Seizure-Induced Sympathoexcitation Is Caused by Activation of Glutamatergic Receptors in RVLM That Also Causes Proarrhythmogenic Changes Mediated by PACAP and Microglia in Rats

Abstract

Cardiovascular autonomic dysfunction in seizure is a major cause of sudden unexpected death in epilepsy. The catecholaminergic neurons in the rostral ventrolateral medulla (RVLM) maintain sympathetic vasomotor tone and blood pressure through their direct excitatory projections to the intermediolateral (IML) cell column. Glutamate, the principal excitatory neurotransmitter in brain, is increased in seizures. Pituitary adenylate cyclase activating polypeptide (PACAP) is an excitatory neuropeptide with neuroprotective properties, whereas microglia are key players in inflammatory responses in CNS. We investigated the roles of glutamate, PACAP, and microglia on RVLM catecholaminergic neurons during the cardiovascular responses to 2 mg/kg kainic acid (KA)-induced seizures in urethane anesthetized, male Sprague Dawley rats. Microinjection of the glutamate antagonist, kynurenic acid (50 nl; 100 mM) into RVLM, blocked the seizure-induced 43.2 ± 12.6% sympathoexcitation (p ≤ 0.05), and abolished the pressor responses, tachycardia, and QT interval prolongation. PACAP or microglia antagonists (50 nl) (PACAP(6-38), 15 pmol; minocycline 10 mg/ml) microinjected bilaterally into RVLM had no effect on seizure-induced sympathoexcitation, pressor responses, or tachycardia but abolished the prolongation of QT interval. The actions of PACAP or microglia on RVLM neurons do not cause sympathoexcitation, but they do elicit proarrhythmogenic changes. An immunohistochemical analysis in 2 and 10 mg/kg KA-induced seizure rats revealed that microglia surrounding catecholaminergic neurons are in a "surveillance" state with no change in the number of M2 microglia (anti-inflammatory). In conclusion, seizure-induced sympathoexcitation is caused by activation of glutamatergic receptors in RVLM that also cause proarrhythmogenic changes mediated by PACAP and microglia.

Significance statement: Sudden unexpected death in epilepsy is a major cause of death in epilepsy. Generally, seizures are accompanied by changes in brain function leading to uncontrolled nerve activity causing high blood pressure, rapid heart rate, and abnormal heart rhythm. Nevertheless, the brain chemicals causing these cardiovascular changes are unknown. Chemicals, such as glutamate and pituitary adenylate cyclase activating polypeptide, whose expression is increased after seizures, act on specific cardiovascular nuclei in the brain and influence the activity of the heart, and blood vessels. Microglia, which manage excitation in the brain, are commonly activated after seizure and produce pro- and/or anti-inflammatory factors. Hence, we aimed to determine the effects of blocking glutamate, pituitary adenylate cyclase activating polypeptide, and microglia in the RVLM and their contribution to cardiovascular autonomic dysfunction in seizure.

Keywords: PACAP; glutamate; microglia; rat; seizure; sympathetic.

Figure 1.

Effect of bilateral RVLM microinjection of (A) PBS (50 nl) and (B) KYNA (50 nl; 100 mM) followed by 2 mg/kg intraperitoneal KA in an anesthetized rat (see Materials and Methods) showing the effect on the following: from the top, (i) HR (bpm), (ii) AP (arterial pressure; mmHg), (iii) SNA (%), and (iv) EEG (μV). Arrow indicates time of RVLM microinjections and intraperitoneal KA. Dotted arrows indicate the starting points for increase in EEG and/or SNA activity. Right side panels, Pre (a) and post (b) KA EEG represents the expanded waveform from respective period. Baseline (a), which is a pre-KA period with desynchronous waves, and (b) post-KA period with increased γ range frequencies and followed by the data from the same EEG, drawn as a power spectrum (c, d, respectively) (during post-KA period γ range frequencies, which is shown between two dotted lines) are increased. Increase in gamma range frequency (25–45 Hz) is characteristic property of KA-induced seizures (Olsson et al., 2006; Gurbanova et al., 2008).

Figure 2.

Effects of KA treatment on induction of seizures in hippocampus and central autonomic nuclei. Change in gamma range frequency (25–45 Hz) in hippocampal EEG and sympathetic nerve recordings every 10 min after 2 mg/kg KA injection. Arrow indicates time of RVLM microinjections and intraperitoneal KA. Dotted arrows indicate the starting points for increase in EEG and/or SNA activity. Induction of seizure activity in sympathetic nerve activity does not start at least until 70 min after KA injection, whereas hippocampal seizure activity starts ∼15–20 min after KA injection followed by an increase in SNA at 25–30 min. Time-dependent increases in hippocampal seizure activity occur up to 110 min after KA injection followed by a fall.

Figure 3.

In vivo effects of RVLM microinjection of PBS, PACAP(6–38), minocycline, and KYNA in 2 mg/kg KA-induced seizure rats. Change in SNA (AUC) between 60 and 120 min after intraperitoneal treatment (A), change in MAP 120 min after intraperitoneal PBS or KA injection (B), change in HR at 120 min after intraperitoneal PBS or KA injection (C), and log transform of percentage change in EEG activity (gamma wave frequency AUC), at 60 min (D) and 120 min (E) after intraperitoneal PBS or KA injection in different groups of rats after development of seizure. Statistical significance was determined using one-way ANOVA followed by t tests with a Holm-Šídák correction. Data are mean ± SEM. ****p ≤ 0.0001 compared with vehicle control group. ***p ≤ 0.001 compared with vehicle control group. **p ≤ 0.01 compared with vehicle control group. *p ≤ 0.05 compared with vehicle control group. ^#p ≤ 0.05 compared with KA control group.^##p ≤ 0.01.

Figure 4.

In vivo effects of PBS and KA-induced (2 and 10 mg/kg) seizures in rats studied for histology. Change in MAP (A) and HR (B), at 120 min after intraperitoneal PBS or KA (2 and 10 mg/kg) injection and percentage change in EEG activity (gamma wave frequency AUC), at 60 min (C) and 120 min (D) after intraperitoneal PBS or KA (2 and 10 mg/kg) injection in different groups of rats. In all groups, n = 5. Statistical significance was determined using one-way ANOVA followed by t tests with a Holm-Šídák correction. Data are mean ± SEM. **p ≤ 0.01 compared with vehicle control group. *p ≤ 0.05 compared with vehicle control group.

Figure 5.

Fluorescence images of RVLM area containing TH⁺-ir (red), Iba1-labeled microglia (yellow), and CD206-labeled M2 microglial cells (green) and their morphological analysis in different treatment groups of rats. Scale bar, 20 μm. TH, Iba1 and CD206 immunoreactivity in RVLM in PBS (A), 2 mg/kg KA (B), and 10 mg/kg KA (C) treated rats. In all of these three groups, TH⁺-ir neurons (red) were surrounded with microglia with its round cell body and normal appearing processes with few ramifications (closed arrow) and no change in number of anti-inflammatory M2 microglia (open arrow). Quantitative analysis of number of microglial cells in mean square area (D), number of end processes/microglia (E), branch length (μm)/microglia of Iba1-labeled microglial cells (F), and percentage of CD206-labeled M2 microglial cells (G) in the RVLM of vehicle-treated and KA-induced seizure (2 and 10 mg/kg i.p.) rats.

Figure 6.

Proarrhythmogenic effects of seizures. Group data showing changes in QTc interval (A) and PR interval (B) 120 min after intraperitoneal injection of PBS or KA in different groups of rats. Statistical significance was determined using one-way ANOVA followed by t tests with a Holm-Šídák correction. Data are mean ± SEM. **p ≤ 0.01 compared with vehicle control group. *p ≤ 0.05 compared with vehicle control group. ^#p ≤ 0.05 compared with KA control group.

Figure 7.

Representative Poincare plots illustrate the changes in QT interval (A) and PR interval (B) following KA-induced seizures in rats. A, Treatment with KA causes a dramatic dispersion in the QT interval (prolongation) and arrhythmic behavior in the heart rate (multiple ellipses) (II). In rats treated with the PACAP antagonist (PACAP(6–38)) or with the microglial antagonist minocycline, prolongation of QT interval and the dysrhythmia is abolished (III, IV). Following treatment with KYNA, the HR and QT are restored to normal (V). HR-triggered ECG was drawn before and after treatment and shown in the right side corner of each box. Continuous black and dotted red lines indicate pretreatment and post-treatment ECG, respectively. B, Induction of seizures with intraperitoneal KA injection shortened the PR interval (II) compared with vehicle control. RVLM microinjection of PACAP(6–38) (III), minocycline (IV), or KYNA (V) did not show changes in seizure-induced short PR interval; however, PACAP(6–38) (III) and minocycline (IV) showed more dispersed PR interval with multiple ellipses. Scale bars are in milliseconds.

Figure 8.

A proposed mechanism by which hippocampal seizures induce increased activity of sympathetic premotor neurons in the RVLM and role of glutamate, PACAP, and microglia. A, Seizure elevates synaptic glutamate release that can act on postsynaptic AMPA or NMDA receptors. Activation of AMPA or NMDA receptors leads to inhibition of cysteine uptake and influx of extracellular calcium, which stimulates production of oxidants, NO and O⁻. Under repetitive and extreme neuronal activation, neurotoxic effects are mediated through increased production of apoptotic factor-like caspase-3. Glutamate transporters are expressed by astrocytes and play an important role in rapid clearance of the synaptically released glutamate, whose expression is downregulated in seizure. Together, increased oxidative stress and cellular excitability cause increased activity of sympathetic premotor neurons. B, Increased PACAP expression can act via cAMP-mediated PKA and/or PKC pathways and produce either excitatory effect through phosphorylation of TH at serine 40 or neuroprotective effect regulated through decreased caspase 3, increased glial-glutamate transporters, and redirecting microglia toward anti-inflammatory M2 phenotype. In neurons, PACAP inhibits MAPK and increases IL-6 production. Microglia are activated by PACAP binding to PAC1 and VPAC1 receptors. Subsequently, microglia increase production and release of IL-10 and TGF-β and decrease production and release of TNF-α, as well as downregulating CD40 and B7 surface protein expression, with a neuroprotective effect. Conversely, the pro-inflammatory phenotype of activated microglia can produce IL-1β and TNF-α that may increase the sensitivity of neurons to activation.