Mechanism of valproic acid induced dysmorphogenesis via oxidative stress and epigenetic regulation at the Hoxa2 gene promoter
Valproic acid (2-propylpentanoic acid, VPA) is a clinically used anti-epileptic drug and an effective mood stabilizer. VPA is also a histone deacetylase inhibitor and can induce embryonic malformations in both humans and mice. The mechanism(s) of VPA-induced teratogenicity are not well characterized. The objectives of my study were three fold, to: (i) investigate the effect of VPA on mouse embryonic development, (ii) characterize the putative mechanism(s) of VPA-induced teratogenicity and, (iii) investigate VPA associated epigenetic regulation of Hoxa2 gene in cell lines and in developing embryos. Whole mouse embryo cultures were treated with VPA at doses of 0, 50 (0.35 mM), 100 (0.70 mM), 200 (1.4 mM), and 400 µg/mL (2.8 mM), encompassing the therapeutic range of 0.35 mM to 0.70 mM. Van Maele-Fabry’s morphologic scoring system was used to quantitatively assess embryonic organ differentiation and development. Hoxa2 gene expression was measured by quantitative real-time RT-PCR (Reverse Transcriptase-Polymerase Chain Reaction). To assess epigenetic changes on the Hoxa2 gene promoter, DNA methylation was determined by bisulfite (BSP) sequencing and pyrosequencing. Histone “bivalent domains” H3K4me3 (histone 3 lysine 4 trimethylation) and H3K27me3 (histone 3 lysine 27 trimethylation) associated with gene activation repression, respectively, analyzed qChIP-PCR (quantitative chromatin immunoprecipitation-PCR). Telomere length and telomerase activity were analyzed in mouse embryos and in NIH3T3 cell line treated with VPA. Results indicate significantly increased incidence of dysmorphogenesis in embryos (11.8%, 35.3%, 47.0% and 88.3%) exposed to increasing doses of VPA (0.35 mM, 0.70 mM, 1.4 mM and 2.8 mM respectively). Van Maele-Fabry’s quantitative differentiation assessment of developing embryos demonstrated a significantly lower score for the circulation system, central nervous system, craniofacial development and limb development in VPA treated embryos (0.35 mM to 2.8 mM) compared to the untreated control group. Glutathione homeostasis was altered as indicated by decreased total glutathione content and increased GSSG/GSH ratio in all VPA treatment groups. In addition, a dose-dependent inhibition of Hoxa2 gene expression was observed in embryos and in the NIH3T3 cell line exposed to VPA. Pre-treatment with ascorbic acid [1000 µg/mL (5 mM)] restored glutathione level and normalized Hoxa2 gene expression in embryos exposed to VPA. DNA methylation status was characterized on the Hoxa2 gene promoter at the three CpG islands; CpG island 1 (-277 to -620 bp), CpG island 2 (-919 to -1133 bp), and CpG island 3 (-1176 to -1301 bp) in the two cells lines (NIH3T3 and EG7) and in developing embryos. CpG sites remained unmethylated on the Hoxa2 gene promoter in the NIH3T3 cell line which expresses the Hoxa2 gene, whereas these same CpG sites were methylated in EG7 cells that did not express Hoxa2. CpG island 1 is closest to Hoxa2 transcription start site and its methylation status was most affected. In developing embryos, CpG island 1 was found to be highly methylated at E6.5 when Hoxa2 is not expressed, whereas the methylation status of CpG sites on the CpG island 1 declined between E8.5 and E10.5 when Hoxa2 expression is present. VPA induced methylation of several CpG sites on CpG island 1 in NIH3T3 cell line and in E10.5 embryos when Hoxa2 expression was down regulated following VPA exposure. In addition, embryos and the NIH3T3 cell line treated with VPA impacted the “bivalent domains” resulting in increased H3K27me3 enrichment and decreased H3K4me3 enrichment on Hoxa2 promoter. Pre-treatment with ascorbic acid normalized Hoxa2 expression and histone bivalent domain changes and prevented increased DNA methylation following VPA exposure. Moreover, the telomerase activity and telomere length were both impacted by changes in glutathione redox potential induced by VPA. Oxidative stress following VPA treatment reduced telomerase activity and accelerated telomere shortening. These results are the first to demonstrate: (i) a correlation between VPA dose and total morphologic score in the developing mouse embryos. VPA impacted embryonic tissue differentiation and neural system development in the dose range of 0.35 mM to 2.8 mM; (ii) VPA altered glutathione homeostasis in cultured mouse embryos and inhibited Hoxa2 gene expression; (iii) Histone bivalent domains of H3K27 and H3K4 trimethylation and DNA methylation status at the Hoxa2 gene promoter region were altered following treatment with VPA. This appears to be the epigenetic event in transcriptional silencing of Hoxa2 gene expression after VPA exposure; and (iv) Ascorbic acid normalizes glutathione homeostasis, H3K27 and H3K4 trimethylation and DNA methylation status, restoring Hoxa2 gene expression following VPA exposure. Taken together our results show VPA- induced altered glutathione homeostasis, telomere shortening and telomerase dysfunction, and an inhibition of Hoxa2 gene expression leads to developmental abnormalities. Exposure to ascorbic acid had a protective effect on developing embryos exposed to VPA.
Valproic acid, dysmorphogenesis, oxidative stress, epigenetic regulation, Hoxa2 gene
Doctor of Philosophy (Ph.D.)
Pharmacy and Nutrition