An assay that cannot separate me2a from me2s cannot test the arginine switch - the two states are isobaric and predict OPPOSITE things, so a non-discriminating readout returns no information rather than an imprecise answer
also called the arginine isomer problem · me2a versus me2s discrimination · isobaric methylarginine constraint
Two different arginine tags on the same histone spot weigh exactly the same but do opposite things, and most antibodies cannot tell them apart. An experiment using such an antibody would report 'no change' even if the biology had completely flipped.
The same thing, for a scientist
Asymmetric and symmetric dimethylarginine on the same residue are **structural isomers with identical mass** - two methyl groups on one terminal guanidino nitrogen versus one on each. At both H3R2 and H4R3 the two states have **opposite** consequences for KMT2-complex access: H3R2me2a prevents WDR5 binding while me0/me1/me2s are equally permissive [E1]; H4R3me2s inhibits KMT2D's tandem PHD(4-6) while H4R3me0/me2a are permissive [E2]. **The consequence is qualitative, not quantitative.** A readout detecting both states reports their sum, which is the sum of an inhibitory and a permissive mark and has no mechanistic interpretation. A perturbation that CONVERTS one state into the other - which is exactly what a PRMT5-versus-PRMT6 or PRMT7-versus-PRMT5 change does - reads as **no change**, so a complete inversion of the mechanism scores as a null result. The field's base rate makes this live: >25% of 200+ antibodies against 57 histone modifications failed specificity by dot or western blot and >20% of the specific ones failed in ChIP [E3]; for the residue **two positions from H3R2**, 52 commercial antibodies against the H3K4 methylforms often fail to distinguish states, with reagent choice changing the biological conclusion [E4]. **Bounded: no published head-to-head me2a-versus-me2s panel was found for either residue** - `"H3R2me2a"[tiab] AND antibod*[tiab]` = 1 and `"H4R3me2s"[tiab] AND antibod*[tiab]` = 3, none a specificity study of this reagent class [E5]. The Histone Antibody Specificity Database exists [E6] and **was not queried in this beat**.
Why this is a different failure mode from ordinary antibody imprecision
This graph already carries several entries about instruments that measure a proxy. This is worse than a proxy, because the sign is not preserved.
| failure | what you get |
|---|---|
| a noisy antibody | the right direction with a wide error bar |
| an antibody measuring a proxy | a real quantity that correlates imperfectly with the one you wanted |
| an antibody that sees both me2a and me2s | the sum of two quantities with opposite meanings - and the sum can be flat while both components move |
So the standard remedies do not apply. More replicates do not help; a spike-in calibration does not help, because calibration fixes the scale and not the identity of what is being counted; and a stronger effect size makes the problem worse, because a larger state conversion produces a larger cancellation.
Worked out on the actual prediction. Suppose a PRMT6 inhibitor converts H3R2me2a into H3R2me0/me1 at a fusion target. WRAD access should go from blocked to permissive - the largest mechanistic change this axis can produce. A pan-H3R2-dimethyl reagent would report the me2a pool falling; a reagent cross-reacting with me2s would report a smaller fall or none; and a reagent that in fact detects me2s preferentially would report an increase. Three published-looking answers, three different conclusions, from the same chromatin.
What this licenses and forbids, as a rule other entries can cite
Forbidden without demonstrated discrimination:
- Asserting a direction of change in an arginine dimethyl state from an antibody-based ChIP, CUT&RUN, western or immunofluorescence.
- Reading a null result as evidence against the switch model.
- Using an arginine-mark antibody as a pharmacodynamic endpoint for a PRMT inhibitor in a trial or a preclinical dose-finding study.
- Comparing two antibody-based studies of the two isomers as though they were two measurements of one system - which is what every distribution result carried in this beat's four mark entries does, including the developmental H4R3 series that IS literally a comparison of two such reagents (PMID 21072194).
Permitted:
- Isomer-resolving mass spectrometry on purified histones or fractionated chromatin. The isomers are chemically distinguishable and MS is the arbiter.
- An antibody whose clone and lot have been shown to discriminate - by peptide microarray against both isomers, or by internally calibrated ChIP with isomer-defined nucleosome standards. Report the clone, not the label: this graph's rule-74 lesson, arrived at independently here.
- Genetic logic that does not require reading the mark: compare a PRMT6 loss to a PRMT5 loss to a PRMT7 loss and read the DOWNSTREAM output - KMT2-complex occupancy or H3K4me3 - rather than the arginine mark itself. This is the cheapest available route and is how the KMT2D trans-tail result was obtained 2.
And the honest self-application. Every mark entry written in this beat carries this constraint in a data.antibody_constraint field, and every distribution claim in them is antibody-derived. This entry does not rescue those claims; it bounds them, and the bound is that the direction of each published distribution result is as reliable as an unreported reagent.
| the constraint | Asymmetric and symmetric dimethylarginine on one residue are structural isomers of identical mass. At H3R2 and at H4R3 the two states have OPPOSITE consequences for KMT2-complex access. A readout that cannot separate them reports their sum, which is the sum of an inhibitory and a permissive mark, and a perturbation converting one into the other reads as NO CHANGE. |
|---|---|
| why it is not ordinary imprecision | The sign is not preserved. More replicates, spike-in calibration and larger effect sizes all fail to help - a larger state conversion produces a larger cancellation. This is a different failure class from a noisy measurement or a proxy measurement. |
| affected results in this graph | Every distribution and perturbation claim in mark-h3r2me2a, mark-h3r2me2s, mark-h4r3me2a and mark-h4r3me2s is antibody-derived: PMID 17898714, 17898715, 30232013, 19234465, 23249737, 28945229, 22689669, 21072194, 35434944. PMID 21072194 is literally a comparison between an anti-H4R3me2a and an anti-H4R3me2s reagent, so its developmental switch would be reproduced in full by a reagent affinity difference. This concept BOUNDS those claims; it does not rescue them. |
| bounded absence | No published head-to-head me2a-versus-me2s discrimination panel was FOUND for either residue: '"H3R2me2a"[tiab] AND antibod*[tiab]' = 1, '"H4R3me2s"[tiab] AND antibod*[tiab]' = 3, and a relevance-ranked keyword search for a methylarginine antibody cross-reactivity study = 0 (2026-09-22). WEAKLY BOUNDED ON PURPOSE: reagent validation data live in supplementary figures, vendor datasheets and the Histone Antibody Specificity Database, none of which PubMed titles and abstracts index. NAMED NEXT STEP: query histoneantibodies.com for the H3R2 and H4R3 reagents - a desk job of under an hour that would either license or retire a decade of distribution claims. |
| generalises to | Any pair of isobaric post-translational modification states of one residue whose consequences differ in SIGN. In this corpus the nearest relative is rule 74's antibody-clone lesson (the clone is the identifier, not the label), reached here independently and from the opposite direction - there the epitope was shared across genes, here it is shared across two chemically distinct states of one residue. |
| absence class | literature-absence-bounded |
| absence class note | MIGRATED 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 - weakly bounded and labelled as such, because the relevant data are not indexed by the instrument used' |
forbidden without demonstrated discrimination
[ "asserting a direction of change in an arginine dimethyl state from antibody-based ChIP, CUT&RUN, western or immunofluorescence", "reading a null result as evidence against the switch model", "using an arginine-mark antibody as a pharmacodynamic endpoint for a PRMT inhibitor", "comparing two antibody-based studies of the two isomers as two measurements of one system" ]
permitted
[ "isomer-resolving mass spectrometry on purified histones or fractionated chromatin - the arbiter", "an antibody whose CLONE AND LOT have been shown to discriminate by peptide microarray against both isomers or by internally calibrated ChIP with isomer-defined nucleosome standards; report the clone, not the label", "genetic logic that does not require reading the mark: compare PRMT6 loss to PRMT5 loss to PRMT7 loss and read the DOWNSTREAM output (KMT2-complex occupancy, H3K4me3) rather than the arginine mark - the route by which the KMT2D trans-tail result was obtained" ]
- A trial that declares target engagement for a PRMT inhibitor on an arginine-mark antibody readout may have measured nothing, and the failure is invisible: a flat result is indistinguishable from a complete state inversion. Dose escalation on such an endpoint is dose escalation on a number with no information in it.
- The constraint cuts both ways commercially. A negative pharmacodynamic could kill a programme whose drug was working perfectly, if the assay sees both isomers - so this is a reason to spend on mass spectrometry early, not a reason to avoid the axis.
- If a published distribution map is an artefact of reagent affinity, then the biology this graph now records on this axis is partly a property of two antibodies. The four mark entries written here name that risk in a dedicated field rather than burying it, but naming it does not remove it, and a reader quoting any distribution claim from them should quote the reagent caveat with it.
- Do the commonly used H3R2me2a, H3R2me2s, H4R3me2a and H4R3me2s reagents in fact discriminate? The Histone Antibody Specificity Database is the place to find out and was not queried in this beat. Until it is, every distribution claim on this axis in this graph is as reliable as an unreported reagent.
- Is there an isomer-resolving mass-spectrometry method already applied to H3R2 and H4R3 in any haematopoietic system? If a published method exists, the decisive experiment for the whole arginine axis is a sample-prep problem rather than a method-development problem, which changes its cost tier from high to low.
- Can the switch be tested entirely without reading the mark? The KMT2D trans-tail result was obtained by perturbing writers and reading the enzyme's catalytic output, which is immune to this constraint. Whether the same design works in a KMT2A-rearranged cell - PRMT6 versus PRMT5 versus PRMT7 loss, read out as H3K4me3 at fusion targets - is the question that decides whether this axis is testable at all with current reagents.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Quantitative binding analysis with high-resolution crystal structures of WDR5 in complex with unmodified, monomethyl and symmetric-dimethyl L-arginine and with the H3R2me1 peptide supports an interaction regulated by a binary arginine methylation switch: H3R2me2a prevents interaction with WDR5, whereas H3R2me0, me1 and me2s are equally permissive - so the two isobaric dimethyl states of one residue have opposite consequences for WDR5 engagement.
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 2020What this evidence cannot show
Abstract read from the eutils record 2026-09-22; full text not read; no PDB accession asserted. Isolated recombinant WDR5 with free amino acids and short peptides, not WDR5 in an assembled WRAD module, not on a nucleosome, not in a cell. The opposite-consequence claim this concept rests on is about WDR5 affinity specifically.
Evidence that argues against this (1)
- The earlier study reported H3R2me2s as ENHANCING WDR5 binding rather than merely permissive. Note that this disagreement does not weaken the concept - under EITHER model me2a and me2s have opposite signs, so a non-discriminating assay fails either way. The concept is robust to the unresolved question. Migliori V et al. 2012
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The tandem plant homeodomain PHD(4-6) of MLL4 (KMT2D) recognises unmethylated or asymmetrically dimethylated histone H4 Arg 3 and is required for the enzyme's nucleosomal methyltransferase activity, while its binding strength is inhibited by H4R3 symmetric dimethylation; PRMT7 but not PRMT5 represses MLL4 target genes by upregulating H4R3me2s, and PRMT7 knockdown increases MLL4-catalysed H3K4me3 - a result obtained by reading the downstream output under genetic perturbation of the writers rather than by discriminating the isomers with antibodies alone.
Trans-tail regulation of MLL4-catalyzed H3K4 methylation by H4R3 symmetric dimethylation is mediated by a tandem PHD of MLL4. · Dhar SS, Lee SH, Kan PY, Voigt P, et al. · Genes Dev 2012What this evidence cannot show
Abstract read from the eutils record 2026-09-22; full text not read. NOMENCLATURE: 'MLL4' is KMT2D, settled by the MLL2/ALR aliases and the Kabuki-syndrome attribution. NT2/D1 embryonal carcinoma differentiation, not haematopoietic and not leukaemic. The abstract does report H4R3me2s LEVELS changing, which is presumably antibody-based, so this paper is not itself free of the constraint - what is cited here is the design principle that the discriminating readout was the enzyme's own catalytic output.
Evidence that argues against this (1)
- The paper's own H4R3me2s level measurements are presumably antibody-based and no isomer-specificity panel is described in the abstract, so its me2s-specific claims inherit the same constraint. Only the PHD-binding assays, done with defined synthetic peptides, are free of it. Egelhofer TA et al. 2011
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
More than 200 antibodies raised against 57 different histone modifications were tested in Drosophila, C. elegans and human cells; more than 25% failed specificity tests by dot blot or western blot, and among the specific antibodies more than 20% failed in chromatin immunoprecipitation, so rigorous per-reagent testing is required before use.
An assessment of histone-modification antibody quality. · Egelhofer TA, Minoda A, Klugman S, Lee K, et al. · Nat Struct Mol Biol 2011What this evidence cannot show
Abstract read from the eutils record 2026-09-22; full text not read. A base rate across 57 modifications; it does NOT report the failure rate for H3R2 or H4R3 methylarginine antibodies specifically, nor whether any tested reagent was assayed for symmetric-versus-asymmetric discrimination. It establishes that assuming specificity is unsafe; it does not establish that any particular arginine reagent fails.
Evidence that argues against this (1)
- Most antibodies in that panel performed well, so this concept must not be read as a blanket prohibition - it is a requirement to demonstrate discrimination per clone and per lot, which is achievable and is a one-experiment cost, not a reason to abandon antibody-based work. Rothbart SB et al. 2015
- E4established 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, resulting in substantial divergence from the literature for numerous H3K4 methylform paradigms - illustrating how poor antibody specificity contributes to the reproducibility crisis.
Examining the Roles of H3K4 Methylation States with Systematically Characterized Antibodies. · Shah RN, Grzybowski AT, Cornett EM, Johnstone AL, et al. · Mol Cell 2018What this evidence cannot show
Abstract read from the eutils record 2026-09-22; full text not read. This concerns H3K4 me1/me2/me3 - a LYSINE, and three degrees rather than two isomers of one degree. It is the residue two positions from H3R2 and is cited to show that discrimination failures among closely related methyl states of one residue are common and change published conclusions. It reports nothing about arginine reagents.
Evidence that argues against this (1)
- The transfer from lysine methylforms to arginine dimethyl isomers is by analogy. It would be refuted as a constraint if a published panel showed arginine reagents discriminate cleanly - and none was found, so the constraint stands as unresolved rather than demonstrated. A beat that queries the Histone Antibody Specificity Database could settle it in an hour. NCBI PubMed 2026
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
PubMed esearch 2026-09-22 with querytranslation read back: '"H3R2me2a"[tiab] AND antibod*[tiab]' = 1 (PMID 17898715, which describes 'an antibody specific for H3R2me2a' used for ChIP-on-chip without a published symmetric-state competition); '"H4R3me2s"[tiab] AND antibod*[tiab]' = 3 (21072194, 25092918, 36230689), none of which is a specificity study of this reagent class; and a keyword search for a methylarginine antibody cross-reactivity study ('symmetric asymmetric dimethylarginine antibody cross-reactivity specificity proteomic') returned 0 records. So no published head-to-head me2a-versus-me2s discrimination panel was found for either H3R2 or H4R3.
PubMed esearch counts, querytranslations and a zero-returning keyword search for methylarginine antibody specificity, run 2026-09-22 · NCBI PubMed 2026What this evidence cannot show
CONFIDENCE IS MODERATE AND THE ABSENCE IS WEAKLY BOUNDED, deliberately. Antibody validation data live in supplementary figures, in vendor datasheets and in the Histone Antibody Specificity Database - none of which PubMed titles and abstracts index. The keyword search that returned 0 was a relevance-ranked phrase search, not a structured field query, and is the weakest instrument used anywhere in this beat. This item records that the discrimination has NOT BEEN FOUND, not that it has not been demonstrated.
Evidence that argues against this (1)
- The Histone Antibody Specificity Database was not queried and is exactly where such data would live; a negative from PubMed titles and abstracts is close to uninformative about reagent validation, which is essentially never the subject of a paper title. This absence should be closed by querying the database, not by widening the PubMed query. Rothbart SB et al. 2015
- E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The Histone Antibody Specificity Database (http://www.histoneantibodies.com) is an online, expanding resource cataloguing the behaviour of widely used, commercially available histone antibodies by peptide microarray, intended as a resource for the research community that routinely uses these antibodies as detection reagents.
An Interactive Database for the Assessment of Histone Antibody Specificity. · Rothbart SB, Dickson BM, Raab JR, Grzybowski AT, et al. · Mol Cell 2015What this evidence cannot show
Abstract read from the eutils record 2026-09-22; full text not read, and THE DATABASE WAS NOT QUERIED IN THIS BEAT. Nothing here asserts what it says about any H3R2 or H4R3 reagent - only that it exists and is where to look. A named next step, not a finding.
Evidence that argues against this (1)
- A peptide microarray reports binding to a synthetic peptide, which is not the same question as behaviour on a nucleosome in a cell; internally calibrated ChIP with nucleosome standards is the stronger test and gives different answers for some reagents that pass on peptides. So even a clean database entry would not fully discharge the constraint. Shah RN et al. 2018
the measurement gate on the whole model: a non-discriminating readout returns no information, and a state inversion reads as no change
bounds every distribution claim in this entry, all of which are antibody-derived
bounds every distribution claim in this entry
bounds this entry's distribution map, which IS a comparison between two antibodies against the isobaric isomers
bounds every distribution and perturbation claim in this entry
designed so that NO conclusion depends on an arginine-mark antibody - mass spectrometry for description, catalytic output for function