Nature Neuroscience
- 9, 1351 - 1352 (2006)
doi:10.1038/nn1106-1351
Inhibiting glycolysis to reduce seizures: how it might workYang Zhong Huang & James O McNamaraThe authors are at the Department of Neurobiology, Department of Medicine (Neurology) and Center for Translational Neuroscience, Duke University Medical Center, Durham, North Carolina 27710, USA. jmc@neuro.duke.edu An inhibitor of glycolysis is shown to have antiepileptic effects in the rat kindling model, possibly through NADH-dependent regulation of gene expression. This may explain how the 'ketogenic diet' treatment works.The epilepsies, disorders of recurrent seizures, affect approximately 1% of the population worldwide. About one-third of patients with epilepsy have recurrent seizures despite optimal therapy with available antiepileptic drugs, making it essential to develop new therapies. Almost 100 years ago, researchers discovered that a diet low in carbohydrates and high in fat reduces seizure frequency in some patients with severe epilepsy1. This 'ketogenic diet' is still used today as an alternative therapy for drug-resistant epilepsy. How the diet inhibits seizures is uncertain and has been variously attributed to acidosis, dehydration, lipids and changes in energy metabolism within the brain. Because restriction of carbohydrates is a key facet of the diet, and glycolysis is the central pathway of glucose catabolism in mammalian cells, one expected consequence of carbohydrate restriction would be inhibition of glycolysis. Garriga-Canut and colleagues therefore hypothesized that inhibition of glycolysis might be the mechanism by which the ketogenic diet exerts its antiepileptic effects. In this issue, they report2 that an inhibitor of glycolysis, 2-deoxy-D-glucose (2DG), has powerful antiepileptic effects in rats. These results advance inhibition of glycolysis as an intriguing and plausible mechanism of the ketogenic diet, and suggest that inhibitors of glycolysis may provide new therapies for epilepsy.
To test the effects of 2DG, the authors used the kindling model of epilepsy, in which a brief, low-intensity electrical stimulation is administered through an intracerebral electrode to produce a brief focal seizure. Repeated application of these low-intensity electrical stimulations results in longer and more widely propagated seizures, evident behaviorally as intense contractions of limb and facial muscles similar to those observed in a grand mal seizure in humans. Treatment with 2DG greatly increased the threshold for stimulation-evoked focal seizures and also partially inhibited the progressive severity of epilepsy in this model.
How might 2DG exert these antiepileptic effects? The authors focused on the neurotrophin BDNF and its receptor TrkB, because a conditional deletion of TrkB similarly increases the focal seizure threshold and eliminates the progressive severity of epilepsy in the kindling model3. They found that, indeed, 2DG inhibited seizure-induced increases in BDNF and TrkB mRNA levels in vivo. The reduction of BDNF and TrkB gene expression by 2DG in turn raised the question as to how exactly 2DG affects transcriptional regulation. One possibility would be that a transcriptional repressor complex is recruited to the BDNF and TrkB promoters in a 2DG-dependent manner. There is a master negative regulator of neuronal genes called neuron-restrictive silencing factor (NRSF) or repressor of expression of sodium type II (REST)4,
5. NRSF functions in neural and non-neural tissue by binding to the neuron-restrictive silencing element (NRSE) and recruiting transcriptional repressor complexes onto chromatin to inhibit gene transcription. Because there is an NRSE sequence in the promoter region of both the BDNF and the TrkB gene, it seemed plausible that NRSF mediates the 2DG-induced inhibition of BDNF and TrkB gene expression. The authors demonstrated that NRSF does indeed bind to the BDNF NRSE, and that 2DG treatment reduces acetylation and increases methylation of histone H3K9, modifications predicted to reduce BDNF transcription. Moreover, distinct inhibitors of glycolysis increased NRSF-mediated gene repression of a Gal4-responsive reporter in JTC-19 cells and reduced BDNF mRNA levels in primary cultures of hippocampal neurons. The effects of 2DG on NRSF function may be mediated by the co-repressor C-terminal binding protein (CtBP), which the authors show to form a complex with NRSF in vitro in an NADH-regulated manner. Dual genetic deletion of CtBP1 and CtBP2 in mouse embryonic fibroblast cells, or knockdown of CtBP or NRSF in normal mouse mammary gland cells by small hairpin RNAs, relieved the repression of NRSF-target genes by 2DG, implicating the NRSF:CtBP complex in 2DG-induced changes in gene expression. Together, these findings support a model in which inhibition of glycolysis by 2DG results in an NADH-labile association of the NRSF:CtBP complex with NRSF target genes such as BDNF and TrkB, and their reduced expression results in the 2DG-mediated antiepileptic effects (Fig. 1).
 | | |
The new study by Garriga-Canut et al. is of interest for a number of reasons. First, it provides a plausible mechanism for how the ketogenic diet exerts its antiepileptic effects, namely by inhibition of glycolysis. It also provides one potential mechanism by which the inhibition of glycolysis exerts these beneficial effects, namely by reduced expression of BDNF and its receptor, TrkB. Finally, the work provides a plausible mechanism by which the inhibition of glycolysis reduces the expression of BDNF and TrkB, namely by recruitment of the co-repressor CtBP to the promoters. That said, some cautionary notes are in order. Was glycolysis actually inhibited in vivo by 2DG under the conditions of this experiment? If so, was inhibition of glycolysis the mechanism by which the 2DG exerted these antiepileptic effects? Given the evidence that glycolysis occurs predominantly in astrocytes, followed by transfer of lactate to neurons to fuel their energy demands6, might the antiepileptic effects of 2DG be mediated by inhibition of glycolysis in astrocytes? Furthermore, alteration of energy metabolism will almost certainly have pleiotropic effects. For example, modifying dendritic mitochondrial activity affects the number of spines and the plasticity of synapses7, suggesting that synaptic energy metabolism may locally influence synaptic plasticity. These mechanisms could influence antiepileptic effects detected in vivo. It is therefore crucial that future follow-up experiments determine whether the reduced expression of BDNF and TrkB is really sufficient to explain the antiepileptic effects of 2DG.
These caveats notwithstanding, the authors should be congratulated for providing a strong and original rationale for how inhibition of glycolysis could reduce neuronal excitability in vivo. Moreover, this work advances small-molecule inhibitors of glycolysis as a new and potentially powerful pharmacological approach for the treatment of drug-resistant epilepsy.
REFERENCES
- Engel, J. & Pedley, T. in Epilepsy: A Comprehensive Textbook 1339–1342 (Lippincott-Raven, Philadelphia, 1997).
- Garriga-Canut, M. et al. Nat. Neurosci. 9, 1382–1387 (2006). | Article |
- He, X. et al. Neuron 43, 31–42 (2004). | Article | PubMed | ISI | ChemPort |
- Chong, J.A. et al. Cell 80, 949–957 (1995). | Article | PubMed | ISI | ChemPort |
- Schoenherr, C.J. & Anderson, D.J. Science 267, 1360–1363 (1995). | PubMed | ISI | ChemPort |
- Magistretti, P.J. J. Exp. Biol. 209, 2304–2311 (2006). | Article | PubMed | ChemPort |
- Li, Z. et al. Cell 119, 873–887 (2004). | Article | PubMed | ISI | ChemPort |
|
