Oncofusion

start / graph / ana-attribution-audit-protein-half-life-kmt2a-aff1

Analysisstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q27b-fusion-attribution

Attribution audit 4 of 5 - the half-life and turnover of KMT2A::AFF1: this class SURVIVES, and it is the only endogenous fusion-attributable protein measurement this beat found

also called MLL-AF4 turnover audit · fusion stability attribution · Skp2 Cdc20 fusion resistance audit

How long the fusion protein lasts inside a cell was measured in real patient-derived leukaemia cells, and it counts, because the fusion is a different size from the normal protein and both were visible in the same lane of the same gel. It is the one thing in this whole audit that was measured on the real fusion in a real leukaemia cell.

The same thing, for a scientist

This is the class that survives, and it survives for a specific reason worth naming: **a Western blot has a size dimension, so a shared-epitope antibody becomes a discriminating reagent the moment the readout separates the species.** **The surviving measurement.** Liu, Cheng and Hsieh separated **SEM** cells - a patient-derived KMT2A::AFF1 line - into cell-cycle fractions by centrifugal elutriation rather than by chemical synchronisation, and blotted. Wild-type MLL shows biphasic expression, peaking at G1/S and G2/M and being degraded at S and late M by SCF(Skp2) and APC(Cdc20) respectively. **MLL-AF4 protein did not fluctuate through the cell cycle** in the same cells, and its level was not markedly increased [E1]. Both species are in the same lysate, resolved by size, read with the same antibody. That is a fusion-attributable statement about the endogenous protein in a human leukaemia cell, and this beat found no other. **A methodological fact hiding inside it.** They tried and failed to synchronise SEM, RS4;11, THP-1, NOMO-1 and MOLM-13 with mimosine, double thymidine or nocodazole, and say so explicitly, inferring inherent cell-cycle defects [E1]. Any pulse-chase or cycloheximide-chase design in these lines that assumes chemical synchronisation works has an unstated failure mode, and elutriation is the published workaround. **What does not survive.** The mechanism - that MLL fusions escape degradation because they lose the Skp2 and Cdc20 contacts - was established on **FLAG-tagged MLL-AF4 transiently expressed in 293T, immunoprecipitated with anti-FLAG and degraded in vitro by synchronised lysates** [E1]. That is attributable to the fusion by tag, but it is a kidney cell line, transient overexpression, and an in vitro degradation rate rather than a cellular half-life. The published **cycloheximide half-life number**, extended roughly eightfold by MG132, was measured on **wild-type MLL in HeLa** [E1] - it is not a fusion number and should never be quoted as one. **The reciprocal product has its own, separate turnover literature.** AF4 wild-type and AF4.MLL differ in their interaction with SIAH proteins [E2]; that work is about the reciprocal and is audited separately. **And the degron kinetics this graph cites are not endogenous.** Every published degron measurement of an MLL fusion is on a retrovirally overexpressed, tagged construct, not on a tagged endogenous allele - see `concept-degron-on-the-fusion-allele-versus-on-an-overexpressed-construct` [E3].

Go deeper
Why this class survived when the other four did not

The discriminator is not the antibody. It is the gel.

readouthas a size dimension?attributable with a shared-epitope antibody?
Western blot of whole-cell or fractionated lysateyesyes
co-IP followed by Westernyes, for the baityes for the bait band
AP-MS / shotgun proteomicsnono
immunofluorescencenono
ChIP / CUT&RUNnono

Four of the five claim classes in this audit are read out by a method in the lower half of that table. Turnover is the one that is read out by the method in the upper half, and that is the whole of the difference.

The practical corollary is cheerful: several of the failing classes could be rescued by changing the readout rather than the reagent. Chromatin fractionation plus blot would attribute localisation. IP plus phospho-blot would attribute a PTM's occupancy. Neither needs a new antibody.

What is still missing in this class

A number. There is no published half-life for KMT2A::AFF1 protein in any cell - only the qualitative statement that it does not oscillate 1. That matters for this project directly, because the decay curve after menin-inhibitor treatment is being interpreted against an assumed protein turnover rate (ana-fusion-halflife-degron-vs-menin), and the rate is assumed rather than measured.

The experiment is a cycloheximide chase in SEM with a size-resolving blot, quantifying the ~240 kDa fusion band and the wild-type MLL-N320 band from the same lanes. It is attributable by the argument above, it needs no new reagent, and it is a day of work. That it has not been done is the most surprising gap this audit found.

Facts and figures
claim classprotein half-life and turnover of KMT2A::AFF1
verdictSURVIVES - the only endogenous, fusion-attributable protein measurement this audit found
method noteSEM, RS4;11, THP-1, NOMO-1 and MOLM-13 could not be synchronised with mimosine, double thymidine or nocodazole; centrifugal elutriation is the published workaround
reagent classes found
[
  {
    "reagent": "anti-MLL Western blot on elutriated SEM fractions",
    "system": "SEM, patient-derived t(4;11)",
    "fusion_attributable": "yes",
    "why": "fusion and wild-type MLL resolved by size in the same lane; endogenous"
  },
  {
    "reagent": "anti-FLAG IP of FLAG-MLL-AF4 plus in vitro degradation",
    "system": "293T transient transfection",
    "fusion_attributable": "yes",
    "why": "tag is unique to the fusion; but not endogenous and an in vitro rate"
  },
  {
    "reagent": "cycloheximide chase plus anti-MLL blot",
    "system": "HeLa",
    "fusion_attributable": "not applicable",
    "why": "no fusion present; the roughly eightfold MG132 half-life extension is a wild-type MLL number"
  },
  {
    "reagent": "dTAG FKBP12-F36V degron plus anti-HA",
    "system": "retrovirally overexpressed MLL::AF9",
    "fusion_attributable": "yes",
    "why": "tag unique to the fusion; overexpressed, not knocked in"
  }
]
minimal fix
[
  "cycloheximide chase in SEM with a size-resolving blot quantifying the ~240 kDa fusion band against wild-type MLL-N320 in the same lanes - no new reagent, and it would supply the half-life number the graph is currently assuming"
]
readout size dimension table
[
  {
    "readout": "Western blot",
    "size_dimension": true
  },
  {
    "readout": "co-IP then Western",
    "size_dimension": true
  },
  {
    "readout": "AP-MS / shotgun",
    "size_dimension": false
  },
  {
    "readout": "immunofluorescence",
    "size_dimension": false
  },
  {
    "readout": "ChIP / CUT&RUN",
    "size_dimension": false
  }
]
How this could bite us back
  • If the fusion does not oscillate through the cell cycle while wild-type MLL does, then a degrader or a destabilising agent has no cell-cycle window of natural vulnerability to exploit, and dosing schedules built on cell-cycle timing will not gain anything.
  • Menin-inhibitor washout and resistance modelling depends on the fusion's protein half-life, which is unmeasured; if the true half-life is long, a drug holiday preserves the oncoprotein and the apparent hysteresis in recovery curves is protein persistence rather than an epigenetic memory effect.
  • Every published degron washout kinetic in this disease comes from an overexpressed construct, so the recovery times the field quotes are set by a promoter someone chose rather than by the endogenous locus - an intervention timed against those numbers may be timed against an artefact.
Still open
  1. What is the half-life of KMT2A::AFF1 protein in SEM? No number exists anywhere, the experiment needs no new reagent, and this graph's menin-inhibitor decay modelling currently assumes a rate rather than using one.
  2. Is the non-oscillation of MLL-AF4 through the cell cycle a property of the fusion or of the elutriation fractions? A second t(4;11) line, or a tagged allele read by flow cytometry, would separate them.
Every source behind this entry (4)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In SEM cells separated by centrifugal elutriation, MLL-AF4 protein did not fluctuate through the cell cycle whereas wild-type MLL showed biphasic expression; the authors were unable to synchronise SEM, RS4;11, THP-1, NOMO-1 or MOLM-13 with mimosine, double thymidine or nocodazole; and the mechanistic claim that MLL fusions resist SCF(Skp2) and APC(Cdc20) degradation was established on FLAG-tagged MLL-AF4 transiently expressed in 293T and degraded in vitro, while the roughly eightfold MG132-dependent half-life extension was measured on wild-type MLL in HeLa.

    PMID 17908926DOI
    What this evidence cannot show

    Read verbatim from the PMC HTML page for PMC1993870 (91,654 characters) after Europe PMC fullTextXML returned HTTP 500 and efetch db=pmc returned a document with no <body> element - the three-failure pattern the brief warns about, recovered by the fourth route. The elutriation result is a single figure panel in one line; it is qualitative (does not fluctuate) rather than a half-life, and it was not replicated in a second t(4;11) line in this paper.

    Evidence that argues against this (1)
    • An alternative reading of a non-oscillating band is that the elutriation fractions were less pure for this line than for HeLa, flattening a real oscillation. The paper reports FACS DNA-content purity for the fractions, but a flattened oscillation and a genuinely constant protein look the same if purity is poor, and the failure of three chemical synchronisation methods in the same cells is a reason to take that possibility seriously rather than to dismiss it. Olsen SN et al. 2022
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Wild-type AF4 and the AF4.MLL fusion protein differ in their interaction with SIAH proteins, giving the reciprocal product a turnover literature distinct from that of the direct fusion.

    PMID 15221006DOI
    What this evidence cannot show

    Cited from the PubMed record for the existence and direction of this literature only; the full text was not retrieved, so the reagents and expression system behind the SIAH interaction are not audited here and no quantity from this paper is asserted. The molecule is AF4::KMT2A, the reciprocal product, not KMT2A::AFF1.

    Evidence that argues against this (1)
    • If the reciprocal product is SIAH-regulated and the direct fusion is Skp2/Cdc20-resistant, the two products of the same translocation have different turnover control, so any pharmacology that stabilises or destabilises one may not touch the other - and about half of KMT2A::MLLT1 patients express no reciprocal product at all, which is the natural experiment that would separate them. Liu H et al. 2007
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Published degron kinetics for an MLL fusion were measured on a retrovirally overexpressed HA-FKBP12-tagged MLL::AF9 construct in murine LSK-derived leukaemias and in human CD34+ cells, not on a degron knocked in at an endogenous fusion allele; the paper's own Limitations section states that future studies should include endogenously degron-tagged MLL::AF9.

    PMID 35245435DOI
    What this evidence cannot show

    Read verbatim from the PMC body (PMC9044330, efetch db=pmc, <body> present). MLL::AF9 rather than KMT2A::AFF1. The degradation kinetics themselves are attributable, because the FKBP12 tag is unique to the fusion; the limitation is that the abundance and the two intact wild-type MLL1 alleles are those of an overexpression model.

    Evidence that argues against this (1)
    • The authors addressed the compensation worry directly and found none: three hours of dTAG treatment had no effect on H3K4me3, and MLL-N versus MLL-C ChIP showed reductions only in the fusion and not in wild-type MLL1 occupancy. So the overexpression caveat is about stoichiometry, not about an uncontrolled wild-type confound. Olsen SN et al. 2022
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The fusion and its wild-type parent are separately resolvable in the same lane: the fusion runs at about 240 kDa and wild-type AF4 at 125 and 145 kDa in t(4;11) lines.

    PMID 9808577
    What this evidence cannot show

    Abstract read in full via the Europe PMC core record; no PMCID and the full text was not reached. This is the load-bearing premise of the whole class verdict: the method supplies the discrimination the reagent lacks.

    Evidence that argues against this (1)
    • Size resolution fails where the two species run close together, which is a real risk for the reciprocal product against wild-type MLL-C, and it fails entirely for a degraded or truncated form. It is a property of a particular fusion and gel, not a general guarantee, and each blot has to show both bands to earn the attribution. Liu H et al. 2007
Connected to