Oncofusion

start / graph / gene-prmt1

Proteinstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q5c-interactome2

PRMT1 — an MLL-fusion cofactor whose first-in-human inhibitor was terminated for lack of efficacy and toxicity

also called PRMT1 · protein arginine N-methyltransferase 1 · HRMT1L2 · ANM1

PRMT1 tags proteins with methyl groups and is an essential part of one particular MLL fusion's machinery. A drug against it reached patients, and the trial was stopped early because it did not work and caused too many problems.

The same thing, for a scientist

PRMT1 (Q99873, 371 aa) is the dominant type I protein arginine methyltransferase. It was shown to be a necessary but not sufficient component of an MLL-EEN oncogenic complex recruited through the adaptor Sam68, and is positioned downstream of the PAFc/HOXA9/beta-catenin axis in later work. It is DepMap common-essential (-1.40, 98.7% of lines depleted) and strongly constrained (LOEUF 0.222). The first-in-human type I PRMT inhibitor GSK3368715 (NCT03666988) was terminated during dose escalation for lack of efficacy and dose-limiting toxicity.

What it looks like
PRMT1371 aa
SAM-dependent methyltransferase, PRMT-typeSAM-dependent methyltransferase, PRMT-type · residues 50–3610100200300
loading AF-Q99873-F1…
AlphaFold AF-Q99873-F1 ↗
colourshape

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

very high (90+)confident (70–90)low (50–70)very low (<50)
I have never read one of these pictures before

A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.

Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.

Anything in yellow sticks is a bound small molecule, which is usually the drug.

Go deeper
Correction: 'essential component' is qualified to 'necessary but not sufficient'

This entry asserted that PRMT1 is an essential component of the MLL-EEN oncogenic complex, while a counter-evidence item in the same file — from the same group that produced the original claim — records that PRMT1 is necessary but not sufficient for transformation, which additionally requires co-recruitment of the H3K9 demethylase KDM4C.

The wording is aligned to the evidence throughout. The practical consequence is the one the counter-evidence already stated and the body did not: a PRMT1-only intervention should be expected to be partially compensated, because the requirement is joint rather than singular.

This is consistent with the cycle-22 strike on PRMT1, which rests on Prmt1 knockdown leaving transformation by most MLL fusions unaffected — and it is a different point from the one I originally relayed for that strike, which was itself wrong.

The biology, with an important boundary

PRMT1 is a necessary but not sufficient component of an MLL-EEN oncogenic transcriptional complex, recruited through the bridging adaptor Sam68 (KHDRBS1). Direct MLL–PRMT1 and MLL–Sam68 fusions enhance self-renewal, and knocking down either PRMT1 or Sam68 suppresses MLL-mediated transformation 1.

The boundary matters. This is MLL-EEN, a rare fusion, not KMT2A::AFF1. The generalisation from 'PRMT1 is required by MLL-EEN' to 'PRMT1 is a KMT2A::AFF1 interactor' has not been demonstrated and should not be asserted. Later work from the same group places Prmt1 as a common downstream mediator of the beta-catenin/Hoxa9 axis in MLL cancer stem cells 2, which supports relevance but is a downstream-mediator claim rather than a complex-membership one.

A related caution for anyone reading this literature: the PAF complex regulates PRMT5, a different enzyme, and the two are routinely conflated because they sit in the same body of work 3.

The clinical result

GSK3368715 was the first type I PRMT inhibitor to enter the clinic, in a phase 1 first-in-human study in solid tumours and DLBCL. The registry record for NCT03666988 shows the study terminated on 2021-03-04 during Part 1 dose escalation with 31 participants enrolled; the Part 2 expansion never ran and the planned response and pharmacokinetic analyses were not performed 4. A later methodology paper characterises the failure plainly: terminated early for lack of clinical efficacy, extensive treatment-emergent effects and dose-limiting toxicities 5.

Two caveats before this is read as a refutation of PRMT1 biology. The trial was in solid tumours and DLBCL, not leukemia, and GSK3368715 inhibits type I PRMTs as a class rather than PRMT1 selectively. So the result establishes that this compound at these doses failed in that population; it does not establish that PRMT1 is unimportant in MLL-fusion leukemia. It does establish that the tolerability problem is real and was encountered in humans.

Why tolerability was the predictable problem

PRMT1 is DepMap common-essential: mean Chronos -1.399, leukemia -1.383, depleted in 98.7% of 1,258 lines — one of the most uniformly required genes in this entire graph 6. gnomAD: pLI ~1.0, LOEUF 0.222, 5 observed loss-of-function variants against 47.4 expected 6.

And there is a specific haematopoietic liability. Conditional Prmt1 deletion (Mx1-Cre) causes anaemia and leukopenia, reduced terminal erythroid and lymphoid differentiation, reduced short-term HSC and granulocyte-macrophage progenitor frequencies, and impaired HSPC self-renewal 7. For a paediatric leukemia programme, that is the wrong compartment to damage.

Expression is high everywhere — bone marrow 61.3, naive B cell 172.5, heart muscle 139.6 nTPM 6 — which offers no selectivity at all.

Leukemia-restricted correlation with MEN1 is r = 0.370 (95% CI 0.058–0.610), failing Holm correction 8. As with SETD1A, this is a pan-essential gene producing a nominal signal, which is worth noting methodologically and acting on not at all.

Facts and figures
uniprotQ99873
gene symbolPRMT1
ensemblENSG00000126457
length AA371
alphafold IDAF-Q99873-F1
PDB note6NT2 is a type I PRMT in complex with the clinical inhibitor GSK3368715, 2.48 A; 9BH4 is a PRMT1 hexamer by cryo-EM at 2.55 A.
roleessential cofactor of the MLL-EEN oncogenic complex via Sam68; downstream mediator of the beta-catenin/HOXA9 axis in MLL cancer stem cells. Not demonstrated as a KMT2A::AFF1 complex subunit.
druggabilityChemically tractable and clinically tested. A co-crystal with the clinical inhibitor exists (6NT2). The result was a terminated phase 1. Degrader approaches are being explored. No ChEMBL count quoted.
subnuclear localizationNONE ASSIGNED - a bounded absence. UniProt Q99873 carries no subnuclear qualifier beyond 'Nucleus, nucleoplasm' (which is mouse-inferred, ECO:0000250 Q9JIF0) and the partner-based phrase 'Colocalizes with CHTOP within the nucleus'. Bounded 2026-09-21: PRMT1[tiab] AND (nucleolar[tiab] OR "nuclear speckles"[tiab] OR subnuclear[tiab]) returns 8 records, and those examined are about PRMT-mediated methylation OF fibrillarin and of nucleolar sub-complexes - the enzyme's substrates being nucleolar rather than the enzyme being nucleolar. A methyltransferase whose substrate lives in a compartment need not live there, and this entry keeps the distinction.
localization shown for fusionNO for PRMT1's own compartment, in any KMT2A fusion, in any lineage. The bounded query PRMT1[tiab] AND (MLL OR KMT2A)[tiab] AND (localization OR localisation OR nuclear OR cytoplasm)[tiab] returns ONE record on 2026-09-21, and it is PMID 40788108 - in which the protein whose localisation is measured is ATG4B, not PRMT1, in a KMT2A::MLLT3 (MLL-AF9) mouse model and AML PDX [E_q89_prmt1_loc2]. So the single hit in the intersection of PRMT1, KMT2A and localisation is about somebody else's translocation, and PRMT1's own compartment in a fusion-bearing cell has never been measured.
PDB IDS
[
  "6NT2",
  "8Z2Z",
  "8Z7H",
  "8Z7O",
  "9BH4",
  "9BHD",
  "9BHG",
  "9JP0"
]
PDB verified
[
  "6NT2",
  "9BH4"
]
domains
[
  {
    "name": "SAM-dependent methyltransferase, PRMT-type",
    "start": 50,
    "end": 361,
    "pfam": null
  }
]
clinical
{
  "compound": "GSK3368715 (type I PRMT inhibitor)",
  "nct": "NCT03666988",
  "status": "TERMINATED 2021-03-04 during Part 1 dose escalation, n = 31",
  "population": "solid tumours and DLBCL, not leukemia",
  "reason": "lack of clinical efficacy, extensive treatment-emergent effects and dose-limiting toxicities",
  "co_crystal": "6NT2"
}
essentiality
{
  "depmap_common_essential": true,
  "chronos_mean": -1.3985,
  "chronos_mean_leukemia_lines": -1.3829,
  "n_lines": 1258,
  "fraction_lines_depleted": 0.9868,
  "gnomad_pLI": 1,
  "gnomad_loeuf": 0.222,
  "gnomad_obs_lof": 5,
  "gnomad_exp_lof": 47.4,
  "mouse_ko": "Conditional deletion in adult mouse haematopoiesis causes anaemia, leukopenia, reduced terminal erythroid and lymphoid differentiation, and impaired HSPC self-renewal",
  "hpa_ntpm": {
    "naive_B_cell": 172.5,
    "heart_muscle": 139.6,
    "lymph_node": 87.9,
    "spleen": 86.8,
    "classical_monocyte": 77,
    "testis": 75.2,
    "thymus": 74.2,
    "bone_marrow": 61.3,
    "liver": 57.8
  },
  "hpa_specificity": "Low tissue specificity"
}
codependency men1
{
  "r_leukemia_lines": 0.3696,
  "ci_leukemia": [
    0.0579,
    0.6102
  ],
  "crosses_zero": false,
  "p_permutation": 0.0054,
  "p_holm": 0.0756,
  "survives_holm": false
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -1.6428,
  "chronos_mean_other_leukaemia": -1.5626,
  "difference": -0.0802,
  "pan_essentiality_all_lines": -1.3985,
  "selectivity_rank": 1733,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
  "how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}
localization
{
  "subcellular": "MOSTLY CYTOPLASMIC, WHICH IS NOT WHAT THIS GRAPH HAS BEEN ASSUMING. UniProt Q99873 fetched 2026-09-21 carries FIVE location values: 'Nucleus' (ECO:0000269 PubMed:10749851, 11387442, 16879614), 'Nucleus, nucleoplasm' (ECO:0000250 BY SIMILARITY TO MOUSE Prmt1 Q9JIF0), 'Cytoplasm' (ECO:0000269 PubMed:26876602), 'Cytoplasm, cytosol' (ECO:0000250, mouse again) and 'LYSOSOME MEMBRANE' (ECO:0000269 PubMed:38006878). Its note opens with four words that matter here: 'MOSTLY FOUND IN THE CYTOPLASM.' It continues: 'Colocalizes with CHTOP within the nucleus. Low levels detected also in the chromatin fraction (By similarity). Upon METHIONINE STIMULATION, localizes to the lysosome membrane in an NPRL2-dependent manner.' [E_q89_prmt1_loc1]",
  "why_that_matters_for_this_entry": "This entry describes PRMT1 as a component of an MLL-EEN oncogenic transcriptional complex and as a downstream mediator of the beta-catenin/HOXA9 axis - both nuclear, chromatin jobs. UniProt says most of the protein is cytoplasmic and that its chromatin presence is at LOW LEVELS and BY SIMILARITY TO MOUSE. Those statements are compatible - an enzyme can do an important nuclear job with a small nuclear pool - but the entry should not leave a reader assuming PRMT1 is a predominantly nuclear protein, because it is not.",
  "subnuclear": "NO COMPARTMENT ASSIGNED, and the one nucleolar thread in the literature is about the SUBSTRATE, not the enzyme. The bounded query PRMT1[tiab] AND (nucleolar OR \"nuclear speckles\" OR subnuclear)[tiab] returns 8 records on 2026-09-21, and the older ones are about arginine methylation OF FIBRILLARIN - a nucleolar protein - by protein arginine methyltransferases. PRMT1 methylating a nucleolar substrate is not PRMT1 being nucleolar, and this entry does not record it as such. UniProt's only within-nucleus qualifier is 'Colocalizes with CHTOP within the nucleus', which names a partner rather than a compartment.",
  "THE_LYSOSOME_ROW_IS_THE_INTERESTING_ONE": "PRMT1 relocalises to the LYSOSOME MEMBRANE on METHIONINE STIMULATION, NPRL2-dependently (ECO:0000269 PubMed:38006878). NPRL2 is a GATOR1 component - this is mTORC1 nutrient sensing. PRMT1 is a SAM-dependent methyltransferase and methionine is the precursor of SAM, so the enzyme is being moved by the availability of its own cosubstrate. This graph already holds cell-state-sam-methionine-dependence, and this is a direct, previously unrecorded mechanical link between that state and a chromatin enzyme it names."
}
cell type
{
  "normal": "WIDELY EXPRESSED (UniProt tissue specificity, PubMed:11097842 cited), with the additional uncoded statement that PRMT1 is strongly expressed in colorectal cancer cells and tissues at protein level. No haematopoietic-specific statement in the record [E_q89_prmt1_loc1].",
  "haematopoietic_requirement": "Conditional Prmt1 deletion in ADULT MOUSE haematopoiesis causes anaemia and leukopenia with reduced terminal erythroid and lymphoid differentiation (this entry's existing E7) - so the cell types that need PRMT1 are erythroid and lymphoid progenitors, and that is also the toxicity that killed the clinical programme.",
  "in_this_disease": "MOUSE AML INDUCED BY MLLT3-KMT2A OVEREXPRESSION, AND AML PATIENT-DERIVED CELLS AND XENOGRAFTS. ATG4B-mediated DNA-repair defects are significantly enhanced in patient-derived AML cells and in mouse AML cells induced by MLLT3-KMT2A overexpression; inhibiting ATG4B enhanced PRMT1-mediated DNA damage responses, suppressed proliferation, reduced mutation burden and prolonged survival in MLLT3-KMT2A mice and in AML PDX [E_q89_prmt1_loc2]. NAMING THE FUSION: MLLT3-KMT2A is KMT2A::MLLT3 / MLL-AF9, myeloid - NOT the index fusion KMT2A::AFF1. This entry's other fusion evidence is MLL-EEN, which is a third fusion again."
}
cell state
{
  "ENERGY_DEFICIENCY": "A newly recorded state, and it acts through localisation - of the partner, not of PRMT1. During energy deficiency ATG4B TRANSLOCATES FROM THE CYTOPLASM TO THE NUCLEUS and disrupts DNA repair by directly interacting with PRMT1, inhibiting PRMT1-dependent methylation of MRE11 and causing genomic instability [E_q89_prmt1_loc2]. Stated precisely because the direction matters: the protein that moves is ATG4B. Nobody reports PRMT1 moving in this system.",
  "METHIONINE_AVAILABILITY": "See cell-state-sam-methionine-dependence. Methionine stimulation relocalises PRMT1 to the lysosome membrane NPRL2-dependently [E_q89_prmt1_loc1] - a nutrient state that moves the enzyme itself, and the only state in this entry that does.",
  "PAN_ESSENTIALITY": "This entry already records PRMT1 as DepMap common-essential at mean Chronos -1.399 with 98.7% of lines depleted, i.e. floor-compressed: almost no dynamic range in which a KMT2A-selective state-dependence could appear."
}
localization bounded absence
{
  "date": "2026-09-21",
  "tool": "NCBI esearch db=pubmed",
  "queries": [
    {
      "query": "PRMT1[tiab] AND (nucleolar[tiab] OR \"nuclear speckles\"[tiab] OR subnuclear[tiab])",
      "querytranslation": "\"PRMT1\"[Title/Abstract] AND (\"nucleolar\"[Title/Abstract] OR \"nuclear speckles\"[Title/Abstract] OR \"subnuclear\"[Title/Abstract])",
      "count": 8,
      "on_topic_for_PRMT1s_own_compartment": 0,
      "note": "Examined titles are substrate papers - arginine methylation of fibrillarin (12523648), a fibrillarin sub-complex containing protein arginine methyltransferases (14583623), hCAF1 as a regulator of PRMT1-dependent methylation (17264152). Substrate location is not enzyme location."
    },
    {
      "query": "PRMT1[tiab] AND (MLL[tiab] OR KMT2A[tiab]) AND (localization[tiab] OR localisation[tiab] OR nuclear[tiab] OR cytoplasm[tiab])",
      "querytranslation": "\"PRMT1\"[Title/Abstract] AND (\"MLL\"[Title/Abstract] OR \"KMT2A\"[Title/Abstract]) AND (\"localization\"[Title/Abstract] OR \"localisation\"[Title/Abstract] OR \"nuclear\"[Title/Abstract] OR \"cytoplasm\"[Title/Abstract])",
      "count": 1,
      "on_topic_for_PRMT1s_own_compartment": 0,
      "the_one": "40788108, in which ATG4B translocates and PRMT1 does not"
    },
    {
      "query": "PRMT1[tiab] AND leukemia[tiab] AND (nuclear[tiab] OR cytoplasmic[tiab] OR localization[tiab])",
      "querytranslation": "\"PRMT1\"[Title/Abstract] AND \"leukemia\"[Title/Abstract] AND (\"nuclear\"[Title/Abstract] OR \"cytoplasmic\"[Title/Abstract] OR \"localization\"[Title/Abstract])",
      "count": 4,
      "role": "operand widening - dropping the MLL clause quadruples the count, confirming the single hit above is a scoping effect and not a dead query"
    }
  ],
  "absence_class": "never-measured"
}
How this could bite us back
  • Essential-function toxicity: required by 98.7% of cell lines, with a conditional mouse knockout that produces anaemia, leukopenia and impaired HSPC self-renewal.
  • The clinical experiment has been run: the first type I PRMT inhibitor in humans was terminated during dose escalation for lack of efficacy and dose-limiting toxicity.
  • Evidence transfer risk: the MLL-fusion case for PRMT1 was built on MLL-EEN and is routinely quoted as if it covered all KMT2A fusions.
  • Literature hazard: PRMT1 and PRMT5 are separate enzymes with overlapping literatures in MLL leukemia, and at least one widely cited paper in this space is about PRMT5.
Still open
  1. Does the MLL-EEN/PRMT1/Sam68 mechanism apply to KMT2A::AFF1 at all? The entire PRMT1 case in this disease rests on a rare fusion, and nobody has tested the common one.
  2. Would a PRMT1-selective agent, rather than a pan-type-I inhibitor, have a different tolerability profile? GSK3368715 inhibits the class, and the haematopoietic toxicity may not be PRMT1's alone.
  3. Is a PRMT1 degrader any more plausible than an inhibitor given 98.7% pan-essentiality? Degradation is usually more, not less, severe than occupancy for an enzyme with scaffolding roles.
  4. What fraction of PRMT1 is nuclear in a KMT2A-rearranged blast? UniProt says most of the protein is cytoplasmic and that its chromatin presence is mouse-inferred and low, yet this graph's PRMT1 argument is entirely a chromatin argument. One fractionation would settle whether the premise holds in the relevant cell.
  5. Methionine stimulation moves PRMT1 to the lysosome membrane NPRL2-dependently, and this graph holds cell-state-sam-methionine-dependence. Does methionine restriction - already of interest here for SAM supply - also change PRMT1's compartment, and would that confound any interpretation of methionine restriction as acting purely through SAM levels?
  6. ATG4B inhibition restores PRMT1 activity and prolongs survival in MLL-AF9 mice, while direct PRMT1 inhibition failed in the clinic. Are these compatible, and does the answer depend on which PRMT1 pool each agent reaches?
Every source behind this entry (10)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    PRMT1 is an ESSENTIAL component (the paper's own word) of an MLL-EEN oncogenic transcriptional complex recruited via the adaptor Sam68; direct MLL-PRMT1 and MLL-Sam68 fusions enhance self-renewal, and knockdown of PRMT1 or Sam68 suppresses MLL-mediated transformation.

    Protein arginine-methyltransferase-dependent oncogenesis. · Cheung N, Chan LC, Thompson A, Cleary ML, So CWE · Nature Cell Biology 2007
    PMID 17891136DOI
    What this evidence cannot show

    The fusion studied is MLL-EEN, a rare partner, not KMT2A::AFF1. Transfer of this mechanism to the infant t(4;11) disease is an assumption, not a demonstration. CORRECTED 2026-09-21 (cycle 46): the phrase 'necessary but not sufficient' is NOT the paper's. PMID 17891136's abstract says PRMT1 is an 'essential component' and contains neither the word 'necessary' nor 'sufficient'; there is no PMC record, so the full text was not read. The 'not sufficient' half was an interpolation of the same shape the coordinator committed this cycle with 'sufficient but not necessary' - completing a familiar logical pair and attaching it to a real citation. If anything the abstract cuts the other way: DIRECT FUSION of MLL with PRMT1 or Sam68 enhanced self-renewal, which is a sufficiency observation. The supported claim is 'essential component'.

    Evidence that argues against this (2)
    • The MLL-EEN/Sam68 story is not the mechanism that operates in the common MLL-rearranged leukemias. In MLL-rearranged ALL, PRMT1's oncogenic function was shown to run through methylation of a non-histone substrate, FLT3 at R972/R973, and to be FLT3-methylation dependent: the ability of the type I PRMT inhibitor MS023 to kill leukemia cells paralleled baseline FLT3 R972/973 methylation, and the benefit came from combination with the FLT3 inhibitor PKC412 in patient-derived xenografts. If the relevant PRMT1 substrate in MLL-r disease is FLT3 rather than histone H4R3 inside a fusion complex, then PRMT1 is not a KMT2A::AFF1 complex member at all, it is an upstream modifier of a receptor kinase that is already drugged, and the actionable node is FLT3. Zhu Y et al. 2019
    • From the same group that produced E1, and qualifying it: PRMT1 is necessary but NOT sufficient for leukemic transformation, which additionally requires co-recruitment of the H3K9 demethylase KDM4C by the chimeric transcription factor. On this account PRMT1 is one arm of a two-enzyme circuit rather than a stand-alone requirement, and a PRMT1-only intervention should be expected to be partially compensated. It also means the E1 claim as written ('essential component') overstates what a single-enzyme perturbation buys. Cheung N et al. 2016
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Prmt1 acts as a common downstream mediator of the beta-catenin and Hoxa9 axis in MLL cancer stem cells, with transcriptional memory of the cell of origin overriding the beta-catenin requirement.

    Transcriptional memory of cells of origin overrides beta-catenin requirement of MLL cancer stem cells. · Siriboonpiputtana T, Zeisig BB, Zarowiecki M, Fung TK, Mallardo M, Tsai CT, et al., So CWE · EMBO Journal 2017
    PMID 28978671DOI
    What this evidence cannot show

    Mouse models; positions PRMT1 as a downstream mediator rather than a fusion-complex subunit.

    Evidence that argues against this (1)
    • A common downstream mediator of beta-catenin and Hoxa9 in MLL cancer stem cells is a murine myeloid result, and 'transcriptional memory of the cell of origin overriding the beta-catenin requirement' is the entry's own statement that the requirement is context-dependent. Oncofusion knowledge graph (internal analysis) 2026
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The PAF complex regulates Prmt5 -- a different enzyme from PRMT1 -- to facilitate the progression and maintenance of MLL-fusion leukemia.

    The PAF complex regulation of Prmt5 facilitates the progression and maintenance of MLL fusion leukemia. · Serio J, Ropa J, Chen W, Mysliwski M, Saha N, Chen L, Wang J, Miao H, Cierpicki T, Grembecka J, Muntean AG · Oncogene 2018
    PMID 28945229DOI
    What this evidence cannot show

    Included here specifically as a disambiguation. This paper is about PRMT5 and must not be cited as PRMT1 evidence; the conflation is common in this literature.

    Evidence that argues against this (1)
    • The entry flags it and it bears repeating: this result is about PRMT5, a different enzyme. It belongs in the dossier as a near-miss that has been mistaken for supporting evidence elsewhere in the literature. Oncofusion knowledge graph (internal analysis) 2026
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.trial_registry

    NCT03666988, the first-in-human phase 1 study of GSK3368715 in solid tumours and DLBCL, was terminated on 2021-03-04 during Part 1 dose escalation with 31 participants enrolled; the Part 2 expansion was never conducted and planned response and pharmacokinetic analyses were not performed.

    NCT03666988
    What this evidence cannot show

    A registry termination record states that a study stopped and when, not always why. The population is solid tumours and DLBCL, not leukemia, and GSK3368715 inhibits type I PRMTs as a class rather than PRMT1 selectively.

    Evidence that argues against this (1)
    • A first-in-human PRMT1 inhibitor terminated during dose escalation with 31 participants, in solid tumours and DLBCL, is the strongest available statement about this target's clinical tractability and it is negative. The termination reason should be read from the raw registry record rather than inferred. Journal of Clinical Oncology 2025
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The phase 1 trial of GSK3368715, the first PRMT1 inhibitor to enter the clinic, was terminated early due to a lack of clinical efficacy, extensive treatment-emergent effects and dose-limiting toxicities.

    Towards the Targeted Protein Degradation of PRMT1. · Martin PL, et al. · ChemMedChem 2024
    PMID 38724444DOI
    What this evidence cannot show

    This is a secondary characterisation of the trial outcome in the introduction of a medicinal-chemistry paper, not the trial report itself. It should be read alongside the registry record rather than in place of it.

    Evidence that argues against this (1)
    • The clinical failure does not close the PRMT1 question for MLL-rearranged leukemia, and this is the paper that keeps it open. A different type I PRMT inhibitor, MS023, suppressed MLL-r ALL cell viability in proportion to baseline FLT3 R972/973 methylation, and combining MS023 with PKC412 eliminated MLL-r ALL cells more effectively than PKC412 alone in patient-derived xenografts. GSK3368715 was tested in solid tumours and DLBCL as monotherapy at maximum-tolerated dose; the preclinical MLL-r ALL case is for a biomarker-selected combination at lower exposure. Terminating one compound in the wrong population is weak evidence about a different regimen in the right one. Zhu Y et al. 2019
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    PRMT1 is DepMap common-essential with mean Chronos -1.399 across 1,258 lines and depletion in 98.7% of lines; gnomAD reports pLI ~1.0 and LOEUF 0.222 with 5 observed against 47.4 expected loss-of-function variants; HPA reports high ubiquitous expression including bone marrow 61.3 and naive B cell 172.5 nTPM.

    What this evidence cannot show

    DepMap release version not stated by the API.

    Evidence that argues against this (1)
    • Searched and not found. Targeted PubMed searches for work showing PRMT1 is dispensable in some cell context, for a replication failure of the DepMap common-essential call, or for a PRMT1-null human or conditional-knockout tissue that tolerates loss, returned nothing contradicting the pan-essentiality profile. The published direction of evidence is uniformly consistent: strong human constraint, near-universal cell-line depletion and a haematopoietic conditional-knockout phenotype. Treat the absence of a dissenting paper here as real rather than as a gap. Broad Institute DepMap; gnomAD; Open Targets Platform; search performed by this project 2026
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Conditional Prmt1 deletion in adult mouse haematopoiesis causes anaemia and leukopenia, reduced terminal erythroid and lymphoid differentiation, reduced short-term HSC and granulocyte-macrophage progenitor frequencies, and impaired HSPC self-renewal.

    Protein arginine methyltransferase 1 is required for maintenance of normal adult hematopoiesis. · Zhu L, et al. · International Journal of Biological Sciences 2019
    PMID 31853216DOI
    What this evidence cannot show

    Published in a lower-profile journal than the other citations here. Complete conditional deletion in mouse, which overstates what partial pharmacological inhibition does; nonetheless it is directionally consistent with the treatment-emergent toxicity seen in the human trial.

    Evidence that argues against this (1)
    • This is a mouse knockout describing what the gene does in normal development or normal haematopoiesis, over the animal's whole life, with the protein entirely absent. A drug gives partial reversible inhibition in an adult for weeks. The experiment bounds the biological requirement and not the therapeutic window. Anaemia, leukopenia and reduced HSC and progenitor output on conditional deletion in ADULT haematopoiesis is the closest model to a treated patient and it is the bite-back for this target. Nature 2012
  8. E8weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset

    Across 54 DepMap leukemia lines PRMT1 correlates with MEN1 at r = 0.370 (95% CI 0.058-0.610, permutation p = 0.005), failing Holm correction across 16 genes (Holm p = 0.076).

    Leukemia-restricted co-dependency with MEN1, bootstrap and permutation testing; analyses/wrad-discrimination/ in this project · Broad Institute DepMap; Open Targets Platform; analysis run by this project · Database and derived analysis 2026
    What this evidence cannot show

    Our own computation in 54 lines, failing correction. Reproduces the cycle-9 value of 0.370. Do not cite PRMT1 as a co-dependency.

    Evidence that argues against this (1)
    • Same objection as for the other sub-threshold correlations in this beat: the deeper 2,743-gene variance-matched null with Holm correction leaves only KMT2A, PSIP1, DOT1L, FLT3, MEF2C and KAT7 standing, with MEIS1 already failing at Holm 0.067. LEO1's r = 0.302 at Holm 0.286 in the 16-gene panel is well below the genes that fail in the deep pool, so it should not be described as a signal of any kind. Broad Institute DepMap; Open Targets Platform; analysis run by this project 2026
  9. E_q89_prmt1_loc1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProt Q99873 (ANM1_HUMAN, primary gene PRMT1, synonyms HMT2/HRMT1L2/IR1B4, 371 aa, read back 2026-09-21) carries FIVE subcellular location values - Nucleus (three ECO:0000269), Nucleus/nucleoplasm (ECO:0000250 from MOUSE Q9JIF0), Cytoplasm (ECO:0000269 PubMed:26876602), Cytoplasm/cytosol (ECO:0000250 from mouse) and LYSOSOME MEMBRANE (ECO:0000269 PubMed:38006878) - and its note states 'MOSTLY FOUND IN THE CYTOPLASM', that it colocalises with CHTOP within the nucleus, that low levels are detected in the chromatin fraction BY SIMILARITY, and that upon METHIONINE STIMULATION it localises to the lysosome membrane in an NPRL2-DEPENDENT manner. Tissue specificity: widely expressed, with strong expression in colorectal cancer cells and tissues at protein level.

    What this evidence cannot show

    Statement about the record. Note the evidence asymmetry inside it: the NUCLEOPLASM and CYTOSOL refinements and the chromatin-fraction statement are all ECO:0000250 inferences from the MOUSE orthologue Q9JIF0, so the two most precise compartment statements about human PRMT1 are not human measurements - the same defect Q87 recorded for NR1H3's cytoplasmic pool in tranche 2. 'Mostly found in the cytoplasm' is uncoded and gives no cell type.

    Evidence that argues against this (1)
    • Taking 'mostly cytoplasmic' at face value would be a mistake in the opposite direction, and the entry should not swing there. Bulk partition says nothing about where the functionally decisive molecules are: PRMT1's known nuclear substrates include histone H4R3 and its named nuclear partner CHTOP, and a methyltransferase with high catalytic turnover can maintain a chromatin mark from a small, rapidly cycling nuclear pool. The DepMap result cuts the same way - a protein depleted in 98.7% of lines is doing something indispensable, and the MLL-EEN complex work this entry cites placed it in a chromatin complex directly. So the correct reading is that the CYTOPLASMIC pool is the majority and the NUCLEAR pool is the one this disease is about, and neither has been quantified in a haematopoietic cell, let alone a fusion-bearing one. What should not survive is any sentence implying PRMT1 is a nuclear protein simpliciter. Wang Z et al. 2025
  10. E_q89_prmt1_loc2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    THE ONE PRMT1-PLUS-KMT2A-PLUS-LOCALISATION PAPER IS ABOUT SOMEBODY ELSE MOVING. During energy deficiency ATG4B translocates from the cytoplasm to the nucleus and disrupts DNA repair by directly interacting with PRMT1, inhibiting PRMT1-dependent methylation of MRE11 and causing genomic instability. ATG4B-mediated DNA-repair defects are significantly enhanced in patient-derived acute myeloid leukaemia cells and in MOUSE AML CELLS INDUCED BY MLLT3-KMT2A OVEREXPRESSION. Inhibiting ATG4B enhanced PRMT1-mediated DNA damage responses, suppressed cell proliferation, reduced mutation burden and prolonged survival in mice with MLLT3-KMT2A-induced AML and in AML patient-derived xenografts.

    Energy Deficiency-Induced ATG4B Nuclear Translocation Inhibits PRMT1-Mediated DNA Repair and Promotes Leukemia Progression. · Wang Z, Zhang X, Zhou Y, Lin S, Fu Y, Liu H, Gao Q, Xiao Y, Yin Z, Liu S, Huang K, Deng S, Du J, Zeng H, Wang J, Zheng J, Zhou Q, Chen T, Gao X, Ju Z, Liu B · Adv Sci (Weinh) 2025
    PMID 40788108DOI
    What this evidence cannot show

    The fusion is MLLT3-KMT2A, i.e. KMT2A::MLLT3 / MLL-AF9, MOUSE and MYELOID - NOT the index fusion KMT2A::AFF1 and not B-lineage; stated in the claim per rule 73. The protein whose localisation was measured is ATG4B; PRMT1 is the binding partner and the enzymatic activity, and no PRMT1 compartment measurement is reported. Abstract read in full, full text not read. 2025 and single-laboratory; graded moderate.

    Evidence that argues against this (1)
    • This paper is easy to over-claim for this entry and the over-claim would be specific: it does NOT show that PRMT1 acts in the nucleus of a KMT2A-fusion cell, it shows that a nuclear ATG4B pool can inhibit PRMT1 activity. An equally consistent reading is that the interaction occurs wherever the two proteins meet and that the energy-deficiency phenotype reflects ATG4B's arrival rather than PRMT1's residence - which matters because PRMT1 is mostly cytoplasmic and could be inhibited in either compartment. There is also a therapeutic tension with this entry's own clinical section: the proposal here is to INHIBIT ATG4B in order to RESTORE PRMT1 activity, whereas the failed first-in-human programme inhibited PRMT1 itself. Those are opposite directions on the same enzyme in overlapping disease settings, and the graph should hold both rather than letting the newer, weaker result quietly supersede a terminated randomised-era trial. UniProt Consortium 2026
Connected to