Ascorbic acid reverses valproic acid-induced inhibition of hoxa2 and maintains glutathione homeostasis in mouse embryos in culture

Abstract

Valproic acid (VPA) has been shown to cause neural tube defects in humans and mice, but its mechanism of action has not been elucidated. We hypothesize that alterations in embryonic antioxidant status and Hoxa2 gene expression play an important role in VPA-induced teratogenesis. A whole embryo culture system was applied to explore the effects of VPA on total glutathione, on glutathione in its oxidized (GSSG) and reduced (GSH) forms [GSSG/GSH ratio] and on Hoxa2 expression in cultured CD-1 mouse embryos during their critical period of organogenesis. Our results show that VPA can (1) induce embryo malformations including neural tube defects, abnormal flexion, yolk sac circulation defects, somite defects, and craniofacial deformities such as fusion of the first and second arches, and (2) alter glutathione homeostasis of embryos through an increase in embryonic GSSG/GSH ratio and a decrease in total GSH content in embryos. Western blot analysis and quantitative real-time RT-PCR show that VPA can inhibit Hoxa2 expression in cultured embryos at both the protein and mRNA level, respectively. The presence of ascorbic acid in the culture media was effective in protecting embryos against oxidative stress induced by VPA and prevented VPA-induced inhibition of Hoxa2 gene expression. Hoxa2 null mutant embryos do not exhibit altered glutathione homeostasis, indicating that inhibition of Hoxa2 is downstream of VPA-induced oxidative stress. These results are first to suggest VPA may, in part, exert its teratogenicity through alteration of the embryonic antioxidant status and inhibition of Hoxa2 gene expression and that ascorbic acid can protect embryos from VPA-induced oxidative stress.

Fig. 1

Embryos treated with VPA. Control (a), 50 μg/ml (b), and 200 μg/ml (c, d). The types of malformations observed included open neural tube (b, c), somite defects (c), and craniofacial malformations such as fusion of first and second arch (c). d The same embryo as c, which is magnified to show the fusion of the first and second arches (arrow). e Western blot analysis of Hoxa2 in embryonic samples treated with VPA. Lane 1 control samples. Lane 2 samples treated with 50 μg/ml [0.35 mM] valproic acid; Lane 3 samples treated with 100 μg/ml [0.7 mM] valproic acid; Lane 4 samples treated with 200 μg/ml [1.4 mM] valproic acid; Lane 5 samples treated with 400 μg/ml [2.8 mM] valproic acid

Fig. 2

Effect of VPA on total glutathione content (a) and on GSSG/GSH ratio (b) in mouse embryos in vitro (n = 8). GSSG glutathione disulfide, GSH reduced glutathione. * Represents a significance P ≤ 0.05 and ** represents a significance P ≤ 0.01 between bars indicated by brackets

Fig. 3

Effect of ascorbic acid on total glutathione content (a) and on GSSG/GSH ratio (b) in cultured mouse embryos treated with VPA. Control; AA, treated with 5 mM l-ascorbic acid; VPA, treated with 400 μg/ml [2.8 mM] of valproic acid; VPA + AA, pretreated with 5 mM l-ascorbic acid prior to addition of 400 μg/ml [2.8 mM] of valproic acid. Embryos were cultured for 48 h. Bars represents mean ± SD, n = 4–5. * represents a significance P ≤ 0.05 and ** represents a significance P ≤ 0.01 between bars indicated by brackets. GSSG glutathione disulfide, GSH reduced glutathione

Fig. 4

a Relative quantitative expression of Hoxa2 in cultured mouse embryos as determined by quantitative real-time RT-PCR using TaqMan primers and probes. b Western blot analysis of Hoxa2 protein in cultured embryos. Control; AA, treated with 5 mM l-ascorbic acid; VPA, treated with 400 μg/ml [2.8 mM] of valproic acid; VPA + AA, pretreated with 5 mM l-ascorbic acid prior to addition of 400 μg/ml [2.8 mM] of valproic acid. Embryos were cultured for 48 h. Bars represents mean ± SD, n = 3–4. ** Represents a significance P ≤ 0.01 between bars indicated by brackets

Fig. 5

Schematic diagram showing the relationship between VPA, glutathione, and Hoxa2. a VPA-induced oxidative stress leads to a decline in glutathione levels which in turn induces a downregulation of Hoxa2 gene expression. VPA’s inhibition of glutathione and glutathione’s effect on Hoxa2 may be direct or via several intermediary factors. The arrow with a cross indicates that the response is in one direction only since Hoxa2−/− mice do not display altered glutathione levels. The broken arrow indicates VPA could directly affect Hoxa2. A combination of reduced glutathione levels and downregulation of Hoxa2 may trigger a teratogenic response. b Pretreatment with 5 mM l-ascorbic acid prior to VPA exposure prevents VPA-induced decline in glutathione and Hoxa2 gene expression, protecting embryos from a teratogenic insult