Oncofusion

start / graph / exp-endogenous-tag-and-degron-knock-in-kmt2a-aff1

Experimentmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q27b-fusion-attribution

Knock a tag-plus-degron cassette into the endogenous KMT2A::AFF1 allele in SEM - one construct that makes five claim classes attributable

also called SEM knock-in tagging · HA-FKBP12 KMT2A::AFF1 knock-in · endogenous degron t(4;11)

Put a small molecular handle, plus a switch that lets you destroy the protein on command, onto the fusion gene in a patient-derived leukaemia cell line - on the real gene, not an extra copy. Every measurement made through that handle can only be seeing the fusion, which fixes five different kinds of claim at once.

The same thing, for a scientist

**Design.** CRISPR knock-in of an in-frame `HA-FKBP12(F36V)` cassette at the 3' end of the **der(11) KMT2A::AFF1** open reading frame in SEM, with RS4;11 as a second host. Tag C-terminal, on the AFF1-derived portion: an N-terminal tag on MLL::AF9 failed to reproduce the parent's leukaemogenesis kinetics, which the authors attribute to interference with the N-terminal DNA-binding and menin-binding elements [E1]. Allele assignment verified by PCR across the fusion junction; at least three independent clones plus an untagged sibling clone. **Readouts, one per failing claim class.** | class | readout on the tagged line | what it replaces | |---|---|---| | composition | anti-HA AP-MS | anti-AF4 AP-MS from K562, which contains no fusion [E2] | | localisation | anti-HA IF and, with a HaloTag variant, live imaging | anti-AF4 IF dominated by the more abundant wild-type pool [E3] | | occupancy | anti-HA CUT&RUN, plus degron subtraction against MLL-N | MLL-N ChIP, which contains wild-type MLL [E4] | | turnover | dTAG-13 washout, and cycloheximide chase read on the tag | degron kinetics from a retroviral construct [E1] | | PTMs | anti-HA IP to concentrate the fusion, then junction-peptide PRM | shotgun phosphoproteomics, which cannot assign a shared peptide | **Controls that decide whether the line is usable.** Growth rate and colony formation against the untagged sibling; menin-inhibitor IC50 against the parent, because a tag that changes menin sensitivity invalidates the line for the project's main question; xenograft engraftment; and an MLL-N ChIP on the tagged line, whose overlap with the anti-HA CUT&RUN is **the first direct measurement of how much of an MLL-N peak set is actually fusion**. That last comparison alone would justify the experiment. **Known practical constraint.** SEM, 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 [E5]. Plan any cell-cycle-resolved arm accordingly. **Cost tier: medium.** A knock-in campaign in a hard-to-transfect suspension line, plus clone validation, is a few months; every readout after that uses standard methods and commercial anti-HA reagents.

Go deeper
Why this and not a junction-specific antibody

A junction antibody would be better, because it would work in primary patient material and would need no engineering. None exists for any KMT2A fusion, and raising one is a multi-year programme with an uncertain endpoint (absence-no-junction-specific-antibody-exists-for-any-kmt2a-fusion).

The knock-in is the available substitute, with one important limitation: it works only in engineered lines, never in a patient sample. The two therefore divide the problem. The knock-in validates the antibody-based methods in a line, which tells the field how much to trust the same methods applied to patient material. The junction peptide (exp-junction-peptide-prm-assay-kmt2a-fusions) is the route that reaches the patient sample, because a peptide needs no reagent at all.

Falsifiers and what a negative result would look like
  • If anti-HA CUT&RUN in the tagged line reproduces the published MLL-N map closely, then the field's occupancy claims were right all along and this audit's occupancy finding is a formal worry rather than a practical one. That is a real possible outcome and it would be worth knowing.
  • If the tagged line's menin-inhibitor IC50 differs from the parent's, the line is not usable for this project's central question and the experiment has failed at the control stage.
  • If anti-HA AP-MS returns essentially the AEP inventory, the composition claims transfer and only the numbers needed qualification.

The experiment is worth running precisely because all three of those are plausible and none is currently known.

Facts and figures
designCRISPR knock-in of an in-frame HA-FKBP12(F36V) cassette at the 3' end of the der(11) KMT2A::AFF1 ORF in SEM, replicated in RS4;11; tag C-terminal on the AFF1-derived portion; allele verified by PCR across the fusion junction; >=3 independent clones plus an untagged sibling clone
readoutanti-HA AP-MS (composition); anti-HA IF and HaloTag live imaging (localisation); anti-HA CUT&RUN plus degron subtraction against MLL-N (occupancy); dTAG-13 washout and cycloheximide chase read on the tag (turnover); anti-HA IP then junction-peptide PRM (PTMs)
controlsgrowth and colony formation versus untagged sibling; menin-inhibitor IC50 versus parent; xenograft engraftment; MLL-N ChIP on the tagged line, overlapped with anti-HA CUT&RUN to measure the fusion fraction of an MLL-N peak set
cost tiermed
timeline3-6 months to a validated clone panel; each readout 2-6 weeks thereafter
constraintchemical synchronisation fails in SEM, RS4;11, THP-1, NOMO-1 and MOLM-13; use centrifugal elutriation or a genetically encoded cell-cycle reporter
falsifiers
[
  "anti-HA CUT&RUN closely reproduces the published MLL-N map, indicating the occupancy attribution concern is formal rather than practical",
  "the tagged line's menin-inhibitor IC50 differs from the parent's, invalidating the line for this project's central question",
  "anti-HA AP-MS returns the AEP inventory, indicating composition claims transfer and only their numbers needed qualification"
]
How this could bite us back
  • If the tag attenuates the fusion even slightly, the tagged line will over-report drug sensitivity, and because it would become the field's reference model that bias would propagate into every subsequent preclinical decision.
  • A degron cassette adds an FKBP12 domain to a protein whose N-terminal region is the drug target; if the added bulk changes menin-pocket accessibility, the line becomes the wrong model for exactly the interface this project cares about, and the IC50 control is the only thing standing between that and a wrong conclusion.
  • Single-cell-derived knock-in clones confound the tag with clonal drift, so a phenotype seen in one clone is not attributable to the tag; three clones and an untagged sibling are the minimum, and a result that holds in only one clone should be reported as clone-specific.
Still open
  1. What fraction of a published MLL-N peak set in SEM is actually fusion? The overlap of anti-HA CUT&RUN with MLL-N ChIP in the same cells would give the first direct number, and every occupancy claim in this field depends on it.
  2. Is a C-terminal tag on the AFF1 portion neutral for AFF4, ENL and P-TEFb recruitment? This must be tested on the tagged line by co-IP before any mechanistic conclusion is drawn from it.
Every source behind this entry (5)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The only published MLL-fusion degron used a retrovirally overexpressed HA-FKBP12(F36V)-tagged MLL::AF9, only the C-terminally tagged construct reproduced the untagged construct's leukaemogenesis kinetics, anti-HA ChIP-seq and anti-HA Western were performed on it, and the authors state 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 in murine and human myeloid systems; the tag-position result may not transfer to KMT2A::AFF1, whose C-terminus is the AFF1 region carrying the AFF4/ENL/P-TEFb contacts.

    Evidence that argues against this (1)
    • If a C-terminal tag on the AFF1 portion disrupts P-TEFb recruitment, the tagged line under-reports the elongation arm of the mechanism - the arm that most distinguishes KMT2A::AFF1 from a menin-interface-only model. An internal tag site, or a small tag such as a 3xFLAG rather than a full FKBP12 domain, is the fallback. Yokoyama A et al. 2010
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The canonical AEP complex inventory that fusion composition claims rest on was purified with an anti-AF4 monoclonal from K562, a line carrying no KMT2A rearrangement, and the fusion hybrid-complex claim rests on shared-epitope reciprocal co-IP in leukaemia lines.

    PMID 20153263DOI
    What this evidence cannot show

    Read verbatim from the PMC body (PMC2824033 via efetch db=pmc). The paper states its systems plainly; the criticism is of downstream citation, not of the work.

    Evidence that argues against this (1)
    • The tagged-overexpression IPs in the same paper are attributable and support the hybrid-complex model, so an anti-HA AP-MS on a knock-in line may simply confirm what is already believed. The value would then be the stoichiometry, which no existing experiment gives. Olsen SN et al. 2022
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The only endogenous immunofluorescence of a KMT2A fusion used anti-AF4 polyclonals and found similar punctate nuclear staining in t(4;11) and non-t(4;11) cells, with the fusion present in lesser quantity than wild-type AF4.

    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. 1998 confocal resolution, though the audit point does not depend on resolution.

    Evidence that argues against this (1)
    • If the fusion really does go where wild-type AF4 goes, a tagged line will show the same picture and the localisation class will have been a false alarm. The experiment is designed so that outcome is informative rather than a failure. Olsen SN et al. 2022
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Published KMT2A::AFF1 occupancy maps are defined from MLL(N) ChIP-seq, which binds wild-type MLL as well as the fusion - 4,427 sites in SEM on that basis.

    PMID 28076791DOI
    What this evidence cannot show

    Read verbatim from the Europe PMC full text XML for PMC5263239. The paper supplements MLL(N) with wild-type MLL profiling and siRNA depletion, so its attribution is better than the antibody alone; the point is that no direct measurement of the fusion fraction of an MLL-N peak exists.

    Evidence that argues against this (1)
    • If wild-type MLL occupies a largely disjoint set of loci in SEM, the MLL-N map is nearly pure fusion and the correction is small. The overlap has never been measured in a rearranged line with a fusion-specific reagent, which is exactly what the tagged line would provide. Yokoyama A et al. 2010
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    SEM, RS4;11, THP-1, NOMO-1 and MOLM-13 could not be synchronised with mimosine, double thymidine or nocodazole, and centrifugal elutriation was used instead.

    PMID 17908926DOI
    What this evidence cannot show

    Read verbatim from the PMC HTML page for PMC1993870 after the XML routes failed. A 2007 observation from one laboratory; modern alternatives such as FUCCI reporters have not been tried in these lines as far as this beat found.

    Evidence that argues against this (1)
    • A knock-in line could carry a cell-cycle reporter alongside the tag, sidestepping the synchronisation problem entirely - which would make the cell-cycle arm easier in the engineered line than in the parent. Olsen SN et al. 2022
Connected to