Glucose is the main source of energy for neurons and changes in glucose levels alter brain function. Hypoglycemia commonly occurs in patients with diabetes mellitus as a side effect of insulin treatment (Herbel and Boyle, 2000), or in hyperinsulinism due to insulinoma (Malouf and Brust, 1985), and it is a critical problem in the management of these diseases. Hypoglycemia presents with a variety of neurological impairments, interferes with everyday life, and creates a serious psychological burden for the patient. Seizures are one of the most common acute symptoms (Malouf and Brust, 1985). To better understand the neurologic complications, animal models have been used. In rats, insulin administration induces hypoglycemia, which may lead to generalized seizures with barrel rotations as the main feature (Gastaut et al., 1968).
The substantia nigra pars reticulata (SNR) has been identified as a critical midbrain structure involved in motor control and modulation of seizures (Iadarola and Gale, 1982; Depaulis et al., 1994; Velíšková and Moshé, 2006). The SNR contains mainly GABAergic neurons with high spontaneous firing rates. The seizure control is achieved when the activity of the SNR GABAergic neurons is decreased. Thus, focal enhancement of GABA-mediated inhibitory effects by localized bilateral SNR microinfusions of muscimol (a GABAA receptor agonist) have anticonvulsant effects in several models of generalized seizures (Iadarola and Gale, 1982; Depaulis et al., 1994; Velíšková and Moshé, 2006). Similarly, treatments that decrease the excitatory glutamatergic neurotransmission in the SNR, such as infusions of AP7 (an NMDA receptor antagonist) also have anticonvulsant effects (Wurpel et al., 1992; Velíšková et al., 2001).
Since fasting or irregular food intake are predisposing factors for seizure expression during hypoglycemic crises in humans (Malouf and Brust, 1985), here we used an animal model to investigate the incidence of barrel rotations, a hallmark of hypoglycemia-induced seizures (Kirchner et al., 2006), in nonfasted and fasted rats. Further, we tested the involvement of the SNR in the regulation of hypoglycemia-induced barrel rotations using pharmacological manipulations by GABAergic and glutamatergic agents.
Adult Sprague-Dawley male rats were used. Rats were housed under standard conditions in our accredited animal facility. Nonfasted rats had a free access to food and water. Fasted rats had food removed the day before testing (for 18 h) with free access to water.
In the first experiment, we tested the effect of fasting on the incidence of hypoglycemia-induced barrel rotations. Insulin was injected intraperitoneally (i.p.) 30 IU/kg to induce severe hypoglycemia (<20 mg/100 ml). Blood glucose levels were monitored every 30 min by precise One Touch Basic glucosemeter (Schwechter et al., 2003). When the rats reached severe hypoglycemia, they were observed up to 6 h and the occurrence of barrel rotations was recorded.
In the second experiment, only fasted rats were used. Nonfasted rats were not used because data from the first experiment indicated that not all nonfasted rats would reliably develop any seizure behavior. This fact would preclude precise evaluation of incidence in this experiment. Rats were implanted with bilateral cannulae into the SNR as described previously (Velíšková and Moshé, 2001) and allowed for 2 days of recovery. Muscimol (a GABAA receptor agonist) at a dose 0.1 μg (Velíšková and Moshé, 2001), baclofen (a GABAB receptor agonist) at a dose 0.2 μg (Sperber et al., 1989a), or AP7 (an NMDA receptor antagonist) at a dose 1.1 μg (Velíšková et al., 2001) were used. The doses were chosen based on efficacy of SNR infusions of respective agents in other seizure models. Muscimol and baclofen were dissolved in saline, AP7 was dissolved in 1N sodium hydroxide and the pH was adjusted to 7.4 with 0.1M sodium phosphate buffer (Velíšková et al., 2001). Just prior to insulin injection (30 IU/kg i.p.), the drugs were infused bilaterally into the SNR in a volume of 0.25 μl/per side. Controls received infusions of 0.25 μl/per side of vehicle. Rats were observed until the first barrel rotation occurred or up to 6 h following insulin injection when the experiment was terminated. The incidence of barrel rotations was monitored. In this experiment, an additional requirement was the symmetrical bilateral cannula placements within the SNR, which has been confirmed under the light microscope in thionin-stained sections.
The incidence of barrel rotations was compared using Fisher's exact test. Statistical significance was set to p < 0.05.
We found that in fasted rats, the incidence of barrel rotations was significantly higher compared to nonfasted rats (p < 0.05). While 100% of fasted rats (n = 25) had barrel rotations, only 43.9% of nonfasted rats (n = 21) had this type of seizure behavior within the observation period. This finding prompted us to use only fasted rats for the second experiment.
The effects of intranigral drug infusions on the incidence of barrel rotations are summarized in Table 1. Muscimol infusions (n = 14) in the SNR had no effect on the incidence of barrel rotations (p > 0.05) compared to saline-infused controls (n = 7). On the other hand, baclofen (n = 11) significantly decreased the incidence of barrel rotations (p < 0.05) compared to saline-infused rats (n = 12). Similarly, SNR infusions of AP7 (n = 10) significantly suppressed the incidence of barrel rotation (p < 0.05) compared to vehicle-infused rats (n = 11). In addition, muscimol had no effect on the onset of barrel rotations (muscimol: 120.1 ± 7.5 min, saline: 144.7 ± 17.6 min; p > 0.05), while both baclofen (baclofen: 183.5 ± 18.0 min, saline: 120.7 ± 10.0 min; p < 0.05) and AP7 delayed the onset of barrel rotations compared to vehicle-infused rats (AP7: 200.0 ± 9 min, vehicle: 126.6 ± 13.8 min; p < 0.05). Please note that in the AP7 group, the statistical analysis on the barrel rotation onset was done on n = 2 due to the incidence suppression.
|Agent||Effect||Dose per SNR (in 0.25 μl)||Incidence of barrel rotation|
|Muscimol||GABAA receptor agonist||0.1μg||14/14||6/7|
|Baclofen||GABAB receptor agonist||0.2μg||5/11*||11/12|
|AP7||NMDA receptor antagonist||1.1 μg||2/10*||10/11|
Fasting is a predisposing factor for seizure expression during hypoglycemic crisis in humans (Malouf and Brust, 1985). We found that overnight fasting in rats increased the incidence of barrel rotations, a characteristic phenotype of hypoglycemic seizures (Kirchner et al., 2006). Our data suggest that fasting predisposes the rats to express seizures during hypoglycemia similarly as in humans. Further, our data demonstrate that the SNR is involved in regulation of hypoglycemic seizures because bilateral intranigral microinfusions of baclofen or AP7 significantly decreased the incidence of barrel rotations. Fasting decreases GABA release in the SNR (During et al., 1995). Thus, we can speculate that the lack of GABA within the SNR during fasting may predispose for the hypoglycemic seizures to occur.
The incidence of hypoglycemic seizures was decreased by SNR infusions of baclofen, but not muscimol. In addition, baclofen but not muscimol delayed the onset of barrel rotations. These results were surprising because in many models of generalized seizures in adult rats, intranigral muscimol infusions have anticonvulsant effects (for review see [Depaulis et al., 1994; Velíšková and Moshé, 2006]), while baclofen infusions are ineffective at least in the flurothyl seizure model (Sperber et al., 1989a). The failure of muscimol to affect the incidence and onset of barrel rotations may be due to lack of available GABAA receptors. During fasting GABAA receptors may be internalized (Chapell et al., 1998). On the other hand, GABAB receptors are very stable at the neuronal membrane and internalization rarely occurs (Fairfax et al., 2004). Thus, we speculate that the GABAB receptor-mediated effects may substitute for the loss or decreased action of GABAA receptors during abnormal or stressful situations, such as fasting and insulin-induced hypoglycemia. Based on this hypothesis, we would expect that SNR muscimol infusions might decrease or prevent the incidence of barrel rotations during hypoglycemia. However, the fact that the incidence of barrel rotation in nonfasted rats is only about 44% would preclude determining exactly the reason, why the seizure did not occur in nonfasted animals. Another explanation for the lack of SNR muscimol effects on the occurrence of barrel rotations may be due to loss of its bioactivity at the time of their onset. However, our pilot experiment using muscimol infusions 30 min prior to the expected barrel rotation onset suggests that this may not be the case because muscimol still did not affect the occurrence or onset of barrel rotation (unpublished observation). It is also possible that fasting leads to changes in GABAA receptor sensitivity. We used a relatively high dose of muscimol since in nonfasted rats, the effective dose of SNR muscimol against different types of seizures is already 25ng/0.25μl (Garant and Gale, 1986; Sperber et al., 1989b). However, more studies such as dose response curves for muscimol or infusions with other GABAA receptor agonists are necessary to disqualify the role of GABAA receptors during hypoglycemic seizures. Finally, a direct testing of possible GABAA receptors internalization and their subunit composition changes, which may lead to alterations in their sensitivity to muscimol following fasting, would bring light on potential mechanisms involved in hypoglycemic seizures.
The incidence of hypoglycemia-induced barrel rotations was also decreased by SNR infusions of AP7. This is in agreement with observations by other investigators using different seizure models (Maggio and Gale, 1989; Velíšková et al., 2001). These data support the idea that a decrease of GABAergic neuron firing within the SNR has general antiseizure effects (Velíšková and Moshé, 2006) including hypoglycemic seizures. In addition, the data show that increased glutamatergic input into the SNR is involved during hypoglycemic seizures. This is indirect evidence that the glutamate release may not be decreased or being decreased to a lesser extent during fasting and hypoglycemia compared to the GABA release (During et al., 1995). Indeed, our pilot experiment using c-Fos expression suggests that the main SNR glutamatergic input from the subthalamic nucleus (Bolam et al., 2000) may be involved during hypoglycemic seizures (unpublished observation).
Our data show that during a metabolic stress, the SNR seizure controlling system has different properties than in rats naïve to such condition. This finding may be important not only for the expression of hypoglycemic seizures but also for the maintenance of seizures during status epilepticus, which is associated with metabolic disturbances including the development of hypoglycemia (Wasterlain et al., 1993).