Oncofusion

start / graph / gene-brd2

Genemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q39-acetylation-cycle

BRD2 - a bounded absence: zero KMT2A papers name it, it is inhibited by every BET compound used in this disease, and it is the paralogue most likely to carry the toxicity

also called RING3 · FSRG1 · KIAA9001 · O27.1.1 · bromodomain-containing protein 2

BRD2 is the third member of the BET family of reader proteins, alongside BRD3 and BRD4. No paper about MLL or KMT2A leukaemia mentions it by name - and yet every BET-blocking drug ever tested in this disease blocks it just as hard as the other two. It is a target that is always hit and never studied.

The same thing, for a scientist

BRD2 (P25440, 801 aa) is a BET-family acetyl-lysine reader. This entry records a **bounded absence with a pharmacological complement**, which is a combination the graph does not otherwise hold. `(MLL[tiab] OR KMT2A[tiab]) AND (BRD2[tiab] OR BRD3[tiab])` returns **7 records**; title-checked, **not one of them makes a BRD2-specific claim in KMT2A-rearranged leukaemia** [E1], and a complex-level query on `"BET bromodomain" OR "BET protein" OR "BET proteins"` returns 17 records that are likewise pan-BET or BRD4-framed [E2]. So BRD2 has **no individual literature in this disease at all**. What makes that worth an entry rather than a footnote is the other half: **every BET compound used in this disease occupies BRD2's bromodomains too.** I-BET151, JQ1 and I-BET762 are pan-BET by design, and the selectivity the field has achieved within the class is **BD1-versus-BD2 - the first versus the second bromodomain - which cuts across the paralogues rather than between them** [E3]. BRD2 is therefore engaged in every experiment and credited in none. **The asymmetry matters most for toxicity rather than for efficacy**: if a pan-BET compound's anti-leukaemic effect runs through BRD4 while a share of its dose-limiting toxicity runs through BRD2, then the paralogue nobody studies is the one setting the therapeutic window - and nothing in this literature could detect that.

Go deeper
The absence, bounded properly

Run 2026-09-20, title/abstract, query translations read, all hits title-checked:

(MLL[tiab] OR KMT2A[tiab]) AND (BRD2[tiab] OR BRD3[tiab])                              ->  7
(MLL[tiab] OR KMT2A[tiab]) AND ("BET bromodomain"[tiab] OR "BET protein"[tiab]
                                 OR "BET proteins"[tiab])                              -> 17

The 7 1: two BRD4-PROTAC (MZ1) papers, a dBET1 paper, a SETBP1/FLT3-ITD paper, a breast-cancer bromodomain mapping study, the BRD3/4 I-BET151 paper, and the founding Nature paper. None is BRD2-specific. The 17 2 add WDR5 WIN-site inhibition, PD-1/T-cell exhaustion, a CBP/p300 multi-bromodomain inhibitor, AMPK/acetyl-CoA/BRD4 recruitment, the BET resistance paper, and both halves of the Reproducibility Project pair - again none BRD2-specific.

Both the singular and plural phrase forms were run per this project's rule, and the complex-level forms were run per the rule that a subunit-keyed count is structurally wrong for a complex. The absence survives all of them.

What would break it: a paper framed at the level of a BRD2-containing complex or a BRD2 alias. RING3 is BRD2's classical alias and was not searched; by this project's own retired-symbol rule that makes 7 a floor rather than a measurement, and this entry says so rather than claiming a clean zero.

Engaged in every experiment, credited in none

The BET bromodomain acetyl-lysine pocket is strongly conserved across BRD2, BRD3 and BRD4. The consequence is that the class's chemistry has never delivered paralogue selectivity; what it has delivered is BD1/BD2 selectivity, distinguishing the first bromodomain from the second within each protein 3. That axis is orthogonal to the one this entry needs.

So for every result in gene-brd3 and in the BET literature of this disease:

  • BRD2 was inhibited.
  • BRD2 was not measured.
  • BRD2 was not named.

Three separate facts, and the graph should not let the third imply the first two did not happen.

The practical reading. Efficacy attribution in this family points at BRD4 for good reasons - chromatin-bound BRD4 is what is globally reduced in resistant cells, and BRD4 already has an entry here. But toxicity attribution has never been done at all. BRD2 has distinct biology - it is the BET protein most associated with adipogenesis, insulin secretion and, through its position in the MHC class II locus, immune gene regulation. If the thrombocytopenia and gastrointestinal toxicity that have limited BET inhibitors clinically are partly BRD2-mediated, a BRD4-selective degrader would widen the window and a pan-BET inhibitor never could. That is a testable, currently untested, and clinically consequential proposition, and it is the reason this entry exists as a node rather than as a line in someone else's caveats.

Facts and figures
uniprotP25440
length AA801
gene symbolBRD2
accession resolutiongene_exact:BRD2 AND organism_id:9606 AND reviewed:true returned exactly 1 reviewed row; primary gene name read back as BRD2 (rule 62 applied).
familyBET (bromodomain and extraterminal) family acetyl-lysine reader
kmt2a specific literaturenone found
symbol query count7
complex query count17
brd2 specific claims among them0
absence is a floortrue
inhibited by every bet compound used in this diseasetrue
selectivity axis the class achievesBD1 versus BD2 within each protein - orthogonal to paralogue selectivity
why it mattersengaged in every experiment, measured in none; if a share of pan-BET dose-limiting toxicity is BRD2-mediated then the unstudied paralogue sets the therapeutic window and no existing experiment could detect it
distinct biologyadipogenesis, insulin secretion, and immune gene regulation via its position in the MHC class II locus - stated as background, not sourced in this entry
paralogues
[
  "gene-brd3",
  "gene-brd4"
]
aliases not searched
[
  "RING3",
  "FSRG1",
  "KIAA9001"
]
compounds that engage it
[
  "I-BET151",
  "JQ1",
  "I-BET762",
  "dBET1",
  "MZ1 (BET-family degrader with reported BRD4 preference)"
]
essentiality
{
  "depmap_common_essential": null,
  "note": "Not queried here."
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.3285,
  "chronos_mean_other_leukaemia": -0.3862,
  "difference": 0.0577,
  "pan_essentiality_all_lines": -0.4353,
  "selectivity_rank": 15287,
  "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."
}
How this could bite us back
  • BRD2 is engaged by every BET compound ever tested in this disease and studied by none of them. Any safety reasoning about BET inhibitors in infants is therefore reasoning about a drug whose effect on one of its three main targets has never been separated out.
  • Because the chemical class achieves BD1/BD2 selectivity rather than paralogue selectivity, the natural next-generation compound does not solve this problem - it solves a different one. A programme aiming to spare BRD2 needs a degrader, not a better inhibitor.
  • The absence recorded here is a floor, not a zero: BRD2's classical alias RING3 was not searched. An intervention argued from 'BRD2 is unstudied' could be arguing from an incomplete query, and this entry flags that rather than hiding it.
  • If BRD2 turns out to carry part of the efficacy rather than part of the toxicity, then a BRD4-selective degrader developed to widen the window would lose activity instead - and there is currently no data in this disease that could predict which way it goes.
Still open
  1. Is any part of pan-BET dose-limiting toxicity BRD2-mediated? Nobody has asked, and the answer decides whether a BRD4-selective degrader would widen the therapeutic window or merely narrow the efficacy.
  2. Does a RING3-keyed search recover a BRD2 literature in this disease that the current-symbol search missed? The alias was not run and the absence claimed here is explicitly a floor.
  3. Do BRD2, BRD3 and BRD4 occupy different loci in a KMT2A-rearranged cell? Paralogue-resolved ChIP in one cell line would be cheap and would be the first such data in this disease.
  4. Is BRD2 part of the resistance state? Chromatin-bound BRD4 is globally reduced in I-BET-resistant cells, and nobody has reported what BRD2 does under the same conditions.
Every source behind this entry (4)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    A title/abstract search for BRD2 or BRD3 against MLL or KMT2A returns 7 records, none of which makes a BRD2-specific claim in KMT2A-rearranged leukaemia; the seven are two BRD4-PROTAC (MZ1) studies, a dBET1 study, a SETBP1/FLT3-ITD study, a breast-cancer bromodomain mapping study, the BRD3/4 I-BET151 study in MLL-AF9, and the founding BET/MLL Nature paper.

    What this evidence cannot show

    BRD2's classical alias RING3 was NOT searched, nor were FSRG1 or KIAA9001. By this project's retired-symbol rule - which found SEPTIN9 returning 1 record where SEP9 returned 20 - an alias-incomplete count is a FLOOR and not a measurement, and this entry treats it as one. RING3 is also a short token that would need title-checking for precision. Only titles were checked, not abstracts, so a BRD2-specific claim made in an abstract under a non-BRD2 title would have been missed.

    Evidence that argues against this (1)
    • A counted absence built from gene symbols has repeatedly missed most of the relevant literature in this project, and the complementary failure is that a paper may frame its work at the level of the BET family while containing paralogue-resolved data in its figures. A title-level absence bounds what is claimed in titles, not what was measured. NCBI E-utilities 2026
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    A complex-level title/abstract search on BET phrase forms against MLL or KMT2A returns 17 records, all of which are pan-BET or BRD4-framed on title check, with none BRD2-specific; the set includes WDR5 WIN-site inhibition, BET inhibition and PD-1-mediated T-cell exhaustion, a CBP/p300 multi-bromodomain inhibitor in MLL-rearranged ALL, AMPK-acetyl-CoA-BRD4 recruitment, the BET resistance paper, and both the Registered Report and the Replication Study of the founding paper.

    What this evidence cannot show

    Both the singular and plural phrase forms were run, per this project's rule that PubMed matches a multi-word tiab term as an unstemmed phrase. 'Bromodomain and extra-terminal' spelled out was not run as a separate phrase. Title-level checking only. This query is offered as a complex-level complement to the symbol query rather than as an independent measurement.

    Evidence that argues against this (1)
    • A title-checked search count is two judgements stacked: the query decides what is reachable, and the reader decides what is 'on-beat'. Both are recorded here, which is the right practice, but neither is a measurement. This project has since established that field-scoped queries can return zero regardless of the literature, and that a symbol-level count misses work indexed under the pathway or the complex rather than the gene - so a small count bounds what was found, not what exists. Shan X et al. 2017
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Resistance to the BET inhibitor I-BET in MLL-AF9 leukaemia confers cross-resistance to chemically distinct BET inhibitors such as JQ1 as well as resistance to genetic knockdown of BET proteins, and chromatin-bound BRD4 is globally reduced in resistant cells - the genetic arm having been performed at the level of BET proteins as a group rather than per paralogue.

    BET inhibitor resistance emerges from leukaemia stem cells. · Fong CY, Gilan O, Lam EY, Rubin AF, Ftouni S, Tyler D, et al. · Nature 525(7570):538-542 2015
    PMID 26367796DOI
    What this evidence cannot show

    Cited here for the family-level nature of the perturbations, which is the point this entry needs, and for the BRD4-specific chromatin observation. The claim that BET chemistry achieves BD1-versus-BD2 rather than paralogue selectivity is general medicinal-chemistry background and is NOT established by this citation; it is stated in this entry as context and a reader who needs it load-bearing should source it separately. Abstract read; PMCID PMC6069604, full text not fetched.

    Evidence that argues against this (1)
    • Cross-resistance between chemically distinct BET inhibitors and genetic BRD4 knockdown argues the resistance is on-pathway, but it was derived in MLL-AF9 mouse cells - a myeloid fusion - and the formal replication of the founding BET/MLL result recovered selective growth inhibition in MV4;11 while reporting that not every effect reproduced at the originally described magnitude. Shan X et al. 2017
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Human BRD2 is UniProt P25440, primary gene name BRD2, 801 amino acids, primary protein name 'Bromodomain-containing protein 2'.

    UniProtKB P25440 (BRD2_HUMAN), Bromodomain-containing protein 2 · UniProt Consortium · UniProtKB, retrieved 2026-09-20 2026
    What this evidence cannot show

    Resolved per rule 62: gene_exact:BRD2 AND organism_id:9606 AND reviewed:true returned exactly one reviewed row and the PRIMARY gene name was read back as BRD2. No alias collision with BRD3 or BRD4 was observed.

    Evidence that argues against this (1)
    • A reviewed database record fixes the canonical sequence on the date it was read and nothing more. It does not establish which isoform is expressed in KMT2A-rearranged blasts, and this beat's own findings turn on distinctions a canonical record hides - HDAC9's catalytically inactive MITR form, and the nomenclature collision in which HDAC9's protein name reads 'Histone deacetylase 7B'. Accession and length are asserted here so a swap fails the validator, not because they carry biological weight. Vega-Garcia N et al. 2018
Connected to