Epigenetic signatures – profiling the epilepsies beyond genetics

What is epigenetics? In a single idea: the molecular memory of a cell. The system stores information of previously external (e.g. environmental) or internal (e.g. developmental) stimuli, learns from this experience and responds. A collection of specific tags tells genes whether to be ON or OFF. Hardcore epigeneticists claim that an epigenetic tag should be meiotically and/or mitotically heritable, self-perpetuating, and reversible. DNA methylation is the mechanism coming closest to this ideal. A more liberal definition not focusing on heritability refers to any structural adaptation of the chromatin template that regulates gene expression. This would also include posttranslational histone tail modifications, incorporation of histone variants, chromatin remodeling processes, and action of non-coding RNAs. The large variety, flexibility, interdependence and potential synergistic effects of epigenetic mechanisms could provide the molecular basis for any phenotypic variation in physiological and pathological conditions. In epilepsy research this is especially interesting with regard to the stimulus-driven activity and connectivity of post-mitotic neurons in the adult brain. We set out to study methylation for the most common form of epilepsy in adults.

Temporal lobe epilepsy and animal models. Seizures originate primarily from the hippocampus, which frequently shows distinct patterns of neuronal cell loss and scarring (hippocampal sclerosis, HS). Clinical history indicates an early onset of the disease process in some patients, as a fraction of TLE patients have suffered from severe febrile seizures, status epilepticus or brain inflammation during the first years of childhood. This initial precipitating injury is usually followed by a clinically silent latent period before the onset of epilepsy. Many patients become drug-resistant during the course of the disease and possibly need surgical treatment to achieve seizure control. The underlying pathomechanisms have not yet been identified. Animal models mimicking the pathogenic time course, typical histopathological changes and some clinical features of TLE are indispensable in our aim to understand the disease and test treatment strategies.

Deep sequencing (Methyl-Seq) revealed increased genomic DNA methylation in chronic rat epilepsy. Heat map displaying hierarchical clustering of samples and genomic regions according to differential methylation profiles (yellow - methylation up, red - methylation down). A specific DNA methylation signature characterized chronic rat epilepsy. CTRL - healthy controls; EPI - chronic epileptic animals. Bisulfite sequencing validates deep sequencing results for two candidate genes. White dots represent unmethylated and black dots methylated CpGs. Middle panel summarizing schematic gene structure with TSS (green arrow), chromosomal region and region covered in Bis-Seq. Right panel presenting gene expression data from RT-PCR. Hypermethylation corresponded with decreased and hypomethylation with increased gene expression. (adapted from Kobow et al., 2013 under a creative commons licence)

Deep sequencing (Methyl-Seq) revealed increased genomic DNA methylation in chronic rat epilepsy. Heat map displaying hierarchical clustering of samples and genomic regions according to differential methylation profiles (yellow – methylation up, red – methylation down). A specific DNA methylation signature characterized chronic rat epilepsy. CTRL – healthy controls; EPI – chronic epileptic animals. Bisulfite sequencing validates deep sequencing results for two candidate genes. White dots represent unmethylated and black dots methylated CpGs. Middle panel summarizing schematic gene structure with TSS (green arrow), chromosomal region and region covered in Bis-Seq. Right panel presenting gene expression data from RT-PCR. Hypermethylation corresponded with decreased and hypomethylation with increased gene expression. (adapted from Kobow et al., 2013 under a creative commons licence)

DNA methylation in chronic rat epilepsy. TLE is not a genetic disorder in the strict sense, but familial cases of TLE and genetic predisposition to febrile seizures or other risk factors have been described. The majority of cases appear to be associated with aquired focal lesions like HS, tumors or certain malformations. To gain some insight into the mechanisms underlying TLE, gene expression profiling studies have been performed for over a decade and hundreds of genes without any conceivable link were identified to be misregulated in human and experimental TLE. It was hypothesized that epigenetics could work as a master switch here. Following this idea, we performed genome-wide analysis of DNA methylation in chronic epileptic rats compared to controls. The question was: Can we distinguish rats with epilepsy from control animals based on their methylation pattern? Indeed, unsupervised clustering of methylated loci throughout the rat genome revealed that we can. We further identified a strong inverse correlation between DNA methylation and gene expression, meaning that genes that were highly methylated showed a reduction in expression and vice versa. The question remains how to explain methylated loci that showed no corresponding gene expression change or, the other way around, gene expression changes that were not associated with a respective methylation signal. One possibility is that competing epigenetic mechanisms beside DNA methylation may be active. Also, the presence of common transcription factor binding sites in genes that were co-regulated may offer some explanation. Further DNA methylation changes outside genic regions may have affected expression of long or short non-coding RNAs, which are master regulators of gene expression themselves. In the end the sum of all activating and repressive signals defines a gene’s activity state.

Relevance for EuroEPINOMICS. Our study establishes genomic DNA methylation as distinguishing feature in chronic rat TLE and is the first report of unsupervised clustering of an epigenetic mark to separate epileptic from non-epileptic animals. Future studies will have to answer whether similar DNA methylation signatures can be found in other epilepsy models as well as human TLE and whether it is the presence of spontaneous seizures that leads to altered methylation, or the other way around, altered methylation influences seizures? It also remains unsolved if and how genetic variation affects  epigenetics and downstream processes in this and other forms of epilepsy. I am confident that we will discover many contributions of epigenetics to our field in the near future.

Epigenetics on the blog. The Channelopathist was a little heavy on genetics. Including some posts on epigenetics seemed reasonable as we study it as part of EuroEPINOMICS. It is also odd that the interplay of genetics, epigenetics and environment is recognized in many other diseases, but does not seem to be discussed much in the context of epilepsy. Once I made the suggestion to write about epigenetics, things had their own dynamics and I was moved from a silent reader’s position towards being a first-time contributor on this blog. Although I feel a little hopeless about competing with Ingo’s very entertaining posts, this one is hopefully going to be the start of an irregular epigenetics series.