Oncofusion

start / graph / hyp-trna-ribosome-biogenesis-kmt2a-r

Hypothesisspeculative confidence 1 of 5: Our own inference or analogy. Not shown to be true.Q10-genome

Hypothesis: transfer-RNA and ribosome biology is a genuinely empty cell in KMT2A-rearranged leukemia, and translational capacity may be the unmeasured dependency

also called ribosome biogenesis KMT2A-r · tRNA modification leukemia · translational control MLL-AF4 · nucleolar biology KMT2A-r

Cells need a supply of protein-building machines (ribosomes) and RNA delivery adaptors (transfer RNAs) to grow. This matters in other blood diseases, and in KMT2A-fusion leukemia it has been looked at only once, as the side effect of a drug rather than as a property of the leukemia itself - but every fast-growing cancer cell needs more of these machines, so finding that this leukemia does too would not by itself mean anything, and the main drug used to test the idea turns out to work a completely different way.

The same thing, for a scientist

**Corrected 2026-09-19.** This entry opened by calling itself 'the sweep's clearest empty cell' and asserting that there is no published study of ribosome biogenesis, nucleolar function, tRNA abundance, tRNA modification or global translation rate in KMT2A-rearranged leukemia. That is withdrawn, and it contradicted a paper the entry itself already cited two paragraphs later: Howard et al. measure ribosome protein gene transcription, ribosome stock depletion, nucleolar stress, p53 activation and a genome-wide translational choke **in MLL-rearranged cells** (PMID 38682900). The cell is not empty. The accurate statement is narrower: **no study characterises ribosome biogenesis, nucleolar function, tRNA pools or tRNA modification in KMT2A-rearranged leukaemia at baseline, or as a dependency created by the fusion**; the one study that measures these variables in MLLr cells measures them as the *response to a drug* (a WDR5 WIN-site inhibitor), and its authors read the phenotype as drug-imposed and p53-dependent rather than fusion-imposed. The surrounding evidence is nevertheless not trivial. **Ribosome biology causes myeloid disease:** partial loss of function of the ribosomal protein RPS14 phenocopies the 5q- myelodysplastic syndrome in normal human haematopoietic progenitors, forced RPS14 expression rescues the patient-derived phenotype, and the underlying defect in pre-rRNA processing is functionally equivalent to Diamond-Blackfan anaemia [E1]. **Translational control is already implicated in KMT2A-fusion cells:** m6A promotes translation of MYC, BCL2 and PTEN in the KMT2A::MLLT3 line MOLM-13, and METTL3 loss produces a translational defect with differentiation [E2]. **tRNA modification is oncogenic in other tissues:** METTL1-dependent m7G tRNA methylation controls translation and tumour growth [E3]. **Downgraded, cycle 35, on three counts.** (i) *Genericity.* Increased Pol I transcription of rDNA is 'a common feature of human cancer' and sensitivity to its inhibition across leukaemia and lymphoma lines tracks p53 status, not genotype or lineage [E1 counter, PMID 22789538]. A KMT2A-r line that is more sensitive than average is explained by p53 and proliferation before the fusion is invoked. (ii) *The one ribosome phenotype actually reported in MLLr cells is drug-imposed and p53-dependent.* WDR5 WIN-site inhibitors suppress ribosome protein gene transcription, induce nucleolar stress and activate p53 in MLLr cells - and the authors explicitly withdrew the earlier MLL1/HSC-programme explanation of WINi action in favour of it [E2 counter, PMID 38682900]. That is a ribosome dependency created by a drug, not one created by a fusion. (iii) *The probe is broken.* CX-5461, named in this node's original prediction, kills primarily by topoisomerase II poisoning rather than Pol I inhibition, and its use as a chemical-genetic probe of Pol I is unreliable [E3 counter, PMID 32041867]. The hypothesis survives as a genuine gap, but its prior has been cut and its diagnostic experiment rewritten.

Go deeper
Why the empty cell is worth recording rather than quietly skipped1 deeper

Two reasons, and one large caveat added in cycle 35.

First, an untested dependency class is exactly where a therapy for a mutation-poor cancer might be found. Infant KMT2A-r ALL has almost no co-mutations to target (see comut-low-mutation-burden-infant-kmt2a-r), so dependencies must come from cell state rather than genotype. Translational capacity is a cell-state dependency with existing chemical matter: eIF4A inhibitors (zotatifin class), and the METTL3 inhibitor STM2457 already discussed in genome-m6a-mettl3-ythdc1-ctcf-tad-kmt2a-r.

Second, a plain statement of absence protects the project from a plausible-sounding false memory. It is easy to assume that a locus as heavily studied as KMT2A has had its nucleolar and translational biology characterised. It has not, and the graph should say so.

The caveat, which is large. 'Nobody has measured it here' and 'it would mean something if they did' are different claims, and only the first is safe. Elevated ribosome biogenesis is a common feature of human cancer, and the selectivity of Pol I inhibition in leukaemia and lymphoma lines is governed by p53 status rather than by genotype [E1 counter-evidence]. A positive result here - KMT2A::AFF1 lines sitting in the more-dependent tail - is the expected result for any fast-proliferating, p53-wild-type blast, and infant KMT2A-r ALL is characteristically p53-wild-type. Every experiment in this node therefore needs a proliferation-matched and p53-isogenic comparator, or it will produce a true finding with no fusion-specific content.

And the CX-5461 line has been removed. This node previously proposed low-dose CX-5461 as a nucleolar-stress probe. CX-5461's primary cytotoxic mechanism is topoisomerase II poisoning, not Pol I inhibition, and CX-5461-resistant cells are cross-resistant to doxorubicin [E3 counter-evidence]. Using it in a leukaemia defined by a TOP2-associated translocation, and reading the result as nucleolar stress, would have been a double error.

The most concrete sub-prediction

HOXA9 and MEIS1 are transcription factors; BCL2 and MYC are the outputs that actually keep the cell alive, and both are m6A-dependent at the translational level in a KMT2A-fusion cell line 2. If the leukemic phenotype is partly translationally gated, then (a) polysome or Ribo-seq profiling of KMT2A::AFF1 cells should show a distinct translationally regulated module, and (b) blocking translation initiation should phenocopy part of the fusion-knockdown phenotype. Neither has been reported.

The discriminating version of (b), added cycle 35, is genetic rather than pharmacological: a degron on POLR1A or RRN3 in isogenic KMT2A::AFF1 and KMT2A-wild-type B-ALL lines, matched for doubling time and p53 status. That separates nucleolar stress from the off-target activities that have made every small-molecule 'Pol I inhibitor' hard to interpret, and it is the only design that could return an answer this project should believe.

Honest grading
  • Direct KMT2A-r evidence on ribosome biogenesis, nucleolar function, tRNA pools or tRNA modification: one study, and it is a drug-response study (corrected 2026-09-19). Howard et al. show WDR5 WIN-site inhibitors deplete the ribosome stock, suppress ribosome protein gene transcription, induce nucleolar stress, activate p53 and impose a broad modest translational choke in MLLr cells [E2 counter, PMID 38682900]. Baseline measurement in KMT2A-r cells, and any measurement of tRNA pools or tRNA modification in this disease: still none found.
  • Indirect evidence that ribosome biology drives myeloid disease: established 1.
  • Indirect evidence that translational control matters in a KMT2A-fusion cell line: strong, but as a by-product of an m6A study rather than a translation study 2.
  • tRNA-modification oncogenesis: strong in prostate cancer, unstudied in leukemia by that group 3.

A competent first pass costs almost nothing: DepMap dependency profiles for ribosome-biogenesis genes, nucleolar proteins, tRNA synthetases and tRNA-modifying enzymes, restricted to KMT2A::AFF1 lines versus other leukemia lines. That is ana-noncoding-gap-ledger-kmt2a-r's first job.

Facts and figures
statementRibosome biogenesis, nucleolar function, tRNA abundance and tRNA modification have never been measured at baseline in KMT2A-rearranged leukemia, and tRNA pools and modifications have not been measured in it at all; the one study that measures ribosome biogenesis and global translation in MLL-rearranged cells (PMID 38682900) measures them as the response to WDR5 WIN-site inhibition and reads the phenotype as drug-imposed and p53-dependent. Because the fusion drives a high-output transcriptional programme whose survival effectors (MYC, BCL2) are translationally gated, translational capacity remains a plausible unmeasured cell-state dependency - but the hypothesis is only interesting if the dependency exceeds what proliferation rate and p53 status already predict.
originliterature-gap
prior probability0.12
tool compound warningCX-5461 must not be used as a nucleolar-stress or Pol I probe in this project: its primary cytotoxic mechanism is topoisomerase II poisoning (PMID 32041867), which in a TOP2-associated translocation disease is also a specific hazard. Use a POLR1A/RRN3 degron instead.
planned experiments noteMoved out of a reference field on 2026-09-19. It named a node that does not exist, which made the pointer read as a real cross-reference and, for hypotheses, silently satisfied the 'needs at least one experiment or analysis' rule. The intent is preserved here; the node still needs writing.
predictions
[
  "KMT2A::AFF1 lines sit in the more-dependent tail for ribosome-biogenesis and tRNA-modification genes in DepMap RELATIVE TO proliferation-matched and p53-status-matched leukemia lines, not relative to the leukemia average",
  "Ribo-seq or polysome profiling of KMT2A::AFF1 cells identifies a translationally regulated module distinct from the transcriptional signature and not shared with matched non-KMT2A B-ALL lines",
  "A POLR1A or RRN3 degron (not CX-5461) is more toxic to KMT2A::AFF1 cells than to isogenic p53-matched controls, and the effect survives p53 knockout",
  "Global translation rate falls on menin inhibition, indicating that the transcriptional and translational programmes are coupled rather than independent"
]
analyses
[
  "ana-noncoding-gap-ledger-kmt2a-r"
]
falsifiers
[
  "KMT2A::AFF1 lines show no selective dependency on ribosome-biogenesis, nucleolar or tRNA-modification genes in DepMap relative to proliferation- and p53-matched leukemia lines",
  "Any apparent nucleolar-stress sensitivity disappears in a p53-null isogenic background, showing the effect is p53-gated and generic rather than fusion-driven",
  "Ribo-seq shows translational output in KMT2A::AFF1 cells is fully explained by transcript abundance",
  "Global translation is unchanged by menin inhibition"
]
evidence grade by class
{
  "direct_kmt2a_r": "none",
  "indirect_haematopoietic_ribosome_disease": "established",
  "translational_control_in_a_kmt2a_fusion_line": "strong",
  "trna_modification_oncogenesis": "moderate, non-haematopoietic",
  "fusion_specificity": "unsupported; elevated ribosome biogenesis is a common feature of human cancer and Pol I inhibitor sensitivity tracks p53 status (PMID 22789538)",
  "ribosome_phenotype_reported_in_mllr_cells": "drug-imposed and p53-dependent (WDR5 WIN-site inhibition, PMID 38682900), not fusion-driven"
}
prior probability history
[
  {
    "cycle": 1,
    "value": 0.25,
    "basis": "clear literature gap plus strong indirect haematopoietic and translational evidence"
  },
  {
    "cycle": 35,
    "value": 0.12,
    "basis": "downgraded: ribosome biogenesis is a general cancer property with p53-gated drug sensitivity (PMID 22789538); the only ribosome phenotype reported in MLLr cells is drug-imposed and p53-dependent, and replaced an earlier MLL1-biology explanation (PMID 38682900); and the tool compound named in the original prediction is a topoisomerase II poison (PMID 32041867). The gap is real; the fusion-specific dependency is not supported."
  }
]
corrections
[
  {
    "cycle": 35,
    "date": "2026-09-19",
    "withdrawn": "the prediction 'Low-dose nucleolar stress (CX-5461) or eIF4A inhibition phenocopies part of the fusion-knockdown differentiation phenotype', and the framing of DepMap dependency relative to 'other leukemia lines' without proliferation or p53 matching",
    "replaced_with": "a POLR1A/RRN3 degron in p53- and doubling-time-matched isogenic lines; DepMap comparisons explicitly matched for proliferation and p53 status",
    "source": "PMID 32041867 (CX-5461 is a topoisomerase II poison); PMID 22789538 (Pol I inhibitor sensitivity tracks p53)"
  }
]
planned experiments
[
  "exp-noncoding-dependency-crispr-screen-kmt2a-aff1"
]
withdrawn claims
[
  {
    "claim": "There is no published study of ribosome biogenesis, nucleolar function, tRNA abundance, tRNA modification or global translation rate in KMT2A-rearranged leukemia.",
    "withdrawn": "2026-09-19",
    "refuted_by": [
      "38682900"
    ],
    "note": "the refuting paper was already cited in this same entry as a counter-item; the lead sentence, the lay summary, the honest-grading section and data.statement had not been updated with it"
  }
]
How this could bite us back
  • Ribosome biogenesis and translation are the definition of an essential function; inhibiting them in a growing infant is the highest essential-function-toxicity risk in this entire beat.
  • Ribosomal protein haploinsufficiency causes bone marrow failure syndromes, so pharmacological mimicry risks precisely the marrow failure these patients cannot survive.
  • CX-5461 is a topoisomerase II poison rather than a clean Pol I inhibitor and confers collateral doxorubicin resistance; in a leukaemia defined by a TOP2-associated translocation, using it at all - let alone reading its effect as nucleolar stress - is hard to justify.
  • If the dependency is generic to proliferation rather than fusion-specific, any observed efficacy in cell lines will not survive a normal-progenitor comparison.
  • If the ribosome route works only in p53-wild-type cells, it is a strategy that selects for p53 loss, and p53 loss in KMT2A-r relapse is already associated with antigen escape and worse outcomes.
Still open
  1. Is there truly no ribosome or tRNA study in KMT2A-rearranged leukemia? This claim of absence is the load-bearing part of the node and should be rechecked each cycle, including preprints.
  2. Are ribosomal protein genes or nucleolar genes among the differentially expressed set in KMT2A::AFF1 versus normal pro-B progenitors in existing RNA-seq? A free first look.
  3. If translational capacity is a dependency, is it fusion-specific or a general property of any fast-proliferating blast? Cycle 35 moved this from an open question to the default expectation: elevated ribosome biogenesis is a common feature of human cancer and Pol I inhibitor sensitivity tracks p53 status, so the burden of proof is now on demonstrating fusion-specificity against proliferation- and p53-matched controls.
  4. Do infants, with fetal-type haematopoiesis and different ribosome biogenesis rates, have a different therapeutic window for translation-directed agents than adults?
  5. Does the WDR5 WIN-site inhibitor result (ribosome protein gene suppression, nucleolar stress, p53 activation in MLLr cells) reflect anything the fusion does, or only what the drug does? A WINi time course in p53-null isogenic KMT2A::AFF1 cells would separate them, and would also tell the menin-inhibitor programme whether WDR5 and menin targeting are mechanistically independent or both funnel into p53.
Every source behind this entry (3)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Partial loss of function of the ribosomal protein RPS14 phenocopies the 5q- syndrome erythroid differentiation defect in normal human haematopoietic progenitors, forced RPS14 expression rescues the phenotype in patient bone marrow cells, and RPS14-deficient cells show a pre-rRNA processing block functionally equivalent to Diamond-Blackfan anaemia.

    Identification of RPS14 as a 5q- syndrome gene by RNA interference screen · Ebert BL, Pretz J, Bosco J, Chang CY, Tamayo P, Galili N, Raza A, Root DE, Attar E, Ellis SR, Golub TR · Nature 2008
    PMID 18202658DOI
    What this evidence cannot show

    Myelodysplastic syndrome, not acute leukemia, and erythroid rather than lymphoid lineage; establishes that ribosome biology is disease-relevant in haematopoiesis but says nothing about KMT2A fusions.

    Evidence that argues against this (1)
    • Ribosome biogenesis is not an untouched niche waiting to be a KMT2A-specific dependency; it is a general property of cancer. Bywater et al. state that increased rDNA transcription by RNA Polymerase I 'is a common feature of human cancer', and show that CX-5461 selectively kills B-lymphoma cells in vivo while sparing wild-type B cells, with the therapeutic effect a consequence of nucleolar disruption and p53-dependent apoptotic signalling. Crucially, sensitivity across human leukaemia and lymphoma lines tracks **p53 mutational status**, not any fusion or lineage. So any differential sensitivity of KMT2A::AFF1 lines to nucleolar stress is predicted by their p53 status and proliferation rate before the fusion is invoked at all, and a p53-isogenic control is mandatory. Bywater MJ et al. 2012
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In the KMT2A::MLLT3 cell line MOLM-13, m6A promotes the translation of MYC, BCL2 and PTEN mRNAs, and METTL3 loss produces a translational defect accompanied by differentiation and apoptosis, showing that translational control is an operative layer in a KMT2A-fusion leukemia cell.

    The N6-methyladenosine (m6A)-forming enzyme METTL3 controls myeloid differentiation of normal hematopoietic and leukemia cells · Vu LP, Pickering BF, Cheng Y, Zaccara S, Nguyen D, Minuesa G, Chou T, Chow A, Saletore Y, MacKay M, Schulman J, Famulare C, Patel M, Klimek VM, Garrett-Bakelman FE, Melnick A, Carroll M, Mason CE, Jaffrey SR, Kharas MG · Nature Medicine 2017
    PMID 28920958DOI
    What this evidence cannot show

    One cell line; the study is about m6A and myeloid differentiation, not about ribosome or tRNA biology, and makes no claim of KMT2A-fusion specificity. B-lineage KMT2A::AFF1 was not examined.

    Evidence that argues against this (1)
    • This is simultaneously the best reason to look at ribosome biology in MLL-rearranged cells and the clearest warning against the causal story this node tells. WDR5 WIN-site inhibitors were originally proposed to work in MLLr cells through dysregulation of a haematopoietic-stem-cell expansion programme; on interrogation with small-molecule WINi the authors found instead that they suppress ribosome protein gene transcription, induce nucleolar stress and activate p53, depleting the ribosome stock with a broad but modest translational choke and an MDM4-inactivating splicing change. Two consequences. (i) The MLL1-biology explanation was withdrawn by the people who proposed it - a direct precedent for a ribosome phenotype in MLLr cells that is not fusion biology. (ii) The ribosome involvement here is *downstream of a drug* and *p53-dependent*; that is not the same as a fusion-driven dependency, and this node must not use it as evidence of one. Howard GC et al. 2024
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The tRNA N7-methylguanosine transferase METTL1 is highly expressed in tumours and its depletion causes loss of m7G tRNA methylation and generates 5'-tRNA-derived small RNAs that redirect translation toward tumour-growth-suppressive and interferon programmes, with Mettl1 knockdown increasing pro-inflammatory immune infiltration and immunotherapy response in preclinical models.

    METTL1 promotes tumorigenesis through tRNA-derived fragment biogenesis in prostate cancer · García-Vílchez R, Añazco-Guenkova AM, Dietmann S, López J, Morón-Calvente V, D'Ambrosi S, Nombela P, Zamacola K, Mendizabal I, García-Longarte S, Zabala-Letona A, Astobiza I, Fernández S, Paniagua A, Miguel-López B, Marchand V, Alonso-López D, Merkel A, García-Tuñón I, Ugalde-Olano A, Loizaga-Iriarte A, Lacasa-Viscasillas I, Unda M, Azkargorta M, Elortza F, Bárcena L, Gonzalez-Lopez M, Aransay AM, Di Domenico T, Sánchez-Martín MA, De Las Rivas J, Guil S, Motorin Y, Helm M, Pandolfi PP, Carracedo A, Blanco S · Molecular Cancer 2023
    PMID 37516825DOI
    What this evidence cannot show

    Prostate cancer, not leukemia; cited only to show that tRNA modification is a real oncogenic axis with an immune consequence. Any transfer to KMT2A-r is entirely speculative.

    Evidence that argues against this (1)
    • The tool compound this node's own experimental plan named does not do what the plan assumed. CX-5461 was developed as a selective RNA Pol I inhibitor, but Bruno et al. show by three data-mining modalities plus biochemical and cell-biological assays that its primary cytotoxic activity is topoisomerase II poisoning, and that CX-5461-resistant lymphoma cells acquire collateral resistance to doxorubicin. The authors state explicitly that CX-5461's use as a specific chemical-genetic probe of RNA Pol I function is challenging to interpret. The prediction 'low-dose nucleolar stress (CX-5461) phenocopies part of the fusion-knockdown phenotype' was therefore untestable as written and has been replaced; and in a leukaemia whose defining lesion is a TOP2-associated translocation, giving a topoisomerase II poison labelled as a nucleolar probe is a specific hazard. Bruno PM et al. 2020
Connected to