Defining the mechanistic role of high mobility group box-1 and its utility as a biomarker in the inflammatory pathogenesis of epilepsy

Epilepsy, which affects 50 million people worldwide, is a chronic neurological condition characterized by a predisposition to generate spontaneous seizures. Antiepileptic drug resistance is a significant problem, the causes of which are poorly understood. Inflammation is purported to play a pathological role in the development of epilepsy following brain insult. High mobility group box-1 (HMGB1) has been implicated in the development of seizures and epilepsy in preclinical models and human studies. HMGB1 undergoes post-translational modifications, including acetylation and redox changes, which dictate its inflammatory extracellular function. Novel inflammatory blood biomarkers such as HMGB1 that are intricately involved in the epilepsy disease process per se may act as stratification markers to identify patients who may benefit from immunomodulatory interventions. This thesis aimed to characterise the role of HMGB1 in seizures and epilepsy and its utility as a clinical biomarker. Analysis of 24 healthy volunteers undergoing a 24-hour blood-sampling study did not demonstrate any significant circadian fluctuations in serum HMGB1. No intra or inter-subject variability was also observed in the biomarker. A further study involving patients with idiopathic intracranial hypertension (IIH, n=18), neuroinfection (n=15) and Rasmussen’s encephalitis (n=10) showed that there was no correlation between serum and cerebrospinal (CSF) fluid levels of HMGB1, regardless of blood brain barrier integrity. Subgroup analysis of bacterial meningitis showed that both CSF and serum HMGB1 was significantly elevated (as compared to IIH). Furthermore, CSF HMGB1 was more than 10-fold higher in those with bacterial (n=6) rather than viral meningitis (n=8). The expression pattern of HMGB1 acetylation and redox isoforms in brain and blood was examined in three distinct preclinical models of seizures and epilepsy including recurrent seizures and status epilepticus in the kainate-model, single seizure in the maximal electroshock test (MES) and chronic spontaneous seizures in the pilocarpine epilepsy model. In response to kainate-induced seizures, in both brain and blood, an early rise in non-acetylated and reduced HMGB1 isoforms was demonstrated consistent with functional chemotaxis. This was followed by a delayed 6-fold rise at 24 hours in brain of the acetylated, disulphide inflammatory form of HMGB1. In serum, significant expression of the inflammatory isoforms was seen after 14 days, possibly coinciding with the onset of spontaneous seizures. Inflammatory isoforms of HMGB1 were not identified within the first 24 hours following isolated MES-seizure in mice. Serum, but not brain, total HMGB1 was significantly elevated (by 311%) in chronic epileptic mice experiencing regular spontaneous seizures; however the contribution of the different isoforms remains to be elucidated. In humans with epilepsy, compared to both healthy controls (1.11±0.07ng/ml,