Three unnamed things at once: the DOT1L-PRC1.1 deposition names the fusion, the cell types and both inhibitors with concentrations
also called GSE260456 provenance · GSE260742 provenance · DOT1L PRC1.1 system · SGC0946 VTP-50469 doses
An entry on DOT1L and Polycomb says the paper names neither its cell systems, nor its MLL fusion, nor the drugs it used. Three deposited datasets name all three: mouse MLL-AF9 cells plus human MOLM13, MV4;11, K562 and erythroid and stem-cell lines, treated with the DOT1L inhibitor SGC0946 and the menin inhibitor VTP-50469 at stated doses and times.
The same thing, for a scientist
`mechanism-dot1l-prc1-antagonism` carries an unusually dense caveat: *"The cell systems, the MLL fusion used and the inhibitors' identities and concentrations are not stated in the abstract, and 'conserved across cell types' is asserted without the types being named"* [E1]. Every clause is answerable from three GEO series recovered from the paper's data-availability statement. **GSE260456** (180 samples: 168 *Mus musculus*, 12 *Homo sapiens*) and **GSE260742** (158 samples: 121 mouse, 37 human) are dominated by `cell line: MLLAF9` (66 and 107 samples), with human `MOLM13`, `MV4;11`, `K562`, `HUDEP-2` and `mESC` arms; **GSE181829** (71 samples, all human) is a K562 CRISPR-screen and CUT&Tag series that also contains H9 human embryonic stem cells [E2][E3]. The treatment fields name the compounds outright - **SGC0946 at 3 uM and 5 uM** (DOT1L) and **VTP-50469 at 300 nM for 8 and 48 hours** (menin) - together with 24 h, 48 h, 72 h, 96 h, 168 h and *7 days + 72 hrs* exposures [E3]. And *"conserved across cell types"* is now bounded: the deposited cell types are myeloid leukaemia (mouse and human), erythroid (HUDEP-2, K562) and pluripotent (mESC, H9) - **no B-lymphoid sample anywhere**, which is precisely the cell type this project's disease occupies.
Every clause of the caveat, answered
| Clause of the caveat | Resolution | Where |
|---|---|---|
| cell systems not stated | mouse MLLAF9; human MOLM13, MV4;11, K562, HUDEP-2; mESC; H9 hESC | GSE260456, GSE260742, GSE181829 23 |
| MLL fusion not stated | MLL-AF9 dominant (66 + 107 samples); MV4;11 is KMT2A::AFF1 | GSE260456, GSE260742 3 |
| inhibitors not identified | SGC0946 (DOT1L), VTP-50469 (menin) | treatment fields 3 |
| concentrations not stated | SGC0946 3 uM and 5 uM; VTP-50469 300 nM | treatment fields 3 |
| conserved across cell types unbounded | myeloid, erythroid and pluripotent; zero B-lymphoid | 23 |
That is five separate unknowns closed from metadata, in a beat where the typical yield is one. It is the strongest single argument in this project for treating the source does not state as a queued question.
The one thing the deposition does not conserve
The phrase the entry flagged - conserved across cell types - turns out to be both supported and misleading. The deposition genuinely spans an unusual range: two species, leukaemic and non-leukaemic, differentiated and pluripotent. That is a stronger conservation claim than most papers earn.
But the range has a hole exactly where this graph needs it. No B-lymphoid cell appears in any of the three series. MV4;11 carries KMT2A::AFF1 - the index fusion - and is a monocytic AML line, so even the one index-fusion arm is myeloid. A DOT1L-PRC1.1 antagonism conserved across myeloid, erythroid and stem cells is a real and interesting generalisation that says nothing about pro-B blasts, and mechanism-dot1l-prc1-antagonism should say so rather than repeating the unbounded phrase.
| script | none - GEO accession-display text endpoint plus PMC data-availability statement |
|---|---|
| status | run |
| result summary | Five separate unknowns closed: cell systems (mouse MLLAF9, human MOLM13/MV4;11/K562/HUDEP-2, mESC, H9), fusion (MLL-AF9 dominant; MV4;11 is KMT2A::AFF1), inhibitor identities (SGC0946, VTP-50469), concentrations (3-5 uM, 300 nM) and the bound on 'conserved across cell types' (myeloid, erythroid, pluripotent - zero B-lymphoid). |
inputs
[ "PMID 41634341", "GSE181829", "GSE260456", "GSE260742", "PMC12904788" ]
counts — the numbers behind the chart
{
"series": 3,
"samples_total": 409,
"samples_mouse": 289,
"samples_human": 120,
"b_lymphoid_samples": 0,
"cell_types_represented": 3,
"inhibitors_named": 2
}entries affected
[ "mechanism-dot1l-prc1-antagonism" ]
import axes
[ "species: mouse dominant", "lineage: myeloid, erythroid and pluripotent - no B-lymphoid", "fusion: MLL-AF9 dominant, with an MV4;11 KMT2A::AFF1 myeloid arm" ]
- If DOT1L restrains PRC1.1 in normal erythroid and pluripotent cells as this deposition suggests, then DOT1L inhibition releases Polycomb repression in non-leukaemic tissue too, which is an on-target toxicity a leukaemia-only experiment would never surface.
- The deposition pairs SGC0946 with VTP-50469, which is exactly the DOT1L-plus-menin combination this project contemplates; if the transcriptional-memory effect requires the combination, then single-agent menin inhibition may leave a memory state that reconstitutes the programme after the drug stops.
- VTP-50469 is a tool compound at 300 nM for hours; a combination schedule designed from this deposition is designed from an acute exposure and may not describe weeks of clinical menin inhibition.
- Does DOT1L-PRC1.1 antagonism hold in a B-lymphoid cell? The deposition's conservation claim spans myeloid, erythroid and pluripotent cells and contains no B-lineage sample at all.
- The MV4;11 arm is KMT2A::AFF1 in a myeloid cell. Does the antagonism behave the same there as in the MLL-AF9 arms, and if so is the determining variable the fusion or the lineage?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The DOT1L-PRC1.1 paper reports that DOT1L provides transcriptional memory through PRC1.1 antagonism, and its abstract names neither the cell systems, the MLL fusion, nor the inhibitors and their concentrations.
DOT1L provides transcriptional memory through PRC1.1 antagonism. · Neville D, Ferguson DT, Heikamp EB, Lai Z, Magor GW, Lam C, Dobbs OG, Levina V, Knezevic K, The JJ, Alex S, Suits SC, Rumler B, Uckelmann M, Talarmain L, Lam EYN, Perkins AC, Armstrong SA, Bell CC, Davidovich C, Gilan O · Nat Cell Biol 2026What this evidence cannot show
Identifiers taken from the NCBI esummary record fetched on 2026-09-19. Full text read from Europe PMC PMC12904788 (open access) on the same date; it names MLLAF9 116 times, MOLM13 62, MV4;11 24 and K562 14, consistent with the depositions.
Evidence that argues against this (1)
- H3K79 methylation also controls enhancer-promoter contacts at loci such as PROM1/CD133 independently of Polycomb, so a memory function attributed to PRC1.1 antagonism competes with a direct architectural role for the mark that this deposition does not separate from it. Godfrey L et al. 2021
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
GEO series GSE260456 contains 180 samples (168 Mus musculus, 12 Homo sapiens) and GSE260742 contains 158 samples (121 Mus musculus, 37 Homo sapiens), dominated by 'cell line: MLLAF9' with human MOLM13, MV4;11, K562, HUDEP-2 and mESC arms annotated 'cell type: Myeloid', 'Erytheroid' or 'Stem'.
DOT1L provides transcriptional memory through PRC1.1 antagonism (GSE260456) · GEO submitters · NCBI Gene Expression Omnibus 2026What this evidence cannot show
Read on 2026-09-19 from the GEO accession-display text endpoint and tallied by characteristic string across both series. Submitter-entered, with the misspelling 'Erytheroid' in the cell-type field. Counts of a characteristic string do not partition samples cleanly because a sample may carry several; the organism counts are exact and the cell-line counts are the number of samples bearing that string. Both accessions were recovered from the paper's data-availability statement; elink returns no GEO link for this PMID.
Evidence that argues against this (1)
- A deposition spanning many cell types can create an impression of generality that the experiments do not support if most arms are shallow; sample counts per cell line here range from 8 to 107, so the conservation claim rests very unevenly across the types it spans. Rogic S et al. 2026
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
The treatment characteristics in GSE260456 and GSE260742 name the DOT1L inhibitor SGC0946 at 3 uM and 5 uM and the menin inhibitor VTP-50469 at 300 nM with 8-hour and 48-hour exposures, alongside 24, 48, 72, 96 and 168-hour and '7 days + 72 hrs' timepoints; sample titles include 'NTVKO PCGF1 KO VTP50469 48h 300nM' and 'HUDEP-2 Control - SGC0946 5uM H2AK119ub'.
DOT1L provides transcriptional memory through PRC1.1 antagonism [CUT and RUN] (GSE260742) · GEO submitters · NCBI Gene Expression Omnibus 2026What this evidence cannot show
Read on 2026-09-19 from the GEO accession-display text endpoint. Submitter-entered free text in sample titles rather than structured dose fields, so the dose-to-sample mapping is read off title strings. VTP-50469 is a tool menin inhibitor and not the clinical agent this project tracks, so a dose recovered here does not transfer to revumenib exposure.
Evidence that argues against this (1)
- VTP-50469 at 300 nM for 8 to 48 hours is a short, tool-compound exposure; conclusions about what menin inhibition does to a chromatin mark on that schedule need not describe what weeks of clinical revumenib do, and this graph has already had to separate the two elsewhere. Soto-Feliciano YM et al. 2023
closes five clauses of the entry's caveat, including both inhibitor identities and concentrations
SGC0946 at 3-5 uM across mouse MLLAF9 and human myeloid, erythroid and pluripotent cells
the deposition pairs DOT1L inhibition with VTP-50469 menin inhibition at 300 nM
H3K79me2 measured against H2AK119ub across cell types with no B-lymphoid arm
resolved source in the unstated-system worklist