Abstracts

Rationale: DBA/1 mice are a relevant model of sudden unexpected death in epilepsy (SUDEP). SUDEP is a major neurological public health problem, but the brain sites and neurochemical mechanisms of SUDEP are unknown, which hinders prevention. Witnessed human SUDEP most often follows generalized seizures and respiratory dysfunction, and DBA/1 mice are susceptible to generalized audiogenic seizures (AGSz) that lead to seizure-induced respiratory arrest (S-IRA). Magnetic resonance imaging (MRI) studies in human SUDEP revealed atrophy at specific brainstem structures implicating them in the neuronal network of SUDEP (Mueller et al., Neuroimage Clin 2014; 5: 208-16). Therefore, we evaluated changes in brain activity using manganese-enhanced MRI (MEMRI) to map the neuronal network of DBA/1 mice.Methods: DBA/1 mice received manganese chloride (80 mg/kg) 10 hr prior to AGSz. Groups included DBA/1 mice that exhibited AGSz-induced S-IRA, or AGSz but no respiratory arrest (NS-IRA) induced by administration of a selective serotonin reuptake inhibitor (SSRI, fluoxetine, 45 mg/kg). Controls included DBA/1 mice that were not exposed to AGSz. Immediately after AGSz the mice were anesthetized, and the brains were fixed and then scanned with a 14.1T Varian microimagerc. A high image quality spin-echo-multi-slice protocol was employed for the T1 mapping. TIFF images acquired were analyzed using Image J software and statistically compared using ANOVA.Results: A significant (p<0.001) overall increase in the MEMRI-based neuronal activity of the NS-IRA group vs. S-IRA and control DBA/1 group 110.24±0.72 vs. 104.34±0.70 and 103.26±0.61. Detailed analysis revealed significant (p<0.05) increases in activity at Kӧlliker-Fuse nuc. (123.6±4.4 vs. 103.4±1.7), pontine raphe nuc. (126.1±6 vs. 104.2±1.8), lateral dorsal tegmental area (125.2± 5.6 vs. 110.9±2.4), and inferior colliculus (IC) (111.5±2.7 vs. 102.1±0.6) at anterior/posterior (AP) -5.3 (Franklin and Paxinos Mouse Stereotaxic Atlas, 1997); activity in the dorsomedial periaqueductal gray (PAG) was also significantly elevated (113.6±1.6 vs. 100.7±1.7) at AP -3.4, in the NS-IRA compared to S-IRA group. The finding that fluoxetine, which blocked S-IRA (but not AGSz), resulted in increased activity in PAG and other brainstem sites suggests that these brain regions may be important to the ability of SSRIs to block S-IRA in DBA/1 mice and to reduce peri-ictal respiratory depression in epileptic patients (Bateman et al., Epilepsia 51:2211-14, 2010).Conclusions: These data indicate that the neuronal network for AGSz and S-IRA in DBA/1 mice include several nuclei that were atrophied in human SUDEP neuroimaging studies, including the PAG, IC, and raphe nuclei (Mueller et al., 2014). The similarity of the brainstem sites in DBA/1 mice and human MRI findings supports an increasing relevance of the DBA/1 mouse model to human SUDEP and suggests brain structures that are important to fluoxetine-mediated inhibition of S-IRA. (Support: Epilepsy Foundation, CURE)