HDAC isoform targeting offers a precision route to treating neurodegeneration

Not all histone deacetylases are villains in the brain, and a new review argues that treating them as one may explain why decades of drug development for Alzheimer’s, Parkinson’s, Huntington’s and motor neuron disease have stalled. Writing in Discover Neuroscience, researchers led by Mayowa Damilola Olorunsesan and Tomiwa Joshua Odeniyi set out a disease-stratified model in which a different HDAC isoform dominates the pathology of each condition, and in which broad-spectrum inhibitors that shut down every member of the enzyme family simultaneously may be doing more harm than good.

Histone deacetylases are epigenetic enzymes that remove acetyl groups from histone lysine residues, promoting chromatin compaction and transcriptional repression. When their activity becomes dysregulated, the review explains, genes that protect neurons are silenced: synaptic proteins, brain-derived neurotrophic factor, anti-apoptotic factors such as BCL-2, and antioxidant enzymes all fall under transcriptional repression. The consequences ripple outward through synaptic gene silencing, neuroinflammation, impaired proteostasis, mitochondrial dysfunction and oxidative stress, mechanisms that recur across the major neurodegenerative diseases.

The core of the review is a mapping of isoform to disease. In Alzheimer’s disease, HDAC2 drives nuclear repression of synaptic genes and memory loss; selective knockdown experiments in rodent models restored cognitive deficits without disrupting the homeostatic functions of the closely related HDAC1. In Parkinson’s disease and amyotrophic lateral sclerosis, the cytoplasmic deacetylase HDAC6 governs proteostasis, aggresome formation, clearance of misfolded proteins such as alpha-synuclein, and axonal transport dynamics. In Huntington’s disease, HDAC1 and HDAC3 amplify transcriptional repression through direct interaction with mutant huntingtin within corepressor complexes. Meanwhile SIRT1, a Class III, NAD+-dependent sirtuin, exerts broadly neuroprotective effects across all four conditions by deacetylating and activating PGC-1α, which promotes mitochondrial biogenesis and attenuates oxidative stress.

That division of labor is precisely what makes pan-HDAC inhibition pharmacologically incoherent, the authors argue. A broad-spectrum inhibitor that suppresses HDAC2 in Alzheimer’s may be beneficial, but the same drug would also inhibit SIRT1, a protective deacetylase whose activity is typically diminished in the Parkinson’s disease substantia nigra. In Parkinson’s, simultaneous inhibition of HDAC6, which would aid alpha-synuclein clearance, and SIRT1, which sustains mitochondrial biogenesis in dopaminergic neurons, risks net harm rather than net benefit. The review presents this therapeutic paradox as a mechanistic explanation for the consistent failure of pan-HDAC inhibitors in clinical trials for neurodegeneration.

The preclinical record is nonetheless substantial. In APP/PS1 mice, oral vorinostat increased H4 acetylation, improved memory and exerted anti-inflammatory effects, and in Tg2576 mice intraperitoneal vorinostat at 12.5 mg/kg for four weeks improved learning and reduced cortical pTau. Valproic acid at 300 mg/kg/day for eight weeks improved memory deficits and reduced amyloid accumulation in APP/PS1 mice, and suppressed microgliosis, astrogliosis and inflammatory cytokine expression in the hippocampus and cortex. Panobinostat alleviated symptoms in an EAE mouse model, entinostat improved motor function in a rat Parkinson’s model, and nicotinamide cleared p-tau and promoted autophagy in AD model mice. These are animal and model findings, and the review itself cautions that results are limited and should be interpreted with caution when extrapolating to human clinical outcomes.

The human data are far thinner. A Phase II trial of valproic acid in mild-to-moderate Alzheimer’s disease over 24 months found no significant cognitive benefit versus placebo. In Huntington’s disease, the CREST-E trial of valproic acid was terminated early due to futility. Vorinostat demonstrated CNS penetrance in a pilot study, with measurable histone acetylation increases in cerebrospinal fluid, providing proof-of-concept for target engagement in the brain. Sodium butyrate and phenylbutyrate showed modest benefits in early HD and ALS trials, primarily as tolerability studies rather than efficacy evaluations. Nicotinamide riboside, targeting SIRT1 pathways, is currently in Phase II evaluation for Parkinson’s disease in the NOPARK trial, with results awaited.

Why has the translation failed? The review identifies four intertwined mechanistic issues. First, pan-HDAC inhibitors simultaneously suppress pathogenic and homeostatic deacetylase activities, producing at best unpredictable and at worst neurotoxic pharmacological effects. Second, the blood-brain barrier is a formidable obstacle: many HDAC inhibitors are substrates of efflux transporters such as P-glycoprotein, and the brain-to-plasma ratios of vorinostat and other hydroxamates are typically less than 0.1, meaning doses effective in animal models cannot be translated to safe, tolerable human doses without undue peripheral toxicity. Third, most preclinical models employ young animals with acutely induced disease, which rarely recapitulates the slow, multi-component and genetically heterogeneous progression of human neurodegeneration. Fourth, there are no validated pharmacodynamic biomarkers of target engagement in the CNS, so clinical dosing has been empirical rather than mechanistically driven, risking underdosing or variable CNS exposure between patients.

Biomarker development is itself at an exploratory stage. CSF global histone H3/H4 acetylation offers the most direct pharmacodynamic readout and was the primary endpoint in a recent Phase 2a nicotinamide trial in early Alzheimer’s disease, which showed drug exposure without a significant reduction in CSF phosphorylated tau, underscoring how far even a well-powered trial remains from establishing target engagement. CSF neurofilament light chain tracks axonal injury across diseases but is not HDAC-specific, and inflammatory cytokine panels provide only an indirect readout of glial normalization. Whether peripheral blood acetylation correlates meaningfully with CNS acetylation has not been systematically validated for any HDAC isoform, which is why CSF- and neuroimaging-based endpoints are increasingly favoured over peripheral surrogates alone. The review also raises a deeper conceptual question: whether HDAC dysregulation is a primary driver of neurodegeneration or a secondary consequence of the disease process, in which case inhibition may offer only limited symptomatic benefit.

Long-term safety adds a further layer of concern. HDACs modulate immune function, metabolism, cell-cycle control and stress responses throughout the body, so prolonged systemic inhibition risks toxicity, and epigenetic changes may persist long after treatment ends, a phenomenon the authors describe as epigenetic memory that could leave patients susceptible to maladaptive transcriptional programs, immune dysfunction or oncogenic transformation. Chronic neurodegenerative diseases may require treatment for years or decades, and brain-targeted agents, intermittent schedules and personalized epigenetic approaches remain largely experimental.

The review closes with three strategic priorities: advancing next-generation isoform-selective inhibitors, such as HDAC2-selective benzamides for Alzheimer’s and HDAC6-selective tubastatin A derivatives for Parkinson’s and ALS, into properly powered phase II/III trials with CNS pharmacodynamic endpoints; systematically comparing novel CNS delivery platforms, including brain-penetrant prodrugs, lipid nanoparticles functionalized with transferrin receptor ligands and intranasal systems, for penetrance and tolerability; and stratifying patients by multi-omics epigenetic profiling, including baseline HDAC isoform expression in accessible biofluids and genome-wide histone acetylation maps from patient-derived iPSC neurons. The authors conclude that the failure of pan-HDAC inhibitors reflects not a failure of the epigenetic hypothesis but an unoptimized targeting strategy, and that isoform-selective inhibitors with defined CNS target engagement biomarkers should be the focus of future research.

Subject of Research: Isoform-selective targeting of histone deacetylases as a precision therapeutic strategy for neurodegenerative diseases

Article Title: Isoform selective targeting of histone deacetylases as a precision therapeutic strategy for neurodegenerative diseases

Article References: Olorunsesan, M. D., Oyedokun, P. A., Gbadero, J. O., Chuka-Utazi, O. L., Adewale, O., Olaniyi, K. S., Olayemi, S. B., & Odeniyi, T. J. (2026). Isoform selective targeting of histone deacetylases as a precision therapeutic strategy for neurodegenerative diseases. Discover Neuroscience, 21(1), Article 106. https://doi.org/10.1186/s13064-026-00331-z

Image Credits: AI Generated

DOI: 10.1186/s13064-026-00331-z

Keywords: histone deacetylases, HDAC2, HDAC6, SIRT1, Alzheimer's disease, Parkinson's disease, Huntington's disease, amyotrophic lateral sclerosis, epigenetics, blood-brain barrier, biomarkers, drug development