Oncofusion

start / graph / localization-wdr5-kmt2a-centrosome-and-midbody

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q57-localization2

WDR5 and KMT2A are at the centrosome and the midbody - a non-chromatin address for both, with a patient-cell phenotype

also called WDR5 localization · MLL centrosome · MLL midbody · pericentriolar material KMT2A · WDR5 Cep72 · WDR5 KIF2C

WDR5 and MLL are thought of as chromatin proteins, but two recent studies put both of them on the machinery that pulls chromosomes apart and then pinches the cell in two. Removing either one leaves cells with crooked spindles, misaligned chromosomes and two nuclei - and cells from patients with an MLL-haploinsufficiency syndrome show the same defect.

The same thing, for a scientist

UniProt P61964 annotates WDR5 as `Nucleus` on three experimental references and nothing else [E1]. That is incomplete. MLL/KMT2A and WDR5 **localise to the pericentriolar material**, interact with the centriolar satellite protein **Cep72** and with gamma-tubulin ring complex proteins, and promote localisation of gamma-TuRCs and AKAP9 to the centrosome in interphase and mitosis; loss of MLL, WDR5 or Cep72 disrupts spindle formation and produces misaligned chromosomes, and **the same centrosomal phenotype is seen in cells derived from patients with Wiedemann-Steiner syndrome** (MLL haploinsufficiency) [E2]. Separately, MLL and WDR5 localise to the **midbody**, target the kinesin-13 depolymerase **KIF2C** there, and their loss gives a wide midzone-like microtubule structure, chromosome bridges and **binucleated cells** [E3]. On chromatin, WDR5's address is narrow and WIN-site-dependent: a picomolar WIN-site inhibitor **displaces WDR5 from chromatin at a small cohort of loci including ribosome protein genes**, causing translational inhibition, **nucleolar stress** and p53 induction [E4][E5].

Go deeper
Two addresses nobody's database records

Centrosome. MLL and WDR5 are in the pericentriolar material. They bind Cep72 and gamma-TuRC components, and they are required for gamma-TuRC and AKAP9 recruitment to the centrosome in both interphase and mitosis. Losing MLL, WDR5 or Cep72 disturbs microtubule nucleation and regrowth, spindle formation and chromosome alignment 2.

Midbody. MLL was already known to localise to the midbody with binucleation on loss; this work adds WDR5, and shows the complex targets KIF2C there. KIF2C's depolymerase activity then positions centralspindlin, compacts midzone microtubules and permits timely furrow initiation 3.

Neither address appears in UniProt for WDR5 1. This is the most direct demonstration in this beat that a graph built from database localisation would have no idea these exist - and they concern the graph's own central protein.

Why this is not a footnote for a leukaemia graph
  1. The phenotype is in patient cells. Wiedemann-Steiner syndrome is caused by KMT2A haploinsufficiency, and WSS patient-derived cells reproduce the centrosomal defect 2. That is a human germline genotype-to-phenotype link running through a non-chromatin function of KMT2A.
  2. Binucleation and chromosome bridges are a mutagenic state. A cell that fails cytokinesis or mis-segregates chromosomes acquires structural genomic change. KMT2A translocations arise from double-strand breaks, and this graph holds a whole break-machinery layer. A KMT2A-dosage-dependent cytokinesis defect is a candidate route from reduced KMT2A function to genomic instability that the break layer does not currently contain.
  3. The fusion question is wide open. KMT2A fusions retain the N-terminal region; whether they retain midbody or centrosome localisation, act as dominant negatives there, or lose it entirely, is untested. If a fusion sequestered WDR5 away from the centrosome, the leukaemic cell would inherit a mitotic defect as a side effect of the oncoprotein.

Rule 73 on all of it: these are cell-biology systems (human cell lines plus WSS patient-derived cells), studying wild-type MLL dosage, not a KMT2A fusion and not leukaemia.

The chromatin address is narrower than the reputation

The WIN site is the pocket that binds the KMT2A/KMT2 WIN motif and is the main WDR5 drug target. Potent WIN-site inhibitors show that the WIN site links WDR5 to chromatin at only a small cohort of loci, prominently ribosome protein genes; inhibition rapidly displaces WDR5 from those loci and lowers their expression, causing translational inhibition, nucleolar stress and p53 induction 4. A later dissection in MLL-rearranged cancer cells finds the dominant response to WIN-site inhibition is ribosome subunit attrition and activation of the p53-MDM4 axis 5.

Two consequences for this graph:

  • The therapeutic effect of WIN-site inhibition in KMT2A-rearranged disease may be a ribosome/p53 effect, not a fusion-complex effect. That is a direct challenge to the assumption that a WDR5 inhibitor works by disabling KMT2A-fusion methyltransferase activity, and it belongs beside concept-a-wdr5-result-is-not-automatically-a-kmt2a-axis-result.
  • Nucleolar stress is a localisation phenotype. If the readout of WDR5 inhibition is nucleolar, the assay for it is imaging, and no one has imaged nucleoli in a KMT2A-rearranged cell under a WIN-site inhibitor.
Facts and figures
genegene-wdr5
also concernsgene-kmt2a
uniprotP61964
length AA334
compartment subcellularnucleus (UniProt) PLUS pericentriolar material / centrosome and midbody (2024-2025 primary literature, not in UniProt)
compartment subnuclearno nuclear-body observation. The chromatin address is WIN-site-dependent and narrow - a small cohort of loci including ribosome protein genes - and WIN-site inhibition produces nucleolar stress, which is itself a compartment-level phenotype.
patient material datatrue
patient material notethe centrosomal phenotype is reproduced in cells from patients with Wiedemann-Steiner syndrome (KMT2A haploinsufficiency)
shown for fusionNO for the centrosome and midbody - both are wild-type MLL dosage experiments. The WIN-site mechanism-of-action work IS in MLL-rearranged cells, and finds the dominant response to be ribosome attrition and p53-MDM4 activation rather than a fusion-complex effect.
independence caveatthe centrosome and midbody results come from a single laboratory and have not been independently replicated
uniprot evidence codes
{
  "Nucleus": "ECO:0000269|PubMed:17355966, 20018852, 34850113",
  "centrosome": "ABSENT from the record",
  "midbody": "ABSENT from the record"
}
non chromatin partners
[
  "Cep72 (centriolar satellite)",
  "gamma-tubulin ring complex proteins",
  "AKAP9",
  "KIF2C (kinesin-13 depolymerase)",
  "centralspindlin (downstream)"
]
loss of function phenotypes
[
  "impaired microtubule nucleation and regrowth",
  "defective spindle formation and misaligned chromosomes",
  "defective midbody with wide midzone-like microtubules",
  "chromosome bridges",
  "binucleation"
]
methods
[
  "immunofluorescence localisation to pericentriolar material and midbody",
  "microtubule regrowth assays",
  "co-immunoprecipitation with Cep72 and gamma-TuRC components",
  "knockdown/knockout with mitotic phenotype scoring",
  "patient-derived cells",
  "picomolar WIN-site chemical probe with chromatin displacement readout"
]
species and systems
[
  "human cell lines",
  "Wiedemann-Steiner syndrome patient-derived cells",
  "MLL-rearranged cancer cell lines (for the WIN-site work)"
]
How this could bite us back
  • If WDR5 is required for spindle formation and cytokinesis, a WDR5 inhibitor or degrader carries a mitotic-toxicity mechanism in every dividing tissue - the classic antimitotic toxicity profile, arriving from an epigenetic drug where it would not be anticipated.
  • If the dominant response of MLL-rearranged cells to WIN-site inhibition is p53-MDM4 activation, efficacy should collapse in TP53-altered disease - which is the relapsed population - and the drug's selectivity for KMT2A rearrangement becomes doubtful.
  • Binucleation and chromosome bridges on MLL/WDR5 loss mean that partial, chronic inhibition could be genome-destabilising rather than merely cytostatic, which is a poor property in a paediatric survivor population.
  • A pharmacodynamic assay measuring H3K4me3 would miss the centrosomal and midbody arms of WDR5 inhibition entirely.
Still open
  1. Do KMT2A fusion proteins retain midbody and centrosome localisation? The fusions keep the N-terminal region; nobody has looked, and a fusion that sequestered WDR5 from the centrosome would give the leukaemic cell a mitotic defect as a side effect.
  2. Is the KMT2A-dosage cytokinesis defect a route to genomic instability? Binucleation and chromosome bridges are mutagenic states, and this graph's break-machinery layer contains no such route.
  3. Is the centrosomal role direct or transcriptional? Only acute degron depletion with a short timecourse, or a localisation-dead separation-of-function allele, can tell, and neither is reported.
  4. Does a second laboratory see MLL and WDR5 at the centrosome and midbody? Both papers are from one group, and this is the kind of claim that needs replication.
  5. Under a WIN-site inhibitor, do nucleoli in KMT2A-rearranged cells actually change? 'Nucleolar stress' is currently a molecular signature; the imaging has not been reported.
Every source behind this entry (5)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProt P61964 (human WDR5, 334 aa) annotates subcellular location as Nucleus only, with three experimental references (PubMed 17355966, 20018852, 34850113), and records no centrosomal, midbody or cytoplasmic pool.

    What this evidence cannot show

    A database record. gene_exact:WDR5 AND organism_id:9606 AND reviewed:true returned one reviewed record, gene_primary WDR5, accession P61964, length 334. The record is not wrong - WDR5 is nuclear - it is incomplete, and the incompleteness is exactly the kind a bulk import cannot detect.

    Evidence that argues against this (1)
    • Database lag is the ordinary explanation: the centrosome and midbody papers are from 2024 and 2025 and curation trails publication. That is a reason to check the recent literature rather than a criticism of the record, and it is the general argument against building localisation fields from a database snapshot.
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    MLL/KMT2A and WDR5 localise to the pericentriolar material and interact with the centriolar satellite protein Cep72 and with gamma-tubulin ring complex proteins; they promote localisation of gamma-TuRCs and structural proteins such as AKAP9 to the centrosome during interphase and mitosis; loss of MLL or WDR5 affects microtubule nucleation and regrowth; during mitosis loss of MLL, WDR5 or Cep72 disrupts spindle formation and causes misaligned chromosomes; and the centrosomal phenotype is also observed in cells derived from patients with Wiedemann-Steiner syndrome.

    PMID 39661677DOI
    What this evidence cannot show

    RULE 73: human cell lines plus Wiedemann-Steiner-syndrome patient-derived cells. The genotype studied is MLL HAPLOINSUFFICIENCY - a loss-of-function germline condition - not a KMT2A fusion and not leukaemia. Loss-of-function perturbations of a nuclear protein can produce centrosomal phenotypes indirectly through transcription of centrosomal genes; the localisation data and the Cep72/gamma-TuRC interactions are the arguments against that reading, and a rapid-degron timecourse would be a stronger one. Read: abstract only (PubMed efetch, 2026-09-20).

    Evidence that argues against this (1)
    • The classic alternative for any chromatin protein with a cytoskeletal phenotype is that the phenotype is transcriptional: MLL loss lowers expression of centrosomal components and the centrosome fails downstream. Knockdown and knockout experiments cannot separate this from a direct role; only acute degradation with a short timecourse, or a localisation-dead separation-of-function allele, can. Neither is reported.
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    MLL and WDR5 localise to the midbody; loss of the MLL/WDR5 complex gives defective midbody formation with a wide midzone-like microtubule structure, chromosome bridges and binucleated cells; MLL and WDR5 interact with the kinesin-13 motor KIF2C and target it to the midbody, where its depolymerase activity promotes correct centralspindlin localisation, midzone microtubule compaction and timely furrow initiation.

    PMID 40958680DOI
    What this evidence cannot show

    RULE 73: cell-line cytokinesis assays; wild-type MLL dosage, no fusion, no leukaemia, no patient material. From the same laboratory as the centrosome work, so the two results are not independent. The abstract does not state which cell lines or which MLL and WDR5 antibodies were used, and the full text was not retrieved for this entry. Read: abstract only (PubMed efetch, 2026-09-20).

    Evidence that argues against this (1)
    • Independent replication is the missing piece: both the centrosome and midbody results come from one group, and a non-chromatin localisation for a famous chromatin protein is exactly the claim that most needs a second laboratory. Until then the correct confidence is moderate, and this entry does not raise it on the strength of the phenotype alone.
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Potent WIN-site inhibitors demonstrate that the WIN site links WDR5 to chromatin at a small cohort of loci including a specific subset of ribosome protein genes; WIN-site inhibition rapidly displaces WDR5 from chromatin and decreases expression of the associated genes, causing translational inhibition, nucleolar stress and p53 induction.

    Displacement of WDR5 from Chromatin by a WIN Site Inhibitor with Picomolar Affinity. · Aho ER, Wang J, Gogliotti RD, Howard GC, Phan J, Acharya P, Macdonald JD, Cheng K, Lorey SL, Lu B, Wenzel S, Foshage AM · Cell Rep 2019
    PMID 30865883DOI
    What this evidence cannot show

    Cancer cell lines; the specific lines and whether any is KMT2A-rearranged are not stated in the abstract and the full text was not retrieved here. 'Nucleolar stress' is inferred from a molecular signature rather than from nucleolar imaging in this abstract. The WIN site is only one of WDR5's interaction surfaces - the WBM site is separate - so displacement from WIN-dependent loci is not displacement of WDR5 from chromatin generally. Read: abstract only (PubMed efetch, 2026-09-20).

    Evidence that argues against this (1)
    • This result narrows rather than supports the usual rationale. If the WIN site holds WDR5 at only a small cohort of loci, then WIN-site inhibition is not a general disruption of WDR5-containing methyltransferase complexes, and any claim that a WIN inhibitor disables KMT2A-fusion function has to show that the fusion's loci are in that small cohort.
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In MLL-rearranged cancer cells, the dominant response to WDR5 WIN-site inhibition is ribosome subunit attrition and activation of the p53-MDM4 axis.

    Ribosome subunit attrition and activation of the p53-MDM4 axis dominate the response of MLL-rearranged cancer cells to WDR5 WIN site inhibition. · Howard GC, Wang J, Rose KL, Jones C, Patel P, Tsui T, Florian AC, Vlach L, Lorey SL, Grieb BC, Smith BN, Slota MJ · Elife 2024
    PMID 38682900DOI
    What this evidence cannot show

    MLL-rearranged cancer cell lines; the abstract as retrieved does not specify which fusions or lineages, and this entry did not retrieve the full text. The finding is a mechanism-of-action result about an inhibitor, not a localisation measurement - it is included here because 'nucleolar stress' and 'ribosome subunit attrition' are compartment-level phenotypes that a localisation dossier should carry. Read: title and abstract (PubMed efetch, 2026-09-20).

    Evidence that argues against this (1)
    • This is the central counter-argument to the WDR5-as-KMT2A-axis-target rationale in this graph: if the dominant response in MLL-rearranged cells is ribosome attrition and p53-MDM4, then efficacy may be p53-dependent and unrelated to the fusion complex - which predicts failure in p53-mutant or TP53-deleted disease and no selectivity for KMT2A rearrangement per se. This graph already holds concept-a-wdr5-result-is-not-automatically-a-kmt2a-axis-result; this item is its strongest published support.
Connected to