A second disease finds the same DOT1L-to-non-canonical-PRC1 axis in the same year - and the mark moves the opposite way
also called DOT1L ncPRC1 in GCB-DLBCL · the sign conflict in the DOT1L-PRC1 axis
The 2026 leukaemia paper that this project's whole downstream story rests on says that blocking DOT1L makes a silencing mark go up everywhere. A second 2026 paper, in lymphoma, finds the same two machines wired together - and reports the mark going down. Both cannot be a general rule.
The same thing, for a scientist
`mechanism-dot1l-prc1-antagonism` rests on a single paper [E1]. An independent group, publishing in the same year in a different malignancy, reaches the same **axis** by a different route and reports the **opposite sign**. Göbel et al. [E2] ran an EZH2-inhibition-anchored genome-wide CRISPRi screen in germinal-centre B-cell-like DLBCL, recovered multiple non-canonical PRC1 components (USP7, KDM2B, RING1, PCGF1), and found that **DOT1L influences the composition of chromatin-bound ncPRC1 complexes and regulates, in part, H2AK119ub1 deposition** at promoters co-occupied by H3K27me3. Their DOT1L inhibition causes **loss** of H2AK119ub1 at those promoters, derepressing plasma-cell signature genes. Neville et al. report DOT1L inhibition producing a **genome-wide increase** in H2AK119ub, driven by direct biochemical antagonism between H3K79 methylation and PRC1 activity. **The axis replicates; the sign does not.** Two readings are available and this entry does not choose between them: either the antagonism is not universal, or the two studies measure different compartments - bulk genome-wide against PRC1/2-target promoters - in which case a global gain and a local loss can coexist. Either way, **this graph should stop treating 'DOT1L inhibition raises H2AK119ub' as a general statement**, and a beat citing the antagonism should say which compartment it means.
Why this matters more than a routine discrepancy
mechanism-dot1l-prc1-antagonism is cited across 14 entries in this graph and is the mechanistic spine of the project's account of what a DOT1L inhibitor does. It rests on one paper and, as of 2026-09-20, had no independent corroboration recorded anywhere here. The Göbel paper is the first independent work on the same axis to appear, and its arrival is unambiguously good for the axis and awkward for the sign.
What replicates, and it is substantial. An independent group, in a different malignancy, in a different lineage, with an unbiased genome-wide screen rather than a candidate approach, recovers non-canonical PRC1 as the partner of DOT1L and measures H2AK119ub1 as the currency. The screen names four of the components this graph has entries for - USP7, KDM2B, RING1, PCGF1 - which is close to the PRC1.1 roster written in protein-prc1-1-complex. That is a strong, independent argument that the DOT1L-ncPRC1 connection is real biology and not a peculiarity of MLL-fusion leukaemia.
What does not replicate. The direction. A mechanism whose sign flips between two diseases is not yet a mechanism you can dose a patient on.
The two readings, and what would decide between them
Reading 1 - compartment. Neville et al. report a genome-wide increase; Göbel et al. report loss at PRC1/2-target promoters co-occupied by H3K27me3. These are different measurements on different subsets, and a redistribution - mark leaving strongly-marked promoters and appearing broadly elsewhere - would satisfy both descriptions simultaneously. This is the more parsimonious reading and it is testable in one experiment: spike-in-normalised H2AK119ub CUT&RUN under DOT1L inhibition, reported both as total signal and as signal at H3K27me3-co-occupied promoters, in one cell system.
Reading 2 - the antagonism is not universal. Neville et al. attribute the gain to direct biochemical antagonism between H3K79 methylation and PRC1 activity, which is a property of the chemistry and should not be lineage-restricted. But they also report the axis as 'conserved across cell types' on evidence not visible in the abstract, and Göbel et al. add an indirect route that would run the other way: they identify USP7 as a potential direct target of DOT1L, and USP7 is required for ncPRC1 integrity 3. If DOT1L inhibition lowers USP7, it lowers PRC1 integrity and therefore H2AK119ub - a second, opposing arm within the same axis. Whichever arm dominates could differ by cell type, which would produce exactly the observed sign flip.
Neither reading is established and this entry asserts neither. What it asserts is that the graph's blanket statement is now unsafe.
| mechanism class | independent corroboration with a sign conflict |
|---|---|
| axis replicated | DOT1L to non-canonical PRC1, with H2AK119ub1 as the currency, in a second malignancy by an independent group using an unbiased genome-wide screen |
| reading 1 compartment | a redistribution - loss at strongly marked promoters, gain broadly elsewhere - satisfies both descriptions; testable by spike-in-normalised H2AK119ub CUT&RUN reported both globally and at H3K27me3-co-occupied promoters in one system |
| reading 2 two opposing arms | direct H3K79me-PRC1 antagonism raises the mark; an indirect DOT1L->USP7->ncPRC1-integrity arm lowers it; whichever dominates could differ by cell type |
| what the graph should stop saying | 'DOT1L inhibition raises H2AK119ub' without naming the compartment |
| entries affected | mechanism-dot1l-prc1-antagonism and the 14 entries citing PMID 41634341 as of 2026-09-20 |
| independent corroboration before this entry | none recorded in this graph |
components recovered by the independent screen
[ "USP7", "KDM2B", "RING1", "PCGF1" ]
sign conflict
{
"kmt2a_r_leukaemia": "DOT1L inhibition -> genome-wide INCREASE in H2AK119ub (PMID 41634341)",
"gcb_dlbcl": "DOT1L inhibition -> LOSS of H2AK119ub1 at H3K27me3-co-occupied promoters, with derepression (PMID 42213647)"
}- This graph's account of DOT1L inhibitor pharmacodynamics is built on one paper's direction of effect, and an independent paper in the same year reports the opposite direction. Any pharmacodynamic assay designed to look for an H2AK119ub increase could be looking for the wrong sign.
- If both arms are real and oppose each other, the net effect of a DOT1L inhibitor on Polycomb marking is dose- and context-dependent in a way that no single-timepoint biomarker would capture.
- The corroboration is genuinely reassuring about the axis and could easily be cited as reassurance about the mechanism as stated here, which it is not.
- If DOT1L inhibition lowers ncPRC1 integrity in some contexts, it would phenocopy the PRC1.1-loss resistance lesion and could antagonise a co-administered menin inhibitor - the reverse of the synergy the field expects from that combination.
- Is the conflict a compartment artefact? One spike-in-normalised experiment reporting both global and promoter-level H2AK119ub under DOT1L inhibition in a single system would answer it.
- Does DOT1L sustain USP7 expression in KMT2A-rearranged leukaemia as it appears to in GCB-DLBCL? If it does, the axis has two arms of opposite sign in the disease this project cares about.
- Which cell types did Neville et al. test for the claimed conservation? The abstract asserts conservation without naming them, and the answer determines whether GCB-DLBCL is an exception or a counterexample.
- Does the plasma-cell-differentiation outcome in DLBCL have a counterpart in KMT2A-rearranged disease? Both stories end in differentiation toward an anti-proliferative state, which may be the more robust shared feature than the direction of the mark.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin inhibition induces PRC1.1-dependent deposition of H2AK119ub silencing a subset of MLL-fusion targets, whereas DOT1L inhibition results in a genome-wide increase in H2AK119ub arising from progressive loss of H3K79 methylation and driven by direct biochemical antagonism between H3K79 methylation and PRC1 activity, a crosstalk the authors state is conserved across cell types.
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
Abstract-level. 'Genome-wide increase' does not specify normalisation, and an H2AK119ub measurement without spike-in cannot report a global change at all - so the claim's compartment is not determinable from the abstract. 'Conserved across cell types' is asserted without the cell types being named in the abstract.
Evidence that argues against this (1)
- An independent group in GCB-DLBCL reports DOT1L inhibition causing LOSS of H2AK119ub1 at PRC1/2-target promoters, with derepression of the genes that lose it. If the antagonism were a direct chemical property conserved across cell types, that observation should not occur. Gobel C et al. 2026
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In GCB-DLBCL, an EZH2-inhibition-anchored genome-wide CRISPR interference screen identified non-canonical PRC1 components including USP7, KDM2B, RING1 and PCGF1; DOT1L influences the composition of chromatin-bound ncPRC1 complexes and regulates in part the deposition of H2AK119ub1 at promoters co-occupied by H3K27me3, and DOT1L inhibition causes loss of H2AK119ub1 at those promoters with 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
Abstract-level, a different malignancy and lineage, and the anchoring perturbation is EZH2 inhibition rather than a menin or DOT1L inhibitor. The H2AK119ub1 regulation is stated as 'in part', and the promoters studied are defined by H3K27me3 co-occupancy - a restricted class that is explicitly not the genome. Graded moderate for that reason: it is a strong result about a defined subset and a weak basis for a global statement.
Evidence that argues against this (1)
- Neville et al. report the opposite direction from a mechanism they attribute to direct biochemical antagonism between H3K79 methylation and PRC1 activity, which is a chemistry-level claim that should not be lineage-restricted. The two results cannot both be general and this entry does not decide which is the special case. Neville D et al. 2026
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
USP7 is an integral component of non-canonical PRC1.1 and is required for its integrity: USP7 inhibition causes PRC1.1 disassembly, loss of chromatin binding and reduced H2AK119ub.
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
Cited here only to supply the second, opposing arm of the proposed reconciliation - if DOT1L sustains USP7 and USP7 sustains ncPRC1, then DOT1L inhibition lowers H2AK119ub by an indirect route that runs against the direct antagonism. That chain has never been tested end to end and this entry presents it as a candidate explanation, not a mechanism.
Evidence that argues against this (1)
- The DOT1L-to-USP7 link is described by its own authors as USP7 being a 'potential direct target of DOT1L' in GCB-DLBCL, and no such link has been reported in leukaemia. The indirect arm may simply not exist in the MLL-fusion setting, in which case there is nothing to oppose the direct antagonism and the sign conflict needs a different explanation. Gobel C et al. 2026
replicates the axis independently in a second disease and reports the opposite sign for the mark
DOT1L shapes chromatin-bound ncPRC1 composition in GCB-DLBCL as well as in MLL-fusion leukaemia
the currency in which the sign conflict is expressed; compartment must be named before the direction is asserted
the proposed opposing arm: DOT1L sustains USP7, USP7 sustains ncPRC1 integrity