Experiment: is the primed state in KMT2A::AFF1 actually bivalent? Multi-mark CUT&Tag on primary infant blasts
also called bivalency CUT&Tag experiment · H3K4me3 H3K27me3 co-occupancy in infant KMT2A-r ALL
Everyone says these leukaemia cells keep two sets of instructions half-switched-on at once, but nobody has actually shown the chemical marks that would prove it in this disease. This experiment measures them directly on real patient cells.
The same thing, for a scientist
A chromatin profiling experiment to test whether lineage-decision promoters in primary infant KMT2A::AFF1 blasts carry canonical H3K4me3/H3K27me3 bivalent domains, using low-input CUT&Tag for H3K4me3, H3K27me3, H3K79me2 and H3K36me3 with matched fetal PreProB and fetal LMPP references, before and after menin inhibition. This exists because the bivalency claim is widely repeated for KMT2A-rearranged leukaemia and no primary demonstration in this disease could be verified. The result discriminates between two combination strategies that are otherwise indistinguishable on current evidence.
Why this specific gap is worth closing
What is verified about chromatin in KMT2A::AFF1 leukaemia is the spreading signature: fusion and menin binding extend into gene bodies with elevated H3K79me2/3, disrupted H3K36me3, and unmethylated CpG regions that the retained CXXC domain can read, and spreading-gene expression is preferentially DOT1L-dependent 1. Separately, t(4;11) is one of the extensively promoter-hypermethylated infant ALL subtypes while t(9;11) is not, and the hypermethylation is reversible in vitro with a demethylating agent 2.
What is not verified is whether the lineage-decision promoters that make this disease plastic are held in H3K4me3/H3K27me3 bivalency. That is a specific, measurable claim, and it points at a different drug:
| if priming is... | rational combination partner |
|---|---|
| H3K4me3/H3K27me3 bivalency | PRC2/EZH2 inhibitor to release the repressive arm |
| DNA hypermethylation plus CXXC-readable unmethylated CpG plus H3K79me2/3 spreading | hypomethylating agent or DOT1L inhibitor |
A lab that assumes the first and is wrong runs the wrong trial. This is a two-to-three-month experiment that redirects a multi-year programme.
Design
Material. Primary infant KMT2A::AFF1 diagnostic blasts, ideally 6 or more cases spanning integrative clusters 3. Low-input or single-cell CUT&Tag is required because infant diagnostic material is precious and often cryopreserved.
Marks. H3K4me3, H3K27me3, H3K79me2, H3K36me3, plus an IgG control. Sequential or single-cell CUT&Tag is essential: bulk co-occurrence of H3K4me3 and H3K27me3 in a population can arise from two cell populations each carrying one mark, which is the classic artefact that has inflated bivalency claims elsewhere.
References. Matched normal fetal CD10-negative PreProB and fetal LMPP populations, so that any bivalency found can be scored as leukaemia-specific or as inherited from the normal fetal cell 4. Without the fetal reference the experiment cannot distinguish a fusion-induced state from a developmental one, which is the central question of this beat.
Perturbation. Repeat after 48 to 96 hours of menin inhibition. The key readout is whether myeloid-lineage promoters resolve toward the active arm (which would support hyp-menin-release-lineage-destination) or the repressive arm.
Target loci to score explicitly. Myeloid regulators (CEBPA, SPI1, CSF1R, ITGAM), B regulators (PAX5, EBF1, IGHM, MME/CD10), and the fusion's own core network nodes including RUNX1 5.
| design | Low-input sequential or single-cell CUT&Tag for H3K4me3, H3K27me3, H3K79me2 and H3K36me3 plus IgG control, on 6 or more primary infant KMT2A::AFF1 diagnostic samples spanning integrative clusters, with matched normal fetal CD10-negative PreProB and fetal LMPP references, repeated after 48-96 hours of menin inhibition. |
|---|---|
| readout | Single-cell or sequential co-occupancy of H3K4me3 and H3K27me3 at myeloid regulators (CEBPA, SPI1, CSF1R, ITGAM), B regulators (PAX5, EBF1, IGHM, MME) and KMT2A-AFF1 core network nodes including RUNX1; direction of resolution after menin inhibition. |
| controls | IgG control; matched fetal PreProB and fetal LMPP references to separate leukaemia-specific from developmentally inherited bivalency; sequential or single-cell design to exclude population-mixing artefacts. |
| cost tier | med |
| timeline | 3-6 months given sample and fetal-tissue access |
| decision value | Discriminates a PRC2/EZH2-inhibitor combination strategy from a hypomethylating-agent or DOT1L combination strategy, which current evidence cannot separate. |
tests hypotheses
[ "hyp-menin-release-lineage-destination" ]
- Bulk co-occurrence of H3K4me3 and H3K27me3 can be produced by two single-marked populations, so a badly designed version of this experiment would manufacture a false positive and send the programme toward the wrong drug.
- Primary human fetal tissue access is the rate-limiting resource and restricts this experiment to a small number of laboratories, which is itself a reason the gap persists.
- A negative result leaves the priming mechanism unexplained, which is scientifically useful but harder to fund than a positive one.
- Has anyone already generated H3K27me3 data on primary infant KMT2A-r blasts as part of an unrelated study, so this could be answered from public data before generating any?
- Is the fetal reference strictly necessary, or would a cord blood CD10-negative progenitor be an acceptable and far more accessible substitute, given that fetal and adult PreProB are transcriptionally distinct?
- If bivalency is absent, what does hold the myeloid program in a poised state? DNA methylation alone would not explain the speed of the reported lineage switches.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MLL-AF4 and menin binding spread into gene bodies with elevated H3K79me2/3, disrupted H3K36me3 and unmethylated CpG regions, and spreading-gene expression is preferentially downregulated by DOT1L inhibition; this paper does not report H3K4me3/H3K27me3 bivalency.
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, Tapia M, Blackledge NP, Ma H, Ballabio E, O'Byrne S, Ponthan F, Heidenreich O, Roy A, Roberts I, Konopleva M, Klose RJ, Geng H, Milne TA · Cell Reports 2017What this evidence cannot show
Cell-line-dominated ChIP-seq with limited primary material; spreading is defined by a computational threshold.
Evidence that argues against this (1)
- CUT&Tag is a bulk assay and cannot establish that two marks sit on the same nucleosome. Single-molecule imaging of millions of individual nucleosomes was required to demonstrate genuinely bivalent nucleosomes and to show that combinatorial modification states vary in prevalence with developmental potency. A CUT&Tag experiment that finds H3K4me3 and H3K27me3 over the same locus in a blast population will not distinguish true bivalency from a mixture of cells each carrying one mark, so the experiment as designed cannot answer the question it poses without a single-molecule or sequential arm. Read: abstract only. Shema E et al. 2016
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
t(4;11) and t(11;19) infant ALL are extensively promoter-CpG-island hypermethylated whereas t(9;11) and MLL-wild-type infant ALL resemble normal bone marrow, the degree of hypermethylation influenced relapse-free survival, and zebularine reversed the methylation and induced apoptosis in vitro.
Specific promoter methylation identifies different subgroups of MLL-rearranged infant acute lymphoblastic leukemia, influences clinical outcome, and provides therapeutic options · Stumpel DJPM, Schneider P, van Roon EHJ, Boer JM, de Lorenzo P, Valsecchi MG, de Menezes RX, Pieters R, Stam RW · Blood 2009What this evidence cannot show
2009-era differential methylation hybridisation with lower resolution than current bisulfite sequencing; zebularine data are in vitro only.
Evidence that argues against this (1)
- There is a competing chromatin explanation with a demonstrated causal link to leukaemogenesis. In MLL fusions, whether specific CpGs at the target locus are protected from DNA methylation — and hence whether H3K9me3 is prevented and the locus stays expressed — determines whether leukaemia develops, while locus targeting itself is preserved even by domains that cannot bind the DNA. Promoter hypermethylation in t(4;11) may therefore be the primary epigenetic abnormality rather than a correlate of a bivalent state. Read: abstract only. Risner LE et al. 2013
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Infant KMT2A-rearranged leukaemias separate into five integrative clusters at distinct differentiation stages with different master transcription factors, so a chromatin experiment must sample across clusters rather than pooling cases.
Multi-omics analysis defines highly refractory RAS burdened immature subgroup of infant acute lymphoblastic leukemia · Isobe T, Takagi M, Sato-Otsubo A, Nishimura A, Nagae G, Yamagishi C, Tamura M, Tanaka Y, Asada S, Takeda R, Tsuchiya A, Wang X, Yoshida K, Nannya Y, Ueno H, Akazawa R, Kato I, Mikami T, Watanabe K, Sekiguchi M, Seki M, Kimura S, Hiwatari M, Kato M, Fukuda S, Tatsuno K, Tsutsumi S, Kanai A, Inaba T, Shiozawa Y, Shiraishi Y, Chiba K, Tanaka H, Kotecha RS, Cruickshank MN, Ishikawa F, Morio T, Eguchi M, Deguchi T, Kiyokawa N, Arakawa Y, Koh K, Aoki Y, Ishihara T, Tomizawa D, Miyamura T, Ishii E, Mizutani S, Wilson NK, Gottgens B, Miyano S, Kitamura T, Goyama S, Yokoyama A, Aburatani H, Ogawa S, Takita J · Nature Communications 2022What this evidence cannot show
Cluster membership is inferred by multi-omics projection and would need to be assigned in the new cohort using the published classifier, which may not transfer cleanly.
Evidence that argues against this (1)
- Five integrative clusters were derived by clustering 84 cases on combined RNA, methylation, exome and targeted sequencing, and the number five is a choice made during the analysis rather than a count of things in the world. Differentiation-stage assignment is by similarity to normal reference profiles, so a leukaemia that expresses a fetal programme will be assigned a fetal stage whether or not it is arrested there. As a design constraint the requirement is right and the cluster count it is built on is soft, so "sample across clusters" should be read as "sample across the disease's heterogeneity" rather than as five defined strata. Oncofusion knowledge graph (internal analysis) 2026
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Matched normal fetal reference populations exist and are defined: a CD10-negative PreProB progenitor that is the first B-lymphoid-restricted progenitor of human fetal life and shares transcriptomic signatures with CD10-negative B-progenitor infant ALL blasts.
Discovery of a CD10-negative B-progenitor in human fetal life identifies unique ontogeny-related developmental programs · O'Byrne S, Elliott N, Rice S, Buck G, Fordham N, Garnett C, Godfrey L, Crump NT, Wright G, Inglott S, Hua P, Psaila B, Povinelli B, Knapp DJHF, Agraz-Doblas A, Bueno C, Varela I, Bennett P, Koohy H, Watt SM, Karadimitris A, Mead AJ, Ancliff P, Vyas P, Menendez P, Milne TA, Roberts I, Roy A · Blood 2019What this evidence cannot show
Access to primary human fetal haematopoietic tissue requires specific ethical approval and is available to few laboratories, which is the main feasibility constraint on this experiment.
Evidence that argues against this (1)
- A transcriptionally matched reference population may still be chromatin-mismatched. In human fetal liver and bone marrow, HSC/MPP showed extensive epigenetic but not transcriptional priming, so populations that look equivalent by expression can differ in accessibility; the same study had to refine the sorting strategy to reach about 90% enrichment, implying the conventional gates carry substantial contamination. A chromatin comparison against a transcriptome-matched normal is not a like-for-like control. Read: abstract only. Ranzoni AM et al. 2021
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The KMT2A-AFF1 core regulatory network includes RUNX1-centred feed-forward loops and cascades and a set of nodes shared between the ALL and AML forms, giving a defined list of loci to score for bivalency.
A KMT2A-AFF1 gene regulatory network highlights the role of core transcription factors and reveals the regulatory logic of key downstream target genes · Harman JR, Thorne R, Jamilly M, Tapia M, Crump NT, Rice S, Beveridge R, Morrissey E, de Bruijn MFTR, Roberts I, Roy A, Fulga TA, Milne TA · Genome Research 2021What this evidence cannot show
Inferred network; the locus list is a prediction and should be treated as a prioritisation rather than a ground truth.
Evidence that argues against this (1)
- The locus list may be the wrong unit. Single-cell mapping places KMT2A-r infant blasts in an early lymphocyte precursor state with hybrid myeloid-lymphoid features rather than in a pro-B cell holding two programmes in tension, and concludes that developmental identity cannot be read off transcriptional state. Scoring a shared ALL/AML regulatory core for bivalency assumes a two-programme model that the single-cell data do not require. Read: abstract only. Khabirova E et al. 2022
competing primary epigenetic explanation raised by the counter-evidence pass: protection of specific CpGs from DNA methylation at MLL-fusion target loci is causally required for leukaemogenesis, while locus targeting itself is preserved even by DNA-binding-dead domain swaps
direction of chromatin resolution at myeloid promoters after menin inhibition
closes the unverified bivalency gap in this node
H3R2me2a holds RUNX1 targets in an INTERMEDIATE state and is not on the planned bivalency panel