The anti-MLL-C ambiguity is not a t(4;11) problem - it has three sources, and one KMT2A fusion family is measurably cleaner than the rest
also called MLL-C ambiguity generalised · how many proteins does an anti-MLL-C antibody see · KMT2A::MLLT1 as the clean arm
An antibody against the back half of KMT2A can be looking at up to three different proteins at once, in any KMT2A leukemia, not just t(4;11). In the KMT2A::MLLT1 family about half of patients make no mirror-image protein at all, so in those patients the antibody is looking at fewer things - which makes that group the cleanest comparison available.
The same thing, for a scientist
`concept-an-mll-c-antibody-cannot-separate-wild-type-from-reciprocal-fusion` establishes the ambiguity for t(4;11). It generalises, because the ambiguity is a property of the **KMT2A C-terminus**, which is retained by every species other than the direct fusion. In any KMT2A-rearranged cell an anti-MLL-C reagent may be reading three proteins: (1) wild-type KMT2A, as MLL-C p180 after taspase1 cleavage; (2) the in-frame reciprocal product, cleaved at the same retained taspase1 sites and releasing a sequence-identical MLL-C fragment; (3) **KMT2A\***, the 5'-truncated product of the gene-internal promoter upstream of KMT2A exon 12, which carries the C-terminus and the SET domain while binding neither menin, LEDGF nor MYB. The direct fusion is the only KMT2A-derived species the reagent cannot see at all. The ambiguity is therefore **worse than the two-way problem already recorded**, and it applies to KMT2A::MLLT3, KMT2A::MLLT1, KMT2A::ELL and every other family equally. The one place it is measurably smaller is **KMT2A::MLLT1**: of 460 MLLT1 fusions, about 50% (n = 241) break upstream of MLLT1 exon 1 and are spliced fusions in which *"only KMT2A::MLLT1 fusion transcripts but no reciprocal one are expressed"* [E1]. In that half, source (2) is absent by construction. The same paper gives smaller reciprocal-negative fractions for ELL (10 of 119) and EPS15 (12 of 51).
Three sources, and what removes each
| source of MLL-C signal | removable by | cost |
|---|---|---|
| wild-type KMT2A allele | acute degron on the wild-type allele; or comparison to a KMT2A-germline line | high / moderate |
| in-frame reciprocal fusion | choosing a reciprocal-negative patient or family - KMT2A::MLLT1 spliced fusions are ~50% reciprocal-negative | free |
| KMT2A* | nothing currently available. No junction, no N-terminus, no assay | - |
The third row is why this cannot be fully solved by reagent choice. KMT2A* is invisible to junction RT-PCR (it has no junction), invisible to an anti-MLL-N antibody (it has no N-terminus), and indistinguishable from wild-type MLL-C by the C-terminal reagent. The only handle on it would be a truncation-specific N-terminal neo-epitope at the internal-promoter start site, which has never been raised.
The practical consequence for experiment design is that the cleanest MLL-C experiment available today is in a KMT2A::MLLT1 spliced-fusion patient or line, where the reciprocal source is absent, leaving two species rather than three. This graph already identifies that population as an untapped control arm for a different question; the attribution argument is an independent reason to use it.
Why generalising matters more than it sounds
The t(4;11)-only framing invites the reading that MLL-AF9 and MLL-ENL work is safe because those fusions retain no KMT2A C-terminus. That is true of the direct fusion and false of the cell. A t(9;11) cell with an in-frame reciprocal MLLT3::KMT2A carries exactly the same taspase1-cleavable KMT2A C-terminus, and an out-of-frame one may carry KMT2A*. Every 'MLL-C complex' or 'MLL occupancy' result in any KMT2A-rearranged system therefore has the same unresolved mixture behind it, and the fraction of patients affected is a function of the partner gene's reciprocal-expression rate, which is published and differs by an order of magnitude between partners.
Nobody has ever stratified an MLL-C experiment by reciprocal status, in any fusion family.
| cleanest available system | KMT2A::MLLT1 spliced-fusion patients or lines - two MLL-C sources rather than three |
|---|---|
| extends | concept-an-mll-c-antibody-cannot-separate-wild-type-from-reciprocal-fusion |
| no stratified experiment exists | true |
sources of mll c signal
[ "wild-type KMT2A (MLL-C p180 after taspase1 cleavage)", "in-frame reciprocal partner::KMT2A fusion (same cleavage, same fragment)", "KMT2A* (5'-truncated internal-promoter product)" ]
species invisible to this reagent
[ "the direct KMT2A::partner fusion" ]
reciprocal negative fractions by partner — the numbers behind the chart
[
{
"partner": "MLLT1",
"reciprocal_negative": 241,
"total": 460,
"mechanism": "breakpoint upstream of MLLT1 exon 1, spliced fusion"
},
{
"partner": "ELL",
"reciprocal_negative": 10,
"total": 119,
"mechanism": "spliced fusion"
},
{
"partner": "EPS15",
"reciprocal_negative": 12,
"total": 51,
"mechanism": "spliced fusion"
},
{
"partner": "AFF1",
"reciprocal_negative": 1,
"total": 1262,
"mechanism": "break upstream of AFF1 exon 1; 0.08%, not ~50%"
}
]- If a therapeutic strategy targets 'MLL-C-containing complexes', it targets the wild-type allele and KMT2A* along with anything fusion-derived - in an infant, and against a protein required for normal haematopoiesis.
- Choosing KMT2A::MLLT1 as the clean system trades one confound for another: MLLT1 fusions have a different age distribution and a different disease biology from t(4;11), so a result obtained there does not transfer to the disease this graph is mainly about without an explicit argument.
- Would an anti-MLL-C experiment run in matched KMT2A::MLLT1 spliced-fusion and non-spliced-fusion samples show a measurable difference in signal? That is a free control arm that has never been used, and a null result would put an upper bound on the reciprocal product's contribution.
- Is KMT2A* protein present in patient blasts at all? Until someone raises a neo-epitope antibody against the internal-promoter start or finds its unique N-terminal peptide, the third source of MLL-C signal is unquantified and unquantifiable.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
About half of KMT2A::MLLT1 fusions (241 of 460) break upstream of MLLT1 exon 1 as spliced fusions in which only the direct fusion transcript, and no reciprocal one, is expressed; the equivalent fractions are 10 of 119 for ELL and 12 of 51 for EPS15.
The KMT2A recombinome of acute leukemias in 2023. · Meyer C, Larghero P, Almeida Lopes B, et al. · Leukemia 2023What this evidence cannot show
Read verbatim from PMC10169636 on 2026-09-20: "Spliced fusions are rare events except for MLLT1 gene fusions (n = 460), where about 50% of the breakpoints (n = 241) localize upstream of MLLT1 exon 1. In these cases, only KMT2A::MLLT1 fusion transcripts but no reciprocal one are expressed." This is a statement about **transcript**, from a diagnostic archive, and it is an inference from breakpoint position rather than a direct measurement of reciprocal transcript in each case. It does not address KMT2A*, which would be produced from the same truncated 3'-KMT2A allele.
Evidence that argues against this (1)
- The same architecture that removes the reciprocal fusion leaves a truncated 3'-KMT2A allele - precisely the substrate the gene-internal promoter acts on to produce KMT2A*. So a spliced-fusion patient may trade one invisible MLL-C source for another, and the swap has never been measured. Scharf S et al. 2007
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Taspase1 cleavage of KMT2A generates the MLL-C p180 fragment, and cleavage status regulates MLL stability and chromatin association - so the MLL-C species an antibody detects is produced by every KMT2A-derived protein that retains the cleavage sites, not only by the wild-type allele.
Regulation of MLL/COMPASS stability through its proteolytic cleavage by taspase1 as a possible approach for clinical therapy of leukemia. · Zhao Z, Wang L, Volk AG, et al. · Genes Dev 2019What this evidence cannot show
The paper establishes the biochemistry of cleavage and its consequences for the wild-type protein; the extension to reciprocal fusions and to KMT2A* is an inference from those species retaining the cleavage sites and is carried at that strength here.
Evidence that argues against this (1)
- If taspase1-dependent destabilisation removes cleaved fragments efficiently, the steady-state contribution of reciprocal-derived and KMT2A*-derived MLL-C may be small enough that the ambiguity is quantitatively minor even though it is qualitatively real. Nobody has measured the relative abundances, so the size of the problem is unknown in both directions. Zhao Z et al. 2019
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A gene-internal promoter within the KMT2A breakpoint cluster region drives an mRNA encoding an N-terminally truncated MLL protein that retains the C-terminal functional regions including the SET domain.
Transcription linked to recombination: a gene-internal promoter coincides with the recombination hot spot II of the human MLL gene. · Scharf S, Zech J, Bursen A, et al. · Oncogene 2007What this evidence cannot show
Transcript- and reporter-level characterisation. Whether the truncated protein is made at meaningful levels in patient blasts has not been demonstrated, so the magnitude of this third contribution to an MLL-C signal is unknown. It is included here because an unmeasured source is still a source, not because its size is established.
Evidence that argues against this (1)
- If KMT2A* protein is not in fact produced at appreciable levels in leukaemic blasts, the MLL-C ambiguity reduces to the two-source problem already recorded and this entry overstates it. No experiment has looked, which is itself the point. Meyer C et al. 2023
extends the two-source ambiguity to three sources and to every KMT2A fusion family
- depends on →KMT2A*: a menin-blind, LEDGF-blind, SET-competent KMT2A protein made from out-of-frame reciprocal alleles
the third, unquantified source of MLL-C signal
- related to →KMT2A::MLLT1 (MLL-ENL), t(11;19)(q23;p13.3)
241 of 460 MLLT1 fusions are reciprocal-negative spliced fusions, making this the cleanest anti-MLL-C system available
- related to →KMT2A::ELL (MLL-ELL), t(11;19)(q23;p13.1)
10 of 119 ELL fusions are reciprocal-negative spliced fusions
the two reagent classes have complementary coverage and neither attributes
the MLLT1 reciprocal-negative arm is the free control that bounds the MLL-C ambiguity empirically
KMT2A* and the reciprocal product are the two C-only species that bias the subtraction
adds the RNA-level half: junction RT-PCR is structurally blind to KMT2A* because KMT2A* has no junction
names the case where size resolution itself breaks down
a junction epitope is the one reagent class that can separate fusion from wild-type and from KMT2A*