Oncofusion

start / graph / concept-cord-blood-failure-is-method-not-cell

Explainerunrated confidence 0 of 5: unrated

Cord blood does transform — when the translocation is engineered rather than expressed

also called cord blood transformation · CRISPR versus retroviral models

This project had settled on the idea that cord-blood cells cannot be turned leukaemic by this fusion. A different way of making the fusion does turn them leukaemic, so the failure belongs to the method, not to the cell.

The same thing, for a scientist

Several entries in this graph rest on the shorthand that MLL-AF4 in human cord-blood CD34+ cells enhances repopulation but fails to initiate leukaemia, with the fetal-liver CRISPR model succeeding where cord blood does not. A CRISPR/Cas9 model that engineers the endogenous translocation in cord-blood HSPCs reports **monoclonal outgrowth with indefinite growth potential**, an immature myelomonocytic phenotype and elevated KMT2A-target expression comparable to patient cells. The one construct that failed was **MLL-AF4 in adult bone marrow**, which ceased proliferating after 80 days. So the axis is method and cell of origin together, not cord blood alone.

Go deeper
What the correction does and does not change

It does not rescue the retroviral cord-blood results. Enforced overexpression of MLL-AF4 in cord blood still does not initiate leukaemia, and this graph's objection to reading that as a statement about the cell still holds — it is a statement about the construct.

It does change the shorthand. Entries that say 'cord blood fails to transform' are wrong as written. The accurate version is that an authentically engineered translocation transforms cord-blood HSPCs and does not sustain adult bone-marrow HSPCs when the partner is AFF1 — which is a sharper and more useful claim, because it is about the index fusion specifically.

It sharpens the fetal-liver argument rather than removing it. Cord blood is neonatal and infant t(4;11) is prenatal in origin, so a model that works in cord blood is still one developmental step late. The difference between the two models is now a question about what those eighty days of adult bone-marrow proliferation mean.

Still open
  1. Does the CRISPR cord-blood MLL-AF4 model initiate leukaemia in vivo, or only grow indefinitely in vitro? The distinction decides how much of this graph's cord-blood shorthand needs rewriting.
  2. Why does MLL-AF4 specifically fail in adult bone marrow while MLL-AF9 does not? The 80-day ceiling is the most concrete partner-by-cell-of-origin interaction anyone has reported.
Every source behind this entry (2)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In a CRISPR/Cas9 KMT2A-rearrangement model applied to HSPCs from human cord blood and from adult bone marrow, both genome-edited infant and adult KMT2A-rearranged cells showed monoclonal outgrowth with an immature morphology, myelomonocytic phenotype and elevated KMT2A-target gene expression comparable to patient cells, and all presented with indefinite growth potential EXCEPT MLL-AF4 bone-marrow cells, which ceased proliferating after 80 days. FFAR2 was nominated as a candidate epigenetic tumour suppressor.

    PMID 32517300DOI
    What this evidence cannot show

    Abstract read; full text not read. Indefinite growth potential in vitro is not leukaemia initiation in vivo, and the abstract does not state xenograft outcomes for each construct.

    Evidence that argues against this (2)
    • Immortalisation and transformation are different endpoints, and the difference matters for exactly this argument. The retroviral cord-blood studies this corrects were scored on leukaemia initiation in immunodeficient mice; indefinite growth in culture is a weaker bar and may be met by cells that would not initiate disease. Neoplasia 2026
    • FFAR2 is invisible to the instrument that would test it: in DepMap it sits at −0.218 with 3.6% of lines depleted, which is what a gene with no dependency looks like — as expected for a proposed tumour suppressor, and a reminder that a dependency screen cannot evaluate a suppressor claim. Oncofusion knowledge graph (internal analysis) 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    A later model from the same group introduced MLL::AF4 or MLL::AF9 with KMT2A breakpoints in intron 9 or 11 into cord-blood-derived CD34+ cells from the SAME donor, finding breakpoint- and partner-specific expression patterns; MLL(intron 9)::AF9 cells formed colonies robustly after prolonged culture, engrafted in NOD scid gamma mice, and switched from myeloid to B-lymphoid identity in vivo.

    PMID 42001795DOI
    What this evidence cannot show

    Abstract read; full text not read. The engrafting and lineage-switching construct is MLL::AF9, not the index fusion MLL::AF4.

    Evidence that argues against this (2)
    • Same-donor comparison is the design this graph's comparative beat has been missing, and it delivers its clearest result for the partner that is NOT the index fusion. That the AF9 construct engrafts and the abstract does not report the same for AF4 is consistent with the earlier finding that AF4 is the construct with the growth problem. Cancers 2020
    • Lineage switching in a xenograft is not the lineage switching seen in patients, which occurs under CD19-directed immunological pressure. An immunodeficient mouse applies no such pressure, so the two phenomena share a name and not a cause. Leukemia 2023
Connected to