Abstract
Stewart LT, Khan AU, Wang K, Pizarro D, Pati S, Buckingham SC, Olsen ML, Chatham JC, McMahon LL. J Neurosci 2017;37:8207–8215.
O-GlcNAcylation is a ubiquitous and dynamic post-translational modification involving the O-linkage of β-N-acetylglucosamine to serine/threonine residues of membrane, cytosolic, and nuclear proteins. This modification is similar to phosphorylation and regarded as a key regulator of cell survival and homeostasis. Previous studies have shown that phosphorylation of serine residues on synaptic proteins is a major regulator of synaptic strength and long-term plasticity, suggesting that O-GlcNAcylation of synaptic proteins is likely as important as phosphorylation; however, few studies have investigated its role in synaptic efficacy. We recently demonstrated that acutely increasing O-GlcNAcylation induces a novel form of LTD at CA3-CA1 synapses, O-GlcNAc LTD. Here, using hippocampal slices from young adult male rats and mice, we report that epileptiform activity at CA3-CA1 synapses, generated by GABAAR inhibition, is significantly attenuated when protein O-GlcNAcylation is pharmacologically increased. This dampening effect is lost in slices from GluA2 KO mice, indicating a requirement of GluA2-containing AMPARs, similar to expression of O-GlcNAc LTD. Furthermore, we find that increasing O-GlcNAcylation decreases spontaneous CA3 pyramidal cell activity under basal and hyperexcitable conditions. This dampening effect was also observed on cortical hyperexcitability during in vivo EEG recordings in awake mice where the effects of the proconvulsant pentylenetetrazole are attenuated by acutely increasing O-GlcNAcylation. Collectively, these data demonstrate that the post-translational modification, O-GlcNAcylation, is a novel mechanism by which neuronal and synaptic excitability can be regulated, and suggest the possibility that increasing O-GlcNAcylation could be a novel therapeutic target to treat seizure disorders and epilepsy. SIGNIFICANCE STATEMENT: We recently reported that an acute pharmacological increase in protein O-GlcNAcylation induces a novel form of long-term synaptic depression at hippocampal CA3-CA1 synapses (O-GlcNAc LTD). This synaptic dampening effect on glutamatergic networks suggests that increasing O-GlcNAcylation will depress pathological hyperexcitability. Using in vitro and in vivo models of epileptiform activity, we show that acutely increasing O-GlcNAc levels can significantly attenuate ongoing epileptiform activity and prophylactically dampen subsequent seizure activity. Together, our findings support the conclusion that protein O-GlcNAcylation is a regulator of neuronal excitability, and it represents a promising target for further research on seizure disorder therapeutics.
Commentary
Glycosylation of cell surface proteins is well known as a regulator of cell–cell and cell–extracellular matrix interactions. Much less studied than these complex glycan structures labeling cell surface proteins is β-N-acetyl glucosamine (O-GlcNAc), a hexosamine that is enzymatically added to select serine and threonine residues on cytosolic, nuclear, mitochondrial, or membrane proteins. O-GlcNAcylation has taken center stage as a posttranslational modification altering protein function in recent years (1) and is believed to be as important as the well-known protein phosphorylation. Indeed, O-GlcNAcylation often competes with phosphorylation on the same amino acid residues (2), adding a powerful level of control of protein function and localization. Several recent studies have begun to shed light on the role of O-GlcNAcylation in synaptic plasticity (3, 4), but so far, the effect of this form of protein glycosylation on epilepsy, neuronal hyperexcitability, and seizure susceptibility is poorly understood.
Stewart and colleagues now provide strong evidence that protein O-GlcNAcylation alters the susceptibility of neurons to epileptiform activity and might be an underrecognized posttranslational modification in epilepsy. Using different approaches to increase O-GlcNAcylation in the hippocampus of mice in vitro and in vivo, the authors show that even a transient surge in O-GlcNAcylated proteins has neuronal activity-reducing effects under basal conditions and after epileptiform activity induced by inhibition of GABAergic signaling.
Two aspects of this study are particularly intriguing and may motivate follow-up studies to understand the underlying molecular mechanisms. First, increasing O-GlcNAcylation dampened epileptiform activity in hippocampal slices, regardless of whether the increase in O-GlcNAcylated proteins occurred before or after inducing hyperexcitability. Second, the approach to add glucosamine or a selective inhibitor of the O-GlcNAc–removing enzyme O-GlcNAcase—leading to an almost 1.5-fold increase of overall O-GlcNAcylation—was successful in modulating neuronal excitability despite a lack of target-specificity. This is surprising because induction of epileptiform activity in mouse and rat hippocampal slices in vitro or in living mice in vivo did not alter overall protein O-GlcNAcylation, suggesting that only a minority of proteins change O-GlcNAcylation levels upon neuronal hyperactivation.
The authors provided interesting first clues about a protein target that might be crucial for the activity-reducing effects of increased protein O-GlcNAcylation, namely, the AMPA receptor subunit GluA2, an important regulator of neuronal plasticity. They showed that administering glucosamine and an O-GlcNAcase inhibitor after induction of epileptiform activity in hippocampal slices from mice that lack GluA2 was not as effective in reducing epileptiform activity as in wild-type slices, suggesting that GluA2 is needed to elicit the full protective effect. An interesting next step would be to assess the consequences of preemptive O-GlcNAc increase on epileptiform activity in the absence of GluA2, which could provide first answers to the question whether the same proteins mediate O-GlcNAc's effect on neuronal excitability, regardless of whether O-GlcNAcylation is increased before or after induction of epileptiform activity. Deletion of GluA2 did not fully abrogate the impact of increased O-GlucNAcylation; therefore, it is highly likely that other proteins are important as well. Proteomic studies to identify candidate proteins that display altered O-GlcNAcylation during epileptiform activity and following glucosamine treatment will be critical to fully understand the mechanistic underpinnings of the activity-dampening effect of O-GlcNAcylation. Other proteins shown to be modulated by O-GlcNAcylation during neuronal activity are involved in regulating synaptic excitability, and may be important players in glucosamine-induced reduction of epileptiform activity are, for example, CaMK2α and synapsin (5).
Notably, glucosamine is currently used as an over-the-counter dietary supplement in the United States to enhance joint health due to its assumed anti-inflammatory actions (6). An appealing conclusion of this study is that glucosamine as a dietary supplement may reduce neuronal hyperexcitability in epilepsy. As speculated by the authors, the same treatment may also alter EEG patterns associated with epileptogenesis that could change disease progression. The latter hypothesis is based on the observation that glucosamine paired with inhibition of O-GlcNAcase changes the dynamic power frequency distribution measured by cortical EEG recordings after pentylenetetrazole-induced seizures. To be confirmed, this theory needs further testing in animal models of epileptogenesis and epilepsy. Certainly, the results reported by Stewart and colleagues justify additional work to evaluate the potential benefits of dietary glucosamine as a disease-modifying treatment in epilepsy.
A rate-limiting step in O-GlcNAcylation is the generation of the metabolite UDP-GlcNAc through the hexosamine biosynthetic pathway (1). This pathway is tightly regulated by nutrient status, suggesting that diet can influence O-GlcNAcylation. Indeed, calorie restriction was shown to increase hippocampal O-GlcNAcylation and improve cognition in a mouse model of obesity-induced diabetes (7). Notably, reduction in carbohydrate intake through the ketogenic diet is an effective but poorly understood strategy to reduce seizure frequency in refractory epilepsy (8). A low-carbohydrate ketogenic diet in obese rats drastically reduced O-GlcNacylation of Akt, and, at the same time, increased Serine 473 phosphorylation. At first sight, this seems to be counterintuitive given the epilepsy-mitigating effects attributed to the ketogenic diet, the hyperexcitability-reducing effect of increasing O-GlcNAcylation, and the well-known connection between overactive signaling through the PI3K/Akt/mTOR pathway and epilepsy (9). However, these apparently conflicting results may be explained by tissue-specific effects: A very recent study analyzed O-GlcNAcylation as a consequence of a ketogenic diet in a mouse model of autism, revealing differential O-GlcNAcylation in the liver with no detectable effect in the brain (10). Nonetheless, it is tempting to speculate that ketose mediates its beneficial effects in epilepsy at least partially through altering O-GlcNAcylation of specific proteins involved in regulation of neuronal activity.
Manipulating O-GlcNAcylation levels through either glucosamine supplementation or the ketogenic diet lacks protein target specificity and might have unwanted side effects. When evaluating the findings of Stewart et al. for their potential as a basis for future development of novel treatment strategies in epilepsy, it is important to acknowledge that, in contrast to protein phosphorylation (which is mediated by many different substrate-selective kinases and phosphatases), current knowledge suggests that there are only two O-GlcNAcylation-regulating enzymes: the O-GlcNAc transferase and the O-GlcNacase (1). Previous studies have shown that target-specificity may be accomplished through enzyme localization, the presence of different splice isoforms, or by other posttranslational modifications, such as phosphorylation on the respective proteins. Yet, as of now, pharmacological interventions to alter O-GlcNAcylation are mainly limited to the manipulation of these two enzymes, a strategy that lacks protein specificity similar to the dietary interventions discussed above. Therefore, future studies are needed not only to identify the specific proteins affected by differential O-GlcNAcylation that mediate the observed reduction in epileptiform activity but also to better understand the mechanisms of how neurons regulate protein-specific O-GlcNAcylation.
The study by Stewart and colleagues is a first step towards unraveling the function of O-GlcNAcylation in taming seizure activity. Although there are still many open questions about the underlying mechanisms and target specificity, their work raises hope that this posttranslational modification could be a sweet deal for epilepsy treatment.