Cks1 and Cks2 hold both halves of KMT2A and set its level through the cell cycle
also called Cks1 Cks2 Mll1 · CKS1B CKS2 MLL · Cdc28 kinase subunit MLL
Two small adaptor proteins, Cks1 and Cks2, grip both halves of KMT2A and control how much of it exists at each stage of the cell cycle. Both are overexpressed in KMT2A-rearranged leukaemia, and drugs that block the disposal machinery they work with slow those cells more than normal ones.
The same thing, for a scientist
Cks1 and Cks2 are Cdk-substrate-targeting subunits that also act Cdk-independently as adaptors for SCF(Skp2) and APC(Cdc20). Grey et al. report that **Cks1 and Cks2 interact with both the MllN and MllC subunits of Mll1** and together define Mll1 levels through the cell cycle. **CKS1B and CKS2 are overexpressed in multiple human cancers including KMT2A-rearranged AML subtypes**, and the degradation-pathway inhibitors MLN4924 and C1 reduced KMT2A-rearranged cell proliferation more than primary controls. **The MllN half is the fusion-retained fragment, so at least part of this contact sits on the region this beat is auditing** - but because the interaction is reported to span both subunits, and because the paper does not resolve which part carries it, the retained/lost call cannot be made from the abstract. This is the clearest example in the beat of a contact that is *partly* retained and therefore *unclassifiable* without the full text.
Retained, lost, or both - and why this one cannot be called
KMT2A is cleaved by taspase1 into MllN and MllC, which persist as a complex. MllN is, approximately, the fragment a fusion retains; MllC is the fragment it loses. A protein reported to bind both subunits therefore has at least one determinant on each side of the breakpoint.
Three possibilities, and nothing in the accessible text distinguishes them:
- Two independent contacts, one retained and one lost - the fusion keeps half the interaction.
- One composite contact requiring both subunits - the fusion keeps none of it, because it has no MllC.
- A contact on the MllN/MllC complex interface - which the fusion cannot form at all.
Only the first would make Cks a non-menin handle on the retained fragment. The beat records this as undetermined, with the specific experiment that would determine it: a co-immunoprecipitation of Cks1 with a KMT2A fusion protein, which has apparently never been reported.
The Cdk connection, which is why it is in a phosphorylation beat
Cks proteins have two jobs: they dock Cdk substrates, increasing the efficiency of multisite phosphorylation, and they act as phospho-adaptors for the SCF and APC/C ligases. So a Cks contact on KMT2A implies KMT2A is a multisite Cdk substrate, which is consistent with the atlas finding that the retained fragment carries 56 observed phosphosites with 51 of them unassigned.
That is a specific, cheap prediction: KMT2A should carry clustered Cdk-consensus (S/T-P) sites near the Cks contact. Nobody has looked, and the phosphosite list this beat assembled makes the search trivial.
| partner a | CKS1B / CKS2 |
|---|---|
| partner b | gene-kmt2a |
| motif b | unmapped; reported to involve both the MllN and MllC subunits |
| fate in fusion | undetermined - the contact spans both taspase1-generated subunits and the abstract does not resolve which determinant sits where |
| tested on fusion | false |
| deciding experiment | co-immunoprecipitation of Cks1/Cks2 with a KMT2A fusion protein, which would show directly whether the retained fragment alone suffices |
structures
[]
- MLN4924/pevonedistat inhibits every cullin-RING ligase in the cell. Any anti-leukaemic effect attributed to the MLL1-Cks axis on the basis of that compound is attributed on very thin evidence.
- Cks1 and Cks2 are required for normal cell-cycle progression; the overexpression that makes them a marker is not by itself a therapeutic window.
- This beat's own literature contains three mutually incompatible recommendations about MLL1 protein levels. Acting on any of them without resolving the contradiction risks a trial whose mechanism runs backwards.
- Does Cks1 co-immunoprecipitate with a KMT2A fusion protein? One experiment converts this entry from undetermined to retained or lost.
- Are there clustered S/T-P sites near the Cks contact on KMT2A? The phosphosite list exists and the search has not been done.
- Is the MLN4924 sensitivity of KMT2A-rearranged lines mediated by MLL1 at all, or by p27 and the other canonical SCF(Skp2) substrates?
- Three papers now recommend opposite manipulations of MLL1 abundance. Which direction helps in a KMT2A-rearranged cell with an intact wild-type allele?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Cks1 and Cks2 proteins interact with both the MllN and MllC subunits of Mll1 and together define Mll1 levels throughout the cell cycle in concert with SCFSkp2 and APCCdc20; CKS1B and CKS2 are overexpressed in multiple human cancers including KMT2A-rearranged AML subtypes, and the small-molecule degradation-pathway inhibitors MLN4924 and C1 specifically reduced proliferation of KMT2A-rearranged cell lines compared with primary controls.
The Cks1/Cks2 axis fine-tunes Mll1 expression and is crucial for MLL-rearranged leukaemia cell viability. · Grey W, et al. · Biochim Biophys Acta Mol Cell Res 2018What this evidence cannot show
Abstract read 2026-09-19; full text not retrieved. The abstract does not resolve which Mll1 subunit carries which determinant, so the retained/lost call cannot be made. MLN4924 (pevonedistat) inhibits NEDD8-activating enzyme and therefore all cullin-RING ligases, so a differential effect on KMT2A-rearranged lines is not evidence that the Mll1-Cks axis mediates it. 'Compared with primary controls' compares immortalised leukaemia lines with primary cells, a comparison that confounds transformation with genotype.
Evidence that argues against this (2)
- The foundational study of this degradation circuit concluded the opposite therapeutic direction: that MLL fusions are already resistant to degradation and that impaired degradation is what causes the disease. Inhibiting the degradation machinery, as this paper proposes, would push wild-type MLL towards the state the earlier work identifies as oncogenic. Liu H et al. 2007
- A separate account of MLL1 stability reaches the opposite conclusion about which direction helps: there, raising full-length MLL1 by blocking its cleavage is what displaces the chimera from chromatin and delays leukaemia in mice. Two papers on MLL1 abundance therefore recommend opposite interventions, and neither tested the other's. Zhao Z et al. 2019
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
KMT2A's fusion-retained fragment carries 56 observed phosphosites of which 51 have no assigned kinase, a distribution consistent with the multisite-substrate role that a Cks adaptor contact implies.
Proteins API proteomics-PTM, Q03164: 130 phosphorylation sites, 56 at or below residue 1400 · EMBL-EBI 2026What this evidence cannot show
This is a consistency argument, not evidence for the Cks contact: a phospho-dense disordered region is consistent with many things besides Cdk multisite phosphorylation. No S/T-P clustering analysis was performed here, and the Cks contact has not been residue-mapped by anyone.
Evidence that argues against this (1)
- Phosphosite density in a disordered region is the expected background for any large nuclear protein and is not by itself evidence of regulated multisite phosphorylation; the same density argument would support a Cdk-substrate claim for most chromatin proteins, which makes it weak support for this specific one. Casado P et al. 2023
Cks1/Cks2 bind both MllN and MllC; retained/lost undetermined
the adaptor that couples MLL1 to SCF(Skp2) and APC(Cdc20)
adjacent contact, fate undetermined