LIN28B has two routes - blocking let-7, and binding mRNAs directly - and in KMT2A::AFF1 every published mechanism is the direct-mRNA one (let-7 IS measured in mouse MLL-fusion AML: in-session correction)
also called LIN28B/let-7 axis · LIN28B let-7-independent mechanism · LIN28B-MYBBP1A-MYB axis · the two arms of LIN28B
LIN28B is famous for one job: holding down a growth-braking RNA called let-7. But it has a second, less famous job - sticking to ordinary messenger RNAs and making them last longer. Every published explanation that names a specific target inside an MLL/KMT2A fusion is the second job, not the famous one.
The same thing, for a scientist
The textbook LIN28B mechanism is repression of let-7 biogenesis: LIN28B binds primary let-7 transcripts and sequesters them in the nucleolus away from the Microprocessor, a ZCCHC11-independent route distinct from LIN28A's cytoplasmic TUTase mechanism [E1]. That route is real and it is the developmental clock of haematopoiesis [E2]. **But in KMT2A-rearranged leukaemia, the mechanism actually measured is the other one.** Eldeeb et al. showed LIN28B impedes MLL-ENL AML by **binding the MYBBP1A mRNA directly**, elevating MYBBP1A protein, which acts as a MYB co-repressor to shut down MYB signalling and leukaemia-initiating-cell activity; overexpressing MYBBP1A phenocopies LIN28B and perturbing it abolishes the LIN28B effect [E3]. Ling et al., in the opposite direction and the opposite lineage, report LIN28B acting by **stabilising** early B-lymphoid, epigenetic-regulator, cell-cycle and anti-apoptotic transcripts in the human fetal-liver CRISPR-KMT2A::AFF1 model [E4]. Both are direct mRNA effects. **In the INDEX fusion, let-7 has never been measured** - zero title/abstract records against `KMT2A`, `"MLL-AF4"` or `"t(4;11)"`. But **that absence does not extend to MLL-rearranged leukaemia generally**, and an earlier draft of this entry wrongly said it did: the operand `MLL` returns three records, and Chen et al. 2016 report the c-Myc/Lin28B/let-7g programme as PARTNER-DISCRIMINATING across mouse MLL-fusion AML models, with forced let-7g differentiating blasts [E6]. So route 1 is live in MLL-rearranged myeloid disease in mouse, and unmeasured in KMT2A::AFF1.
The two routes, side by side
| route 1: let-7 repression | route 2: direct mRNA binding | |
|---|---|---|
| what LIN28B binds | primary let-7 transcripts | mature mRNAs, via GGAG-type elements recognised by the cold-shock domain and CCHC zinc knuckles |
| compartment | nucleolus (sequestration away from Microprocessor) | cytoplasm, on translating/stored mRNA |
| output | mature let-7 falls, so let-7 targets (HMGA2, IGF2BP1, MYC, CDK6, ...) rise | the bound transcript is stabilised or better translated |
| paralogue difference | LIN28A uses ZCCHC11/TUT4 on pre-let-7 in the cytoplasm instead - a genuinely different mechanism, not a redundancy | shared in principle by both paralogues |
| status in KMT2A-r leukaemia | measured in MOUSE MLL-fusion AML (let-7g, partner-discriminating); never in KMT2A::AFF1 or human KMT2A-r ALL | the only route traced to a NAMED TARGET inside a KMT2A fusion, twice, in opposite directions |
Why the distinction is not pedantic. Route 1 is indirect, slow, family-wide and would be phenocopied by a let-7 mimetic. Route 2 is direct, transcript-specific, and would not be. They imply different drugs, different biomarkers and different resistance mechanisms. The graph has been drawing route 1 because it is the famous one - and route 1 is not fictitious in this disease family, it is simply unmeasured in the index fusion.
The MYBBP1A route in detail, because it is the best-characterised LIN28B mechanism in this disease family
Eldeeb et al. used a dual-transgenic mouse co-expressing MLL-ENL and LIN28B 3. Findings, in the authors' own logical order:
- LIN28B impedes MLL-r AML progression, via compromised leukaemia-initiating-cell activity and suppression of MYB signalling.
- Mechanistically, LIN28B directly binds MYBBP1A mRNA, raising MYBBP1A protein.
- MYBBP1A is a MYB co-repressor; overexpressing it phenocopies LIN28B's tumour-suppressor effect.
- Perturbing MYBBP1A omits the LIN28B effect - the necessity arm.
- Conclusion: developmentally restricted LIN28B expression provides a layer of protection against MYB-dependent AML.
That is a four-step causal chain with both sufficiency and necessity, in a KMT2A fusion, and it never invokes let-7. gene-myb already exists in this graph; gene-mybbp1a is written by this beat.
The direction is the awkward part and must not be smoothed over. Here LIN28B is a tumour suppressor. In Ling et al. it is required for transformation. Those are opposite signs of the same protein in the same fusion family, and the honest resolution offered by the data is lineage and species, not a reconciliation of mechanism. See hyp-fetal-state-is-permissive-for-lymphoid-and-protective-for-myeloid-kmt2a-transformation.
Where let-7 DOES have a leukaemia measurement - including one in MLL-rearranged disease that this entry first missed
CORRECTED IN-SESSION 2026-09-23. An earlier draft of this entry said let-7 had never been measured in any KMT2A rearrangement. That was a NAME-LIST ARTEFACT: the operand MLL was not tried and returns three records, one of which (Chen et al., Oncotarget 2016, PMID 27007052) measures let-7g across MLL-fusion AML models and shows forced let-7g differentiates blasts. The surviving, narrower absence is: no let-7 measurement in KMT2A::AFF1, in t(4;11), or in human infant ALL. See noncoding-let-7-microrna-family for the full query table.
Zhou et al. report that knocking down LIN28B in AML cell lines causes cell-cycle arrest and loss of proliferation and colony formation, that the effect is induced by de-repression of let-7a, and that IGF2BP1 is a downstream target of LIN28B via let-7 in AML; LIN28B overexpression increased tumorigenicity in vivo and knockdown slowed xenograft growth 5.
So route 1 does operate in myeloid leukaemia cells. Three reasons it does not settle the KMT2A::AFF1 question:
- The cells are unselected AML lines, not KMT2A-rearranged fetal B-lineage cells; the abstract names no fusion.
- The sign is opposite to Eldeeb and Li in the same broad disease - LIN28B as oncogene there, as tumour suppressor here. Three papers, three signs, one protein (see
concept-lin28b-has-three-published-signs-in-myeloid-leukaemia). - Cell-line dependency on an overexpressed RBP is the weakest of the three designs. Eldeeb and Li both used genetic mouse models with defined developmental timing.
And the MLL-rearranged measurement that the correction surfaced 6. Chen et al. compared mouse AML transformed by several different MLL fusion proteins and found Lin28B expression tracking c-Myc, let-7g highest in the longest-latency leukaemias, and forced let-7g inducing differentiation. This is route 1 operating inside a KMT2A rearrangement, in mouse myeloid disease, and it is partner-discriminating - which makes 'where does AFF1 sit on this axis' a specific question rather than a general one. It does not change the fact that both mechanisms traced inside a KMT2A fusion to a named molecular target (MYBBP1A; the Ling et al. stabilised transcripts) are route-2 events.
| mechanism class | post-transcriptional RNA regulation |
|---|---|
| effector protein | gene-lin28b |
| the discrepancy | Route 2 has been measured twice inside KMT2A fusions and the two results have OPPOSITE SIGNS. Route 1 has been measured in mouse MLL-fusion AML and in human AML, but NEVER in KMT2A::AFF1 or any human KMT2A-rearranged ALL - which is the narrower and correct statement of the gap. |
| shown for fusion | Route 2: YES for KMT2A::MLLT1 (mouse, tumour-suppressive via MYBBP1A) and YES for KMT2A::AFF1 (human, transformation-permissive, abstract-only). Route 1: YES in mouse MLL-fusion AML as a partner-discriminating let-7g signal (partners unnamed in the abstract); NO in KMT2A::AFF1 or any human KMT2A-r ALL. |
| subnuclear localization | The route-1 step is nucleolar; see localization-lin28b-nucleolus-pri-let-7-sequestration. |
| read depth | E3, E5, E1, E2 abstract-only; E4 abstract-only and paywalled; the companion Li 2024 letter was read in FULL from PMC10912017. |
route 1
{
"name": "let-7 biogenesis repression",
"compartment": "nucleolus",
"substrate": "primary let-7 transcripts",
"cofactor": "none required (ZCCHC11-independent, unlike LIN28A)",
"downstream": [
"noncoding-let-7-microrna-family",
"gene-hmga2",
"gene-igf2bp1"
],
"measured_in_kmt2a_rearranged_leukaemia": "CORRECTED: yes in MOUSE MLL-fusion AML (PMID 27007052, let-7g, partner-discriminating), and as an expression anticorrelation in human infant AML (PMID 34793513). Still NO in KMT2A::AFF1, t(4;11) or human infant ALL.",
"measured_in_any_leukaemia": "yes - mouse MLL-FP AML (let-7g, PMID 27007052); unselected AML cell lines (let-7a, PMID 28693523); paediatric t(8;21) AML where it is operative and MLL-rearranged AML where it is reported NOT operative (PMID 24270738); human infant AML as anticorrelation (PMID 34793513)"
}route 2
{
"name": "direct mRNA binding and stabilisation",
"compartment": "cytoplasm",
"substrate": "mature mRNAs",
"downstream": [
"gene-mybbp1a",
"gene-myb",
"early B-lymphoid transcripts (unnamed)",
"epigenetic regulators (unnamed)",
"cell-cycle and anti-apoptotic transcripts (unnamed)"
],
"measured_in_kmt2a_rearranged_leukaemia": true,
"studies": [
"36763502 (MLL-ENL mouse, MYBBP1A, tumour-suppressive)",
"42213620 (human fetal liver KMT2A::AFF1, transformation-permissive)"
]
}- A let-7 mimetic developed on the assumption of route 1 would miss entirely if KMT2A::AFF1 uses route 2 - and the only mechanistic data in the index fusion point at route 2.
- Conversely, an inhibitor of LIN28B's RNA-binding surface blocks BOTH routes, which in a myeloid context would remove a tumour suppressor. Lineage would decide the sign of the therapeutic effect.
- MYBBP1A raises MYB repression; MYB is required for normal haematopoiesis. Any intervention that moves this axis is moving a normal developmental regulator in an infant.
- The graph cannot currently discriminate the two routes in the index disease, so any target dossier written on this axis is resting on a mechanism nobody has measured in the relevant cell.
- Which MLL fusion partners did Chen et al. 2016 compare, and was AF4/AFF1 among them? One full text would tell this project where the index fusion sits on a partner-discriminating let-7g axis.
- Does LIN28B bind let-7 precursors at all in a human fetal liver CD34+ cell, or is its RNA target repertoire there dominated by mRNAs? A CLIP experiment in the Ling et al. material would answer route-1-versus-route-2 in one dataset.
- Is MYBBP1A among the transcripts LIN28B stabilises in the human fetal B-lineage KMT2A::AFF1 model? If it is, the same molecular event is producing opposite phenotypes in two lineages, which would be the most informative single result available on this axis.
- Which of the 'epigenetic regulators' Ling et al. report LIN28B stabilising are KMT2A::AFF1 complex members? That list is invisible behind the paywall and would immediately connect this beat to the core interactome.
- Is the nucleolar pool of LIN28B present in a fetal B progenitor, or is its localisation cell-type-dependent? Route 1 requires the nucleolar pool and nobody has looked in a haematopoietic cell.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
LIN28B represses let-7 through a ZCCHC11-independent mechanism, functioning in the nucleus by sequestering primary let-7 transcripts in the nucleolus and inhibiting their processing by the Microprocessor, whereas LIN28A recruits the TUTase ZCCHC11/TUT4 to let-7 precursors in the cytoplasm to block Dicer.
Lin28A and Lin28B inhibit let-7 microRNA biogenesis by distinct mechanisms. · Piskounova E, Polytarchou C, Thornton JE, LaPierre RJ, Pothoulakis C, Hagan JP, Iliopoulos D, Gregory RI · Cell 2011What this evidence cannot show
Abstract read; full text not opened. Human colon and breast cancer cells and xenografts - no haematopoietic cell and no leukaemia. The paralogue distinction is drawn from over-expression and depletion experiments in those systems.
Evidence that argues against this (1)
- This mechanism is a statement about let-7 biogenesis, not about what LIN28B does in a leukaemia cell. Both LIN28B mechanisms measured in KMT2A-rearranged leukaemia are direct mRNA effects that do not require it. Cell Rep 2023
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The Lin28b-let-7-Hmga2 axis sets the developmental timing of mouse haematopoietic output: Lin28b falls and mature let-7 rises with maturation, let-7 inhibition in adults recapitulates fetal erythroid-dominant haematopoiesis, and Lin28b deletion or let-7 activation in the fetal state produces adult-like myeloid-dominant output.
Developmental regulation of myeloerythroid progenitor function by the Lin28b-let-7-Hmga2 axis. · Rowe RG, Wang LD, Coma S, Han A, Mathieu R, Pearson DS, Ross S, Sousa P, Nguyen PT, Rodriguez A, Wagers AJ, Daley GQ · J Exp Med 2016What this evidence cannot show
Abstract read; full text not opened. Mouse, normal haematopoiesis, myeloerythroid readout. Establishes route 1 as physiologically causal for developmental timing - which is why it is reasonable to expect it in fetal B-lymphopoiesis - but does not measure it there or in any leukaemia.
Evidence that argues against this (1)
- Establishing the axis as a developmental clock in normal mouse myeloerythroid progenitors is not evidence that route 1 operates in a fetal B progenitor, and this entry's central contrast between a famous route and a measured one depends on route 1 being genuinely available in the relevant cell - which no localisation or let-7 measurement in a haematopoietic cell has ever established. Cell 2011
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In a dual-transgenic mouse co-expressing MLL-ENL and LIN28B, LIN28B impedes MLL-rearranged AML progression through compromised leukaemia-initiating cell activity and suppression of MYB signalling; mechanistically LIN28B directly binds MYBBP1A mRNA, elevating MYBBP1A protein, and overexpression of MYBBP1A phenocopies the tumour-suppressor effect of LIN28B while perturbing MYBBP1A omits it.
A fetal tumor suppressor axis abrogates MLL-fusion-driven acute myeloid leukemia. · Eldeeb M, Yuan O, Guzzi N, Thi Ngoc PC, Konturek-Ciesla A, Kristiansen TA, Muthukumar S, Magee J, Bellodi C, Yuan J, Bryder D · Cell Rep 2023What this evidence cannot show
Abstract read; full text not opened. Mouse, transgenic MLL-ENL (the KMT2A::MLLT1 analogue), MYELOID disease, with LIN28B supplied by a transgene rather than at endogenous fetal levels. 'Directly binds' is the authors' wording; the binding assay is not named in the abstract and was not verified here. The abstract explicitly frames the influence of prenatal signals on leukaemogenesis as CONTROVERSIAL, which is a useful signal from inside the field.
Evidence that argues against this (1)
- A companion study from the same collaboration found that conditional Lin28b DELETION in fetal progenitors did not accelerate MLL::ENL AML initiation, so LIN28B is not the sole effector of fetal protection even within this myeloid model - the sufficiency arm is much stronger than the necessity arm. Leukemia 2024
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In the human fetal-liver CRISPR-Cas9 KMT2A::AFF1 model, mechanistic studies showed LIN28B acting by stabilising key early B-lymphoid genes, epigenetic regulators, and cell-cycle and anti-apoptotic genes - a direct transcript-level effect, with no let-7 measurement reported in the abstract.
The fetal specific gene LIN28B is essential for human fetal B-lymphopoiesis and initiation of KMT2A::AFF1 infant leukemia. · Ling RE, Welsh AM, Hamer L, Jackson T, Neil E, Elliott N, et al. · Blood 2026What this evidence cannot show
ABSTRACT ONLY; the paper is paywalled and absent from PMC and Europe PMC (verified 2026-09-23). Crucially, ABSENCE OF A LET-7 MEASUREMENT IN AN ABSTRACT IS NOT ABSENCE OF ONE IN THE PAPER (rule 80), and this entry's claim that 'no KMT2A study has measured let-7' is a claim about titles and abstracts. If the paper contains a let-7 panel, this entry's headline weakens. The word 'stabilising' is the abstract's; which assay established stabilisation is not stated, and no transcript is named.
Evidence that argues against this (1)
- The full text could contain a let-7 figure that would move this paper onto route 1. It cannot be checked, because the paper is paywalled and not in PMC - so this is a caveat that cannot currently be discharged. source 2026
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In AML cells, targeting LIN28B caused cell-cycle arrest and inhibition of proliferation and colony formation, induced by de-repression of let-7a; IGF2BP1 was confirmed as a downstream target of LIN28B via let-7 in AML; ectopic LIN28B increased tumorigenicity in vivo while silencing it slowed xenograft growth.
Inhibition of LIN28B impairs leukemia cell growth and metabolism in acute myeloid leukemia. · Zhou J, Bi C, Ching YQ, Chooi JY, Lu X, Quah JY, Toh SH, Chan ZL, Tan TZ, Chong PS, Chng WJ · J Hematol Oncol 2017What this evidence cannot show
Abstract read; full text not opened. AML cell lines plus a xenograft; NO FUSION IS NAMED in the abstract, so this is not a KMT2A-rearranged result and cannot be counted as one. LIN28B was modulated by overexpression and silencing in lines that may not express it physiologically. The sign of the LIN28B effect here is opposite to Eldeeb 2023 and Li 2024 in the same broad disease.
Evidence that argues against this (1)
- Li et al. report that most human paediatric AML, including MLLr AML, do NOT express LIN28B and express let-7 family miRNAs at high levels - which makes a LIN28B dependency in unselected AML lines hard to reconcile with the patient data, and suggests the lines are not representative. Leukemia 2024
- E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Across mouse AML models transformed with different MLL fusion proteins, Lin28B was differentially expressed in agreement with c-Myc, let-7g was significantly increased in the leukaemias with the longest disease latency, and forced let-7g expression induced differentiation of leukaemic blasts; the authors conclude the c-Myc/Lin28/let-7g programme is differentially regulated by different MLL fusion partners and relates to disease latency and BET-inhibitor resistance.
Differential regulation of the c-Myc/Lin28 axis discriminates subclasses of rearranged MLL leukemia. · Chen L, Sun Y, Wang J, Jiang H, Muntean AG · Oncotarget 2016What this evidence cannot show
Abstract read; full text not opened, so WHICH MLL fusion partners were compared is unknown - and whether AFF1/AF4 was among them is the single fact this project most needs from the paper. Mouse, myeloid, retroviral MLL-FP models. Oncotarget's editorial standards over this period have been questioned; treated as a strong pointer rather than as established. THIS ITEM WAS ADDED AS A SELF-CORRECTION - it refutes an earlier draft of this entry.
Evidence that argues against this (1)
- In human paediatric AML material the let-7/LIN28B/HMGA2 axis was reported NOT to be operative in MLL-rearranged blasts, in contrast to t(8;21) blasts - so the mouse MLL-FP result may not transfer. Leukemia 2014
the two mechanistic routes LIN28B can use
route 1 - nucleolar sequestration of pri-let-7, never measured in any KMT2A fusion
route 2 - LIN28B stabilises MYBBP1A mRNA, MYBBP1A co-represses MYB
the index-fusion experiment reports a route-2 mechanism (transcript stabilisation)
IGF2BP1 as a route-1 output via let-7 in AML cells
the effector of LIN28B route 2, the let-7-independent one
route 1 needs the nucleolar pool; if absent in fetal B progenitors only route 2 is available
route 1 versus route 2 in the index fusion
- related to →KMT2A::MLLT3 (MLL-AF9), t(9;11)(p21.3;q23.3)
the let-7g/Lin28B/c-Myc programme is partner-discriminating across mouse MLL-fusion AML models