Fermenting Seizures With Lactate Dehydrogenase

Commentary

Epilepsy Treatment. Targeting LDH Enzymes With a Stiripentol Analog to Treat Epilepsy.

Sada N, Lee S, Katsu T, Otsuki T, Inoue T. Science 2015;47, 1362–1367.

Neuronal excitation is regulated by energy metabolism, and drug-resistant epilepsy can be suppressed by special diets. Here, we report that seizures and epileptiform activity are reduced by inhibition of the metabolic pathway via lactate dehydrogenase (LDH), a component of the astrocyte-neuron lactate shuttle. Inhibition of the enzyme LDH hyperpolarized neurons, which was reversed by the downstream metabolite pyruvate. LDH inhibition also suppressed seizures in vivo in a mouse model of epilepsy. We further found that stiripentol, a clinically used antiepileptic drug, is an LDH inhibitor. By modifying its chemical structure, we identified a previously unknown LDH inhibitor, which potently suppressed seizures in vivo. We conclude that LDH inhibitors are a promising new group of antiepileptic drugs.

Metabolic control of seizures has been recognized since the early 20th century based on the beneficial effects of fasting. Later, the high fat/low carbohydrate ketogenic diet (KD) was designed as a long-term feasible method to mimic the metabolic effects of chronic food restriction (1). Accordingly, variations of the KD are now used in the clinic for the treatment of certain childhood epilepsies. Metabolic research in epilepsy is largely focused on discerning the mechanism(s) by which KDs exert antiseizure effects with the hope of identifying drug targets and developing “the ketogenic diet in a pill” (2). Key candidate mechanisms underlying the efficacy of the KD involve bypassing glycolysis, utilization of fatty acids and ketone bodies as alternate fuels, enhanced opening of K_ATP channels to reduce neuronal firing rates, and redox control (1). However, the exact mechanisms underlying the dietary control of neuronal excitability remain unknown.

The high energy demands of the brain are met by a complex integration of both oxidative and non-oxidative energetic processes. Neurons maintain low energy reserves, which require the use of alternative fuel sources and metabolic coordination with astrocytes to maintain ATP demands. A key element to achieving this is the astrocyte-neuron lactate shuttle (ANLS), which is initiated in astrocytes with glucose uptake and oxidation via glycolysis to pyruvate with subsequent reduction to lactate via lactate dehydrogenase (LDH). Lactate is then exported into the extracellular space, where it is taken up by neurons and converted by cytosolic LDH to pyruvate for further oxidation in the mitochondrial TCA cycle. This “fermentation” of glucose to lactate via glycolysis and LDH has been shown to correlate with enhanced neuronal activity (3). Since astrocytes (but not neurons) use glycogen for energy storage, the ANLS pathway allows astrocytic glycogen breakdown and export of lactate to sustain neuronal activity during prolonged activation. Stimulation of the ANLS pathway by neuronal activity has also been attributed to astrocytic glutamate uptake and the regeneration of NAD+ to maintain glycolytic flux in astrocytes (4).

Recent work by Sada et al. has identified a novel metabolic pathway to control seizures via inhibition of LDH and the ANLS. As discussed below, the authors provide four key pieces of evidence to support their findings. They establish 1) neuronal hyperpolarization following glucose restriction, 2) the ability of LDH inhibition in astrocytes to mimic effects of glucose restriction, 3) suppression of seizure activity in vivo by LDH inhibition, and 4) a novel action of analogs of stiripentol, an antiseizure drug (ASD), on LDH activity.

Glucose Restriction Induces Hyperpolarization in Excitatory Cells

To investigate the direct effects of changes in metabolism on neuronal membrane potentials, the authors performed patch-clamp recordings in slices from the subthalamic nucleus (STN) of the basal ganglia. Replacement of glucose with ketone bodies (β-hydroxybutyrate [BHB] or acetoacetate) elicited hyperpolarization in STN cells, which was then abolished by lactate. Ketone bodies are directly oxidized in the mitochondria via the TCA cycle, whereas neuronal glucose has two metabolic fates: 1) glycolysis in neurons, where it is converted to pyruvate, and 2) uptake by astrocytes for glycolytic metabolism to produce lactate, which is then released to the extracellular space via the ANLS.

LDH Inhibition in Neurons and Astrocytes Controls Neuronal Excitability

The authors probed the mechanism behind the attenuation found with lactate and focused on LDH. Selective inhibition of neuronal LDH via intracellular administration of oxamate similarly hyperpolarized STN cells and decreased membrane excitability; these effects were also reversed by K_ATP channel blockers and the downstream metabolites, pyruvate and oxaloacetate. However, additional substrates and products of LDH activity (lactate, NADH, acetyl-CoA, alanine), as well as other energy substrates (BHB and ATP) had no effect. Interestingly, the hyperpolarization induced by LDH inhibition was recovered by α-ketobutyrate (α-KB), a compound structurally similar to pyruvate and oxaloacetate yet not an energetic substrate. The authors suggest this might indicate a possible direct action of pyruvate, oxaloacetate, and α-KB to close K_ATP channels versus a change in the metabolic pathways elicited by downstream metabolites of LDH.

As isoforms of LDH are expressed in both neurons and astrocytes, Sada et al. further investigated metabolic control of neuronal membrane potential by selectively inhibiting LDH in astrocytes. Double recordings of hippocampal CA1 pyramidal cells and juxtaposed astrocytes demonstrated hyperpolarization of pyramidal cells with inhibition of LDH in astrocytes; however, this did not occur with lactate in the ACSF. This suggests control of membrane excitability by LDH via astrocytic lactate production and export with subsequent uptake by neurons for energy production.

Interestingly, the effects of oxamate and downstream metabolites of lactate on membrane potential were specific to excitatory cells, as LDH inhibition and glucose removal did not induce hyperpolarization in fast-spiking inhibitory cells. However, as a K_ATP channel agonist did elicit hyperpolarization in the inhibitory cells, the authors speculated that a link between LDH metabolism and K_ATP channels does not exist in this cell type. Also consistent with a specific effect of LDH inhibition on excitatory hippocampal neurons, oxamate was more effective at blocking excitatory postsynaptic currents (EPSCs) than inhibitory postsynaptic currents (IPSCs) in pyramidal cells.

Seizure Suppression by LDH Inhibition In Vivo

The effects of LDH inhibition on neuronal excitability was subsequently tested in rodent models of acute and chronic seizures. In the mouse pilocarpine model, LDH inhibition with oxamate decreased acute electrographic and behavioral seizures. In accordance with their in vitro results, the decrease in electrographic seizures with LDH inhibition was reversed with systemic administration of pyruvate, the product of neuronal LDH, but not the upstream substrate lactate. In the mouse kainate model, a single intraperitoneal injection of oxamate acutely attenuated spontaneous electrographic seizures. The results of pharmacological inhibition of LDH were then verified by administration of an antisense oligodeoxynucleotide to knockdown a subunit of LDH, LDHA, in the mouse hippocampus two weeks after kainate. A 30 to 35 percent decrease in hippocampal LDHA expression was sufficient to suppress acute electrographic seizures. However, the authors did not confirm concurrent decreases in LDH activity and lactate concentrations in this paradigm.

To test their theory that inhibition of LDH and the ANLS could mimic the KD, Sada et al. then assessed lactate levels and LDH activity in mice fed a KD. A reduction of hippocampal lactate concentrations was found, with no effect on LDH activity or protein expression. However, levels of pyruvate and oxaloacetate were not measured; therefore, it is unknown if the KD decreases in vivo concentrations of the two downstream metabolites of lactate, which are proposed by the authors to mediate closure of K_ATP channels to enhance neuronal excitability.

Stiripentol and Its Structural Analogs Inhibit LDH Activity In Vitro

Sada et al. then investigated the possibility that LDH inhibition could be an additional target of current ASDs. A screen of ASDs for effects on LDH activity identified stiripentol—a drug clinically used for the treatment of Dravet syndrome—as a modest inhibitor of both directions of the LDH reaction, the oxidation of lactate to pyruvate and the reduction of pyruvate to lactate. In the kainate model, acute administration of stiripentol weakly suppressed electrographic seizures. A search of compounds with structural similarity to stiripentol revealed isosafrole as a more potent LDH inhibitor. Isosafrole potently suppressed lactate formation by neuronal LDH1 and, importantly, LDH5, the isoform expressed in astrocytes, indicating that it may effectively block the production and transfer of lactate to neurons via the ANLS. Further, acute intraperitoneal injection of isosafrole strongly attenuated spontaneous spike discharges in the kainate model.

The intriguing study by Sada et al. suggests LDH and the ANLS as novel metabolism-based targets for the treatment of epilepsy. Although the authors suggest that partial LDH inhibition by stiripentol may contribute to its antiseizure efficacy, it is questionable whether any meaningful inhibition of brain LDH occurs at therapeutically relevant doses of stiripentol or isosafrole.

Further, the authors did not provide any in vivo measures of LDH inhibition of oxamate, stiripentol, or isosafrole. In the mouse pilocarpine model, pyruvate was shown to reverse the inhibitory effect of oxamate on seizures; however, this was not demonstrated with stiripentol or isosafrole. Additionally, neither were in vivo data provided on lactate levels or LDH activity in response to oxamate, stiripentol, or isosafrole (as was shown in mice fed the KD) or the ability of these drugs to affect the activity of other metabolic enzymes. Oxamate has been shown to stimulate the reverse reaction of pyruvate carboxylase (decarboxylation of oxaloacetate) in vitro (5); therefore, the specificity of these compounds for LDH inhibition in vivo remains to be fully determined.

The authors also provide evidence that short-term treatment with LDH inhibitors did not affect normal brain function in control animals, and body weights and spontaneous locomotion were not altered in mice with spontaneous electrographic seizures in the kainate model. However, since lactate is known to be neuroprotective and acts as a signaling molecule to affect other cellular targets and functions (6), including the formation of long-term memories (7), the chronic effects of LDH inhibition will need to be further evaluated.

In cancer models, LDH inhibition has been shown to simultaneously inhibit glycolysis and increase mitochondrial oxidation (8), similar to the metabolic effects of the KD (1). Further studies could determine if LDH inhibition impacts mitochondrial function by enhancing flux through the TCA cycle and electron transport chain. Additionally, it has been suggested that astrocytic production of lactate by LDH is necessary to regenerate NAD+ from the NADH produced during glycolysis (4). Hence, the effect of LDH inhibition on the redox state and energy metabolism in astrocytes also remains to be determined. Delineation of the bioenergetic consequences of LDH inhibition and the ANLS can guide the translational path of this target for chronic treatment of epilepsy.

Notwithstanding, the work by Sada et al. brings to light a new mechanism for the known control of seizure activity by glycolytic restriction via inhibition of LDH in astrocytes. The authors provide evidence that production of pyruvate from lactate in neurons enhances neuronal excitability—possibly through inhibition of K_ATP channels by pyruvate and other structurally similar downstream metabolites—and show the efficacy of acute LDH inhibition on attenuating seizure activity in vivo. Further, they identify a novel metabolic target for the ASD stiripentol and demonstrate the efficacy of isosafrole (a structurally similar and more potent LDH inhibitor) on the suppression of seizure activity in vivo. Collectively, their results highlight the potential utility of LDH inhibitors as a novel class of ASD and the importance of continued research into the metabolic control of epilepsy.