Human Molecular Genetics, 2000, Vol. 9, No. 2 227-236
© 2000 Oxford University Press
A mouse model for valproate teratogenicity: parental effects, homeotic transformations, and altered HOX expression
DIBIT and 1Department of Neuroscience, San Raffaele Scientific Institute, Milan, Italy, 2Institut für Säugetiergenetik, GSF-Forschungszentrum für Umwelt und Gesundheit, D-85758, Neuherberg, Germany, 3Department of Genetics, Case Western Reserve University School of Medicine and Center for Human Genetics, University Hospitals of Cleveland, Cleveland, OH 44106, USA
Received 31 August 1999; Revised and Accepted 12 November, 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Valproate (VPA) is one of several effective anti-epileptic and mood-stabilizing drugs, many of which are also potent teratogens in humans and several other mammalian species. Variable teratogenicity among inbred strains of laboratory mice suggests that genetic factors influence susceptibility. While studying the genetic basis for VPA teratogenicity in mice, we discovered that parental factors influence fetal susceptibility to induced malformations. Detailed examination of these malformations revealed that many were homeotic transformations. To test whether VPA, like retinoic acid (RA), alters HOX expression, pluripotent human embryonal carcinoma cells were treated with VPA or RA and Hox expression assessed. Altered expression of specific Hox genes may thus account for the homeotic transformations and other malformations found in VPA-treated fetuses.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Valproic acid (VPA) is a widely used and effective treatment for seizures and bipolar disorder (1–3). As many as 10% of the 12 000 infants that are exposed to anti-epileptic drugs during pregnancy every year show malformations (4–6). Use of VPA during the first trimester of pregnancy significantly increases the risk for spina bifida as well as other malformations such as heart defects, limb abnormalities, cleft palate and craniofacial abnormalities (4–10). Together these abnormalities constitute the fetal valproate syndrome (11). Many other structurally unrelated anti-epileptic and anti-manic drugs, such as lithium, barbiturates and carbamazepine, are also teratogenic when used during pregnancy (4,12,13). Unfortunately, the mechanism of VPA teratogenicity is not known. Insight into its molecular basis might lead to identification of the genetic factors controlling susceptibility to teratogenicity, understanding of its mode of neurological action, and design of alternative drugs that minimize adverse side-effects.
Laboratory mice have been used to study the basis for VPA teratogenicity. Administration of VPA on days 8–9 of gestation results in failure of cranial neural tube closure and spina bifida, as well as limb abnormalities such as syndactyly and oligodactyly (14–20). Among the physiological factors that have been implicated are folate and zinc metabolism, embryonic pH and lipid metabolism (21–24). The variable susceptibility to malformations among inbred mouse strains suggests that genetic factors influence VPA teratogenicity (4,19,25). It is unclear whether parental or embryonic factors influence these differences. Studies based on whole-embryo in vitro culture suggest that embryonic factors are important (18,19,26), but do not address the possible contribution of parental factors.
We therefore examined VPA teratogenicity among fetuses resulting from reciprocal crosses between a susceptible and a resistant strain of mice. These crosses revealed that parental factors strongly influence susceptibility to VPA-induced malformations. While examining these malformations, we discovered that many resulted from homeotic transformations of the vertebral column. These transformations are similar to those found in mice treated with retinoic acid (RA) (27,28) and in mice with certain engineered mutations (29–42). We therefore compared HOX expression patterns in human embryonal cells that were treated with VPA or RA. The ability of VPA to alter HOX expression patterns, as well as to induce homeotic transformations, suggests a molecular basis for VPA teratogenicity.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Embryo lethality, weight reductions and embryonic malformations
Two VPA doses were used: 200 and 400 mg VPA/kg body wt. The higher dose was used to assess embryo lethality, weight reductions and embryonic malformations. In general, malformations were more severe at the higher dose. The lower dose produced more subtle malformations and was used to assess homeotic transformations, which had escaped detection previously.
With the high dose of 400 mg VPA/kg body wt, embryotoxic effects were observed in all strains that were tested (Table 1). Embryo lethality ranged from 34% in C57BL/6J to 100% in A/J, and fetal weight reductions ranged from 62% for C3H/HeJ to 89% for BALB/cByJ. This variability among inbred strains suggests genetic influences on embryotoxicity.
|
Abnormalities of the axial skeleton were observed, including fusion of ribs, neural arches and vertebral bodies as well as spina bifida occulta (Table 2; Fig. 1). In many cases, the vertebral column was completely disrupted and individual vertebrae were not identifiable. Cranial abnormalities included cleft palate and delay in ossification of cranial bones (data not shown).
|
|
Abnormalities of the appendicular skeleton usually occurred after treatment with the highest tested dose of 400 mg VPA/kg body wt (Table 3). The most frequent abnormality was post-axial digit loss (one to three digits), sometimes in combination with absence of the ulna with a relatively intact radius. Less frequent malformations were syndactylies and dysplasia of the fourth phalange including the fourth metacarpal bone (Fig. 2). Control mice did not show any of the malformations that were found after VPA treatment.
|
|
Several traits were significantly correlated with each other, suggesting related developmental or physiological origins of the malformations. In treated DBA/2J fetuses, the number of fused neural arches was strongly correlated with both the severity of fused arches (r = 0.77, P < 0.01) and the number of missing post-axial digits on the left paw (r = 0.76, P < 0.01). In addition, the weight of DBA/2J fetuses was inversely correlated with occurrence of spina bifida (r = –0.79, P < 0.01), suggesting that small fetuses were more likely to show neural tube defects. In treated P/J fetuses, the occurrence and severity of fused ribs were strongly correlated (r = 0.87, P < 0.01) and the number of missing post-axial digits on the left paw was strongly correlated with ulnar defects (r = 0.84, P < 0.01), whereas the severity of fused ribs was inversely correlated with ulnar defects (r = –0.71, P < 0.01). The absence of a correlation among all other malformations was striking given the considerable expectation that they might be developmentally related, e.g. the number of missing post-axial digits on the left and right paws.
Susceptibility to VPA-induced malformations varied sub- stantially among inbred strains. For example, BALB/cByJ and C3H/HeJ showed the highest number and greatest severity of fused ribs, AKR/J, DBA/2J and P/J showed intermediate rates, and C57BL/6J showed the lowest rate (Table 2). In general, strains such as P/J were highly susceptible to induced malformations, whereas DBA/2J was relatively resistant to malformations (Tables 2 and 3).
Parental effects on VPA teratogenicity
To test for parental effects, reciprocal F1 hybrid fetuses were made by crossing the susceptible P/J strain to the resistant DBA/2J strain. Although P/J mice showed greater overall susceptibility to VPA teratogenicity, DBA/2J mice were more susceptible to malformations such as fused ribs. Pregnant control and experimental females were treated with saline or VPA, respectively, according to the same treatment regimen. The frequency of many malformations differed significantly among fetuses resulting from reciprocal crosses (Tables 2 and 3). An example is the lower frequency of fused ribs in P/J and (P x D)F1 hybrid fetuses than in DBA/2J and (D x P)F1 hybrids (Table 2). Other examples are the higher frequencies of fused neural arches, vertebral bodies and split vertebrae in P/J and (P x D)F1 hybrids than in DBA/2J and (D x P)F1 hybrids (Table 2). These results suggest that maternal genotype is more strongly associated than paternal genotype with VPA-induced teratogenicity. In contrast, neither fetal weight nor embryo lethality differed significantly among fetuses from reciprocal crosses (Table 1).
Homeotic transformations
A detailed morphological analysis of VPA-treated fetuses revealed the presence of homeotic transformations. In general, these transformations were found in the lower thoracic and lumbar regions but not in the cervical or upper thoracic segments. To study this unexpected observation systematically, five region-specific skeletal features were used to identify individual vertebrae and evaluate homeotic transformations in control and treated DBA/2J fetuses. The analysis revealed little or no variation among control fetuses, and a clearly increased frequency of anterior homeotic transformations after VPA treatment (Fig. 3).
|
These transformations, from anterior to posterior, were as follows. (i) The transition from an angular to a round neural arch defines the thoracic vertebra 9 (Th9)/Th10 boundary (Fig. 1e and f). Most control mice presented Th9 with an angular neural arch and Th10 vertebra with a round arch (Fig. 1e and e'). In treated mice, this transition was shifted one segment posterior in a greater fraction of treated fetuses than control fetuses (Fig. 1f and f'). (ii) The next anatomical feature that was evaluated was the fusion of transversal and superior articular processes (Fig. 1g and h'). In the upper thoracic region of untreated control mice, the transversal process and the superior articular process were clearly separated. In more caudal thoracic segments, the distance between these two processes was reduced and more caudally they were fused. In control fetuses, fusion of these two processes usually occurred at the level of Th11 (Fig. 1g and g'), whereas after treatment many more fetuses showed this fusion shifted one segment posterior (Fig. 1h and h'). (iii) The third feature was the number of ribs, independent of their length. All control fetuses had 13 thoracic rib-bearing segments (Fig. 1a), whereas treated embryos had an increased frequency of additional ribs (Fig. 1b and k). (iv) The next feature analyzed was the location of the most caudal anapophysis (Fig. 1i and k'). Anterior, but not posterior, lumbar (L) segments have a small lateral process, called an anapophysis. L2 is the last segment that bears this process in control fetuses (Fig. 1i and i'). Again, in many treated mice, an anapophysis was found one segment posterior (Fig. 1h and h'). (v) The final feature analyzed was the number of lumbar vertebrae, identified by the existence of a costarious process. Control fetuses had five lumbar vertebrae (Fig. 1i), whereas treated fetuses showed a posterior shift of this process (Fig. 1i'). In contrast to low VPA doses, intraperitoneal injections of high doses regularly led to severe skeletal defects, such as spina bifida, fusions of ribs and vertebrae, and delay in ossification. Perhaps massive sclerotomal cell death masked subtle patterning abnormalities at high VPA doses that are evident at lower doses.
Penetrance and localization of VPA-induced transformations varied among strains (Fig. 3). BALB/cByJ and C57BL/6J showed the lowest transformation rate and other strains showed higher rates, including 90–100% penetrance in A/J, AKR/J and C3H/HeJ. With respect to localization, most transformations in BALB/cByJ occurred in the lower thoracic region, in A/J, AKR/J and C3H/HeJ in the upper lumbar region, and in C57BL/6J, DBA/2J and P/J in the lower lumbar region. SWR/J showed uniform transformation rates along the vertebral axis. These strain differences imply genetic predisposition to particular transformations.
VPA alters HOX expression in human embryonal carcinoma cells
RA both induces homeotic transformations and alters Hox expression (43,44). The homeotic transformations found in treated fetuses raise the possibility that VPA also alters HOX expression patterns. To test this hypothesis, we treated NT2/D1 embryonal carcinoma (EC) cells with RA or VPA and assessed HOX expression patterns. NT2/D1 cells are a pluripotent EC cell line that differentiates into various cell types, including neuronal cells, in response to RA treatment (45,46). This cell line has been used to characterize the transcription of Hox genes following RA treatment (47–49). Systematic analysis of all four clusters of 39 Hox genes demonstrated that none is expressed in untreated EC cells at levels that are detectable by northern analysis, whereas transcription of 24 genes is induced after 7 days of treatment with 10 µM RA. Expression of the remaining genes, all belonging to the 5' paralogy groups, remains undetectable even after longer exposures to RA. The boundary between inducible and uninducible genes is at the level of the ninth paralogy group. Expression studies with more sensitive procedures, such as RNase protection assays, revealed that the low baseline transcript levels of the most 5' Hoxc genes and of the five most 5' Hoxd genes that were detectable in EC cells remain unchanged or decreased after RA treatment (47–49). Besides RA, many other agents have been tested in various attempts to modulate HOX gene expression in EC cells, but none were successful (E. Boncinelli, unpublished data).
We treated NT2/D1 cells with low VPA concentrations (5 µg/ml), comparable to therapeutic plasma concentrations (30–100 µg/ml) (5–11) and in contrast to the high concentrations of RA used in the same experiments (47–49). Classical studies with RA used concentrations (10 mM) that are 10- to 1000-fold greater than those that are sufficient to induce maximal transcription levels of Hox (43–46). As predicted, VPA affected HOX expression (Fig. 4). The strongest VPA responses were observed with Hoxd1, -d8, -d10 and -d11. Both RA and VPA upregulated Hoxd1 and Hoxd8 and downregulated Hoxd11 and Hoxd12 to various extents after prolonged in vitro treatment (168 h). Interestingly, Hoxd10 responded to RA treatment in a biphasic manner, being substantially downregulated after a short treatment, and upregulated subsequently. In contrast to RA, VPA treatment failed to downregulate Hoxd10 significantly, but instead led to an increase in its transcription levels after 48 and 72 h of treatment. The other genes either failed to respond or responded weakly to VPA treatment (Fig. 4). These data suggest that in some cases changes in HOX transcript levels in VPA-treated cells were similar to those induced with RA treatment. In other cases, including Hoxd10, changes in gene expression differed from those observed after RA treatment.
|
Many of the axial homeotic effects found in VPA-treated fetuses are consistent with dysregulation of HOX gene expression observed in EC cells treated with the same compound, in particular with respect to Hoxd10 and Hoxd11. The Hoxd10 gene has an anterior expression boundary at the third lumbar vertebra (28). Mice with a targeted disruption of the Hoxd10 gene show defects in lumbar and sacral neuromeres as well as malformations of the axial mesoderm in the sacrum (50). Similarly, mice with a targeted disruption of the Hoxd11 gene, whose anterior expression boundary lies at the sixth lumbar vertebra (28), show homeosis in the sacrum (50). Malformations found in the lumbar, sacral and appendicular regions of VPA-treated fetuses are consistent with those expected after dysregulation of 5' Hox genes during early development.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Teratogenicity of anti-epileptic and mood-stabilizing drugs
Treatment of epilepsy and bipolar disorder during pregnancy is greatly complicated by the teratogenicity of most medications. Tens of thousands of pregnant women may be exposed to medications during pregnancy to manage seizures and mood disorders. However, use of these medications places fetuses at significantly higher risk for drug-induced birth defects. Valproate, lithium, barbiturates, carbamazepine and vigabatrin are effective pharmacotherapies to prevent seizures and in some cases as mood stabilizers (1–3,51,52), but all cause similar axial, appendicular and organ defects in mice or rats (14–17,53–56). The therapeutic and teratogenic similarities suggest shared mechanisms of action. However, the frequency of affected fetuses is often modest, sometimes involving only 1–10% of those exposed (4). The low rate of affected fetuses in humans and in segregating crosses of mice raises the possibility that genetic factors predispose to malformation (4–6). The variable sensitivity among inbred strains of mice supports the argument that genetic factors are important (4,18,19). Obviously, understanding these genetic factors would greatly aid the planning of treatment programs for pregnant women. The present study provides insight into the nature of these genetic factors and the mechanism of teratogenic action. We also found that VPA induced homeotic transformations and altered HOX expression patterns, suggesting a mechanism of teratogenicity.
Parental effects
Genes controlling susceptibility to VPA-induced malformations could reside in the fetal or parental genomes, or both. Previous in vitro experiments showed that VPA, like several other anti-epileptic drugs, acts directly on embryos (18,19,57,58). However, embryonic factors alone are not sufficient to account for the contrasting frequency of VPA-induced malformations in genetically identical embryos resulting from reciprocal crosses (Table 1). If embryonic factors were a sufficient explanation, these genetically identical fetuses derived from reciprocal crosses should have been similarly affected. In fact, hybrid embryos showed patterns of malformations that were generally similar to their maternal genotype. For example, oligodactylies, fused neural arches, split vertebrae and spina bifida were frequent in P/J and (P x D)F1 hybrids, but not in DBA/2J or (D x P)F1 hybrids (Table 3).
The nature of the parental factor is uncertain and could involve genes on the X chromosome, imprinting or mitochondrial genes that modulate VPA pharmacokinetics. An X-linked gene(s) with recessive effects could have contrasting effects on male offspring from the reciprocal crosses. Two important tests involve determining whether hybrid males are more susceptible than females and evaluating susceptibility to VPA-induced malformation among progeny of crosses with these hybrid males. Imprinting could also account for the parental effects. Two other important tests involve nuclear transfers between DBA/2J and P/J conceptuses and embryo transfers between DBA/2J and P/J dams. Together, these transfers would test the imprinting and mitochondrial hypotheses.
Given their maternal inheritance, mitochondria are a provocative explanation. Chronic administration of VPA in humans and in rats results in hepatotoxicity including damage to mitochondria and peroxisomes (59–61). Medium-chain acyl-CoA dehydrogenase (where CoA is coenzyme A), a mitochondrial protein involved in intermediary metabolism, is strongly elevated in retinoid X receptor alpha (RXRA)-deficient mice (62). Valproate can be converted to valproyl-CoA, thereby trapping CoA and leading to functional CoA deficiency (63) as well as dramatically reduced levels of acetyl-CoA and acyl-CoA (64). Moreover, valproate affects adult but not embryonic acetyl-CoA levels (65). These observations may explain why in vivo but not in vitro folinic acid (5-formyl tetrahydrofolate) treatment ameliorates many of VPA’s teratogenic effects (66,67, but see ref. 68). Folinic acid treatments should increase 5,10-methylene-tetrahydrofolate levels and promote synthesis of CoA (69). Low maternal folate levels are associated with increased risk for developmental anomalies after treatment with anti-epileptic drugs (21). We propose that DBA/2J and P/J differ genetically in their mitochondrial response to VPA and these mitochondrial anomalies contribute to many VPA-induced birth defects.
Another dimension to the adverse effects of VPA on development involves the recent discovery that VPA interferes with AP1 transcription factor activity, GSK3ß phosphorylation of c-fos and c-jun and ß-catenin translocation to the nucleus (70–72). Mutated genes in these pathways can cause diverse developmental defects. VPA teratogenesis may in part involve dysregulation of key genes in the hedgehog pathway.
Homeotic transformations
Homeotic transformations are unique types of birth defect involving specific genes and developmental processes. We demonstrated anterior homeotic transformations in VPA-exposed mouse fetuses. Carbamazepine and vigabatrin treatment can also result in additional ribs, which can be interpreted as homeotic transformations (52–54). The occurrence of subtle but unique developmental alterations suggests that these anti-epileptic and mood-stabilizing drugs do not simply disrupt development in a non-specific manner, but instead act by altering HOX gene expression. Homeotic transformations are often observed in mice that have engineered mutations in Hox genes (29–42). In addition, RA induces homeotic transformations (27,28) and alters HOX expression patterns (27,28,43–49). We discovered that VPA causes similar birth defects and HOX expression changes, although subtle quantitative and qualitative differences in their effects were found. VPA-induced dysregulation of HOX gene expression might disturb positional cues required for closure of the posterior neuropore. Interestingly, many of the Hox genes that are altered by VPA in vitro define posterior positional identities, as is the case for Hoxd8 (Th11–Th12), Hoxd10 (L3–L5) and Hoxd11 (S3) (27,28). We therefore propose that VPA teratogenicity is mediated in part through changes in Hox gene expression.
Summary
Parental effects, homeotic transformation and disrupted HOX expression patterns may account for the teratogenicity of VPA and perhaps that of related drugs. Because VPA shares many functional and teratogenic similarities with other neuroleptic drugs, it is now important to test whether these other drugs also induce homeotic transformations and alter HOX expression, and whether drugs used for treating epilepsy as well as stabilizing moods act by modulating the expression of various Hox genes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Mice
Eight inbred strains, A/J, AKR/J, BALB/cByJ, C57BL/6J, C3H/HeJ, DBA/2J, P/J and SWR/J, were maintained on a 12 h light cycle in the mouse facilities of the Jackson Laboratory or the Department of Genetics, Case Western Reserve University School of Medicine. The mice were pathogen free and without obvious health problems. Nulliparous females that were 6–12 weeks old were mated overnight to appropriate breeder males. Females were examined at 07:00 and 12:00 h for the presence of a vaginal plug. The onset of gestation (day 0, h 0, designated E0.0) was set at midnight of the previous night.
Crosses
To obtain reciprocal F1 hybrids, P/J females were mated to DBA/2J males and DBA/2J females were mated to P/J males, respectively. The resulting hybrids were designated (P x D)F1 and (D x P)F1, respectively.
VPA treatments
Sodium valproate (Sigma, St Louis, MO) was dissolved in distilled water and administered at doses of 200, 300 or 400 mg VPA/kg body wt, the latter being a high teratogenic dose for mice (73). Control animals received an equivalent volume of distilled water. VPA was injected intraperitoneally to pregnant primiparous females.
The therapeutic concentration of VPA in humans is 50–150 µg/ml, but peak values are >200 g/ml. In mice, a single injection of 200 mg/kg results in 307 ± 41 g/ml in maternal plasma and 153 ± 18 µg/g in gestational material (73). The half-life of VPA is only ~1 h in the mouse and ~10 h in humans (73), which means that after 1 h the concentration is only 150 µg/ml and is abolished quickly. Injections were therefore made three times in each mouse at E9.25, E9.5 and E9.75. The peak concentration is ~30 min after injection (73). Peak values are important for teratogenicity—a constant delivery with minipumps reduces exencephaly but increases lethality (73).
Skeletal preparations
Fetuses were examined at E18.5. Dams were killed and uteri removed. Fetuses were fixed in 95% ethanol for at least 4 days, scored for external malformations, and then they were skinned and eviscerated. Cartilage was stained with alcian blue, bone with alizarin red and fetuses cleared according to established methods (28).
Morphological assessment
A panel of 16 traits was used to assess each fetus: (i) number of missing post-axial digits on the left paw; (ii) number of missing post-axial digits on the right paw; (iii) syndactyly (presence versus absence); (iv) polydactyly; (v) dysplasia of the fourth digit; (vi) spina bifida (distance between tips of the opposing neural arches of the first lumbar vertebrae); (vii) cleft palate (presence versus absence); (viii) number of hemi-vertebrae; (ix) number of fused vertebral bodies; (x) number of split vertebrae; (xi) number of fused neural arches; (xii) number of fused ribs; (xiii) severity of neural arch fusions; (xiv) severity of rib fusions; (xv) viability (alive or dead); and (xvi) fetal weight in grams. These traits are based on standard vertebral features (74–76). Visceral anomalies were not assessed.
Units of measurement
Depending on the trait, either quantitative or qualitative measures were used to summarize data for embryotoxicity, fetal weight and the various malformations. Embryo lethality was measured as the number of absorbed embryos as a function of the total number of implantation sites per female, and summarized as the average rate of embryo lethality per female; fetal weight was measured in day 18.5 fetuses immediately after dissection and summarized as average fetal weight; fused ribs, vertebral arches or vertebral bodies were measured as the total number of fused ribs, arches or bodies per fetus and summarized as the average of each kind of fusion over all analyzed fetuses; severity of fusions was measured as the total number of ribs in a single fusion in each embryo and summarized as the average of each of these numbers; split vertebrae or hemi-vertebrae were scored as positive if a fetus had split vertebrae or hemi-vertebrae and negative if an anomaly was not observed, and summarized as the percentage of affected fetuses; spina bifida occulta was measured in arbitrary units where 1 unit = 20 µm; oligodactyly was scored as positive if at least one digit was missing and summarized as the percentage of affected fetuses; syndactyly was scored as positive if at least two digits were fused and summarized as the percentage of affected fetuses; polydactyly was scored as positive if more than the normal number of digits was present and summarized as the percentage of affected fetuses; dysplasia of the fourth digit was scored as positive if the fourth digit was malformed, primarily due to shortening of the os metacarpal IV and summarized as the percentage of affected fetuses; and cleft palate was scored as positive if a fetus presents with cleft palate and summarized as the percentage of affected fetuses.
Cell cultures
The human embryonal carcinoma cells, Ntera-2 clone D1 (NT2/D1) (45,46), were maintained at high density in Dulbecco’s modified minimal essential medium supplemented with 10% fetal bovine serum. Cultures to be treated with RA and VPA were established by seeding cells at a density of 106 cells/75 cm2 tissue culture flask. RA (10 mM solution in dimethyl sulfoxide, all-trans; Eastman Kodak, Rochester, NY) was added to a final concentration of 10 fM; VPA (Sigma) was added to a final concentration of 50 µg/ml. Cells were refreshed every 48 h with fresh medium containing RA or VPA. The percentage of undifferentiated stem cells, as monitored by immunochemical staining, decreased steadily after RA addition, reaching values of 2% within 7 days of treatment. Differentiation was characterized by the appearance of several cell types, including neurons, at 7–28 days after induction (45,46). Although several morphologically distinct cell types may be found in treated cultures, only neurons were positively identified. The heterogeneity of the cultures increased with time after the addition of RA and recognizable neuronal cells appeared relatively late (3–4 weeks) after most morphological changes had already occurred.
RNAs extracted at various time points after treatment with either RA or VPA were analyzed by RNase protection assays using probes corresponding to human homeobox genes Hoxb1–b3, -d1, -d8 and -d10–d13. The lengths of the HOX probes ranged from 150 to 350 bp. Inserts were obtained from the second or third exon of various Hox genes, and none of the probes contained the homeobox sequence, which could produce cross-hybridization due to lack of specificity. For normalization, we used a 150 bp probe from a human ß-actin plasmid, yielding a protected band of 86 bp. RNase protection assay results reported in Figure 4 are representative of at least three independent assays.
RNA isolation and analysis
Cytoplasmic RNA was extracted from 75 cm2 tissue culture flasks of confluent cells at different time points after treatment with RA or VPA, and hybridized overnight in 30–50 µg aliquots to radiolabeled antisense RNA probes corresponding to the analyzed Hox genes [single DNA fragments including the 3' portion of the homeobox were subcloned in a pGEM3 vector (Promega, Madison, WI)] or human ß-actin RNA probes. Mixtures were then digested with RNase A and T1 (1 h at 32°C) and proteinase-K (Promega), extracted with phenol–chloroform and precipitated in ethanol. Electrophoresis was carried out on 6% urea–polyacrylamide sequencing gels that were dried and exposed for 8–96 h at –70°C to Kodak X-AR5 films. Quantitation of the relative amounts of protected mRNA was performed by standard densitometry scanning of at least three autoradiographs representing independent experiments.
Statistical analysis
Correlation analysis was used to identify independent traits and the correlation coefficient was calculated for each of the 120 = [16(15)/2] pairwise morphological assessments for each strain. The threshold correlation coefficient was set at r = 0.707, which would explain at least 50% of the variation. The statistical package SPSS 6.0 was used for the other statistical tests. For weight data, the statistical analysis included Student’s t-tests and analysis of variance followed by Scheffé’s test. For the remaining traits, either the non-parametric Mann–Whitney U/Wilcoxon Rank Sum W test or the likelihood ratio test, both corrected for ties, were used.
|
ACKNOWLEDGEMENTS |
|---|
We thank Terry Magnuson and Bryan Roth for many helpful comments on a draft of this manuscript. R.B. was funded by EC grant BMH4-CT96-1418 and a grant from the GSF. G.G.C. and E.B. were funded by grants from the Italian Telethon and by the Armenise-Harvard Foundation. J.H.N. was funded by grants from the March of Dimes, NIH NHLBI 58982 and NIH NICHD 37843. This research was also supported by a grant from the Howard Hughes Medical Institute to Case Western Reserve University School of Medicine.
|
FOOTNOTES |
|---|
+ These authors contributed equally to this work
§ To whom correspondence should be addressed at: Department of Genetics, BRB-630, CWRU, 10900 Euclid Avenue, Cleveland, OH 44106, USA. Tel: +1 216 368 0581; Fax: +1 216 368 3432; Email: jhn4@po.cwru.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Guay, D.R. (1995) The emerging role of valproate in bipolar disorder and other psychiatric disorders. Pharmacotherapy, 15, 631–647.[Web of Science][Medline]
2 Feely, M. (1999) Fortnightly review: drug treatment of epilepsy. Br. Med. J., 318, 106–109.
3 Tohen, M. and Grundy, S. (1999) Management of acute mania. J. Clin. Psychiat., 60(Suppl. 5), 31–34.
4 Finnell, R.H. (1991) Genetic differences in susceptibility to anticonvulsant drug-induced developmental defects. Pharmacol. Toxicol., 69, 223–227.[Web of Science][Medline]
5 Robert, E. (1991) Risques teratogenes de l’epilepsie et des antiepileptiques. Pediatrie, 46, 579–583.[Web of Science][Medline]
6 Lindhout, D. and Omtzigt, J.G. (1992) Pregnancy and the risk of teratogenicity. Epilepsia, 33(Suppl. 4), S41–S48.
7 Robert, E. and Guibaud, P. (1982) Maternal valproic acid and congenital neural tube defects. Lancet, ii, 937.
8 Robert, E. and Rosa, F. (1983) Valproate and birth defects. Lancet, ii, 1142.
9 Lindhout, D. and Schmidt, D. (1986) In-utero exposure to valproate and neural tube defects. Lancet, i, 1392–1393.
10 Lammer, E.J., Sever, L.E. and Oakley, G.P. (1987) Teratogen update: valproic acid. Teratology, 35, 465–473.[Web of Science][Medline]
11 Clayton-Smith, J. and Donnai, D. (1995) Fetal valproate syndrome. J. Med. Genet., 32, 724–727.
12 Jones, K.L., Lacro, R.V., Johnson, K.A. and Adams, J. (1989) Patterns of malformations in the children of women treated with carbamazepine during pregnancy. N. Engl. J. Med., 320, 1661–1666.[Abstract]
13 Koch, S., Losche, G., Jager-Roman, E., Jakob, S., Rating, D., Diechl, A. and Helge, H. (1992) Major and minor birth malformations and antiepileptic drugs. Neurology, 42(Suppl. 5), 83–88.[Web of Science][Medline]
14 Ehlers, K., Sturje, H., Merker, H.J. and Nau, H. (1992) Spina bifida aperta induced by valproic acid and by all-trans-retinoic acid in the mouse: distinct differences in morphology and periods of sensitivity. Teratology, 46, 117–130.[Web of Science][Medline]
15 Nau, H., Hauck, R.S. and Ehlers, K. (1991) Valproic acid-induced neural tube defects in mouse and human: aspects of chirality, alternative drug development, pharmacokinetics and possible mechanisms. Pharmacol. Toxicol., 69, 310–321.[Web of Science][Medline]
16 Ehlers, K., Stürje, H., Merker, H.-J. and Nau, H. (1992) Valproic acid-induced spina bifida: a mouse model. Teratology, 45, 145–154.[Web of Science][Medline]
17 Narotsky, M.G., Francis, E.Z. and Kavlock, R.J. (1994) Developmental toxicity and structure–activity relationships of aliphatic acids, including dose–response assessment of valproic acid in mice and rats. Fundam. Appl. Toxicol., 22, 251–265.[Web of Science][Medline]
18 Kao, J., Brown, N., Schmid, B., Goulding, E. and Fabro, S. (1981) Teratogenicity of valproic acid: in vivo and in vitro investigations. Teratogen. Carcinogen. Mutagen., 1, 367–382.
19 Naruse, I., Collins, M.D. and Scott, W.J. (1988) Strain differences in the teratogenicity induced by sodium valproate in cultured mouse embryos. Teratology, 38, 87–96.[Web of Science][Medline]
20 Finnell, R.H., Bennett, G.D., Karras, S.B. and Mohl, V.K. (1988) Common hierarchies of susceptibility to the induction of neural tube defects in mouse embryos by valproic acid and its 4-propyl-4-pentenoic acid metabolite. Teratology, 38, 313–320.[Web of Science][Medline]
21 Dansky, L.V., Rosenblatt, D.S. and Andermann, E. (1992) Mechanisms of teratogenesis: folic acid and antiepileptic therapy. Neurology, 42, 32–42.
22 Bui, L.M., Taubeneck, M.W., Commiso, J.F., Uriu-Hare, J.Y., Farber, W.D. and Keen, C.L. (1998) Altered zinc metabolism contributes to the developmental toxicity of 2-ethylhexanoic acid, 2-ethylhexanol and valproic acid. Toxicology, 126, 9–21.[Web of Science][Medline]
23 Wegner, C., Drews, E. and Nau, H. (1990) Zinc concentrations in mouse embryo and maternal plasma. Effect of valproic acid and non-teratogenic metabolite. Biol. Trace Elem. Res., 25, 211–217.[Web of Science][Medline]
24 Clarke, D.O. and Brown, N.A. (1987) Valproic acid teratogenesis and embryonic lipid metabolism. Arch. Toxicol., 11(suppl.), 143–147.
25 Nau, H. (1986) Species differences in pharmacokinetics and drug teratogenesis. Environ. Health Perspect., 70, 113–129.[Web of Science][Medline]
26 Bruckner, A., Lee, Y.J., O’Shea, K.S. and Henneberry, R.C. (1983) Teratogenic effects of valproic acid and diphenylhydantoin on mouse embryos in culture. Teratology, 27, 29–42.[Web of Science][Medline]
27 Kessel, M. and Gruss, P. (1991) Homeotic transformations of murine vertebrae and concomitant alteration of Hox codes induced by retinoic acid. Cell, 67, 89–104.[Web of Science][Medline]
28 Kessel, M. (1992) Respecification of vertebral identities by retinoic acid. Development, 115, 487–501.[Abstract]
29 Mouellic, H., Lallemand, Y. and Brulet, P. (1992) Homeosis in the mouse induced by a null mutation in the Hox-3.1 gene. Cell, 69, 251–254.[Web of Science][Medline]
30 Fromental-Ramain, C., Warot, X., Lakkaraju, S., Favier, B., Haack, H., Birling, C., Dietrich, A., Dollé, P. and Chambon, P. (1996) Specific and redundant functions of the paralogous Hoxa-9 and Hoxd-9 genes in the forelimb and axial skeleton patterning. Development, 122, 461–472.[Abstract]
31 Horan, G.S., Ramirez-Solis, R., Featherstone, M.S., Wolgemuth, D.J., Bradley, A. and Behringer, R.R. (1995) Compound mutants for the paralogous hoxa-4, hoxb-4, and hoxd-4 genes show more complete homeotic transformations and a dose-dependent increase in the number of vertebrae transformed. Genes Dev., 9, 1667–1777.
32 Jegalian, B.G. and DeRobertis, E.M. (1992) Homeotic transformations in the mouse induced by over-expression of a human Hox3.3 transgene. Cell, 71, 901–910.[Web of Science][Medline]
33 Balling, R., Mutter, G., Gruss, P. and Kessel, M. (1989) Craniofacial abnormalities induced by ectopic expression of the homeobox gene Hox-1.1 in transgenic mice. Cell, 58, 337–347.[Web of Science][Medline]
34 Pollack, R.A., Sreenath, T., Ngo, L. and Bieberich, C.J. (1995) Gain of function mutations for paralogous Hox genes: implications for the evolution of Hox gene function. Proc. Natl Acad. Sci. USA, 92, 4492–4496.
35 Lohnes, D., Kastner, P., Dierich, A., Mark, M., LeMeur, M. and Chambon, P. (1993) Function of retinoic acid receptor gamma in the mouse. Cell, 73, 643–658.[Web of Science][Medline]
36 Lohnes, D., Mark, M., Mendelsohn, C., Dollé, P., Dierich, A., Gorry, P., Gansmuller, A. and Chambon, P. (1994) Function of the retinoic acid receptors (RARs) during development (I). Craniofacial and skeletal abnormalities in RAR double mutants. Development, 120, 2723–2748.[Abstract]
37 Akasaka, T., Kanno, M., Balling, R., Mieza, M.A., Taniguchi, M. and Koseki, H. (1996) A role for mel-18, a Polycomb group-related vertebrate gene, during the anteroposterior specification of the axial skeleton. Development, 122, 1513–1522.[Abstract]
38 Core, N., Bel, S., Gaunt, S.J., Aurrand-Lions, M., Pearce, J., Fisher, A. and Djabali, M. (1997) Altered cellular proliferation and mesoderm patterning in Polycomb-M33-deficient mice. Development, 124, 721–729.[Abstract]
39 Van der Lugt, N.M.T., Domen, J., Linders, K., van Roon, M., Robanus-Maandag, E., te Riele, H., van der Valk, M., Deschamps, J., Sofroniew, M., van Lohuizen, M. et al. (1994) Posterior transformation, neurological abnormalities, and severe hematopoietic defects in mice with a targeted deletion of the bmi-1 proto-oncogene. Genes Dev., 8, 757–769.
40 Schumacher, A., Faust, C. and Magnuson, T. (1996) Positional cloning of a global regulator of anterior-posterior patterning in mice. Nature, 383, 250–253.[Medline]
41 Davis, A.P. and Capecchi, M.R. (1996) A mutational analysis of the 5' HoxD genes: dissection of genetic interactions during limb development in the mouse. Development, 122, 1175–1185.[Abstract]
42 Dolle, P., Izpisua-Belmonte, J.C., Tickle, C. and Duboule, D. (1993) Hox genes and the morphogenesis of the vertebrate limb. Prog. Clin. Biol. Res., 383A, 11–20.
43 Mavilio, F. (1993) Regulation of vertebrate homeobox-containing genes by morphogens. Eur. J. Biochem., 212, 273–288.[Web of Science][Medline]
44 Morris-Kay, G.M. and Ward, S.J. (1999) Retinoids and mammalian development. Int. Rev. Cytol., 188, 73–131.[Web of Science][Medline]
45 Andrews, P.W. (1984) Retinoic acid induces neuronal differentiation of a cloned human embryonal carcinoma cell line in vitro. Dev. Biol., 103, 285–293.[Web of Science][Medline]
46 Thompson, S., Stern, P.L., Webb, M., Walsh, F.S., Engstrom, W., Evans, E.P., Shi, W.-K., Hopkins, B. and Graham, C.F. (1984) Cloned human teratoma cells differentiate into neuron-like cells and other cell types in retinoic acid. J. Cell Sci., 72, 37–64.[Abstract]
47 Simeone, A., Acampora, D., Arcioni, L., Andrews, P.W., Boncinelli, E. and Mavillo, F. (1990) Sequential activation of HOX2 homeobox genes by retinoic acid in human embryonal carcinoma cells. Nature, 346, 763–766.[Medline]
48 Simeone, A., Acampora, D., Negro, V., Faiella, A., D’Esposito, M., Stornaiuolo, A., Mavilo, F. and Boncinelli, E. (1991) Differential regulation by retinoic acid of the homeobox genes of the four HOX loci in human embryonal carcinoma cells. Mech. Dev., 33, 215–227.[Web of Science][Medline]
49 Faiella, A., Zappavigna, V., Mavilio, F. and Boncinelli, E. (1994) Inhibition of retinoic acid-induced activation of 3' human HOXB genes by antisense oligonucleotides affects sequential activation of genes located upstream in the four HOX clusters. Proc. Natl Acad. Sci. USA, 91, 5335–5339.
50 Carpenter, E.M., Goddard, J.M., Davis, A.P., Nguyen, T.P. and Capecchi, M.R. (1997) Targeted disruption of Hoxd10 affects mouse hindbrain development. Development, 124, 4505–4514.[Abstract]
51 Solomon, D.A., Keitner, G.I., Miller, I.W., Shea, M.T. and Keller, M.B. (1995) Course of illness and maintenance treatments for patients with bipolar disorder. J. Clin. Psychiatr., 56, 5–13.
52 Retzow, A. and Emrich, H.M. (1998) Therapy for bipolar affective illnesses with valproate. A review of the literature. Psychiatr. Prax., 25, 163–171.[Web of Science][Medline]
53 Vorhees, C.V., Acuff, K.D., Weisenburger, W.P. and Minck, D.R. (1990) Teratogenicity of carbamazepine in rats. Teratology, 41, 311–317.[Web of Science][Medline]
54 Abdulrazzaq, Y.M., Bastaki, S.M. and Padmanabhan, R. (1997) Teratogenic effects of vigabatrin in TO mouse fetuses. Teratology, 55, 165–176.[Web of Science][Medline]
55 Yasuda, M. and Maeda, H. (1972) Teratogenic effects of 4-chloro-2-methylphenoxyacetic acid ethylester (MCPEE) in rats. Toxicol. Appl. Pharmacol., 23, 326–333.[Web of Science][Medline]
56 Marathe, M.R. and Thomas, G.P. (1986) Embryotoxicity and teratogenicity of lithium carbonate in Wistar rats. Toxicol. Lett., 34, 115–120.[Web of Science][Medline]
57 Hansen, D.K., Diak, S.L., Terry, K.K. and Grafton, T.F. (1996) In vitro embryotoxicity of carbamazepine and carbamazepine-10,11-epoxide. Teratology, 54, 45–51.[Web of Science][Medline]
58 Piersma, A.H., Verhoef, A., Opperhuizen, A., Klaassen, R., van Eijkeren, J. and Olling, M. (1998) Embryotoxicity of carbamazepine in rat postimplantation embryo culture after in vitro exposure via three different routes. Reprod. Toxicol., 12, 161–168.[Web of Science][Medline]
59 Powell-Jackson, P.R., Tredger, J.M. and Williams, R. (1984) Hepatotoxicity to sodium valproate: a review. Gut, 25, 673–681.
60 Cotariu, D. and Zaidman, J.L. (1988) Valproic acid and the liver. Clin. Chem., 34, 890–897.
61 Sobaniec-Lowtowska, M.E. (1997) Effects of long-term administration of the antiepileptic drug sodium valproate upon the ultrastructure of hepatocytes in rats. Exp. Toxicol. Pathol., 49, 225–232.[Web of Science][Medline]
62 Ruiz-Lozano, P., Smith, S.M., Perkins, G., Kabaiak, S.W., Boss, G.R., Sucov, H.M., Evans, R.M. and Chien, K.R. (1998) Energy deprivation and a deficiency in downstream metabolic targets genes during the onset of embryonic heart failure in RXR
–/– embryos. Development, 125, 533–544.[Abstract]
63 Ponchaut, S., van Hoof, F. and Veitch, K. (1992) In vitro effects of valproate and valproate metabolites on mitochondrial oxidations. Relevance of CoA sequestration to the observed inhibitions. Biochem. Pharmacol., 43, 2435–2442.[Web of Science][Medline]
64 Krahenbuhl, S., Mang, G., Kupferschmidt, H., Meier, P.J. and Krause, M. (1995) Plasma and hepatic carnitine and coenzyme A pools in a patient with fatal, valproate hepatotoxicity. Gut, 37, 140–143.
65 Coakley, M.E., Rawlings, S.J. and Brown, N.A. (1986) Short-term carboxylic acids, a new class of teratogens: studies of potential biochemical mechanisms. Environ. Health Perspect., 70, 105–111.[Web of Science][Medline]
66 Trotz, M., Wegner, C. and Nau, H. (1987) Valproic acid-induced neural tube defects: reduction by folinic acid in the mouse. Life Sci., 41, 103–110.[Web of Science][Medline]
67 Hansen, D.K. and Grafton, T.F. (1991) Lack of attenuation of valproic acid-induced effects of folinic acid in rat embryos in vitro. Teratology, 43, 575–582.[Web of Science][Medline]
68 Hansen, D.K., Grafton, T.F., Dial, S.L., Gehring, T.A. and Siitonen, P.H. (1995) Effect of supplemental folic acid on valproic acid-induced embryotoxicity and tissue zinc levels in vivo. Teratology, 52, 277–285.[Web of Science][Medline]
69 Michal, G. (1999) Biochemical Pathways. John Wiley and Sons, New York, NY.
70 Chen, G., Yuan, P.X., Jiang, Y.M., Huang, L.D. and Manji, H.K. (1999) Valproate robustly enhances AP-1 mediated gene expression. Brain Res. Mol. Brain Res., 64, 52–58.[Medline]
71 Chen, B., Wang, J.F., Hill, B.C. and Young, L.T. (1999) Lithium and valproate differentially regulate brain regional expression of phosphorylated CREB and c-FOS. Brain Res. Mol. Brain Res., 70, 45–53.[Medline]
72 Klein, P.S. and Melton, D.A. (1996) A molecular mechanism for the effect of lithium on development. Proc. Natl Acad. Sci. USA, 93, 8455–8459.
73 Nau, H. (1985) Teratogenic valproic acid concentrations: infusion by implanted minipumps vs conventional injection regimen in the mouse. Toxicol. Appl. Pharmacol., 80, 243–250.[Web of Science][Medline]
74 Washburn, S.L. and Buettner-Janusch, J. (1952) The definition of thoracic and lumbar vertebrae. Am. J. Phys. Anthrop., 10, 251–252.
75 Struthers, J. (1875) On variations of the vertebrae and ribs in man. J. Anat. Phys., IX, 17–96.
76 Pollack, R.A., Gilbert, J. and Bieberich, C.J. (1992) Altering the boundaries of Hox3.1 expression: evidence for antipodal gene regulation. Cell, 71, 911–923.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. M. Turner Epigenetic responses to environmental change and their evolutionary implications Phil Trans R Soc B, November 27, 2009; 364(1534): 3403 - 3418. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J Craig and S. J Hunt Treating women with juvenile myoclonic epilepsy Practical Neurology, October 1, 2009; 9(5): 268 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Towers and C. Tickle Growing models of vertebrate limb development Development, January 15, 2009; 136(2): 179 - 190. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Di Renzo, M. L. Broccia, E. Giavini, and E. Menegola Relationship between Embryonic Histonic Hyperacetylation and Axial Skeletal Defects in Mouse Exposed to the Three HDAC Inhibitors Apicidin, MS-275, and Sodium Butyrate Toxicol. Sci., August 1, 2007; 98(2): 582 - 588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. G. Costa, L. Steardo, and V. Cuomo Structural Effects and Neurofunctional Sequelae of Developmental Exposure to Psychotherapeutic Drugs: Experimental and Clinical Aspects Pharmacol. Rev., March 1, 2004; 56(1): 103 - 147. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Iezzi, G. Cossu, C. Nervi, V. Sartorelli, and P. L. Puri Stage-specific modulation of skeletal myogenesis by inhibitors of nuclear deacetylases PNAS, May 28, 2002; 99(11): 7757 - 7762. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J C S Dean, H Hailey, S J Moore, D J Lloyd, P D Turnpenny, and J Little Long term health and neurodevelopment in children exposed to antiepileptic drugs before birth J. Med. Genet., April 1, 2002; 39(4): 251 - 259. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M R JOHNSON, M R JOHNSON, J W A S SANDER, and J W A S SANDER The clinical impact of epilepsy genetics J. Neurol. Neurosurg. Psychiatry, April 1, 2001; 70(4): 428 - 430. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||