Abstracts

Rationale: Depression represents one of the most common, disabling, and serious comorbidities in patients with temporal lobe epilepsy (TLE). Our goal is to clarify the network that underlies depressive-like behavior in TLE, leading to more effective anatomical structures to target (through use of such neurosurgical techniques as lesions and neurostimulation) for the treatment of depression in these patients. Our central hypothesis is that the central nucleus of the amygdala (CeA) and the dorsolateral bed nucleus of the stria terminalis (dlBNST) are part of a shared neural network underlying anhedonic behavior in TLE, and that selective activation of dlBNST neurons will effectively decrease this depressive-like behavior.

Methods: Male Sprague Dawley rats demonstrating an average sucrose preference of above 90% over a two-week period, were implanted with an osmotic pump infusing glutamine synthetase inhibitor and proconvulsant, methionine sulfoximine (MSO), unilaterally into the posterior CeA. Recording EEG depth electrodes were placed in the anterior CeA and screw electrodes were implanted into the subdura. Stimulating electrodes were placed in the dlBNST. Rats were given one week to recover from the surgery prior to EEG recording. Pre-stimulation sucrose preference testing was recorded from all rats daily on the third week post-MSO surgery. The three rats in the sham stimulation (no current passed) group received sham stimulation daily on the 4^th week, 5^th week, and 6^th week post-surgery. In rats receiving BNST stimulations (n=5), these stimulations occurred daily during only a one-week period. The daily average during the one week of stimulation was compared to the matched daily average during one week of sham stimulation. To match stimulation time points with the sham stimulated rats, three BNST rats were stimulated on week 4, two on week 5, and one BNST rat was stimulated on week 6.

Results: Repeated measures ANOVA indicated a significant effect of time (F=12.48, p< 0.0001), with post hoc test demonstrating significantly lower sucrose preference following MSO treatment when compared to pre-MSO baseline (p < 0.01 in the designated sham stimulation group and p=0.003 in the designated BNST stimulation group). Following sham stimulation, sucrose preference remained significantly lower than baseline levels, (p < 0.01). However, rats receiving dlBNST stimulation demonstrated no significant difference in sucrose preference from baseline levels, but significantly differing from the post-MSO pre-stimulation anhedonic period (p < 0.03). All rats demonstrated spontaneous seizures.