Oncofusion

start / graph / mark-h3r2me2s

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q103-arginine-marks

H3R2me2s - the symmetric mark on the SAME residue as H3R2me2a, which permits rather than enhances WRAD access, and whose 'enhancing' half this graph inherited from the paper that was later revised

also called H3R2me2s · symmetric dimethylation of histone H3 arginine 2 · H3 Arg-2 symmetric dimethylation

H3R2me2s is the mirror-image version of the tag that blocks the KMT2A machine: same spot on the histone, same two methyl groups, arranged differently - and it does not block. Whether it actively pulls the machine in, or merely fails to keep it out, is the point on which two papers disagree.

The same thing, for a scientist

H3R2me2s is the symmetric dimethyl mark on histone H3 arginine 2, reported to be written by PRMT5 and PRMT7 and to be present in euchromatic regions [E1]. Its discovery paper made two claims by SILAC MS profiling of H3-tail interactors plus a WDR5/H3R2me2s co-crystal: that H3R2me2s **excludes** RBBP7, a central component of the Sin3a, NuRD and PRC2 co-repressor complexes, and that it **enhances** binding of WDR5, a shared component of the MLL, SET1A, SET1B, NSL1 and ATAC coactivator complexes [E1]. **The second claim was revised in 2020.** Quantitative binding with high-resolution crystal structures of WDR5 bound to unmodified, monomethyl and symmetric-dimethyl arginine and to the H3R2me1 peptide supports a **binary** switch in which H3R2me2a prevents the WDR5 interaction while **H3R2me0, me1 and me2s are equally permissive** [E2]. So on the current best evidence H3R2me2s is the *relief* of a block, not a gain over unmodified chromatin - which changes what an experiment should expect to see. The one in-cell demonstration that the permissive state carries a consequence is outside leukaemia: in bladder cancer, HSF1 recruits PRMT5, the two jointly induce H3R2me1 and H3R2me2s, and this recruits the WDR5/MLL complex and raises H3K4me3 at LEF1, MMP9, CCL20 and E2F2 [E3]. **No eraser is established** and citrullination provably cannot remove a dimethyl state [E4][E5]. **In a KMT2A-rearranged context there is nothing at all**: `"H3R2me2s"[tiab] AND KMT2A[tiab]` = 0, `AND "MLL-AF9"[tiab]` = 0, `AND menin[tiab]` = 0, all re-run, against an operand control of 1652 for KMT2A [E6].

Go deeper
One residue, two dimethyl isomers, and why that is the whole argument

H3R2 can carry three methylarginine states: monomethyl (me1), asymmetric dimethyl (me2a), symmetric dimethyl (me2s). PRMT5 catalyses me1 and me2s; PRMT6 catalyses me1 and me2a 2.

me2a and me2s have the same mass. They differ only in whether the two methyl groups sit on one terminal nitrogen of the guanidino group or one on each. And they sit on the residue immediately adjacent to the H3K4 the entire KMT2/COMPASS field is about.

The consequence, stated as sharply as the evidence allows:

  • me2a is the OFF state. It prevents WDR5 binding to the H3 tail 2, prevents H3K4 methylation and prevents binding of an ASH2/WDR5/MLL-family complex (mark-h3r2me2a).
  • me2s is a permissive state, indistinguishable from unmodified and from me1 in WDR5 affinity 2.

This is a one-sided switch, not a symmetric two-way one, and the distinction is not pedantry. Under the two-way reading, raising PRMT5 activity at a locus should increase WRAD occupancy above baseline. Under the binary reading, it can only restore baseline - and the way to raise occupancy is to remove me2a, which means inhibiting PRMT6, not activating PRMT5. Those two models prescribe opposite drugs.

What this graph already said, and the correction

gene-prmt5 carries a table whose H3R2me2s row reads "euchromatic; ENHANCES WDR5 binding; excludes RBBP7", sourced correctly to PMID 22231400. gene-prmt6's title turns on the same axis. Neither entry cites PMID 32207970, which is the quantitative and crystallographic re-examination and which reports me2s is merely permissive.

This is recorded as a correction to a claim, not a defect in those entries: their citation says what they say it says, the revision is eight years later, and the revising paper is not a retraction - it is a different measurement (isothermal/quantitative binding on defined amino acids and peptides) reaching a narrower conclusion than the SILAC-plus-co-crystal original.

What survives from Migliori et al. unchallenged: the identification of H3R2me2s as a real histone mark, its euchromatic localisation, its writers, and the RBBP7 exclusion 1. That last one is worth noting separately, because RBBP7 has an entry in this graph (gene-rbbp7) and is a shared subunit of PRC2, NuRD and Sin3a - so the mark has an unrevised co-repressor-exclusion arm that nobody has re-tested, and that arm is the half nothing here has ever reasoned about.

What is now contested: whether H3R2me2s is a gain-of-recruitment mark for WDR5. Treat as moderate, and never as the load-bearing half of a two-way switch argument.

The one place the permissive state has been shown to do work in a cell - and it is not this disease

The gap between "permissive in vitro" and "does something in chromatin" is closed once, in bladder cancer 3. HSF1 co-immunoprecipitates with PRMT5; together they induce H3R2me1 and H3R2me2s; that recruits the WDR5/MLL complex; H3K4me3 rises; LEF1, MMP9, CCL20 and E2F2 go up; and the phenotype is lymphatic metastasis, reversed by the HSF1 inhibitor KRIBB11.

Read carefully, this supports the binary model rather than the graded one - the induced species are me1 and me2s, which Lorton et al. place in the same permissive class as unmodified 2. What the experiment shows is that moving H3R2 out of the me2a state is sufficient to bring WDR5/MLL in at specific promoters in a cell. It does not show that me2s does more than me1, and it does not separate either from a simple loss of me2a, because the me2a state was not measured.

And its transfer value to this project is limited in exactly the way this graph keeps finding. Bladder urothelium, a solid tumour, an HSF1-driven program, wild-type MLL. "H3R2me2s"[tiab] AND leukemia[tiab] returns 1 record and it is this same bladder paper 6. There is no leukaemia measurement of this mark.

Writers: PRMT5 is established, PRMT7 is contested, and UniProt does not agree with the paper

PRMT5. Established. Type II enzyme forming monomethyl- and symmetric dimethylarginine (gene-prmt5, UniProt O14744), named as an H3R2me2s writer in the discovery paper 1 and as the me1/me2s writer in the structural re-examination 2, and shown in cells to induce both states with HSF1 3.

PRMT7 is the interesting case, and it is a three-way disagreement.

sourcewhat it says about PRMT7
Migliori 2012 1H3R2 "is also symmetrically dimethylated (H3R2me2s) by PRMT5 and PRMT7"
Zurita-Lopez 2012 7full-length human PRMT7 is a type III enzyme forming only omega-N-monomethylarginine, on peptides, GST-GAR, myelin basic protein and recombinant human histones H2A, H2B, H3 and H4, under every condition tested - no symmetric and no asymmetric dimethylarginine
UniProt Q9NVM4 8PRMT7 "can both catalyze the formation of omega-N monomethylarginine (MMA) and symmetrical dimethylarginine (sDMA), with a preference for the formation of MMA", and the histone mark it "specifically mediates" is H4R3me2s - H3R2 is not named at all; the catalytic-activity reaction annotated is the monomethyl one

The two 2012 papers are the same year. So the brief's premise that PRMT7 is "a second H3R2me2s writer" rests on a single sentence in one paper, is contradicted by a dedicated enzymological study of PRMT7 product specificity published the same year, and is not carried by the database. That is not a reason to leave PRMT7 out of the graph - gene-prmt7 is written in this beat - but the H3R2me2s attribution is recorded as weak and the H4R3me2s attribution as the one UniProt and the KMT2D trans-tail paper both support (see mark-h4r3me2s).

"H3R2me2s"[tiab] AND PRMT7[tiab] returns 2 records, and both of them are accounted for: PMID 22231400 itself, and PMID 33980975 6.

Bounded absences, with translations, warninglists and re-runs

Run 2026-09-22; querytranslation read back from each esearch response; every zero re-run and every re-run identical.

querynwarninglistre-run
"H3R2me2s"[tiab]28none-
H3R2[tiab]75none-
"H3R2me2s"[tiab] AND WDR5[tiab]8none-
"H3R2me2s"[tiab] AND MLL[tiab]3none-
"H3R2me2s"[tiab] AND leukemia[tiab]1none-
"H3R2me2s"[tiab] AND KMT2A[tiab]0outputmessages: ["No items found."], translation non-empty0
"H3R2me2s"[tiab] AND "MLL-AF9"[tiab]0same0
"H3R2me2s"[tiab] AND menin[tiab]0same0
operand control KMT2A[tiab]1652none-

The identifier trap, measured on ID SETS rather than counts. H3R2[tiab] returns 75 records; the union of "H3R2me2a", "H3R2me2s" and "H3R2me1"[tiab] returns 78; the intersection is 15. Sixty records are reachable only through the residue spelling and sixty-three only through a compact spelling - Jaccard 10.9%. For H3R2me2s specifically, 19 of its 28 records are invisible to H3R2[tiab]. Neither form bounds the other, so a bounded absence on an arginine mark must run both and take the union. This corrects the direction stated in this beat's own dispatch brief and sharpens the note in gene-prmt5: it is not that one form returns fewer records, it is that the two forms barely overlap.

And the bound is titles and abstracts only. No Europe PMC full-text sweep was run. A quantitative histone-MS table in a KMT2A-fusion paper would report both H3R2 dimethyl isomers and be invisible to every query above.

Facts and figures
markH3R2me2s
histoneH3
residueR2 (Arg-2), N-terminal tail, two residues N-terminal of K4
chemistrysymmetric omega-N,N'-dimethylarginine (sDMA) - one methyl group on EACH terminal guanidino nitrogen. Isobaric with H3R2me2a and predicting the opposite WRAD outcome; the two cannot be separated by mass and must be separated by isomer-resolving chromatography/MS or by a demonstrably discriminating antibody.
contested writerPRMT7's H3R2me2s attribution rests on one sentence in PMID 22231400. It is contradicted in principle by PMID 22241471, which found full-length human PRMT7 to be a TYPE III enzyme forming only monomethylarginine on peptides, GST-GAR, myelin basic protein and recombinant human histones H2A/H2B/H3/H4 under all conditions tested - published the same year. UniProt Q9NVM4 records PRMT7 as capable of sDMA with a preference for MMA and assigns its specific histone mark as H4R3me2s, not H3R2me2s; H3R2 is absent from the FUNCTION block and the annotated catalytic reaction is the monomethyl one. This graph records PRMT7 as a weak H3R2me2s writer and a better-supported H4R3me2s writer. THIS PARTIALLY REFUTES A PREMISE IN THIS BEAT'S DISPATCH BRIEF, which described PRMT7 as 'a second H3R2me2s writer' without qualification.
eraser statusNO ESTABLISHED ERASER, and citrullination is ruled out for this state specifically: dimethylation prevents deimination by PADI4 while monomethylation still permits it (PMID 15339660). JMJD6's arginine-demethylase assignment did not hold (PMID 17947579 vs 19574390, 25997831); purified JmjC KDMs show in vitro arginine-demethylase capability with the authors calling bona fide RDMs controversial (PMID 27337104). Removal is therefore by histone turnover or nucleosome eviction.
switch statePERMISSIVE state of the H3R2 binary switch, indistinguishable in WDR5 affinity from me0 and me1 (PMID 32207970). NOT established as an ON state above baseline. The two models differ in what they prescribe: a graded switch says activate PRMT5 to gain WRAD occupancy; a binary switch says inhibit PRMT6 to remove the block, because PRMT5 activity can at best restore baseline.
effect on h3k4 methylationPERMISSIVE. Does not block H3K4 methylation and, in the one cellular demonstration, coincides with WDR5/MLL recruitment and a RISE in H3K4me3 - but in that experiment me1 was induced alongside me2s and me2a was not measured, so the causal state is not resolved (PMID 35434944).
antibody constraintLOAD-BEARING. me2s and me2a are isobaric isomers of one residue predicting opposite WRAD outcomes; an assay that cannot separate them reports an uninterpretable sum, and a drug that converts one state into the other would read as NO CHANGE. Base rates for histone antibodies make assumed specificity unsafe (PMID 21131980, 30244833). BOUNDED: no published head-to-head me2a-versus-me2s H3R2 panel was found; the Histone Antibody Specificity Database (PMID 26212453) exists and was NOT queried in this beat.
relationship to menin and the win motifINDEPENDENT OF BOTH. The state of H3R2 changes whether the H3 tail is available to WDR5 and to a WRAD-containing complex; it does not touch the KMT2A polypeptide, the WIN motif or the menin interface. Parallel control point, not a rival hypothesis.
cell type and stateMeasured in HeLa and NT2 differentiation (PMID 22231400) and bladder urothelial carcinoma lines and xenografts (PMID 35434944). The functional framing in the discovery paper is cell-cycle WITHDRAWAL and differentiation. NOT measured in any haematopoietic cell, normal or leukaemic; not in fetal liver, LMPP, pro-B, or any KMT2A-rearranged line or patient sample.
shown for fusionNO - nobody has looked in any fusion. Zero records pair this mark with KMT2A, with MLL-AF9 or with menin. The three MLL-co-occurring records are the discovery paper, a genotoxic-stress beta-catenin/JDP2/PRMT5 paper and the bladder-cancer HSF1 paper.
absence classliterature-absence-bounded
absence class noteMIGRATED 2026-09-23, mechanically and not by judgement: the class string itself begins 'literature'. The field previously held the whole verdict as free text and that string is preserved here verbatim because it says more than the token can: 'literature - bounded with counts, querytranslations, warninglists, re-runs, an operand control and an ID-set comparison'
writers
[
  "PRMT5 (type II; established)",
  "PRMT7 (claimed by PMID 22231400; WEAK - see contested_writer)"
]
erasers
[]
readers
[
  "WDR5 - binds, but NOT more tightly than unmodified or monomethyl H3R2 (PMID 32207970); the earlier 'enhanced binding' claim (PMID 22231400) is the revised half"
]
excluded readers
[
  "RBBP7 - binding to the H3 tail is EXCLUDED by H3R2me2s; RBBP7 is a shared subunit of the Sin3a, NuRD and PRC2 co-repressor complexes (PMID 22231400). This arm has NOT been revised and, as far as this beat found, has not been independently re-tested. gene-rbbp7 exists in this graph and nothing in it reasons about an arginine mark."
]
genomic distribution
{
  "migliori_2012": "euchromatic regions; proposed to keep genes poised for activation on cell-cycle withdrawal and differentiation (HeLa, NT2)",
  "huang_2022_bladder": "HSF1/PRMT5-induced at specific promoters (LEF1, MMP9, CCL20, E2F2) with WDR5/MLL recruitment and rising H3K4me3",
  "note": "No genome-wide H3R2me2s map was found for any haematopoietic system, normal or malignant."
}
literature bounds
{
  "date": "2026-09-22",
  "operand_control": {
    "query": "KMT2A[tiab]",
    "n": 1652
  },
  "counts": [
    {
      "query": "\"H3R2me2s\"[tiab]",
      "translation": "\"H3R2me2s\"[Title/Abstract]",
      "n": 28
    },
    {
      "query": "\"H3R2me2s\"[tiab] AND WDR5[tiab]",
      "n": 8,
      "ids": [
        "22231400",
        "22810296",
        "26972221",
        "31434880",
        "32207970",
        "33045527",
        "35434944",
        "37390022"
      ]
    },
    {
      "query": "\"H3R2me2s\"[tiab] AND MLL[tiab]",
      "n": 3,
      "ids": [
        "22231400",
        "31434880",
        "35434944"
      ]
    },
    {
      "query": "\"H3R2me2s\"[tiab] AND leukemia[tiab]",
      "n": 1,
      "ids": [
        "35434944"
      ]
    },
    {
      "query": "\"H3R2me2s\"[tiab] AND PRMT7[tiab]",
      "n": 2,
      "ids": [
        "22231400",
        "33980975"
      ]
    },
    {
      "query": "\"H3R2me2s\"[tiab] AND KMT2A[tiab]",
      "translation": "\"H3R2me2s\"[Title/Abstract] AND \"KMT2A\"[Title/Abstract]",
      "n": 0,
      "warninglist": "phrasesignored [], quotedphrasesnotfound [], outputmessages [\"No items found.\"]",
      "rerun_n": 0
    },
    {
      "query": "\"H3R2me2s\"[tiab] AND \"MLL-AF9\"[tiab]",
      "n": 0,
      "warninglist": "outputmessages [\"No items found.\"]",
      "rerun_n": 0
    },
    {
      "query": "\"H3R2me2s\"[tiab] AND menin[tiab]",
      "n": 0,
      "warninglist": "outputmessages [\"No items found.\"]",
      "rerun_n": 0
    }
  ],
  "identifier_trap": "ID-SET comparison, 2026-09-22, retmax=500 so nothing truncated: H3R2[tiab] = 75; union of \"H3R2me2a\"/\"H3R2me2s\"/\"H3R2me1\"[tiab] = 78; INTERSECTION = 15 (Jaccard 10.9%). 19 of the 28 \"H3R2me2s\" records are NOT returned by H3R2[tiab], and 60 H3R2[tiab] records are returned by no compact spelling. The two forms are nearly disjoint; neither bounds the other; every arginine-mark absence must run both and take the union.",
  "bound_type": "titles and abstracts only"
}
How this could bite us back
  • PRMT5 inhibitors are in and past the clinic, and if the binary model is right, inhibiting PRMT5 to lower H3R2me2s achieves nothing at H3R2 - because me1 and unmodified arginine are equally permissive for WDR5. A programme that justified a PRMT5 inhibitor on 'removing a WDR5-recruiting mark' would be resting on the revised half of a 2012 result.
  • The opposite risk is worse. PRMT5 is DepMap near-pan-essential (mean Chronos -1.19 across 1,258 lines, per gene-prmt5) and its dominant activities are the methylosome and snRNP core assembly, not chromatin. Any anti-leukaemic effect of a PRMT5 inhibitor is more likely to run through splicing than through H3R2, so attributing efficacy to this mark would misdirect the pharmacodynamic and the combination strategy. NCT03614728 (GSK3326595 +/- azacitidine, MDS/AML) was TERMINATED after 30 participants on internal review of clinical data.
  • If H3R2me2s excludes RBBP7 as originally reported, a drug that lowers it would be RESTORING a PRC2/NuRD/Sin3a docking site at euchromatic promoters. In a disease where menin inhibition already paradoxically activates bivalent genes by shifting the KMT2A-polycomb balance, adding a second perturbation to the same balance in an unmeasured direction is a real combination hazard, not a theoretical one.
  • Every pharmacodynamic proposed for this axis is antibody-dependent and the two isomers are isobaric. A trial reporting 'H3R2me2s unchanged' with a cross-reactive reagent cannot distinguish that from a complete state inversion, and would license a dose decision on a measurement with no information in it.
Still open
  1. Is H3R2me2s a gain over unmodified H3R2, or only the absence of a block? The two published answers prescribe opposite interventions - activate PRMT5 versus inhibit PRMT6 - and the discriminating experiment is a quantitative occupancy measurement with all three H3R2 states resolved by mass spectrometry on the same chromatin, not by antibody.
  2. Does the RBBP7 exclusion hold, and does it matter here? It is the unrevised half of Migliori's model, RBBP7 is a shared PRC2/NuRD/Sin3a subunit with its own entry in this graph, and nothing in the corpus connects an arginine mark to co-repressor exclusion. If it holds, H3R2me2s is simultaneously WRAD-permissive and PRC2-excluding, which is a bivalency-resolution mechanism nobody has tested in a KMT2A-rearranged cell.
  3. Do PRMT5 inhibitors move H3R2me2s in a KMT2A-rearranged cell, and does anything downstream care? PRMT5 inhibitors have reached myeloid patients (NCT03614728, terminated, 30 participants) and PRMT5 is a near-pan-essential DepMap dependency, so the pharmacology exists - but no trial or preclinical study in this disease has reported an H3R2 measurement of any state.
  4. Which enzyme makes H3R2me2s in a haematopoietic cell? PRMT5's attribution is solid but PRMT7's is contested and its better-supported symmetric mark is H4R3me2s. If PRMT7 is type III in cells as well as in vitro, then PRMT5 is the sole writer and PRMT5 inhibition should collapse the permissive state - a prediction with a cheap test.
Every source behind this entry (9)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    H3R2 is symmetrically dimethylated by PRMT5 and PRMT7 to give H3R2me2s, present in euchromatic regions; SILAC MS profiling of H3-tail interactors showed H3R2me2s excludes binding of RBBP7, a central component of the co-repressor complexes Sin3a, NuRD and PRC2, and enhances binding of WDR5, a common component of the coactivator complexes MLL, SET1A, SET1B, NSL1 and ATAC; a crystallographic structure of WDR5 with the H3R2me2s peptide was solved, and the mark is proposed to keep genes poised in euchromatin for activation upon cell-cycle withdrawal and differentiation in human cells.

    Symmetric dimethylation of H3R2 is a newly identified histone mark that supports euchromatin maintenance. · Migliori V, Müller J, Phalke S, Low D, et al. · Nat Struct Mol Biol 2012
    PMID 22231400DOI
    What this evidence cannot show

    Abstract read from the eutils record 2026-09-22; full text not read. HeLa and NT2 differentiation - not haematopoietic, not leukaemic, no fusion protein. Confidence is MODERATE and not established BECAUSE the 'enhances WDR5' half is specifically revised by E2, which finds me2s merely permissive. The PRMT7 attribution is weaker still (see E7, E8). The RBBP7 exclusion appears not to have been independently re-tested and is carried here at the same moderate confidence.

    Evidence that argues against this (1)
    • Quantitative binding plus crystal structures of WDR5 with unmodified, me1 and me2s arginine report a BINARY switch in which H3R2me0, me1 and me2s are equally permissive and only me2a prevents the interaction - i.e. me2s is not a gain over unmodified chromatin. This is the specific half of the claim this graph's PRMT5 and PRMT6 entries lean on. Lorton BM et al. 2020
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    H3R2 methylarginine isoforms are PRMT5-catalysed me1 and me2s and PRMT6-catalysed me1 and me2a; using quantitative binding analysis with high-resolution crystal structures of WDR5 in complex with unmodified and me1/me2s L-arginine amino acids and with the H3R2me1 peptide, the interaction is regulated by a binary arginine methylation switch in which H3R2me2a prevents interaction with WDR5 whereas H3R2me0, me1 and me2s are equally permissive - the authors' stated purpose being to address long-standing discrepancies about this interaction.

    A Binary Arginine Methylation Switch on Histone H3 Arginine 2 Regulates Its Interaction with WDR5. · Lorton BM, Harijan RK, Burgos ES, Bonanno JB, et al. · Biochemistry 2020
    PMID 32207970DOI
    What this evidence cannot show

    Abstract read from the eutils record 2026-09-22; full text not read, and the PDB accessions of the reported structures were NOT fetched, so no structure id is asserted. Isolated recombinant WDR5 with free amino acids and short peptides: not WDR5 inside an assembled WRAD module, not on a nucleosome, not in a cell. The paper addresses WDR5 affinity only - it does not test the RBBP7 exclusion, which is the other half of Migliori's model and remains unrevised.

    Evidence that argues against this (1)
    • A free-peptide affinity measurement need not predict chromatin occupancy by a six-subunit complex, and the earlier study's evidence was a proteome-scale interactor profile plus a co-crystal, which is a different and complementary design. The disagreement is genuinely unresolved for the me2s state; what both agree on is that me2a is the non-permissive state. Migliori V et al. 2012
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In bladder cancer, HSF1 interacts with PRMT5 and the two jointly induce monomethylation and symmetric dimethylation of histone H3 arginine 2 (H3R2me1 and H3R2me2s); this recruits the WDR5/MLL complex to increase H3K4me3, upregulating LEF1, MMP9, CCL20 and E2F2, and the pharmacological HSF1 inhibitor KRIBB11 inhibited lymphatic metastasis in popliteal lymph-node and patient-derived xenograft models without significant toxicity.

    PMID 35434944DOI
    What this evidence cannot show

    Abstract read from the eutils record 2026-09-22; full text not read. Bladder urothelial carcinoma - not leukaemia, not haematopoietic, wild-type MLL, no fusion protein. The induced species are me1 AND me2s together, so the experiment cannot attribute the WDR5/MLL recruitment to the symmetric dimethyl state specifically, and it does not report H3R2me2a, so it cannot distinguish 'me2s recruits' from 'loss of me2a relieves a block'. The ChIP is antibody-based and no clone or specificity panel is named in the abstract, which for isobaric arginine isomers is the constraint that matters.

    Evidence that argues against this (1)
    • Because me1 and me2s were induced together and me2a was not measured, this result is equally consistent with the binary model in which both induced species are merely permissive - and under that model the causal event is the departure of me2a, which this design does not observe. A version of the experiment that separated the three states by mass spectrometry would distinguish them. Lorton BM et al. 2020
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Peptidyl arginine deiminase 4 (PADI4) deiminates arginine residues R2, R8, R17 and R26 in the H3 tail and antagonises arginine methylation, but dimethylation of arginines PREVENTS deimination by PADI4 while monomethylation still allows it - so citrullination cannot remove a symmetric or asymmetric dimethylarginine mark.

    Histone deimination antagonizes arginine methylation. · Cuthbert GL, Daujat S, Snowden AW, Erdjument-Bromage H, et al. · Cell 2004
    PMID 15339660DOI
    What this evidence cannot show

    Abstract read from the eutils record 2026-09-22; full text not read. The substrate-requirement result is in vitro on H3 tails; the cellular arm is one oestrogen-responsive promoter in MCF-7. H3R2 is explicitly among the deiminated residues, which makes the dimethyl exclusion directly load-bearing for this mark. Deimination is not demethylation in any case: it destroys the arginine and produces citrulline.

    Evidence that argues against this (1)
    • PAD4 does measurably lower histone arginine methylation levels in cells, which can look like erasure. The reconciliation is that it removes the unmethylated and monomethylated substrate pool rather than the dimethyl product, so an experiment seeing H3R2me2s fall after raising PAD4 is most likely seeing reduced deposition; distinguishing the two requires a pulse-chase or a mass-spectrometry read of the citrulline species. Wang Y et al. 2004
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review

    No bona fide arginine demethylase is established for this mark: JMJD6 was reported as a histone arginine demethylase acting on H3R2 and H4R3, was subsequently shown to be a lysyl-5-hydroxylase of the splicing factor U2AF65, and a dedicated review states the demethylase assignment is subject to conflicting reports; separately, purified JmjC lysine demethylases can catalyse arginine demethylation in vitro while the authors describe bona fide arginine demethylases as controversial.

    The oxygenase Jmjd6 - a case study in conflicting assignments. · Böttger A, Islam MS, Chowdhury R, Schofield CJ, et al. · Biochem J 2015
    PMID 25997831DOI
    What this evidence cannot show

    A review, cited as a map of a controversy rather than as primary evidence, per the QUEEN_BRIEF rule. Abstract read from the eutils record 2026-09-22. Ten years old; this beat's searches found no subsequent cellular assignment for an H3R2 eraser, but a review cannot bound the later literature and no claim here says it does. The primary papers on both sides are cited in full on mark-h3r2me2a (E7, E8, E15).

    Evidence that argues against this (1)
    • Detailed biochemistry shows a subset of JmjC KDMs can act as arginine demethylases in purified form with formaldehyde release, so the chemistry is possible and a cellular enzyme may yet be assigned. The graph's position should be 'no established eraser', not 'no eraser can exist'. Walport LJ et al. 2016
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    PubMed esearch, run 2026-09-22 with querytranslation and warninglist read back and every zero re-run: '"H3R2me2s"[tiab]' = 28; AND WDR5[tiab] = 8; AND MLL[tiab] = 3 (22231400, 31434880, 35434944); AND leukemia[tiab] = 1 (35434944, bladder cancer); AND PRMT7[tiab] = 2 (22231400, 33980975); AND KMT2A[tiab] = 0; AND 'MLL-AF9'[tiab] = 0; AND menin[tiab] = 0. Operand control KMT2A[tiab] = 1652. Every zero carried a non-empty translation with warninglist.outputmessages = ['No items found.'] and returned the same zero on re-run. ID-SET comparison: H3R2[tiab] = 75 records, union of the three compact H3R2 spellings = 78, intersection = 15 (Jaccard 10.9%); 19 of the 28 '"H3R2me2s"' records are NOT returned by H3R2[tiab].

    What this evidence cannot show

    Bounds TITLES AND ABSTRACTS only; no Europe PMC full-text sweep was run, and a quantitative histone-MS supplementary table would report both dimethyl isomers invisibly to every query here. The non-zero counts were title-checked but the papers were not all read. The set comparison was run once per spelling on 2026-09-22 with retmax=500, which exceeds every count involved, so no result is truncated.

    Evidence that argues against this (1)
    • This project has a measured instance of exactly this absence being false in the other direction: a bounded H3K79 query on SET-NUP214 returned a clean zero across four spellings while an H3K79me2 ChIP sits in Figure 6A of PMID 18299449. Read these zeros as 'no paper announces the measurement in its title or abstract'. Bernt KM et al. 2011
  7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Full-length human PRMT7 expressed as a GST fusion with robust activity produced only omega-N-monomethylarginine on every substrate tested - synthetic arginine-containing peptides, GST-GAR (the fibrillarin N-terminal domain), myelin basic protein and recombinant human histones H2A, H2B, H3 and H4 - regardless of incubation time, reaction volume or substrate concentration, and neither asymmetric nor symmetric dimethylarginine was formed, while control GST-PRMT1 and Myc-PRMT5 in the same experiments dimethylated their peptides as expected; the authors conclude PRMT7 is a type III methyltransferase capable of forming only monomethylarginine.

    PMID 22241471DOI
    What this evidence cannot show

    Abstract read from the eutils record 2026-09-22; full text not read. In vitro with a bacterially expressed GST fusion; the authors themselves note a prior study characterised bacterially and mammalian-expressed PRMT7 fusions as type II/type I, so the discrepancy may be construct- or condition-dependent, and 'under the conditions tested' is their own qualifier. A negative in vitro result on recombinant histone octamer components does not exclude symmetric dimethylation of a nucleosomal substrate in a cell with cofactors present.

    Evidence that argues against this (1)
    • UniProt's curated FUNCTION for PRMT7 records that it CAN form symmetric dimethylarginine with a preference for monomethyl, and specifically assigns it H4R3me2s deposition - a symmetric dimethyl histone mark. And a Genes Dev study shows PRMT7 knockdown lowers H4R3me2s in cells while PRMT5 knockdown does not, which is a cellular symmetric-dimethyl phenotype this in vitro negative does not account for. Dhar SS et al. 2012
  8. E8established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProt Q9NVM4 (ANM7_HUMAN, PRMT7, 692 aa, two tandem SAM-dependent MTase PRMT-type domains at 14-345 and 358-684) records that PRMT7 can form both monomethylarginine and symmetrical dimethylarginine with a preference for MMA, that the histone mark it specifically mediates is the symmetric dimethylation of histone H4 'Arg-3' to form H4R3me2s, and that it is recruited by CTCFL at the H19 imprinted control region; the annotated catalytic-activity reaction is the monomethyl one, and H3R2 is not named anywhere in the FUNCTION block.

    What this evidence cannot show

    A curated annotation, not a measurement, fetched live rather than recalled. The absence of H3R2 from the FUNCTION block is an absence in a database, which is weaker than a negative experiment - it means no curator has entered the Migliori attribution, not that it is wrong. The 'can form sDMA' statement and the monomethyl-only catalytic reaction sit side by side in the same record, which is itself a sign of an unsettled assignment rather than a settled one.

    Evidence that argues against this (1)
    • The discovery paper for H3R2me2s names PRMT5 AND PRMT7 as its writers in the abstract, so the database's silence on H3R2 is a curation gap as much as a judgement. The safe reading is that PRMT7's H4R3me2s activity is the better-supported of its two claimed symmetric histone marks, and this graph records the H3R2me2s attribution as weak rather than absent. Migliori V et al. 2012
  9. E9established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Fifty-two commercial antibodies purported to distinguish the H3K4 methylforms were characterised by histone peptide array and internally calibrated ChIP; many widely used antibodies poorly distinguish the methylforms and high- versus low-specificity reagents yield dramatically different biological interpretations - and separately, of more than 200 antibodies against 57 histone modifications, more than 25% failed specificity by dot or western blot and more than 20% of the specific ones failed in ChIP.

    Examining the Roles of H3K4 Methylation States with Systematically Characterized Antibodies. · Shah RN, Grzybowski AT, Cornett EM, Johnstone AL, et al. · Mol Cell 2018
    PMID 30244833DOI
    What this evidence cannot show

    Abstract read from the eutils record 2026-09-22; full text not read. This study is about the H3K4 methylforms, not about H3R2 methylarginine isomers - it is the residue two positions away, cited to establish that failures to discriminate closely related methyl states of one residue are common and change conclusions. The second half of the claim (25%/20%) is from PMID 21131980, cited in full on mark-h3r2me2a E12. No published head-to-head me2a-versus-me2s H3R2 antibody panel was found in this beat.

    Evidence that argues against this (1)
    • Most antibodies in the broader survey performed well, so a blanket presumption of failure is also wrong; the constraint is per-clone and per-lot. The Histone Antibody Specificity Database is the place to check, and this beat did NOT query it - so the arginine-mark reagents' actual behaviour is an open item rather than a finding. Rothbart SB et al. 2015
Connected to