Settle the WRAD question: tagged co-IP plus degron-timed proximity across three fusion classes
also called the experiment nobody has run · WRAD versus the fusion, decisively
One set of experiments answers a twenty-two-year-old question. Tag each kind of KMT2A fusion protein, pull it down, and look for the four WRAD proteins - then repeat with the normal KMT2A switched off, so a positive result cannot be explained away as WRAD hitching a ride.
The same thing, for a scientist
Three fusion classes, two assays, one decisive control. **Classes:** (i) direct 5' fusion KMT2A::AFF1 in SEM or RS4;11; (ii) reciprocal AFF1::KMT2A in the same cell, tagged on the der(4) allele - the only construct carrying AFF1 N-terminal sequence upstream of the KMT2A portion, which is what makes it taggable and distinguishable from wild-type MLL-C; (iii) a menin-blind, SET-retaining comparator, ideally NUP98::KMT2A expressed in an isogenic background. **Assays:** stringent anti-tag co-immunoprecipitation blotted for WDR5, RBBP5, ASH2L, DPY30 and menin, and TurboID from the same tagged alleles. **Control that decides the interpretation:** an auxin- or dTAG-inducible degron on the endogenous wild-type KMT2A allele, with proximity labelling performed in a window short enough that the cell is still viable. Under `hyp-wrad-reaches-kmt2a-fusions-only-as-a-second-shell-passenger`, WRAD appears in the proximity map of the direct fusion and vanishes when wild-type KMT2A is degraded; under a direct-contact model it survives; under a no-route model it never appears. No published experiment has run any arm of this.
Design
| Arm | Construct | Assay | What a positive means |
|---|---|---|---|
| A | endogenously tagged KMT2A::AFF1, SEM | stringent anti-tag co-IP, blot WDR5/RBBP5/ASH2L/DPY30/menin | a direct contact exists and the 2004 inference is wrong |
| B | arm A plus wild-type KMT2A degron | TurboID, 30-min labelling, degron on for 4 h | WRAD present with degron off and absent with it on = second-shell passenger |
| C | endogenously tagged AFF1::KMT2A (der(4)) in a reciprocal-positive line | co-IP and TurboID | the expected positive; its magnitude calibrates arms A and B |
| D | NUP98::KMT2A in an isogenic myeloid background | co-IP and TurboID | menin-blind, SET-retaining control |
| E | parental untagged and a KMT2A-wild-type leukaemia line | both assays | the specificity floor |
Readouts: blot intensity normalised to bait; TMT-labelled TurboID with a bait-normalised enrichment score; CUT&RUN for WDR5 and RBBP5 at HOXA9, MEIS1 and the RUNX1-dependent loci (SPI1, MEF2D, JUNB) in each arm, before and after degron.
Decision rule, declared in advance. Arm A positive for any subunit at >3-fold over arm E, reproduced in two independent clones, overturns the founding inference and reopens that subunit. Arm A negative and arm B positive-then-collapsing supports the second-shell model and closes the direct question. Arm B positive and degron-insensitive means a third route exists that nobody has proposed, which is the most informative outcome available.
Why the reciprocal arm is not optional
Arm C is the arm that makes the rest interpretable. It supplies a genuine positive control for a KMT2A-fusion-to-WRAD association within the same cell, the same lysis conditions and the same antibodies. Without it, a negative in arm A is uninterpretable: it could mean no association, or it could mean the assay does not detect WRAD on a KMT2A-derived protein under those conditions. Every published attempt to reason about this question has lacked such a control, because nobody was thinking about the der(4) product.
It also requires resolving something this beat could not: whether the commonly used t(4;11) lines are reciprocal-positive. A single RT-PCR across SEM, RS4;11 and MV4-11 precedes everything else here.
Feasibility
All reagents are commercial or published. Endogenous tagging in SEM and RS4;11 is routine; a CRISPR-engineered t(4;11) model in primary human HSPCs already exists if a cleaner background is wanted 3. dTAG and auxin degron systems for endogenous alleles are standard. The wild-type KMT2A degron is the only risky element, because the wild-type allele is required for maintenance 2 and the viability window may be short - which is why the design specifies a 30-minute labelling pulse and a 4-hour degron.
| design | Endogenous tagging of three KMT2A fusion classes, stringent anti-tag co-IP and TurboID, with an inducible degron on the wild-type KMT2A allele as the interpretive control |
|---|---|
| readout | Immunoblot for WDR5/RBBP5/ASH2L/DPY30/menin normalised to bait; TMT-quantified TurboID enrichment; WDR5 and RBBP5 CUT&RUN at HOXA9, MEIS1, SPI1, MEF2D and JUNB |
| controls | Parental untagged line; a KMT2A-wild-type leukaemia line; menin as the internal positive for the direct fusion; the tagged der(4) product as the internal positive for a WRAD-competent KMT2A fusion |
| prerequisite | RT-PCR for AFF1::KMT2A transcript in SEM, RS4;11 and MV4-11 to establish reciprocal status - this beat could not find it stated |
| decision rule | Arm A >3-fold over parental in two independent clones overturns the founding inference; arm A negative with a degron-collapsing arm B supports the second-shell model; arm B degron-insensitive indicates an unproposed third route |
| cost tier | med |
| timeline | 9-12 months including clone generation |
- A positive in arm A would reopen WDR5, RBBP5 and ASH2L as candidate fusion interactors, and all three are pan-essential or near-pan-essential, so the reopening would deliver targets with very poor therapeutic windows.
- The degron arm removes wild-type KMT2A, which is itself proposed as a therapeutic stabilisation target; a result showing wild-type KMT2A delivers WRAD to fusion loci would argue against stabilising it, putting two of this graph's strategies in direct conflict.
- If the answer is 'no association by any route', the experiment closes a question without opening a target, which is the correct outcome and the hardest one to fund.
- Is there a viability window in which the wild-type KMT2A allele can be degraded in a t(4;11) line long enough to label but short enough to survive? A titration precedes the whole design.
- Can the der(4) product be endogenously tagged without destabilising it? Taspase1-dependent cleavage means a C-terminal tag would end up on the released MLL-C fragment and an N-terminal tag on the AFF1 portion, and only the latter is diagnostic.
- Should the comparator in arm D be NUP98::KMT2A or an engineered SET-restored KMT2A::AFF1? The latter is not a natural species but is a cleaner isogenic control.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
No published experiment has assayed any WRAD subunit for association with any KMT2A fusion protein: the founding paper blotted its fusion immunoprecipitates for menin, HCF-1, MLL-N, MLL-C and ENL only, and alias-expanded PubMed queries pairing each WRAD subunit with the named fusions return 26, 6, 3 and 2 records with no association experiment among them.
Leukemia proto-oncoprotein MLL forms a SET1-like histone methyltransferase complex with menin to regulate Hox gene expression. · Yokoyama A, Wang Z, Wysocka J, et al. · Mol Cell Biol 2004What this evidence cannot show
The absence claim rests on titles and abstracts plus the founding paper's figures; a co-IP in a supplementary figure of a paper about something else would be invisible. This beat's own counted absence for proximity labelling was falsified by a paper already in this graph, so the bound should be read conservatively.
Evidence that argues against this (1)
- Proximity labelling of a NUP98 oncofusion does recover WDR5 and RBBP5, and it is invisible to every WRAD-keyed query because its abstract names the methyltransferases at complex level - proof that this class of absence claim under-counts. Ahn JH et al. 2025
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The wild-type Mll allele is required for MLL-AF9-induced leukemogenesis and for maintenance of transformed cells, which both motivates the degron arm and bounds how long it can run.
MLL-AF9-induced leukemogenesis requires coexpression of the wild-type Mll allele. · Thiel AT, Blessington P, Zou T, et al. · Cancer Cell 2010What this evidence cannot show
Mouse MLL-AF9 genetics; the tolerable degron window in a human t(4;11) line has not been measured and would need a viability titration first.
Evidence that argues against this (1)
- Stabilising rather than degrading wild-type MLL1 displaces MLL chimeras from chromatin and delays leukaemic progression, so wild-type KMT2A has opposing roles and the degron arm may produce a mixture of effects rather than a clean subtraction. Liang K et al. 2017
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
CRISPR/Cas9-engineered t(4;11) translocations can be generated in primary human haematopoietic stem and precursor cells with both reciprocal products concomitantly present, supplying an isogenic background for the tagging arms.
CRISPR/Cas9-mediated t(4;11) translocation in human hematopoietic stem/precursor cells demonstrates plasticity to differentiate into either the myeloid or lymphoid lineage. · Benz T, Larghero P, Meyer C, et al. · Leukemia 2026What this evidence cannot show
A recent model; primary HSPC cultures are short-lived and lower-throughput than cell lines, and whether an endogenous tag can be placed on the der(4) product without perturbing its biology is untested.
Evidence that argues against this (1)
- An engineered translocation does not solve the attribution problem by itself, because the resulting cells contain the same three sources of chromatin-bound WRAD; the tag on the der(4) allele, not the engineering, is what resolves it. Wilhelm A et al. 2021
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Proximity labelling output is an abundance-weighted, compartment-structured census rather than a membership test, so a proximity positive for WRAD around a KMT2A fusion requires the degron control to mean anything.
The phenylalanine-and-glycine repeats of NUP98 oncofusions form condensates that selectively partition transcriptional coactivators. · Ahn JH, Guo Y, Lyons H, et al. · Mol Cell 2025What this evidence cannot show
Cited here for the general property demonstrated in a fusion system, not for a KMT2A result. This graph's concept-proximity-labelling-is-not-co-purification states the four rules in full.
Evidence that argues against this (1)
- Proximity labelling is uniquely good at transient and chromatin-bound contacts that affinity purification misses, so treating a proximity positive as uninformative is the opposite error and the degron control exists precisely to avoid having to choose. Terlecki-Zaniewicz S et al. 2021
The degron arm is the experiment designed to decide this hypothesis in one pass.
Arm A tags the direct fusion in SEM or RS4;11 and blots for all four WRAD subunits with menin as the internal positive.
Arm C tags the der(4) product, which is the only available positive control for a WRAD-competent KMT2A fusion in the same cell.
The design exists because an anti-MLL-C antibody cannot assign the species it detects; endogenous tagging of the der(4) allele is the fix.
Arm D uses NUP98::KMT2A as the menin-blind, SET-retaining comparator.
- ← supportsNo proximity, degron or tiling-CRISPR study has ever been run on WRAD in KMT2A-fusion leukaemia
The absence names exactly the experiment this entry designs.