Oncofusion

start / graph / gene-lin28b

Genemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q132-interactome-tail-promotion

LIN28B - a one-row BioID prey of KMT2A that a 2026 Blood paper shows is REQUIRED for KMT2A::AFF1 to transform human fetal cells

also called lin-28 homolog B · CSDD2 · FLJ16517 · Protein lin-28 homolog B

LIN28B is a gene switched on in unborn babies and switched off after birth; it blocks a family of small RNAs called let-7 that would otherwise force blood cells to grow up. In 2026 a Blood paper showed that human fetal blood stem cells cannot be turned into KMT2A::AFF1 infant leukaemia at all if LIN28B is missing - which makes it the rare case of a protein in this project's unwritten interactome tail that already has a dedicated experiment in the exact fusion the lab studies.

The same thing, for a scientist

LIN28B (HGNC:32207, ENSG00000187772, UniProt Q6ZN17, 250 aa, 6q16.3-q21) is a fetal-restricted RNA-binding protein that sequesters primary let-7 transcripts in the nucleolus and prevents their processing, sustaining a fetal proliferative programme [E3]. It reached this beat's candidate list as one of the weakest rows in the interactome: a single spoke-expanded BioID2 proximity row with KMT2A as prey, IntAct IM-30059, K562, MI-score 0.27 [E1][E2]. **The literature check inverted the priority.** Ling et al., *Blood* 2026, derived a fetal-specific signature by comparing human fetal and adult HSPC transcriptomes, found LIN28B among the top hits, and then used a human fetal-liver-derived **CRISPR-Cas9 KMT2A::AFF1 model**: *human FL CD34+ cells fail to transform upon induction of KMT2A::AFF1 translocation in the absence of LIN28B*, and LIN28B-expressing CRISPR-KMT2A::AFF1 leukaemias were more proliferative in vitro and in vivo with that advantage lost on knockdown [E4]. This is a **dedicated human KMT2A::AFF1 experiment** - the class this project measured at 0 of 24 sampled load-bearing experiments - and it is about a gene the graph held only as an un-promoted prey row.

Go deeper
Two pieces of evidence of very different weight, kept apart on purpose

The interaction row (weak). IntAct IM-30059, from Khoroshkin et al. 2024: MI:1314 BioID2, MI:2364 proximity, MI:1060 spoke expansion, LIN28B bait / KMT2A prey, MI-score 0.27, host taxid:9606(Homo sapiens K562 myeloid leukemia cell), curator annotation Data file S6 2. One row, one paper, one method, no orthogonal validation, in a BCR::ABL1 line with no KMT2A rearrangement. On its own it would not justify an entry beyond a stub.

The functional requirement (strong, and about the index fusion). Ling et al. 2026 4. Human fetal liver CD34+ cells, CRISPR-Cas9-induced KMT2A::AFF1 translocation, LIN28B knockout: no transformation. LIN28B knockdown in established CRISPR-KMT2A::AFF1 leukaemias: loss of the proliferative advantage, in vitro and in vivo. Mechanistically the authors report LIN28B stabilising early B-lymphoid genes, epigenetic regulators, and cell-cycle and anti-apoptotic genes.

These are not the same claim and this entry does not let them merge. Nothing in Ling et al., as read from the abstract, reports a physical LIN28B-KMT2A or LIN28B-fusion contact. The BioID row is not evidence for the requirement, and the requirement is not evidence for the contact.

Cell type and cell state - the axis that makes LIN28B unusual here

LIN28B is fetal-restricted, and the requirement was demonstrated in human fetal liver CD34+ cells, the cell of origin this project's cell-type-fetal-liver-hsc and cell-state-fetal-program-retention entries already model. Ling et al. frame infant ALL as initiated in utero with a very poor prognosis despite a lack of additional oncogenic driver mutations common in childhood ALL, and propose LIN28B activity as part of the answer to why KMT2A::AFF1 leukaemias are so aggressive 4.

That makes LIN28B a cell-state entry as much as an interactome entry: it is the state of the target cell, not a property of the fusion, and it is the clearest published statement this beat found that the fetal state is permissive rather than incidental. The graph already records LIN28B as marking a fetal-like subgroup in juvenile myelomonocytic leukaemia, and this beat did not re-verify that claim.

Localization

UniProt Q6ZN17 assigns nucleus, nucleolus (ECO:0000269) and cytoplasm (ECO:0000269) 3. The nucleolar pool is the mechanistically loaded one: it is where pri-let-7 is sequestered away from the microprocessor. A nucleolar RNA-binding protein recovered as prey of a KMT2A neighbourhood raises the standing worry in this project's proximity work - that nucleolar and ribosome-biogenesis proteins are an AP-MS and BioID contaminant signature - and this entry does not claim the row escapes it.

Counter-evidence that is directly on point

A 2024 Leukemia paper reports that LIN28B promotes differentiation of fully transformed AML cells but is dispensable for fetal leukemia suppression 5. Read together with Ling et al., the honest position is that LIN28B's role is stage- and lineage-dependent: required for initiation in human fetal B-lineage KMT2A::AFF1, dispensable and even differentiation-promoting in an established myeloid context. Anyone proposing LIN28B as a target has to say which of those states the patient is in.

Facts and figures
hgnc IDHGNC:32207
symbolLIN28B
ensembl gene IDENSG00000187772
uniprotQ6ZN17
length AA250
chromosomal location6q16.3-q21
identifier provenanceHGNC REST fetch/symbol/LIN28B and UniProtKB Q6ZN17 retrieved live 2026-09-23; PMIDs 42213620, 38321106 and 39251607 fetched from PubMed efetch the same day and their titles read back before citation.
subnuclear locationnucleolus - the compartment where pri-let-7 is sequestered
cell type of the observationinteraction row: K562. Functional requirement: human fetal liver CD34+ HSPC - the index cell of origin.
cell statefetal-restricted; the requirement is a property of the fetal state, not of the fusion
epigenetic roleIndirect: blocks let-7 biogenesis and, per Ling et al., stabilises transcripts of epigenetic regulators among other classes. No chromatin-binding activity is claimed.
shown for fusionYES, KMT2A::AFF1, and in HUMAN fetal cells - the strongest fusion-specific claim in this beat. The PHYSICAL interaction, by contrast, is shown for nobody: it is a K562 proximity row against wild-type KMT2A.
evidence classinteraction: single publication, single method, spoke-expanded BioID prey row. Function: a dedicated human KMT2A::AFF1 CRISPR model, abstract-read only.
prev and alias symbols
[
  "FLJ16517",
  "CSDD2"
]
interaction with project baits
[
  {
    "bait": "KMT2A",
    "bait_uniprot": "Q03164",
    "partner_experimental_role": "bait",
    "scaffold_experimental_role": "prey",
    "method": "MI:1314 proximity-dependent biotin identification (BioID2)",
    "interaction_type": "MI:2364 proximity",
    "expansion": "MI:1060 spoke",
    "mi_score": 0.27,
    "imex": "IM-30059",
    "pmid": "39251607",
    "host_cell": "K562 (BCR::ABL1 CML, no KMT2A rearrangement)",
    "n_rows": 1,
    "n_publications": 1
  }
]
functional requirement in the index fusion
{
  "model": "human fetal liver CD34+ cells, CRISPR-Cas9-induced KMT2A::AFF1 translocation",
  "result": "no transformation in the absence of LIN28B; knockdown removes the proliferative advantage of established CRISPR-KMT2A::AFF1 leukaemias in vitro and in vivo",
  "pmid": "42213620",
  "venue": "Blood",
  "year": 2026,
  "read_depth": "abstract only"
}
subcellular location uniprot
[
  "Nucleus",
  "Nucleus, nucleolus (ECO:0000269)",
  "Cytoplasm (ECO:0000269)"
]
How this could bite us back
  • LIN28B is required for normal human fetal B- and NK-lymphopoiesis in the same paper that shows it is required for transformation. Any LIN28B-directed agent given to an infant is being given to a patient whose normal lymphopoiesis uses the same gene - the therapeutic window is narrow by construction.
  • The requirement demonstrated is for INITIATION. Infant ALL is diagnosed after initiation, so a drug that blocks initiation may do nothing to established disease; the knockdown-in-established-leukaemia arm is the only part of the result that is therapeutically load-bearing and it reports a proliferative disadvantage, not regression.
  • A 2024 Leukemia report has LIN28B promoting differentiation of fully transformed AML cells. If that generalises, LIN28B loss in an established myeloid clone could be the wrong direction, and the lineage would decide the sign of the effect.
  • LIN28B acts through let-7 suppression, and let-7 restoration has broad effects on proliferation across normal tissues; the compensation risk is that other LIN28B-independent let-7 regulators take over.
  • Delivery: there is no clinical-grade LIN28B inhibitor, and RNA-binding proteins of this class have been repeatedly nominated and rarely drugged.
Still open
  1. Is there any PHYSICAL relationship between LIN28B and KMT2A::AFF1, or is the requirement entirely a property of the cell state the fusion lands in? The K562 proximity row is the only physical datum and it is against wild-type KMT2A in the wrong cell.
  2. Which of the transcripts LIN28B stabilises in the CRISPR-KMT2A::AFF1 model are fusion targets, and which are fetal-programme genes that would be stabilised in any fetal cell? That separation decides whether LIN28B is a co-operating oncogene or a permissive background.
  3. Does let-7 restoration phenocopy LIN28B loss in the fetal-liver KMT2A::AFF1 model? If it does, the druggable node moves from LIN28B to the let-7 axis.
  4. Is LIN28B expression retained in diagnostic KMT2A::AFF1 infant ALL blasts at a level that would support targeting, and does it fall at relapse?
  5. Does the same requirement hold for KMT2A::MLLT3 and KMT2A::MLLT1 in fetal cells, or is it specific to the AFF1 partner?
Every source behind this entry (5)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Khoroshkin et al. generated in vivo proximity-dependent biotinylation datasets for 50 human RNA-binding proteins in K562 cells, building a large-scale map of RBP protein neighbourhoods, and combined it with CRISPR-interference single-cell readouts and eCLIP target atlases to annotate RBP regulatory modules.

    Systematic identification of post-transcriptional regulatory modules. · Khoroshkin M, Buyan A, Dodel M, et al. · Nat Commun 2024
    PMID 39251607DOI
    What this evidence cannot show

    Abstract read here; this graph's ana-rbp-bioid2-proximity-map-k562 entry records that the full text was read from PMC11385195 in an earlier cycle and that the elongation factors appear nowhere in the body text - the pairs are reachable only through the supplementary data and the IMEx deposition IM-30059. K562 is a BCR::ABL1 chronic myeloid leukaemia line with NO KMT2A rearrangement, so nothing here is measured in the index disease. The authors note that RBP over-expression, BioID2 fusion and the use of a single isoform per protein may bias the data.

    Evidence that argues against this (1)
    • No super elongation complex component, no KMT2A-complex component and no fusion protein is a BAIT in this deposition; KMT2A, MEN1, PSIP1, AFF1, AFF4, ELL, MLLT1, CDK9, CCNT1 and DOT1L appear only in the prey column. A prey-side row in a 2,682-prey map carries much less information than a bait-side one, and the median prey in this map is recovered by five baits. source 2026
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    IntAct serves one curated human LIN28B-KMT2A row: MI:1314 proximity-dependent biotin identification, MI:2364 proximity, MI:1060 spoke expansion, LIN28B experimental role MI:0496 bait and KMT2A MI:0498 prey, MI-score 0.27, host taxid:9606 Homo sapiens K562 myeloid leukemia cell, IMEx IM-30059, publication PMID 39251607, curator annotation 'Data file S6'.

    What this evidence cannot show

    Retrieved live 2026-09-23. Rule 93: PSICQUIC is a floor relative to IntAct REST. MI-score 0.27 is the deposition-wide value for this screen and carries no pair-specific information.

    Evidence that argues against this (1)
    • K562 has no KMT2A rearrangement, LIN28B is a fetal gene and K562 is an adult-derived CML line, so the physiological relevance of a LIN28B neighbourhood measured there to infant KMT2A::AFF1 biology is unestablished. source 2026
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProt Q6ZN17 is 250 aa and records LIN28B as a suppressor of microRNA biogenesis that binds primary let-7 transcripts and sequesters them in the nucleolus away from the microprocessor complex, with experimental subcellular locations nucleus, nucleolus and cytoplasm.

    What this evidence cannot show

    Fetched live 2026-09-23. Rule 91: the ECO:0000269 codes point at papers a curator read, not necessarily the founding or best observation; those papers were not opened.

    Evidence that argues against this (1)
    • The nucleolar pool that makes this annotation mechanistically interesting is also the standing worry about LIN28B's presence in a proximity dataset: this graph's own Q129 beat spent a cycle on whether nucleolar and ribosome-biogenesis proteins in these interactomes are genuine biology or an AP-MS and BioID contaminant signature, and declined all thirteen candidates it examined. source 2026
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Ling et al. derived a fetal-specific gene signature from human fetal versus adult HSPC transcriptomes, identified LIN28B among the top hits and found it expressed in infant ALL; in a human fetal-liver-derived CRISPR-Cas9 KMT2A::AFF1 model, human FL CD34+ cells fail to transform upon induction of the KMT2A::AFF1 translocation in the absence of LIN28B, and LIN28B-expressing CRISPR-KMT2A::AFF1 leukaemias were more proliferative in vitro and in vivo with that advantage lost upon LIN28B knockdown.

    PMID 42213620DOI
    What this evidence cannot show

    ABSTRACT ONLY - the full text was not read and no figure was inspected, so the effect sizes, the number of donors, the knockout efficiency and the in vivo n are all unknown to this entry. The claim quoted is the authors' own summary sentence. A CRISPR-induced translocation in primary fetal liver cells is a demanding model with low transformation efficiency, and 'fail to transform' in such a system needs the paper's own controls to interpret. VERIFY BEFORE BUILDING ON IT: PMID, DOI, journal, year and author list were fetched from PubMed efetch on 2026-09-23 and the PMID resolves to this exact title, but a 2026 Blood paper is new enough that this graph has no independent corroboration of its findings.

    Evidence that argues against this (1)
    • LIN28B promotes differentiation of fully transformed AML cells but is dispensable for fetal leukemia suppression - a directly opposing functional reading in a myeloid, already-transformed context. Leukemia 2024
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    A 2024 Leukemia report titled 'LIN28B promotes differentiation of fully transformed AML cells but is dispensable for fetal leukemia suppression' exists and states, in its title, a role opposite in sign to the initiation requirement reported for fetal B-lineage KMT2A::AFF1.

    PMID 38321106
    What this evidence cannot show

    TITLE ONLY. This record was found by a PubMed title search and its abstract was not retrieved, so the model system, the fusion involved and the strength of the dispensability claim are all unread. It is cited as declared counter-evidence and as a pointer, and must not be quoted as a measurement. Status of this entry is nonetheless 'verified' because the PMID resolves to this exact title; the read-depth limit is recorded here rather than hidden.

    Evidence that argues against this (1)
    • This item is cited from its TITLE alone with no abstract read, so it cannot be weighed against the Blood 2026 result. A title-only citation is the weakest admissible evidence in this graph and it is recorded as a pointer for the next reader rather than as a finding; anyone using it must open the paper first. source 2024
Connected to