LIN28B and IGF2BP3 are co-regulated members of ONE onco-fetal module in infant leukaemia - which strengthens the graph's existing scepticism about IGF2BP3 rather than weakening it

concept-igf2bp3-is-a-marker-of-a-state-not-a-maintainer-of-one poses: does removing IGF2BP3 collapse the fetal-programme signature, or only proliferation? Three things this beat can add.

1. The module framing raises the bar for that question, it does not lower it. If BRD4/POU2AF1 sit upstream of both LIN28B and IGF2BP3 1, then the experiment that would answer it must ask whether IGF2BP3 loss collapses the rest of the module - LIN28B included - or only its own downstream. A signature readout is not enough; the module members must be measured individually.

2. LIN28B does have the class-2 evidence IGF2BP3 lacks in the fetal-programme sense. Ling et al. show LIN28B loss abolishes transformation and damages normal fetal B/NK lymphopoiesis, which is a functional requirement in the fetal programme itself. That is precisely the kind of result the IGF2BP3 entry says is missing for IGF2BP3. So within the module, one member now has it and the others do not, and the honest reading is that LIN28B is currently the better-evidenced node, not that IGF2BP3 has been rescued by association.

3. And there is a result pointing the opposite way for IGF2BP3 in the index fusion specifically. Hanewald et al. report that both reciprocal t(4;11) fusion transcripts are bound by IGF2BP3, resulting in REDUCED KMT2A::AFF1 protein abundance and increased AFF1::KMT2A 3. An oncofetal protein that lowers the oncoprotein it is supposed to support is not a straightforward maintainer of anything, and this graph already records the tension as unresolved in ana-unstated-system-igf2bp3-t411-unresolved and localization-igf2bp3-cytoplasmic-rnp-granules-p-bodies-stress-granules.

     BRD4  +  POU2AF1          <- epigenetic regulators, hypomethylated in infant AML
           |
     +-----+-----+
     |           |
  LIN28B      IGF2BP3          <- onco-fetal targets, also hypomethylated
     |
   (-| let-7  (Let7a-2 anticorrelates with LIN28B in infant AML)
     |
  IGF2BP1                       <- reported let-7 target downstream of LIN28B in AML

Every arrow here is from a different paper and most are correlative. Bolouri et al. is an integrative expression/methylation analysis across ~1500 paediatric AML samples with no perturbation 1; the LIN28B→let-7→IGF2BP1 arrow is from AML cell lines 2. The diagram is a hypothesis about module structure, and this entry labels it as one.

The single most useful test is cheap and has never been run: does BET inhibition, which this graph already tracks for KMT2A-rearranged disease, collapse LIN28B and IGF2BP3 together in an infant-derived sample? Bolouri et al. themselves suggest infant AML may respond to bromodomain inhibitors on the strength of the BRD4 node. If the module collapses as a unit under BET inhibition, then the therapeutic target is the module's switch and not any individual member - which would reframe both this beat and the existing IGF2BP entries.

The module was characterised in infant AML, and the B-cell developmental identity of the network in a myeloid disease is the paper's own headline surprise. Meanwhile the LIN28B requirement is in infant B-ALL biology, and this beat's other entries argue that LIN28B has opposite signs in the two lineages.

Those two facts do not sit comfortably together and should not be smoothed. If LIN28B is protective in myeloid disease [see concept-lin28b-has-three-published-signs-in-myeloid-leukaemia], then a module that switches LIN28B on in infant AML is switching on something that restrains the disease - and Li et al. report that most paediatric AML including MLLr AML does not express LIN28B, which is in tension with Bolouri's infant-AML finding unless infant AML is a distinct subset. It probably is - Bolouri's whole argument is that infant AML is genomically and transcriptionally distinct from older-child AML - but the reconciliation is inference, not measurement, and this entry does not assert it.