PROPOSED: swap the partner, hold everything else - KMT2A::MLLT3 versus KMT2A::AFF1 in LIN28B-null human fetal liver CD34+ cells
also called the partner-swap discriminator · is the LIN28B requirement AFF1-specific
The fetal gene LIN28B is needed for one MLL fusion to cause leukaemia in human fetal cells, but in mice a different MLL fusion is actually held back by the same gene. Putting both fusions into the same human fetal cells, with and without LIN28B, would tell us whether the difference is the fusion partner or the species and lineage.
The same thing, for a scientist
The LIN28B literature in KMT2A-rearranged leukaemia is split by sign, and **five variables move together**: species, lineage, partner, lesion type (endogenous translocation vs transgene) and perturbation direction. The human lymphoid AFF1 result says required-for-transformation [E1]; the mouse myeloid MLLT1/MLLT3 results say tumour-suppressive [E2][E3]. **This experiment holds four of the five fixed and varies the partner.** Same human fetal liver CD34+ cells, same CRISPR endogenous-translocation approach, same LIN28B-null background, same readout - only KMT2A::MLLT3 in place of KMT2A::AFF1. It is the discriminating experiment for `hyp-fetal-state-is-permissive-for-lymphoid-and-protective-for-myeloid-kmt2a-transformation`, and it also answers an open question already standing in `gene-lin28b` and `cell-state-fetal-program-retention`.
Design and what each outcome means
| arm | fusion | LIN28B | expected if the hypothesis is right |
|---|---|---|---|
| 1 | KMT2A::AFF1 | intact | transforms (published reference) |
| 2 | KMT2A::AFF1 | null | does not transform (published) |
| 3 | KMT2A::MLLT3 | intact | transforms |
| 4 | KMT2A::MLLT3 | null | still transforms, possibly better |
Readout: transformation frequency by the source study's own operational definition, plus immunophenotype of the product, because KMT2A::MLLT3 in a fetal human progenitor may produce myeloid or mixed output and the lineage of the product is itself part of the answer.
Outcomes:
- Arms 2 and 4 both blocked -> the LIN28B requirement is partner-independent and is a property of the fetal progenitor's transformability. That kills the lineage hypothesis and strongly supports the competence reading in
exp-lin28b-null-fetal-progenitor-second-oncogene-specificity-control. - Arm 2 blocked, arm 4 not -> the requirement is partner- or lineage-specific, and the mouse myeloid literature is reconciled rather than contradicted. This is the result the hypothesis predicts.
- Arm 4 transforms BETTER than arm 3 -> LIN28B is actively protective against a myeloid KMT2A fusion in human fetal cells too, which would be the strongest possible confirmation and would make lineage the whole story.
Caveat built into the design: partner and lineage are still confounded, because KMT2A::MLLT3 is the myeloid-skewed partner. A cleaner third arm would be a lymphoid-skewed non-AFF1 partner, but the graph holds no partner that is both clearly lymphoid and clearly distinct from AFF1 in complex composition. This limitation is stated rather than designed away.
| status | planned |
|---|---|
| design | Four arms in human fetal liver CD34+ HSPC from >=3 matched donors: KMT2A::AFF1 and KMT2A::MLLT3, each in LIN28B-null and LIN28B-intact backgrounds, using the same CRISPR-Cas9 endogenous-translocation approach for both fusions. |
| readout | Transformation frequency by the source study's operational definition, plus immunophenotype (lymphoid/myeloid/mixed) of the transformed product, plus xenograft initiation for arms that transform. |
| controls | Vector/no-induction arms; donor-matched pairs; protein-level LIN28B knockout efficiency; translocation induction efficiency measured per fusion, because CRISPR translocation efficiency differs by partner locus and an efficiency difference would masquerade as a biological one. |
| cost tier | high |
| timeline | 9-18 months |
| decision rule | Requirement is partner-specific if KMT2A::MLLT3 transformation frequency falls by less than 2-fold on LIN28B loss while KMT2A::AFF1 falls by 5-fold or more. Requirement is partner-independent if both fall by 5-fold or more. |
| tests hypothesis | hyp-fetal-state-is-permissive-for-lymphoid-and-protective-for-myeloid-kmt2a-transformation |
| known confound not removed | Partner and lineage remain confounded, because KMT2A::MLLT3 is the myeloid-skewed partner. No available partner separates the two cleanly. |
| prerequisite | A published or in-house CRISPR endogenous-translocation protocol for KMT2A::MLLT3 in human fetal liver cells. This beat did not verify that one exists; if it does not, the experiment's cost rises substantially. |
- If the requirement turns out to be partner-independent, the clinical framing shifts from 'a KMT2A::AFF1 target' to 'a fetal-progenitor growth requirement', which is far less selective and much harder to justify in an infant.
- If LIN28B loss makes a myeloid KMT2A fusion transform BETTER, then a LIN28B inhibitor would be contraindicated in KMT2A-rearranged AML and in any patient whose disease might switch lineage - a real risk in this disease under lineage-directed immunotherapy.
- Does a CRISPR endogenous-translocation protocol for KMT2A::MLLT3 in human fetal liver CD34+ cells exist? This beat did not check and it is the gating question for feasibility.
- What immunophenotype does KMT2A::MLLT3 produce in a human FETAL liver progenitor? If it produces lymphoid output there, the partner/lineage confound partly dissolves and the experiment becomes much more informative.
- Is CRISPR translocation efficiency comparable between the AFF1 and MLLT3 loci in the same cells? An efficiency difference would produce a false partner effect, and this is the control most likely to be omitted.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Human fetal liver CD34+ cells fail to transform upon induction of the KMT2A::AFF1 translocation in the absence of LIN28B, in a CRISPR-Cas9 endogenous-translocation model.
The fetal specific gene LIN28B is essential for human fetal B-lymphopoiesis and initiation of KMT2A::AFF1 infant leukemia. · Ling RE, Welsh AM, Hamer L, Jackson T, Neil E, Elliott N, et al. · Blood 2026What this evidence cannot show
ABSTRACT ONLY and paywalled (Europe PMC: isOpenAccess N, inEPMC N, inPMC N, verified 2026-09-23). The operational definition of transformation, the donor number and the effect size are all unknown, which means the comparator arm of this proposed experiment cannot yet be specified exactly.
Evidence that argues against this (1)
- The comparator arm of this proposed experiment cannot be specified until the source study's operational definition of transformation, its effect size and its donor number are known - and they are unreadable because the paper is paywalled and absent from PMC. A proposed experiment whose reference arm is undefined is not yet designable. source 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Enforced LIN28B in established mouse MLL::AF9 (KMT2A::MLLT3) AML significantly extended survival with 80% of secondary recipients surviving beyond 4 months, induced partial monocytic differentiation by 6 days and severely depleted colony-forming potential in 5 biologically independent AMLs without increasing apoptosis.
LIN28B promotes differentiation of fully transformed AML cells but is dispensable for fetal leukemia suppression. · Li Y, Mendoza-Castrejon J, Patel RM, Casey EB, Denby E, Bryder D, Magee JA · Leukemia 2024What this evidence cannot show
FULL TEXT READ (PMC10912017). Mouse, transgene, supraphysiological LIN28B by the authors' own statement, and the perturbation is a GAIN whereas this proposed experiment is a LOSS. It motivates arm 4 but does not predict it: nothing in it shows that removing LIN28B helps a myeloid fusion.
Evidence that argues against this (1)
- This is a GAIN-of-function result and the proposed experiment is a LOSS-of-function one. Nothing in Li et al. shows that removing LIN28B helps a myeloid KMT2A fusion, and the companion fetal-deletion arm found no effect at all - so arm 4's most interesting predicted outcome has no supporting precedent. Leukemia 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In inducible transgenic MLL-ENL mice, enforced LIN28B suppressed AML initiation rather than accelerating it, and AML initiation efficiency peaked neonatally rather than fetally.
The efficiency of murine MLL-ENL-driven leukemia initiation changes with age and peaks during neonatal development. · Okeyo-Owuor T, Li Y, Patel RM, Yang W, Casey EB, Cluster AS, Porter SN, Bryder D, Magee JA · Blood Adv 2019What this evidence cannot show
Abstract read; full text not opened. Mouse, transgene, myeloid, MLLT1 rather than MLLT3. Transplantation stress is reported to mitigate the developmental barrier, so the age effect is fragile to handling.
Evidence that argues against this (1)
- Man et al. found MLL-AF9 in mouse FETAL LIVER cells to have a PRO-leukaemic Id1 dependency, i.e. the fetal compartment supplying something a myeloid KMT2A fusion needs. If the fetal compartment is permissive for myeloid fusions too, the partner-swap experiment may return 'both arms blocked' for reasons unrelated to lineage. Blood 2016
the discriminating experiment: partner swap with species, cells and lesion held fixed
KMT2A::MLLT3 arm in LIN28B-null human fetal liver
KMT2A::AFF1 reference arm