Experiment: induce CyP33 in a KMT2A-rearranged cell and ask, allele by allele, whether the brake still works at HOXA9
also called CyP33 induction CUT&RUN · allele-resolved restraint test
Turn up the enzyme that normally makes KMT2A switch genes off, in a leukemia cell, and watch separately what happens to the broken copy of the protein and the healthy one.
The same thing, for a scientist
A doxycycline-inducible PPIE/CyP33 construct in a KMT2A-rearranged line that retains a wild-type allele (e.g. SEM for KMT2A::AFF1, MV4-11 or MOLM-13 for KMT2A::MLLT3), read out by allele-resolved CUT&RUN. The discriminating measurement is occupancy of wild-type KMT2A versus fusion at HOXA9 and HOXC8, using an antibody or tag that distinguishes the C-terminus (wild-type only) from the N-terminus (both), plus HDAC1 CUT&RUN and HOXA9/MEIS1 expression. This is the single experiment that separates 'the brake was removed' from 'the fusion is exempt from a brake that still operates', and it is a one-line-one-vector experiment that has never been done because the field has treated the question as settled.
The design
Cells. A KMT2A-rearranged line with a documented retained wild-type allele, plus - critically - a comparator line that has lost it, if one can be identified from existing characterisation.
Perturbation. Dox-inducible PPIE/CyP33. Include a PPIase-dead point mutant as the key control, because the published mechanism is isomerase-dependent, and an RRM mutant that cannot bind PHD3.
Readouts, in order of discriminating power:
- Allele-resolved CUT&RUN. An N-terminal-directed antibody sees both fusion and wild-type; a C-terminal-directed antibody sees only wild-type. The difference is the fusion. Better still, tag the endogenous wild-type allele's C-terminus so the two can be pulled separately.
- HDAC1 CUT&RUN at HOXA9. Prediction P3: HDAC1 arrives at the locus on CyP33 induction even though the fusion is unaffected.
- HOXA9, HOXC8, MEIS1 expression.
- H3K4me3 and H3 acetylation at the same loci.
The discriminating result. Repression with asymmetric occupancy change - wild-type KMT2A altered, fusion unchanged - supports exemption. No repression at all supports removal. Repression with symmetric change would mean the fusion is somehow brakeable in cis after all, which nothing predicts and would be the most interesting outcome of the three.
Why this is cheap and why nobody has done it
Every reagent exists. CyP33 constructs and the PPIase-dead mutant are published; CUT&RUN in KMT2A-rearranged lines is routine in this field; the antibodies distinguishing MLL-N from MLL-C are the same ones used to detect fusion proteins diagnostically.
The reason it has not been done is framing, not difficulty. Once 'the fusion lost the switch' became the standard sentence, there was no obvious reason to ask what the switch was doing on the other allele. The graph's own entries, including its predecessors in this beat, repeat the same framing.
Design flaw found in the cycle-31 counter-evidence pass, and how to fix it
The flaw. This design's discriminating rule was: repression at HOXA9 with asymmetric occupancy change - wild-type altered, fusion unchanged - supports exemption; symmetric change 'is what nothing predicts'. That rule is no longer safe. Something does predict symmetric change. Stabilising wild-type MLL1 raises its chromatin occupancy and displaces MLL-AFF1, AFF4 and ELL2 from the fusion's own loci by competition 4. So a perturbation acting only on the wild-type molecule can move fusion occupancy without ever touching the fusion, and under this design that result would be read as 'the fusion is brakeable in cis', which it would not be.
The fix, which is cheap. Add wild-type KMT2A abundance as a measured covariate, not just occupancy - Western blot and, better, a degron or dox titration of wild-type KMT2A run as its own arm. Then:
- CyP33 induction that represses HOXA9 without changing total wild-type KMT2A abundance or its bulk occupancy, while local character changes (HDAC1 arrives, H3ac falls), is exemption.
- Repression that tracks a rise in wild-type KMT2A occupancy across the genome is competition, i.e. the Zhao mechanism, and says nothing about brakes.
- The two have different genome-wide signatures - a locus-specific effector arrival versus a global occupancy shift - so a single CUT&RUN experiment can separate them if the global normalisation is done with spike-in rather than to total signal. Spike-in normalisation should be specified; without it, a global occupancy shift is invisible by construction.
The arm that may not exist. The design asks for 'a KMT2A-rearranged line that has lost the wild-type allele' as the cleanest internal control. Thiel et al. found the wild-type Mll allele is genetically required for MLL-AF9 leukemogenesis and maintenance 3. That comparator may therefore be unobtainable, and the open question about which lines have lost it should be read as possibly having the answer 'none, because such cells do not survive'. A dosage arm - partial knockdown of the wild-type allele - substitutes for it and is available immediately.
| design | Dox-inducible PPIE/CyP33 (wild-type, PPIase-dead, RRM-mutant) in a KMT2A-rearranged line retaining a wild-type KMT2A allele, with a wild-type-allele-null comparator where available |
|---|---|
| readout | Allele-resolved CUT&RUN for KMT2A N-terminus (fusion + wild-type) versus C-terminus (wild-type only) at HOXA9/HOXC8; HDAC1 CUT&RUN; HOXA9/HOXC8/MEIS1 expression; H3K4me3 and H3ac |
| controls | PPIase-dead CyP33 (mechanism control), RRM mutant (binding control), empty-vector dox, a non-KMT2A-rearranged leukemia line, and endogenous C-terminal tagging of the wild-type allele to make the allele split unambiguous ADDED cycle 31: spike-in normalised CUT&RUN (without it a global occupancy shift cannot be distinguished from a local one); total wild-type KMT2A protein abundance measured in every arm; and a wild-type-allele knockdown DOSAGE arm in place of the wild-type-null comparator line, which may not exist. |
| cost tier | low |
| timeline | 3-4 months |
discriminating outcomes
{
"supports_exemption": "repression at HOXA9 with asymmetric occupancy change (wild-type altered, fusion unchanged)",
"supports_removal": "no repression despite wild-type allele present and occupied",
"unexpected": "symmetric change, implying the fusion is brakeable in cis",
"confounder_added_cycle_31": "A wild-type-only perturbation CAN move fusion occupancy, by competition for chromatin (PMID 30573454). 'Symmetric change' therefore no longer implies the fusion is brakeable in cis, and the design must measure wild-type KMT2A abundance and use spike-in normalisation to tell repression-in-trans from displacement-by-competition."
}- Overexpressed CyP33 is a general prolyl isomerase and a spliceosome-associated protein; induction at non-physiological levels will have effects far beyond KMT2A, so the PPIase-dead and RRM-mutant controls carry the whole interpretation.
- A negative result would be ambiguous if the chosen line's wild-type allele is expressed at low level, so allele expression must be measured before the experiment rather than after.
- Which KMT2A-rearranged lines have lost the wild-type allele - and do any exist? Thiel et al. show the wild-type Mll allele is required for MLL-AF9 leukemogenesis, so the honest prior is that the comparator may be unobtainable rather than merely unidentified.
- Can endogenous C-terminal tagging of KMT2A be done in these lines without perturbing SET function, given that FYRC and SET are adjacent to any C-terminal tag site - and given that the FYRN/FYRC clamp is what holds the cleaved fragments together?
- Does CyP33 induction change wild-type KMT2A ABUNDANCE as well as its occupancy? If it does, the experiment is measuring two things at once and the exemption readout is confounded from the start.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The MLL1-CyP33 interaction is required for repression of HOXA9 and HOXC8 in vivo, and CyP33's PPIase activity acting on the PHD3-Bromo linker is what licenses PHD3 to bind the CyP33 RRM - making a PPIase-dead mutant the correct control for this design.
Pro isomerization in MLL1 PHD3-bromo cassette connects H3K4me readout to CyP33 and HDAC-mediated repression · Wang Z, Song J, Milne TA, Wang GG, Li H, Allis CD, Patel DJ · Cell 141(7):1183-1194 2010What this evidence cannot show
ABSTRACT ONLY. The repression experiments in that work were not performed in a fusion-driven KMT2A-rearranged cell, which is the gap this experiment addresses.
Evidence that argues against this (1)
- The requirement was demonstrated for repression of HOXA9 and HOXC8, two loci. A CUT&RUN experiment reads genome-wide occupancy, and whether the CyP33 axis governs MLL1 across its targets or only at these is not established - the design should say which it expects. Cell 2010
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MLL1 fusion proteins lack the PHD fingers and require prebinding of a wild-type MLL1 complex for stable HoxA9 association, establishing that wild-type MLL1 is present at the loci the fusion occupies.
Multiple interactions recruit MLL1 and MLL1 fusion proteins to the HOXA9 locus in leukemogenesis · Milne TA, Kim J, Wang GG, Stadler SC, Basrur V, Whitcomb SJ, Wang Z, Ruthenburg AJ, Elenitoba-Johnson KSJ, Roeder RG, Allis CD · Molecular Cell 38(6):853-863 2010What this evidence cannot show
ABSTRACT ONLY. The premise of the experiment - that wild-type MLL1 is at the locus - rests on this single study's recruitment model and would itself be tested by readout 1.
Evidence that argues against this (1)
- This experiment's discriminating rule assumed that a perturbation acting only on wild-type KMT2A cannot move fusion occupancy, so that a symmetric occupancy change would be unexplained. That assumption is false: stabilising wild-type MLL1 raises its chromatin occupancy and displaces MLL-AFF1, AFF4 and ELL2 from the fusion's own loci by competition. A symmetric change is therefore fully explained without the fusion being brakeable in cis, and the design needs wild-type KMT2A abundance measured and spike-in normalisation before its readout means anything. Zhao Z et al. 2019
- E3strong 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 MLL-AF9-transformed cells, with menin recruiting both wild-type MLL and the fusion to HOX loci.
MLL-AF9-induced leukemogenesis requires coexpression of the wild-type Mll allele · Thiel AT, Blessington P, Zou T, Feather D, Wu X, Yan J, Zhang H, Liu Z, Ernst P, Koretzky GA, Hua X · Cancer Cell 17(2):148-159 2010What this evidence cannot show
ABSTRACT ONLY; mouse genetics in one fusion. This strengthens the experiment's premise and weakens one of its arms: if the wild-type allele is genetically required, the 'wild-type-allele-null comparator line' this design asks for may not exist, and the design must not be written as though finding one is a matter of searching harder.
Evidence that argues against this (1)
- If the leukaemia requires the wild-type allele, an allele-resolved CUT&RUN cannot treat the wild-type signal as a clean internal control: perturbing CyP33 may change the wild-type allele's contribution to the fusion's own stability and occupancy, so the two readouts are coupled rather than independent. Cancer Cell 2010
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A perturbation that acts only on wild-type MLL1 - blocking its taspase1 cleavage, by taspase1 knockout or CKII inhibition - raises wild-type MLL1 chromatin occupancy and thereby displaces MLL-AFF1, AFF4 and ELL2 from chromatin in SEM cells, lowering target-gene mRNA, with survival benefit in an MLL-AF9 mouse.
Regulation of MLL/COMPASS stability through its proteolytic cleavage by taspase1 as a possible approach for clinical therapy of leukemia · Zhao Z, Wang L, Volk AG, Birch NW, Stoltz KL, Bartom ET, Marshall SA, Rendleman EJ, Nestler CM, Shilati J, Schiltz GE, Crispino JD, Shilatifard A · Genes & Development 33(1-2):61-74 2019What this evidence cannot show
FULL TEXT read via PMC6317322. Cited here because it changes what a 'fusion occupancy unchanged' result would mean in this experiment. The design's discriminating logic assumed that a wild-type-only perturbation cannot move the fusion. It can - by competition for chromatin - so the readout has to distinguish repression-in-trans from displacement-by-competition, and the current design cannot.
Evidence that argues against this (1)
- Neither arm of this perturbation is specific to wild-type MLL1. Taspase1 cleaves substrates besides MLL, and CKII inhibition by CX-4945 reaches a very large substrate set, so 'acts only on wild-type MLL1' is an assumption the experiment inherits rather than one it tests. Genes & Development 2019
the cellular experiment that would bear on which model holds
[requires] the competition confound means this design needs the dosage curve to interpret a symmetric occupancy change
the discriminating experiment: asymmetric occupancy change supports exemption, no repression supports removal
inducible CyP33 with PPIase-dead and RRM-mutant controls
HOXA9 and HOXC8 are the readout loci