The HDAC7 direction conflict is scaffold versus catalysis: removing the protein and blocking its catalysis are opposite interventions, because class IIa HDACs barely have any catalysis
also called HDAC7 induction versus knockdown · class IIa scaffold hypothesis · why two groups disagree about HDAC7
Two groups report opposite things about the same enzyme in the same infant leukaemia: one says put HDAC7 back, the other says take it away. The proposed explanation is that class IIa enzymes like HDAC7 barely work as enzymes at all - their real job is to hold other proteins in place - so deleting the protein and blocking its chemistry do completely different things, and the two experiments were never measuring the same intervention.
The same thing, for a scientist
**The conflict.** de Barrios et al.: HDAC7 is **low** in t(4;11) infant B-ALL, silenced by EZH2 and KMT2A::AFF1; low HDAC7 associates with poor response; **restoring** it via MI-538 plus chidamide differentiates leukaemic pro-B cells and impairs proliferation in an HDAC7-dependent manner [E1]. Verbeek et al.: shRNA **knock-down** of HDAC4, HDAC5 and HDAC7 induces **apoptosis and cell-cycle arrest** in KMT2A-rearranged ALL cells [E2]. Same gene, same disease, opposite sign. **The hypothesis.** Class IIa HDACs (HDAC4, 5, 7, 9) carry a **tyrosine-to-histidine substitution in the catalytic pocket that renders them essentially inactive against acetyl-lysine**; their established function is as **adaptors** - binding MEF2 family transcription factors through an N-terminal domain and recruiting the catalytically competent NCoR/SMRT-HDAC3 complex. If that is the operative function here, then: **shRNA removes the adaptor** (losing MEF2 repression, losing HDAC3 recruitment, and freeing MEF2 sites), while **re-expression restores the adaptor**, and **a catalytic inhibitor such as LMK-235 does neither** - it blocks a catalysis that was never doing much. The three experiments are three different perturbations, and the literature treats them as one. **This predicts the in vivo failure too.** LMK-235, MC1568 and TMP195 all failed at maximum achievable dose in a KMT2A-rearranged ALL xenograft [E2], which the authors attributed to compound stability. The scaffold hypothesis offers a rival explanation that costs nothing to test: **the compounds reached the target and inhibited the wrong function.** The HDAC9 result supports the frame - HDAC9's KMT2A signature travels with **MEF2D**, its partner transcription factor, and HDAC9's classical alias is **MITR, 'MEF2-interacting transcription repressor'** [E3].
What each experiment actually perturbs
| experiment | adaptor function | catalytic function | MEF2 sites | HDAC3 recruitment |
|---|---|---|---|---|
| shRNA knockdown 2 | removed | removed | freed | lost |
| re-expression / induction 1 | restored | restored | re-repressed | restored |
| LMK-235 / MC1568 / TMP195 2 | intact | blocked | still occupied | still recruited |
Read across the rows. The knockdown and the induction are genuine opposites and should give opposite phenotypes - which they do. The inhibitor is not the opposite of either; it is a fourth thing, and on this hypothesis it is close to a no-op on the function that matters. That is a complete account of all three published observations, including the in vivo null, without any of them being wrong.
It also explains a detail that otherwise looks like noise: the knockdown result was reported for HDAC4, HDAC5 and HDAC7 together, and all three are adaptors for the same MEF2/NCoR axis. A class-level adaptor loss is a coherent perturbation in a way that a class-level catalytic inhibition is not.
Why this matters beyond one gene
If the hypothesis holds, three things follow that reach past HDAC7.
- The field's response to pan-HDAC toxicity is aimed wrong for class IIa. More selective catalytic inhibitors will not deliver the phenotype the genetics predicts. The right modality is a degrader, which removes the scaffold - and no class IIa degrader has been tested in this disease.
- Two of this graph's instruments become suspect for this class. A DepMap or CRISPR score removes the protein (adaptor-relevant); a drug-sensitivity dataset blocks catalysis (adaptor-irrelevant). They are measuring different things and should not be read as corroborating each other.
- It sharpens the menin-inhibitor story rather than threatening it. The induction arm's mechanism runs through menin inhibition - MI-538 de-represses HDAC7 1. So the hypothesis is compatible with, and would extend, the menin-interface position: menin inhibition would have a lineage-differentiation consequence in the lymphoid index disease that is currently invisible in a graph that describes menin inhibitors almost entirely through HOXA/MEIS1 eviction.
What would make this hypothesis wrong in an interesting way: if HDAC7's catalytic activity turns out to be non-trivial at the relevant substrates. Class IIa enzymes do have measurable activity against trifluoroacetyl-lysine substrates and there is a live argument that their physiological substrates are not acetyl-lysine at all. If the relevant mark is a different acyl group, the scaffold/catalysis dichotomy is too clean and the real answer is a substrate question.
| statement | The opposite published directions on HDAC7 in KMT2A-rearranged ALL are not a contradiction but a difference in what was perturbed: shRNA and re-expression move the class IIa ADAPTOR function (MEF2 binding and NCoR/SMRT-HDAC3 recruitment), while a catalytic inhibitor blocks a catalysis that class IIa enzymes barely possess. The three experiments are three different interventions and the literature reads them as one. |
|---|---|
| origin | literature-gap |
| prior probability | 0.4 |
| cheapest first step | Read the full text of PMID 40722046 (PMC12305908) and determine whether the HDAC7 rescue construct was catalytically active. If a catalytic-dead construct was used and worked, the hypothesis is substantially supported before any new experiment is run; if it was used and failed, the hypothesis is dead. |
| what it would change | The modality. If correct, class IIa in this disease needs a degrader rather than a better inhibitor, and the three published in vivo failures are explained by the wrong modality rather than by poor pharmacokinetics. |
| relationship to the menin hypothesis | Supportive and extending, not threatening. The induction arm's mechanism runs THROUGH menin inhibition, so this would give menin inhibitors a second, lineage-differentiation consequence in the lymphoid index disease that the graph currently does not describe. |
predictions
[ "A catalytic-dead HDAC7 knock-in or rescue construct will reproduce the benefit of HDAC7 re-expression in t(4;11) infant B-ALL, because the benefit is adaptor-mediated.", "A class IIa DEGRADER will phenocopy the shRNA knockdown (apoptosis, cell-cycle arrest) where LMK-235, MC1568 and TMP195 did not.", "LMK-235's in vitro activity against KMT2A-rearranged cells will show class I off-target engagement at the concentrations used, measurable as global histone hyperacetylation that a genuinely class IIa-selective perturbation would not produce.", "MEF2D or MEF2C ChIP occupancy will change on HDAC7 or HDAC9 depletion and will NOT change appreciably on class IIa catalytic inhibition.", "Menin inhibition with a CLINICAL agent (revumenib, ziftomenib, bleximenib, enzomenib) will restore HDAC7 expression as MI-538 does, making HDAC7 mRNA a pharmacodynamic biomarker for menin inhibition in the lymphoid disease." ]
experiments
[ "exp-catalytic-dead-versus-degrader-test-of-class-iia-hdac-function-in-kmt2a-r-all" ]
analyses
[ "ana-no-hdac-inhibitor-result-in-this-disease-attributes-to-an-enzyme" ]
falsifiers
[ "A catalytic-dead HDAC7 rescue construct FAILS to reproduce the benefit of wild-type HDAC7 re-expression. That would show the catalysis is the operative function and kills this hypothesis outright.", "A class IIa degrader fails to phenocopy the shRNA knockdown. The adaptor-removal prediction is then wrong.", "LMK-235 shows no class I off-target engagement at its in vitro active concentrations, leaving its cell-killing to be explained by genuine class IIa catalytic inhibition - which this hypothesis says should be close to a no-op.", "HDAC7 is shown to have substantial catalytic activity against a physiologically relevant acyl-lysine substrate in these cells, making the scaffold/catalysis dichotomy the wrong frame.", "The de Barrios rescue experiments already used a catalytic-dead construct and it failed. The full text was not read in this beat and this is the cheapest possible falsification - it may already be published." ]
- If this hypothesis is right, the correct modality is a degrader - and degrading a MEF2 adaptor in an infant removes a transcriptional brake in cardiac and skeletal muscle and in neurons, where MEF2 derepression is the established toxicity of class IIa loss. The hypothesis makes the biology clearer and the safety worse.
- The hypothesis is an explanation of a contradiction, which is the most seductive and least reliable kind of hypothesis. It was constructed after seeing both results, it fits both by design, and it has not predicted anything yet. The catalytic-dead rescue is the only reason to believe it and it has not been done here.
- Its weakest point is stated rather than hidden: LMK-235 does eradicate KMT2A-rearranged cell lines and primary patient samples in vitro, which a pure scaffold model does not predict. Explaining that away as class I off-target engagement is an unfalsified convenience until someone measures it.
- If the hypothesis is wrong and catalysis is the operative function, then the three in vivo failures mean class IIa inhibition genuinely does not work in this disease, and an effort redirected toward degraders would waste years on a target that the available evidence already rejected.
- Was the de Barrios HDAC7 rescue construct catalytically active? The answer may already be in the full text and would move this hypothesis a long way in either direction for the cost of one read.
- Is there any published class IIa HDAC degrader, and has it been tested in any leukaemia?
- If class IIa HDACs act through MEF2, does the KMT2A-rearranged MEF2D overexpression reported alongside high HDAC9 mean the adaptor axis is amplified in this disease specifically?
- Do HDAC7 and HDAC9 compete for the same MEF2 sites? If so, HDAC7 low and HDAC9 high would be one coherent state rather than two separate observations.
- What are class IIa HDACs' real physiological substrates? If they are not acetyl-lysine, the whole acetylation framing of this sub-class - including its place in this beat - is misfiled.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
HDAC7 is epigenetically silenced by EZH2 and the KMT2A::AFF1 fusion protein in t(4;11) infant B-ALL where low HDAC7 associates with poor therapeutic response; the menin inhibitor MI-538 restores HDAC7 expression, enhanced by chidamide; and the treatment drives leukaemic pro-B cells toward a more differentiated state and impairs aberrant proliferation in an HDAC7-dependent manner.
HDAC7 induction combined with standard-of-care chemotherapy provides a therapeutic advantage in t(4;11) infant B-cell acute lymphoblastic leukemia. · de Barrios O, Ocon-Gabarro I, Gusi-Vives M, Collazo O, Meler A, Romecin PA, et al. · Biomarker Research 13(1):99 2025What this evidence cannot show
The therapeutic effect is stated to be only PARTIALLY mediated by HDAC7. MI-538 is a tool menin inhibitor. Whether the rescue construct used was catalytically active, and whether a catalytic-dead HDAC7 would rescue, is not stated in the abstract - and that single detail is the most direct test of this hypothesis. Abstract read; PMCID PMC12305908, full text not fetched in this beat, so the rescue-construct question may already be answered in the paper.
Evidence that argues against this (1)
- The opposing report is the reason this hypothesis exists and is its primary counter-evidence: shRNA knock-down of HDAC4, HDAC5 and HDAC7 induces apoptosis and cell-cycle arrest in KMT2A-rearranged ALL cells, which on a naive reading means the gene is required rather than protective. Verbeek TCAI et al. 2024
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
shRNA-mediated knock-down of the class IIA HDAC isoforms HDAC4, HDAC5 and HDAC7 results in apoptosis induction and cell cycle arrest in KMT2A-rearranged ALL cells, and the HDAC4/5-selective inhibitor LMK-235 eradicates KMT2A-rearranged ALL cell lines and primary patient samples in vitro; but the maximum achievable dose of LMK-235 was insufficient to induce anti-leukaemic effects in a KMT2A-rearranged ALL xenograft mouse model, with similar results for the class IIA inhibitors MC1568 and TMP195, and LMK-235 plus venetoclax did not prolong survival.
Selective inhibition of HDAC class IIA as therapeutic intervention for KMT2A-rearranged acute lymphoblastic leukemia. · Verbeek TCAI, Vrenken KS, Arentsen-Peters STCJM, Castro PG, van de Ven M, van Tellingen O, et al. · Communications Biology 7(1):1257 2024What this evidence cannot show
The in vitro LMK-235 result is a problem for the strong form of this hypothesis, not merely a neutral datum: if catalytic inhibition were a near-no-op on the operative function, LMK-235 should not eradicate cell lines and primary samples in vitro. The hypothesis must therefore accommodate an in vitro catalytic effect that does not translate, which is a weaker and less elegant position than the one-line version. The authors attribute the in vivo failure to compound stability and efficiency, and no target-engagement measurement distinguishes their explanation from this one. Abstract read; PMCID PMC11450098, full text not fetched.
Evidence that argues against this (1)
- LMK-235 eradicating KMT2A-rearranged cell lines and primary patient samples in vitro is direct evidence that class IIa CATALYTIC inhibition does something substantial in these cells, which the scaffold hypothesis in its simplest form does not predict. High-concentration in vitro activity of a hydroxamate could equally be off-target on class I enzymes, and that alternative is untested. Verbeek TCAI et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The KMT2A-rearrangement HDAC signature in 211 paediatric patients is high HDAC9 together with high MEF2D, and HDAC9's UniProt record (Q9UKV0) carries MITR - MEF2-interacting transcription repressor - among its names, consistent with class IIa HDACs acting through MEF2 binding.
Paediatric patients with acute leukaemia and KMT2A (MLL) rearrangement show a distinctive expression pattern of histone deacetylases. · Vega-Garcia N, Malatesta R, Estella C, Perez-Jaume S, Esperanza-Cebollada E, Torrebadell M, et al. · British Journal of Haematology 182(4):542-553 2018What this evidence cannot show
The HDAC9-MEF2D co-occurrence is an expression correlation in a patient cohort and is consistent with the MEF2-adaptor frame without demonstrating it; the paper does not test MEF2 binding. The MITR naming is a UniProt record fact (Q9UKV0, retrieved 2026-09-20) and is offered as corroborating context, not as evidence. The claim that class IIa HDACs are catalytically near-inactive against acetyl-lysine and act as NCoR/SMRT-HDAC3 adaptors is general enzymology background stated in this entry and is NOT established by this citation; a reader who needs it load-bearing should source it separately, and doing so is the first step of the analysis proposed below.
Evidence that argues against this (1)
- MITR's presence in the UniProt record establishes that a catalytically inactive HDAC9 form exists, not that it is the form the 211-patient signature measured: the assay was mRNA expression, which does not resolve splice forms unless designed to, and the paper does not claim to have done so. The scaffold-versus-catalysis reading is therefore a hypothesis the signature is consistent with, not one it supports. Verbeek TCAI et al. 2024
the hypothesis exists to reconcile the two opposite published directions on HDAC7
HDAC5 is one of the three class IIa isoforms knocked down together in the opposing report
HDAC9's KMT2A signature travels with MEF2D and its alias is MITR, consistent with the MEF2-adaptor frame
the adaptor model requires class IIa HDACs to recruit the catalytically competent NCoR/SMRT-HDAC3 complex
the wild-type versus catalytic-dead rescue comparison is the decisive test