🧬 Aging Wiki

hdac-inhibitors

Properties1
aliasesHDAC inhibitors, HDACi, histone deacetylase inhibitors, pan-HDAC inhibitors, HDAC6 inhibitors

12 min read

HDAC inhibitors (HDACi)

HDAC inhibitors are small molecules that block the active site of Zn2+-dependent histone deacetylases (Class I, IIa, IIb, and IV; hdac HDAC1–11), raising acetylation levels on histone tails and — equally importantly — on ~3,600 non-histone substrates including alpha-tubulin, HSP90, p53, and NF-κB p65. In aged tissues, global histone acetylation at stress-response and metabolic gene promoters declines; HDAC inhibition is hypothesized to partially reverse this drift. The class has five FDA-approved members, all oncology-indicated. No HDAC inhibitor has an approved aging or healthspan indication as of 2026-05-07, and translation of invertebrate lifespan results to mammals remains undemonstrated.

This is a class page. Compound-level PK, dosing, and individual trial data belong on dedicated compound pages when seeded. Mechanistic protein anchor: hdac (verified R31). NAD+-dependent Class III deacetylases (sirtuins) are a distinct family — see sirtuin and sirtuin-activators.

Chemotype overview

Five structural classes achieve HDAC inhibition through distinct binding modes 1:

ChemotypeExamplesSelectivityClinical status
Hydroxamic acidsVorinostat (SAHA), panobinostat, belinostat, trichostatin A (TSA)Pan-HDAC (Class I/II)Vorinostat, panobinostat, belinostat FDA-approved (oncology); TSA research probe only
Cyclic depsipeptidesRomidepsin (FK228, Istodax)Class I-selective (prodrug; thiol-activated)FDA-approved 2009 (CTCL + PTCL)
BenzamidesEntinostat (MS-275), chidamide (tucidinostat)Class I-selective (HDAC1/2/3)Entinostat Phase 3 failed (E2112, breast cancer, 2019); chidamide China NMPA-approved 2014 (PTCL)
Aliphatic acidsSodium butyrate, phenylbutyrate, valproateBroad / weak (Class I/IIa)Valproate FDA-approved (epilepsy); butyrate food-grade; not approved for aging
HDAC6-selectiveTubastatin A, ricolinostat (ACY-1215), citarinostat (ACY-241)HDAC6 >> Class IRicolinostat Phase I/II (myeloma, CLL); no completed aging trial

FDA-approved agents

All five FDA-approved HDAC inhibitors target oncology indications — NOT aging-validated:

  • Vorinostat (SAHA, Zolinza) — hydroxamic acid; FDA-approved 2006 for cutaneous T-cell lymphoma (CTCL); the first approved HDAC inhibitor. Origin compound for the class pharmacology 1.
  • Romidepsin (Istodax, FK228) — cyclic depsipeptide prodrug; FDA-approved 2009 (CTCL) and 2011 (PTCL); Class I-selective after thiol activation.
  • Belinostat (Beleodaq) — hydroxamic acid; FDA-approved 2014 (PTCL).
  • Panobinostat (Farydak) — hydroxamic acid; FDA-approved 2015 (multiple myeloma, conditional); conditional approval withdrawn EMA 2020 after confirmatory trial failed.
  • Chidamide (Tucidinostat, Epidaza) — benzamide; China NMPA-approved 2014 (PTCL); not FDA-approved.

HDAC6-selective inhibitors (primary aging-context candidates)

HDAC6 is a cytoplasmic deacetylase with two tandem catalytic domains; its primary substrates are alpha-tubulin (K40) and HSP90 rather than histones. HDAC6 inhibition increases tubulin acetylation, stabilizing microtubules and restoring axonal transport — a mechanism implicated in neurodegeneration and proteostasis 2. This selectivity profile separates HDAC6 inhibitors from the pan-HDAC toxicity pattern (cytopenias, GI) that limits oncology agents for healthy-aging use.

AgentStatusPrimary indication explored
Tubastatin APreclinicalNeurodegeneration models
Ricolinostat (ACY-1215)Phase I/II (completed, oncology)Multiple myeloma, CLL; no aging trial completed
Citarinostat (ACY-241)Phase I/II (oncology)Multiple myeloma; not aging-indicated

Aliphatic acids (endogenous HDAC inhibitors)

Sodium butyrate, phenylbutyrate, and valproate inhibit Class I/IIa HDACs weakly and non-selectively. Beta-hydroxybutyrate (BHB), produced during fasting and ketogenic diet at 1–2 mmol/L, is an endogenous Class I HDAC inhibitor — a proposed link between caloric restriction and epigenetic rejuvenation 2. See ketogenic-diet for the BHB-HDAC inhibition mechanism.

Mechanism of action

All approved HDAC inhibitors chelate the active-site Zn2+ ion in the HDAC catalytic domain (the DAC domain), blocking deacetylase activity 1. The primary aging-relevant consequences are:

  1. Histone hyperacetylation — H3K9ac, H3K14ac, H4K16ac increase at gene promoters, reopening chromatin at loci silenced with age. Stress-response, autophagy, and metabolic gene networks that are hypo-acetylated in aged tissues are the intended targets 2.
  2. Non-histone substrate effects — see hdac for the full substrate landscape. Key aging-relevant non-histone effects:
    • HDAC6 inhibition → alpha-tubulin K40 hyperacetylation → microtubule stabilization → restored axonal transport → improved proteostasis and autophagosome flux
    • HDAC1 inhibition → OGG1 hyperacetylation → impaired BER AP-lyase activity (note: HDAC1 activation, not inhibition, rescued OGG1 in the Pao 2020 brain-aging context — this is a nuance where HDAC1 inhibition would be counterproductive; see below)
    • HDAC3 inhibition → NF-ÎşB p65 K310 hyperacetylation → altered inflammatory SASP gene output

Important nuance — HDAC1 and DNA repair: Pao et al. 2020 showed that HDAC1 deacetylates OGG1 to stimulate its AP-lyase DNA repair activity in aged neurons 3. In this context, HDAC1 inhibition would hyperacetylate OGG1 and impair BER — the opposite of the intended effect. This illustrates that pan-HDAC inhibition in the brain may have complex substrate-specific consequences not captured by the “histone hyperacetylation = good” heuristic. contradictory-evidence

Aging biology evidence

Invertebrate lifespan extension

Pharmacological HDAC inhibition extends lifespan in multiple invertebrate models 2:

  • Drosophila melanogaster: Vorinostat (SAHA) administered during mid- and late-life extended longevity in Canton-S and Oregon-R flies 4. Sodium butyrate and phenylbutyrate also extend fly lifespan (Kang et al. 2002; Zhao et al. 2005 — see 2).
  • C. elegans: Valproate, trichostatin A (TSA), and beta-hydroxybutyrate extend lifespan in multiple studies, in part via hda (HDAC) knockdown phenocopying — consistent with epigenetic derepression of stress-response genes 2.
DimensionStatus
Pathway conserved in humans?yes — Zn2+-dependent HDACs are structurally conserved; histone acetylation patterns are conserved across eukaryotes
Phenotype conserved in humans?unknown — invertebrate lifespan extension not replicated in mammals
Replicated in humans?no — no aging-endpoint human trial

Caveats: (1) Invertebrate lifespan has not been extended by selective HDAC inhibitors in mammals under controlled ITP-style conditions. (2) Short-chain fatty acids (butyrate, valproate) act on multiple targets beyond HDACs (GPR41/43, mTOR, mitochondria); HDAC-specificity of the lifespan effects is not established. (3) Doses required for HDAC inhibition in flies may not translate to safe chronic dosing in mammals. needs-human-replication needs-replication

Mammalian aging models

The Pao 2020 study established a specific HDAC1 pro-repair role in aged mouse neurons: exifone (a P/Q-HDAC1 activator, 50 mg/kg/day IP, 4 weeks) restored OGG1 AP-lyase activity in 17-month-old C57BL/6J mice and improved contextual fear conditioning and hippocampal LTP in 8-month-old 5XFAD AD-model mice 3. Critically, this is HDAC1 activation, not inhibition — a mechanism that runs opposite to the canonical HDAC inhibitor class. The 5XFAD behavioral rescue arm was AD-model mice, not aged WT.

No pan-HDAC inhibitor has been shown to extend lifespan in mice under controlled multi-center conditions (NIA ITP has not tested any HDAC inhibitor). needs-human-replication

HDAC6 and neurodegeneration models

Tubastatin A and ricolinostat improve tubulin acetylation and axonal transport deficits in Alzheimer’s and Parkinson’s disease mouse models 5. The mechanism — HDAC6 inhibition → tubulin hyperacetylation → restored microtubule stability → autophagosome trafficking — connects to the loss-of-proteostasis and disabled-macroautophagy hallmarks. These are preclinical findings; no clinical neurodegeneration trial using HDAC6-selective inhibitors has reported aging-relevant outcomes. needs-human-replication

Human evidence

No HDAC inhibitor has an approved aging, healthspan, or geroprotective indication. Human evidence is confined to:

  1. Oncology Phase 2/3 trials (FDA-approved agents) — establish safety and PK profiles useful for extrapolation, but not aging endpoints.
  2. Entinostat E2112 Phase 3 failure (2019) — entinostat + exemestane failed to improve overall survival vs. exemestane alone in postmenopausal hormone-receptor-positive, HER2-negative advanced breast cancer. This was an oncology failure, but it illustrates Class I-selective HDAC inhibitor limitations in a human setting.
  3. NCT07404085 (EPIC trial) — a single recruiting trial as of 2026-05-07: sodium butyrate (dietary supplement; aliphatic acid HDACi) + virtual-reality cognitive training in 40–75-year-old adults with remitted mood disorders (bipolar disorder or major depressive disorder; n=160 estimated). Primary endpoint: global cognitive composite score at 3 weeks. This is the only active aging-adjacent HDAC inhibitor trial identified; it is not a longevity or healthspan primary endpoint trial, and it uses a non-selective aliphatic acid rather than a clinical-grade HDAC inhibitor drug.
  4. Ricolinostat (ACY-1215) oncology trials — Phase I/II in myeloma (NCT02787369, active not recruiting) and CLL; not aging-indicated.

clinical-trials-active: 1 reflects the EPIC trial (NCT07404085, sodium butyrate arm) as the sole active aging-adjacent registered trial (ClinicalTrials.gov v2 API search, 2026-05-07; queries: “HDAC inhibitor aging”, “histone deacetylase inhibitor older adults”, “HDACi longevity”). Note: NCT02787369 (ricolinostat in CLL; ACTIVE_NOT_RECRUITING) is oncology-only and excluded from this count. long-term-unknown

Safety profile and translation barriers

Oncology doses are incompatible with aging prophylaxis. At therapeutic oncology doses, pan-HDAC inhibitors cause:

  • Thrombocytopenia and neutropenia (cytopenias — dose-limiting)
  • GI toxicity (nausea, diarrhea, fatigue)
  • QTc prolongation (especially panobinostat)
  • Increased infection risk

These class toxicities reflect the broad epigenetic impact of pan-HDAC inhibition on rapidly dividing hematopoietic and GI cells. At the chronic low doses required for aging prevention, the safety profile is unknown. long-term-unknown

HDAC6-selective inhibitors (ricolinostat, tubastatin A) avoid the histone-disruption toxicity of pan agents by targeting the cytoplasmic tubulin/HSP90 substrates rather than nuclear chromatin broadly. This selectivity makes them the most credible aging-indication candidates, but chronic safety data in healthy older adults remains entirely absent.

Aliphatic acids (sodium butyrate, valproate, BHB at nutritional ketosis levels) are the most tolerable chronic HDAC inhibitors, but their HDAC selectivity is poor and their in-vivo HDAC-specific contribution to any aging benefit is unestablished.

Safety tierAgentsRationale
Incompatible with aging prophylaxisVorinostat, panobinostat, romidepsin, belinostatOncology-dose cytopenias + GI; chronic QTc risk
Investigational (requires safety profiling)Entinostat, ricolinostat, citarinostatClass I / HDAC6 selective; Phase 1/2 data in cancer patients; chronic aging-dose safety not established
Tolerated but non-specificSodium butyrate, valproate, BHBBroad targets; poor HDAC class selectivity; chronic dietary exposure acceptable

Assessment: translation gap

HDAC inhibitors as an aging-intervention class face a stack of unresolved barriers:

  1. Species-translation gap: Invertebrate lifespan extension (flies, worms) has not been extended to mammals under controlled conditions. The translational distance from Drosophila to human is large for epigenetic interventions. needs-human-replication
  2. Selectivity problem: Pan-HDAC inhibitors change acetylation at thousands of loci simultaneously. Beneficial effects on aging-relevant loci may be outweighed by adverse epigenetic perturbations at oncogene promoters or immune gene networks. The chromatin-dose window for aging benefit vs. harm has not been characterized.
  3. HDAC1 brain nuance: The Pao 2020 finding that HDAC1 activation (not inhibition) rescues DNA repair in aged neurons suggests that the simplistic “HDAC inhibition → epigenetic rejuvenation” frame is incomplete. Tissue-specific and member-specific effects complicate class-level predictions.
  4. No aging-dose PK: Pharmacokinetics at sub-oncology doses in older adults are not established. Bioavailability, half-life, and tissue distribution at chronic low doses are unknown for most agents.
  5. No human aging biomarker data: No trial has reported whether any HDAC inhibitor moves epigenetic clock biomarkers (e.g., grimage-2019, dunedinpace-2022) in human aging.
ClassMechanismRelation to HDAC inhibitors
sirtuin-activatorsSIRT1–7 deacetylase activationClass III deacetylases; mechanistically distinct from Zn2+-HDACs; same acetylome substrates overlap
nad-precursorsNAD+ replenishment → SIRT1/3/6 activityIndirect support for sirtuin (Class III) activity; complementary to HDAC inhibition at different substrate sets
mtor-inhibitorsmTOR inhibition → autophagy inductionConvergence on autophagy: HDAC6 inhibition → tubulin acetylation → autophagosome flux; mTOR inhibition → ULK1 activation
caloric-restrictionEndogenous BHB → HDAC inhibitionBHB is an endogenous Class I HDAC inhibitor at nutritional ketosis; CR is a confounded HDAC inhibitor + mTOR inhibitor + AMPK activator
ketogenic-dietBHB production → HDAC inhibitionMost physiological chronic HDAC inhibitor; low selectivity

Limitations and open questions

  • No mammalian lifespan extension data. ITP has not tested any HDAC inhibitor. Invertebrate results do not establish mammalian aging efficacy. needs-human-replication
  • HDAC1 activation vs. inhibition in the brain. Pao 2020 shows that HDAC1 function is required for OGG1-mediated BER in aged neurons — pan-HDAC inhibitors would impair this protective mechanism. The right aging intervention for brain HDAC biology may be activation of HDAC1 or selective inhibition of other members, not pan-inhibition. contradictory-evidence
  • Epigenetic clock response unknown. Whether any HDAC inhibitor reverses DNA methylation age in human tissues has not been tested in a controlled trial. needs-human-replication
  • Chronic safety at sub-oncology doses entirely uncharacterized. long-term-unknown
  • Dosing for aging: no established target. The dose that achieves histone hyperacetylation at aging-relevant loci without off-target chromatin disruption in healthy adults is unknown. dose-response-unclear
  • HDAC6-selective candidates are the most tractable but have no completed aging-indication trial data beyond preclinical models.

Cross-references

  • hdac — verified protein family page (R31); canonical source for class taxonomy, substrate biology, mechanistic claims, and pharmacology table
  • epigenetic-alterations — primary targeted hallmark
  • loss-of-proteostasis — secondary target (HDAC6 → tubulin acetylation → axonal transport → proteostasis)
  • disabled-macroautophagy — tertiary target (HDAC6 → autophagosome trafficking)
  • sirtuin — Class III deacetylases; mechanistically distinct but substrate-overlapping
  • sirtuin-activators — sibling pharmacological class
  • caloric-restriction — lifestyle intervention with overlapping HDAC-inhibitor mechanism via BHB
  • ketogenic-diet — endogenous BHB-mediated HDAC inhibition
  • scfa-signaling — gut-derived butyrate as endogenous HDAC inhibitor; microbiome link to epigenome
  • cbp-p300 — the HAT counterpart to HDACs; the acetylation balance is the core regulatory axis
  • p53 — HDAC1/SIRT1 deacetylate p53 K382; relevant to senescence threshold

Footnotes

Footnotes

  1. doi:10.1038/sj.onc.1210204 · Marks PA · Oncogene 2007 · review · ~537 citations · Discovery and development of SAHA (vorinostat) as the first approved HDAC inhibitor; covers hydroxamic acid chemotype, mechanism of Zn2+-chelation, and early clinical rationale · archive: not_oa ↩ ↩2 ↩3

  2. doi:10.15252/emmm.201809854 · McIntyre RL, Daniels EG, Molenaars M, Houtkooper RH, Janssens GE · EMBO Molecular Medicine 2019 · review · 112 citations · HDAC inhibitors in aging — covers lifespan extension in C. elegans (valproate, BHB via hda RNAi) and Drosophila (sodium butyrate, phenylbutyrate, TSA, vorinostat/SAHA); preclinical results in neurodegeneration, cardiac, and metabolic aging · archive: downloaded (gold OA); PDF verified against hdac.md R31 ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  3. doi:10.1038/s41467-020-16361-y · Pao PC et al. · Nature Communications 2020 · in-vivo · model: Hdac1 cKO (Nestin-Cre), 5XFAD AD model, aged (17-month) C57BL/6J mice · 199 citations · HDAC1 deacetylates OGG1 (p300-acetylated lysines), stimulating AP-lyase BER activity; HDAC1 loss → elevated 8-oxoG at gene promoters; exifone (50 mg/kg/day IP, 4 wk) rescued OGG1 activity in 17-mo WT mice and improved fear conditioning + LTP in 8-mo 5XFAD mice; note: HDAC1 activation is the mechanism — the opposite direction from pan-HDAC inhibition · archive: downloaded (gold OA); PDF verified against hdac.md R31 ↩ ↩2

  4. doi:10.1016/j.exger.2012.09.006 · McDonald P, Maizi BM, Arking R · Experimental Gerontology 2013 · in-vivo · model: Drosophila melanogaster · 35 citations · Vorinostat (SAHA) extended mid- and late-life longevity in Canton-S and Oregon-R flies; mechanism attributed to HDAC inhibition and altered histone acetylation · archive: download failed (green OA); drug identity (vorinostat/SAHA) confirmed via McIntyre 2019 Table 1; verified against hdac.md R31 ↩

  5. doi:10.3233/JAD-140066 · Zhang L et al. (11 authors; last/corresponding: Qin C) · Journal of Alzheimer’s Disease 2014 · in-vivo · model: Alzheimer’s disease transgenic mice · title: “Tubastatin A/ACY-1215 Improves Cognition in Alzheimer’s Disease Transgenic Mice” — both agents tested; HDAC6 inhibition → tubulin hyperacetylation → microtubule stabilization mechanism · archive: not_oa (closed-access; full text not verified) · no-fulltext-access needs-replication — single preclinical study in AD model, not aged WT mice; exact n, strain, doses, and quantitative outcomes not independently verified ↩

Graph View

Backlinks