H2AK119ub - the mark every PRC1.1 claim in this graph rests on, named in fifteen entries and written in none until now
also called H2AK119ub · H2AK119ub1 · H2AK119Ub · H2Aub · histone H2A lysine 119 monoubiquitination · ubiquitylated H2A
Polycomb complex 1 silences genes by hanging a small protein tag called ubiquitin on histone H2A at position 119, and that tag - not the better-known H3K27me3 - is the direct product of the machine this graph keeps calling PRC1.1. Fifteen entries here already argue about what happens to this tag when a menin or DOT1L inhibitor is given, and none of them had ever said what the tag is.
The same thing, for a scientist
H2AK119ub1 is the monoubiquitin mark deposited on histone H2A by the RING1A/RING1B (RNF2) catalytic core of every PRC1 complex, canonical and variant alike. It is the direct enzymatic output of PRC1.1 and therefore the correct pharmacodynamic readout for every PRC1.1 claim in this graph, which have until now been written against H3K27me3 by habit. Three results make the distinction load-bearing rather than pedantic. (1) H2AK119ub1 dosage is **largely disconnected from H3K27me3**: complete loss of the PRC1 catalytic core and of H2AK119ub1 has only a limited effect on H3K27me3 [E1]. (2) Genetic depletion of both Ring1a and Ring1b in mice shows that PRC1 activity and H2AK119ub deposition are required for the leukemogenic potential of Hox-activating fusions **and** of non-Hox-activating fusions, by preventing differentiation independently of Cdkn2a [E2]. (3) The two drugs at the centre of this project move the mark in opposite geometries: menin inhibition induces PRC1.1-dependent H2AK119ub at a **subset** of MLL-fusion targets, while DOT1L inhibition produces a **genome-wide** increase [E3]. The mark is also the axis on which menin-inhibitor resistance turns, because the non-canonical menin targets that reactivate on PRC1.1 loss are defined by bivalent co-occupancy of active menin with repressive H2AK119ub [E4]. A DOT1L-to-H2AK119ub1 link has now been reproduced in a second, non-KMT2A disease [E5], and the mark is pharmacologically movable from the deubiquitinase side [E6].
Why an entry for the mark and not just for the enzyme
This graph already has gene-rnf2 (the writer), gene-pcgf1, gene-kdm2b, gene-bcor, gene-bcorl1, gene-skp1 and protein-prc1-1-complex. What it did not have is the measurable quantity, and that is what every proposed experiment in this beat actually assays. A word-boundary count run on 2026-09-20 over all 1,510 node files found H2AK119ub in 15 entries, none of them an entry about the mark. The parallel repressive mark H3K27me3 has had its own entry since cycle 78 (mark-h3k27me3), and the asymmetry matters because the two are not interchangeable: Sijm et al. show that even complete loss of the PRC1 catalytic core, and with it of H2AK119ub1, has only a limited effect on H3K27me3 1. An entry that measures H3K27me3 and concludes something about PRC1 is measuring the wrong molecule.
That is not a hypothetical failure in this project. analogy-brg1-to-ring1b-swap-at-myc-enhancers-inv16 records that the single cleanest published demonstration of a SWI/SNF-to-PRC1 exchange at MYC enhancers assayed H3K27me3, not H2AK119ub, and therefore does not establish which PRC1 flavour did the work.
The two drugs move this mark in different geometries, and that is the whole pharmacodynamic argument
Neville et al. separate the two inhibitors at the level of this mark 3: menin inhibition induces PRC1.1-dependent deposition of H2AK119ub to silence a subset of MLL-fusion targets, whereas DOT1L inhibition produces a genome-wide increase in H2AK119ub. Two drugs with a shared clinical indication produce opposite-shaped chromatin responses in the same currency.
The operational consequence: a bulk western blot for H2AK119ub is an adequate pharmacodynamic assay for a DOT1L inhibitor and is close to useless for a menin inhibitor, because a subset-restricted gain is invisible in bulk. A menin-inhibitor pharmacodynamic assay has to be locus-resolved (CUT&RUN or CUT&Tag with spike-in normalisation) and has to name its loci in advance. mechanism-dot1l-prc1-antagonism states the biology; this entry states what you would have to measure to see it in a patient.
What is not established
No H2AK119ub measurement has been published in KMT2A::AFF1 infant B-ALL. The PRC1.1 literature in this disease family is myeloid (MLL-AF9, KMT2A-r AML) and T-lymphoid; the index fusion of this project is absent from it. protein-prc1-1-complex records the same gap from the complex side and it is not closed here.
No eraser is established for this mark in this disease. BAP1 (in the PR-DUB complex) and USP16 deubiquitinate H2AK119ub in other systems; neither has been measured under a menin or DOT1L inhibitor in a KMT2A-rearranged cell, and this entry deliberately asserts nothing about them beyond naming them as the untested candidates.
The mark's relationship to H3K27me3 is disputed in direction of causation, not only in magnitude. Sijm et al. report the disconnection from the PRC1 side 1; the classical hierarchical model (H2AK119ub recruits PRC2 via JARID2) predicts coupling. mark-h3k27me3 in this graph carries the PRC2 side of the same argument and should be read against this entry, not instead of it.
| mark | H2AK119ub1 |
|---|---|
| histone | H2A |
| residue | K119, C-terminal tail |
| chemistry | monoubiquitination (a ~8.5 kDa protein adduct, not a methyl or acetyl group) - so it is detected by a shifted band, not by a mass difference |
| writer complex | every PRC1 flavour, canonical and variant; PRC1.1 is the variant this project cares about (protein-prc1-1-complex) |
| catalytic stoichiometry | RING1A or RING1B provides the E3 RING domain; a PCGF paralogue sets the flavour (PCGF1 for PRC1.1) |
| stabiliser | gene-usp7 (via TRIM27; USP7 inhibition disassembles PRC1.1 and lowers the mark) |
| relationship to h3k27me3 | largely uncoupled in dosage (PMID 36332031); classically coupled via JARID2-mediated PRC2 recruitment - both on record, not reconciled |
| menin inhibitor effect | PRC1.1-dependent deposition at a SUBSET of MLL-fusion targets (PMID 41634341) |
| dot1l inhibitor effect | genome-wide increase (PMID 41634341); opposite sign reported at PRC1/2-target promoters in GCB-DLBCL (PMID 42213647) |
| pharmacodynamic implication | bulk immunoblot is adequate for a DOT1L inhibitor and inadequate for a menin inhibitor, because a subset-restricted gain does not move the bulk signal |
| named in entries when written | 15 |
| count method | word-boundary regex /\bH2AK119ub/ over all 1,510 node JSON files, run 2026-09-20 |
| measured in kmt2a aff1 infant b all | false |
writers
[ "gene-rnf2", "gene-ring1" ]
candidate erasers untested in this disease
[ "BAP1 (PR-DUB)", "USP16" ]
- Deliberately raising H2AK119ub to help a menin inhibitor would require activating PRC1.1, and PRC1.1 activity in normal haematopoiesis restrains differentiation - the same activity that prevents leukaemic differentiation prevents normal differentiation, so the therapeutic window is not obviously open.
- Conversely, inhibiting PRC1.1 (for example with a USP7 inhibitor) removes the very machine through which a menin inhibitor silences its targets. A USP7 inhibitor and a menin inhibitor given together could be pharmacologically antagonistic at the chromatin level, and nobody has tested that combination.
- Using bulk H2AK119ub as a pharmacodynamic biomarker for a menin inhibitor would generate a confident negative - no change - from a drug that is working, because the reported effect is subset-restricted.
- PRC1.1 loss is already an established menin-inhibitor resistance route in AML models; any intervention that lowers H2AK119ub therapeutically is phenocopying that resistance route.
- Has H2AK119ub ever been measured in a KMT2A::AFF1 infant B-ALL sample or model? A word-boundary search of this graph on 2026-09-20 found no entry citing such a measurement, and the published PRC1.1 leukaemia literature is myeloid and T-lymphoid.
- Which subset of MLL-fusion targets gains H2AK119ub after menin inhibition? Until those loci are named, a locus-resolved pharmacodynamic assay cannot be designed, and the bulk assay cannot see the effect.
- Is there an eraser worth measuring? BAP1 and USP16 are the obvious candidates and neither has been assayed under a menin or DOT1L inhibitor in a KMT2A-rearranged cell.
- Does the genome-wide H2AK119ub gain after DOT1L inhibition occur in normal haematopoietic progenitors at the same exposure? If it does, it is a candidate mechanism for the cytopenias attributed to DOT1L loss, and it would be measurable in the same samples.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
H2AK119ub1 dosage is largely disconnected from H3K27me3: USP7 promotes genomic deposition of H2AK119ub1 by non-canonical PRC1 complexes, and even complete loss of the PRC1 catalytic core and of H2AK119ub1 has only a limited effect on H3K27me3.
USP7 regulates the ncPRC1 Polycomb axis to stimulate genomic H2AK119ub1 deposition uncoupled from H3K27me3. · Sijm A, Atlasi Y, van der Knaap JA, Wolf van der Meer J, Chalkley GE, Bezstarosti K, et al. · Sci Adv 2022What this evidence cannot show
The core experiments are in human cell lines with quantitative proteomics/epigenomics, not in leukaemia and not in a KMT2A-rearranged background. USP7 stabilises ncPRC1.6 strongly and ncPRC1.1 only 'to a lesser extent', so the paper's headline is more about PRC1.6 than about the complex this project cares about. 'Limited effect on H3K27me3' is a magnitude statement whose threshold is set by the assay's dynamic range.
Evidence that argues against this (1)
- The hierarchical-recruitment model predicts coupling, not disconnection: PRC2 reads H2AK119ub through JARID2 and is recruited to ubiquitylated nucleosomes, which is why H3K27me3 has been used as a PRC1 proxy for two decades. Rossi et al. likewise read out PRC1 activity through H2AK119Ub but interpret it inside a Polycomb system where both marks co-occur at Cdkn2a; the disconnection claim is a dosage claim and does not abolish the mechanistic link. Rossi A et al. 2016
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Genetic depletion of the H2A ubiquitin ligase subunits Ring1a and Ring1b in mice shows that PRC1 activity and H2AK119 monoubiquitination are required for the leukemogenic potential of both Hox-activating and non-Hox-activating fusion proteins, by preventing differentiation of leukaemic cells independently of the Cdkn2a locus.
Maintenance of leukemic cell identity by the activity of the Polycomb complex PRC1 in mice. · Rossi A, Ferrari KJ, Piunti A, Jammula S, Chiacchiera F, Mazzarella L, et al. · Sci Adv 2016What this evidence cannot show
Mouse genetics, and it deletes BOTH catalytic subunits, which removes every PRC1 flavour at once - canonical PRC1 and all six variant complexes including PRC1.1. It therefore establishes that H2AK119ub matters and says nothing about which complex deposits the functionally relevant fraction. The direction is also opposite to the menin-inhibitor-resistance literature, where LOSS of PRC1.1 subunits helps the leukaemia.
Evidence that argues against this (1)
- In KMT2A-rearranged AML the sign is reversed for the PRC1.1 flavour specifically: loss of PCGF1 or BCOR does not kill the leukaemia, it makes it RESISTANT to menin inhibition by reactivating non-canonical menin targets including MYC. So 'PRC1 activity is required for leukaemogenesis' (whole-PRC1, mouse, untreated) and 'PRC1.1 loss is a survival route' (PRC1.1-specific, human and mouse, under drug) are both on record and are not reconciled anywhere. Zhou X et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin inhibition induces PRC1.1-dependent deposition of H2AK119ub that silences a subset of MLL-fusion-protein targets, whereas DOT1L inhibition results in a genome-wide increase in H2AK119ub.
DOT1L provides transcriptional memory through PRC1.1 antagonism. · Neville D, Ferguson DT, Heikamp EB, Lai Z, Magor GW, Lam C, et al. · Nat Cell Biol 2026What this evidence cannot show
Read at abstract level only for this entry; the full text was not available through the routes tried on 2026-09-20. The 'subset' is not enumerated in the abstract, so the loci a menin-inhibitor pharmacodynamic assay would have to target are not yet nameable from this source. Cell systems are not stated in the abstract beyond 'conserved across cell types'.
Evidence that argues against this (1)
- Sijm et al. show H2AK119ub1 can be moved substantially without moving H3K27me3, which means a genome-wide H2AK119ub increase after DOT1L inhibition need not produce a genome-wide silencing phenotype - the mark's presence and its transcriptional consequence are separable, and the abstract does not report the transcriptional consequence of the genome-wide gain. Sijm A et al. 2022
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A group of non-canonical menin targets in KMT2A-rearranged AML is defined by bivalent co-occupancy of active menin with repressive H2AK119ub marks, and these targets are typically downregulated after menin inhibition.
Epigenetic regulation of noncanonical menin targets modulates menin inhibitor response in acute myeloid leukemia. · Zhou X, Zhang L, Aryal S, Veasey V, Tajik A, Restelli C, et al. · Blood 2024What this evidence cannot show
Defined by chromatin-focused CRISPR screens plus profiling in human and mouse KMT2A-r AML models; no patient-derived confirmation that this bivalent class exists in vivo, and no lymphoid or infant data. 'Bivalent' here means menin plus H2AK119ub, which is a different usage from the classical H3K4me3/H3K27me3 bivalency this graph uses elsewhere - a reader who conflates the two will reach the wrong experiment.
Evidence that argues against this (1)
- Neville et al. place the H2AK119ub gain AFTER menin inhibition as the silencing mechanism, i.e. the mark is largely drug-induced at these loci; Zhou et al. describe H2AK119ub as already co-occupying the targets before treatment. Whether the repressive mark is pre-existing or drug-deposited at the same class of loci is not settled between the two papers, and it changes whether a baseline H2AK119ub map could predict response. Neville D et al. 2026
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
DOT1L influences the composition of chromatin-bound non-canonical PRC1 complexes and regulates, in part, the deposition of H2AK119 monoubiquitination at promoters co-occupied by H3K27me3 in germinal-centre B-cell-like diffuse large B-cell lymphoma, where DOT1L inhibition causes loss of H2AK119ub1 and derepression of plasma-cell signature genes.
DOT1L Shapes ncPRC1-Target Gene Repression to Maintain Germinal Center B Cell Identity of Diffuse Large B cell Lymphoma. · Gobel C, Niccolai R, Gregoricchio S, de Groot MHP, Kneefel SFK, Kreft M, et al. · Blood 2026What this evidence cannot show
A different disease (GCB-DLBCL), a different partner enzyme (EZH2 co-dependency rather than a KMT2A fusion), and the direction is OPPOSITE to Neville's leukaemia result: here DOT1L inhibition LOSES H2AK119ub1, whereas in MLL-fusion leukaemia it produces a genome-wide INCREASE. Cited here as independent evidence that a DOT1L-to-H2AK119ub link exists outside KMT2A biology, not as confirmation of its sign.
Evidence that argues against this (1)
- The sign conflict is itself the counter-evidence and it is unresolved: Neville et al. report DOT1L inhibition raising H2AK119ub genome-wide through direct biochemical antagonism between H3K79 methylation and PRC1 activity, which predicts a gain, not the promoter-level loss reported in GCB-DLBCL. Either the antagonism is not universal, or the two studies are measuring different compartments (genome-wide bulk versus PRC1/2-target promoters). Neville D et al. 2026
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
USP7 inhibition disassembles PRC1.1 and causes loss of chromatin binding with reduced H2AK119ub, impairs proliferation of AML cells independently of the USP7-MDM2-TP53 axis, and delays MLL-AF9-induced leukaemia in vivo in human xenografts.
The USP7-TRIM27 axis mediates non-canonical PRC1.1 function and is a druggable target in leukemia. · Maat H, Atsma TJ, Hogeling SM, Rodriguez Lopez A, Jaques J, Olthuis M, et al. · iScience 2021What this evidence cannot show
USP7 inhibitors are pleiotropic; the paper explicitly has to exclude the MDM2-TP53 route, which is the standard confound for this target class and is only partly excludable. The MLL-AF9 in vivo result is a delay, not a cure, and is in a xenograft. Reducing H2AK119ub by disassembling PRC1.1 would, on the resistance model, be expected to PROMOTE menin-inhibitor resistance rather than help - the two literatures have not been run against each other.
Evidence that argues against this (1)
- The same USP7-PRC1.1 axis in neuroblastoma is reported to work by reducing H2AK119Ub and inducing differentiation, i.e. PRC1.1 there is the pro-tumour machine whose removal helps; in KMT2A-rearranged AML under a menin inhibitor, PRC1.1 removal is a resistance route. The therapeutic sign of PRC1.1 inhibition is therefore disease- and context-dependent and cannot be carried across. Cmarik EA et al. 2026
bounds how far this mark can be trusted in GSE260742: 2 material disagreements in 68 comparable ratio statistics
same-condition replicate floor for this mark: 15.97% at the peaks, 3.03% at D1
RING1B is the E3 that deposits H2AK119ub1 in every PRC1 flavour
RING1A is the redundant paralogous E3; only double Ring1a/Ring1b loss abolishes the mark
PRC1.1 deposits this mark at non-methylated CpG islands via KDM2B targeting
dosage largely uncoupled (PMID 36332031) despite the classical hierarchical model
menin inhibition induces PRC1.1-dependent deposition at a SUBSET of MLL-fusion targets
- ← causesPinometostat (EPZ-5676) — the DOT1L inhibitor, and the most instructive failure in this field
DOT1L inhibition produces a genome-wide increase
this mark is the measurable output of the antagonism that entry describes
- ← depends onMYC reactivation through PRC1.1 loss — a named, druggable bypass out of menin dependence
the reactivated non-canonical menin targets are defined by bivalent menin/H2AK119ub occupancy
USP7 stabilises ncPRC1 and its inhibition lowers the mark
PRC1 activity and H2AK119ub are required for the leukaemogenic potential of MLL fusions in mice
the assay reads the mark directly rather than substituting H3K27me3, which is what the inv(16) study did
subset-restricted deposition after menin inhibition, against a genome-wide gain after DOT1L inhibition
the currency in which the sign conflict is expressed; compartment must be named before the direction is asserted
destabilising arm of the PRC1.1 stability switch; its activity lowers complex integrity and therefore the mark
measures whether the repressive mark leaves as H3K79me returns, and with what lag
the count that motivated the entry: named in 15 entries, written in none
the series' only spike-in-normalised experiment is an H2AK119ub panel (Extended Data Fig. 4d)
the ligase machinery measured here writes H2AK119ub; the mark itself was not downloaded in this beat
share ratio at the fusion's own sites, stratified by eviction depth
the channel in which the replicate spread was measured
the in-filled mark, graded by eviction depth
BMI1 is a required non-catalytic subunit of the H2AK119ub writer (hPRC1L: Ring1, Ring2, BMI1, HPH); catalysis is RING1B's
the module's substrate list includes histone H2A, so the SAGA eraser and this graph's PRC1.1 writer may share a substrate - untested in leukaemia