Lost from the KMT2A half: the SET domain (aa 3829-3945) - the lost module whose loss demonstrably does not matter
also called MLL1 SET domain · H3K4 methyltransferase activity · KMT2A catalytic domain
The enzyme part of KMT2A - the bit that actually writes a chemical mark onto DNA packaging - is deleted from every leukemia fusion. Deleting it from mice does not stop the leukemia, so this is a loss that appears not to matter.
The same thing, for a scientist
The KMT2A SET domain (UniProt Q03164, aa 3829-3945, with post-SET at 3953-3969) carries the protein's H3K4 methyltransferase activity and is absent from every fusion. It is also the clearest case in this beat of **lost but not load-bearing**: Mishra et al. show the SET domain, and therefore MLL1's HMT activity, is dispensable both for maintaining haematopoietic stem cells and for supporting MLL-AF9-driven leukemogenesis, and that the functionally critical output at MLL1 target genes is instead recruited MOF/H4K16 acetylation, opposed by SIRT1 [E1]. The corollary matters for how this graph reads its own domain maps: the fusion is a neomorph, not a KMT2A-catalysis-deficiency state, and 'the fusion cannot methylate H3K4' is a true statement with, as far as anyone has shown, no consequence. The honest caveat is that a pharmacological block of the same activity (MM-401) does kill MLL-rearranged cells [E3], so 'dispensable' has been established genetically and contradicted pharmacologically, and the discrepancy is unresolved.
Coordinates, and why they are not in doubt
UniProt Q03164 gives SET 3829-3945 and post-SET 3953-3969 2. That matches the value already stored in gene-kmt2a, so unlike the bromodomain range this one needs no correction. The domain sits ~2,500 residues C-terminal to the breakpoint cluster region; no fusion retains any part of it.
The genetic result: dispensable1 deeper
Mishra et al. deleted the SET domain in mice and asked two questions 1:
- Is MLL1 catalysis needed for HSC maintenance? No.
- Is it needed for MLL-AF9 leukemogenesis? No.
And then the mechanistic why: on Mll1 deletion, H4K16 acetylation is selectively depleted at MLL1 target genes together with reduced transcription, and inhibiting SIRT1 is sufficient to prevent that loss and to prevent the drop in target expression. The conclusion is that recruited MOF activity, not intrinsic MLL1 HMT activity, is the central output at MLL1 targets.
This is why the KMT2A-side subtraction beat cannot simply be a list of deletions. The single largest, most famous lost module is, on the best genetic evidence available, not carrying load.
Note the direction of the MOF result
Mishra's mechanism runs through MOF/KAT8, which binds the MLL1 C-terminal region 4 - also lost in the fusion. So the finding is not 'the SET domain is irrelevant and everything else is fine'; it is 'what matters at MLL1 targets is a recruited acetyltransferase that the fusion also cannot recruit'. See subtraction-kmt2a-mof-contact. The load the SET domain is not carrying is being carried by another module the fusion also lacks, and only the wild-type allele can supply either.
The pharmacological result: not dispensable
MM-401 blocks MLL1's H3K4 methyltransferase activity by preventing WDR5-dependent core-complex assembly, and it selectively blocks proliferation of MLL-rearranged cells with cell-cycle arrest, apoptosis and myeloid differentiation, sparing normal bone marrow and non-MLL cells 3.
The two results are hard to hold at once. Candidate explanations, all untested head to head:
- Assembly is not catalysis. MM-401 removes WRAD from MLL1, which removes a scaffold as well as an activity; a SET deletion removes only the activity. If WRAD does structural work, the two perturbations are not the same perturbation.
- Acute versus developmental. A germline/conditional deletion permits compensation that an acute small-molecule block does not.
- Model mismatch. Human MLL-rearranged cell lines versus mouse MLL-AF9.
- Off-target. MM-401 is a peptidomimetic tool compound; selectivity beyond the MLL family is asserted rather than profiled in the abstract.
The graph should carry this as an open contradiction rather than resolving it by preference. Its practical consequence: 'the fusion lost the SET domain, therefore KMT2A catalysis is irrelevant here' is not a safe inference.
| module | SET (H3K4 methyltransferase) domain |
|---|---|
| uniprot | Q03164 |
| start | 3829 |
| end | 3945 |
| fate in fusion | lost |
| demonstrated functional consequence of loss | none in the MLL-AF9 mouse model: SET deletion does not impair HSC maintenance or leukemogenesis (PMID 24813891) |
| contradicting result | pharmacological block of the same activity selectively kills MLL-rearranged cells (PMID 24389101) |
| loss class | lost, and demonstrably tolerated genetically; contradicted pharmacologically; unresolved |
adjacent
[
{
"name": "Post-SET",
"start": 3953,
"end": 3969
}
]position confirmed by
[ "UniProt Q03164 domain feature table", "agrees with the range already stored in gene-kmt2a" ]
- If SET-domain catalysis is genuinely dispensable, then DOT1L- and SET-adjacent 'restore the enzyme' strategies target something the disease does not need, and a negative trial would be uninformative about the underlying model rather than about the drug.
- If instead MM-401 is right and MLL1 catalysis is required, the required copy is the wild-type allele - so the effective therapy would be one that damages a gene under extreme human constraint, in infants, during development.
- Genetic SET deletion says dispensable; MM-401 says required. Nobody has run both perturbations in the same cells with the same readout. That single head-to-head experiment would resolve one of the clearest open contradictions in KMT2A biology and needs no new reagents.
- Is the SET domain dispensable in a lymphoid KMT2A::AFF1 model as it is in myeloid MLL-AF9? The infant disease this project targets has never been tested.
- If the functional output at KMT2A targets is MOF-dependent H4K16 acetylation opposed by SIRT1, is SIRT1 inhibition a partner-agnostic lever in KMT2A-rearranged leukemia, and has anyone tried it in that context rather than in normal HSCs?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The SET domain, and thus the histone methyltransferase activity of MLL1, is dispensable for maintaining haematopoietic stem cells and for supporting leukemogenesis driven by the MLL-AF9 fusion oncoprotein; upon Mll1 deletion, H4K16 acetylation is selectively depleted at MLL1 target genes with reduced transcription, SIRT1 inhibition is sufficient to prevent that loss, and recruited MOF activity rather than intrinsic MLL1 HMT activity is central to maintaining HSC target genes.
The histone methyltransferase activity of MLL1 is dispensable for hematopoiesis and leukemogenesis · Mishra BP, Zaffuto KM, Artinger EL, Org T, Mikkola HKA, Cheng C, Djabali M, Ernst P · Cell Reports 7(4):1239-1247 2014What this evidence cannot show
ABSTRACT ONLY. Mouse genetics; MLL-AF9 only, which is a myeloid model and not the infant KMT2A::AFF1 lymphoid disease this project centres on. A germline or conditional deletion allows compensation - including by the paralogue MLL2/KMT2B - that an acute inhibitor would not, so 'dispensable' here means 'dispensable in this model over this timescale'. **The universal is false and the qualification matters.** The SET domain and its WRAD contacts are lost from the **direct 5' product** only. Two classes keep them: the **reciprocal der(4) product AFF1::KMT2A**, which carries KMT2A exons 12-37 including the WIN motif and SET domain and is expressed in a large minority of t(4;11) infants; and **NUP98::KMT2A** (inv(11)(p15q23)), which deletes the exon-1 menin motif while keeping WIN and SET, and is menin-inhibitor resistant. Both counterexamples were already in this graph when the universal was written — a linking failure, not a knowledge gap. Corrected 2026-09-20 after the Q23 adjudication; verified by the coordinator against both abstracts.
Evidence that argues against this (2)
- Pharmacological blockade of the same activity gives the opposite answer: MM-401, which blocks MLL1-WDR5 assembly and hence MLL1 H3K4 methyltransferase activity, selectively blocks proliferation of MLL-rearranged cells. Genetic dispensability and pharmacological requirement are both on record and have never been reconciled experimentally. Cao F et al. 2014
- The reciprocal AFF1::KMT2A product is the exact mirror of the one this entry describes: **WRAD-competent and menin-blind**. In the subset of t(4;11) infants expressing it, the cell contains a second KMT2A-derived oncoprotein with an intact methyltransferase module that a menin inhibitor cannot engage by construction. Any claim about what 'the fusion' retains has to say which one. Oncogene 2021
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt Q03164 places the KMT2A SET domain at residues 3829-3945 and the post-SET region at 3953-3969.
UniProtKB entry Q03164 (KMT2A_HUMAN) - domain feature table, retrieved via the UniProt REST API · UniProt Consortium · UniProtKB 2026What this evidence cannot show
Accession verified by reading back the primary gene name (KMT2A) and length (3,969 aa) from the fetched record. Profile-derived curatorial boundaries, consistent here with the value already stored in gene-kmt2a.
Evidence that argues against this (1)
- A UniProt feature table is a curated model rather than a measurement. Profile-based domain calls for KMT2A differ between UniProt, Pfam, SMART, InterPro and CATH, and the sibling atlas beat was corrected last cycle for quoting one source's range as though it were observed. Quoting coordinates without the spread states a choice as a fact. Molecular Cell 2014
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MM-401 inhibits MLL1 activity by blocking the MLL1-WDR5 interaction and complex assembly, does not affect other MLL-family methyltransferases, and blocks proliferation of MLL-rearranged cells through cell-cycle arrest, apoptosis and myeloid differentiation without general toxicity to normal bone marrow or non-MLL cells.
Targeting MLL1 H3K4 methyltransferase activity in mixed-lineage leukemia · Cao F, Townsend EC, Karatas H, et al. (Wang S, Dou Y, senior authors) · Molecular Cell 53(2):247-261 2014What this evidence cannot show
ABSTRACT ONLY. Tool compound characterised by the group that made it, with an enantiomer control. Because the fusion has no WIN motif, MM-401's target in a KMT2A-rearranged cell must be the wild-type allele; that inference is this Queen's, not the paper's.
Evidence that argues against this (1)
- MM-401 acts by blocking the MLL1-WDR5 interaction, so it removes complex ASSEMBLY rather than catalysis specifically, and its MLL1 selectivity is the problem: if the paralogue sustaining MLL-rearranged AML is MLL2, a probe that spares other family members is testing the wrong protein. The entry treats it as evidence about SET-domain function. Cancer Cell 2017
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A stable MLL1-MOF complex was immunoaffinity purified and the interaction sites mapped to MLL1 C-terminal and MOF zinc finger domains; both the MLL1 H3K4 methyltransferase and the MOF H4K16 acetyltransferase activities are required for optimal transcriptional activation on a chromatin template in vitro and at the endogenous MLL1 target HOXA9 in vivo.
Physical association and coordinate function of the H3 K4 methyltransferase MLL1 and the H4 K16 acetyltransferase MOF · Dou Y, Milne TA, Tackett AJ, Smith ER, Fukuda A, Wysocka J, Allis CD, Chait BT, Hess JL, Roeder RG · Cell 121(6):873-885 2005What this evidence cannot show
ABSTRACT ONLY. Biochemistry in HeLa cells and in-vitro chromatin templates; 'MLL1 C-terminal' is not resolved to residue numbers in the abstract, so this entry does not assign the MOF contact a coordinate range. Note this paper reports BOTH activities as required for optimal activation, which sits awkwardly beside E1's conclusion that MLL1's own catalysis is dispensable in vivo.
Evidence that argues against this (1)
- Mapping the MLL1-MOF interaction to the MLL1 C-terminus places it beyond every breakpoint, which is the entry's point - but it also means the contact is absent from the fusion by deletion, so no experiment on the fusion can test what its loss does without re-inserting it. The claim is structural, not functional. Cell 2005
the fusion allele cannot supply the H3K4me3 this mechanism requires - stated as a prediction, not a finding
AN INDEPENDENT LINE REACHING THE SAME PLACE: a genome-wide CRISPR screen puts ASH2L at 17,248 and RBBP5 at 14,622 of 17,728 for KMT2A-selectivity, which is what this entry's mouse-genetics argument for SET-domain dispensability predicts.
THE PAIRED SUBTRACTION: MLL1's own SET domain is dispensable where recruited MOF activity is not, which relocates the decisive histone output at MLL1 targets.
the catalytic module the fusion loses, and the dispensability result
- ← related toLost from the KMT2A half: the WIN motif and the whole WRAD interface (aa 3762-3767, Arg3765)
[in_tension_with] MM-401 kills MLL-r cells by blocking assembly of a complex whose catalytic output is genetically dispensable in the same disease
- part of →KMT2A / MLL1 — the gene that breaks
SET aa 3829-3945, post-SET 3953-3969; lost in every fusion
same verdict from the subtraction side, plus the unresolved MM-401 contradiction that entry does not record
- depends on →Lost from the KMT2A half: the WIN motif and the whole WRAD interface (aa 3762-3767, Arg3765)
SET catalysis requires WRAD assembly through R3765; both are lost together
the functional output at KMT2A targets runs through recruited MOF, not through SET catalysis
SIRT1 inhibition prevents the H4K16ac loss that follows Mll1 deletion
- ← related toLost from the KMT2A half: the FYRN-FYRC clamp (2018-2074 / 3666-3747) that holds wild-type KMT2A together
FYRN binds FYRC and the SET domain; the clamp and the catalytic domain are lost as one piece
explains why SET loss is tolerated: the load at KMT2A targets is carried by recruited MOF, not by SET catalysis
row: SET 3829-3945
Correction to 'deleted from every leukemia fusion': the SET domain is deleted from the direct 5'-KMT2A product and retained in reciprocal der products and in NUP98::KMT2A.
- ← refutesThe graph's universal 'every KMT2A fusion loses the SET domain' is a silent, propagating error class
Names the sentence: 'the enzyme part of KMT2A is deleted from every leukemia fusion' should read 'from the direct 5'-KMT2A fusion product'.