Oncofusion

start / graph / model-pluripotent-stem-cell-route-mll-af4

Explainerstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q21-index-fusion-models

The pluripotent stem-cell route has failed from both ends, and the failures are informative

also called hESC MLL-AF4 model · AND-1 MLLAF4 · iPSC reprogramming of MLL-rearranged B-ALL

Two ways of using stem cells to study this leukaemia have both failed: putting the fusion into embryonic stem cells stops them making blood, and patients' leukaemia cells refuse to be turned back into stem cells at all.

The same thing, for a scientist

**Forward route.** MLL-AF4 was expressed in human embryonic stem cells (the AND-1 MLLAF4 line, CVCL_V162, and its FLT3-ITD and FLT3-D835 derivatives CVCL_V164 and CVCL_V163, all in the Spanish National Stem Cell Bank). MLL-AF4 acts as a global transcriptional activator and a positive regulator of homeobox gene expression in hESCs, and **impairs early hemato-endothelial specification** — it skews haematoendothelial development but is insufficient for transformation. Adding FLT3 activation, the candidate cooperating event given that MLL-AF4 pro-B ALL expresses enormous FLT3, **abrogates** hematopoietic specification rather than rescuing it. **Reverse route.** Reprogramming MLL-rearranged human B-ALL **patient** blasts into induced pluripotent stem cells failed across mono- and polycistronic OKSM(L) constructs delivered by retroviral, lentiviral, episomal and Sendai-virus vectors — a developmental refractoriness rather than a technical shortfall.

Go deeper
Two failures pointing the same way

The fusion blocks haemato-endothelial specification when expressed in pluripotent cells 1, and blasts carrying it cannot be returned to pluripotency 3. Both say that MLL-AF4 and the pluripotent/early-specification state are incompatible, which is a substantive statement about the fusion and not merely a record of two dead ends.

It also argues, with the fetal-liver CRISPR result, that the fusion needs a specific, already-committed developmental compartment — fetal haematopoietic cells — and is either inert or actively obstructive earlier. That narrows the cell-of-origin question from the direction opposite to the usual transduce-and-transplant approach, which this graph already holds cannot settle it.

The FLT3 result is the sharpest single fact here

FLT3 is massively expressed in MLL-AF4 pro-B ALL and is a standing therapeutic target. In the hESC system, FLT3 activation combined with MLL-AF4 abrogates haematopoietic specification 2. So the two lesions that co-occur in the patient are, in an early developmental context, jointly incompatible with making blood at all.

That is either a statement about developmental timing — both lesions must arrive after specification — or a statement that the hESC system is the wrong context. Either reading constrains the model; the first is the more interesting and is testable in the fetal-liver system.

Facts and figures
model classhuman pluripotent stem cells
obtainablehESC derivatives are banked and distributable (Spanish Stem Cell Bank); no PATIENT-DERIVED iPSC exists, but an engineered inducible KMT2A-AFF1 iPSC does (CVCL_C9HE)
forward route
{
  "lines": [
    {
      "name": "AND-1 MLLAF4",
      "accession": "CVCL_V162",
      "construct": "pRRL-EF1a-MLLAF4-PGKp-Neo"
    },
    {
      "name": "AND-1 MLLAF4 FLT3 D835",
      "accession": "CVCL_V163"
    },
    {
      "name": "AND-1 MLLAF4 FLT3 ITD",
      "accession": "CVCL_V164"
    }
  ],
  "parent": "CVCL_B526 (AND-1 hESC)",
  "bank": "Spanish Stem Cell Bank (Banco Nacional de Lineas Celulares)",
  "outcome": "impaired hemato-endothelial specification; insufficient for transformation; FLT3 activation abrogates specification"
}
reverse route
{
  "attempt": "iPSC reprogramming of MLL-rearranged primary B-ALL blasts",
  "vectors_tried": [
    "retroviral OKSM",
    "lentiviral OKSM(L)",
    "episomal",
    "Sendai virus"
  ],
  "outcome": "failed"
}
How this could bite us back
  • If MLL-AF4 obstructs haematopoietic specification at early developmental stages, a therapy that reactivates or mimics early developmental programmes could be antagonistic rather than differentiating.
  • FLT3 activation plus MLL-AF4 abolishes blood specification in this system. FLT3 inhibitors are an active combination hypothesis in this disease and this result gives no support for the combination's mechanism as usually stated.
  • No PATIENT-DERIVED iPSC of this disease exists, so the design iPSC modelling is prized for — fusion-corrected versus fusion-carrying from one patient — is unavailable for KMT2A::AFF1. (Correction added in the same cycle: an ENGINEERED doxycycline-inducible KMT2A-AFF1 iPSC from a healthy donor does exist, IMSUTi002-A-2 / CVCL_C9HE; see `model-inducible-kmt2a-aff1-ipsc`. It supplies an isogenic on/off control from the opposite direction and does not supply a patient background.)
Still open
  1. Does the reciprocal AF4-MLL fusion rescue specification in the same hESC system? One paper says the combination enhances it, which if robust reverses the main result of the forward route.
  2. Is the reprogramming refractoriness specific to MLL rearrangement or general to B-ALL? The paper tested several B-ALL subtypes; the abstract does not say whether any non-MLL subtype succeeded.
  3. Would the hESC lines behave differently if the fusion were expressed from the endogenous KMT2A locus rather than from EF1a? Every result in this route is an overexpression result.
Every source behind this entry (3)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Expression of MLL-AF4 in human embryonic stem cells has a phenotypic, functional and gene-expression impact: MLL-AF4 acts as a global transcriptional activator and a positive regulator of homeobox gene expression in hESCs, and impairs early hemato-endothelial specification.

    PMID 22212479DOI
    What this evidence cannot show

    Abstract read; full text not read. The construct is a transduced MLL-AF4 cDNA expressed from an EF1a promoter (Cellosaurus records the AND-1 MLLAF4 line as pRRL-EF1a-MLLAF4-PGKp-Neo), so this is overexpression in a non-physiological context and from a single hESC background with a recorded abnormal karyotype (46,XY,+i(7)p(11),-7).

    Evidence that argues against this (1)
    • A later study from the same group reports that developmental cooperation between AF4-MLL and MLL-AF4 — the reciprocal fusion together with the direct one — ENHANCES hemato-endothelial specification during human embryonic differentiation, the opposite direction to MLL-AF4 alone. So the impairment reported here is a property of expressing one half of the translocation, and about half of t(4;11) patients express both. Bueno C et al. 2019
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In human embryonic stem cells MLL-AF4 skewed haematoendothelial specification but was insufficient for transformation; FLT3 activation, a candidate cooperating event because MLL-AF4 pro-B ALL expresses enormous levels of FLT3, cooperates with MLL-AF4 to abrogate the haematopoietic specification of human ESCs.

    PMID 23479570DOI
    What this evidence cannot show

    Abstract read; full text not read. Two FLT3-activating alleles were modelled (ITD and D835, as the Cellosaurus lines CVCL_V164 and CVCL_V163) and the abstract does not separate their effects. Abrogation of specification in a directed-differentiation assay is not the same as an effect on haematopoiesis in vivo.

    Evidence that argues against this (1)
    • FLT3-ITD is in fact uncommon in infant KMT2A::AFF1 ALL even though FLT3 expression is high — MV4-11, the line that carries an ITD, is a 10-year-old's monocytic AML. So the cooperating event modelled here may not be the cooperating event the disease uses, and the abrogation result may be about a genotype patients largely do not have. Bairoch A et al. 2025
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Attempts to generate induced pluripotent stem cells from different subtypes of B-cell acute lymphoblastic leukaemia, using mono- and polycistronic OKSM(L)-expressing retroviral, lentiviral, episomal and Sendai-virus reprogramming strategies, failed; iPSC generation from acute myeloid or lymphoid leukaemias has not been achieved, and the authors frame this as a developmental refractoriness of MLL-rearranged human B-cell acute leukaemias to reprogramming into pluripotency.

    Development Refractoriness of MLL-Rearranged Human B Cell Acute Leukemias to Reprogramming into Pluripotency · Muñoz-López A, Romero-Moya D, Prieto C, et al. · Stem Cell Reports 2016
    PMID 27666791DOI
    What this evidence cannot show

    Abstract read; full text not read; the abstract is truncated in the PubMed record mid-sentence. A negative reprogramming result across several vector systems in one laboratory is a strong technical negative but is still one laboratory, and reprogramming efficiency for primary leukaemia is very low even when it succeeds, so a null could be a power problem.

    Evidence that argues against this (1)
    • The paper's own framing concedes that reprogramming leukaemias is an extremely inefficient process and that a few studies did generate iPSCs from primary chronic myeloid leukaemia. An inefficiency that is overcome in one leukaemia and not another is evidence of a difference in degree; calling it developmental refractoriness is an interpretation that a larger attempt could overturn. Tan YT et al. 2018
Connected to