Teratogenic effects of valproic acid and diphenylhydantoin on mouse embryos in culture

Spatiotemporal evaluation of the mouse embryonic redox environment and histiotrophic nutrition following treatment with valproic acid and 1,2-dithiole-3-thione during early organogenesis

Redox regulation during metazoan development ensures that coordinated metabolic reprogramming and developmental signaling are orchestrated with high fidelity in the hypoxic embryonic environment. Valproic acid (VPA), an anti-seizure medication, is known to increase markers of oxidation and also increase the risk of neural tube defects (NTDs) when taken during pregnancy. It is unknown, however, whether oxidation plays a direct role in failed neural tube closure (NTC). Spatial and temporal fluctuations in total glutathione (GSH) and total cysteine (Cys) redox steady states were seen during a 24 h period of CD-1 mouse organogenesis in untreated conceptuses and following exposure to VPA and the Nrf2 antioxidant pathway inducer, 1,2-dithiole-3-thione (D3T). Glutathione, glutathione disulfide (GSSG), and Cys, cystine (CySS) concentrations, measured in conceptal tissues (embryo/visceral yolk sac) and fluids (yolk sac fluid/amniotic fluid) showed that VPA did not cause extensive and prolonged oxidation during the period of NTC, but instead produced transient periods of oxidation, as assessed by GSH:GSSG redox potentials, which revealed oxidation in all four conceptal compartments at 4, 10, and 14 h, corresponding to the period of heartbeat activation and NTC. Other changes were tissue and time specific. VPA treatment also reduced total FITC-Ab clearance from the medium over 3 h, indicating potential disruption of nutritive amino acid supply. Overall, these results indicated that VPA's ability to affect cellular redox status may be limited to tissue-specific windows of sensitivity during the period of NTC. The safety evaluation of drugs used during pregnancy should consider time and tissue specific redox factors.

Retinal alterations in a pre-clinical model of an autism spectrum disorder

Background

Autism spectrum disorders (ASD) affect around 1.5% of people worldwide. Symptoms start around age 2, when children fail to maintain eye contact and to develop speech and other forms of communication. Disturbances in glutamatergic and GABAergic signaling that lead to synaptic changes and alter the balance between excitation and inhibition in the developing brain are consistently found in ASD. One of the hallmarks of these disorders is hypersensitivity to sensory stimuli; however, little is known about its underlying causes. Since the retina is the part of the CNS that converts light into a neuronal signal, we set out to study how it is affected in adolescent mice prenatally exposed to valproic acid (VPA), a useful tool to study ASD endophenotypes.

Methods

Pregnant female mice received VPA (600 mg/kg, ip) or saline at gestational day 11. Their male adolescent pups (P29–35) were behaviorally tested for anxiety and social interaction. Proteins known to be related with ASD were quantified and visualized in their retinas by immunoassays, and retinal function was assessed by full-field scotopic electroretinograms (ERGs).

Results

Early adolescent mice prenatally exposed to VPA displayed impaired social interest and increased anxiety-like behaviors consistent with an ASD phenotype. The expression of GABA, GAD, synapsin-1, and FMRP proteins were reduced in their retinas, while mGluR5 was increased. The a-wave amplitudes of VPA-exposed were smaller than those of CTR animals, whereas the b-wave and oscillatory potentials were normal.

Conclusions

This study establishes that adolescent male mice of the VPA-induced ASD model have alterations in retinal function and protein expression compatible with those found in several brain areas of other autism models. These results support the view that synaptic disturbances with excitatory/inhibitory imbalance early in life are associated with ASD and point to the retina as a window to understand their subjacent mechanisms.

Electronic supplementary material

The online version of this article (10.1186/s13229-019-0270-8) contains supplementary material, which is available to authorized users.

Metabolite profiling of whole murine embryos reveals metabolic perturbations associated with maternal valproate-induced neural tube closure defects

BACKGROUND

Valproic acid (VPA) is prescribed therapeutically for multiple conditions, including epilepsy. When taken during pregnancy, VPA is teratogenic, increasing the risk of several birth and developmental defects including neural tube defects (NTDs). The mechanism by which VPA causes NTDs remains controversial and how VPA interacts with folic acid (FA), a vitamin commonly recommended for the prevention of NTDs, remains uncertain. We sought to address both questions by applying untargeted metabolite profiling analysis to neural tube closure (NTC) stage mouse embryos.

METHODS

Pregnant SWV dams on either a 2 ppm or 10 ppm FA supplemented diet were injected with a single dose of VPA on gestational day E8.5. On day E9.5, the mouse embryos were collected and evaluated for NTC status. Liquid chromatography coupled to mass spectrometry metabolomics analysis was performed to compare metabolite profiles of NTD-affected VPA-exposed whole mouse embryos with profiles from embryos that underwent normal NTC from control dams.

RESULTS

NTDs were observed in all embryos from VPA-treated dams and penetrance was not diminished by dietary FA supplementation. The most profound metabolic perturbations were found in the 10ppm FA VPA-exposed mouse embryos, compared with the other three treatment groups. Affected metabolites included amino acids, nucleobases and related phosphorylated nucleotides, lipids, and carnitines.

CONCLUSION

Maternal VPA treatment markedly perturbed purine and pyrimidine metabolism in E9.5 embryos. In combination with a high FA diet, VPA treatment resulted in gross metabolic changes, likely caused by a multiplicity of mechanisms, including an apparent disruption of mitochondrial beta-oxidation. Birth Defects Research 109:106-119, 2017. © 2016 Wiley Periodicals, Inc.

Adverse effect of valproic acid on an in vitro gastrulation model entails activation of retinoic acid signaling

Valproic acid (VPA), an antiepileptic drug, is a teratogen that causes neural tube and axial skeletal defects, although the mechanisms are not fully understood. We previously established a gastrulation model using mouse P19C5 stem cell embryoid bodies (EBs), which exhibits axial patterning and elongation morphogenesis in vitro. Here, we investigated the effects of VPA on the EB axial morphogenesis to gain insights into its teratogenic mechanisms. Axial elongation and patterning of EBs were inhibited by VPA at therapeutic concentrations. VPA elevated expression levels of various developmental regulators, including Cdx1 and Hoxa1, known transcriptional targets of retinoic acid (RA) signaling. Co-treatment of EBs with VPA and BMS493, an RA receptor antagonist, partially rescued axial elongation as well as gene expression profiles. These results suggest that VPA requires active RA signaling to interfere with EB morphogenesis.

Neuronal developmental gene and miRNA signatures induced by histone deacetylase inhibitors in human embryonic stem cells

Human embryonic stem cells (hESCs) may be applied to develop human-relevant sensitive in vitro test systems for monitoring developmental toxicants. The aim of this study was to identify potential developmental toxicity mechanisms of the histone deacetylase inhibitors (HDAC) valproic acid (VPA), suberoylanilide hydroxamic acid (SAHA) and trichostatin A (TSA) relevant to the in vivo condition using a hESC model in combination with specific differentiation protocols and genome-wide gene expression and microRNA profiling. Analysis of the gene expression data showed that VPA repressed neural tube and dorsal forebrain (OTX2, ISL1, EMX2 and SOX10)-related transcripts. In addition, VPA upregulates axonogenesis and ventral forebrain-associated genes, such as SLIT1, SEMA3A, DLX2/4 and GAD2. HDACi-induced expression of miR-378 and knockdown of miR-378 increases the expression of OTX2 and EMX2, which supports our hypothesis that HDACi targets forebrain markers through miR-378. In conclusion, multilineage differentiation in vitro test system is very sensitive for monitoring molecular activities relevant to in vivo neuronal developmental toxicity. Moreover, miR-378 seems to repress the expression of the OTX2 and EMX2 and therefore could be a regulator of the development of neural tube and dorsal forebrain neurons.

Mouse models of neural tube defects: investigating preventive mechanisms

Neural tube defects (NTD), including anencephaly and spina bifida, are a group of severe congenital abnormalities in which the future brain and/or spinal cord fail to close. In mice, NTD may result from genetic mutations or knockouts, or from exposure to teratogenic agents, several of which are known risk factors in humans. Among the many mouse NTD models that have been identified to date, a number have been tested for possible primary prevention of NTD by exogenous agents, such as folic acid. In genetic NTD models such as Cart1, splotch, Cited2, and crooked tail, and NTD induced by teratogens including valproic acid and fumonisins, the incidence of defects is reduced by maternal folic acid supplementation. These folate-responsive models provide an opportunity to investigate the possible mechanisms underlying prevention of NTD by folic acid in humans. In another group of mouse models, that includes curly tail, axial defects, and the Ephrin-A5 knockout, NTD are not preventable by folic acid, reflecting the situation in humans in which a subset of NTD appear resistant to folic acid therapy. In this group of mutants alternative preventive agents, including inositol and methionine, have been shown to be effective. Overall, the data from mouse models suggests that a broad-based in utero therapy may offer scope for prevention of a greater proportion of NTD than is currently possible.

The teratology of autism

Autism spectrum disorders affect behaviors that emerge at ages when typically developing children become increasingly social and communicative, but many lines of evidence suggest that the underlying alterations in the brain occur long before the period when symptoms become obvious. Studies of the behavior of children in the first year of life demonstrate that symptoms are often detectable in the first 6 months. The environmental factors known to increase the risk of autism have critical periods of action during embryogenesis. Minor malformations that occur frequently in people with autism are known to arise in the same stages of development. Anomalies reported from histological studies of the brain are consistent with an early alteration of development. Congenital syndromes with high rates of autism include somatic that originate early in the first trimester. In addition, it is possible to duplicate a number of anatomic and behavioral features characteristic of human cases by exposing rat embryos to a teratogenic dose of valproic acid at the time of neural tube closure.

Molecular basis of environmentally induced birth defects

Exposure of the developing conceptus to selected environmental agents can lead to deleterious and often times lethal birth defects. These malformations result in serious emotional and financial consequences to families and societies worldwide. As we continue to progress technologically, we face challenges from the introduction of new pharmacological agents and chemical compounds into the environment. This results in a concomitant need to more fully understand the relationship between in utero exposure to environmental teratogens and the risk of congenital malformations. The goal of this review is to provide a current perspective of the major concepts related to the molecular basis of environmentally induced birth defects. Starting with a discussion of commonly occurring birth defects, we consider important fundamental facets of embryonic development, teratology, and gene-environment interactions. The review then summarizes our current understanding of the molecular mechanisms involved in selected birth defects following exposure to pharmacological compounds, including thalidomide, retinoids, and valproic acid. Understanding these signaling pathways may lead to the development of safer pharmaceutical compounds and a reduction in the number of infants born with preventable birth defects.

Valproate enhances N-cadherin production in Xenopus embryos

Xenopus embryos were exposed to valproate (0, 20, 40, 80 mg/l) either before or after neural tube closure. The embryos were then homogenized and fractionated by gel electrophoresis, and N-cadherin was detected and measured with quantitative immunoblotting. Findings indicated that valproate exposure increased N-cadherin production in a dose-dependent manner. Embryos exposed prior to neural tube closure tended to be more sensitive to the effects of valproate. These findings suggest that alterations in N-cadherin-mediated adhesion or morphogenesis may partially explain the teratogenic mechanism of valproate.

Valproic acid-induced alterations in growth and neurotrophic factor gene expression in murine embryos [corrected]

Although the teratogenicity of valproic acid (VPA) has been well established, the mechanism(s) by which this anticonvulsant drug induces malformations remains controversial. Using the combined molecular techniques of in situ-transcription (IST) and antisense RNA (aRNA) amplification we analyzed VPA-induced alterations in the gene expression for 10 genes within the neural tubes of embryos from two murine strains that have been shown to differ in their susceptibility to VPA-induce neural tube defects (NTD). Pregnant dams from both SWV (susceptible) and LM/Bc (resistant) strains were either treated with saline (control) or VPA (600 mg/kg) on gestational day (GD) 8:12 (day:hour). Neural tubes were isolated from control or VPA exposed embryos at three gestational time points, which represented the beginning (GD 8:18), middle (GD 9:00), and end (GD 9:12) of neural tube closure (NTC) in both of these murine strains. Using univariant statistics we demonstrated that in LM/Bc embryos with NTDs, the expression of bdnf, ngf, and trk, ngf-R were significantly elevated at all three time points, and the cytokine, cntf was significantly decreased at GD 9:00. In contrast, the major gene alterations observed in SWV embryos were a significant increase in tfgalpha and tgfbeta1-3 at GD 9:00. In an effort to better define the more intricate interactions between VPA exposure and the expression of these genes, we analyzed our data using Principal Component Analysis. The results from this analysis demonstrated that embryos from these two stains behaved differently, not only in response to a VPA exposure, but also under control conditions, which may explain the multifactorial nature of NTDs in these mice.

Blepharoptosis and central nervous system abnormalities in combined valproate and hydantoin embryopathy

PURPOSE

To report a case of intrauterine anticonvulsant exposure with subsequent ocular adnexal manifestations.

METHODS

Case report.

RESULTS

An 18-month-old child with known anticonvulsant embryopathy was referred for the management of bilateral congenital blepharoptosis. Physical examination confirmed ocular and nonocular external manifestations of valproate and hydantoin embryopathies. Cavum septum pellucidum, mild sulcation defects, and cerebellar atrophy were identified on neuroimaging.

CONCLUSIONS

To our knowledge, our patient represents the second reported case of anomalous septum pellucidum after intrauterine valproate exposure. Clinicians evaluating patients with craniofacial features associated with intrauterine valproate exposure should recognize that concomitant anomalies of the central nervous system, including the septum pellucidum, might exist.

DNA oxidation as a potential molecular mechanism mediating drug-induced birth defects: phenytoin and structurally related teratogens initiate the formation of 8-hydroxy-2'-deoxyguanosine in vitro and in vivo in murine maternal hepatic and embryonic tissues

A considerable number of teratogens, including the anticonvulsant drug phenytoin and structurally related drugs and environmental chemicals, may be bioactivated by peroxidases, such as prostaglandin H synthase (PHS) and lipoxygenases (LPOs), to a reactive free radical intermediate that initiates birth defects. However, the molecular targets of the reactive free radical intermediates mediating chemical teratogenesis, and hence the fundamental determinants of susceptibility, are poorly understood. In these studies, a teratogenic dose of phenytoin (65 mg/kg), when injected into pregnant CD-1 mice during organogenesis on gestational day 12, initiated the oxidation of DNA in maternal hepatic and embryonic nuclei, forming 8-hydroxy-2'-deoxyguanosine. Significant maternal and embryonic DNA oxidation occurred at 6 and 3 h, respectively, suggesting relative embryonic deficiencies in free radical-related cytoprotective enzymes, although the rates appeared similar. Maximal DNA oxidation in both maternal and embryonic tissues occurred at 6 h, presumably reflecting the balance of DNA oxidation and repair, the latter of which appeared similar in both tissues. Inhibition of phenytoin-initiated embryonic DNA oxidation by the free radical spin trapping agent alpha-phenyl-N-t-butylnitrone (41.5 mg/kg), and by acetylsalicylic acid (10 mg/kg), an inhibitor of the cyclooxygenase component of PHS, was consistent with the previously reported reduction by these inhibitors of phenytoin-initiated murine birth defects. In vitro studies using a horseradish peroxidase (0.5 mg/ml)-H2O2 (5.45 micrograms/ml) bioactivating system for drug-initiated oxidation of 2'-deoxyguanosine (3.74 mM), indicated that the potency of xenobiotic-initiated formation of 8-hydroxy-2'-deoxyguanosine for the structurally related drugs and metabolites phenytoin, 5-(p-hydroxyphenyl)-5-phenylhydantoin, trimethadione, dimethadione, l-mephenytoin, l-nirvanol, d-nirvanol (80 microM each), or thalidomide (64 microM), reflected their murine teratogenic potency. Given the relatively low activities of cytochromes P450, compared to PHS and LPOs, in human and rodent embryonic tissues, these data support the potential teratological importance of peroxidase-catalysed bioactivation of xenobiotics with structural similarities to phenytoin. These studies provide the first evidence that peroxidase-catalysed embryonic DNA oxidation may constitute a critical molecular mechanism mediating the teratogenicity of phenytoin and related drugs and environmental chemicals, and suggest the potential teratological importance of additional embryonic processes, such as DNA repair and tumor suppressor genes, as determinants of susceptibility.

Mouse Whole-embryo Culture in Serum Diluted with Waymouth Medium: A Study of Valproic Acid Teratogenicity

The culture of whole post-implantation rodent embryos has become an important tool in developmental biology and toxicology. Since the establishment of this system, rat serum has been the main culture medium used. In this study, we demonstrate that medium composed of 50% serum and 50% chemically defined medium (Waymouth 705/1) permits satisfactory development of mouse embryos over a 26-hour period. In addition, our data demonstrate that, with less than 50% serum, the frequency of malformations increases and growth and differentiation decrease with dilution in a dose-dependent way. Little information is available in the literature on the types of abnormalities induced by inadequate dilutions of serum. This study shows that a chemically defined medium, supplemented with amounts of serum that are below threshold levels, interferes mainly with the normal development of the head (neural tube, eyes and maxillary processes) and with the growth of the embryo. The second part of this work compares the teratogenicity of valproic acid (VPA) in this new medium to that in undiluted serum. We show that the effects of VPA are similar in terms of the type of alterations observed, the reduction of growth and the differentiation. However, the concentrations necessary to induce these effects in Waymouth/human serum/rat serum are half of those needed in human serum/rat serum.

In vitro neuroteratogenicity of valproic acid and 4-en-VPA

Mouse embryos displaying 8 to 9 pairs of somites were cultured during 26 h in presence of 0.75 mM of VPA, or of 1 mM of 4-en-VPA. These concentrations induced approximately 50% of dysmorphogenic embryos. Irregular suture of caudal neural tube, abnormal head shape, cranial neural tube defects, and deformed optic vesicles were the most common defects observed with both compounds. The main differences in the types of dysmorphogeneses detected between the two compounds concerned the suture of the caudal neural tube and the telencephalic region. Other macroscopic effects induced by the two compounds were similar. Several of the observed abnormalities can be correlated with defects reported after in vivo exposure. The major alteration of the histological structure of the neural tube concerned a specific area in the hindbrain : VPA and 4-en-VPA induced an abnormal and irregular budding of the neuroepithelium at this level. Immunohistology with an antibody specific for radial glial fibers (RC-2) as well as SEM analysis showed a moderate effect on glial development, mainly after exposure to VPA.

Effects of anticonvulsant drugs on fetal mouse palates cultured in vitro

Maxillary regions of day-12.5 ICR mouse fetuses were dissected and cultured in a chemically defined serumless medium, and the effects of anticonvulsant drugs on in vitro palatogenesis were studied. The explants were treated for 72 h in vitro with 50 to 200 micrograms/mL diphenylhydantoin (DPH), 200 to 800 micrograms/mL sodium phenobarbital (PB), 12.5 to 400 micrograms/mL sodium valproate (VPA), and 3 to 100 micrograms/mL diazepam (DAZ). During the culture, the secondary palatal shelves of control explants elevated, grew medially, and fused after 72-h culture in a manner similar to the palatogenetic process in vivo. The fusion of palatal shelves was inhibited dose-dependently by treatments with DPH, VPA, and DAZ. PB showed no significant inhibitory effects on palatal fusion at concentrations up to 800 micrograms/mL. The in vitro toxicity of the anticonvulsants tested appeared to correlate with the relative in vivo teratogenic potential of the drugs. The present study demonstrated that the in vitro organ culture system should be useful for screening teratogenic agents, especially those causing cleft palate, and for exploring the mechanisms of cleft palate formation.

The teratogenic effects of valproic acid in human chondrogenesis in vitro

The anticonvulsant drug valproic acid (VPA) is a known teratogen in humans. In general, anticonvulsants effect major systems in the embryo causing craniofacial, cardiovascular, neurological, urogenital, and major and minor skeletal defects. The limb defects associated with in utero VPA exposure include digital hypoplasia, ectrodactyly, radial ray aplasia, and proximal phocomelia. Human studies are limited to case reports and to retrospective and/or prospective studies. Although animal studies have demonstrated a teratogenic effect of VPA on skeletogenesis, these doses were well above the human therapeutic dose which makes extrapolation from these studies to humans difficult. The purpose of this research was to evaluate the potential deleterious effects of VPA on chondrogenesis, a process that occurs in human limb formation. To accomplish this goal, human chondrocytes were cultured in a three dimensional agarose gel and treated with VPA. The use of this model system was a novel approach to evaluate the teratogenic potential of VPA during chondrogenesis. The influence of VPA on human chondrocytes was monitored using histochemical, immunocytochemical, and morphological techniques. There was a decrease in mitotic activity and the extracellular matrix was modified. At human therapeutic doses, immunofluorescence revealed that type II collagen was reduced, while type I collagen increased. In addition, the alcian blue-staining matrices (i.e., sulfated proteoglycans) were reduced. Moreover, the Golgi apparatus had swelling in the trans-face cisternae suggesting that proteoglycan synthesis may be altered.

Methionine decreases the embryotoxicity of sodium valproate in the rat: in vivo and in vitro observations

Methionine provided in the drinking water of pregnant rats injected with sodium valproate reduced the frequency of resorptions but did not improve embryo growth. Rats drinking methionine supplemented water had approximately twice the level of serum-free methionine and consumed only one-half the volume of water of controls. Using whole rat embryo cultures, the simultaneous addition of methionine and sodium valproate to the medium provided no protection from neural tube defects, nor did the addition of methionine to a medium of serum obtained from rats previously dosed with sodium valproate. However, protection from the teratogenic effects of sodium valproate was afforded by methionine when the culture medium was sera from rats consuming methionine and was particularly striking when embryos for culture were taken from pregnant rats that had been consuming methionine. These observations along with those of others indicated the importance of dietary and culture media methionine levels in evaluating experimental and regulatory teratology studies and suggested the possibility that methionine may play an important role in human teratology where multifactorial causes have been implicated in problems such as neural tube closure defects.

Spina bifida aperta induced by valproic acid and by all-trans-retinoic acid in the mouse: distinct differences in morphology and periods of sensitivity

The antiepileptic drug valproic acid (VPA) has been implicated as a human teratogen causing spina bifida aperta. Recently, we developed a mouse model inducing spina bifida aperta with VPA. To elucidate the pathogenesis of VPA-induced spina bifida aperta we now investigated the anatomy and histology of this defect in the mouse. The morphology of spina bifida aperta induced by all-trans-retinoic acid (RA) was used for comparison. Various doses of VPA and RA were administered at different times to determine the periods of sensitivity for inducing spina bifida aperta with these drugs. Each administration regimen consisted of three doses applied at intervals of 6 hr. RA induced spina bifida aperta during an earlier developmental period (day 8 of gestation) than VPA (day 9 of gestation). The most effective regimens for induction of spina bifida aperta in mice were injections of 3 x 500 mg VPA-Na/kg body weight (b.w.) intraperitoneally on day 9 of gestation at 0, 6, and 12 hr; RA (12.5 mg/kg b.w.) was given orally on day 8 of gestation at 12 and 18 hr, day 9 at 0 hr. VPA did not induce spina bifida aperta on day 8 of gestation and RA did not induce this effect on day 9 of gestation. Histological studies of day 18 fetuses carrying spina bifida aperta were performed. The spina bifida aperta induced by VPA shows a disorganized and necrotic spinal cord. In the vertebral canal were observed cell debris, blood cells, capillaries, macrophages, and rests of meninges. These results indicate that the spinal cord is almost destroyed at the affected section. In contrast, the spina bifida aperta induced by RA demonstrates a spinal cord organized in the gray and white matter, the dorsal and ventral horn. But the neural canal does not exist, only a layer of ependymal cells lies on the surface of the spinal cord. Our results indicate that the morphology of spina bifida aperta induced by VPA differed distinctly from that induced by RA in the mouse fetus. Moreover VPA produced a spina bifida aperta with a specific morphology. Also the period of sensitivity for induction of this lesion differed and occurred earlier for RA than for VPA. VPA and RA may possibly induce spina bifida aperta via different mechanisms in the mouse.

Valproic acid induced abnormal development of the central nervous system of three species of amphibians: Implications for neural tube defects and alternative experimental systems

Embryos of Ambystoma mexicanum, Xenopus laevis, and Hyperolius viridiflavus taeniatus were exposed to various concentrations of valproic acid (VPA: 0.1, 1.5, 10 mM) from blastula stage (S) 9 on up to advanced gastrulation of control embryos (S 11 1/2-12). At 10 and 5 mM VPA early development was affected in all species tested. However, the most pronounced effects occurred in Ambystoma: the neural folds appeared delayed and showed a flattened and wavy shape; the neural tube was not formed and embryos successively died. In Xenopus and Hyperolius (10, 5 mM VPA) the beginning of gastrulation was delayed up to neurulation of control embryos. In Xenopus many of the embryos completed neurulation, whereas some embryos exposed to 10 mM VPA showed neural tube defects (NTDs) of different type and degree (open neural tube at different regions of the dorsum). In Hyperolius neural folds arose around the blastoporus and fused later on (earlier in embryos treated with 5 mM VPA), but the shape of these embryos was abnormal and the development was not continued (pronounced effect at 10 mM VPA). Comparing the three species, Xenopus proved to be the least sensitive species (at 5 mM VPA 14.2% NTDs of total malformations compared to 100% in the other species). The most sensitive species, Ambystoma, developed head-oedema at 1 mM VPA, whereas the anurans were not affected. Our results suggest a similar mechanism of VPA-induced NTDs in mammals and amphibians.

Studies on valproate-induced perturbations of neurulation in the explanted chick embryo

The effect of the anticonvulsant sodium valproate on in vitro neurulation of the chick embryo, explanted after a 25-h in ovo incubation period, is described. Sodium valproate, at concentrations of 0.5-1.5 mM did not appear to have any profound effect on embryo growth when assessed by light microscopy. However scanning electron microscopy revealed a dose-dependent increase in the incidence of open anterior and posterior neuropores after 20 h of in vitro development (Stage 11). Concentrations of sodium valproate which were greater than 1.5 mM markedly increased the number of gross malformations, which were manifested as a complete disruption of the neural tube along its entire length. Failure of neuropore closure could not be attributed to a drug-induced neurodevelopmental delay as these defects were still apparent following 27 h of in vitro culture, a time coincident with the onset of embryo torsion.

Morphological differences elicited by two weak acids, retinoic and valproic, in rat embryos grown in vitro

We compared in rat whole-embryo culture the morphological changes elicited by valproic acid (VPA) with those elicited by trans-retinoic acid (RA). Rat embryos explanted on day 9.5 of gestation were treated on day 10 with RA or VPA at concentrations producing equivalent reductions in embryonic protein. The concentrations selected for morphological assessment by scanning and transmission electron microscopy, 2.3 and 800 microM, respectively, for RA and VPA, produced approximately a 50% incidence of abnormally open anterior neuropores in initial range-finding experiments in the culture system. Protein and DNA analyses were also performed on corresponding groups of embryos at three different doses. With concurrent control groups used as reference standards, the two treatment groups were compared for differences in external and internal morphology, protein and DNA contents, and growth indices. While certain variables responded similarly in the two treatment groups, e.g., the growth variables, protein and DNA contents, each drug produced selective morphological effects. Whereas treatment with RA produced underdeveloped branchial arches, symmetrically cleft cranial defects resulting in openings in rhombencephalic and prosencephalic regions, and exteriorized neural tissue in the caudal neuropore region, VPA produced irregular clefts with wavy margins along the entire length of the neural tube, and an open caudal neuropore without eversion of the neuroepithelium, while producing no detectable effect on the branchial arches. The similar effects of these two drugs on protein and DNA contents suggest comparable degrees of overall toxicity; however, the dissimilar effects on neural tube and branchial arches, coupled with the large difference in concentration of the drug required to produce the effects, add to the evidence that their mechanisms for elicitation of abnormal development are qualitatively different.

Developmental toxicity of valproic acid

Valproic acid is a very effective anticonvulsant agent widely used in the management of various forms of epilepsy. Administration of the drug during pregnancy results in increased incidence of congenital abnormalities in both humans and experimental animals. In recent years, a significant number of research efforts have attempted to define the contributory role of valproic acid to the impairment of normal prenatal growth and development. The present report summarizes current knowledge that has emerged from clinical and research studies. The specific topics include: the placental transfer of valproic acid; the teratogenic potential; structure-teratogenicity and dose-response relationships; species and strain differences; biochemical changes evoked by the drug in the fetus.

Whole embryo culture: interpretation of abnormal development in vitro

An overview will be given of a number of problems that arise when we attempt an interpretation of data obtained with the whole embryo culture method (WEC). The following aspects will be considered: 1) Types of deviation from normal development in the whole-embryo culture system: (a) problem of "artifacts" (for example, due to preparation mistakes and suboptimal culture conditions), (b) types and frequency of abnormal development in controls, retardation as an abnormal outcome, and the difficulty of assessing its significance, (c) gross-structural defects and, specifically, attempts to answer the following questions: (i) What are the limitations for an assessment of abnormal development with the WEC? (ii) Is a specific abnormality pattern recognizable under specific pathologic conditions? (iii) Is the rate of induced abnormal development concentration-dependent? (2) Problems encountered when interpreting concentration-effect relationships (for example, statistical and pharmacokinetic aspects). (3) Necessity of verifying in vitro effects with in vivo studies.

Effects of valproic acid, some of its metabolites and analogues on prenatal development of rats in vitro and comparison with effects in vivo

Using a whole-embryo culture system valproic acid (VPA) and some of its metabolites (2-en-VPA, 4-en-VPA, 4,4'-dien-VPA) and analogues (ethyl-propyl-acetic acid, propyl-butyl-acetic acid, di-butyl-acetic acid, 2-methyl-2-ethyl-hexanoic acid, 1-methyl-1-cyclohexanoic acid) were tested for their potential to induce abnormal development. With regard to embryonic growth, development and abnormality rate, the tested compounds showed a wide range of "teratogenic potency" in vitro. In order to verify some of the in vitro results, in vivo experiments were performed. Pregnant rats were treated subcutaneously on day 10 of gestation with 2 x 330 mg VPA/kg, or 2 x 400 mg 2-en-VPA/kg, respectively. Evaluation of the embryos was performed on day 11.5 of gestation, corresponding to the in vitro experiments. VPA showed a high potential to induce abnormal development in vivo as well as in vitro, whereas 2-en-VPA was inactive under our experimental conditions. Problems connected with the evaluation of the predictive value of an in vitro test system for the detection of embryotoxic effects, such as "validation" and significance of pharmacokinetic data, are discussed.

Methionine and neural tube closure in cultured rat embryos: morphological and biochemical analyses

When headfold-stage rat embryos were cultured on cow serum, their neural tubes failed to close unless the serum was supplemented with methionine. Methionine deficiency did not appear to affect the ability of the neural epithelium to fuse as a type of fusion was observed between anterior and posterior regions of the open neural tube in methionine-deficient embryos. Although methionine deficiency reduced the cell density and mitotic indices of cranial mesenchyme and neural epithelial cells, this did not appear to be a factor in failure of the neural tube to close. For example, embryos cultured on diluted cow serum also had fewer mesenchymal cells yet could complete neural tube closure if provided with methionine. Examination of the tips of the neural folds suggested that microfilament contraction could be involved; in the absence of methionine the neural folds failed to turn in. This possibility was supported by the reductions in neurite extension of isolated neural tubes cultured without methionine and by the reductions in microfilament associated methylated amino acids contained in embryo neural tube proteins.

Marked accumulation of valproic acid in embryonic neuroepithelium of the mouse during early organogenesis

Valproic acid, an antiepileptic drug, causes neural tube defects in mice and man. 14C-labeled valproic acid (sodium-salt) was administered to pregnant mice on days 8 and 9 of gestation (period of high sensitivity in regard to formation of neural tube defects in this species). Two dose levels of valproic acid (1 and 400 mg/kg) were used; in each case the total radioactivity administered was the same: 400 microCi/kg or 14.7 MBq/kg. Autoradiography combined with computerized densitometry revealed that in low-dose animals most of the radioactivity was confined to maternal liver and kidney, while at high doses more activity was observed in soft tissues and fluids, including amniotic fluid. In the embryo, the neuroepithelium showed the highest concentration, irrespective of dose and survival interval (30 min, 3 h, and 6 h). Upon administration of the high dose, up to five times more radioactivity (approximately 2,000 times more valproic acid) was recovered in embryonic tissues than after the low dose. It is concluded that high doses of VPA saturate the capacities of metabolism, excretion, and protein binding in the maternal organism, resulting in a higher proportion of the dose reaching the embryo, allowing more of the drug to be accumulated by the target organ, the neuroepithelium.

Teratogenic effects on the neuroepithelium of the CD-1 mouse embryo exposed in utero to sodium valproate

A causal association has now been recognized between the use of the anticonvulsant drug sodium valproate during pregnancy and the increased incidence of spina bifida in the human population. The objective of this study was to investigate the teratogenic effects of sodium valproate on the cephalic 1) neuroepithelium, 2) extracellular matrix, and 3) embryonic protein content in the CD-1 mouse embryo. Nulliparous female CD-1 mice were dosed intraperitoneally on day 8 of gestation with 340 mg/kg of sodium valproate. On day 10 of gestation, females were killed by cervical dislocation, and all live embryos were assigned to one of the following groups and processed accordingly for: 1) head measurements, 2) scanning electron microscopy, 3) total protein determination, 4) two-dimensional polyacrylamide gel electrophoresis, 5) immunohistochemistry, and 6) light microscopy. Exposure to sodium valproate at the selected dosage resulted in a 30% incidence of neural tube defects in the cranial region of these embryos. Treated embryos showed a significant reduction in head size, indicating a drug-induced microcephaly. No major differences were seen in the total embryonic protein patterns between control and treated embryos. Immunoreactivity to laminin and fibronectin showed a similar distribution in control and treated embryos except in the vasculature pattern of the hindbrain neuroepithelium. The neuroepithelium of the treated embryos showed marked disorganization when it was examined histologically, particularly in the forebrain region. Cells were disoriented, and there was a noticeable loss of intercellular adhesion in the juxtaluminal region. Increased cellular blebbing was apparent at the ependymal surface, and large protrusions of cells were seen invading the neural tube lumen. The lumen was distorted in shape and frequently contained blood cells. Irregularities and gaps were observed in the underlying basal lamina. These results suggest that treatment with sodium valproate during a critical time in neurogenesis in the CD-1 mouse embryo alters the normal architecture of the neuroepithelium, with a loss of integrity at both the basal and apical surfaces. The alterations seen in the neuroepithelium at any of these sites in this animal model could help explain the increased incidence of spina bifida seen in children of epileptic mothers receiving sodium valproate.

Common biochemical pathway of dysmorphogenesis in murine embryos: use of the glucocorticoid pathway by phenytoin

Phenytoin (5,5-diphenylhydantoin), a common anticonvulsant drug, is known to produce anomalies in the craniofacial region of animals and humans. Furthermore, recent evidence suggests that phenytoin disrupts craniofacial and neural tube morphogenesis by inhibiting the arachidonic acid cascade, a pathogenesis already implicated for glucocorticoids and hyperglycemia in the palate. This study tested the hypothesis that phenytoin interferes with the arachidonic acid cascade via the same biochemical pathway demonstrated for glucocorticoids. The proposed pathway was tested at two levels. First, indomethacin, an inhibitor of the enzyme cyclooxygenase, was used in culture to block the correction of phenytoin-induced defects by arachidonic acid. Second, cortexolone, an anti-glucocorticoid that binds at the glucocorticoid receptor binding site, was tested for its ability to prevent phenytoin-induced teratogenicity. Eighty-four percent of the embryos cultured in phenytoin and 93% of those cultured in phenytoin plus arachidonic acid and indomethacin had neural tube and/or craniofacial deformities. In contrast, only 14% of the embryos cultured in phenytoin plus cortexolone were affected. Indomethacin itself produced anomalies in 83% of the exposed embryos. These data are consistent with the hypothesis that the teratogenic action of phenytoin in murine embryo cultures occurs via the glucocorticoid anti-inflammatory pathway. Thus, the glucocorticoid receptor appears to be responsible for mediating phenytoin-induced teratogenicity.

Acetaminophen toxicity to cultured rat embryos

We tested the effects of acetaminophen on cultured rat embryo development. When added directly to culture media at 300 microM, a concentration approximately twice the human therapeutic blood level, acetaminophen caused abnormalities in the cultured embryos. Sera from both rats and monkeys following gavage with acetaminophen were also toxic to cultured embryos. The sera toxicities were related to acetaminophen concentrations, and the toxicity could be removed by serum dialysis. With regard to the metabolism of acetaminophen, glutathione levels in the yolk sac decreased in a concentration related fashion with addition of the drug. Also, buthionine sulfoximine, an inhibitor of glutathione synthesis, appeared to enhance acetaminophen embryo toxicity, and N-acetylcysteine, a glutathione precursor, appeared to protect embryos from acetaminophen toxicity. These results suggested that acetaminophen embryo toxicity resulted from direct exposure of embryos to acetaminophen and not a maternal metabolite.

Phenytoin embryotoxicity: role of enzymatic bioactivation in a murine embryo culture model

A murine embryo culture model was developed to study the potential contribution of enzymatic bioactivation to the teratogenicity of phenytoin. To assess the relative embryonic and maternal contributions to bioactivation, embryos were cultured respectively alone or in the presence of an exogenous source of cytochromes P-450 (P-450), which are thought to bioactivate phenytoin to a teratogenic reactive intermediate. Embryological development from gestational day 9 to day 10 was assessed, and bioactivation was quantified by the irreversible binding of radiolabeled phenytoin to embryonic protein. Embryos cultured with phenytoin and an exogenous P-450 bioactivating system showed a significant decrease in the incidence of turning and closure of the anterior neuropore, yolk sac diameter, and protein content as well as growth retardation. In the absence of an exogenous P-450 system, phenytoin did not decrease the incidence of turning or anterior neuropore closure but did cause growth retardation and a lesser but significant reduction in yolk sac diameter and embryonic protein content. An exogenous P-450 system enhanced the bioactivation of phenytoin, although significant activity also was detectable in embryos cultured without an exogenous bioactivating system. These results suggest that the embryo itself can enzymatically bioactivate embryotoxically significant amounts of phenytoin, and that bioactivation and embryotoxicity can be further enhanced, qualitatively and quantitatively, by an exogenous P-450 system, implicating a possible maternal contribution to phenytoin teratogenicity.

Effects of sodium valproate on development and social behaviour in the Mongolian gerbil

Sodium valproate is an anticonvulsant widely prescribed because of its broad spectrum of activity. While acute toxicity from high doses is well recognized, there have been few animal studies of its chronic toxicity at therapeutic dose levels. Sodium valproate given continuously in drinking fluid (600 mg/l) throughout pregnancy and lactation to breeding gerbils caused developmental delay of the self-righting reflex in their pups. Dams ingested 97 mg/kg daily during gestation and 151 mg/kg on average during lactation, a dose in the lower range of anticonvulsant effectiveness. Reproductive performance, birth weights and subsequent growth of the pups remained normal, as did brain weights in adulthood. Drug-treated offspring, continuing to receive valproate as drinking fluid after weaning (600 mg/l; 82 to 111 mg/kg) showed negligible behavioural alteration at 6 weeks of age as assessed by ethological procedures, although behavioural change did occur at 20 weeks in the female animals. These females were characterised by significant enhancement of exploration and scanning during dyadic encounters in an unfamiliar cage, and showed a concomitant reduction of other nonsocial activities. Short-term administration of this dose of the drug did not affect behaviour. These results suggest an increased reactivity to the environment which becomes evident only after long-term treatment with valproate and to which female animals are more susceptible than males. These findings of developmental delay and of modifications to behaviour later in life points to the need for more detailed clinical assessments of the effects of valproate in human patients.

Modulation of phenytoin teratogenicity and embryonic covalent binding by acetylsalicylic acid, caffeic acid, and alpha-phenyl-N-t-butylnitrone: implications for bioactivation by prostaglandin synthetase

Teratogenicity of the anticonvulsant drug phenytoin is thought to involve its bioactivation by cytochromes P-450 to a reactive arene oxide intermediate. We hypothesized that phenytoin also may be bioactivated to a teratogenic free radical intermediate by another enzymatic system, prostaglandin synthetase. To evaluate the teratogenic contribution of this latter pathway, an irreversible inhibitor of prostaglandin synthetase, acetylsalicylic acid (ASA), 10 mg/kg intraperitoneally (ip), was administered to pregnant CD-1 mice at 9:00 AM on Gestational Days 12 and 13, 2 hr before phenytoin, 65 mg/kg ip. Other groups were pretreated 2 hr prior to phenytoin administration with either the antioxidant caffeic acid or the free radical spin trapping agent alpha-phenyl-N-t-butylnitrone (PBN). Caffeic acid and PBN were given ip in doses that respectively were up to 1.0 to 0.05 molar equivalents to the dose of phenytoin. Dams were killed on Day 19 and the fetuses were assessed for teratologic anomalies. A similar study evaluated the effect of ASA on the in vivo covalent binding of radiolabeled phenytoin administered on Day 12, in which case dams were killed 24 hr later on Day 13. ASA pretreatment produced a 50% reduction in the incidence of fetal cleft palates induced by phenytoin (p less than 0.05), without significantly altering the incidence of resorptions or mean fetal body weight. Pretreatment with either caffeic acid or PBN resulted in dose-related decreases in the incidence of fetal cleft palates produced by phenytoin, with maximal respective reductions of 71 and 82% at the highest doses of caffeic acid and PBN (p less than 0.05). Caffeic acid and PBN also significantly reduced the incidence of fetal resorptions produced by phenytoin, but not the fetal weight loss. In viable embryos, ASA pretreatment reduced the covalent binding of phenytoin to embryonic protein by 43% (p less than 0.05). Binding of phenytoin to embryonic resorptions was equally high with and without ASA pretreatment, and within each treatment group was 3- to 10-fold higher than that in the respective placentas and associated viable embryos (p less than 0.05). These results suggest that prostaglandin synthetase may contribute to the enzymatic bioactivation of phenytoin to a teratogenic free radical intermediate.

Postimplantation embryo culture for the assessment of the teratogenic potential and potency of compounds

Whole rat embryos cultured during the early stages of organogenesis were subjected to a panel of selected chemicals. Of seventeen known in vivo teratogens, seventeen also induced specific malformations in embryos grown in culture. Of ten chemicals which were reported to be negative in in vivo rat teratogenicity studies, eight also did not provoke dysmorphogenic effects in vitro. Of five additionally tested retinoids, all induced multiple malformations. However, concentrations used to induce these effects varied considerably, isotretinoin inducing malformations at 10(-5) M and arotinoid at 10(-11) M. The results indicate qualitatively as well as quantitatively a high predictability of this in vitro system and suggest that the postimplantation embryo culture system may also be useful in the prospective testing of new drugs and environmental chemicals.

Congenital abnormalities in two sibs exposed to valproic acid in utero

Neural tube, craniofacial, and other congenital abnormalities have been described in infants born to mothers with epilepsy who were treated with valproic acid (VPA) during pregnancy. The pathogenetic relationship between the congenital abnormalities and exposure to VPA is not clear. We describe 3 sibs born to a mother with epilepsy. Only the 2 sibs who were exposed to VPA in utero had certain of the characteristic craniofacial changes described in previous reports of children with similar exposure. In addition, the 2 affected children had other craniofacial and skeletal abnormalities which have not been reported previously as teratogenic sequelae of VPA exposure. Many of the minor anomalies present in the 2 affected patients resemble those observed in rhesus monkeys exposed to VPA in utero. This suggests that the abnormalities observed in the 2 children may have been a consequence of intrauterine exposure to VPA.

The anticonvulsant sodium valproate specifically induces the expression of a rat glial heat shock protein which is identified as the collagen type IV receptor

A potential mechanism for valproate (VPA)-induced increases in glial cell-substratum adhesivity has been demonstrated. Metabolically labelled glioma (C6) and primary astrocytes showed a statistically significant accumulation of protein when cultured in the presence of therapeutic concentrations of VPA (1 mM). This was mainly accounted for by a 10-fold increase in the production of a single polypeptide of 43 kDa molecular weight. Fractionation studies and metabolic labelling with N-acetyl-D-mannosamine showed this to be a sialoglycoprotein which was plasma membrane-bound. VPA-induction of the polypeptide was apparently specific to glioma and primary astrocytes and was not observed in neuroblastoma (neuro-2a), fibroblasts (3T3), pituicytes (GH3) and epithelial cells (NCTC). The 43 kDa component of glia was demonstrated to be the receptor for type IV collagen by binding metabolically labelled and solubilised cells to Sepharose beads which had been individually coated with laminin, fibronectin and type IV collagen. The protein has also been shown to be a heat shock product as metabolically labelled glioma showed a 10-fold increase in its expression when cultured at 42 degrees C. This heat shock induced expression was transient and was in marked contrast to that seen with VPA where it increased with time and was sustained. The expression of 43 kDa is suggested to arise by VPA and heat shock induced delays in cell cycle progression and this is discussed in relation to teratogenic action.

Strain differences in the teratogenicity induced by sodium valproate in cultured mouse embryos

Strain differences in the teratogenicity of valproic acid (VPA) have been reported in mice. Finnell and Chernoff (Proc. Grnwd. Genet. Ctr. 5:162-163, 1985) showed that 300 mg/kg of VPA twice a day on days 6-8 of gestation induced exencephaly in 82% of SWV embryos but in 0% of C57BL/6J embryos. In the present experiment, we have collected similar results and investigated this strain difference using whole embryo culture in an attempt to determine whether maternal or embryonic factors are responsible for the difference. Mouse embryos were explanted on day 8.5 (plug day 0), and embryos at the 6-8-somite stage were cultured for 48 hours in rat serum containing various doses of sodium valproate (NaVP). All the embryos died within 24 hours with 4.5-mM and higher doses of NaVP in C57BL/6NCr1BR (C57) and with 3.0-mM and higher doses in SWV. Unfused brain folds were recognized in embryos treated with 3.0-mM and higher doses in C57, and with 1.0-mM and higher doses in SWV. Irregular somite formation was observed in many embryos treated with 1.6-mM and higher doses in C57 and with 1.0-mM and higher doses in SWV. These results indicate that SWV embryos have 1.5-3 times the sensitivity of C57 embryos to the embryolethal and teratogenic effects of NaVP. Furthermore, the results suggest that the basis of the strain difference resides within the embryo rather than the mother.

Post-implantation embryo culture: validation with selected compounds for teratogenicity testing

1. Some chemical compounds selected by experts for the validation of in vitro teratogenicity testing were investigated in whole rat embryos cultured during the early stages of organogenesis. All sixteen known in vivo teratogens tested also induced specific malformations in embryos grown in culture. 2. Of the nine compounds which were negative in in vivo rat teratogenicity studies, none provoked dysmorphogenic effects in cultured embryos. Abnormal development of the embryos was only observed with these compounds at concentrations also high enough to affect significantly overall growth and/or differentiation. 3. The results showed a high predictability of this system for the compounds tested and suggest that the post-implantation embryo culture system may also be useful in the prospective testing of new drugs and environmental chemicals.

Glutathione status and the incidence of neural tube defects elicited by direct acting teratogens in vitro

Valproic acid (VPA), cytochalasin D (CD) and 7-hydroxy-2-acetylaminofluorene (7-OH-AAF) each caused abnormal closure of the anterior neuropore in rat embryos cultured in vitro in the absence of an exogenous bioactivation system. Morphological comparisons showed that although all three compounds prevented normal neural tube closure, each did so in a distinctive manner. Modulation of GSH in cultured rat conceptuses was evaluated to determine whether common responses occurred relative to the ability of different chemicals to elicit neural tube defects. Malformation incidence in embryos (10-14 somites) varied widely following exposure to CD (44%), 7-OH-AAF (29%) or VPA (17%). The incidence of CD-elicited malformations was increased by 50% following GSH depletion by L-buthionine-S, R-sulfoximine (BSO) and was decreased by nearly 60% when the cysteine pro-drug 2-oxothiazolidine-4-carboxylate (OTC) was added to the culture medium. GSH modulation also exerted significant effects on the incidence of abnormal neurulation caused by VPA or 7-OH-AAF. A relatively low incidence of open neural tubes produced by VPA or 7-OH-AAF alone was shown to be a function of the state of maturation in the embryos. Conceptuses cultured in the presence of VPA or 7-OH-AAF from an earlier gestational age (6-10 somites) showed 2-3 fold increases in the number of embryos with open neural tubes. Differential alterations in protein and DNA content were observed among embryos and yolk sacs after various treatments indicating possible differences in the site of embryotoxicity. These results demonstrate the role of GSH status on the capacity of three chemically diverse compounds to elicit abnormal neurulation in cultured rat embryos and suggest some possible mechanisms by which normal neurulation may be compromised.