Competing model 1: the reciprocal product starts the leukaemia and then leaves
also called hit-and-run model · AF4-MLL as initiator · 48 hours is enough
One model says the mirror-image protein is the match, not the fire. Switching it on for only two days permanently changed which genes a cell could reach - and then it was no longer needed.
The same thing, for a scientist
The initiator model has two published legs. First, AF4.MLL expressed in murine Lin-/Sca1+ progenitors induced pro-B ALL, B/T biphenotypic acute leukaemia or mixed-lineage leukaemia, while MLL.AF4 alone caused no disease over 13 months [E1] - the only in vivo experiment in which either t(4;11) product alone produces leukaemia. Second, in a cell-culture model with MLL-AF4 constitutive and AF4-MLL inducible for 48 hours only, the transient reciprocal expression produced roughly eight times more differentially expressed protein-coding genes plus hundreds of pseudogenes and non-annotated genes, with transcriptional and chromatin-accessibility changes still detectable at day 28 after the reciprocal protein was gone [E2]. The authors call this hit-and-run: necessary to set up pre-leukaemic conditions, dispensable afterwards. The model explains the prognostic paradox elegantly - a protein that is only needed at the start can be lost later without the leukaemia caring - and it has one severe weakness that the paper states in its own methods: the experiment is in HEK293T cells [E2].
Why HEK293T is not a small caveat here
The claim is about chromatin state in a haematopoietic progenitor being permanently reconfigured by a transient transcription factor. HEK293T is an adenovirus-transformed embryonic kidney line with its own aberrant chromatin, no B-lymphoid programme and no capacity to become leukaemic. The finding that transient expression leaves a durable transcriptional trace is interesting in any cell; the finding that it sets up pre-leukaemic conditions is not tested there at all. The deposited sequencing confirms the system: the study's GEO series carry the MACE-Seq and ATAC-Seq data from that model 4.
The human experiment that fails the model
AF4-MLL expressed in human cord-blood CD34+ HSPCs enhanced long-term multilineage engraftment in primary immunodeficient recipients but conferred no serial self-renewal and initiated no leukaemia within 36 weeks 3. That is the same protein in the right species and a plausible cell of origin, and it does not initiate. So the initiator model currently rests on a murine experiment in which about one third of transplanted mice developed disease 1, plus a chromatin experiment in kidney cells 2, against a negative in human cord blood 3.
What would settle it
A CRISPR-engineered balanced t(4;11) in human HSPCs, with the der(4) allele subsequently deleted or silenced at a defined timepoint. The engineering half already exists - balanced t(4;11) translocations have been generated in primary human haematopoietic stem and precursor cells with both reciprocal products concomitantly present 5 - so the missing step is the conditional removal, not the model.
| kind | initiation model |
|---|---|
| claim | the reciprocal product is required to establish pre-leukaemic chromatin and dispensable thereafter |
| induction window hours | 48 |
| durability days | 28 |
| system provenance warning | The keystone chromatin experiment is in HEK293T, not a haematopoietic cell; the keystone leukaemia experiment is murine. No load-bearing experiment for this model is in a human KMT2A::AFF1 leukaemia. |
supporting systems
[ "murine Lin-/Sca1+ progenitors", "HEK293T transient induction" ]
opposing systems
[ "human cord-blood CD34+ xenograft", "siRNA in human t(4;11) lines" ]
geo
[ "GSE178569", "GSE178567", "GSE178568" ]
- If the reciprocal product acts only at initiation, then no drug against it can treat established disease, and the entire therapeutic interest in Taspase1 and FYRN/FYRC evaporates into a prevention indication with no trial design.
- A hit-and-run initiator is also undetectable at diagnosis, so a patient could have been initiated by a protein their blasts no longer make - which is one reading of the 40% allele-loss figure.
- The model's strongest evidence is a murine sufficiency experiment; the human experiment with the same construct is negative.
- Does transient AF4-MLL expression reconfigure chromatin in a human haematopoietic progenitor, as opposed to in HEK293T? The identical experiment in CD34+ cells has not been done.
- In the CRISPR-engineered human t(4;11) system, does deleting the der(4) allele after transformation change anything? That is the direct test of hit-and-run and the platform exists.
- Why did leukaemia arise in only about a third of mice transplanted with AF4.MLL? A partially penetrant initiator implies a second event that has never been sought.
- Is the durable day-28 signature a chromatin memory or simply the outgrowth of a subclone selected during the 48-hour induction? The paper's own language about clonal evolution suggests the second, which is a different mechanism with the same readout.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
AF4.MLL expressed in murine Lin-/Sca1+ progenitors induced pro-B ALL, B/T biphenotypic acute leukaemia or mixed-lineage leukaemia, whereas MLL.AF4 alone caused no disease over a 13-month observation period.
The AF4.MLL fusion protein is capable of inducing ALL in mice without requirement of MLL.AF4. · Bursen A, Schwabe K, Rüster B, et al. · Blood 2010What this evidence cannot show
Murine progenitors, retroviral over-expression, and leukaemia in only about one third of transplanted animals as the same group later described it; the construct is one breakpoint version; a positive result for sufficiency in mouse says nothing about necessity in human disease.
Evidence that argues against this (1)
- The same protein in human cord-blood CD34+ cells enhanced engraftment but conferred no serial self-renewal and initiated no leukaemia within 36 weeks. Prieto C et al. 2017
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In a model with MLL-AF4 constitutively expressed and AF4-MLL induced for only 48 hours, transient reciprocal expression produced a signature roughly eight times larger in protein-coding genes plus hundreds of pseudogenes and non-annotated genes, with MACE-Seq and ATAC-Seq changes still present at day 28, which the authors interpret as a hit-and-run mechanism setting up pre-leukaemic conditions but dispensable for later progression.
The role of reciprocal fusions in MLL-r acute leukemia: studying the chromosomal translocation t(4;11). · Wilhelm A, Marschalek R · Oncogene 2021What this evidence cannot show
The experiments are in HEK293T cells - non-haematopoietic, non-leukaemic, adenovirus-transformed - with plasmid-borne transgenes and reporter-based tracking of plasmid out-segregation rather than a genomic knockout. Three biological replicates. Nothing in the design tests leukaemogenesis.
Evidence that argues against this (1)
- A fusion pair expressed in immature haematopoietic cells has been shown to fail to recapitulate the transcriptional changes seen in human AML, so a transcriptional signature from a model system is weak evidence about what the fusion does in a patient. Rapin N et al. 2014
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The AF4-MLL fusion expressed in human cord-blood CD34+ cells enhanced long-term multilineage haematopoietic engraftment in primary immunodeficient recipients but conferred no serial self-renewal and initiated no leukaemia within 36 weeks.
The AF4-MLL fusion transiently augments multilineage hematopoietic engraftment but is not sufficient to initiate leukemia in cord blood CD34+ cells. · Prieto C, Marschalek R, Kühn A, et al. · Oncotarget 2017What this evidence cannot show
A negative result over a 36-week window in a xenograft system whose latency for infant ALL is not established; lentiviral expression levels may not match the endogenous allele; absence of leukaemia in this window does not exclude a longer latency or a requirement for the direct fusion in cis.
Evidence that argues against this (1)
- The murine experiment with the same fusion did produce leukaemia, so the discrepancy could be species, cell type, expression level or assay duration rather than a real negative. Bursen A et al. 2010
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The transcriptomic and chromatin-accessibility data behind the hit-and-run model are deposited as GEO series GSE178569 (MACE-Seq) and GSE178567 and GSE178568 (ATAC-Seq).
GEO series GSE178569 (MACE-Seq) and GSE178567/GSE178568 (ATAC-Seq), deposited for the t(4;11) hit-and-run study · Wilhelm A, Marschalek R · NCBI GEO 2021What this evidence cannot show
The deposition is recorded in the paper's methods. This entry cites it to make the experimental system auditable per sample rather than to support a biological claim; per-sample SOFT records were not parsed for this beat.
Evidence that argues against this (1)
- Deposition confirms what was sequenced, not what it means; the same paper's stated cell system is HEK293T, so the depositions cannot supply a haematopoietic context the experiment did not have. Wilhelm A et al. 2021
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
CRISPR/Cas9-mediated t(4;11) translocations can be generated in primary human haematopoietic stem and precursor cells with both reciprocal fusion proteins concomitantly present, and the resulting cells show plasticity to differentiate into either myeloid or lymphoid lineage.
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
Establishes the engineering platform; it does not itself perform conditional removal of one allele, which is the experiment the initiator model needs.
Evidence that argues against this (1)
- A genome-editing model of a related translocation, t(9;11), produced leukaemia in mice and was interpreted as showing the genetic sufficiency of the translocation as a whole, without attributing the effect to either product - so an engineered balanced translocation does not by itself separate the two. Schneidawind C et al. 2018
Model in which this protein is required transiently to establish pre-leukaemic chromatin.
Proposed initiating event for t(4;11) infant ALL; demonstrated in mouse, not in human cells.
Supplies the temporal half of the chromatin-opening account: 48 hours of expression, 28 days of effect.
The model requires an early progenitor target; the human cord-blood test of it was negative.
A protein needed only at initiation is free to be lost later; the two models are designed to fit together.
Maintenance and initiation give opposite answers for the same protein; this is the maintenance side.
A second, independent system in which reciprocal products show transforming activity.