KMT2A partial tandem duplication: an addition that only works once the normal copy has been switched off — 'recessive gain-of-function'
also called MLL-PTD · KMT2A-PTD · recessive gain-of-function · wild-type MLL allele silencing
This version of the disease has no partner gene at all — a chunk of KMT2A is simply copied and pasted back into itself. The surprise is that the copy is not enough: the cell also switches off its remaining normal copy of the gene, and forcing that normal copy back on kills the leukemia cells.
The same thing, for a scientist
KMT2A-PTD is an in-frame internal duplication of the 5' exons encoding the AT-hooks and CXXC, arising by Alu-mediated homologous recombination; nothing is deleted and no partner is involved. Its subtraction component is allelic rather than structural: blasts with the MLL-PTD/WT genotype express the PTD allele but not the wild-type allele, silencing is reversed by DNMT and HDAC inhibitors, and wild-type MLL induction is associated with selective sensitivity to cell death. The authors' own term is 'recessive gain-of-function' — addition and subtraction jointly required. This is the third distinct sub-type of subtraction in this beat and the only one that is already pharmacologically actionable.
What the duplication adds, and what it does not remove
The founding reports describe a direct tandem duplication of a portion of ALL-1/MLL spanning exons 2-6 1 or exons 2-8 2, transcribed into an mRNA encoding a partially duplicated protein. The duplication arises by homologous recombination between Alu elements in MLL introns 2. The duplicated segment covers the 5' region encoding the AT-hook DNA-binding motifs and the CXXC domain.
Two points of care, both corrections to assumptions this beat started with:
- Exon numbering. The primary literature says exons 2-6 or 2-8 in the Schichman/Caligiuri numbering, not 3-9 or 5-11. MLL exon numbering has been revised, so any statement of the duplicated interval must name its convention.
- SET domain retention is an inference, not a citation. PTD is uniformly described as an in-frame internal duplication producing a partially duplicated protein, and MLL's SET domain lies far 3' of the duplicated interval, so the SET domain should be retained once and intact. No paper located in a dedicated search says this in so many words. It is recorded here as inference from gene structure and should not be cited as a fact.
The subtraction: the wild-type allele is silenced, and turning it back on is lethal1 deeper
Whitman et al. found that normal-karyotype AML blasts with the MLL(PTD/WT) genotype express MLL PTD but not MLL WT; that silencing of the wild-type allele is reversed by DNMT and HDAC inhibitors; and that wild-type MLL induction was associated with selective sensitivity to cell death. Their conclusion is that loss of wild-type MLL function via monoallelic repression contributes to the leukemic phenotype 3.
This clears the bar that most entries in this beat do not. It is not merely that something is missing; the missing thing was restored, and restoring it killed the cells. Together with the CALM-AF10 PZP add-back, it is one of only two restoration experiments in the entire subtraction set, and it is the only one performed in human patient blasts rather than a mouse model.
Mouse genetics adds a dosage dimension: Mll(PTD/-) mice die at birth while Mll(PTD/WT) mice do not, and although both genotypes show increased HoxA expression and myeloid progenitors, only Mll(PTD/WT) shows increased pluripotent progenitors 4. The wild-type allele's presence is not a neutral background; it changes which compartment expands.
Why 'recessive gain-of-function' is the right phrase and a warning label
Whitman et al.'s own term captures the structure exactly: the PTD behaves as a gain-of-function lesion whose full leukemogenic effect requires the loss of the wild-type allele. Addition and subtraction are not alternatives here, they are conjuncts.
That matters for the framing this beat audits, because it shows the addition/subtraction dichotomy is a false one at least once. A partner or lesion can be scored 'addition' on the strength of a real gain and still be substantially driven by a loss that nobody looked for, because the loss is not in the sequence of the oncoprotein.
The one subtraction in this beat that already implies a therapy
Everywhere else in this beat the therapeutic conclusion is deflating: you cannot inhibit an absence. KMT2A-PTD is the exception, because the absence is an epigenetically silenced allele and epigenetic silencing is reversible with approved drugs. Whitman et al. explicitly framed this as identifying a patient subgroup for molecular-targeted therapy, with DNMT and HDAC inhibitors as the tools 3.
The generalisable lesson for target selection is not 'use hypomethylating agents' — it is that when the subtraction is a silenced gene rather than a deleted domain, restoration is tractable. That distinction should be the first branch in any triage of a loss-driven mechanism.
| lesion | KMT2A partial tandem duplication (no partner gene) |
|---|---|
| duplicated interval | exons 2-6 or exons 2-8 (Schichman/Caligiuri 1994 numbering) |
| exon numbering caveat | MLL/KMT2A exon numbering has been revised since 1994. The interval must always be stated with its convention; renderings of this lesion as 'exons 3-9' or 'exons 5-11' do not match the founding reports. |
| duplicated content | AT-hook DNA-binding motifs and the CXXC domain |
| mechanism of formation | homologous recombination between Alu elements in KMT2A introns |
| discarded region | none — an in-frame internal duplication, not a truncation |
| set domain retained | inferred, not directly cited |
| subtraction component | transcriptional silencing of the wild-type KMT2A allele in PTD/WT blasts |
| loss shown to contribute to transformation | true |
| restoration experiment done | yes — DNMT + HDAC inhibitors reverse wild-type silencing and are associated with selective cell death (Whitman 2005); not allele-specific |
| verdict | both |
| verdict rationale | The authors' own framing of 'recessive gain-of-function' is precisely right: the duplication is the addition, the epigenetic silencing of the remaining normal allele is the subtraction, and the evidence indicates both are required. |
| experiment to settle | Allele-specific re-expression of wild-type KMT2A in MLL-PTD blasts using CRISPRa targeted to the wild-type promoter, scoring apoptosis and colony formation against a scrambled-guide control. This separates wild-type MLL restoration from the genome-wide effects of hypomethylating and HDAC-inhibitor treatment, which is the one gap in an otherwise strong result. |
| subtraction subtype | allelic (epigenetic silencing of the second allele), distinct from discarded-module and sequestration subtypes |
- Hypomethylating agents and HDAC inhibitors de-repress far more than the wild-type KMT2A allele. A therapy justified by this mechanism will have effects that have nothing to do with the mechanism, and attributing response to allele restoration is exactly the error the evidence base cannot yet exclude.
- Restoring wild-type KMT2A restores a methyltransferase with broad developmental roles. In an infant or young child, the toxicity of globally re-activating epigenetically silenced loci is not a theoretical concern.
- Is wild-type KMT2A also silenced in translocation-driven KMT2A-r leukemia? If the second allele is repressed there too, the recessive component would apply across the whole class and not just to PTD — and nobody appears to have asked.
- Can allele-specific CRISPRa separate wild-type KMT2A re-expression from the pleiotropy of hypomethylating agents? Without it the strongest subtraction result in this beat rests on a drug that changes everything.
- Menin inhibitors are active in KMT2A-PTD AML and revumenib's development included this genotype. Does menin inhibition alter wild-type allele silencing, or does it act purely on the PTD product?
- Does the duplicated AT-hook/CXXC cassette compete with wild-type KMT2A for chromatin, making the silencing a consequence of competition rather than an independent event?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Two AML patients with trisomy 11 carried a direct tandem duplication of a portion of the ALL-1 (MLL) gene spanning exons 2-6, transcribed into an mRNA capable of encoding a partially duplicated protein.
ALL-1 partial duplication in acute leukemia · Schichman SA, Caligiuri MA, Gu Y, Strout MP, Canaani E, Bloomfield CD, Croce CM · Proceedings of the National Academy of Sciences USA 1994What this evidence cannot show
Two patients, both with trisomy 11. Exon numbering follows the 1994 MLL convention, which is not the modern one. Identifier and abstract verified via PubMed on 2026-09-18.
Evidence that argues against this (1)
- Both index patients carried trisomy 11, so the founding report cannot separate the duplication from a gene-dosage change - and the mouse experiment that removes that confound found the PTD acts in the PRESENCE of an intact wild-type allele. Dorrance et al. knocked murine Mll exons 5-11 into intron 4 of the endogenous locus; Mll(PTD/WT) mice, with one PTD allele, one wild-type allele and normal copy number, showed altered Hox expression boundaries in embryogenesis, axial skeletal defects, increased haematopoietic progenitors, and overexpression of Hoxa7, Hoxa9 and Hoxa10 in spleen, marrow and blood, with increased H3/H4 acetylation and H3K4 methylation at the Hoxa7 and Hoxa9 promoters. That is gain of function with the second allele intact, which is what the 'recessive' half of 'recessive gain-of-function' says should not suffice. Dorrance AM et al. 2006
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
A normal-karyotype AML case carried an ALL-1/MLL tandem duplication spanning exons 2-8, arising by homologous recombination between Alu elements.
ALL-1 tandem duplication in acute myeloid leukemia with a normal karyotype involves homologous recombination between Alu elements · Schichman SA, Caligiuri MA, Strout MP, Carter SL, Gu Y, Canaani E, Bloomfield CD, Croce CM · Cancer Research 1994What this evidence cannot show
Single case for the breakpoint mapping. Identifier and abstract verified via PubMed on 2026-09-18. Connects to this graph's `genome-alu-repeat-landscape-kmt2a-bcr`, which covers the same repeat architecture for translocation breakpoints.
Evidence that argues against this (1)
- A mechanism claim resting on one mapped case is weak where a systematic series exists. Meyer et al. sequenced MLL breakpoints in 2,345 acute leukaemia patients and found the breakpoint distribution itself is age- and lineage-dependent - intron 11 in ALL and younger patients, intron 9 in AML and older patients - which they interpret as different aetiologies in different age groups, with 135 distinct rearrangements catalogued. Alu-mediated homologous recombination inferred from a single normal-karyotype AML case is one candidate aetiology among several, and the PTD's breakpoint mechanism has never been re-examined at that scale or against the age structure the series reveals. Meyer C et al. 2018
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Normal-karyotype AML blasts with the MLL(PTD/WT) genotype express MLL PTD but not MLL WT; silencing of MLL WT is reversed by DNA methyltransferase and histone deacetylase inhibitors, and MLL WT induction is associated with selective sensitivity to cell death, supporting the hypothesis that loss of wild-type MLL function via monoallelic repression contributes to the leukemic phenotype.
The MLL partial tandem duplication: evidence for recessive gain-of-function in acute myeloid leukemia identifies a novel patient subgroup for molecular-targeted therapy · Whitman SP, Liu S, Vukosavljevic T, Rush LJ, Yu L, Liu C, Klisovic MI, Maharry K, Guimond M, Strout MP, Becknell B, Dorrance A, Klisovic RB, Plass C, Bloomfield CD, Marcucci G, Caligiuri MA · Blood 2005What this evidence cannot show
Patient blasts and cell lines. DNMT and HDAC inhibitors are pleiotropic, so the association between wild-type MLL induction and cell death is correlative within a treatment that changes thousands of genes; the paper does not isolate wild-type MLL re-expression from the rest of the drug effect. Journal is Blood, not Cancer Research, a misattribution caught and corrected during verification. Identifier and abstract verified via PubMed on 2026-09-18.
Evidence that argues against this (1)
- Because DNMT and HDAC inhibitors de-repress a large fraction of the genome, the selective cell death they cause in MLL-PTD blasts need not be attributable to wild-type MLL re-expression specifically. An allele-specific re-expression tool (CRISPRa targeted to the wild-type allele) would be required to make the causal claim cleanly, and has not been reported. Whitman SP et al. 2005
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Mll(PTD/-) mice die at birth, unlike Mll(PTD/WT), Mll(WT/-) and Mll(WT/WT) mice; Mll(PTD/-) and Mll(PTD/WT) fetal liver cells both show increased HoxA expression and myeloid progenitors, but only Mll(PTD/WT) shows increased pluripotent progenitors.
The Mll partial tandem duplication: differential, tissue-specific activity in the presence or absence of the wild-type allele · Dorrance AM, Liu S, Chong A, Pulley B, Nemer D, Guimond M, Yuan W, Chang D, Whitman SP, Marcucci G, Caligiuri MA · Blood 2008What this evidence cannot show
Mouse knock-in genetics; germline null of Mll has its own severe developmental phenotype, so the Mll(PTD/-) lethality is not cleanly attributable to the leukemic mechanism. Identifier and abstract verified via PubMed on 2026-09-18.
Evidence that argues against this (1)
- The neonatal lethality is more parsimoniously read as Mll hemizygosity than as a statement about the PTD, and the same laboratory's earlier knock-in is the reason. Dorrance et al. reported that Mll(PTD/WT) animals - wild-type allele intact - already display the Hox-boundary shift, the axial skeletal defects, the increased haematopoietic progenitors and Hoxa7/Hoxa9/Hoxa10 overexpression with the corresponding H3/H4 acetylation and H3K4 methylation gains. The biological output of the PTD is therefore present without any silencing of the second allele, so the Mll(PTD/-) versus Mll(PTD/WT) contrast measures what the wild-type allele contributes on its own - and germline Mll loss has its own severe developmental phenotype - rather than measuring a requirement for wild-type silencing in leukaemogenesis. Dorrance AM et al. 2006
Counter-evidence to 'all frequent KMT2A rearrangements lose PHD/bromodomain/SET': the KMT2A partial tandem duplication retains all of them, is found in 11% of AML with normal cytogenetics, and is generated by Alu-mediated homologous recombination joining intron 6 or 8 to intron 1 - outside the major BCR entirely (Strout 1998, PMID 9482895).
- ← related toA three-question triage to detect subtraction mechanisms, applied to every KMT2A partner in the graph
Counter-evidence (Dorrance 2008, PMID 18617636): Mll(PTD/-) mice die at birth, HoxA up-regulation is dominant with or without the wild-type allele, and increased pluripotent CFU-GEMM requires wild-type Mll - so the recessive gain-of-function reading behind the Q3 positive is contested
- ← confers resistance toSilmitasertib (CX-4945) - a CKII inhibitor that works on KMT2A-rearranged leukemia by strengthening the allele the fusion does not have
[limited_by] loss or silencing of the wild-type allele defeats the mechanism entirely
- refutes →Lost from the KMT2A half: PHD3 and the bromodomain are one cassette, not two modules (aa 1566-1765)
the PTD retains the whole cassette and still deregulates Hox, so the switch can be bypassed rather than only deleted
- related to →KMT2A / MLL1 — the gene that breaks
recessive gain-of-function challenged: Mll(PTD/WT) knock-in mice show the Hox and progenitor phenotype with the wild-type allele intact (Dorrance 2006)
- related to →Alu repeats and nuclear-matrix anchors in the KMT2A breakpoint region — the genome-feature side of the break
the duplication arises by homologous recombination between Alu elements in KMT2A introns
wild-type allele silencing is reversed by DNMT and HDAC inhibitors
KMT2A-PTD is a normal-karyotype AML lesion with a distinct recessive gain-of-function mechanism
contrast case: PTD retains the whole protein and loses function by silencing the second allele rather than by truncation
the allelic question, asked for the first time, returned positive for PTD
per-partner entry feeding the synthesis