Oncofusion

start / graph / localization-lin28b-nucleolar-pri-let-7-sequestration

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q133-lin28b-fetal-ontogeny

LIN28B localization - a nucleolar pool that sequesters pri-let-7, a predominantly cytoplasmic steady state with cell-cycle-dependent nuclear entry, and NO measurement in any haematopoietic cell

also called LIN28B subcellular localization · LIN28B nucleolus · Q6ZN17 subcellular location

LIN28B is mostly found in the body of the cell, but a pool of it goes into the nucleolus - a dense blob inside the nucleus - and that is where it traps the growth-braking let-7 RNA before it can be made. Nobody has ever looked at where LIN28B sits inside a blood cell, let alone a leukaemia cell.

The same thing, for a scientist

UniProt Q6ZN17 (250 aa) records three subcellular locations for LIN28B: **Nucleus** (uncoded), **Nucleus, nucleolus** (ECO:0000269, PubMed 22118463) and **Cytoplasm** (ECO:0000269, PubMed 16971064) [E1]. Unusually for this graph's experience with rule 91, **both coded citations were opened and both are genuinely about localisation of this protein**. The nucleolar call comes from Piskounova et al., where the nucleolar pool is not incidental but *is the mechanism*: LIN28B functions in the nucleus by sequestering primary let-7 transcripts in the nucleolus, away from the Microprocessor, by a route that does not use ZCCHC11 and is therefore different from LIN28A's [E2]. The cytoplasmic call comes from Guo et al., the paper that cloned and named LIN28B, which reports it **predominantly cytoplasmic with cell-cycle-dependent nuclear translocation in Huh7 hepatocellular carcinoma cells** [E3] - a conditional statement UniProt's flat 'Nucleus' line does not carry. **The gap that matters for this project: every one of these observations is in a cancer epithelial or unspecified cell line. There is no published LIN28B localisation measurement in a haematopoietic cell, a fetal liver CD34+ cell, or a KMT2A-rearranged blast.**

Go deeper
The three calls, their sources, and what each can actually support
UniProt callevidence codecited PMIDwhat the paper iscan it support the call?
Nucleusnone--uncoded; a curator inference, not a measurement
Nucleus, nucleolusECO:000026922118463Piskounova 2011 Cell - the paper that established nucleolar pri-let-7 sequestration as LIN28B's mechanismyes, and it is the founding observation
CytoplasmECO:000026916971064Guo 2006 Gene - cloning and characterisation of LIN28B in hepatocellular carcinomayes, and it adds that localisation is predominantly cytoplasmic and that nuclear entry is cell-cycle-dependent in Huh7

Two things to take from this table. First, rule 91 is satisfied here where it usually is not: the coded citations point at papers about the fact. Second, the flat UniProt list loses the most interesting thing Guo et al. found - that the compartment is conditional on cell-cycle phase. A protein whose nuclear pool appears cyclically is a protein whose let-7-repressive activity may be periodic, and no flat annotation can say that.

The absence, stated plainly

Cell types in which LIN28B localisation has been measured, per the citations UniProt carries: Huh7 (hepatocellular carcinoma), and the colon and breast cancer lines of the Piskounova work.

Cell types in which it has NOT been measured, and which this project cares about: human fetal liver CD34+ HSPC, fetal pro-B / pre-pro-B, any KMT2A-rearranged blast or line, any normal haematopoietic cell at any ontogenetic stage.

This is not a pedantic complaint. The whole LIN28B argument for infant leukaemia runs through a fetal haematopoietic cell, and:

  • Route 1 (let-7 repression) requires the nucleolar pool. If LIN28B is purely cytoplasmic in a fetal B progenitor, route 1 is not available there and the mechanism must be the direct mRNA-stabilising one - which is, in fact, the only mechanism reported in the index fusion (see mechanism-lin28b-has-two-routes-and-the-leukaemia-evidence-is-all-on-the-non-let-7-one).
  • The one physical datum linking LIN28B to KMT2A is a K562 BioID2 proximity row with LIN28B as bait and KMT2A as prey (IntAct IM-30059), recorded in gene-lin28b. A nucleolar bait recovering nuclear prey in a proximity screen is exactly the configuration this project's own nucleolar-contaminant analyses warn about, and the compartment question is therefore load-bearing for whether that row means anything.

The experiment is cheap. Immunofluorescence with nucleolar and nuclear counterstains, plus subcellular fractionation, on the CRISPR-KMT2A::AFF1 fetal-liver cells that already exist. It would separate route 1 from route 2 in the correct cell in a week.

Facts and figures
mechanism classsubcellular and subnuclear localization
proteingene-lin28b
uniprotQ6ZN17
length AA250
subnuclear localizationnucleolus - and it is functional, not descriptive: pri-let-7 is held there away from the Microprocessor
paralogue contrastLIN28A acts in the CYTOPLASM on pre-let-7 via ZCCHC11/TUT4. Same output, different compartment, different cofactor requirement. The two paralogues are not interchangeable and a compartment observation for one says nothing about the other.
localization shown for fusionNOBODY HAS LOOKED IN ANY FUSION OR ANY BLOOD CELL. Not 'shown in MLL-AF9 but not AFF1' - there is no haematopoietic localisation measurement for LIN28B at all.
why it matters hereRoute 1 of the LIN28B mechanism requires the nucleolar pool. If LIN28B is cytoplasmic-only in a fetal B progenitor, route 1 is unavailable there and the index-fusion mechanism must be the direct mRNA one - which is the only mechanism the index-fusion paper reports.
method warningThe nucleolar call and the K562 BioID2 proximity row that connects LIN28B to KMT2A sit in the same compartment-shaped hazard: nucleolar and ribosome-biogenesis proteins are a recognised AP-MS and BioID contaminant signature, and this graph has a dedicated analysis that declined thirteen nucleolar candidates on exactly that ground.
assay not yet doneImmunofluorescence with nucleolar (e.g. fibrillarin/nucleolin) and nuclear counterstains, plus biochemical fractionation, in human fetal liver CD34+ cells and in the CRISPR-KMT2A::AFF1 derivatives - material that already exists in the Ling et al. laboratory.
compartments
[
  {
    "compartment": "cytoplasm",
    "status": "predominant at steady state",
    "evidence_code": "ECO:0000269",
    "source_pmid": "16971064",
    "system": "Huh7 and other hepatocellular carcinoma lines",
    "source_is_about_this_fact": true
  },
  {
    "compartment": "nucleus",
    "status": "cell-cycle-dependent entry reported in Huh7; UniProt's own 'Nucleus' line is UNCODED",
    "evidence_code": "none (uncoded)",
    "source_pmid": null,
    "system": null,
    "source_is_about_this_fact": null
  },
  {
    "compartment": "nucleolus",
    "status": "minority pool, and the site of the let-7 mechanism",
    "evidence_code": "ECO:0000269",
    "source_pmid": "22118463",
    "system": "human colon and breast cancer cells",
    "source_is_about_this_fact": true
  }
]
never measured in
[
  "human fetal liver CD34+ HSPC",
  "fetal pro-B / pre-pro-B",
  "any KMT2A-rearranged blast or cell line",
  "any normal haematopoietic cell at any ontogenetic stage"
]
How this could bite us back
  • A small molecule designed against LIN28B's cold-shock domain / zinc-knuckle RNA-binding surface would act wherever the protein is, and the dominant pool is cytoplasmic - so the on-target effect would be mostly on direct mRNA binding, not on let-7, which is the effect the rationale usually cites.
  • If the nucleolar pool is small and conditional, assays that read out let-7 levels will be insensitive readouts of target engagement, and a drug could engage LIN28B fully while moving let-7 barely at all.
  • Nucleolar localisation is a known confound in proximity proteomics. Any interaction claim for LIN28B built on BioID or AP-MS needs a nucleolar-contaminant control, and the one row this graph holds does not have one.
Still open
  1. Where is LIN28B in a human fetal liver CD34+ cell - cytoplasmic only, or with a nucleolar pool? This single image decides whether the let-7 route is even geographically available in the cell where the requirement was demonstrated.
  2. Does the nucleolar pool change when KMT2A::AFF1 is induced? If the fusion redistributes LIN28B, that would be the first physical link between the two beyond a single K562 proximity row.
  3. Is the cell-cycle dependence of nuclear entry reproduced outside Huh7? If LIN28B only enters the nucleus in a particular phase, its let-7 repression is periodic, and a quiescent fetal HSC would experience little of it.
  4. What is the short LIN28B isoform detected in fetal liver tissue by Guo et al., and does it exist in fetal haematopoietic cells? An isoform with a different compartment would change the whole picture and nobody has followed it up.
Every source behind this entry (3)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProt Q6ZN17 (LN28B_HUMAN, 250 aa) records three subcellular locations for LIN28B: Nucleus with no evidence code, Nucleus/nucleolus with ECO:0000269 citing PubMed 22118463, and Cytoplasm with ECO:0000269 citing PubMed 16971064.

    What this evidence cannot show

    Fetched live 2026-09-23 and the evidence codes and source PMIDs were read from the JSON rather than from a rendered page. The bare 'Nucleus' line is UNCODED and should be treated as a curator inference, consistent with this graph's existing finding that uncoded UniProt localisation annotations are not measurements in roughly a tenth of its entries. UniProt lists no isoform- or condition-scoped location comment for LIN28B, so the cell-cycle dependence found in the cited cytoplasm paper is not represented in the database at all.

    Evidence that argues against this (1)
    • UniProt carries no isoform- or condition-scoped location comment for LIN28B, so the cell-cycle-dependent nuclear entry reported in the cited cytoplasm paper is invisible in the database record. Reading the flat three-line annotation without opening the sources gives a static picture of a protein whose compartment is reported as conditional. Gene 2006
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    LIN28B functions in the nucleus by sequestering primary let-7 transcripts and inhibiting their processing by the Microprocessor, a ZCCHC11-independent mechanism distinct from LIN28A's cytoplasmic recruitment of ZCCHC11/TUT4 to pre-let-7; the nucleolus is the compartment of sequestration.

    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 2011
    PMID 22118463DOI
    What this evidence cannot show

    Abstract read; full text and figures not opened, so the imaging or fractionation method behind the nucleolar call was not inspected. Human colon and breast cancer cells and xenografts. NO HAEMATOPOIETIC CELL. This is the paper UniProt cites for the nucleolar location and it is genuinely about that fact - rule 91 is satisfied here, which is worth recording because in this graph's audits it usually is not.

    Evidence that argues against this (1)
    • The founding characterisation paper reports LIN28B as PREDOMINANTLY CYTOPLASMIC at steady state, so the nucleolar pool is a minority pool whose size is likely condition-dependent. A mechanism located in a minority compartment is not thereby the dominant activity of the protein. Gene 2006
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In the paper that cloned and characterised LIN28B, the protein was found to be predominantly localised in the cytoplasm but to show cell-cycle-dependent nuclear translocation in Huh7 hepatocellular carcinoma cells; the same paper reports a short LIN28B isoform detected by western blot in non-tumour liver and in FETAL LIVER, and an unusually long 3'UTR containing let-7 complementary sites, giving indirect evidence that LIN28B is itself a let-7 target.

    Identification and characterization of lin-28 homolog B (LIN28B) in human hepatocellular carcinoma. · Guo Y, Chen Y, Ito H, Watanabe A, Ge X, Kodama T, Aburatani H · Gene 2006
    PMID 16971064DOI
    What this evidence cannot show

    Abstract read; full text not opened. Huh7 and other HCC lines plus clinical HCC samples - an EPITHELIAL, hepatic context. 'Fetal liver' here means the organ as a tissue lysate, NOT the fetal liver CD34+ haematopoietic compartment this project cares about, and conflating the two would be a real error. The short isoform detected in fetal liver is a western-blot band whose identity and function are not established in the abstract, and this graph should not treat it as a characterised isoform.

    Evidence that argues against this (1)
    • A cell-cycle-dependent nuclear pool in a hepatoma line says nothing about a quiescent or cycling fetal haematopoietic progenitor, and no localisation measurement exists in any blood cell. This entry's central claim is that absence, not the HCC result. source 2026
Connected to