Menin and LEDGF hold WILD-TYPE KMT2A on chromatin; for the fusion LEDGF is functionally required but positionally dispensable
also called menin-MLL-LEDGF module · MBM1/MBM2 and IBM2 · chromatin tethering of KMT2A fusions
The fusion cannot grip chromatin strongly on its own. It borrows two helper proteins — menin and LEDGF — that together clamp it in place, and menin is the handle that today's approved drugs grab.
The same thing, for a scientist
The extreme N-terminus of KMT2A, retained in every leukemogenic fusion, carries two menin-binding motifs (MBM1/MBM2) and a separate LEDGF integrase-binding-domain-binding motif (IBM2). Menin does not contact DNA; it acts as a molecular adaptor that bridges the fusion to LEDGF/PSIP1, whose PWWP domain reads H3K36me2/3 on nucleosomes. The combination of a CXXC read-out of unmethylated CpG (step 1) plus a PWWP read-out of H3K36me2/3 gives stable, high-avidity occupancy of transcribed CpG-island promoters. Menin is required for initiation AND maintenance of KMT2A-fusion leukemia but not for other oncogene-driven myeloid transformation, and LEDGF is dispensable for steady-state haematopoiesis while essential for KMT2A-fusion leukemogenesis — the two facts that make this step the best-validated therapeutic node in the entire chain.
Menin is an adaptor, not a DNA-binding protein2 deeper
Menin binds the extreme N-terminus of KMT2A through a high-affinity conserved motif, and oncogenic KMT2A fusions retain that motif and that interaction. The genetics are unusually clean: the interaction is required for the initiation of KMT2A-mediated leukemogenesis, and menin is required for the maintenance of KMT2A-associated but not other oncogene-induced myeloid transformation. Acute genetic ablation of menin reverses aberrant Hox expression and specifically abrogates the differentiation arrest and oncogenic properties of KMT2A-transformed blasts 1.
What menin does mechanistically is bridge. It physically links KMT2A (and its fusions) to LEDGF/PSIP1, a chromatin-associated protein. LEDGF is required for both KMT2A-dependent transcription and leukemic transformation. The elegant confirmation comes from human genetics: a subset of MEN1 mutations found in MEN1-syndrome patients abrogate the LEDGF interaction while preserving the KMT2A interaction, and still compromise KMT2A/menin-dependent function — proving that menin's job here is the bridge itself, not KMT2A binding alone 2.
Two independent interfaces, not one
Menin is not the only contact. A distinct motif in KMT2A, IBM2, binds a well-defined site on the LEDGF integrase-binding domain (IBD); the solution structure of the KMT2A-IBD complex defines it, and point mutations within IBM2 abolish leukemogenic transformation by MLL-AF9 3. The same IBD site is used by HIV-1 integrase, which can efficiently sequester IBD out of the menin-KMT2A complex, and a short IBM2 peptide inhibits both interactions 3.
For the lab's hypothesis this is important in two ways. It is a second druggable protein-protein interface on the same module, independent of the menin pocket that revumenib and ziftomenib occupy — so MEN1 gatekeeper mutations that defeat menin inhibitors would not obviously defeat an IBM2/IBD-directed agent. And it means the module has some internal redundancy: partial loss of one contact may be buffered by the other, which predicts steep rather than shallow dose-response curves for any single-interface drug.
Why the two readers together are the point
LEDGF's PWWP domain is an H3K36me2/3 reader. Cryo-EM of the LEDGF PWWP domain on an H3K36-methylated nucleosome shows multivalent binding — to the methylated histone tail and to both gyres of nucleosomal DNA — which explains the known cooperativity of the interaction 4.
Stack that on step 1 and the targeting logic becomes clear: CXXC reads unmethylated CpG (a promoter/CpG-island signal) while PWWP reads H3K36me2/3 (a transcribed-gene-body signal). Only loci that are simultaneously CpG-island-bearing and actively transcribed satisfy both, which is why the fusion sits on the active/poised developmental promoter set of the cell of origin and why occupancy there is so stable. A drug that removes menin collapses the composite grip rather than one contact.
The therapeutic-index argument, stated precisely
The reason this step is the field's favourite target is not that it is the most essential step — it is that the normal and the oncogenic requirements are separable.
A conditional Psip1 (LEDGF) knockout in the mouse haematopoietic compartment is viable. It shows measurable defects (reduced colony-forming activity in vitro, decreased Hox expression in HSCs, decreased KMT2A occupancy at target genes) but is dispensable for steady-state haematopoiesis while essential for the initiation of KMT2A-mediated leukemia 5. That is the same shape of result as the menin data in gene-men1. Two independent components of the clamp both show the dissociation.
The honest caveat is that both experiments test initiation more strongly than maintenance in a heavily pre-treated human patient, and both are complete genetic deletions rather than partial reversible occupancy.
Correction: LEDGF does not retain the fusion, and the measurement is direct
This entry's title and summary said menin plus LEDGF hold the fusion on chromatin in "stable, high-avidity occupancy". That is correct for wild-type KMT2A and wrong for the fusion.
PMID 27154821, read in full (PMC4930721 via efetch db=pmc, 61,478 characters, after fullTextXML returned HTTP 500), measured it directly in MLL-transformed HSPCs. Verbatim:
"These results indicate that LEDGF is required for retention of WT MLL, but not MLL fusion proteins, at target gene loci in MLL-transformed HSPCs."
The supporting observations point the same way and one of them is the opposite of what this entry predicted: on LEDGF knockdown there were "decreased occupancies of WT MLL and increased occupancies of MLL fusion proteins", reproduced for MLL-AF9, MLL-AF10 and MLL-ENL. And "a missense mutant of MLL-AF9 (F129A) that cannot interact with LEDGF retained an ability to associate with MLL target genes" - F129 being the IBM1 hot spot.
So what does LEDGF do for the fusion? It is still required for leukaemogenesis - the Vav-iCre knockout is protective - but what it supplies is the elongation machinery (AF4, CDK9, BRD4, P-TEFb, Ser2P Pol II, H4K16ac) rather than the address. A licence, not a tether.
This matters for the drug argument in both directions. A LEDGF-directed agent will not evict the fusion from chromatin, so an occupancy readout is the wrong endpoint for it; and the prediction that such an agent should be epistatic with BET or CDK9 inhibition rather than synergistic is a testable consequence that would normally be written off as a failed combination.
| step | 2 |
|---|---|
| chain | mechanism-kmt2a-fusion-causal-chain |
| fusion specific surface | false |
| rate limiting | yes — clinically validated; menin inhibition collapses the program |
| separable from normal function | true |
module components
[ "gene-men1", "gene-psip1", "gene-kmt2a" ]
kmt2a motifs
[ "MBM1 (approx aa 6-10, RWRFP)", "MBM2", "IBM2 (LEDGF-IBD-binding motif)" ]
reader domains
[
{
"protein": "KMT2A",
"domain": "CXXC",
"mark_read": "unmethylated CpG DNA"
},
{
"protein": "LEDGF/PSIP1",
"domain": "PWWP",
"mark_read": "H3K36me2/3 nucleosome"
}
]druggable interfaces
[ "interface-menin-kmt2a-mbm1", "LEDGF-IBD / KMT2A-IBM2" ]
- Redundancy inside the clamp: menin, LEDGF-IBM2, CXXC and the AT-hooks are four contacts. Blocking one may only lower avidity, permitting residual target-gene expression that is enough to sustain self-renewal.
- MEN1 interface mutations arise under menin-inhibitor pressure and preserve KMT2A binding, restarting the program with the drug still present (see gene-men1).
- Menin is a tumour suppressor in endocrine tissue and LEDGF has independent roles in transcription coupling and DNA repair; chronic disruption of either has toxicity risk that months-long relapse trials cannot detect.
- Because the clamp also positions wild-type KMT2A, disrupting it perturbs normal HOX regulation in HSCs — the reason stress haematopoiesis, not steady-state counts, is the toxicity readout to watch.
- Is the LEDGF-IBD / KMT2A-IBM2 interface small-molecule tractable, and would an IBM2-directed agent retain activity against MEN1 gatekeeper-mutant leukemia? This is the single most valuable unexplored interface in the module.
- Do menin inhibitor and LEDGF disruption synergise, or are they epistatic because they collapse the same clamp? The answer decides whether this is one target or two.
- Is LEDGF, like menin, required for long-term HSC function under transplantation stress? Nobody has run the competitive-repopulation experiment that exposed the menin defect.
- In a menin-inhibitor-treated cell, does the fusion remain on chromatin via IBM2/LEDGF and AT-hooks at reduced occupancy, or does it leave entirely? Residual occupancy would explain non-genetic escape.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Oncogenic KMT2A fusions retain a high-affinity N-terminal menin-binding motif; this interaction is required to initiate KMT2A-mediated leukemogenesis, and menin is essential for maintenance of KMT2A-associated but not other oncogene-induced myeloid transformation, with acute menin ablation reversing aberrant Hox expression and abrogating differentiation arrest.
The menin tumor suppressor protein is an essential oncogenic cofactor for MLL-associated leukemogenesis · Yokoyama A, Somervaille TCP, Smith KS, Rozenblatt-Rosen O, Meyerson M, Cleary ML · Cell 123(2):207-18 2005What this evidence cannot show
Mouse bone-marrow retroviral transformation and conditional Men1 deletion; 'acute genetic ablation' is a Cre-mediated knockout, not a drug, so kinetics and partial-occupancy behaviour are not addressed. The fusions tested are myeloid-context KMT2A fusions rather than KMT2A::AFF1 in a lymphoid context.
Evidence that argues against this (1)
- "Essential for maintenance" comes from acute genetic ablation in mouse models, and the pharmacological test in patients gives a different number: revumenib in relapsed or refractory KMT2A-rearranged acute leukaemia produced complete remission with full or partial haematological recovery in roughly a quarter of patients. Genetic essentiality and clinical dependency are not the same measurement, and the gap between them is the central open question for this whole hypothesis. Journal of Clinical Oncology 2025
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin acts as a molecular adaptor physically linking KMT2A proteins to LEDGF/PSIP1; LEDGF is required for both KMT2A-dependent transcription and leukemic transformation, and MEN1 patient mutations that abrogate the LEDGF interaction while preserving KMT2A binding still compromise KMT2A/menin-dependent function.
Menin critically links MLL proteins with LEDGF on cancer-associated target genes. · Yokoyama A, Cleary ML · Cancer Cell 14(1):36-46 2008What this evidence cannot show
Largely RNAi and overexpression in cell lines plus mouse myeloid progenitor assays; the MEN1 separation-of-function mutants are a small set and their effect on protein stability is a confounder.
Evidence that argues against this (1)
- The separation-of-function MEN1 mutations are patient alleles selected in endocrine tumours, so their behaviour in a leukaemia assay is being read across a disease boundary. The adaptor model also has to accommodate menin's other roles: the same pocket binds JUND with the opposite transcriptional effect, so "molecular adaptor" describes one of at least two mutually exclusive occupancies. Nature 2012
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A distinct KMT2A motif, IBM2, binds a defined site on the LEDGF integrase-binding domain; point mutations within IBM2 abolish leukemogenic transformation by MLL-AF9, the same IBD site is used by HIV-1 integrase, and a short IBM2 peptide inhibits both the IBD-KMT2A/menin and IBD-integrase interactions.
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, He S, et al. (Grembecka J, Cierpicki T, senior authors) · Blood 124(25):3730-7 2014What this evidence cannot show
NMR solution structure of a peptide-domain complex; the transformation readout is MLL-AF9 colony formation in mouse progenitors. The IBM2 peptide is a tool compound, not a lead — the interface is shallow and peptide-like, so its small-molecule tractability is unproven.
Evidence that argues against this (1)
- The IBM2 result is scored on MLL-AF9 transformation, a murine myeloid assay, and the shared HIV-integrase site is a double-edged observation: a peptide that blocks both IBD-KMT2A and IBD-integrase is by construction not selective for the leukaemic interaction, and LEDGF is broadly required as a chromatin adaptor. Cancer Research 2014
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Cryo-EM at 3.2 A of the LEDGF PWWP domain bound to an H3K36-methylated nucleosome shows multivalent binding to the methylated H3 tail and to both gyres of nucleosomal DNA, explaining the cooperativity of PWWP recruitment to H3K36-methylated chromatin.
Structure of H3K36-methylated nucleosome-PWWP complex reveals multivalent cross-gyre binding · Wang H, Farnung L, Dienemann C, Cramer P · Nature Structural & Molecular Biology 27(1):8-13 2020What this evidence cannot show
Reconstituted nucleosome and isolated PWWP domain in vitro; no leukemia system and no KMT2A fusion present. The inference that this read-out positions the fusion complex in leukemic cells is mechanistic reasoning, not a measurement in that system.
Evidence that argues against this (1)
- A 3.2 A structure of an isolated PWWP domain on a reconstituted nucleosome explains cooperativity in vitro and does not address the tethering question this entry is about: H3K36 methylation marks most transcribed gene bodies, so PWWP affinity cannot select the specific loci the fusion occupies. The selectivity has to come from elsewhere in the complex. Cancer Research 2014
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Conditional Psip1/LEDGF-p75 knockout in the mouse haematopoietic compartment is dispensable for steady-state haematopoiesis but essential for the initiation of KMT2A-fusion-mediated leukemia, with reduced in-vitro colony-forming activity, decreased Hox expression in HSCs and decreased KMT2A occupancy at target genes.
LEDGF/p75 is dispensable for hematopoiesis but essential for MLL-rearranged leukemogenesis · El Ashkar S, Schwaller J, Pieters T, et al. (De Rijck J, senior author) · Blood 131(1):95-107 2018What this evidence cannot show
'Dispensable' is measured at steady state; the paper itself reports several haematopoietic defects and reduced colony-forming activity, so the claim is about survival and lineage output, not about unimpaired stem-cell fitness. Initiation was tested more rigorously than maintenance of an established leukemia.
Evidence that argues against this (1)
- The analogous menin experiment shows that a 'dispensable at steady state' cofactor can still be required for long-term HSC function under competitive transplantation stress. The same caveat should be assumed for LEDGF until tested, which matters for a patient recovering from intensive chemotherapy. Maillard I et al. 2009
- E-retentionestablished confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.journal-article
LEDGF knockdown in MLL-transformed HSPCs decreases wild-type MLL occupancy and increases MLL-fusion occupancy at target loci, and an MLL-AF9 F129A mutant unable to bind LEDGF still associates with MLL target genes. LEDGF is required for retention of wild-type MLL but not of MLL fusion proteins.
ASH1L Links Histone H3 Lysine 36 Dimethylation to MLL Leukemia · Cancer Discovery 2016What this evidence cannot show
Read in full 2026-09-20. **Established as a result within the reported study** under the grading rule. The system is murine MLL-AF9/AF10/ENL-transformed HSPCs - **no KMT2A::AFF1 and no human cells** - and the result is a secondary finding in a paper whose subject is ASH1L, so it has had far less scrutiny than its headline. Knockdown is partial, and an increase in fusion occupancy on losing a partner is unusual enough to want independent replication.
Evidence that argues against this (1)
- Kerry 2017 reproduced the direction at only 30% knockdown, which is weak corroboration, and no one has tested KMT2A::AFF1. The entry's original claim was not baseless - the clamp is real and structurally characterised - it was **applied to the wrong species of KMT2A**, which is the same error this graph has now made several times. Blood 2014
- ← related toWhere KMT2A and its fusions sit inside the nucleus — SNL1/SNL2, nuclear speckles and the condensate question
chromatin tethering by menin and LEDGF/PSIP1 as the biochemical counterpart of subnuclear positioning
- ← related toThe same pocket suppresses JUND: menin inhibition's unmeasured endocrine liability in infants
the tethering mechanism the programme exploits runs through the pocket whose disruption defines MEN1 syndrome
menin selects between the two IBD paralogues rather than only stabilising one
independent evidence for the tether module from a different data type
the tether co-dependency survives lineage adjustment and strengthens in leukemia
tether module survives bootstrap and Holm correction with intervals far from zero
- depends on →Step 1 — How the fusion finds its genes: AT-hooks and the CXXC domain reading unmethylated CpG
step 2 follows step 1 in the causal chain
the elongation machinery only acts once the fusion is anchored
the molecular contact underneath the tethering step
how the tether reads the chromatin it sits on
breaks the tether, evicting the fusion from chromatin
the maintenance half of that entry's therapeutic-index argument has never been tested
direct occupancy measurement contradicts the fusion-tether claim
qualifies that entry's therapeutic-index argument
proposes a rival model for the fusion half of the clamp
CK2 phosphorylation of the KMT2A IBD-binding motif raises LEDGF affinity; paralogue assignment untested
The authors' own proposed explanation for menin outlasting the fusion is LEDGF anchorage, citing Yokoyama & Cleary 2008 (PMID 18598942). This is the first degron-based observation consistent with that tether being fusion-independent.
step 2 - menin/LEDGF; required, positionally contradicted for the fusion