Forty-five percent of t(4;11) patients express a KMT2A fusion transcript that keeps the whole WRAD interface
also called AF4-MLL keeps the SET domain · reciprocal fusion WRAD · der(4) product and WRAD
A t(4;11) swap makes two hybrid genes, not one. The famous one throws away the enzyme end of KMT2A; the mirror-image one keeps it, together with every surface the WRAD proteins grip. So in roughly half of these children there is a KMT2A fusion protein in the nucleus that should bind WRAD perfectly well - and nobody has ever tested it.
The same thing, for a scientist
The brief for this beat, and several entries in this graph, state that the SET domain and its WRAD contacts are lost in EVERY KMT2A fusion. That is false as stated. A balanced t(4;11) produces der(11) KMT2A::AFF1 and der(4) AFF1::KMT2A (AF4-MLL). The reciprocal product is AF4 exons 1-3 fused to KMT2A exons 12-37 [E1], i.e. it carries the KMT2A C-terminus through the final exon: PHD1-4, the bromodomain, FYRC, the TAD, the WIN motif at 3762-3773 and the SET domain at 3829-3945. Every published WRAD contact surface on KMT2A is inside that fragment. The reciprocal transcript is expressed in about 45% of t(4;11) infant B-ALL [E2] and its presence predicts 4-year event-free survival of 62.4% versus 11.7% [E2]; Marschalek's group puts the loss of the AF4-MLL allele at about 40% of patients [E1]. AF4-MLL alone is sufficient to induce acute leukaemia in transplanted murine progenitors where MLL-AF4 alone is not [E3]. Its complex was affinity-purified and reported to contain H3K4-trimethyltransferase activity among sixteen constituents [E4]. So the honest answer to 'do WRAD subunits associate with a KMT2A fusion protein?' is: for the direct der(11) fusion, almost certainly not, and for the reciprocal der(4) fusion, structurally they must and nobody has published the subunit-resolved experiment.
The domain arithmetic
| Product | KMT2A portion | menin-binding motif (aa 6-25) | WIN motif (3762-3773) | SET (3829-3945) |
|---|---|---|---|---|
| der(11) KMT2A::AFF1 | exons 1-9/10/11 | retained | lost | lost |
| der(4) AFF1::KMT2A | exons 12-37 | lost | retained | retained |
| inv(11) NUP98::KMT2A | almost the entire ORF minus exon 1 | lost | retained | retained |
The two fusion products of one translocation are complementary in exactly the dimension this graph cares about. The direct product is menin-competent and WRAD-blind; the reciprocal product is WRAD-competent and menin-blind. A menin inhibitor engages one of them and by construction cannot engage the other.
How much of the disease this covers
t(4;11) is about 57% of KMT2A-rearranged pro-B ALL 1. Within t(4;11), the reciprocal transcript was detected in 45% of 124 Interfant-treated infants 2, and Marschalek's group states the AF4-MLL allele is absent or silent in about 40% of patients 1. Two mechanisms were proposed for the loss: complex translocations that never generate the allele, and breaks upstream of AFF1 exon 1 that leave no reciprocal reading frame. The second is negligible in this partner - 1 of 1,262 cases - so complex rearrangements, out-of-frame reciprocal alleles and transcriptional silence have to carry the whole gap between the ~80-94% who have the allele and the 45% who express it. Taking the published numbers at face value, roughly half of t(4;11) infant B-ALL expresses a KMT2A fusion protein with an intact WRAD interface, which is not a rare corner case.
One complication the same group flags: breakpoints in KMT2A intron 11, the commonest in infant ALL, disrupt the first PHD finger of the reciprocal protein and change its CYP33/PPIE binding 1. The PHD disruption does not touch the WIN motif or the SET domain, which sit roughly 2,200 residues further C-terminal, so the WRAD interface survives the infant-typical breakpoint.
What has actually been shown, and what has not
Shown. AF4-MLL alone induces pro-B ALL, B/T biphenotypic or mixed-lineage leukaemia in mice; MLL-AF4 alone produced no disease over 13 months in the same experiment 3. The AF4-MLL complex was affinity-purified and described as containing at least sixteen constituents including P-TEFb, an H3K4me3 methyltransferase, an H3K79me3 methyltransferase, a protein arginine N-methyltransferase and a histone acetyltransferase 4. In SEM and RS4;11 nuclear extracts, anti-RUNX1, anti-MLL-C and anti-AF4-N each co-immunoprecipitate a complex containing RUNX1, MLL-C, wild-type AF4 and a band assigned to cleaved AF4-MLL, and RBBP5 is described in the same paper as an MLL-C complex component whose chromatin occupancy at those target genes is RUNX1-dependent 5.
Not shown. No published experiment names WDR5, RBBP5, ASH2L or DPY30 as a constituent of an affinity-purified AF4-MLL complex, or blots any of them on an AF4-MLL immunoprecipitate. A Europe PMC query combining 'AF4-MLL' with 'WDR5' returns 29 records and all are reviews or unrelated primaries; combining it with RBBP5, ASH2L or DPY30 returns 28, likewise. The one primary paper in either set is the RUNX1 paper 5, and its RBBP5 datum is chromatin occupancy, not association with the fusion.
Why this changes the menin argument rather than merely adding to it
This graph's position is that menin is very nearly the only retained handle on the fusion. That holds for the der(11) product. It does not hold for the cell, because in 45% of t(4;11) cases - by transcript; 80-94% by DNA - the cell also contains a second KMT2A-derived oncoprotein that menin cannot reach and that carries an intact methyltransferase module. If AF4-MLL contributes to maintenance rather than only to initiation, a menin inhibitor is by construction a partial intervention in those patients - and the published survival data cut the other way, with reciprocal-positive patients doing better, which is the observation that most needs explaining before this is treated as a therapeutic gap.
| reciprocal product | AFF1::KMT2A (AF4-MLL), der(4) of t(4;11)(q21;q23) |
|---|---|
| kmt2a exons retained | 12-37 |
| reciprocal transcript positive PCT | 45 |
| reciprocal transcript cohort n | 124 |
| efs 4y reciprocal positive PCT | 62.4 |
| efs 4y reciprocal negative PCT | 11.7 |
| t411 share of kmt2a r prob all PCT | 57 |
| wrad subunit resolved test on reciprocal | none published |
kmt2a domains retained
[ "PHD1-4 (PHD1 disrupted by intron 11 breakpoints)", "bromodomain", "FYRC", "TAD", "WIN motif 3762-3773", "SET 3829-3945" ]
kmt2a domains lost
[ "MBM menin-binding motif aa 6-25", "LEDGF IBM1/IBM2", "AT-hooks", "CXXC" ]
- If AF4-MLL matters for maintenance in the roughly half of t(4;11) patients who express it, a menin inhibitor cannot reach it, and the menin-inhibitor eligibility criteria in use make no distinction between reciprocal-positive and reciprocal-negative patients.
- A WRAD-directed agent would engage the reciprocal fusion, the wild-type KMT2A allele, all five other SET1/MLL enzymes and the NSL complex simultaneously, so any efficacy could not be attributed to the fusion without a reciprocal-negative comparator arm.
- The prognostic direction is the wrong way round for a second-oncoprotein model: reciprocal-positive patients survive better. An intervention designed to remove AF4-MLL could therefore plausibly worsen outcome, and that possibility has to be excluded before the reciprocal product is treated as a target.
- Does WDR5, RBBP5, ASH2L or DPY30 co-immunoprecipitate with AF4-MLL in SEM or RS4;11? This is a one-week experiment that has never been reported and it is the direct answer to the question this beat was sent to settle.
- Is AF4-MLL H3K4 methyltransferase-active in cells, and if so at which loci? The complex was reported to carry H3K4me3 activity, but whether that activity is the reciprocal protein's own SET domain or a co-purified SETD1A/KMT2B has never been separated.
- Why do reciprocal-positive patients do better? If AF4-MLL supplies an intact, WRAD-loaded methyltransferase to a cell that already has one from the wild-type allele, the simplest explanations are dosage or differentiation, and neither has been tested.
- Is the reciprocal product present at the protein level in patient blasts, or only as transcript? Every clinical number in this entry is RT-PCR.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The reciprocal t(4;11) fusion AF4-MLL comprises AF4 exons 1-3 fused to MLL exons 12-37, so it retains the entire KMT2A C-terminus including the WIN motif and SET domain; t(4;11) is about 57% of KMT2A-rearranged pro-B ALL; the AF4-MLL allele is eliminated in about 40% of patients, which worsens outcome; and KMT2A intron 11 breakpoints, the commonest in infant ALL, disrupt the first PHD finger of the reciprocal protein.
The role of reciprocal fusions in MLL-r acute leukemia: studying the chromosomal translocation t(4;11). · Wilhelm A, Marschalek R · Oncogene 2021What this evidence cannot show
The construct description is of the Sleeping Beauty vectors used in that study, engineered to represent the infant-typical breakpoint, not a survey of patient breakpoints; the 40% figure is stated by the authors with a citation rather than measured here. The experiments in this paper are in HEK293T cells, not haematopoietic cells, and are transcriptomic/chromatin-accessibility rather than proteomic.
Evidence that argues against this (1)
- AF4-MLL expressed in human cord-blood CD34+ HSPCs transiently augmented multilineage engraftment but conferred no serial self-renewal and initiated no leukaemia within 36 weeks, so the reciprocal product's sufficiency in mouse does not transfer to human cells. Prieto C et al. 2017
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 124 infants with B-ALL treated on Interfant protocols the reciprocal AFF1::KMT2A transcript was expressed in only 45% of t(4;11) cases, and its expression predicted 4-year event-free survival of 62.4% versus 11.7%, retaining significance in a Cox model adjusted for risk stratification.
Unraveling the cellular origin and clinical prognostic markers of infant B-cell acute lymphoblastic leukemia using genome-wide analysis. · Agraz-Doblas A, Bueno C, Bashford-Rogers R, et al. · Haematologica 2019What this evidence cannot show
Transcript detection by RT-PCR is not protein detection, and a negative RT-PCR can reflect assay design rather than allele absence. The prognostic association is from a retrospective cohort and the direction - reciprocal-positive patients doing better - is the opposite of what a second oncoprotein would naively predict.
Evidence that argues against this (1)
- A CRISPR/Cas9-engineered t(4;11) in primary human haematopoietic stem/precursor cells, in which both reciprocal products are concomitantly present, shows lineage plasticity rather than a fixed phenotype, so the reciprocal product's contribution may be to plasticity rather than to aggressiveness, complicating any prognostic reading. Benz T et al. 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
AF4.MLL expressed in murine Lin-/Sca1+ progenitors induced pro-B ALL, B/T biphenotypic acute leukaemia or mixed-lineage leukaemia, whereas MLL.AF4 alone caused no disease over a 13-month observation period.
The AF4.MLL fusion protein is capable of inducing ALL in mice without requirement of MLL.AF4. · Bursen A, Schwabe K, Rüster B, et al. · Blood 2010What this evidence cannot show
Retroviral overexpression in mouse progenitors, non-physiological levels, and a small transplant cohort. Other groups have not reproduced the sufficiency of AF4-MLL in human cells.
Evidence that argues against this (1)
- AF4-MLL in human cord-blood CD34+ cells did not initiate leukaemia within 36 weeks despite augmenting engraftment, which bounds the mouse sufficiency result to the mouse system. Prieto C et al. 2017
- E4weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
The AF4-MLL protein complex was affinity-purified and reported to consist of at least sixteen constituents including P-TEFb kinase, H3K4(me3) and H3K79(me3) histone methyltransferases, a protein arginine N-methyltransferase and a histone acetyltransferase.
The leukemogenic AF4-MLL fusion protein causes P-TEFb kinase activation and altered epigenetic signatures. · Benedikt A, Baltruschat S, Scholz B, et al. · Leukemia 2011What this evidence cannot show
KNOWN BUT UNREADABLE. The full text is behind the Nature/Leukemia paywall: no PMCID exists in Europe PMC, fullTextXML is unavailable, and the publisher redirects to an authentication endpoint. Only the abstract was read. The abstract names activity classes, not gene symbols, so it is NOT established from this record whether any WRAD subunit is among the sixteen constituents, and this entry does not assert that it is. Resolving this requires the paper's supplementary table.
Evidence that argues against this (1)
- A Europe PMC search combining 'AF4-MLL' with 'WDR5' (29 records) or with RBBP5/ASH2L/DPY30 (28 records) returns no primary paper naming any WRAD subunit in an AF4-MLL purification, so if the sixteen constituents include WRAD, the fact has not propagated into the citing literature in twenty-five years. Europe PMC; query run by this project 2026
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In the t(4;11) lines SEM and RS4;11, anti-RUNX1, anti-MLL-C and anti-AF4-N immunoprecipitations each recover a complex containing RUNX1, MLL-C, wild-type AF4 and a band corresponding to the cleaved ~194 kDa AF4-MLL protein; RUNX1 knockdown reduces chromatin binding of AF4-N, MLL-C and the MLL-C complex component RBBP5 at the affected target genes.
RUNX1 is a key target in t(4;11) leukemias that contributes to gene activation through an AF4-MLL complex interaction. · Wilkinson AC, Ballabio E, Geng H, et al. · Cell Rep 2013What this evidence cannot show
The RBBP5 result is ChIP occupancy at target genes, not co-immunoprecipitation with the fusion, so it does not by itself establish an AF4-MLL/RBBP5 association. The anti-MLL-C antibody cannot distinguish wild-type MLL-C from the MLL-C fragment released from AF4-MLL by taspase1 cleavage, so 'MLL-C complex' in this paper is an ambiguous species. The authors state that their AF4-MLL-specific siRNAs failed to reduce AF4-MLL protein, so the causal test was not achieved.
Evidence that argues against this (1)
- The same authors note that an earlier group reported an AF4-MLL-specific siRNA had no effect on SEM cell growth; the dispute was resolved only by showing the siRNA did not knock the protein down, so the functional requirement for AF4-MLL in an established t(4;11) line remains formally untested in either direction. Wilkinson AC et al. 2013
- E-linesstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.journal-article
The two standard t(4;11) cell lines used for essentially all mechanistic work in this disease, RS4;11 and SEMK2-M1 (a SEM subline), both express the reciprocal AF4-MLL transcript as well as MLL-AF4. There is therefore no reciprocal-negative t(4;11) line in routine use.
t(4;11) leukemias display addiction to MLL-AF4 but not to AF4-MLL · Leukemia Research 2011What this evidence cannot show
Abstract read verbatim 2026-09-19: *"In the human leukemia cell lines - RS4;11 and SEMK2-M1, both of which express MLL-AF4 and AF4-MLL, we knocked down the expression of AF4-MLL using siRNA."* This is an RT-PCR-level statement about transcript, not protein, and it is made in passing as a premise rather than tested. SEMK2-M1 is a subline, so it does not strictly settle SEM itself, and **MV4-11 is not addressed**.
Evidence that argues against this (1)
- The same paper reports the reciprocal product is **dispensable** in both lines - siRNA knockdown of AF4-MLL had no effect on growth, cell cycle or apoptosis, while MLL-AF4 knockdown arrested and killed them. So 'every line is reciprocal-positive' is a weaker complaint than it sounds: the one experiment that removed the reciprocal product from those lines found nothing changed. The unstratified preclinical package matters for initiation and for drug response, not obviously for maintenance. Leukemia Research 2011
- E-assaystrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.journal-article
The frequency of the reciprocal allele depends entirely on which assay is named. At the DNA level roughly 80% of t(4;11) patients carry a reciprocal AFF1::KMT2A allele, and 94% of all KMT2A rearrangements generate a reciprocal fusion gene of some kind; at the transcript level only 45% of t(4;11) infants express it.
Complex MLL rearrangements in t(4;11) leukemia patients with absent AF4.MLL fusion allele · Leukemia 2007What this evidence cannot show
This entry previously said 'about half' without naming an assay, which happens to sit between the two real numbers and matches neither. The 80% figure is DNA-level from PMID 17410185, which also shows the other ~20% are not simply negative - they carry reciprocal KMT2A fused to NFKB1 or RABGAP1L and transcribed, or out-of-frame to PBX1/ELF2/DSCAML1/FXYD6, or a solitary 3' KMT2A. The 94% is from the Frankfurt archive (PMID 32294539) and covers all KMT2A rearrangements, not t(4;11). Cohorts, eras and referral patterns differ across all three.
Evidence that argues against this (1)
- The DNA-versus-transcript gap is itself unexplained, and it is the load-bearing number: a WRAD interface retained on an allele that is never transcribed does nothing. Nobody has shown the protein exists in patient blasts by any assay. Leukemia 2023
Supplies the residue-level domain inventory behind that entry's WRAD-competence claim.
Supplies the denominator behind that entry's 'about half of patients' framing and corrects it upward at the DNA level.
Reports H3K4me3 methyltransferase activity in the reciprocal complex, without separating the fusion's own SET domain from a co-purified enzyme.
Adds the WRAD consequence the existing entry did not draw: because the reciprocal product carries the WIN motif and SET domain, it is a KMT2A fusion protein with an intact WRAD interface.
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
The der(4) counterpart of this fusion, expressed in about 45% of t(4;11) infant B-ALL, is menin-blind and WRAD-competent - the mirror image of the der(11) product.
Supplies the architectural reason the remainder matters: what it retains is the whole methyltransferase module, and what it loses is the menin tether.
t(4;11) is about 57% of KMT2A-rearranged pro-B ALL and the reciprocal transcript is present in about 45% of those infants.
The reciprocal product fuses AFF1 exons 1-3 to KMT2A exons 12-37.
The reciprocal der product is the commoner of the two SET-retaining classes.
The ambiguity exists only because the reciprocal product is taspase1-cleaved to release an MLL-C fragment indistinguishable from the wild-type one.
The hypothesis exists because the der(4) product is menin-blind and present in about half of t(4;11) patients.
- ← testsSettle the WRAD question: tagged co-IP plus degron-timed proximity across three fusion classes
Arm C tags the der(4) product, which is the only available positive control for a WRAD-competent KMT2A fusion in the same cell.