLL fusions are resistant to degradation. The degron machinery therefore engages the retained region in the wild-type protein and fails to on the fusion - which is a specific, testable difference located exactly where this beat is looking 2.
This entry does not claim S516 is a therapeutic handle. It claims the residue is retained, the contact it gates is on the retained fragment, and nobody has tested either on a fusion.
| kinase | ATR |
|---|---|
| substrate | gene-kmt2a |
| uniprot | Q03164 |
| site | S516 |
| fate in fusion | retained |
| split residue | 1400 |
| gated contact | SCF(Skp2) E3 ligase - phosphorylation disrupts the interaction |
| tested on fusion | No. The paper tests fusions as dominant negatives against wild-type MLL; it does not report whether S516 on a fusion is phosphorylated. |
| observed in proteomics | true |
| observed site ambiguity | UniProt annotates the adjacent S518; PRIDE re-analysis reports S516 at localisation probability 0.82 |
- This axis argues for caution with ATR inhibitors in KMT2A-rearranged disease, not for enthusiasm: if the fusion already compromises an MLL-dependent S-phase checkpoint, adding an ATR inhibitor may hit the normal marrow harder than the leukaemia.
- A degrader that restored SCF(Skp2) engagement of the fusion would also degrade the wild-type allele, whose stabilisation is the mechanism by which CK2 inhibition is proposed to help. Two of this beat's therapeutic ideas push the same protein in opposite directions.
- S516 is retained in sequence and untested in function. Treating a retained residue as a retained handle is the exact inference this beat exists to check, and here it does not yet hold.
- Is S516 phosphorylated on a KMT2A fusion protein in a t(4;11) or t(9;11) line after replication stress? One phospho-specific western would answer it and nobody appears to have done it.
- Which residue actually carries the site, S516 or S518? The two major databases disagree and the functional literature has only ever tested S516A.
- Do KMT2A-rearranged leukaemias show the radioresistant DNA synthesis phenotype, and is it reversed by menin inhibition?
- If the fusion is resistant to SCF(Skp2) because the degron is masked by the 3' partner rather than deleted, is it unmaskable - and is that a degrader strategy?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MLL is phosphorylated at serine 516 by ATR in response to genotoxic stress in S phase, which disrupts its interaction with and degradation by the SCF(Skp2) E3 ligase; stabilised MLL accumulates on chromatin, methylates H3K4 at late replication origins and inhibits CDC45 loading, and reconstitution of Mll-null MEFs with wild-type but not S516A or SET-deleted MLL rescues the S-phase checkpoint defect.
Phosphorylation of MLL by ATR is required for execution of mammalian S-phase checkpoint. · Liu H, et al. · Nature 2010What this evidence cannot show
Abstract read 2026-09-19; full text not retrieved in this session. The rescue experiments are in mouse embryonic fibroblasts and the fusion experiments use an Mll-CBP knock-in allele modelling t(11;16), which is not one of the prevalent fusions this graph centres on. The abstract does not state whether S516 phosphorylation was assayed on a fusion protein.
Evidence that argues against this (2)
- The same paper concludes that MLL fusions act as dominant-negative mutants abrogating ATR-mediated phosphorylation and stabilisation of wild-type MLL. On that reading the S516 switch is something the fusion breaks rather than something it carries, and a drug aimed at it would be aimed at a pathway the disease has already disabled. Liu H et al. 2010
- ATR inhibition is being pursued across oncology on the opposite logic - that abrogating the S-phase checkpoint kills replication-stressed tumour cells. A KMT2A-rearranged leukaemia already carrying a compromised MLL-dependent S-phase checkpoint could be either unusually sensitive or already adapted, and the direction is not predictable from this mechanism alone. Liu H et al. 2007
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
SCF(Skp2) and APC(Cdc20) mark MLL for degradation at S phase and late M phase respectively; MLL degradation initiates at its N-terminal approximately 1400 amino acids, and tested prevalent MLL fusions are resistant to degradation, which the authors propose as a mechanism common to all MLL leukaemias.
Bimodal degradation of MLL by SCFSkp2 and APCCdc20 assures cell cycle execution: a critical regulatory circuit lost in leukemogenic MLL fusions. · Liu H, et al. · Genes Dev 2007What this evidence cannot show
KNOWN BUT NOT FULLY READ. Abstract only. Three routes were tried on 2026-09-19 and all failed: Europe PMC fullTextXML returned HTTP 500; efetch db=pmc returned a well-formed record carrying the note 'The publisher of this article does not allow downloading of the full text in XML form' and no body element; the publisher HTML at genesdev.cshlp.org and the Europe PMC article page both returned HTTP 403. So which fusions were tested, and whether the Skp2 contact is a classical phosphodegron, are not established here and must not be inferred.
Evidence that argues against this (1)
- The claim that impaired degradation 'likely constitutes the universal mechanism underlying all MLL leukemias' is the authors' proposal, not a demonstrated result, and it competes directly with this graph's dominant account in which the fusion acts by hijacking transcriptional elongation. A stability mechanism and an elongation mechanism make different predictions about what a degrader would do. Grey W et al. 2018
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
S516 is independently observed as a phosphorylation site on human KMT2A in the PRIDE/PTMeXchange re-analysis served by the EMBL-EBI Proteins API, with a localisation probability of 0.82 in project PXD012174.
What this evidence cannot show
Fetched 2026-09-19. A localisation probability of 0.82 is acceptable but not decisive; the neighbouring S518 is also reported, and at this probability the phosphate could in principle be on the neighbour. The observation is in whatever cell types PXD012174 sampled, with no relation to the genotoxic-stress condition in which the ATR event was characterised.
Evidence that argues against this (1)
- UniProt's own curated table annotates S518 as a phosphoserine and does not annotate S516, so the two principal databases disagree about which of the adjacent serines carries the site. That is exactly the ambiguity a 0.82 localisation probability admits, and it means the ATR site and the observed site are not certainly the same residue. UniProt Consortium 2026
the kinase named in that mechanism; identity and length asserted here
- regulates →KMT2A / MLL1 — the gene that breaks
ATR phosphorylation of S516 blocks SCF(Skp2)-mediated degradation
one of five retained sites with a named kinase
- related to →KMT2A-rearranged acute lymphoblastic leukaemia
checkpoint dysfunction proposed as a contributor to KMT2A-rearranged leukaemogenesis
the contact ATR's S516 phosphorylation gates