Enhancer usage in KMT2A-rearranged leukemia — occupancy and amplification, not hijacking, and not super-enhancers
also called KEE · H3K79me2 enhancer elements · enhancer hijacking KMT2A · super-enhancers in MLL-r · PROM1 intragenic enhancer
The fusion does not build new control switches for genes. It parks on switches the cell already has — including hidden ones inside genes — and turns their volume up by wiring them more tightly to the gene's start site.
The same thing, for a scientist
Three claims commonly made about enhancers in KMT2A-r do not survive checking. (1) Enhancer hijacking in the classical sense — a translocation juxtaposing a foreign enhancer to an oncogene — has no verified primary literature in KMT2A-r; the mechanism is occupancy and amplification of pre-existing elements. (2) Super-enhancers are not the operative unit: KMT2A::AFF1 spreading sites are explicitly distinct from super-enhancers, and BRD4/Mediator have little impact on the relevant enhancer-promoter contacts. (3) The genuinely fusion-specific element class is the KEE, an H3K79me2/3-marked enhancer whose activity depends on the fusion-deposited mark — including intragenic examples such as the PROM1/CD133 element. The fusion's 3D effect is to drive broad, high-density enhancer-promoter interaction domains through PAF1C and FACT.
What the fusion actually does to enhancers2 deeper
It occupies pre-existing elements. At BCL2, KMT2A::AFF1 binds a BCL-2-specific enhancer consisting of two clusters of H3K27ac at the 3' end of the gene; losing the fusion reduces gene-body H3K79me3 and 3' enhancer H3K27ac 1. The enhancer exists first; the fusion binds and amplifies it.
It creates a mark-dependent enhancer class (KEEs). DOT1L inhibition defined H3K79me2/3-marked enhancers (KEEs) across multiple cell types. In KMT2A::AFF1 cells, H3K79me2/3 is required to maintain chromatin accessibility, histone acetylation and transcription-factor binding specifically at KEEs and not at non-KEE enhancers, and to maintain enhancer-promoter interactions at a subset 2. This is the one enhancer class whose dependence is genuinely a fusion-created dependence, because the mark is a fusion product.
It reaches enhancers inside genes. The clearest single-locus case is PROM1/CD133: aberrant expression is essential for leukemic growth, driven by direct fusion binding at an intragenic H3K79me2/3 KEE, which raises enhancer-promoter contacts between PROM1 and the neighbouring gene TAPT1; in non-expressing cells the locus is PRC2-repressed instead 3. Note how this fits step 1: intragenic unmethylated CpG islands are latent CXXC-readable sites, so 'intragenic enhancer' and 'spreading' are the same phenomenon seen from two directions.
It builds high-density 3D interaction domains. Micro-Capture-C shows fusion binding promotes broad, high-density regions of enhancer-promoter interactions at a subset of key targets. Those enhancers are enriched for PAF1C and FACT, and losing those factors abolishes enhancer-promoter contact — while BRD4 and Mediator have little impact 4.
The super-enhancer claim is actively contradicted
This matters because 'KMT2A-r is a super-enhancer disease' is a widely repeated framing that would justify BET inhibitors as a rational mechanism-based therapy.
The spreading sites that govern DOT1L sensitivity in KMT2A::AFF1 are explicitly distinct from super-enhancers 5. And the enhancer-promoter contacts that carry the fusion's output depend on PAF1C and FACT, not on BRD4 or Mediator 4. Both results point the same way: the operative unit is fusion spreading and elongation-machinery-dependent contact, not H3K27ac-dense super-enhancer formation.
The therapeutic reading is that BET inhibition in KMT2A-r is an empirical, not mechanistic, strategy — consistent with the fact that BET inhibitors kill these cells without releasing the differentiation block (see mechanism-self-renewal-differentiation-block) and that BET resistance arises through WNT/beta-catenin-mediated MYC restoration in exactly this model (see mechanism-rate-limiting-steps-and-compensation).
Enhancer hijacking: a negative result worth recording
This Queen and a dedicated verification worker both searched for primary literature on enhancer hijacking in KMT2A-r and found none. Every enhancer mechanism that is documented in this disease is occupancy or mark-dependent amplification of a pre-existing element in cis on the same locus, not juxtaposition of a foreign enhancer by the translocation.
This is a real biological difference between fusion classes, and it is worth stating explicitly in the graph because the absence is informative:
- KMT2A-r: fusion protein is the effector; the translocation matters because of the protein it makes. Enhancers are read, not relocated.
- Classical enhancer hijacking (e.g. IGH-juxtaposed rearrangements, MYC or GFI1 enhancer hijacking in other tumours): the rearrangement matters because of the regulatory DNA it moves. No fusion protein required.
A cycle-2 caveat: absence of verified evidence is not evidence of absence, and nobody appears to have specifically looked for hijacking in KMT2A-r using modern breakpoint-resolved 3D data. Given that the reciprocal AFF1::KMT2A product exists and that mechanism-reciprocal-kmt2a-fusion-chromatin-opening is already in the graph from the Q1/Q2 beats, a targeted search for enhancer relocation across the der(4)/der(11) junctions is a cheap, genuinely novel analysis.
| chain | mechanism-kmt2a-fusion-causal-chain |
|---|---|
| 3d architecture | broad high-density enhancer-promoter interaction domains dependent on PAF1C and FACT, not BRD4/Mediator |
| enhancer hijacking in kmt2a r | no verified primary literature; mechanism is occupancy and amplification in cis |
| repressive default | PRC2 occupies the PROM1 locus when silent |
| therapeutic reading | BET inhibition in KMT2A-r is empirical rather than mechanism-based |
element classes
[
{
"name": "KEE (H3K79me2/3 enhancer element)",
"fusion_created_dependency": true,
"example": "PROM1 intragenic element"
},
{
"name": "pre-existing H3K27ac enhancer",
"fusion_created_dependency": false,
"example": "BCL2 3' enhancer (two H3K27ac clusters)"
},
{
"name": "super-enhancer",
"fusion_created_dependency": false,
"status": "explicitly distinct from fusion spreading sites"
}
]- PAF1C and FACT are broadly essential, so disrupting the fusion's 3D contacts via those factors is a general transcription toxicity, not a selective strategy.
- PROM1/CD133 is a normal stem-cell marker across tissues, so CD133-directed immunotherapy carries on-target toxicity to normal stem compartments.
- The PROM1 locus is PRC2-repressed in non-expressing cells, so a PRC2-perturbing therapy could de-repress a stem-cell program the fusion had not yet switched on.
- If BET inhibition is not mechanism-based in this disease, its clinical failures should not be read as evidence about enhancer biology — and its successes elsewhere should not be read as transferable.
- Does enhancer hijacking occur at all in KMT2A-r? Nobody appears to have looked with breakpoint-resolved Hi-C/Micro-C across the der(4) and der(11) junctions. Given the reciprocal AFF1::KMT2A product already in this graph, this is a cheap and genuinely novel analysis.
- Are KEEs a fusion-specific vulnerability that could be exploited without inhibiting DOT1L globally — for example by targeting the reader that interprets H3K79me2 at those elements? No H3K79me2 reader is firmly established, which is itself a notable gap.
- PAF1C and FACT are required for the fusion's 3D contacts but are broadly essential. Is there a fusion-selective surface on PAF1C (the CXXC-flanking KMT2A-PAF1 interface) that separates the two?
- Is CD133/PROM1 expression a stratifier for CD133-directed immunotherapy in KMT2A-r, and does its PRC2-repressed state in non-expressors mean it can be pharmacologically induced — or lost under selection?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
KMT2A::AFF1 binds a BCL-2-specific enhancer composed of two clusters of H3K27ac at the 3' end of the gene; loss of fusion activity reduces both gene-body H3K79me3 and 3' enhancer H3K27ac, and the fusion directly activates BCL2 and MCL1 while repressing BIM via CBX8/PRC1.
MLL-AF4 binds directly to a BCL-2 specific enhancer and modulates H3K27 acetylation · Godfrey L, Kerry J, Thorne R, et al. (Milne TA, senior author) · Experimental Hematology 47:64-75 2017What this evidence cannot show
Capture-C plus knockdown in t(4;11) cell lines; the enhancer is pre-existing and modulated rather than created, and the finding is single-locus.
Evidence that argues against this (1)
- A promoter-centric account of the same biology exists and is not reconciled with the enhancer account. Okuda et al. showed that two targeting modules - non-methylated CpG binding and H3K36me2/3 binding - are necessary and sufficient for KMT2A-fusion target recognition, and that the fusion activates *previously active CpG-rich promoters* by recruiting coactivators there. On that model the 3' BCL2 H3K27ac clusters are a consequence of the gene already being transcribed rather than an enhancer the fusion selects. Okuda H et al. 2014
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
DOT1L inhibition defines H3K79me2/3-marked enhancer elements (KEEs); in KMT2A::AFF1 cells H3K79me2/3 is required to maintain chromatin accessibility, histone acetylation and transcription-factor binding specifically at KEEs but not at non-KEE enhancers, and to maintain enhancer-promoter interactions at a subset of KEEs.
DOT1L inhibition reveals a distinct subset of enhancers dependent on H3K79 methylation · Godfrey L, Crump NT, Thorne R, et al. (Milne TA, senior author) · Nature Communications 10(1):2803 2019What this evidence cannot show
KEE is an operationally defined class from a small number of t(4;11) lines; causality from mark loss to accessibility loss rests on co-kinetics under inhibitor rather than on a separation-of-function mutant.
Evidence that argues against this (1)
- KEEs are defined by a mark the fusion deposits, so the class is partly circular: an element is a KEE because it carries H3K79me2/3, and H3K79me2/3 is deposited across transcribed bodies. Guenther et al. had already shown that KMT2A::AFF1-bound regions carry trithorax- and DOT1-catalysed modifications *extending across large domains* at stem-cell-regulator genes - a domain-spreading picture in which 'enhancer' and 'gene body' are not cleanly separable. Guenther MG et al. 2008
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Aberrant PROM1/CD133 expression is essential for leukemic growth and is activated by direct KMT2A::AFF1 binding at an intragenic H3K79me2/3 enhancer element, increasing enhancer-promoter interactions between PROM1 and the nearby gene TAPT1; in PROM1-non-expressing cells the locus is repressed by PRC2 binding with reduced TAPT1 expression.
H3K79me2/3 controls enhancer-promoter interactions and activation of the pan-cancer stem cell marker PROM1/CD133 in MLL-AF4 leukemia cells · Godfrey L, Crump NT, O'Byrne S, et al. (Roy A, Milne TA, senior authors) · Leukemia 35(1):90-106 2021What this evidence cannot show
Cell lines plus primary expression correlation; the dual PROM1/TAPT1 regulation is a single-locus mechanism whose generality is asserted rather than demonstrated across targets.
Evidence that argues against this (1)
- PROM1/CD133 is one locus and the evidence is knockdown in lines, where "essential for leukemic growth" means growth of that line in culture. The increased PROM1-TAPT1 contacts are measured by chromosome conformation, which reports contact frequency in a population and cannot show that the intragenic element acts as an enhancer rather than being a by-product of transcription through the gene body. Oncofusion knowledge graph (internal analysis) 2026
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Micro-Capture-C shows KMT2A::AFF1 binding promotes broad, high-density regions of enhancer-promoter interactions at a subset of key targets; these enhancers are enriched for PAF1C and FACT and loss of those factors abolishes enhancer-promoter contact, whereas BRD4 and Mediator have little impact on these contacts.
MLL-AF4 cooperates with PAF1 and FACT to drive high-density enhancer interactions in leukemia · Crump NT, Smith AL, Godfrey L, et al. (Milne TA, senior author) · Nature Communications 14(1):5208 2023What this evidence cannot show
Highest-resolution 3D data available but applied to a subset of targets in t(4;11) lines; degradation/knockdown of PAF1C and FACT has broad transcriptional consequences, so 'abolishes contact' may be partly a downstream effect of general transcription loss.
Evidence that argues against this (1)
- PAF1C and FACT are general elongation and chromatin-transaction factors. Okuda et al.'s result that coactivator recruitment to already-active CpG-rich promoters is sufficient for KMT2A-fusion function offers an alternative reading of the Micro-Capture-C data: high-density contacts may be a signature of high transcriptional output at those promoters rather than a fusion-built 3D architecture. The entry's own caveat on E4 makes the same point about downstream effects of general transcription loss. Okuda H et al. 2014
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The KMT2A::AFF1 spreading sites that govern DOT1L-inhibitor sensitivity are distinct from super-enhancers, arguing against super-enhancer formation as the operative regulatory unit in KMT2A-rearranged leukemia.
MLL-AF4 Spreading Identifies Binding Sites that Are Distinct from Super-Enhancers and that Govern Sensitivity to DOT1L Inhibition in Leukemia · Kerry J, Godfrey L, Repapi E, et al. (Milne TA, senior author) · Cell Reports 18(2):482-495 2017What this evidence cannot show
'Distinct from super-enhancers' is a genomic-overlap statement in a small number of lines; it does not exclude a role for super-enhancers elsewhere in the same cells, only at the spreading/DOT1L-sensitive sites.
Evidence that argues against this (1)
- "Distinct from super-enhancers" depends on where the super-enhancer cut-off was drawn. Super-enhancers are defined by ranking stitched enhancers on signal and taking those above an inflection point, which is a threshold on a continuum; a set of spreading sites can fall just below it and still be the same class of element. The claim is about a classification, not about a biological distinction. eLife 2021
one of four entries already citing the paper this absence missed
the mark that defines the enhancers that entry is about
- related to →KMT2A / MLL1 — the gene that breaks
Counter-account (Okuda 2014, PMID 24465000): KMT2A fusions recruit coactivators to previously active CpG-rich promoters via two chromatin-reading modules, which explains much of the 'enhancer' signal as a consequence of prior promoter activity.
a named, functionally validated distal enhancer module for a KMT2A-relevant oncogene
modifier, not a step - constraint on the elements the chain acts through
KEEs are the one element class whose dependency this disease creates