**1. Introduction**

Hypoglycemia is a potentially serious condition that occurs when blood glucose levels quickly fall below a specific threshold concentration. The most common cause of hypoglycemia is misuse of medications such as insulin and sulfonylureas to control blood glucose levels [1,2]. Medications that cause blood glucose levels to fall sharply or repeatedly can lead to hypoglycemic shock. In these cases, the first treatment for patients who present at the hospital in a hypoglycemic state of shock is to rapidly raise blood glucose levels, which is essential for resuscitating the patient and impossible to avoid. Clinically, patients display decreased cognitive abilities over time following such episodes [3], and our

previous paper revealed that this condition causes an increase in neuronal cell death [4]. Therefore, this study is important to identify means to minimize neuronal cell death caused by hypoglycemia and glucose reperfusion, which is performed as an essential step in treating hypoglycemia.

Therefore, type 1 and type 2 diabetic patients are at an increased risk of experiencing hypoglycemia. Hypoglycemia is classified into three categories depending on the level of blood glucose: mild, moderate, and severe. Mild hypoglycemia occurs when the concentration of blood glucose falls below 70 mg/dL, leading to symptoms such as headache, sweating, and irritation. Moderate hypoglycemia (below 55 mg/dL) is more severe; thus, people need to regulate this condition carefully. The most serious form of hypoglycemia occurs when blood glucose levels fall below 35 mg/dL. Severe hypoglycemia can lead to seizure, loss of consciousness, and death in the most extreme cases. Previous studies demonstrated that some degree of the observed neuronal death is caused by hypoglycemia itself, but glucose reperfusion to recover the patient from critically low glucose levels acts to drive secondary injury, thereby inducing a second, more severe wave of neuronal death [5,6]. This process is called "glucose-reperfusion injury" after hypoglycemia by our laboratory [4,7]. This injury promotes an imbalance in several cellular programs that can cause neuronal cell death, including the production of reactive oxygen species (ROS), destruction of the blood–brain barrier (BBB), microglial activation associated with inflammation [8], and excessive zinc release [9].

The enzyme pyruvate dehydrogenase (PDH) has three main components: E1, decarboxylates pyruvate and acetylates lipoic acid; E2, dihydrolipoamide acetyltransferase, which uses covalently bound lipoic acid; and lipoic acid is reoxidized by E3, dihydrolipoyl dehydrogenase. In addition, there are other subunits, the E3 binding protein and two complex-controlling enzymes: PDH kinases (PDKs), which inactivate PDH, and PDH phosphatase, which reactivates the PDH. PDH is a key enzyme that catalyzes the oxidative decarboxylation of pyruvate to form acetyl-coenzyme A (acetyl-CoA) under normal conditions. PDH controls the influx of pyruvate into the mitochondria to initiate oxidative metabolism and is an important regulator of the citric acid cycle [10,11]. PDH dysfunction or deficiency most often arise due to mutations or inactivation by PDKs via E1, which causes malfunction of the citric acid cycle due to PDH inactivation and leads to cell death [12]. Previous studies reported that starvation and diabetes lead to enhanced PDK activity, leading to decreased activity of PDH [13–18]. This decreased PDH activity results from increased PDK expression. Starvation increases the level of PDK2 expression in liver and kidney [19].

PDKs have four isoforms: PDK1, PDK2, PDK3, and PDK4. PDK1 is expressed in the heart, pancreatic islet cells, and in muscles [20–22]. PDK3 is only been found in the testis and kidneys in small quantities [23]. PDK4 expression was observed in the heart, skeletal muscles, liver, and brain [19,22,24]. PDK2 is the most abundant isoenzyme in the rat brain [23], with PDK4 hardly being expressed under basal physiological conditions. PDKs phosphorylate the E1 subunit, thus inactivating PDH [25].

In general, previous studies showed that PDK inhibits PDH via phosphorylating PDH after brain injuries, such as ischemia or epilepsy, thereby restricting the pyruvate oxidation (PO) pathway available to brain cells. As a result, under this pathological condition, the citric acid cycle cannot proceed and the production of ATP decreases, leading to cell death [26,27].

Sodium dichloroacetate (DCA) has been studied for a long time, especially in cancer research. DCA, a known activator of PDH and inhibitor of PDK, is a mitochondria-targeting small molecule that can penetrate most tissues even after oral administration [28]. It is responsible for maintaining PDH in the dephosphorylated active form and improving the oxidation of pyruvate via inhibition of PDK [29]. DCA has been shown to instigate inhibition of fatty acid oxidation, which eventually can lead to an increase in acetyl-CoA concentration [30]. DCA stimulates glucose and lactate oxidation, thereby improving recovery of injury after ischemia [31–33]. Although DCA has toxic side e ffects in high doses, it has highly beneficial e ffects on outcomes for many diseases such as ischemia and cardiac arrest [34,35].

In our laboratory, we have suggested a number of possible methods for aiding in the prevention of hippocampal neuron death after hypoglycemia [4,8,9,36–39]. In our previous studies, we revealed that administration of pyruvate reduced severe hypoglycemia-induced neuronal cell death and cognitive impairment [37]. Therefore, we questioned whether supplementation of DCA (100 mg/kg, i.v., two days) could improve the oxidation of pyruvate and reduce severe hypoglycemia-induced neuronal cell death.

We found that DCA treatment inhibited PDK and significantly reduced oxidative stress, microglial activation, BBB disruption, and thus hippocampal neuronal death after severe hypoglycemia via enhancing PDH activation.

#### **2. Materials and Methods**

## *2.1. Ethics Statement*

This study was conducted in strict accordance with the recommendation of the Institutional Animal Studies Care and Use Committee of the Hallym University in Chuncheon, Korea (Protocol # Hallym-2017-3). Animal sacrifice was conducted using isoflurane anesthesia and all attempts were made to minimize pain or distress.
