- Top of page
Purpose: The ketogenic diet (KD) has been used as an effective antiepileptic treatment for nearly a century. Inhibition of glycolysis and increased levels of ketone bodies are both known to contribute to the antiepileptic effects of the KD. Neuron-restrictive silencer factor (NRSF), also known as RE-1 silencing transcription factor (REST), is implicated in the antiepileptic effects of the glycolytic inhibitor 2-deoxy-d-glucose (2DG). Glycolytic inhibition is a common feature of the KD and 2DG treatment, leading to the hypothesis that NRSF might also be involved in the antiepileptic effect of the KD. To test this hypothesis, the present study was designed to investigate the role of NRSF in the antiepileptic effect of 2DG, the KD, and acetone in vivo.
Methods: Kindling was used as a model to test the antiepileptic effects of 2DG, the KD, and acetone on control and NRSF conditional knockout mice (NRSF-cKO; from the intercross of CamKIIα-iCre and NRSF exon 2 floxed mice). After recovery from electrode implantation, adult mice were stimulated twice a day at afterdischarge threshold (ADT) current intensity. In the 2DG- (500 mg/kg) and acetone- (10 mmol/kg) treated groups, drugs were injected intraperitoneally 20 min before each stimulus. In the 2DG group, mice were pretreated with intraperitoneal injections for 3 days in addition to the injections administered before the regular kindling stimulation. In the KD group, mice were fed the KD instead of a control diet until the end of stimulations.
Key Findings: Compared with control mice, the antiepileptic effect of 2DG was abolished in NRSF-cKO mice, indicating that NRSF is required for the antiepileptic effect of 2DG. In the KD-fed group, kindling development was retarded in both control and NRSF-cKO mice. In the acetone-treated group, inhibition of kindling-induced epileptogenesis was observed in both control and NRSF-cKO mice, similar to the action of the KD.
Significance: These findings imply that NRSF repression complex is not essential for the antiepileptic effect of the ketogenic diet.
Despite the development of various antiepileptic drugs, a large number of patients (about 30%) are resistant to current pharmacotherapies (Loscher, 2002). For many drug-resistant patients with severe epilepsy, dietary interventions such as the ketogenic diet (KD) can provide an effective alternative (Vicente-Hernandez et al., 2007). In the KD, carbohydrate intake is replaced by high levels of long-chain fatty acids, which leads to inhibition of glucose metabolism and overproduction of ketone bodies in the liver, making ketones the main energy source for the glucose-deprived brain (Sankar & Sotero de Menezes, 1999). In this way, the diet mimics biochemical responses associated with prolonged starvation, or fasting, which has long been known to have an antiepileptic value (Veech, 2004). Although the KD has been used clinically for nearly a century, the molecular mechanisms underlying its antiepileptic effect remain unclear.
As mentioned above, the KD leads to exuberant production of ketone bodies and to reduced glycolysis (Hartman et al., 2007). The respective up- and down-regulation of these two metabolic processes results in changes in the profiles of small metabolic derivatives, some of which serve as ligands or coenzymes in the regulation of gene transcription and the modulation of neuronal excitability (Bough et al., 2006). Previous works suggest that both decreased glycolytic activity and increased ketone body levels contribute to the antiepileptic effects of the KD (Likhodii et al., 2003; Garriga-Canut et al., 2006; Gasior et al., 2007; Ma et al., 2007).
Neuron-restrictive silencing factor (NRSF) (Schoenherr & Anderson, 1995), also known as REST (Chong et al., 1995), regulates transcription by recruiting the corepressors CoREST and mSin3A and forming a repression complex with HDAC1/2 that condenses chromatin structure (Chen et al., 1998; Andres et al., 1999; Huang et al.,1999; Grimes et al., 2000; Roopra et al., 2000; Ballas et al., 2005; Lunyak & Rosenfeld, 2005). NRSF mRNA has been shown to be upregulated after kainic acid stimulation (Palm et al.,1998). Moreover, NRSF target genes, such as brain-derived neurotrophic factor (BDNF) and tropomyosin-related kinase B (TrKB), play critical roles in the progression of kindling-induced epileptic seizure activity (Kokaia et al.,1995; He et al., 2004; Garriga-Canut et al., 2006; Chen et al., 2007). In addition, a recent study by our group demonstrated that conditional knockout of NRSF in forebrain excitatory neurons accelerates the development of epilepsy in the kindling model (Hu et al., 2011). These findings suggest that NRSF might serve as an endogenous moderator that inhibits epileptogenesis. Whether the KD works by recruiting the NRSF repression complex to mediate the inhibition of epileptogenesis remains an unanswered question.
Recent work demonstrated that the glycolytic inhibitor 2-deoxy-d-glucose (2DG) inhibits the progression of epilepsy by enhancing afterdischarge threshold (ADT) values in the kindling model (Garriga-Canut et al., 2006). Glycolytic inhibition leads to reduced levels of NADH, thereby strengthening the interaction of NRSF with C-terminal binding protein (CtBP) to suppress the expression of NRSF target genes. These experiments indicate a correlation between NRSF-mediated epigenetic regulation and the inhibition of the glycolytic metabolism by 2DG, and led to the hypothesis that the antiepileptic effect of 2DG, and possibly the KD, relies on increased repressive activity of the NRSF complex.
Here, we conditionally deleted NRSF in forebrain excitatory neurons using a previously described Cre-loxP strategy (Hu et al., 2011). We employed the kindling model to test the relationship between NRSF and glycolytic inhibition, the KD, or increased acetone levels. We found that 2DG, the KD, and acetone treatment all had an antiepileptic effect on control mice. The antiepileptic effect of 2DG was abolished in NRSF-cKO mice. However, NRSF-cKO mice injected with acetone or fed the KD showed an attenuation in epilepsy progression similar to that in control mice. These results demonstrate that NRSF is required for the antiepileptic effect of 2DG, but not for the antiepileptic effect of the KD or acetone treatment.
- Top of page
In this study, we investigated the role of NRSF in impaired epileptogenesis mediated by treatment with 2DG, acetone, and the KD. We found that NRSF was required for 2DG-mediated antiepileptic effects, but not those of acetone and the KD. A previous report revealed that the glycolytic inhibitor 2DG impairs kindling development by recruiting the transcriptional repression complex NRSF-CtBP, thereby repressing BDNF and TrKB expression (Garriga-Canut et al., 2006). Because the KD mediates antiepileptic effects as well as glycolytic inhibition, similar to 2DG (Szot et al., 2001; Kennedy et al., 2007), we wondered whether the NRSF complex is required for the action of 2DG and KD in the development of epilepsy. The current study offers several conclusions: first, NRSF is required for the antiepileptic effect of 2DG, because 2DG treatment impaired kindling in control mice but not in NRSF-cKO mice. Second, NRSF is not required for the antiepileptic effect of the KD, because the KD impaired kindling in both control and NRSF-cKO mice. Third, the antiepileptic effect of acetone is similar to that of the KD, because acetone treatment inhibited kindling in both control and NRSF-cKO mice.
2DG is a glucose molecule in which the 2-hydroxyl group is replaced by hydrogen, so that it can be taken up by cells, but cannot undergo further glycolysis (Lazzarini et al., 1988). Radioactively labeled 2DG is a popular marker for detecting the rate of glucose uptake. 2DG can also be used as an antitumor agent, because uptake of 2DG hampers cell growth (Pelicano et al., 2006; Ben Sahra et al., 2010). Recently, 2DG has been proposed to mimic the effects of the KD, functioning as an antiepileptic agent by recruiting the transcriptional repressor NRSF-CtBP complex (Garriga-Canut et al., 2006; Stafstrom et al., 2009). Our current study revealed that 2DG treatment impairs kindling development in control mice but not in NRSF-cKO mice (Fig. 1), indicating that NRSF is required for 2DG-mediated antiepileptic effects. This finding confirms the previously reported effect of 2DG in recruiting the NRSF-CtBP complex to suppress kindling-induced epileptogenesis. Previous data established an association between the recruitment of the NRSF-CtBP complex by 2DG and kindling development. Our current data in transgenic mice provide direct in vivo evidence to confirm the causality of the NRSF suppression complex in the antiepileptic effect of 2DG or glycolytic inhibition.
Our finding that the antiepileptic effect of 2DG is lost in NRSF-cKO mice indicates that NRSF is the main molecular player in this function of 2DG. Although 2DG delayed epileptogenesis in an NRSF-dependent manner, higher dosages of 2DG (750 mg/kg or above) have been shown to cause significant limpness in skeletal muscles of both control and NRSF-cKO mice, with the mice exhibiting a lack of motion and cold extremities. As a glycolytic inhibitor, 2DG reduce glycolysis both in the brain and in the skeletal muscles. The above effects might result from its glycolytic inhibition in skeletal muscles. Acute 2DG treatment (10 mmol) has been reported to reduce interictal epileptiform bursts induced in several in vitro models of epilepsy, indicating an acute anticonvulsant effect of 2DG (Stafstrom et al., 2009). Chronic in vivo effects of 2DG against progression of kindled seizures have been tested in perforant path and in olfactory bulb of rats (Garriga-Canut et al., 2006; Stafstrom et al., 2009). Similar results in the current study with applied kindling test in amygdala of mice indicated that 2DG is a broadly suppressive reagent against epileptogenesis.
The KD has been used as an effective therapeutic approach to treat refractory epilepsy for nearly a century (Lefevre & Aronson, 2000; Freeman et al., 2006). The metabolic consequences of the KD in animal have two major aspects: glycolytic inhibition and ketosis (Veech, 2004; Bough & Rho, 2007; Kennedy et al., 2007; Ma et al., 2007). If the NRSF complex is known to be required for the antiepileptic effect of the glycolytic inhibitor 2DG (Garriga-Canut et al., 2006), how NRSF is also required for KD-mediated antiepileptic effects is still up for debate. Because both glycolytic inhibition and ketone body upregulation have been shown to suppress epileptogenesis (Likhodii et al., 2003; Garriga-Canut et al., 2006), we set out to determine which of these two mechanisms mediates the therapeutic action of the KD on epilepsy. Our results revealed that the NRSF complex is not required for the antiepileptic effect of the KD, because the KD inhibited kindling development in both control and NRSF-cKO mice. Furthermore, the antiepileptic effect of the KD on NRSF-cKO mice (accelerated epileptogenesis) was even more pronounced than on control mice (Fig. 2A,B,D,E). These results indicate that the KD leads to powerful suppression of epileptogenesis, especially in a brain susceptible to epilepsy. We also observed that long-term KD feeding increased the stimulus intensity required for induced afterdischarges (ADT).
Because ketosis has been reported to play important role in antiepileptic processes (Rho et al., 2002; Likhodii et al., 2003; Gasior et al., 2007; Ma et al., 2007; Likhodii et al., 2008), we employed acetone (direct systemic injection) to test its antiepileptic effect on control and NRSF-cKO mice. Of interest to note is that we observed that acetone mimics the effect of the KD by inhibiting kindling, both in control mice and in NRSF-cKO mice. Similar to the action of the KD, acetone exerted a more powerful antiepileptic effect in cKO mice than in control mice, indicating that acetone is more effective in susceptible or epileptic brains than in a healthy control brain. These results suggest that the KD and acetone are similar in their antiepileptic effects, and that the two potentially share the same underlying molecular mechanisms. CD-fed mice treated with 2DG for 2 weeks did not exhibit ketosis, whereas KD-fed mice had elevated ketone bodies in serum and urine (Fig. 4). This result indicates that inhibition of glycolysis impairs kindling development without leading to ketosis. Together with the results of the powerful antiepileptic effects of the KD and acetone, we demonstrate that the antiepileptic effects of the KD and acetone are not dependent on NRSF-mediated transcriptional repression. In summary, these findings reveal that the NRSF transcriptional repression complex is not required for the antiepileptic effects of the KD and acetone, but is required for the antiepileptic effects of 2DG and glycolytic inhibition.
Figure 4. Ketone body production was increased in mice fed the KD for 2 weeks, but not in 2DG-treated mice. β-hydroxybutyrate (bHB), one of the major ketone bodies, offers high stability for detection. bHB concentrations in the serum and urine of CD, KD, or 2DG-treated mice were analyzed by bHB assay kit (BioVision #K632-100, Mountain View, CA, U.S.A). All data are presented as mean ± standard error of the mean (SEM). ***p < 0.001; *p < 0.05; n = 6 mice.
Download figure to PowerPoint