IGF2BP3 marks the fetal programme AND is required for B-ALL survival - the trap is not expression-for-requirement, it is requirement-for-state-maintenance
also called IGF2BP3 marker versus maintainer · oncofetal expression trap
IGF2BP3 is switched on in infant leukaemia and its level predicts how patients do. Removing it does kill B-ALL cells, so the leukaemia does need it - but nobody has checked whether what it maintains is the fetal programme or just growth.
The same thing, for a scientist
**TITLE AND FRAMING CORRECTED 2026-09-21 (Q68).** IGF2BP3 is an oncofetal RNA-binding protein whose expression associates with a proliferative phenotype and with prognostic features in B-cell acute lymphoblastic leukaemia [E1], and the IGF2BP family is already listed as a marker of `cell-state-fetal-program-retention` in this graph. A B-cell developmental gene regulatory network including IGF2BP3 is activated in infant AML [E2]. This entry previously called all of that evidence class 1 and asserted that no class-2 measurement supported a requirement. **That was wrong.** Palanichamy et al. (PMID 26974154, J Clin Invest 2016) report IGF2BP3 as specifically overexpressed in MLL-rearranged B-ALL and **required for the survival of B-ALL cell lines** - knockdown decreases proliferation and increases apoptosis - with enforced expression giving murine bone-marrow cells a survival advantage, expanding haematopoietic stem and progenitor cells and skewing development to the B-cell/myeloid lineage, and with MYC and CDK6 as direct CLIP-identified 3'UTR targets. That is class 2 in the index lineage and in the index fusion class, plus a gain-of-function lineage effect. The class-2 measurement this entry cited - IGF2BP3 depleted below -0.5 in 1.6% of 1,258 screened lines, -0.087 selective for the KMT2A-rearranged arm [E3] - therefore **disagrees with a knockdown in the relevant lineage**, and the disagreement is the finding rather than the null. What survives, and is still the real gap: **no experiment removes IGF2BP3 and scores the fetal-programme signature.** The requirement is established; what it is a requirement *for* is not. The trap this entry documents is therefore a class-2-for-class-3 substitution, not a class-1-for-class-2 one.
Why this is the trap and not merely a gap
The temptation with IGF2BP3 is specific and strong. It is oncofetal — off in normal adult tissue, on in the tumour — which is exactly the profile a therapeutic-window argument wants. It correlates with proliferation and with outcome. It is a marker of a cell state this graph already considers central.
Each of those is a class-1 observation. Stacking three class-1 observations does not produce a class-2 conclusion, and the one class-2 measurement available says the screened cells do not need it 3. The graph should record the gap rather than let the accumulation of correlations stand in for a requirement.
The counter-evidence that complicates it, honestly
A small-molecule IGF2BP3 inhibitor shows anti-leukaemic activity 4. That is a pharmacological requirement claim and it is not consistent with the genetic null. Three readings are available and the graph should not choose without evidence: the compound has off-target activity; acute chemical inhibition and chronic knockout differ (a well-documented divergence); or the cell lines in DepMap are not the ones where the requirement lives. The third is testable — none of the eleven KMT2A-rearranged lines is infant material.
Correction, 2026-09-21 (Q68)
What was withdrawn. The previous title - 'IGF2BP3 marks the fetal programme and has not been shown to maintain it - the beat's clearest expression-for-requirement trap' - and the summary sentence 'the experiment that would show a need has not been done'. A perturbation experiment showing a need exists and predates this entry by ten years.
Why. PMID 26974154 knocks IGF2BP3 down in B-ALL cell lines (decreased proliferation, increased apoptosis) and enforces its expression in murine bone marrow (survival advantage, HSPC proliferation, B-cell/myeloid skewing). Both arms are class 2. The paper is framed on MLL-rearranged B-ALL specifically, i.e. the index lineage.
Why it was missed, which is the transferable part. The entry's evidence base was an association study, a network analysis and a DepMap lookup. None of the three is a perturbation, and the DepMap lookup was allowed to stand as the class-2 measurement because it was the only perturbation-shaped number in view. A genome-wide screen is not a substitute for asking whether anyone has done the knockdown, and when a screen and a targeted knockdown disagree the screen is not automatically right - DepMap's panel contains few MLL-rearranged B-ALL lines and none of infant origin.
What is unchanged. Evidence items E1-E4 and their claims stand as written; only the entry's synthesis changed. The genuinely open question - does removing IGF2BP3 collapse the fetal-programme signature, or only proliferation? - is unaffected and is now the entry's whole point.
| gene | gene-igf2bp3 |
|---|---|
| uniprot | O00425 |
| evidence class available | 1 |
| evidence class claimed elsewhere | 2 |
| marks | cell-state-fetal-program-retention |
| the missing experiment | remove IGF2BP3 from a KMT2A::AFF1 infant-derived model and score the fetal-programme signature (LIN28B, HMGA2, let-7) rather than viability |
depmap — the numbers behind the chart
{
"pct_all_lines_depleted": 1.6,
"mean_all_lines": -0.137,
"mean_kmt2ar": -0.224,
"contrast_A": -0.087
}- An oncofetal expression profile is the most persuasive-looking class-1 evidence there is, and this graph has already been bitten once by expression standing in for requirement; IGF2BP3 is where it would happen again.
- If the chemical inhibitor's activity is off-target, a programme built on it would fail late and expensively.
- IGF2BP3 is expressed in normal fetal tissue, so in the youngest patients - the ones with the worst outcome - the therapeutic window is narrowest exactly where it is most needed.
- Does removing IGF2BP3 collapse the fetal-programme signature, or only proliferation? PMID 26974154 establishes a survival requirement and scores MYC and CDK6, not a fetal-programme signature. This is now the entry's central question rather than one of three.
- Why does the DepMap CRISPR null (1.6% of lines depleted) disagree with a knockdown that decreases proliferation and increases apoptosis in B-ALL lines? Panel composition, knockout-versus-knockdown kinetics and RNA-binding-protein buffering are all testable and none has been tested here.
- Is IGF2BP3 required in infant-derived material specifically? No infant line is in DepMap, and the murine gain-of-function arm used adult bone marrow.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
IGF2BP3 expression associates with a proliferative phenotype and with prognostic features in B-cell acute lymphoblastic leukaemia.
IGF2BP3 Associates with Proliferative Phenotype and Prognostic Features in B-Cell Acute Lymphoblastic Leukemia. · Mäkinen A, Nikkilä A, et al. · Cancers 2021What this evidence cannot show
An association study in human B-ALL. Correlation with phenotype and outcome; no perturbation, so this is evidence class 1 and cannot support a requirement claim. Abstract read in this beat.
Evidence that argues against this (1)
- PUBLISHED CONFLICTING RESULT that moves IGF2BP3 out of evidence class 1 and refutes this entry's headline. Palanichamy et al. report that IGF2BP3 is SPECIFICALLY OVEREXPRESSED in MLL-rearranged B-ALL, that it is REQUIRED FOR THE SURVIVAL of B-ALL cell lines - knockdown gives decreased proliferation and increased apoptosis - and that enforced expression gives murine bone-marrow cells a strong survival advantage, drives proliferation of haematopoietic stem and progenitor cells, and skews haematopoietic development toward the B-cell/myeloid lineage. CLIP-seq identifies MYC and CDK6 as direct 3'UTR targets and enforced expression raises Myc and Cdk6 in marrow. That is a loss-of-function requirement plus a gain-of-function lineage effect, in the index lineage and in the index fusion class. The association study cited here is class 1; this paper is not. Palanichamy JK et al. 2016
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A B-cell developmental gene regulatory network including IGF2BP3 is activated in infant AML.
A B-cell developmental gene regulatory network is activated in infant AML. · Bolouri H, Ries R, et al. · PloS one 2021What this evidence cannot show
A computational network analysis of expression data in human infant AML - the right patient population, the wrong evidence class for a requirement. Abstract read in this beat.
Evidence that argues against this (1)
- PUBLISHED CONFLICTING RESULT that inverts the direction this network analysis is read in. Palanichamy et al. show that ENFORCED IGF2BP3 expression in mice skews haematopoietic development to the B-cell/myeloid lineage and expands stem and progenitor cells. If IGF2BP3 can drive a B-lineage skew, then finding a B-cell developmental gene regulatory network activated in infant AML alongside IGF2BP3 is at least as consistent with IGF2BP3 activity PRODUCING the B-cell programme as with the programme being an inherited feature of the cell of origin. The computational network analysis cannot orient that arrow, and this entry currently treats the network as context for a passive marker. Palanichamy JK et al. 2016
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
In DepMap fetched 2026-09-20, IGF2BP3 is depleted below -0.5 in 1.6% of 1,258 screened lines (mean -0.137) and its mean across eleven KMT2A-rearranged lines is -0.224.
Open Targets Platform - DepMap CRISPR gene essentiality (depMapEssentiality), queried by Ensembl gene id 2026-09-20 · Open Targets Platform GraphQL API 2026What this evidence cannot show
Cancer lines in culture, none of them infant material. A three-week viability screen may miss a requirement that acts on differentiation or on long-term self-renewal rather than on survival.
Evidence that argues against this (1)
- AUTHORS' OWN STATED LIMITATION of the screen, and here it matters because the screen and a knockdown disagree. IGF2BP3 knockdown decreases proliferation and increases apoptosis in B-ALL cell lines, which is exactly the phenotype a fitness screen exists to detect - yet DepMap scores IGF2BP3 as depleted below -0.5 in 1.6% of lines. The Chronos authors state that CRISPR loss-of-function screens 'exhibit a number of biases and artifacts that can confound the results' and that their model infers fitness from cell proliferation dynamics; they also report residual copy-number and screen-quality bias even in the best-performing method. When a knockdown in the relevant lineage and a genome-wide knockout screen over a panel containing few MLL-rearranged B-ALL lines disagree, the screen is not automatically the arbiter. Dempster JM et al. 2021
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A small-molecule inhibitor of the RNA-binding protein IGF2BP3 shows anti-leukaemic activity.
A small molecule inhibitor of RNA-binding protein IGF2BP3 shows anti-leukemic activity. · Jaiswal AK, Scherer GM, et al. · Haematologica 2026What this evidence cannot show
2026 report in Haematologica; abstract read in this beat, full text not opened. Chemical-probe selectivity is not established here and the result is not consistent with the genetic null, which is why it is carried as counter-evidence to this entry's own framing rather than resolved.
Evidence that argues against this (1)
- The genetic measurement in E3 finds no requirement in 98.4% of screened lines, so either the compound acts off target, or acute inhibition differs from chronic knockout, or the relevant cells are not in the screen. Open Targets Platform GraphQL API 2026
the cell-context dossier for this gene
the beat's clearest case of three class-1 observations being read as a class-2 conclusion
the expression association is in B-cell precursor ALL
contrast case: unlike IGF2BP3, LIN28B is reported necessary for transformation, not merely a marker of the fetal state
module co-regulation STRENGTHENS the expression-is-not-requirement scepticism
TITLE CORRECTED: IGF2BP3 knockdown decreases proliferation and increases apoptosis in B-ALL lines and enforced expression expands HSPCs and skews lineage, in MLL-rearranged B-ALL specifically (PMID 26974154) - a class-2 requirement in the index lineage that this entry had recorded as absent
enforced IGF2BP3 skews murine haematopoietic development to the B-cell/myeloid lineage, so a B-cell programme found alongside IGF2BP3 may be its CONSEQUENCE rather than evidence of cell of origin - the network analysis cannot orient that arrow