Gene profile of electroconvulsive seizures: induction of neurotrophic and angiogenic factors

Abstract

Electroconvulsive seizure therapy (ECS) is a clinically proven treatment for depression and is often effective even in patients resistant to chemical antidepressants. However, the molecular mechanisms underlying the therapeutic efficacy of ECS are not fully understood. One theory that has gained attention is that ECS and other antidepressants increase the expression of select neurotrophic factors that could reverse or block the atrophy and cell loss resulting from stress and depression. To further address this topic, we examined the expression of other neurotrophic-growth factors and related signaling pathways in the hippocampus in response to ECS using a custom growth factor microarray chip. We report the regulation of several genes that are involved in growth factor and angiogenic-endothelial signaling, including neuritin, stem cell factor, vascular endothelial growth factor (VEGF), VGF (nonacronymic), cyclooxygenase-2, and tissue inhibitor of matrix metalloproteinase-1. Some of these, as well as other growth factors identified, including VEGF, basic fibroblast growth factor, and brain-derived neurotrophic factor, have roles in mediating neurogenesis and cell proliferation in the adult brain. We also examined gene expression in the choroid plexus and found several growth factors that are enriched in this vascular tissue as well as regulated by ECS. These data suggest that an amplification of growth factor signaling combined with angiogenic mechanisms could have an important role in the molecular action of ECS. This study demonstrates the applicability of custom-focused microarray technology in addressing hypothesis-driven questions regarding the action of antidepressants.

Figure 1.

Dual color image of the custom growth factor microarray chip. PCR products of 645 genes were printed in duplicate on Corning UltraGAPS slides. The glass chip was simultaneously hybridized to cDNA from sham and ECS-treated animals that were indirectly labeled using Cy3 and Cy5 dendrimers, respectively. Four micrograms of hippocampal total RNA was used to generate cDNA for each experiment. The magnified inset shows the housekeeping genes β-actin, cyclophilin, and β-tubulin (yellow box), an upregulated gene, Cox-2 (red box), and a downregulated gene, 3′ untranslated region of cyclin D1 (green box).

Figure 2.

Classification of genes upregulated by ECS. Upregulated genes from both acute and chronic ECS-treated rats from two time points (2 and 6 hr) are classified on the basis of known functional roles into the following four categories: (1) growth factor signaling, (2) angiogenesis and vasodilation, (3) neurotransmitter signaling, and (4) transcription factors and kinases. Graphs show the mean ratio (ECS:sham) of fold change in gene upregulation. Dotted horizontal line indicates a ratio of 1 (i.e., no regulation). Error bars represent SEM of four replicates, each from a different animal.

Figure 3.

Secondary confirmation of microarray data by in situ hybridization and RNA blot assay. Representative photomicrographs of hippocampal sections from ISH and RNA blot assays using radiolabeled riboprobes are shown, and quantified expression from the indicated cell layers is shown by bar graphs on the right. Results are expressed as a percentage of sham and are the mean ± SEM of four separate animals, each analyzed in duplicate brain sections. Lanes 1-4 in the RNA blots indicate individual samples, each from a separate animal, spotted in duplicate. The housekeeping gene cyclophilin was used to normalize the signal from sham and ECS groups. With the exception of NPY, all ISH images are from acute ECS-treated rats. Upregulation of the neuritin gene shown in A (top) is most evident in the DG granule cell layer after either acute or chronic treatments. A significant increase was also observed in the CA1 pyramidal cell layer after chronic ECS. B, Expression of VEGF in the choroid plexus and induction in the CA1 and CA3 pyramidal cell layers and DG granule cell layer. C, VGF was significantly upregulated in the DG with acute ECS and in the DG and CA3 pyramidal cell layers with chronic ECS. D, Regulation of BDNF was confirmed using the RNA blot assay only. E, FGF-2 shows prominent expression in the CA2 of both sham and ECS groups with maximal induction seen in the DG.

Figure 4.

Secondary confirmation of Cox-2, Egr-3, and NPY. Representative photomicrographs of hippocampal sections from ISH and RNA blot assays using radiolabeled riboprobes are shown. Quantified expression from the indicated cell layers is shown by bar graphs on the right. Results are expressed as a percentage of sham and are the mean ± SEM of four separate animals, each analyzed in duplicate brain sections. Lanes 1-4 in the RNA blots indicate individual samples, each from a separate animal spotted in duplicate. A housekeeping gene, cyclophilin, was used to normalize the signal from sham and ECS groups. A, Cox-2 exhibits low basal expression but is robustly induced in the DG, amygdala, and outer layer of the cerebral cortex (L1). B, Egr3 is induced exclusively in the DG. C, NPY, which shows a punctuate pattern of expression in sections from sham-treated animals, showed maximal regulation only with chronic treatment.

Figure 5.

Analysis of TIMP-1 expression and regulation. A, B, Representative photomicrographs of TIMP-1 after acute (ISH and RNA blot) or chronic ECS (ISH). The results are expressed as a percentage of sham and are the means ± SEM of four separate determinations, each representing a different animal. A, Robust induction of TIMP-1 after acute ECS is seen in the DG, outer layer of the cerebral cortex (L1), and a blood vessel (BV) just below the DG. Note that a significant expression of TIMP-1 in the choroid plexus is shown. B, After chronic ECS, there is further induction in the L1 region and the BV. Significant expression is seen in the molecular layer (stratum moleculare) of the DG (SM). C, An enlarged view of the hippocampus is shown. D, Emulsion autoradiography shows high grain density over cresyl violet-stained cells.

Figure 6.

Analysis of gene expression and regulation in the choroid plexus. A, A comparison of growth factor signaling genes expressed in three regions, choroid plexus, hippocampus, and cortex, is shown by a venn diagram, and the 10 most enriched genes in the choroid plexus are listed. The regional expression of midkine (Mdk), IGF2, IGFBP2, cyclophilin, TGFb1, and TGFb3 was measured by RNA blot assay. B, The bar graph represents signal intensity of corresponding spots. Inset shows the high level of Mdk and IGF2BP expression by ISH. Gene regulation in the choroid plexus in response to chronic ECS was examined from three sets of pooled samples (n = 6 animals for each set). C, D, Regulated genes were classified into the following two categories: growth factor and angiogenesis signaling (C), and other signaling pathways (D). The results are presented as fold change relative to sham. The dotted horizontal line indicates a ratio of 1 or no regulation. E, Secondary confirmation by RNA blot assay is shown for TIMP-1. Error bars represent SEM from three replicates.

Figure 7.

Immunohistochemical analysis demonstrating ECS regulation of corresponding protein levels for several genes. Immunohistochemistry was performed on sections from chronic ECS or sham animals using fresh frozen cryocut sections. A-C, E, Representative images are shown from chronic ECS (6 hr after ECS) for NPY (A), VEGF (B), neuritin (C), and TIMP-1 (E). D, Cox-2 immunohistochemistry was 2 hr after ECS. D, Maximal increases in Cox-2 protein were observed in the DG and amygdala and shown with higher power magnification in the inset. E, TIMP-1 expression in blood vessel (BV) and outer layer of cerebral cortex corresponds to regions of mRNA regulation. Upregulation of NPY, VEGF, and neuritin was most evident in the hilus and is shown with magnified insets. Scale bars: A-C, 25 μm; D, E, 50 μm.