Depletion of Glutathione Induces 4-Hydroxynonenal Protein Adducts and Hydroxyurea Teratogenicity in the Organogenesis Stage Mouse Embryo

Glutathione (γ-glutamylcysteinylglycine, GSH), synthesized de novo by the sequential actions of γ-glutamylcysteine synthetase (γ-GCS) and glutathione synthetase, is the principal nonprotein sulfhydryl in mammalian cells. GSH serves mainly as a thiol substrate for enzymes to reduce free radicals and hydroperoxides and regulate protein functions. Disturbances in GSH homeostasis may cause oxidative stress, inducing damage to cellular macromolecules (lipids, proteins, and DNA).

GSH homeostasis is critical to maintenance of cellular redox status and essential for normal embryo development. Targeted disruption of the catalytic subunit of γ-GCS, the rate-limiting enzyme for GSH synthesis, induces embryo lethality between gestational day(s) (GD) 7.5 and 8.5 (Shi et al., 2000). Depletion of GSH with l-buthionine-S,R-sulfoximine (BSO) between GD 10 and 11 in rats delays embryo growth and enhances embryonic deaths and malformations, both in vivo and in vitro (Slott and Hales, 1987a; Hales and Brown, 1991; Ozolins et al., 2002). Furthermore, GSH homeostasis is involved in defense of the embryo against teratogens. Thalidomide preferentially depletes GSH in thalidomide-sensitive species but not in thalidomide-resistant species (Hansen et al., 2002). During organogenesis, the addition of GSH protects cultured rat embryos from the embryotoxicity of the aldehyde acrolein (Slott and Hales, 1987b); disruption of GSH homeostasis exaggerates the teratogenicity of many drugs, such as 5-fluorouracil and phenytoin (Wong et al., 1989; Naya et al., 1990). Although oxidative stress has been proposed as a common mechanism of teratogen action, the underlying mechanism is not fully understood (Wells et al., 1997).

When oxidative stress occurs, the polyunsaturated fatty acids within the cellular membrane are a primary target of free radicals. 4-Hydroxynonenal (4-HNE), an α,β-unsaturated aldehyde, is a major lipid peroxidation product of n-6 polyunsaturated fatty acids. Relatively more stable than free radicals, 4-HNE passes easily among subcellular compartments to react with a variety of biomolecules bearing thiol and amino groups (Schaur, 2003). Through the formation of protein adducts, 4-HNE interferes with the activities of various signal kinases, such as protein kinase C and mitogen-activated protein kinases (MAPKs), to regulate cellular processes from proliferation to differentiation and apoptosis (Leonarduzzi et al., 2004). To the best of our knowledge, the impact of teratogen exposure on the formation of 4-HNE protein adducts in the conceptus has not been investigated previously.

Hydroxyurea (HU) is a model teratogen used to elucidate the relationship between embryotoxicity and oxidative stress (DeSesso, 1979). As a DNA synthesis inhibitor, HU destroys the free radicals in the catalytic center of ribonucleotide reductase and induces oxidative stress by generating free radicals. HU exposure induces extensive cell death in the neural tube region and limb buds (DeSesso, 1981; Zucker et al., 1999); antioxidants delay the onset of HU-induced cell death and reduce the incidence of external abnormalities (DeSesso, 1981; DeSesso et al., 1994). We have shown previously that exposure to teratogenic doses of HU on GD 9 in CD1 mice induces mainly hind limb and curly tail defects (indicative of neural tube defects), as well as a dose-dependent activation of activator protein 1 (AP-1) (Yan and Hales, 2005). AP-1 is a redox sensitive transcription factor that consists of jun/jun (c-jun, junB, and junD) or jun/fos (c-fos, fosB, fra-1, and fra2) dimeric nuclear proteins (Angel and Karin, 1991). AP-1 activation is important in the embryo during organogenesis and in mediating the response to stress (Wisdom, 1999; Jochum et al., 2001).

The goal of this study was to elucidate the impact of disturbances in GSH homeostasis on HU embryotoxicity; BSO, an irreversible inhibitor of γ-GCS, was used to deplete GSH (Griffith and Meister, 1979). The formation of 4-HNE-protein adducts and AP-1 DNA binding activity was assessed as indicators of the response of the embryo to oxidative stress.

Materials and Methods

Animals and Treatments. Timed-pregnant CD1 mice (20–25 g) were purchased from Charles River Canada Ltd. (St. Constant, QC, Canada) and housed in the McIntyre Animal Resource Centre (McGill University, Montreal, QC, Canada). All animal protocols were conducted in accordance with the guidelines outlined in the Guide to the Care and Use of Experimental Animals, prepared by the Canadian Council on Animal Care. Female mice, mated between 10:00 AM and 12:00 AM (GD 0), were treated with vehicle (saline) or BSO (Aldrich Chemical Co., Milwaukee, WI) at 600 mg/kg by i.p. injection at 7:00 AM on GD 9. After 4 h, female mice were treated with saline or HU (400 or 600 mg/kg) by i.p. injection. Dams were euthanized on GD 9 (0.5 or 3 h after treatment with HU; 6–10 litters/treatment group) or GD 18 (7–10 litters/treatment group) by cervical dislocation. On GD 9, the embryos were dissected out in Hanks' balanced salt solution (Invitrogen Canada, Inc., Burlington, ON, Canada) for the subsequent assessment of GSH and GSSG concentrations, the formation of 4-HNE protein adducts, and c-Fos-dependent AP-1 DNA binding activity. On GD 18, the uteri were removed, and the numbers of implantations, resorption sites, and live and dead fetuses were recorded. All of the live fetuses were weighed, inspected for external malformations, and then fixed in 95% ethanol for skeletal double staining and evaluation. Fetuses were skinned and double-stained with Alcian Blue (cartilage) and alizarin red S (bone) for the analysis of skeletal malformations, as described previously (Yan and Hales, 2005).

GSH and GSSG Determinations. At the time of collection on GD 9, four embryos from each litter were fixed in paraformaldehyde (4%) for immunostaining. The remaining embryos from each litter were placed in 40 μl of modified radioimmunoprecipitation buffer (150 mM NaCl; 1% Nonidet P-40; 0.5% sodium deoxycholate; 0.1% SDS; 50 mM Tris, pH 7.5) containing 10 μl/ml protease inhibitor cocktail (Active Motif Inc., Carlsbad, CA). The samples were homogenized with an ultrasonicator (Sonics & Materials, Inc., Newtown, CT) and centrifuged at 10,000g for 10 min at 4°C. From each sample, 30 μl of supernatant was removed and prepared for the measurement of GSH and GSSG, as described previously (Yan and Hales, 2005). The remaining supernatant from each sample was aliquoted, flash-frozen in liquid nitrogen, and stored at –80°C for protein assays (Bradford, 1976) (Bio-Rad Canada Ltd., Mississauga, ON, Canada), enzyme-linked immunosorbent assays (ELISA) tests, and Western blot analysis.

Immunofluorescence Staining. GD 9 embryos were fixed for 5 h at 4°C in 4% paraformaldehyde. After fixation, the embryos were dehydrated in ethanol, embedded in paraffin, and serially sectioned (5 μm sections). 4-HNE immunoreactivity was detected using a M.O.M immunodetection kit (Vector Laboratories, Burlingame, CA) as follows. Tissue sections were deparaffinized and hydrated. After rinsing twice for 2 min each with PBS, sections were incubated in the working solution of M.O.M. mouse IgG blocking reagent for 1 h. After further rinses with PBS, two times for 2 min each, sections were incubated in the working solution of M.O.M diluent for 5 min. Excess diluent solution was tipped off the slides, and the sections were incubated for 30 min at room temperature with a mouse monoclonal anti-4-HNE antibody (OXIS Research, Inc., Portland, OR) at 1 μg/ml diluted in M.O.M. diluent. After washing two times for 2 min in PBS, the sections were incubated in the working solution of M.O.M. biotinylated anti-mouse IgG Reagent for 10 min, followed by washing two times for 2 min in PBS. The sections were stained with Fluorescein Avidin DCS for 5 min, washed two times for 5 min in PBS, and then mounted with propidium iodide antifade solution (Chemicon International, Temecula, CA). As a negative control for 4-HNE staining, the primary antibody was preadsorbed with 4-hydroxy-2-nonenal-diethylacetal (OXIS Research, Inc.) as described by the manufacturer.

c-Fos ELISAs. The DNA binding activity of the c-Fos heterodimer complex was detected using ELISA transcription factor assay kits (Active Motif), as described previously (Yan and Hales, 2005).

Western Blot Analysis. Fifteen micrograms of protein from each sample was separated with 10% SDS-polyacrylamide gel electrophoresis and then transferred onto equilibrated polyvinylidene difluoride membranes (Amersham Biosciences, Buckinghamshire, UK) by electroblotting. Membranes were blocked in 5% skim milk and then probed with primary antibodies against 4-HNE (1:1000; OXIS Research, Inc.) or β-actin (1:500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) overnight at 4°C. After incubation with horseradish peroxidase-conjugated secondary antibodies (1:1000), proteins were detected by enhanced chemiluminescence (Amersham Biosciences). The bands were quantified by densitometric analysis using a Chemi-Imager 400 imaging system (Alpha Innotech, San Leandro, CA); the peak area represents the intensity of the band.

Statistical Analysis. Statistical analyses were done by chisquare or by two-way ANOVA, one-way ANOVA, or one-way ANOVA on ranks, as appropriate, using the SigmaStat computer program, followed by a post hoc Holm-Sidak or Dunn's multiple range test. The a priori level of significance was p < 0.05.

Results

Effects of BSO and HU on GSH Homeostasis. Whereas exposure to BSO alone did not significantly decrease GSH concentrations in the embryo by 0.5 h post-treatment, a decrease was observed at 3 h, compared with vehicle control (Fig. 1A). GSH concentrations in embryos exposed to HU alone (400 or 600 mg/kg) were lower at 3 h post-treatment than at 0.5 h (p < 0.05). BSO pretreatment enhanced the GSH depletion in low-dose (400 mg/kg) HU-exposed embryos at 0.5 h post-treatment, and in high-dose (600 mg/kg) HU-exposed embryos at 3 h post-treatment.

The ratios of GSSG/GSH in the embryos were not altered by treatment with BSO alone or HU alone (400 or 600 mg/kg) at either 0.5 or 3 h, compared with vehicle control (Fig. 1B). However, BSO exposure preceding high-dose HU (600 mg/kg) dramatically increased the GSSG/GSH ratio compared with vehicle control at both 0.5 and 3 h post-treatment. Thus, although neither HU nor BSO alone induced oxidative stress, as assessed by the GSSG/GSH ratio, the combination did.

Localization of Oxidative Stress in the Embryo: The Formation of 4-HNE Protein Adducts. The formation of 4-HNE protein adducts was assessed to elucidate the tissue specificity of the response of the embryo to BSO and HU exposure-induced oxidative stress. Embryos were examined 3 h after HU treatment, at the time when GSH homeostasis was significantly affected. In the control embryo, low amounts of 4-HNE immunoreactivity (green color) were detected in the neural epithelium, otic pit, branchial arch, mid-gut, and the caudal region of the tail (Fig. 2A). Exposure to HU alone (400 mg/kg) slightly increased the 4-HNE immunoreactivity in these regions (green color) (Fig. 2B); furthermore, 4-HNE reactivity was detected in nucleated blood cells (mainly yellow color) in these embryos (Fig. 2G, inset). The green fluorescence indicates that 4-HNE is localized in the cytoplasm, whereas the yellow fluorescence suggests that the 4-HNE adducts are present in nuclei. Treatment with high-dose HU (600 mg/kg) dramatically enhanced 4-HNE reactivity in all of the regions described in control embryos (Fig. 2C); in addition, intense staining was observed in blood cells (inset, Fig. 2H), in the somites, and in the neural tube in the caudal region of the embryo (Fig. 2C). Most of the 4-HNE immunoreactivity in embryos exposed to 600 mg/kg HU was in the cytoplasm but some appeared to be localized in nuclei, as shown in high magnification views of the neural tube (Fig. 2I, inset) and blood cells (Fig. 2H, inset).

Compared with vehicle control, BSO alone enhanced 4-HNE immunoreactivity in the neural tube, particularly in the neural epithelium in the midbrain region (Fig. 2D). Pretreatment with BSO increased the 4-HNE staining in embryos exposed to 400 mg/kg HU, specifically in the neural epithelium in the forebrain, in the mid-gut, and in the somites (Fig. 2E); this 4-HNE immunoreactivity was mainly cytoplasmic, although some nuclear localization, shown in yellow, is apparent in the forebrain neural epithelium (Fig. 2E). In the embryos exposed to both BSO and high-dose HU (600 mg/kg), the extent of nuclear localization of 4-HNE immunoreactivity was dramatically increased in all the regions described above (Fig. 2F), as illustrated in the high magnification inset of the neural tube close to the caudal region of the tail (Fig. 2J, inset). However, the integrity of the tissue in this region was affected, probably as a consequence of cytotoxicity (Fig. 2F).

There is little information on the nature of the 4-HNE protein adducts formed in any tissue. We used Western blot analysis to elucidate the molecular weight range of the 4-HNE-protein adducts formed in embryos exposed to oxidative stress. Two high-intensity bands were displayed in the molecular mass ranges of 150 and 100 kDa (Fig. 2K). Exposure to high-dose (600 mg/kg) HU alone increased the amounts of the 4-HNE-protein adducts found in both the high and lower molecular weight bands (Fig. 2, K, L, and M). Pretreatment with BSO tended to increase the formation of 4-HNE-protein adducts in embryos exposed to either dose of HU.

Effects of BSO Pretreatment on the AP-1 c-Fos Heterodimer DNA Binding Activity Induced by HU. We reported previously that maternal exposure to HU (400 or 600 mg/kg) induced c-Fos heterodimer-dependent AP-1 DNA binding activity in the embryo (Yan and Hales, 2005). Our goal here was to determine the impact of BSO pretreatment on HU-induced activation of c-Fos binding activity. The relative binding activity of c-Fos heterodimers in the embryo was enhanced by exposure to HU alone (400 or 600 mg/kg) at 3 h in a dose-dependent manner (Fig. 3). BSO alone did not influence c-Fos DNA binding activity. Furthermore, pretreatment with BSO had no effect on the extent to which HU exposure induced the activation of AP-1 c-Fos dimers in embryos (Fig. 3).

Effects of BSO Pretreatment on HU Embryotoxicity. Exposure to BSO alone on GD 9 did not affect progeny outcome, as assessed on GD 18 by fetal mortality, live fetal body weights, and external or skeletal abnormalities (Fig. 4, A–D). Exposure to 400 mg/kg HU alone was not embryo-lethal or teratogenic but did induce a reduction in fetal weights (Fig. 4B). In contrast, exposure to high-dose HU (600 mg/kg) increased the incidence of fetal deaths, external malformations, and skeletal deformities, in addition to causing growth retardation. BSO pretreatment did not affect the incidence of fetal mortality or growth retardation induced by either dose of HU (Fig. 4, A and B). Dead fetuses were observed as embryo resorptions, with the exception of two fetuses in the 400 mg/kg HU group and one fetus in the 600 mg/kg HU group, which appeared as late fetal deaths.

Although BSO pretreatment did not significantly increase the overall incidence of malformed fetuses per litter, a 2.4-fold increase in the percentage of malformed fetuses per litter was observed after pretreatment with BSO in the low-dose HU (400 mg/kg) group and a 1.4-fold increase was observed among litters exposed to high-dose HU (600 mg/kg) (Fig. 4C). However, both the incidence and the spectrum of specific external malformations were enhanced significantly by BSO pretreatment of HU-exposed dams (Tables 1 and 2).

No external malformations were apparent in control fetuses (Fig. 5A). One fetus with hydrocephaly (1 of 63) was observed among the litters exposed to BSO alone (Fig. 5D). Tail defects (10 of 108) were observed among the fetuses exposed to low-dose HU alone (400 mg/kg) (Table 1; Fig. 5B); the combination of BSO with this dose of HU resulted in fetuses with curly tail, hind limb malformations (Fig. 3E), hydrocephaly, exencephaly, open eye, spina bifida, and gastroschisis (Table 1). Tail and hind limb abnormalities predominated in the fetuses exposed to 600 mg/kg HU either alone or with BSO pretreatment (Fig. 5, C and F). Low incidences of hydrocephaly, spina bifida, and open eye defects were found in the 600 mg/kg HU group; forelimb oligodactyly, spina bifida, and gastroschisis were observed among the fetuses exposed to 600 mg/kg HU and BSO (Table 1). BSO pretreatment particularly enhanced the incidence of hind limb defects induced by HU. Although no hind limb defects were observed in the group exposed to low-dose HU alone, 11.5% were observed after exposure to 400 mg/kg HU in combination with BSO. BSO pretreatment also increased the incidence of hind limb defects in the 600 mg/kg high-dose HU treatment group, from 31.0 to 68.9%. In addition, three fetuses with forelimb defects (ectrodactyly or hemimelia) (Table 1; Fig. 5, F and J) were observed uniquely in the litters exposed to BSO and 600 mg/kg HU.

Strikingly, the hind limb defects observed occurred mainly at the first digit and included agenesis, truncation, or displacement. Defects (aplasia/hypoplasia) of more than the first digit were observed at a very low frequency, in one fetus (1.7%) in the group treated with high-dose HU alone and six fetuses (9.8%) in the group exposed to this dose of HU with BSO (Table 1; Fig. 5K). Interestingly, in the fetuses exposed to BSO in combination with 400 mg/kg HU, three cases of tibial polydactyly were observed as an extra digit at the great toe, in addition to the predominant aplasia or hypoplasia at the first toe (Fig. 5, E and I).

No apparent skeletal abnormalities were observed in vehicle controls (Fig. 6A). Similarly to the external malformations, BSO pretreatment exaggerated the skeletal malformations induced by HU. In the group treated with 400 mg/kg HU, the skeletal malformation rate per litter increased 2.3-fold when the dams were pretreated with BSO (Fig. 4D); exposure to 600 mg/kg HU in combination with BSO increased the rate of skeletal deformities in live fetuses from 78.8 to 100%. The predominant skeletal abnormalities observed consisted of lumbarsacral vertebral and hind limb defects (Table 2). The vertebral defects were partial ossification, fusion, or misalignment of centra or misalignment of the vertebral arch; more severe deformities were found in the fetuses exposed to HU (400 or 600 mg/kg) in combination with BSO (Fig. 6, B, C, E, and F). The hind limb defects occurred mainly at the anterior axis, displayed as ectrodactyly at the first digit and tibial aplasia or hypoplasia (Fig. 6); in the few instances of forelimb defects, radius agenesis occurred (Table 2). Low frequencies of tail malformations (aplasia/hypoplasia), sternal defects (partial ossification or misalignment), or forked ribs were also observed in the HU-treated groups in the presence or absence of BSO (Table 2); pretreatment with BSO did not influence either the incidence or the severity of these defects.

Discussion

In the organogenesis-stage embryo, the inhibition of GSH synthesis with BSO induced oxidative stress as assessed by the GSSG/GSH ratio, the formation of 4-HNE protein adducts, and the activation of redox-sensitive transcription factors such as AP-1. Interestingly, BSO enhanced HU teratogenicity without affecting fetal mortality or weights. The effects of the combination of BSO and HU were region-specific in the embryo, both with respect to the malformations that resulted and the localization of 4-HNE protein adduct immunoreactivity. Interestingly, region-specific 4-HNE-protein adducts were found in early organogenesis-stage embryos even in the absence of an exposure to exogenous chemicals; these adducts were localized mainly to the neural tube, otic pit, branchial arch, mid-gut, and caudal region of the tail. These “naturally occurring” 4-HNE protein adducts were found predominantly in the cytoplasm. Increases in the formation of 4-HNE adducts may indicate the presence of higher levels of free radicals in these regions. Alternatively, differences in the composition of membrane phospholipids may alter susceptibility to the induction of lipid peroxidation. Finally, region-specific differences in the capacity of the embryo to detoxify 4-HNE by conjugating it with GSH, catalyzed by the glutathione S-transferases, may be important in detoxifying 4-HNE in target cells (Awasthi et al., 2005).

The impact of 4-HNE on the fate of the cell is dose-dependent; at low levels, 4-HNE promotes cell proliferation, whereas at higher concentrations, it induces cell cycle arrest, differentiation, and finally apoptosis (Awasthi et al., 2005). Exposure to 4-HNE induces vascular smooth muscle growth, accompanied by the activation of MAPKs (extracellular signal-regulated kinase, c-Jun NH₂-terminal kinase, and p38), the induction of c-fos and c-jun gene expression, and AP-1 DNA binding activity (Kakishita and Hattori, 2001). The generation of 4-HNE in specific regions of the embryo may suggest the involvement of 4-HNE in the regulation of proliferation, differentiation, and apoptosis during normal development.

4-HNE regulates cell signal cascades mainly due to the formation of 4-HNE-protein adducts (Leonarduzzi et al., 2004). Our Western blot data show that there are two predominant 4-HNE-protein adduct bands in the 100- and 150-kDa regions. Characterization of the identities of the proteins in these bands would help to elucidate the pathways regulated by 4-HNE during normal organogenesis. In the retina, triosephosphate isomerase, α-enolase, heat shock cognate 70, and βB2 crystallin were identified as proteins that were frequently modified and had the highest molar content of 4-HNE (Kapphahn et al., 2006), whereas heat shock protein 90 was found to be consistently modified by 4-HNE in the liver of alcohol-treated rats (Carbone et al., 2005). 4-HNE-protein adducts influence a variety of signal cascades, including signals for limb patterning such as the MAPKs (Kawakami et al., 2003; Zuzarte-Luis et al., 2004). Although the possibility of increased membrane leakage by oxidative damage cannot be excluded, an increase in the nuclear localization of 4-HNE with the increase in HU doses suggests that 4-HNE may transduce signals from the membrane to nuclear compartment. Identification of the proteins conjugated with 4-HNE should help to elucidate the underlying embryopathy of limb malformations.

The tissue-specific localization of 4-HNE protein adducts, either as part of a signal pathway during normal development or as a response to oxidative stress during abnormal development, may indicate that these specific regions are more susceptible to oxidant insults. We propose that the specific external and skeletal malformations induced by exposure to HU are related to the localized increases in 4-HNE immunoreactivity that we observed in the somites (vertebral column and hind limb defects), the caudal end of the neural tube (tail defects), and the optic vesicle (open eye). Intense 4-HNE-protein adduct immunofluorescence was also observed in nucleated blood cells, suggesting that hematopoietic cells are highly susceptible to oxidative stress during the switch from glycolysis to aerobic metabolism. Interestingly, glucose-6-phosphate dehydrogenase, important in generating the reduced NADPH needed to maintain GSH in its reduced form, is essential for the establishment of blood circulation at this stage of development (Longo et al., 2002); a deficiency in glucose-6-phosphate dehydrogenase enhances the sensitivity of organogenesis-stage embryos to oxidative stress (Nicol et al., 2000).

One of the mechanisms by which oxidative stress may have differential effects, depending on the endpoint assessed, is via the regulation of redox sensitive transcription factors such as AP-1. Oxidative stress regulates the activation of AP-1 through a variety of mechanisms (Abate et al., 1990; Hirota et al., 1997). We report that a BSO pretreatment, which decreased GSH and increased the formation of 4-HNE protein adducts, did not enhance the effects of HU on c-Fos-dependent AP-1 DNA binding activity in mouse embryos. This is interesting because 4-HNE has been reported to interfere with the activities of protein kinases that regulate AP-1 activity, namely MAPKs, such as c-Jun NH₂-terminal kinases, p38, and extracellular signal-regulated kinase1/2 (Leonarduzzi et al., 2004). In studies with vascular smooth muscle cells, the effects of 4-HNE on the expression of AP-1 constituents was concentration- and AP-1 family member-specific (Kakishita and Hattori, 2001). Exposure of these cells to 4-HNE concentrations ≤2.5 μM dramatically induced c-fos mRNA expression without influencing that of c-jun; however, exposure to 4-HNE concentrations from 2.5 to 10 μM resulted in an increase in c-jun expression and a decrease in c-fos mRNA concentrations (Kakishita and Hattori, 2001). Concomitantly, AP-1 DNA binding activity peaked after exposure to 2.5 μM 4-HNE, then declined with increases in 4-HNE concentration (Kakishita and Hattori, 2001). We have reported previously that HU-induced c-Fos immunoreactivity in the GD 9 mouse embryo is localized in the brain region, branchial arch, somites, neural tube, heart, and areas around the blood cells in embryos (Yan and Hales, 2005). Here, we demonstrate that 4-HNE protein adduct immunoreactivity is detected in most of the same regions.

There is extensive interest in the development of strategies to protect the conceptus against oxidative stress-mediated insults during organogenesis. Although maternal dietary antioxidant supplementation has clearly been successful in improving fetal outcomes in animal models of experimental diabetes (Cederberg et al., 2001) or after exposure to specific teratogens, including HU (DeSesso, 1981; Wells et al., 2005), in some instances, high doses may be pro-oxidative and enhance adverse effects, such as tumorigenesis in p53 null mice (Chen and Wells, 2006). If 4-HNE plays an important role in mediating the toxicity of reactive oxygen species in the embryo, an alternate approach would be to enhance 4-HNE detoxification by inducing glutathione and the glutathione S-transferases; there is evidence that this approach is effective in the protection by 3H-1,2-dithiole-3-thione of cultured cardiomyocytes against 4-HNE (Li et al., 2005). If, as suggested above, 4-HNE plays a role in mediating the effects of reactive oxygen species by activating MAPKs, such as p38 MAPK, treatment with a MAPK inhibitor may protect the embryo against insult. Interestingly, inhibition of p38 MAPK rescued hematopoietic stem cells from reactive oxygen species-induced defects in repopulating capacity (Ito et al., 2006).

Clearly, both the formation of 4-HNE protein adducts and the induction of c-Fos immunoreactivity represent responses of the embryo to insult. We propose that such localized stress responses play a role in determining the pattern of malformations induced by exposure of the embryo to a teratogen.