Gene expression changes induced by mood stabilizer drugs in a serotonergic cell line

Abstract

Valproic acid (VPA) is widely used in the treatment and prophylaxis of bipolar disorder and epilepsy. Despite its extensive use, the mechanisms of action of this drug are still largely unclear. VPA is known to enhance GABA-mediated neurotransmission, and it modulates a variety of molecular signalling networks although no single mechanism has been implicated in its therapeutic action. Tetrahydrobiopterin (BH4) is an essential cofactor in the biosynthesis of neurotransmitters like serotonin, dopamine and noradrenalin, and the BH4 pathway may thus be important in mood biology. Prior work from this laboratory demonstrated that VPA caused a marked upregulation of Spr, the gene for sepiapterin reductase, which is a key enzyme in the BH4 synthesis pathway. This work therefore defined a potentially novel mechanism of action for VPA, operating through the BH4 pathway, which is further explored here in an in vitro cell culture setting. Using the rat serotonergic cell line RN46A, in which VPA induction of Spr was first observed, this effect was confirmed and further examined using real-time quantitative PCR (qPCR). About 8-fold induction of Spr was observed after 72 h exposure of RN46A cells to 0.5 mM VPA. The generality of this effect was then explored by treating three different rat neural cell lines (RN46A, C6 and H19-7) with various concentrations of VPA for 72 h and measuring Spr expression changes. With the aim of attaining greater understanding of this effect, a range of other drugs that are used as mood stabilizer drugs such as the anticonvulsants carbamazepine, lamotrigine and the classic mood stabilizer drug lithium were examined for Spr expression changes in RN46A cells. Of these, only lithium showed significant upregulation of Spr at therapeutically relevant doses. The action of both VPA and lithium on Spr expression suggested a potential mechanism of action shared by these two widely used mood stabilizer drugs. VPA has recently been recognized as an epigenetic drug, due to its ability to inhibit histone deacetylases and induce demethylation in specific genes. These effects offer new avenues for understanding the therapeutic actions of this drug. With the hypothesis that the HDAC inhibitory activity of VPA causes chromatin changes in the Spr promoter leading to upregulation of the gene, chromatin immunoprecipitation assays to analyse histone modifications at the Spr promoter were carried out. Histone acetylation at H3K9/K14ac histone marks near the Spr transcription start site in RN46A, after exposure to VPA, was found to be significantly increased. Correlation of increased histone acetylation with Spr upregulation suggests involvement of histone modifications in VPA action on this gene, thus defining one possible epigenetic mechanism of action for this mood stabilizer drug. The role of methylation changes at the Spr promoter in the upregulation of this gene in response to VPA exposure was also explored in this study. Bisulfite sequencing was carried out on DNA extracted from RN46A cells exposed to either VPA or 5′-aza-deoxycytidine, a classical DNA methylation inhibitor. Although no methylation changes in the Spr promoter in response to VPA were found, 5′-aza-deoxycytidine resulted in demethylation of some methylated CpG sites. Therefore, although the method used was sensitive to methylation changes, such changes did not appear to be involved in Spr induction after VPA exposure. An attempt was also made to extend the observations in the cell culture model into an in vivo model using samples available from rats previously treated with VPA. This study involved immunohistochemistry of rat brain sections for Spr as well as measurement of BH4 levels from plasma derived from the same animals. These experiments did not reveal any significant variation in either Spr-immunoreactivity in the dorsal raphe nucleus, or concentration of BH4 in plasma, after exposure to VPA. However, despite lack of differences in BH4 concentration, this pilot study showed significant differences in plasma biopterin levels (BH4 is the reduced form of biopterin) between the treated and untreated animals. Finally, RNA-Seq was used to examine the global profile of gene expression in RN46A cells in response to therapeutically relevant concentrations of VPA and lithium. When compared to untreated control cells, differential expression and alternative splicing of a number of genes was observed, particularly after exposure to VPA. A total of 136 genes were upregulated and 41 genes downregulated in response to VPA (fold change ≥1.5 fold, unadjusted p<0.05). In comparison, lithium exposure resulted in upregulation of 10 and downregulation of 11 genes using the same criteria. Functional annotation revealed significant enrichment of Gene Ontology terms such as neurogenesis and neural differentiation in response to VPA. Both these processes are implicated in the aetiology and treatment of depression and other mood disorders, suggesting this is a useful experimental approach to help understand the effect of the drug on these pathways. Unlike the qPCR analyses, RNA-Seq did not detect a significant increase in Spr transcripts after exposure to either VPA or lithium. This could be because Spr is expressed at relatively low levels in RN46A cells and therefore was probably not accurately represented in the RNA-Seq dataset. However, RNA-Seq detected a range of novel genes and transcription factors that were significantly modulated by VPA and lithium. These data, combined with the gene expression and epigenetic studies carried out in this thesis provide a basis for further study of the mood stabilizing effects of VPA as well as lithium.