Abstract
Reddy SD, Clossen BL, Reddy DS. J Pharmacol Exp Ther 2018;364:97–109.
Epilepsy is a chronic brain disease characterized by repeated unprovoked seizures. Currently, no drug therapy exists for curing epilepsy or disease modification in people at risk. Despite several emerging mechanisms, there have been few studies of epigenetic signalling in epileptogenesis, the process whereby a normal brain becomes progressively epileptic because of precipitating factors. Here, we report a novel role of histone deacetylation as a critical epigenetic mechanism in epileptogenesis. Experiments were conducted using the histone deacetylase (HDAC) inhibitor sodium butyrate in the hippocampus kindling model of temporal lobe epilepsy (TLE), a classic model heavily used to approve drugs for treatment of epilepsy. Daily treatment with butyrate significantly inhibited HDAC activity and retarded the development of limbic epileptogenesis without affecting after-discharge signal. HDAC inhibition markedly impaired the persistence of seizure expression many weeks after epilepsy development. Moreover, subchronic HDAC inhibition for 2 weeks resulted in a striking retardation of epileptogenesis. HDAC inhibition, unexpectedly, also showed erasure of the epileptogenic state in epileptic animals. Finally, butyrate-treated animals exhibited a powerful reduction in mossy fiber sprouting, a morphologic index of epileptogenesis. Together these results underscore that HDAC inhibition prevents the development of TLE, indicating HDAC's critical signalling role in epileptogenesis. These findings, therefore, envisage a unique novel therapy for preventing or curing epilepsy by targeting the epigenetic HDAC pathway.
Commentary
Epigenetic modifications broadly encompass a range of diverse chemical alterations that influence the accessibility of DNA for binding partners. These ‘marks’, which include DNA methylation, histone modifications, and noncoding RNA along with many more, thereby influence expression of many gene products in the genome. Histone proteins are perhaps the most versatile of all epigenetic substrates. Histones tightly associate with DNA, and can undergo a huge variety of posttranslational modifications to influence the potential for transcription factor binding and subsequent gene expression or repression. Histone acetylation is one of the most widely studied histone modifications, and the consequence of this is generally to open chromatin structure to facilitate gene transcription (1). This process is catalyzed by histone acetyltransferases and reversed by histone deacetylases (HDACs). Histone acetylation is an attractive therapeutic target for disease since several pharmacological tools have been developed to influence the activity of these enzymes.
The search is on in earnest to identify disease-modifying therapies for epilepsy. Epigenetic modifications present as excellent target candidates: genome-wide changes in gene expression are often observed in epileptic brain (2), and these broad gene expression changes may participate within a molecular network to result in epilepsy, with no single gene change by itself being critical. Also, acquired epilepsy develops following a latent period after a brain insult, potentially underpinned by the emergence of stable alterations to chromatin structure leading to enduring patterns of gene expression. Further, several recent studies detail the presence of epigenetic alterations in models of epilepsy (3) and in humans (4), and there is evidence that manipulation of these can impact the epileptogenic process (5, 6). Finally, one of the most commonly used antiepileptic drugs with a mysterious mechanism of action is sodium valproate—an HDAC inhibitor. The paper by Reddy et al. sets out to explore epigenetic modifications in epilepsy, focusing on histone acetylation as a potential therapeutic target.
The hippocampal kindling model of epileptogenesis was the model of choice, as opposed to other rodent models with enhanced face validity, such as the post-SE models. Kindling has largely fallen out of favor, perhaps due to inherent challenges separating antiseizure effects from antiepileptogenic potency. However, with careful study designs, as adopted in the current paper, these challenges can be overcome. First, the authors tested the efficacy of their drug treatment on the biological target—enzymatic HDAC activity—and this was sufficiently impaired. It would have been valuable to validate the target by assessing HDAC activity directly after kindled seizures. Next, they conducted a series of kindling studies with complex designs. An important issue for kindling studies is separating antiseizure from antiepileptogenic effects. This potential confound was not a concern for the current study. Using fully kindled mice, they demonstrated a complete lack of antiseizure efficacy with the chosen dose of sodium butyrate. Study 1 was a standard kindling study—twice daily injections of sodium butyrate markedly delayed the progression of hippocampal kindling, suggesting that HDAC inhibition influences epileptogenesis. Kindling is thought to represent a stable and persistent change in brain architecture and seizure susceptibility, akin to the chronic epileptic condition. Intriguingly, 8 weeks after the conclusion of kindling, mice were rekindled. Whereas most control mice exhibited convulsive seizures as expected, less than 20% of butyrate-treated mice experienced a Stage 4/5 seizure. This suggests that the biological mechanisms underlying the kindling process were less stable when they occurred during HDAC inhibition (although this finding may also be related to the less robust generation of the kindled state).
A second kindling study partially replicated the first, but this time drug treatment ceased half way through the process. Again, butyrate treatment significantly delayed epileptogenesis, but surprisingly, the drug effects persisted for at least 3 weeks longer, when no noticeable progression of kindling occurred in mice previously exposed to the HDAC inhibitor. One would anticipate that, after the drug had been removed from the system, kindling would progress much more rapidly (as shown with a positive control treatment group). Instead, prolonged retardation of kindling may be caused by sustained inhibition of HDAC activity, despite the drug no longer being present, or by the stable development of gene expression changes, which contribute first to the antiepileptogenic activity, but are also sustained long after the direct drug action has eroded.
Studies of epileptogenesis are valuable for understanding the biological mechanisms that lead to the chronic epileptic state. However, unless biomarkers of epileptogenesis are identified, these studies are not particularly valuable from a translational perspective, because patients are usually referred to clinicians after experiencing their first seizure—when the epileptic state is already established. To address this, the final study took fully kindled mice—akin to a chronically established disease state—and treated them with sodium butyrate for 2 weeks. After treatment had completed, subsequent kindling of these mice showed that the drug exposure impaired the previously established ‘epileptic’ state, such that the mice experienced less severe and shorter seizures than control-treated mice, who all experienced Stage 5 seizures. This remarkable result suggests that HDAC inhibition was able to reverse the biological underpinnings of the kindling process—a true example of disease modification.
Although not necessary for this study, it would be valuable to link improvements in epileptogenesis with the associated molecular modifications. If sodium butyrate is truly acting via HDAC inhibition, then the gene programs that underpin the disease-modifying effects should be readily identifiable. The authors chose to use sodium butyrate as their HDAC inhibitor, as opposed to others, such as sodium valproate, due to its broad spectrum of inhibition (targeting all HDACs). This is a sound strategy, but may have the consequence of increasing the likelihood of off-target effects. It stands to reason that treatment with such broad-spectrum compounds like epigenetic modifiers will influence not only pathological processes, but may also interfere with normal physiologic functions (7, 8). These have not been characterized from a molecular level, but should be considered in future studies. Still, the work from Reddy et al. is a powerful example of the ability of epigenetic modifiers—specifically an HDAC inhibitor—to interfere with the epileptogenic process, and even reverse the condition once established. Future studies, perhaps employing a model exhibiting spontaneous seizures, demonstrating the same effects with other drugs of the same class, and identifying downstream mediators will enhance the translational capacity of this exciting therapeutic angle.