Oncofusion

start / graph / localization-chromocentres-pericentromeric-kmt2a-r

Mechanismweak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.Q4b-localization

Chromocentres and pericentromeric satellite heterochromatin — no KMT2A evidence, but the AT-hook argument says there should be

also called chromocenter · pericentromeric heterochromatin · satellite DNA · HMGA1 chromocentre · D1 satellite bundling

Chromocentres are dense clumps where the AT-rich repetitive DNA near chromosome centres from different chromosomes bundles together. Other proteins with the same DNA-gripping hooks as KMT2A pile up there and hold the clumps together — but nobody has ever reported KMT2A or its fusions at a chromocentre, and that silence is itself informative.

The same thing, for a scientist

Direct evidence for KMT2A or any KMT2A fusion at chromocentres, pericentromeric or satellite heterochromatin: **none**. Searches for MLL AND pericentromeric heterochromatin, KMT2A AND satellite heterochromatin, and MLL AND chromocenter each returned zero relevant primary studies. The indirect argument, however, is unusually strong and points the other way. Jagannathan, Cummings & Yamashita showed that **multi-AT-hook satellite-DNA-binding proteins — Drosophila D1 and mouse HMGA1 — bundle pericentromeric satellite DNA from heterologous chromosomes into chromocentres**, and that loss of this function causes micronucleation, DNA damage and cell death; they propose this as a conserved function for pericentromeric satellite DNA [E1]. KMT2A carries three AT-hooks (169-180, 217-227, 301-309) [E2] and is therefore a multi-AT-hook nuclear protein of exactly that class. Its absence from chromocentres needs explaining rather than merely noting, and there is a candidate explanation already in this graph: the only direct assay of MLL's own AT-hooks reported binding to **cruciform DNA — structure rather than sequence** [E3], not to AT-rich sequence. A structure-selective tether would not concentrate at AT-rich satellite. If that is right, the chromocentre negative is a prediction confirmed rather than an oversight, and it is the best available independent evidence that MLL's hooks are not conventional AT-sequence readers.

Go deeper
What chromocentres are and why an AT-hook protein should be there1 deeper

Pericentromeric satellite DNA is highly repetitive and AT-rich. Jagannathan et al. found that in Drosophila the multi-AT-hook protein D1 and in mouse cells HMGA1 bind satellite DNA and bundle it from heterologous chromosomes into chromocentres; disrupting this causes micronuclei, DNA damage and death, which is why they argue satellite DNA has a conserved function rather than being junk 1. The mechanism is explicitly multivalency: several AT-hooks on one polypeptide cross-link several satellite arrays.

KMT2A has three AT-hooks in a 140-residue stretch. Structurally it belongs to the same class. HMGA1 and KMT2A are both abundant nuclear AT-hook proteins expressed in haematopoietic cells. There is no a-priori reason KMT2A should be excluded from satellite DNA on motif grounds alone.

Three possible reasons for the negative, with different consequences

(a) Nobody looked. Plausible: chromocentres are not on anyone's KMT2A hypothesis list, and the standard anti-MLL immunofluorescence images were read for 'punctate nucleoplasm' without scoring DAPI-dense foci. A DAPI-dense-focus overlap score on any existing image set is a zero-cost test.

(b) The hooks are structure-selective, not AT-selective. The 1994 EMSA on the MLL AT-hook region reported cruciform binding and explicitly 'structure rather than sequence' 3. Cruciform extrusion requires inverted repeats under supercoiling; monotonous satellite arrays are not the obvious substrate. On this reading the negative is real and expected.

(c) Occupancy is competed away. HMGA1 is present at very high copy number and binds AT-rich DNA avidly. Even an AT-competent KMT2A would lose the competition at satellite while winning at CpG islands, where HMGA1 has no advantage and CXXC does. This predicts that HMGA1 depletion would reveal KMT2A at chromocentres — a one-experiment discriminator between (b) and (c).

Option (b) is the one this graph currently favours, because it also explains the nuclear-matrix/SAR association recorded elsewhere in the beat.

Why this matters beyond bookkeeping

Chromocentre integrity is a genome-stability function: losing the bundling protein produces micronuclei and DNA damage 1. KMT2A-rearranged infant leukemia is the classic quiet genome — very low mutation burden, essentially one driver lesion. If KMT2A had a chromocentre-maintenance role, its heterozygous disruption by the translocation would predict genome instability, and the disease's defining feature is the opposite. The absence of KMT2A from chromocentres is therefore concordant with the low-mutation-burden phenotype recorded in comut-low-mutation-burden-infant-kmt2a-r, and a positive finding here would sit awkwardly with it. That consistency is worth recording, because it means the negative is not merely an unexamined corner: it fits.

Facts and figures
compartmentchromocentre / pericentromeric satellite heterochromatin
evidence class for kmt2anone direct; indirect motif-class argument points toward presence, direct MLL binding data point away
direct evidencefalse
indirect evidence via partnerfalse
indirect evidence via motif classtrue
marker proteins
[
  "HP1 alpha (CBX5)",
  "H3K9me3",
  "DAPI-dense foci",
  "SUV39H1"
]
searches returning zero
[
  "MLL AND pericentromeric heterochromatin",
  "KMT2A AND satellite heterochromatin",
  "MLL AND chromocenter",
  "HMGA2 AT-rich heterochromatin localization"
]
How this could bite us back
  • Any drug designed against the AT-hook-DNA interaction risks hitting HMGA1's chromocentre-bundling function, whose loss causes micronucleation and DNA damage — a genotoxicity that would be unacceptable in infants and would not show up in short-term cell-viability screens.
  • If KMT2A does contribute to chromocentre organisation in some cell states and this has been missed, then degrading the fusion could destabilise pericentromeric heterochromatin in the surviving normal cells that also carry the wild-type allele — a class of toxicity nobody is monitoring.
Still open
  1. Score DAPI-dense-focus overlap on any existing endogenous KMT2A immunofluorescence image set. If KMT2A is excluded from chromocentres, that is independent evidence its AT-hooks are not AT-sequence readers, and it costs nothing.
  2. Does HMGA1 depletion reveal KMT2A at chromocentres? This discriminates 'structurally incapable' from 'competed away' and is a single knockdown plus a co-stain.
  3. Is there any satellite-derived transcript signature in KMT2A-rearranged blasts? Satellite derepression is a general cancer phenomenon; its absence in a quiet-genome leukemia would be worth documenting either way and is a free reanalysis of existing RNA-seq.
  4. Does the fusion's three-AT-hook multivalency contribute avidity at its real targets in a way a single hook would not? The triple-alanine mutant proposed in the AT-hook entry answers this and the chromocentre question at the same time.
Every source behind this entry (3)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Multi-AT-hook satellite-DNA-binding proteins — Drosophila D1 and mouse HMGA1 — bundle pericentromeric satellite DNA from heterologous chromosomes into chromocentres, and loss of this function causes micronucleation, DNA damage and cell death, indicating a conserved function for pericentromeric satellite DNA.

    A conserved function for pericentromeric satellite DNA · Jagannathan M, Cummings R, Yamashita YM · eLife 2018
    PMID 29578410DOI
    What this evidence cannot show

    Drosophila and mouse cells; no human data and no haematopoietic cells. The bundling function is demonstrated for D1 and HMGA1, and transfer of the property to any other multi-AT-hook protein is an inference the authors do not make.

    Evidence that argues against this (1)
    • The analogy from multi-AT-hook satellite bundlers to KMT2A fails at the point where it would have to predict localisation. Yano et al. mapped the determinants of MLL/ALL-1 punctate subnuclear distribution to two short TRITHORAX-conserved elements in the N-terminal region, not to the AT hooks, and interpreted the resulting speckles as complexes of ALL-1 positioned at its TARGET SITES ON CHROMATIN - euchromatic target genes - rather than at heterochromatin. Drosophila D1 and mouse HMGA1 carry roughly ten or more AT hooks and bundle satellite; KMT2A carries three, and its punctate pattern is accounted for by elements that are not the hooks. Read: abstract only. Yano T et al. 1997
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Human KMT2A carries three annotated A.T hook motifs at residues 169-180, 217-227 and 301-309, placing it structurally in the multi-AT-hook protein class.

    UniProtKB Q03164 (KMT2A_HUMAN) — Histone-lysine N-methyltransferase 2A · UniProt Consortium · UniProt Knowledgebase 2026
    What this evidence cannot show

    Motif-matched curated annotation; membership of a motif class does not entail shared binding behaviour, which is precisely the point at issue here.

    Evidence that argues against this (1)
    • Searched and empty, recorded explicitly, and this is the honest state of the entry. No published study localises KMT2A, any KMT2A fusion, or the MLL N-terminal fragment to chromocentres, pericentromeric heterochromatin or satellite DNA, in any species. PubMed searches on 2026-09-18 combining MLL/KMT2A/ALL-1 with pericentromeric, chromocentre/chromocenter, satellite DNA, major satellite, HP1 and constitutive heterochromatin returned nothing. The `weak` grade is therefore generous rather than harsh: the entry records a structural analogy with no observation behind it, and should be read as a hypothesis for a two-colour FISH/immunofluorescence experiment rather than as a localisation claim. No citation is attached because there is nothing to cite: this note is a bounded absence recording searches that returned nothing, not a sourced fact. Uncited and unciteable look identical in this field, so it is said here explicitly (Q42 citation-integrity repair, 2026-09-20).
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The MLL AT-hook region binds cruciform DNA and recognises the structure rather than the sequence of the target DNA, which would not predict accumulation on monotonous AT-rich satellite arrays.

    PMID 7938000DOI
    What this evidence cannot show

    A 1994 gel-shift with a limited probe panel on a bacterially expressed fragment; the inference that this predicts absence from satellite DNA is the graph's, not the paper's, and has never been tested.

    Evidence that argues against this (1)
    • The inference runs the wrong way once the structure is read. This item argues that structure-reading (cruciform) binding predicts absence from monotonous AT-rich satellite arrays - but the protein whose AT-hooks define the motif, HMG-I(Y)/HMGA1, is one of the two multi-AT-hook satellite-bundling proteins in the chromocentre result this entry cites, and Huth et al. show its binding to ordinary B-form AT-rich duplex is exactly a structural function: a central Arg-Gly-Arg plus two accessory modules, one protein bridging two DNA molecules, stabilising B-DNA rather than recognising an unusual conformation. Structure-reading and satellite binding are therefore not alternatives, and a one-to-two protein-to-DNA stoichiometry is a bundling architecture. The 1994 cruciform result does not predict KMT2A's absence from chromocentres; the prediction is the graph's own and remains untested. Huth JR et al. 1997
Connected to