The WRAD degron absence survives, and the reason is worth more than the verdict: a 3.7 nM chemical degron of WDR5 with 2-hour kinetics already exists IN MV4;11, and it still cannot see the fusion
also called MS67 near-miss · Q159 WRAD re-test
This project records that nobody has used rapid protein-removal methods to ask whether a particular protein machine touches the leukemia fusion. Searching by drug name rather than by disease name turned up a compound that removes one of those proteins from the exact leukemia cell line in question within two hours - and it still does not answer the question, because the readout it uses cannot tell the fusion protein apart from the normal one.
The same thing, for a scientist
`absence-no-proximity-degron-or-crispr-study-of-wrad-in-kmt2a-fusion-leukaemia` states that no proximity, degron or tiling-CRISPR experiment has been designed to ask whether WRAD reaches a KMT2A fusion. Re-tested with compound names instead of method words, the nearest object is **MS67**, a VHL-recruiting WDR5 PROTAC with **DC50 3.7 +/- 1.4 nM and Dmax 94 +/- 1% in MV4;11** - the KMT2A::AFF1 index-fusion line - degradation apparent at **2 hours**, maximal at 4, rebounding at 48 and fully recovered by 72 hours after washout, **GI50 15 +/- 8 nM** [E1]. It has an association readout: MS67 decreases the **chromatin-bound fraction of MLL, RBBP5 and menin** and reduces H3K4me2 by ChIP-seq at HOX cluster genes and BCL2 in MV4;11 [E1]. **The absence still survives**, because the chromatin-fractionation blot uses an MLL antibody with no stated ability to distinguish the ~180 kDa wild-type C-terminal species from the fusion, so the experiment cannot attribute the loss to either. This corpus already knew the paper - it is cited in six entries - and `concept-wdr5-what-is-established-for-wild-type-kmt2a-and-what-for-the-fusion` already states the narrow form of the claim including the word *degron* [E2]. **The actionable output is not a verdict, it is a reagent**: the entry's own `feasible_now` list asks for 'TurboID from a tagged KMT2A::AFF1 with a wild-type-allele degron', and MS67 supplies an off-the-shelf, washout-reversible, 2-hour chemical degron of a WRAD subunit already characterised in that line.
This is real experimental data — atoms placed by crystallography or cryo-EM, deposited by the people who did the experiment.
I have never read one of these pictures before
A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.
Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.
Anything in yellow sticks is a bound small molecule, which is usually the drug.
What MS67 measures in MV4;11, precisely
From the full text (PMC8500670) 1:
| quantity | value | line |
|---|---|---|
| DC50 | 3.7 +/- 1.4 nM | MV4;11 |
| Dmax | 94 +/- 1% | MV4;11 |
| time to apparent degradation | 2 h | MV4;11 |
| time to maximal degradation | ~4 h | MV4;11 |
| washout rebound / full recovery | ~48 h / ~72 h | MV4;11 |
| GI50 | 15 +/- 8 nM | MV4;11 |
| GI50 | 38 +/- 1 nM | EOL-1 |
| GI50 (WIN-site inhibitor OICR-9429, same lines) | > 2500 nM | MV4;11, EOL-1 |
Sensitivity panel: MV4;11, EOL-1, MOLM13, KOPN8, RS4;11 and THP-1 are sensitive; K562, HL60 and a murine Hoxa9+Meis1 line are insensitive - a clean fusion-versus-no-fusion contrast that is itself a rarity in this literature. Ternary-complex crystal structures are deposited as 7JTO (WDR5-MS33-VCB, 1.7 A) and 7JTP (WDR5-MS67-VCB, 2.1 A).
The authors state as a limitation that the variation in MLL-r line sensitivity - MV4;11 and EOL-1 most sensitive, RS4;11 and THP-1 least - could not be explained by any genetic background they examined. That unexplained gradient is itself a fusion-partner question this project is equipped to ask: MV4;11 and RS4;11 are both KMT2A::AFF1 and sit at opposite ends of it, which argues against a partner explanation and for something else.
Why this is a near-miss and not an overturn
Three reasons, in decreasing order of how much they would matter to a reviewer:
- The readout cannot see the fusion. 'Chromatin-bound fraction of MLL complex components' is a fractionation immunoblot. Nothing in the paper states that the MLL antibody discriminates the fusion protein from the wild-type C-terminal fragment, and KMT2A's own UniProt entry models the two taspase1 products separately for exactly this reason. A decrease in chromatin-bound 'MLL' after WDR5 removal is compatible with the fusion being untouched.
- The design is therapeutic, not associative. The study's stated research goal is 'to find and characterize highly effective WDR5 degraders'. The chromatin panel is a mechanism-of-action confirmation, not an attribution experiment, and the absence entry's claim is explicitly about what has been designed to ask.
- The corpus had already decided this. PMID 34586829 is cited in six node files including two in the same
wrad-adjudicationfolder as the absence. One of them states outright that no published degron experiment reports WDR5 with any KMT2A fusion protein 2. Filing this as a new overturn would be rediscovering a decision rather than making one - the exact failure this project has recorded as 'a gap list cannot see what was decided'.
Against all three: the absence entry's own data.degron_records annotation reads 'all WDR5 degrader chemistry and none in a KMT2A-fusion-association design'. The first half of that is not quite right - MS67 is degrader chemistry and carries a chromatin-association panel in the index-fusion line. This is a refinement of wording, not an error of fact, and the entry's headline claim is unaffected.
The experiment this unlocks, stated so it is scoreable
The absence entry lists as feasible: 'TurboID from a tagged KMT2A::AFF1 with a wild-type-allele degron'. MS67 replaces the hard half of that with a purchase.
Design. MV4;11 (KMT2A::AFF1). Treat with MS67 at 100 nM (>25x DC50) for 4 h - the paper's own time to maximal degradation - against MS67N (the VHL-non-binding diastereoisomer) and OICR-9429 at matched concentration as occupancy-only and inactive controls, all three of which the paper validates in this line. Fractionate chromatin and blot with an N-terminal KMT2A antibody, which detects the fusion and the wild-type N320 fragment, alongside a C-terminal antibody, which detects only wild-type; the fusion is the species present in the N-terminal blot and absent from the C-terminal one.
Scoreable prediction. If WRAD reaches the fusion, chromatin-bound fusion signal falls by >= 2-fold at 4 h, in the same direction and within 2-fold of the magnitude of the wild-type C-terminal species. If WRAD is a wild-type-only passenger, the C-terminal species falls >= 2-fold and the fusion species changes by < 1.5-fold. If both fall and the fusion falls less, that is the second-shell model and it is the outcome the current evidence most predicts.
What makes it cheap: no tagging, no knock-in, no degron construct, 4 hours of drug, one fractionation and two antibodies - and the washout arm at 48 h supplies a reversibility control for free.
| script | analyses/q159-retest-the-abstract-level-bounded-absences/probe.py |
|---|---|
| status | run |
| result summary | absence-no-proximity-degron-or-crispr-study-of-wrad-in-kmt2a-fusion-leukaemia SURVIVES. MS67 is a 3.7 nM, 2-hour, washout-reversible chemical degron of WDR5 validated in MV4;11 with a chromatin-association panel, but its MLL blot cannot distinguish fusion from wild-type, and this corpus had already cited the paper in six entries. |
| verdict | SURVIVES |
| reagent unlocked | MS67 (and its controls MS67N, OICR-9429) as an off-the-shelf chemical degron of a WRAD subunit in the index-fusion line |
| absence class | never-measured |
| absence class basis | The fusion-versus-wild-type attribution has not been made by anyone; the reagent to make it now exists and is commercially characterised. |
inputs
[ "PMC8500670 full text", "data/graph/nodes/wrad-adjudication/*" ]
PDB IDS
[ "7JTO", "7JTP" ]
measurements — the numbers behind the chart
[
{
"quantity": "DC50",
"subject": "MS67 degradation of WDR5 | MV4;11 (KMT2A::AFF1)",
"value": 3.7,
"units": "nM",
"n": 1,
"universe": "18 h treatment, immunoblot, concentration series; SD 1.4 nM",
"null": 260,
"note": "null is MS33, the unoptimised parent degrader, in the same line and assay"
},
{
"quantity": "Dmax",
"subject": "MS67 degradation of WDR5 | MV4;11 (KMT2A::AFF1)",
"value": 94,
"units": "percent",
"n": 1,
"universe": "18 h treatment, immunoblot; SD 1%",
"null": 71,
"note": "null is MS33 in the same line and assay"
},
{
"quantity": "GI50",
"subject": "MS67 growth inhibition | MV4;11 (KMT2A::AFF1)",
"value": 15,
"units": "nM",
"n": 1,
"universe": "in vitro proliferation; SD 8 nM",
"null": 2500,
"note": "null is OICR-9429, the WIN-site occupancy inhibitor, in the same line (>2500 nM)"
},
{
"quantity": "time_to_apparent_degradation",
"subject": "MS67 degradation of WDR5 | MV4;11 (KMT2A::AFF1)",
"value": 2,
"units": "hours",
"n": 1,
"universe": "time course, immunoblot; maximal at ~4 h",
"null": null,
"note": "washout: substantial rebound ~48 h, near-full recovery ~72 h"
}
]- If WRAD turns out to reach the fusion, MS67-like degradation still removes WDR5 from KMT2B, SETD1A, SETD1B and NSL simultaneously, so a therapeutic window argument cannot be made from a fusion-attribution result alone.
- This corpus's own DepMap reading is that WDR5 is depleted LESS in KMT2A-rearranged lines than in leukaemia lines generally (+0.191 lineage-controlled selectivity, the worst of 41 candidates scored here). A 15 nM GI50 in MV4;11 sits uneasily beside that, and the resolution is probably that MS67 kills by a MYC/ribosomal route common to many lineages rather than by a KMT2A-axis route.
- WDR5 P173L has already been selected in vitro as an occupancy-inhibitor resistance mutation. A degrader is predicted to evade it, but VHL loss and linker-independent resistance routes are the standard PROTAC liabilities and neither has been screened for here.
- Why are MV4;11 and RS4;11 - both KMT2A::AFF1 - at opposite ends of the MS67 sensitivity range? The authors state they could not explain the variation from any genetic background they examined, and a fusion-partner explanation is excluded by the two AFF1 lines disagreeing with each other. (The paper groups EOL-1 with the MLL-r lines; this project has not independently verified that line's rearrangement status and should before building on it.)
- Does an N-terminal versus C-terminal KMT2A blot under MS67 in MV4;11 separate the fusion from wild-type? Four hours of drug and two antibodies would settle a question this project has kept open across several cycles.
- MS67 degrades WDR5 out of the NSL complex as well as out of COMPASS. Is there a WIN-site-selective degrader, or a WDR5 separation-of-function allele, that would make the attribution clean?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MS67 is a VHL-recruiting WDR5 PROTAC that degrades WDR5 in MV4;11 with DC50 3.7 nM, Dmax 94%, apparent degradation at 2 hours and full washout recovery by 72 hours, and it decreases the chromatin-bound fraction of MLL, RBBP5 and menin and reduces H3K4me2 at HOX cluster genes and BCL2 in that line.
A selective WDR5 degrader inhibits acute myeloid leukemia in patient-derived mouse models. · Yu X, Li D, Kottur J, Shen Y, Kim HS, Park KS, Tsai YH, Gong W, Wang J, Suzuki K, Parker J, Herring L, Kaniskan HU, Cai L, Jain R, Liu J, Aggarwal AK, Wang GG, Jin J · Sci Transl Med 2021What this evidence cannot show
Full text read via PMC8500670 on 2026-09-24. Numeric values are quoted from the results prose; the figures themselves are not in the XML, so the DC50/Dmax point estimates and their standard deviations come from the text and not from re-reading a curve. The paper never uses the words KMT2A::AFF1 or MLL-AF4 - it says 'MLL-r' throughout - so MV4;11's identity as the index fusion is supplied by this project, not by the paper.
Evidence that argues against this (1)
- Degradation removes every WDR5 function at once, including its mutually exclusive NSL-complex binding, so a chromatin-bound MLL decrease under MS67 is not attributable to the MLL/COMPASS pool specifically. This corpus states that objection already and it is the reason a degrader result is LESS attributable to the KMT2A axis than an occupancy result, not more. Oncofusion Knowledge Graph 2026
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
This corpus already states, in the same folder as the absence entry, that no published co-immunoprecipitation, mass spectrometry, structure, degron or proximity experiment reports WDR5 with any KMT2A fusion protein.
concept-wdr5-what-is-established-for-wild-type-kmt2a-and-what-for-the-fusion (this corpus) · Oncofusion Knowledge Graph 2026What this evidence cannot show
Read in full from disk on 2026-09-24. It is a project entry, so it is a record of a prior decision and not independent evidence about the literature; its own bound is an alias-expanded PubMed query returning 26 records, which is the same class of instrument this beat set out to test.
Evidence that argues against this (1)
- That entry's 26-record bound is itself an abstract-level literature search of exactly the type under audit here, so citing it as the reason not to overturn is partly circular. The non-circular part is the readout argument: the MS67 chromatin blot cannot distinguish the fusion, which is a fact about the paper and not about any query. Sci Transl Med 2021
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
WDR5 is covered in this graph as gene-wdr5, so the near-miss does not create a missing-symbol gap; scripts/is-symbol-covered.py returns COVERED with no DECLINED ON PURPOSE verdict.
scripts/is-symbol-covered.py WDR5, oncofusion-knowledge-graph, run 2026-09-24 · this project 2026What this evidence cannot show
The script resolves symbols against a 2,928-id index and reports coverage, not quality; a COVERED verdict says an entry exists, not that it records MS67. gene-wdr5 does cite OICR-9429 but the MS67 numbers quoted here are not in it.
Evidence that argues against this (1)
- The script's own output warns that a no-match answer is not a gap, and by symmetry a COVERED answer is not a guarantee that the specific fact is recorded. MS67's DC50 and washout kinetics in MV4;11 appear in no node file in this corpus. this project 2026
- supports →No proximity, degron or tiling-CRISPR study has ever been run on WRAD in KMT2A-fusion leukaemia
SURVIVES: MS67 is a 3.7 nM chemical degron in MV4;11 whose MLL blot cannot distinguish the fusion
- related to →WDR5 — the partner that is not a partner: it binds the half of KMT2A the fusion throws away
adds MS67 DC50 3.7 nM, Dmax 94%, 2 h kinetics and 48-72 h washout in MV4;11, none of which gene-wdr5 records
independent re-test by compound name reaches the same verdict this concept already recorded
every parameter of this design comes from the MS67 characterisation recorded there