Oncofusion

start / graph / gene-hdac6

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

HDAC6 - the one HDAC family member this graph refused, whose refusal named a falsifier that has been in the literature since 2020 and fires in the index fusion's own cell line

also called HDAC6 · histone deacetylase 6 · KIAA0901 · Q9UBN7 · tubulin deacetylase

HDAC6 is the enzyme that strips acetyl groups from tubulin rather than from histones. This graph decided not to write it up because a literature search found nothing - but the search could only see titles and abstracts, and several HDAC6-blocking drugs have in fact been tested, at very low doses, on the exact leukaemia cell line that carries this project's index fusion.

The same thing, for a scientist

**HDAC6** (UniProt **Q9UBN7**) is the cytoplasmic, two-catalytic-domain class IIb deacetylase whose principal substrate is **alpha-tubulin**, not histones. This graph carries `gene-hdac1`, `gene-hdac2`, `gene-hdac3`, `gene-hdac5`, `gene-hdac7` and `gene-hdac9` and **explicitly declined HDAC6**, on the ground that `(MLL[tiab] OR KMT2A[tiab]) AND HDAC6[tiab]` returns **0 records**. That refusal named its own falsifier: *"any HDAC6-selective agent tested in a KMT2A-rearranged model."* 🔴 **The falsifier fires, repeatedly, and the evidence predates the refusal.** | agent | HDAC6 potency / selectivity | KMT2A-rearranged model | source | |---|---|---|---| | **PTG-0861** (JG-265) | IC50 **5.92 nM**, ~**36x** isozyme-selective | **MV4-11** | PMID 32615502 | | artemisinin-hybrid lead | HDAC6 IC50 **12 nM**, selective | **MV4-11 IC50 28 nM**, **MOLM-13 IC50 89 nM**, plus a **MOLM-13-Luc xenograft** | PMID 41701111 | | quinazoline-cap series | HDAC6-selective | MV4-11 | PMID 26443078 | | tetrahydro-beta-carboline series | HDAC6-selective | MV4-11 | PMID 39067437 | **MV4-11 is KMT2A::AFF1** - the index fusion of this entire project - as this graph records in `drug-chidamide`. **MOLM-13 is KMT2A::MLLT3.** None of these seven papers is cited anywhere in this corpus, and no HDAC6-selective agent (PTG-0861, tubastatin, ricolinostat, citarinostat, WT161) is named in a single node file [E1]. 🟠 **What this does NOT show, and the counter-evidence is strong.** These are **antiproliferative** results, not interaction results. **Nothing here places HDAC6 in a complex with KMT2A::AFF1, at a fusion target locus, or in any fusion-specific role.** The same compounds are comparably potent in **HL60** (IC50 92 nM, not KMT2A-rearranged) and **MM1S** (multiple myeloma), so the sensitivity is **not fusion-selective** - which is precisely the conclusion the parent HDAC entry reached for every other HDAC and which survives here [E2]. 🔵 **The sharpest counter-evidence is a dedicated isoform comparison in the right lineage.** Stubbs *et al.* (Armstrong lab, *Clin Cancer Res* 2015, PMID **25688158**) tested isoform-selective HDAC inhibitors across a **B-ALL** panel with shRNA validation and concluded that **targeting HDAC1-3 was sufficient**, and that an HDAC1/2-specific compound was the therapeutically useful one. **In B-lineage ALL, the isoform evidence points at HDAC1 and HDAC2, not HDAC6** [E3]. 🟢 **The precise remaining gap, stated as a bound rather than a vibe.** HDAC6-selective agents have been tested in **KMT2A-rearranged myeloid/biphenotypic** lines (MV4-11, MOLM-13) and in **non-KMT2A B-ALL** lines (BALL-1, NALM6, PMID 41249257). **No HDAC6-selective agent has been tested in a KMT2A-rearranged B-lineage line - RS4;11 or SEM - which is this graph's index population** [E4]. That is a one-experiment gap at the intersection of two literatures that both already exist.

Go deeper
Why the original query could not see any of this

The refusal used (MLL[tiab] OR KMT2A[tiab]) AND HDAC6[tiab]. Every paper above fails that query for the same reason: they name the cell line, never the rearrangement. A medicinal chemistry paper reports MV4-11 IC50 = 28 nM in a potency table; it has no reason to write KMT2A or MLL in its title or abstract.

🔴 This corpus had already classified the defect. Its own bounded-absence audit found 23 of 39 bounded absences were ABSTRACT-LEVEL literature search, recorded as not a bound, and named a defect class of "entries that ask PubMed a question PubMed cannot answer." The HDAC6 refusal is a member of that class, and it was written anyway.

🔵 The fix is to search the reagent, not the disease. Replacing the disease terms with cell-line names and drug names turned 0 records into 15. The general rule: when the claim is "no agent has been tested in model X", query the agents and the model, because a paper about a compound is indexed under the compound.

🟠 Two false positives worth recording, because they would have inflated the result: SEM matched a duckweed paper and a catfish paper (PubMed expanded it to scanning electron microscopy and standard error), and THP-1 matched four inflammation papers using it as a monocyte model rather than a leukaemia. 9 of the 15 hits were discarded on reading the titles.

Facts and figures
uniprotQ9UBN7
gene symbolHDAC6
classclass IIb histone deacetylase
principal substratealpha-tubulin (cytoplasmic), not histones
selective inhibitor existstrue
shown for fusionfalse
fusion specificfalse
interaction with fusion complexNOT ESTABLISHED - no interaction data in any source found
index population testedfalse
cell typetested in myeloid and biphenotypic KMT2A-rearranged lines; index population is infant B-ALL
localizationpredominantly cytoplasmic
prior status in this graphDECLINED ON PURPOSE, on a [tiab] query returning 0 records
overturnsthe HDAC6 clause of ana-no-hdac-inhibitor-result-in-this-disease-attributes-to-an-enzyme
papers previously cited in corpus0
query that produced the refusal(MLL[tiab] OR KMT2A[tiab]) AND HDAC6[tiab] -> 0 records
query that overturned it(tubastatin OR ricolinostat OR ACY-1215 OR citarinostat OR ACY-241 OR nexturastat OR "HDAC6 inhibitor" OR "HDAC6-selective") AND (MOLM-13 OR MV4-11 OR RS4;11 OR THP-1 OR NOMO-1) -> 15 records, 6 on-beat after title reading
ptg0861 selectivity vs clinical standards~4x more isozyme-selective than citarinostat and ricolinostat per PMID 32615502
ptg0861 tolerabilitylimited cytotoxicity against non-malignant NHF and HUVEC cells and against CD-1 mice (abstract-level)
measurements
HDAC6-selective agents te…0.000HDAC6-selective agents tested in a KMT2A-rearranged B-lineage line: 0antiproliferative IC50 of…28.0antiproliferative IC50 of a selective HDAC6 inhibitor in MV4-11 (KMT2A::AFF1): 28isozyme selectivity of PT…36.0isozyme selectivity of PTG-0861 for HDAC6: 36
selective inhibitors named
[
  "PTG-0861 (JG-265)",
  "WT161",
  "tubastatin A",
  "ricolinostat (ACY-1215)",
  "citarinostat (ACY-241)",
  "tubacin",
  "nexturastat"
]
kmt2a rearranged lines tested
[
  "MV4-11 (KMT2A::AFF1, myeloid/biphenotypic)",
  "MOLM-13 (KMT2A::MLLT3)",
  "THP-1 (KMT2A::MLLT3)"
]
kmt2a rearranged b lineage lines tested
[]
papers newly cited
[
  "32615502",
  "41701111",
  "26443078",
  "39067437",
  "22464548",
  "40370107",
  "38148433",
  "25688158",
  "41249257",
  "29643474",
  "41484929"
]
measurements — the numbers behind the chart
[
  {
    "quantity": "antiproliferative IC50 of a selective HDAC6 inhibitor in MV4-11 (KMT2A::AFF1)",
    "value": 28,
    "unit": "nM",
    "subject": "artemisinin-derived HDAC6-selective lead, HDAC6 enzymatic IC50 12 nM",
    "universe": "same compound: MOLM-13 89 nM (KMT2A::MLLT3), HL60 92 nM (not KMT2A-rearranged)"
  },
  {
    "quantity": "isozyme selectivity of PTG-0861 for HDAC6",
    "value": 36,
    "unit": "fold",
    "subject": "PTG-0861 (JG-265), HDAC6 IC50 5.92 nM, active against MV4-11 and MM1S",
    "universe": "limited cytotoxicity reported against non-malignant NHF and HUVEC cells"
  },
  {
    "quantity": "HDAC6-selective agents tested in a KMT2A-rearranged B-lineage line",
    "value": 0,
    "unit": "reports",
    "subject": "RS4;11 and SEM, the index-population lines",
    "universe": "abstract-level PubMed bound, 2026-09-24; agents have been tested in MV4-11, MOLM-13, BALL-1 and NALM6"
  }
]
How this could bite us back
  • Do not read this as a therapeutic proposal. Every result here is antiproliferative potency in cell lines, the same compounds are comparably potent in leukaemias that carry no KMT2A rearrangement, and there is no evidence that HDAC6 has any relationship to the fusion complex. The honest summary is that an absence claim was wrong, not that a target was found.
  • The lineage does not match the patients. MV4-11 and MOLM-13 are myeloid or biphenotypic; this graph's index population is infant KMT2A::AFF1 B-ALL, and this corpus already records that results in THP-1 and MV4-11 do not transfer to infant B-lineage disease.
  • The best-powered isoform comparison in the right lineage points elsewhere. In a B-ALL panel with shRNA validation, HDAC1 and HDAC2 were the actionable isoforms, and a trial rationale built on HDAC6 would have to explain that result rather than ignore it.
  • HDAC6-selective agents are not free of toxicity in patients despite the 'selective' label. Ricolinostat and citarinostat have been through clinical evaluation in haematological malignancy, and any claim of a favourable therapeutic window here rests on in vitro non-malignant comparators (NHF, HUVEC), not on clinical data in children.
Still open
  1. Test an HDAC6-selective agent in RS4;11 and SEM against a non-KMT2A B-ALL comparator. This is the single missing cell in a 2x2 that two existing literatures already half-fill, and it is the only version of the experiment that speaks to the index population.
  2. Ask whether HDAC6 sensitivity survives an isogenic fusion-on/fusion-off comparison. Every result here is cross-line, and this graph has been wrong before by reading a cross-line potency difference as fusion-specificity.
  3. Determine whether HDAC6 has any physical or functional relationship to the fusion complex at all. Its substrate is cytoplasmic tubulin, so the prior should be low, and the honest position is that no interaction evidence was found rather than that none exists.
  4. Re-run the other bounded absences in the parent entry with reagent-and-model queries. HDAC6 was overturned by changing the query, and the same [tiab] pattern bounds several sibling claims.
Every source behind this entry (4)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.journal-article

    At least four HDAC6-selective inhibitor series have been tested against MV4-11, a KMT2A::AFF1 line: PTG-0861 (JG-265) has HDAC6 IC50 5.92 nM with ~36-fold isozyme selectivity and relevant potency against MV4-11, and is reported as ~4x more selective than the clinical standards citarinostat and ricolinostat; an artemisinin-hybrid lead is a potent selective HDAC6 inhibitor (IC50 12 nM) with antiproliferative IC50 of 28 nM in MV4-11 and 89 nM in MOLM-13 (KMT2A::MLLT3) and antileukaemic activity in a MOLM-13-Luc xenograft. None of these papers is cited in this corpus and no HDAC6-selective agent is named in any node file.

    PTG-0861: A novel HDAC6-selective inhibitor as a therapeutic strategy in acute myeloid leukaemia · Gawel JM, Shouksmith AE, Raouf YS, Nawar N, Toutah K, Bukhari S, Manaswiyoungkul P, Olaoye OO, Israelian J, Radu TB, Cabral AD, Sina D, Sedighi A, de Araujo ED, Gunning PT · Eur J Med Chem 201:112411 2020
    PMID 32615502DOI
    What this evidence cannot show

    Both abstracts were fetched and read via PubMed on 2026-09-24; full texts were not obtained, so the MV4-11 IC50 for PTG-0861 is described in its abstract only as 'relevant potency' without a number, and only the artemisinin paper gives explicit MV4-11 and MOLM-13 IC50 values. The artemisinin abstract renders compound numbers as blanks in the PubMed record, so the specific lead compound cannot be named from this source. MV4-11's KMT2A::AFF1 status is taken from this graph's own drug-chidamide entry, not re-verified against a cell-line authority this cycle. The author list and volume for PMID 32615502 were re-fetched from PubMed and corrected after a first draft of this entry attributed it to the wrong first author, carried over from an adjacent paper in the same result set.

    Evidence that argues against this (2)
    • Antiproliferative potency in a cell line is not evidence of a molecular relationship with the fusion. This graph has repeatedly recorded the failure mode in which a real biochemical or pharmacological signal in KMT2A-rearranged cells turns out not to be fusion-specific - KAT8/MOF being the worked example - and nothing here distinguishes HDAC6 sensitivity in MV4-11 from HDAC6 sensitivity in any proliferating leukaemia line.
    • The same compounds are similarly potent outside KMT2A-rearranged disease: the artemisinin lead gives IC50 92 nM in HL60, which carries no KMT2A rearrangement, and PTG-0861 is reported against MM1S multiple myeloma. A 3-fold window between MV4-11 and HL60 is not selectivity.
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.journal-article

    The HDAC6 sensitivity reported in KMT2A-rearranged lines is not fusion-specific: the artemisinin-derived selective HDAC6 inhibitor gives antiproliferative IC50 values of 28 nM in MV4-11 and 89 nM in MOLM-13, both KMT2A-rearranged, but also 92 nM in HL60, which is not KMT2A-rearranged, and PTG-0861 is active against MM1S multiple myeloma alongside MV4-11.

    PMID 41701111DOI
    What this evidence cannot show

    This is a cross-line comparison assembled from potency values in a single medicinal-chemistry abstract, not a designed fusion-specificity experiment; the lines differ in doubling time, p53 status and lineage, any of which moves a 3-fold IC50 window. HL60 is p53-null while the paper reports p53-dependent apoptosis, so the comparison may be confounded by mechanism rather than by fusion status. No isogenic fusion-on/fusion-off comparison exists for any HDAC6 agent.

    Evidence that argues against this (1)
    • A 3-fold IC50 window is within the range that a genuine but modest selectivity could produce, so 'not fusion-specific' is the conservative reading rather than a demonstrated negative. Refuting fusion-specificity properly needs an isogenic or degron system, which no HDAC6 paper has run.
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.journal-article

    A dedicated isoform-specificity study in B-lineage ALL concluded that HDAC1 and HDAC2 are the actionable isoforms: across a B-ALL cell line panel with isoform-selective inhibitors and isoform-specific shRNA validation, targeting HDAC1-3 was sufficient to inhibit growth, knockdown of HDAC1 or HDAC2 individually inhibited growth, and a compound specific for HDAC1 and HDAC2 suppressed growth and induced apoptosis in vitro and in vivo.

    Selective Inhibition of HDAC1 and HDAC2 as a Potential Therapeutic Option for B-ALL · Stubbs MC, Kim W, Bariteau M, Davis T, Vempati S, Minehart J, Witkin M, Qi J, Krivtsov AV, Bradner JE, Kung AL, Armstrong SA · Clin Cancer Res 21(10):2348-58 2015
    PMID 25688158DOI
    What this evidence cannot show

    The abstract does not state how many of the B-ALL panel lines were KMT2A-rearranged, so this is evidence about B-ALL as a class rather than about KMT2A-rearranged B-ALL specifically; the paper's MeSH terms include Histone Deacetylase 6, but whether HDAC6 was tested with a selective agent or only profiled cannot be determined from the abstract. Full text was not obtained.

    Evidence that argues against this (1)
    • This study is from 2015 and predates most HDAC6-selective chemistry, including PTG-0861 (2020) and WT161 as applied to ALL (2025), so its silence on HDAC6 may reflect the tool compounds available at the time rather than a tested negative. This graph has a standing rule against publishing a negative that is really a fact about a weak reagent.
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.journal-article

    HDAC6-selective agents have been tested in KMT2A-rearranged myeloid and biphenotypic lines and in non-KMT2A B-lineage ALL lines, but a targeted search returns no report of an HDAC6-selective agent tested in a KMT2A-rearranged B-lineage line such as RS4;11 or SEM: the selective HDAC6 inhibitor WT161 was evaluated in B-ALL lines BALL-1 and NALM6 and T-ALL lines Jurkat and MOLT-4, none of which is KMT2A-rearranged.

    PMID 41249257DOI
    What this evidence cannot show

    This is an absence claim bounded by PubMed searches run on 2026-09-24 over title/abstract/all-fields with drug names, isoform names and cell-line names; both queries returned non-empty query translations, so the zero is a genuine zero for that query and not an instrument failure. It remains an abstract-level bound - exactly the class this corpus records as NOT a true bound - because a paper could test RS4;11 in a potency table without naming it in the abstract. Treating this as settled would repeat the error this entry documents. The KMT2A status of BALL-1 and NALM6 was not re-verified against a cell-line authority this cycle.

    Evidence that argues against this (1)
    • The gap may be uninteresting rather than unexplored. If HDAC6 sensitivity is lineage-independent and not fusion-specific, as E2 suggests, then testing RS4;11 would be expected to reproduce the MV4-11 result and would add little. The experiment is cheap, but its informativeness depends on a specificity claim that is not currently supported.
Connected to