Oncofusion

start / graph / concept-wdr5-what-is-established-for-wild-type-kmt2a-and-what-for-the-fusion

Explainerstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q23-wrad-adjudication

WDR5: an established wild-type KMT2A contact, and no fusion contact ever measured

also called WDR5 versus the fusion · WIN motif and the breakpoint · WDR5 evidence split

WDR5 grips a six-residue handle near the far end of KMT2A. Every leukemia fusion of the usual kind throws that handle away, so WDR5 should not be able to hold the fusion - and no one has ever run the experiment either way. What WDR5 drugs do to these cells turns out to run through ribosomes and p53, not through KMT2A.

The same thing, for a scientist

**(a) Wild-type.** Established at atomic resolution: WDR5 binds the KMT2A WIN motif at approximately aa 3762-3773 with Arg3765 inserted into the WDR5 arginine pocket [E1], and endogenous WDR5 coprecipitates with transfected full-length MLL but not with a SET-domain-deleted construct [E2]. **(b) Fusion.** Nothing measured. No published co-immunoprecipitation, mass spectrometry, structure, degron or proximity experiment reports WDR5 with any KMT2A fusion protein; an alias-expanded PubMed query pairing WDR5 with the named common fusions returns 26 records, of which 21 are medicinal chemistry on the WDR5-MLL1 interface, 3 are reviews and none is a fusion-protein association experiment [E6]. What does exist is genome-wide separation: in murine MLL-AF9 leukaemia cells, an antibody that detects only the 180 kDa MLL1 C-terminal fragment and therefore cannot see MLL-AF9 shows wild-type MLL1 and WDR5 co-occupying enhancers enriched for H3K4me1/2, while MLL-AF9 binds regions with no H3K4me1 enrichment [E3]. **(c) Domain.** WIN motif, aa 3762-3773, roughly 2,400 residues 3' of the breakpoint cluster region - LOST from every direct 5'-KMT2A fusion, RETAINED in reciprocal der products and in NUP98::KMT2A.

Go deeper
The special case warning this beat was given, and it holds

WDR5 is not a KMT2A-specific protein and a WDR5 result is not automatically a KMT2A-axis result. Three published facts make that concrete.

  1. The WIN site is shared. SETD1A, SETD1B and KMT2B all carry their own WIN motifs competing for the same WDR5 pocket, and WDR5 also binds MYC and the NSL complex through distinct surfaces 4.
  2. The phenotype of WIN-site inhibition is not a KMT2A phenotype. In MLL-rearranged cancer cells, picomolar WIN-site inhibitors act by suppressing ribosome protein gene transcription, inducing nucleolar stress and activating p53, with MDM4-inactivating splicing changes; the authors state explicitly that their data displace the earlier haematopoietic-stem-cell-expansion explanation 5.
  3. The genetic result and the pharmacological result agree on the requirement but not on the mechanism. Genetic deletion of Wdr5 impairs MLL-AF9 leukaemogenesis in vivo, and the authors attribute it to direct regulation of ribosomal genes, not to fusion biology 7.

So WDR5 is required in KMT2A-rearranged cells and the requirement has not been shown to run through the fusion.

The dependency data point the other way from the pharmacology

WDR5 is DepMap common-essential (mean Chronos -1.159 across 1,258 lines, 98.3% of lines depleted) and is depleted LESS in KMT2A-rearranged lines than in leukaemia lines generally, giving a lineage-controlled selectivity of +0.191, the worst of 41 candidate partners scored in this project 8. Its leukaemia-restricted correlation with MEN1 is r = 0.004, a confidence interval straddling zero 8. A protein that every cell needs and that KMT2A-rearranged cells need slightly less than other leukaemia cells is a poor candidate for a fusion-specific handle, whatever the chemistry says.

What would change the answer

A subunit-resolved immunoprecipitation of an epitope-tagged KMT2A::AFF1 from SEM or RS4;11 blotted for WDR5, or the same experiment on AF4-MLL. If WDR5 came down with the direct fusion, the retained-domain argument would need a non-WIN explanation - for example an indirect bridge through PAF1C, which does contact the retained fragment, or through the wild-type allele that KMT2A fusions are known to require 9. That is a real possibility and it has never been excluded, because the experiment has never been done.

Facts and figures
subunitWDR5
uniprotP61964
wild type contactWIN motif, KMT2A aa 3762-3773, Arg3765 in the WDR5 arginine pocket
contact retained in direct fusionfalse
contact retained in reciprocal fusiontrue
contact retained in nup98 kmt2atrue
fusion association evidencenone published, in either direction
pubmed records subunit x fusion26
pubmed queryWDR5 AND (MLL-AF4 OR MLL-AF9 OR MLL-ENL OR MLL-AF10 OR MLL-AF6 OR MLL-ELL OR "MLL fusion" OR "MLL fusion protein" OR "KMT2A fusion" OR "MLL-rearranged" OR "KMT2A-rearranged")
verdictwild-type KMT2A interactor; functionally required in MLL-rearranged cells through a ribosome/p53 route; no measured association with any KMT2A fusion protein
aliases searched
[
  "WDR5",
  "BIG-3",
  "BIG3"
]
How this could bite us back
  • A WIN-site inhibitor displaces WDR5 from every WIN motif at once - KMT2A, KMT2B, SETD1A, SETD1B - so an efficacy signal cannot be attributed to KMT2A without a separation-of-function control.
  • The published mechanism of WIN-site inhibition requires functional p53, so the modality is predicted to fail exactly in the TP53-altered relapsed disease where a new agent is most needed.
  • Resistance to a potent WDR5 inhibitor has already been produced in vitro by a single target mutation, WDR5 P173L, which is the classic occupancy-inhibitor liability and an argument for degradation over occupancy.
  • WDR5 is common-essential in 98.3% of cell lines; a therapeutic window here would have to come from dosing and tissue distribution rather than from target selectivity.
Still open
  1. Does WDR5 co-immunoprecipitate with a tagged KMT2A::AFF1? Twenty-two years after the question was framed, the experiment has not been published in either direction.
  2. Two 2024 papers disagree about whether wild-type MLL1 and MLL-AF9 occupy the same chromatin. One reports distinct distributions with WDR5 following MLL1 only; the other reports identical patterns. That disagreement decides how the genome-wide argument in this entry should be read and nobody has adjudicated it.
  3. Does the WDR5 requirement in KMT2A-rearranged cells differ between reciprocal-positive and reciprocal-negative t(4;11) lines? That is the one comparison in which a fusion-mediated WDR5 requirement could appear, and no published dataset is stratified that way.
Every source behind this entry (9)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The crystal structure of WDR5 bound to the MLL1 WIN-motif peptide identifies Arg-3765 of MLL1 as the arginine occupying the WDR5 pocket and shows the WIN peptide is preferentially recognised over histone H3.

    Structure of WDR5 bound to mixed lineage leukemia protein-1 peptide. · Patel A, Dharmarajan V, Cosgrove MS · J Biol Chem 2008
    PMID 18829459DOI
    What this evidence cannot show

    A 1.7 A peptide co-crystal; it establishes the contact and its chemistry but not its occupancy, competition or stability in a cell, and the peptide is taken out of the context of the full SET-domain module.

    Evidence that argues against this (1)
    • Cryo-EM of the MLL1-WRAD complex on an H2B-ubiquitinated nucleosome shows variable docking of MLL1-WRAD and ambiguity in the positions of ASH2L and DPY30, so the static peptide picture understates how conditional the assembly is. Rahman S et al. 2022
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Endogenous WDR5 coprecipitates with transfected full-length MLL in 293 cells and is lost when the SET domain is deleted; WDR5 was never blotted on any fusion protein in the paper that established this.

    PMID 15199122DOI
    What this evidence cannot show

    Transient overexpression in 293 cells, an epithelial line with no relevance to haematopoiesis, and the deletion construct retains the WIN motif yet loses WDR5, which the paper does not address because the WIN motif was not known in 2004.

    Evidence that argues against this (1)
    • The same paper's siRNA experiment found that >80% knockdown of WDR5 had no measurable effect on HoxA9 expression in HeLa cells, so even in the wild-type setting the functional consequence of removing WDR5 from the MLL complex was not demonstrated by the founding work. Yokoyama A et al. 2004
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In primary murine MLL-AF9 leukaemia cells, wild-type MLL1 (detected with an antibody specific for the 180 kDa MLL1 C-terminal fragment, which cannot detect MLL-AF9) and WDR5 co-occupy enhancers enriched for H3K4me1/2, while MLL-AF9 binds chromatin regions with no H3K4me1 enrichment; blocking the MLL1-WDR5 interaction with MM-401 displaces WDR5 from about 2,500 loci and displaces MLL1 with it.

    PMID 27462455DOI
    What this evidence cannot show

    ChIP-seq co-occupancy is not physical association, and the MLL-AF9 ChIP relies on a separate antibody. The system is a retrovirally transduced murine model plus MOLM-13 confirmation; no t(4;11) line was used. Genome-wide separation of binding sites is consistent with, but does not prove, absence of a WDR5-fusion contact.

    Evidence that argues against this (1)
    • A 2024 ChIP study reports the opposite conclusion about localisation, finding that MLL1 and MLL-AF9 occupy promoters and distal intergenic regions with identical chromatin occupancy patterns in MLL cells, explicitly contrary to what the field had believed. Yang Z et al. 2024
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.review

    WDR5 is the shared WD40 hub of the WRAD module common to all six SET1-family H3K4 methyltransferases, so its WIN pocket is engaged by several catalytic subunits besides KMT2A.

    WRAD: enabler of the SET1-family of H3K4 methyltransferases. · Ernst P, Vakoc CR · Brief Funct Genomics 2012
    PMID 22652693DOI
    What this evidence cannot show

    A review, cited for complex composition only.

    Evidence that argues against this (1)
    • WDR5 additionally partitions into the NSL complex through a KANSL1 linear motif that is mutually exclusive with its MLL/COMPASS binding, so the shared-hub picture is broader than the SET1 family and a WIN-site agent redistributes WDR5 rather than simply removing it from KMT2A. Yang Z et al. 2018
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In MLL-rearranged cancer cells, WDR5 WIN-site inhibition acts by suppressing ribosome protein gene transcription, inducing nucleolar stress and activating p53, with ribosome-subunit attrition, a broad modest translational choke and alternative splicing that inactivates MDM4; the authors state that the earlier haematopoietic-stem-cell explanation for WIN-site inhibitor action is displaced by these data.

    PMID 38682900DOI
    What this evidence cannot show

    Cell-line pharmacology; the p53 dependence means the mechanism as described cannot operate identically in TP53-mutant disease, which is a meaningful fraction of relapsed KMT2A-rearranged leukaemia.

    Evidence that argues against this (1)
    • Genetic deletion of Wdr5 in an MLL-AF9 murine model reproduces the growth and leukemogenesis defect, so the requirement is not an artefact of the chemical probe even if the downstream mechanism is ribosomal. Liu L et al. 2023
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    An alias-expanded PubMed query pairing WDR5 with the named common KMT2A fusions and with 'MLL fusion'/'MLL-rearranged' returns 26 records, none of which reports a physical association between WDR5 and a KMT2A fusion protein; the set is dominated by medicinal chemistry on the WDR5-MLL1 interface.

    What this evidence cannot show

    A counted absence is only as good as its symbol set. WDR5's alias BIG3 returns 0 records paired with MLL or KMT2A and is in any case a word-like acronym. Titles were checked; abstracts were read only for the records classified as primary. A co-IP buried in a supplementary figure of a paper whose title and abstract do not mention WDR5 would be invisible to this query.

    Evidence that argues against this (1)
    • Complex-name phrases are unreachable from subunit symbols in PubMed, a trap this project has been caught by before; a paper framing its result at the level of 'the MLL1 core complex' rather than WDR5 would not appear in this count. Schmidt L et al. 2019
  7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Genetic deletion of Wdr5 impairs cell growth and colony-forming ability of MLL-AF9 leukaemia cells in vitro and ex vivo and attenuates leukemogenesis in vivo, acting through direct regulation of ribosomal genes; pharmacological inhibition recapitulates the genetic result in the same model.

    Loss of Wdr5 attenuates MLL-rearranged leukemogenesis by suppressing Myc targets. · Liu L, Guo X, Wang Y, et al. · Biochim Biophys Acta Mol Basis Dis 2023
    PMID 36402263DOI
    What this evidence cannot show

    A single murine MLL-AF9 model; no t(4;11) or infant-derived system, no human primary cells, and the ribosomal-gene mechanism is inferred from expression data rather than from a separation-of-function allele.

    Evidence that argues against this (1)
    • A selective WDR5 degrader outperformed WDR5 occupancy inhibitors on WDR5-regulated transcription, chromatin-bound MLL complex, c-MYC and MLL-rearranged AML patient cells in vivo, which means occupancy and degradation of the same protein give different answers and a genetic deletion is the degradation-like arm. Yu X et al. 2021
  8. E8moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    WDR5 is DepMap common-essential with mean Chronos -1.159 across 1,258 lines, is depleted less in KMT2A-rearranged lines (-1.051) than in leukaemia lines generally (-1.242) giving a lineage-controlled selectivity of +0.191, and has a leukaemia-restricted correlation with MEN1 of r = 0.004 with a confidence interval crossing zero.

    What this evidence cannot show

    Restated from this project's own earlier runs rather than recomputed here. A gene depleted in 98.3% of lines has almost no dynamic range in which selectivity could appear, so the +0.191 is floor-compressed and close to uninterpretable; the KMT2A-rearranged arm in DepMap is myeloid-enriched; DepMap release version unverified.

    Evidence that argues against this (1)
    • WDR5 nevertheless has by far the largest chemical-matter footprint of the MLL-complex proteins and a resistance mutation, WDR5 P173L, has been isolated in a drug-adapted line - features of a target that is being engaged, not one that is irrelevant. Liu L et al. 2023
  9. E9strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    KMT2A fusion proteins lack the PHD fingers and require prebinding of a wild-type KMT2A complex, together with CXXC recognition of DNA, for stable association with HOXA9 - so a WRAD-loaded wild-type complex is a documented indirect route from WDR5 to the fusion's target loci.

    PMID 20541448DOI
    What this evidence cannot show

    A locus-specific study at HOXA9 in a limited set of models; the generality of the prebinding requirement across partners and across the genome has not been established.

    Evidence that argues against this (1)
    • If the fusion's chromatin association ran through a WRAD-loaded wild-type complex, WDR5 loss should phenocopy fusion loss, and it does not: the reported WDR5 phenotype in MLL-rearranged cells is ribosomal and p53-driven rather than a collapse of the fusion programme. Howard GC et al. 2024
Connected to