Chronically altered NMDAR signaling in epilepsy mediates comorbid depression

Abstract

Depression is the most common psychiatric comorbidity of epilepsy. However, the molecular pathways underlying this association remain unclear. The NMDA receptor (NMDAR) may play a role in this association, as its downstream signaling has been shown to undergo long-term changes following excitotoxic neuronal damage. To study this pathway, we used an animal model of fluoxetine-resistant epilepsy-associated depression (EAD). We determined the molecular changes associated with the development of depressive symptoms and examined their response to various combinations of fluoxetine and a selective neuronal nitric oxide synthase inhibitor, 7-nitroindazole (NI). Depressive symptoms were determined using the forced swim test. Furthermore, expression and phosphorylation levels of markers in the ERK/CREB/ELK1/BDNF/cFOS pathway were measured to determine the molecular changes associated with these symptoms. Finally, oxidative stress markers were measured to more clearly determine the individual contributions of each treatment. While chronic fluoxetine (Flxc) and NI were ineffective alone, their combination had a statistically significant synergistic effect in reducing depressive symptoms. The development of depressive symptoms in epileptic rats was associated with the downregulation of ERK2 expression and ELK1 and CREB phosphorylation. These changes were exactly reversed upon Flxc + NI treatment, which led to increased BDNF and cFOS expression as well. Interestingly, ERK1 did not seem to play a role in these experiments. NI seemed to have augmented Flxc's antidepressant activity by reducing oxidative stress. Our findings suggest NMDAR signaling alterations are a major contributor to EAD development and a potential target for treating conditions associated with underlying excitotoxic neuronal damage.

Keywords: Brain-derived neurotrophic factor; Depression; Epilepsy; Immediate early genes; Intracellular signaling; MAP kinase signaling system; N-methyl-D-aspartate receptors; Nitric oxide; Nitric oxide synthase type I; Nitrosative stress.

Fig. 1

Validation of EAD development in post-SE rats. a The hippocampal insertion site of the LFP electrode and a sample recording of one of the spontaneous seizures observed 30 days after SE induction. b Comparison of depressive symptoms in the FST between post-SE (Saline) and Sham rats. c & d Comparison of the number of NeuN-immunoreactive (c) and TUNEL-stained (d) cells in 3 subregions of the hippocampus: CA1, CA3, and DG; between post-SE (Saline) and Sham rats. All comparisons were done using the Welch’s t-test and statistical significance is shown as follows: *p < .05, **p < .01

Fig. 2

Assessing the therapeutic efficacy of the treatment combinations based on their effects on change scores of depressive symptoms between the FSTs. Change scores are calculated for each animal by applying the following formula to the relevant values: (FST2–FST1)/FST1. a & d Establishing the baseline change of depressive symptoms between FST1 and FST2 without administration of any treatments. b & e The chronic treatment setting: comparing the effects of Flxc, NI, and their combination on immobility (b) and latency (e) times. c & f The acute treatment setting: comparing the effects of Flxa, NI, and their combination on immobility (c) and latency (f) times. a & d Change scores were compared using the Welch’s t-test and no statistically significant differences were found. b–c & e–f While change scores are shown in the plots for easier interpretation, the data was actually analyzed using two-way ANCOVA using values from FST1 as the covariate. Details of this analysis are presented in the main text

Fig. 3

Analysis of the protein changes underlying the observed therapeutic effects in the behavioral assays. a–l For each marker, representative bands from the western blot studies are shown when applicable. Band intensities were corrected based on their corresponding GAPDH expression (a) and the fold change of all markers were calculated compared to the Saline group. The Sham and Saline groups are compared using the Welch’s t-test to determine the changes associated with EAD development and statistical significance is shown as follows: *p < .05. Afterwards, the Saline, NI, Flxc, and Flxc + NI groups are compared using two-way ANOVA to determine the changes brought about by the treatments. Details of the results of this analysis are presented in the main text

Fig. 4

Comparison of the change profiles of ERK1/2, ERK1, and ERK2. a–i For each marker, band intensities were corrected based on their corresponding GAPDH expression and their fold changes were calculated compared to the Saline group. The Sham and Saline groups are compared using the Welch’s t-test to determine the changes associated with EAD development and statistical significance is shown as follows: *p < .05. Afterwards, the Saline, NI, Flxc, and Flxc + NI groups are compared using two-way ANOVA to determine the changes brought about by the treatments. Details of the results of this analysis are presented in the main text

Fig. 5

Comparison of oxidative stress markers between the study groups. a Comparison of nitric oxide concentrations. b Comparison of malondialdehyde concentrations. c Comparison of reduced glutathione concentrations. a–c The Sham and Saline groups are compared using the Welch’s t-test to determine the changes associated with EAD development and statistical significance is shown as follows: *p < .05, **p < .01. Afterwards, the Saline, NI, Flxc, and Flxc + NI groups are compared using two-way ANOVA to determine the changes brought about by the treatments. Details of the results of this analysis are presented in the main text

Fig. 6

Comparison of physiological and pathological NMDAR signaling as suggested by the findings in this study. Long-term pathological changes following excitotoxic injury result in the predominance of the NR2B subunit of NMDAR which inhibits downstream ERK signaling upon overstimulation. Furthermore, these long-term changes lead to the constitutive activity of nNOS and increased NO levels, resulting in mitochondrial impairment, inhibition of protein synthesis, and direct inhibition of ERK signaling through nitrosylation