The working attribution reagent in this literature is not an antibody, it is the gel - and that single fact predicts which claim classes survive
also called size resolution attribution · why the Western survives · readout dimension rule
Nobody has an antibody that sees only the fusion protein, so at first it looks as if nothing about the fusion can be trusted. But a fusion is a different size from either of the normal proteins it is made from, and any method that sorts proteins by size can therefore tell them apart even with a non-specific antibody.
The same thing, for a scientist
Across the five claim classes audited in this beat, whether a claim survives is predicted almost perfectly by one property of the **readout**, not of the reagent: does the method retain a size dimension? | readout | size dimension | claim classes it carries | audit verdict | |---|---|---|---| | Western blot of lysate or fraction | **yes** | half-life, turnover, abundance ratio | **survives** | | co-IP followed by Western | **yes**, for the bait | single-partner composition | **survives** | | AP-MS, shotgun proteomics | no | complex inventory, PTM discovery | fails | | immunofluorescence, live imaging | no | localisation | fails | | ChIP, CUT&RUN, CUT&Tag | no | chromatin occupancy | fails unless subtracted or tagged | The mechanism is simple arithmetic. KMT2A::AFF1 runs at about 240 kDa; wild-type AF4 runs at 125 and 145 kDa; wild-type KMT2A is cleaved by taspase1 into MLL-N320 and MLL-C180 [E1][E2]. Those are far apart on a 3-8% gel. **A shared-epitope antibody applied to a separated sample is a discriminating reagent**, because the separation supplies the specificity the antibody lacks. **Three consequences worth acting on.** 1. **Several failing classes can be rescued without any new reagent, by changing the readout.** Localisation: fractionate and blot instead of imaging. PTM occupancy: IP then phospho-blot instead of shotgun MS. Both are described in the relevant audits and neither has been published for KMT2A::AFF1. 2. **The two designs that rescue the no-size-dimension readouts both work by making a species unique rather than by resolving it** - an epitope tag, or a degron whose removal subtracts the fusion's contribution [E3]. That is why the knock-in is the highest-leverage single experiment in this beat. 3. **The junction peptide is the same idea applied at peptide level.** Mass spectrometry loses the protein's size but a junction-spanning peptide has a unique mass, so targeted MS restores attribution by uniqueness where shotgun MS lost it by fragmentation (`ana-junction-peptide-detectability-kmt2a-fusions`). **The general rule, stated so it transfers to other fusions:** *an assay attributes to a fusion if and only if it preserves some dimension - size, mass, sequence or an introduced tag - in which the fusion differs from both parents. Specificity does not have to live in the antibody.*
Where size resolution itself fails
It is not a universal escape and three cases break it.
- Species of similar mass. The reciprocal AF4::KMT2A product and wild-type KMT2A both carry the MLL-C epitope, and an MLL-C blot must separate them by size; where a breakpoint puts them close, it cannot. This graph records the resulting ambiguity for t(4;11) (
concept-the-mll-c-ambiguity-across-fusion-families-and-kmt2a-star). - Degradation products. A truncated or clipped fusion can run anywhere, and a band attributed to wild-type by position could be fusion-derived. Every blot that claims attribution has to show both species in the same lane so the assignment is visible.
- KMT2A*, which carries the C-terminus and no partner sequence 4 - a third band whose position nobody has established, because nobody has looked for it.
So size resolution earns attribution per blot, not in general, and the earning is visible: both bands must be shown. An entry citing a co-IP where only the partner was blotted has inherited an attribution that was never demonstrated.
| rule | an assay attributes to a fusion if and only if it preserves a dimension - size, mass, sequence, or an introduced tag - in which the fusion differs from both parents |
|---|
readout table
[
{
"readout": "Western blot of lysate or fraction",
"size_dimension": true,
"attributes": true
},
{
"readout": "co-IP then Western",
"size_dimension": true,
"attributes": "for the bait band"
},
{
"readout": "AP-MS / shotgun proteomics",
"size_dimension": false,
"attributes": false
},
{
"readout": "immunofluorescence / live imaging",
"size_dimension": false,
"attributes": false
},
{
"readout": "ChIP / CUT&RUN / CUT&Tag",
"size_dimension": false,
"attributes": "only with a tag or degron subtraction"
},
{
"readout": "targeted MS on the junction peptide",
"size_dimension": "mass uniqueness",
"attributes": true
},
{
"readout": "junction RT-PCR",
"size_dimension": "sequence uniqueness",
"attributes": true
}
]species and apparent masses
[
{
"species": "KMT2A::AFF1 fusion",
"apparent_kDa": 240
},
{
"species": "wild-type AF4",
"apparent_kDa": [
125,
145
]
},
{
"species": "wild-type KMT2A after taspase1",
"fragments": [
"MLL-N320",
"MLL-C180"
]
}
]where size resolution fails
[ "reciprocal product versus wild-type MLL-C at similar mass", "degradation products migrating unpredictably", "KMT2A*, whose band position has never been established" ]
rescues available without new reagents
[ "fractionate and blot instead of imaging, for localisation", "IP then phospho-blot instead of shotgun MS, for PTM occupancy" ]
- Pharmacodynamic assays in trials are moving towards antibody-based and imaging-based readouts for convenience, both of which discard the size dimension; a trial biomarker that cannot separate fusion from wild-type will report target engagement that may be on the wrong protein.
- If a clinical assay is built on an apparent molecular weight and a patient's breakpoint shifts the fusion's mass, the band moves and the assay reads negative - a false negative driven by a genotype the assay never measures.
- The cheapest fixes named here are all changes of readout rather than of reagent, so they are within reach of any laboratory already working on this disease; a programme that instead waits for a junction-specific antibody defers attribution indefinitely for no methodological reason.
- Where does KMT2A* run on a gel, and is it separable from wild-type MLL-C180? If the two co-migrate, the size-resolution escape fails for every MLL-C measurement in this disease and nobody would know.
- Does KMT2A::AFF1 migrate anomalously? Its computed length is 2270 residues and the literature figure is an apparent 240 kDa; nobody has reported a calibrated comparison, and a systematic offset would change how bands are assigned in every blot in the field.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In t(4;11) lines the MLL-AF4 fusion is detected at about 240 kDa, resolved from wild-type AF4 at 125 and 145 kDa, using the same rabbit polyclonal reagents on the same blot.
AF4 encodes a ubiquitous protein that in both native and MLL-AF4 fusion types localizes to subnuclear compartments. · Li Q, Frestedt JL, Kersey JH. · Blood 1998What this evidence cannot show
Abstract read in full via the Europe PMC core record; no PMCID and the full text was not reached. The molecular weights are apparent weights on SDS-PAGE, not computed masses, and breakpoint variation changes the fusion's true mass between patients.
Evidence that argues against this (1)
- Apparent molecular weight on SDS-PAGE is not a reliable identifier on its own: large, disordered, heavily modified proteins migrate anomalously, and KMT2A is all three. A band at the expected position is consistent with the fusion rather than proof of it, which is why the junction peptide remains the stronger identification. Domer PH et al. 1993
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Wild-type MLL is processed into MLL-N320 and MLL-C180 fragments whose levels can be followed independently on a blot, and a cell-cycle expression study read both fragments separately alongside the fusion in the same system.
Bimodal degradation of MLL by SCFSkp2 and APCCdc20 assures cell cycle execution: a critical regulatory circuit lost in leukemogenic MLL fusions. · Liu H, Cheng EH, Hsieh JJ. · Genes & development 2007What this evidence cannot show
Read verbatim from the PMC HTML page for PMC1993870 after the XML routes failed. The fragment nomenclature and the separation are the paper's own; the audit use made of it here - that separation supplies the attribution - is this beat's argument rather than the paper's claim.
Evidence that argues against this (1)
- MLL-N320 and MLL-C180 remain associated as a complex after taspase1 cleavage, so a co-IP with an N-terminal antibody pulls down the C-terminal fragment too. Separation on a denaturing gel is therefore essential; a native pulldown does not separate the species at all. Yokoyama A et al. 2010
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Where a readout has no size dimension, attribution is restored either by an epitope unique to the fusion - anti-HA ChIP-seq on HA-FKBP12-tagged MLL::AF9 - or by subtracting the fusion's contribution through degron-induced degradation.
MLL::AF9 degradation induces rapid changes in transcriptional elongation and subsequent loss of an active chromatin landscape. · Olsen SN, Godfrey L, Healy JP, et al. · Molecular cell 2022What this evidence cannot show
Read verbatim from the PMC body (PMC9044330, efetch db=pmc, <body> present). An overexpression system, MLL::AF9 rather than KMT2A::AFF1. The generalisation offered here - that these are the only two ways to restore attribution in a dimensionless readout - is this beat's argument, not a claim of the paper.
Evidence that argues against this (1)
- There is at least a third route that this beat did not find used in KMT2A leukaemia: a fusion-junction-specific antibody, which restores attribution in the reagent itself. Synovial sarcoma's SS18-SSX E9X9V shows the class is achievable, and this graph records that no such reagent exists for any KMT2A fusion. Kerry J et al. 2017
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The canonical KMT2A::AFF1 chimera joins KMT2A residue 1406 to AFF1 residue 347, giving a chimeric protein of 2270 residues whose mass differs from both parents.
Acute mixed-lineage leukemia t(4;11)(q21;q23) generates an MLL-AF4 fusion product. · Domer PH, Fakharzadeh SS, Chen CS, et al. · Proc Natl Acad Sci U S A 1993What this evidence cannot show
The junction anchor is GenBank AAC37520.1 from this paper; the residue arithmetic and the chimeric length are this project's own computation in analyses/junction-peptide-detectability, independently reproduced by the coordinator. Breakpoint variation moves the junction between patients and therefore the mass.
Evidence that argues against this (1)
- A computed mass of 2270 residues is not the apparent weight on a gel: KMT2A migrates anomalously, and the 240 kDa figure in the 1998 literature is an apparent weight rather than a computed one. The two numbers should not be conflated when assigning a band. Li Q et al. 1998
the 2007 fusion measurement is anti-AF4-C-terminal IP blotted with anti-MLL aa 720-780 and discriminated by gel size — another instance of the size dimension doing the attribution work
explains why the turnover class is the one that survives
explains why the PTM class is the one with no partial survivor
names the case where size resolution itself breaks down
the absence of a specific antibody is survivable because specificity can live in the readout instead
the single property that predicts the verdict table
applies the size-dimension rule to the localisation class and names the unrun experiment