The LEDGF hole, measured: every published LEDGF-removal experiment removes it BEFORE or AT transformation, and none treats an established leukaemia in an animal
also called LEDGF hole · PSIP1 removal prevention versus treatment
Scientists have shown many times that leukaemia cells cannot get started if you take LEDGF away first. Nobody has ever taken LEDGF away from a leukaemia that already exists in a living animal, which is the only version of the experiment that tells you whether a LEDGF drug would help a patient.
The same thing, for a scientist
This project's coverage ledger has carried a standing claim: no published experiment removes LEDGF from an *established* human KMT2A::AFF1 leukaemia. Worked here for the first time, the claim survives, and the honest boundary is sharper than the ledger drew it. Published LEDGF-removal experiments in KMT2A-rearranged systems split cleanly. **Prevention** - LEDGF absent before or at the moment of transformation - covers the conditional Psip1 knockout bone-marrow transplant (PMID 29084774), the IBM2 point mutants (PMID 25305204) and the F148A/F151A fusion mutant (PMID 29084774); all are unambiguously positive. **Treatment of an already-transformed population** exists only *in vitro*: dominant-negative LEDGF fragments (PMID 23318960), the CP65 cyclic peptide in primary murine MLL-AF9 blasts (PMID 25082813), shRNA in MLL-AF9-transformed HSPCs (PMID 27154821) and knockdown in human MLLr lines (PMID 33795806). **Treatment of an established leukaemia in an animal: zero.** `(PSIP1 OR LEDGF OR DFS70) AND (xenograft OR PDX OR NSG) AND (leukemia)` returns 2 records and neither removes LEDGF. The narrower KMT2A::AFF1 claim is starker: the whole PubMed intersection of the LEDGF alias set with AFF1/AF4 is **2 records**, one of them a review.
The inventory, split by disease stage
| Study | System | LEDGF removed how | Stage | Readout |
|---|---|---|---|---|
| El Ashkar 2018 1 | murine lin- cells, MLL-ENL, BMT | Vav-iCre conditional KO, excised from early embryonic HSC | before transduction | CFU; survival to 285 d |
| Murai 2014 2 | murine MLL-AF9 | IBM2 point mutation in the fusion | at transformation | transformation abolished |
| El Ashkar 2018 1 | murine lin- cells | F148A/F151A mut.MLL-ENL | at transformation | no colonies |
| Mereau 2013 3 | human + murine MLL-fusion transformed cells | dominant-negative eGFP-LEDGF(325-530), LEDGF(375-386), LEDGF(424-435) | already transformed in vitro; the in vivo arm transduces before BMT | clonogenic growth; latency |
| Cermakova 2014 4 | primary murine MLL-AF9 blasts | CP65 cyclic peptide; LEDGF binding mutants | already transformed, in vitro | clonogenic growth |
| Zhu 2016 5 | murine MLL-AF9/AF10/ENL transformed HSPCs | LEDGF shRNA | already transformed, in vitro | ChIP occupancy, expression |
| Canella 2022 6 | human MLLr cell lines | LEDGF knockdown | already transformed, in vitro | chemosensitivity, apoptosis |
| nothing | established leukaemia in an animal | - | - | - |
The in vivo arm of Mereau 2013 is the closest miss and it is still prevention: the dominant-negative fragment is transduced into the cells before transplantation, and the reported effect is an increase in latency to disease development, which is what a prevention experiment measures 3.
The counted queries, with translations
Run 2026-09-20 against live PubMed. Aliases searched: PSIP1, LEDGF, LEDGF/p75, DFS70, lens epithelium-derived growth factor. p75 alone was deliberately not used in the counted queries - it collides with p75NTR/NGFR and returns neurobiology (a 2000 J Neurosci paper on p75LNTR came back in the loose sweep), which is the alias-precision failure this project has logged before. p52 was excluded for a different reason: it is a distinct isoform lacking the IBD, so a p52 hit is not a p75 hit (concept-psip1-p52-is-not-p75).
| Query | Count | Note |
|---|---|---|
| LEDGF-alias set AND (KMT2A OR MLL) AND leukaemia | 34 | the whole literature of this beat |
| LEDGF-alias set AND (xenograft OR PDX OR NSG) | 11 | none is a leukaemia LEDGF-removal xenograft |
| LEDGF-alias set AND (xenograft OR PDX OR NSG) AND leukaemia | 2 | PMID 41279447 (a menin+LSD1 PDX study that does not remove LEDGF) and PMID 27154821 |
| LEDGF-alias set AND (AFF1 OR AF4 OR MLL-AF4) | 2 | PMID 28076791 and one review |
| LEDGF-alias set AND (revumenib OR ziftomenib OR bleximenib OR enzomenib OR SNDX-5613 OR KO-539 OR JNJ-75276617 OR MI-503 OR MI-3454 OR VTP-50469) | 0 | translation checked, see below |
The zero was checked against querytranslation before being recorded. It translates to a clean disjunction of drug names with no [UID] term and no ignored phrase, and the warning list was {'phrasesignored': [], 'quotedphrasesnotfound': [], 'outputmessages': ['No items found.']} - an honest empty set, not a parser failure. No PubMed-indexed paper mentions LEDGF or PSIP1 alongside any named menin inhibitor. A bioRxiv preprint does, and it is handled in mechanism-ledgf-occupancy-falls-under-menin-inhibition.
Why the distinction is not pedantry
Every menin inhibitor in the clinic is given to a patient whose leukaemia already exists. A target validated only in prevention has been validated in the one setting no drug will ever occupy. This graph already holds the worked precedent: menin was shown to be required for both initiation and maintenance, and it is the maintenance arm that made it a drug target. LEDGF has only ever had the initiation arm done in an animal.
There is a mechanistic reason as well as a translational one. In prevention experiments LEDGF is absent while the fusion is trying to build its programme. In treatment the programme already exists, is reinforced by HOXA9/MEIS1 output and by marks already deposited, and may no longer need the factor that built it. mechanism-ledgf-dispensable-for-fusion-chromatin-retention gives a published reason to expect exactly that.
Bounded rejection: what was read and not used
- PMID 22334021 (Mol Ther 2012), LEDGF/p75 knockdown in NSG mice: the only LEDGF-removal xenograft in the literature, and it is HIV gene therapy. It establishes that LEDGF-depleted human CD34+ cells engraft NSG mice, which is a feasibility datum for the proposed experiment, and nothing about leukaemia.
- PMID 27013196 (JPO2/CDCA7L and LEDGF in medulloblastoma) and PMID 34685704 (IBD interactome in docetaxel-resistant prostate cancer): LEDGF-removal experiments in established tumours, in the wrong disease. They are why the absence is stated about leukaemia and not about cancer.
- PMID 39105355, 34165568, 30787394, 22006329, 19934313, 19724686, 17363574: the remaining xenograft-query hits - renal carcinoma, HIV integration mapping, PSIP1 editing for HIV protection, STAT3 splicing, ovarian lymphangiogenesis, VEGF-C regulation, lysosomal stability. All off-beat.
- PMID 41778782, 39965614, 37675289, 36673033, 33749163: LEDGF papers in HIV integration site mapping, colorectal cancer, pan-cancer/renal, autoantibody diagnostics and lymphoma prognostic signatures. On the protein, not on this beat.
| claim tested | No published experiment removes LEDGF from an established human KMT2A::AFF1 leukaemia. |
|---|---|
| verdict | survives, and generalises: none removes LEDGF from an established KMT2A-rearranged leukaemia in any animal, of any fusion, of any species |
| date run | 2026-09-20 |
| absence class | never-measured |
| absence class basis | Every published LEDGF-removal experiment removes it before or at transformation; the established-disease experiment has not been run. |
| absence class assigned | 2026-09-21 (cycle 47, king) |
queries run — the numbers behind the chart
[
{
"query": "(PSIP1 OR LEDGF OR \"LEDGF/p75\" OR p75 OR DFS70 OR \"lens epithelium-derived growth factor\") AND (KMT2A OR MLL) AND (leukemia OR leukaemia)",
"count": 34
},
{
"query": "(PSIP1 OR LEDGF OR \"LEDGF/p75\" OR DFS70) AND (xenograft OR \"patient-derived\" OR PDX OR NSG OR NSGS)",
"count": 11
},
{
"query": "(PSIP1 OR LEDGF OR DFS70) AND (xenograft OR PDX OR NSG) AND (leukemia OR leukaemia)",
"count": 2
},
{
"query": "(PSIP1 OR LEDGF OR DFS70) AND (AFF1 OR AF4 OR \"MLL-AF4\" OR \"KMT2A::AFF1\")",
"count": 2
},
{
"query": "(PSIP1 OR LEDGF OR DFS70) AND (revumenib OR ziftomenib OR bleximenib OR enzomenib OR \"SNDX-5613\" OR \"KO-539\" OR \"JNJ-75276617\" OR \"MI-503\" OR \"MI-3454\" OR \"VTP-50469\")",
"count": 0
}
]aliases searched
[ "PSIP1", "LEDGF", "LEDGF/p75", "DFS70", "lens epithelium-derived growth factor" ]
aliases excluded
[
{
"alias": "p75",
"reason": "collides with p75NTR/NGFR; returns neurobiology"
},
{
"alias": "p52",
"reason": "a different isoform lacking the IBD entirely"
}
]- Validating a target only in prevention produces a confident target dossier that predicts nothing about the clinical setting, and this axis has carried one for a decade.
- If the inducible-knockdown xenograft returns negative it will not be a clean negative: an shRNA achieving 70% knockdown is not a drug occupying 95% of a pocket, and a null result would be attributable to residual protein.
- LEDGF-depleted human CD34+ cells engraft NSG mice, which makes the proposed experiment feasible but also removes the easy safety argument, because engraftment is not long-term competitive repopulation.
- Does removing LEDGF from an established KMT2A::AFF1 leukaemia in a xenograft reduce leukaemic burden, or does the programme run without it? The experiment is written as exp-psip1-inducible-knockdown-established-kmt2a-aff1-xenograft.
- The whole in vivo requirement rests on MLL-ENL and MLL-AF9 in mice. Is the LEDGF requirement partner-dependent? Nothing in the literature tests KMT2A::AFF1 in an animal at all.
- Canella 2022 used human MLLr lines that cannot be identified from the abstract. Which lines, and did any carry t(4;11)? One paywalled methods section decides whether the in-vitro half of this absence is narrower than stated here.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The conditional Psip1 knockout leukaemia experiment removes LEDGF before the fusion is introduced: bone marrow was harvested from Psip1 Vav-iCre mice, lineage-depleted, then transduced with MLL-ENL and transplanted; recipients of Psip Fl/Fl.MLL-ENL cells succumbed to leukaemia within 60 to 85 days while recipients of Psip Vav/Vav.MLL-ENL cells showed no sign of disease at 285 days despite PCR-confirmed fusion expression. No arm of the study deletes Psip1 in an animal that already has leukaemia.
LEDGF/p75 is dispensable for hematopoiesis but essential for MLL-rearranged leukemogenesis. · El Ashkar S, Schwaller J, et al. · Blood 2018What this evidence cannot show
Read in full via the PMC HTML route (route 4; Europe PMC fullTextXML returned HTTP 500 and efetch db=pmc returned a document with no <body>), 61,618 characters of stripped text. The model is murine MLL-ENL in a myeloid-biased BMT, not human KMT2A::AFF1. Vav-iCre excises from early embryonic HSC, so the cells were never LEDGF-sufficient at any point.
Evidence that argues against this (1)
- The same paper's haematopoiesis arm is not a clean negative either: Psip1 knockout mice showed an approximately twofold decrease in peripheral white cell counts, perturbed HSC and progenitor numbers and impaired serial colony formation. 'Dispensable for steady-state hematopoiesis' summarises a phenotype that is real but non-lethal. El Ashkar S et al. 2018
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Point mutations in the KMT2A IBM2 motif, which abolish the LEDGF IBD contact, abolish leukaemogenic transformation by MLL-AF9 - a removal applied at the moment of transformation, not to established disease.
The same site on the integrase-binding domain of lens epithelium-derived growth factor is a therapeutic target for MLL leukemia and HIV. · Murai MJ, Pollock J, et al. · Blood 2014What this evidence cannot show
Transformation assay in murine progenitors. The mutation is in the fusion rather than in LEDGF, so it removes one contact rather than the protein, and the fusion retains whatever LEDGF-independent functions it has.
Evidence that argues against this (1)
- An MLL-AF9 F129A mutant that cannot interact with LEDGF nonetheless retained the ability to associate with MLL target genes by ChIP, so an IBM mutation that blocks transformation does not necessarily block chromatin binding. Zhu L et al. 2016
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Dominant-negative LEDGF fragments impaired the clonogenic growth of already-transformed human and mouse MLL-fusion cells in vitro, and in vivo the same fragments only increased the latency period to disease development in a mouse bone-marrow-transplant model - a prevention readout, because the fragment was expressed before transplantation.
Impairing MLL-fusion gene-mediated transformation by dissecting critical interactions with the lens epithelium-derived growth factor (LEDGF/p75). · Méreau H, De Rijck J, et al. · Leukemia 2013What this evidence cannot show
Abstract only. Europe PMC returns no PMCID for this record (inEPMC N, isOpenAccess N), so routes 2-4 of the four-route unreadable test are unavailable and the in-vitro cell identities could not be checked. Latency extension, not cure, is what the abstract reports.
Evidence that argues against this (1)
- A latency increase in a transduce-then-transplant design is compatible with a purely initiation-stage effect, so this in vivo result does not bear on established disease. El Ashkar S et al. 2018
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The cyclic peptide CP65 and LEDGF mutants defective for KMT2A binding selectively impaired the clonogenic growth of primary murine MLL-AF9 blasts - removal applied to an already-transformed population, but in vitro.
Validation and structural characterization of the LEDGF/p75-MLL interface as a new target for the treatment of MLL-dependent leukemia. · Cermáková K, Tesina P, et al. · Cancer Res 2014What this evidence cannot show
Abstract only; Europe PMC returns no PMCID for this record. Colony formation in primary murine blasts is not survival of an established leukaemia, and no animal arm appears in the abstract.
Evidence that argues against this (1)
- The same paper reports that the IBD surface used by KMT2A is also used mutually exclusively by HIV-1 integrase, POGZ and JPO2, so a peptide occupying it removes more than the KMT2A contact and a clonogenic defect cannot be attributed to that interface alone. Cermáková K et al. 2014
- E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
LEDGF shRNA knockdown in already-transformed murine MLL-AF9, MLL-AF10 and MLL-ENL HSPCs is published, with chromatin-occupancy and expression readouts, establishing that LEDGF removal from an established transformed population has been done in vitro.
ASH1L Links Histone H3 Lysine 36 Dimethylation to MLL Leukemia. · Zhu L, Li Q, et al. · Cancer Discov 2016What this evidence cannot show
Read in full via efetch db=pmc (route 3; fullTextXML returned HTTP 500), 36,284 characters with a <body> element present. Murine, myeloid, in vitro, and the readouts are molecular rather than survival.
Evidence that argues against this (1)
- This study's own result cuts against the tethering rationale for removing LEDGF: occupancy of the MLL fusion protein at target genes was consistently INCREASED on LEDGF knockdown. Zhu L et al. 2016
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
LEDGF knockdown in human MLL-rearranged cell lines is published and reports LEDGF-dependent chemoresistance with effects on cell-cycle and survival pathways - again removal from an already-transformed population, in vitro.
LEDGF/p75-mediated chemoresistance of mixed-lineage leukemia involves cell survival pathways and super enhancer activators. · Canella A, Van Belle S, et al. · Cancer Gene Ther 2022What this evidence cannot show
Abstract only. Europe PMC returns no PMCID for this record, so the cell lines used could not be identified and it cannot be established whether any carried t(4;11)/KMT2A::AFF1. No claim beyond the abstract is made here.
Evidence that argues against this (1)
- If LEDGF removal principally sensitises to chemotherapy rather than killing outright, the axis is a combination target rather than a single agent, which is a weaker claim than the ledger's framing. Canella A et al. 2022
- E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The only LEDGF-knockdown xenograft in the literature is an HIV gene-therapy study in NSG mice, not a leukaemia study.
Lens epithelium-derived growth factor/p75 qualifies as a target for HIV gene therapy in the NSG mouse model. · Vets S, Kimpel J, et al. · Mol Ther 2012What this evidence cannot show
Cited to bound a counted absence rather than for its own conclusion. Its relevance to the proposed leukaemia experiment is that LEDGF-depleted human haematopoietic cells engraft NSG mice at all.
Evidence that argues against this (1)
- A gene-therapy engraftment study is not powered to detect a subtle fitness cost of LEDGF depletion in human haematopoiesis, so it bounds the absence without licensing a safety claim. Vets S et al. 2012
bounds what removing this gene's product has actually been shown to do
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
the LEDGF literature intersects this fusion in 2 PubMed records
the maintenance half of that entry's therapeutic-index argument has never been tested
bounds the dossier's genetic-validation claim to prevention
closes the counted absence
the alias exclusion the counted queries rest on
the closest existing approach to removing LEDGF from established disease, and why it does not close the hole
SURVIVES at full text on the tight query (12 records, 6 proceedings, 0 on topic); the 124-record in-vivo intersection is an unscreened upper bound
The PSIP1 arm of this design is the chromatin-occupancy counterpart to that absence: LEDGF removal has never been tested in established disease, and it has never been tested against a degraded fusion either.