The Journal of Nutrition Vol. 128 No. 7 July 1998, pp. 1239-1246

Zinc Pretreatment Inhibits Isotretinoin Teratogenicity and Induces Embryonic Metallothionein in CD-1 Mice^1,2

Danielle Blain^*, Stan Kubow^*, and Hing Man Chan^{*, , 3}

^* School of Dietetics and Human Nutrition,  Centre for Indigenous Peoples' Nutrition and Environment, Macdonald Campus of McGill University, Ste. Anne de Bellevue, Quebec, Canada H9X 3V9

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Isotretinoin (ITR), a teratogen in many species, is associated with increased oxidative stress. Metallothionein (MT) is an important tissue antioxidant whose concentrations are induced by zinc. To study the role of supplemental Zn as an inducer of embryonic MT, we injected pregnant CD-1 mice subcutaneously with saline vehicle, or 20 or 40 mg/kg Zn on gestational day (GD) 6.5. After 48 h, embryonic MT concentrations increased in a dose-related manner (r = 0.64, P < 0.05) with Zn treatment. The possible protective role of Zn pretreatment against ITR teratogenicity was investigated in vivo and in vitro. CD-1 mice were pretreated with saline or Zn (20 and 40 mg/kg) on GD 8.5 and 9.5. ITR was administered to both groups of mice via three intragastric intubations of 100 mg ITR/kg at 4 h intervals on GD 10.5. On GD 18.5, Zn pre-treated mice demonstrated decreased ITR-mediated growth retardation, cleft palates and postpartum mortality. A reduction in embryonic MT concentrations was observed in mice exposed to ITR. Mouse embryos cultured on GD 8.5 with an addition of 15 µmol/L Zn for 48 h had a sixfold greater MT concentration (688 µg/g protein) than controls. The Zn pretreatment of cultured embryos prevented malformations and lessened growth retardation caused by 24 h exposure to 17 µmol/L ITR. These results suggest that Zn-mediated induction of MT in mouse embryos could protect against ITR teratogenicity.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Isotretinoin (13-cis-retinoic acid, ITR),⁴ prescribed for the treatment of severe inflammatory, cystic and recalcitrant acne is a proven human teratogen when used at therapeutic doses during early pregnancy (Canadian Pharmaceutical Association 1994). ITR-induced abnormalities include craniofacial, cardioaortic, central nervous system and skeletal birth defects, as well as thymic hypoplasia and parathyroid deficiency (Canadian Pharmaceutical Association 1994). Abnormalities observed in several mouse strains include craniofacial and limb abnormalities, as well as skeletal alterations (Creech-Kraft and Juchau 1992, Yuschak and Gautieri 1993). Embryo culture studies indicate a lower viability of mouse embryos and a high incidence of craniofacial, limb and tail malformations after ITR exposure (Ritchie and Webster 1991).

The teratogenic activity of ITR may be related in part to its ability to generate toxic oxygen free radicals (OFR) (Davis et al. 1990, Kubow 1992). Davis et al. (1990) found that cultures of chick neural crest cells incubated with ITR were associated with an increased liberation of reactive oxygen species including superoxide anion (O₂^·), hydrogen peroxide (H₂O₂) and hydroxyl radical (^·OH). Addition of the scavenging enzymes superoxide dismutase and catalase to the culture medium reduced the quantity of OFR generated and increased cell viability. OFR can oxidize important macromolecules, leading to metabolic and structural modifications that can ultimately cause cell death. These modifications include the following: lipid peroxidation, protein denaturation and cross-linking, enzyme inhibition, DNA strand scission, base modifications and mutations, and changes in cell surface receptor and cell permeability (Freeman and Crapo 1982).

Metallothionein (MT) is a single-chain polypeptide of low molecular weight (6000-8000 Da), containing 61 amino acids with 20 cysteine residues. In animals, the protein is most abundant in parenchymal tissues, i.e., liver, kidney, pancreas and intestine (Kagi and Schaffer 1988). Functions of MT include homeostasis of essential minerals such as Zn and Cu and detoxification of heavy metals such as Cd and Hg (Cherian and Chan 1993). Because of the high number of thiol groups that MT contains, the protein is involved in cellular defense against oxidative stress by acting as a free radical scavenger (Templeton and Cherian 1991). Studies performed with different models such as cell lines (Greenstock et al. 1987, Mello-Filho et al. 1988, Ochi 1988), rat liver slices (Chan and Cherian 1992) or live mice (Matsubara 1987) exposed to various types of oxidative stress have shown that Zn pretreatment significantly induced MT expression and decreased toxicity.

The role of embryonic MT in normal and abnormal development, however, has not been studied. Because of its ability to scavenge reactive oxygen intermediates (Sato and Bremner 1993), embryonic MT may play an important protective role against teratogenic compounds known to induce OFR. Although induction of MT mRNA has been shown at different stages of mouse embryonic development (Andrews et al. 1991), the effect of Zn pretreatment on enhancement of embryonic MT content and on the prevention of the toxicity of teratogenic agents has not been previously explored.

This investigation was designed to study the following in CD-1 mice: 1) whether midgestational mouse embryos are capable of MT synthesis after Zn treatment either in vivo or in whole-embryo culture, 2) whether exposure to ITR affected embryonic MT concentrations indicative of ITR-induced oxidative stress; and 3) the possible role of Zn treatment and embryonic MT induction in protection against the teratogenic effects of ITR. The hypothesis of this study was that Zn-induced embryonic MT expression plays a protective role in the growth and development of midgestational mouse embryos exposed to ITR both in vivo and in vitro.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Chemicals.  ZnCl₂ (Sigma Chemical, St. Louis, MO) was kept in a desiccator, at room temperature, during the time of the experiments. ITR (isotretinoin, 13-cis retinoic acid, Sigma Chemical) was stored at 20°C in the dark in a sealed, lightproof container. ITR was handled in subdued light. 1,2-Phenylenediamine dihydroxide and hydrogen peroxide were purchased from Sigma Chemical. Goat anti-rabbit immunoglobulin G (IgG) and horseradish peroxidase conjugated Avidin D were obtained from Vector Lab (Burlingame, CA). All other chemicals were of reagent grade and were obtained from commercial suppliers.

Animal maintenance and mating procedure.  CD-1 mice (20-22 g) (Charles River Canada, St. Constant, Canada) were housed in a temperature-controlled room (24 ± 1°C) with a 12-h light:dark cycle (light from 0800 to 2000 h). The mice were kept in plastic shoe box cages with Beta-Chip bedding (Beta-Chip, Northeastern Products, Warrensburg, NY). The mice had free access to Purina Mouse Chow (Ren's Feed and Supply, Oakville, Canada) and tap water. Zn concentration was 10 µg/g in the feed and <77 µmol/L in drinking water. Proximate composition of the feed as stated by the manufacturer was 22% protein, 4% fat and 5% fiber. Females were housed three or four per cage and males were caged individually. For mating purposes, three or four females were housed at random with a male from 0900 to 1100 h. The presence of a vaginal plug indicated insemination, and this day was designated as gestational day (GD) 0.5.

Animal treatments.  The investigation was conducted in two parts. Experiments 1 through 3 were conducted to test the hypothesis that midgestational embryos are capable of MT synthesis after Zn exposure. Embryonic MT was assessed after exposure to Zn in vitro, in vivo and in vivo followed by a 48-h whole-embryo culture period. Experiments 4 through 6 studied the protective effect of Zn supplementation during mouse gestation against subsequent ITR exposure. Embryonic growth and development was assessed following exposure to both Zn and ITR in vitro, both Zn and ITR in vivo, and Zn in vivo and ITR in vitro. All experiments were conducted in conformance with the guidelines for experimental procedures set forth by McGill University and by the Canadian Council on Animal Care (1984).

Experiment 1. Modulation of embryonic MT and tissue Zn concentrations after in vitro Zn and ITR treatment.  The responsiveness of tissue Zn and MT in embryos and in yolk sacs to Zn or ITR exposure was assessed by whole-embryo culture. On GD 8.5, five pregnant mice were killed by cervical dislocation and embryos containing four to six somites were randomly allocated to five culture groups as follows: 1) 24 h with saline (control) (n = 4); 2) 24 h with Zn (15 µmol/L) (n = 4); 3) 48 h with saline (control) (n = 9); 4) 48 h with Zn (15 µmol/L) (n = 4); and 5) 48 h with ITR (17 µmol/L) (n = 9). At the end of the culture period, embryonic and yolk sac protein, and MT and Zn concentrations were measured. The intrinsic Zn level of the culture medium was 40 µmol/L. The choice of Zn and ITR doses was based on results of preliminary experiments. The additional 15 µmol/L of Zn in the culture medium, yielding a final concentration of 55 µmol/L, was adequate to modulate MT levels in the embryos. Although ITR is reported to induce birth defects in vitro at a threshold concentration of 1.7 µmol/L (Ritchie and Webster 1991), our preliminary results showed that a threshold concentration of 17 µmol/L ITR was required to induce major teratogenic effects in CD-1 mouse embryos, thus the dosage of 17 µmol/L ITR was selected.

Experiment 2. Induction of embryonic MT by in vivo Zn treatment.  The ability of supplemental Zn to induce embryonic MT in vivo was tested, and the optimal dosage of Zn for induction of embryonic MT was determined in Experiment 2. Concentrations of 20 and 40 mg Zn/kg, administered 48 h before MT measurement were used in this study to investigate their potency on embryonic MT induction.

On GD 6.5, 12 pregnant mice were randomly allocated to three Zn concentration groups and were injected subcutaneously with 0 (saline) (n = 4), 20 (n = 4) or 40 (n = 4) mg Zn/kg body mass. On GD 8.5, mice were killed, and protein and MT concentrations were measured in embryos containing four to six somites and their Reichert's membrane. Retained Zn was measured in decidua and maternal liver.

Experiment 3. Comparison of the effects of embryo culture vs. in vivo Zn treatment on MT and Zn concentrations in embryos and yolk sacs.  To assess the effect of culturing embryos on tissue MT concentrations relative to in vivo Zn treatment, eight pregnant mice were randomly allocated to two concentration groups on GD 6.5 and were injected subcutaneously with 0 (saline) (n = 4) or 40 (n = 4) mg Zn/kg body mass. On GD 8.5, half of the mice from each group (n = 4) were killed, and the embryos containing four to six somites were cultured for 48 h. The remaining mice were kept alive for the same period of time. On GD 10.5, both cultured and in vivo embryos were recovered; embryonic and yolk sac protein, and MT and Zn concentrations were measured. Four experimental groups were thus obtained: saline treatment + embryo culture (S + EC), Zn treatment + embryo culture (Zn + EC), saline treatment + in vivo (S + IV) and Zn treatment + in vivo (Zn + IV).

Experiment 4. Protection by Zn after in vitro ITR treatment.  To evaluate the protective role of Zn on ITR teratogenicity in the absence of maternal tissues, 16 pregnant mice were killed on GD 8.5, and embryos of four to six somites were randomly allocated to the pretreatment groups. The Zn-treated embryos (n = 37) were exposed to Zn (15 µmol/L additional to the 40 µmol/L intrinsic in the medium) for 24 h; the control embryos (n = 40) were exposed to saline under the same conditions. On GD 9.5, embryos were cultured for 24 h with either saline (control) or 17 µmol/L ITR treatments. At the end of the culture period, embryos were examined for growth and development, and embryonic and yolk sac protein and MT and Zn concentrations were measured.

Experiment 5. Protection by Zn after in vivo ITR treatment.  To examine the protective role of supplemental Zn on in vivo ITR teratogenicity, eight pregnant mice were randomly allocated to pretreatment groups. On GD 8.5 and 9.5, the Zn-treated mice (n = 4) received subcutaneous injections of 40 and 20 mg Zn/kg body mass, respectively; the control mice (n = 4) were injected with saline. On GD 10.5, all animals received three intragastric intubations of 333 µmol ITR/kg body mass, in the morning (1100 h), in the afternoon (1500 h) and in the evening (1900 h). This dose regime was selected on the basis of a treatment regime previously used with CD-1 mice to induce ITR teratogenicity (Kochhar and Penner 1987, Kubow 1992). On GD 18.5, mice were killed and the fetuses recovered and examined for growth and development.

Experiment 6. Effect of in vivo Zn pretreatment on in vitro ITR teratogenicity.  To study the possible protection of in vivo Zn pretreatment against the teratogenic effects of ITR on cultured embryos, 20 pregnant mice were randomly allocated to pretreatment groups. On GD 6.5 and 7.5, the Zn-exposed mice (n = 12) received subcutaneous injections of 40 and 20 mg Zn/kg body mass, respectively; the control group (n = 8) was injected with saline. On GD 8.5, mice were killed, and the embryos of four to six somites from each litter were exposed in vitro to either 17 µmol/L ITR or to saline treatments. After 48 h in culture, embryos were examined for growth and development and embryonic protein, and MT and Zn concentrations were measured.

Whole-embryo culture.  The techniques we used for explanting and culturing embryos are described by New (1978). Mouse embryos were cultured in rat serum, obtained from exsanguinated male Sprague-Dawley rats. On GD 8.5, mice were killed by cervical dislocation and the uteri excised immediately. Using aseptic techniques, the decidua were dissected from uteri in Hank's balanced salt solution (Gibco, Burlington, Canada) under a stereomicroscope. Embryos were explanted by removing the decidual tissue and Reichert's membrane, and embryos containing four to six somites were chosen for culture. The ectoplacental cone, amnion and visceral yolk sac were left intact. Embryos were placed into 60-mL glass bottles containing 1.55 mL of warm sterile male rat serum per embryo. Streptomycin sulfate and penicillin G potassium were added to the medium (0.5%). For control treatment, sterile saline (9 g/L) was added for a final volume of 1.6 mL of culture medium per embryo. For Zn treatment, a stock solution of 69.4 mg ZnCl₂/L was prepared in sterile saline, and was added to the medium for a final concentration of 55 µmol/L. Isotretinoin was dissolved in 70% ethanol to prepare a stock solution of 1.25 g ITR/L and was added to the medium for a final concentration of 17 µmol/L. All solutions were prepared immediately before addition to the culture medium. The bottles were gassed with 5% O₂/5% CO₂/90% N₂ at the start of the culture, and with 20% O₂/5% CO₂/75% N₂ and 40% O₂/5% CO₂/45% N₂ at 18 and 26 h of culture, respectively. They were placed in a 37°C incubator and rotated at 30 rpm for either 24 or 48 h.

Developmental assessment of cultured embryos.  After culture, embryos were transferred in 9 g/L sterile saline solution and examined under a dissecting microscope. Yolk sac diameter, crown-rump length and head diameter of live embryos were measured by using an eyepiece micrometer. Somite number was also counted. Morphology was scored with the system described by Brown and Fabro (1981) in which 13 morphological criteria are each assigned a ranking score from zero to five, corresponding to a given stage of differentiation. The sum of the scores for each individual feature gives a total morphological score that reflects the stage of development of each embryo. For the purpose of determining malformations, the following features were considered abnormal for midgestational (GD 8-10) embryos: unfused neural tube, hindbrain, midbrain or forebrain; malrotated, kinked or ventral convex flexion of the tail; and missing forelimb. To calculate the mean number of malformations per embryo, the number of defects was counted separately; therefore, the sum of abnormalities per embryo may exceed one.

In vivo developmental assessment.  On GD 18.5, mice were weighed, killed by cervical dislocation and their uteri excised immediately. The number and mass of resorptions, the number of implantation sites, pup location and viability were recorded. Placentas were matched to the pups and weighed. Fetuses were weighed, sexed and examined for external malformations. Animals were examined for craniofacial defects, such as micrognathia, eye, ear or nose underdevelopment or malformation and cleft palate, and for skeletal defects, such as limb and tail underdevelopment or malformation. Postpartum mortality was recorded 2 h after fetus recovery.

ITR and Zn administration.  ITR was suspended in food-grade safflower oil (ICN Biochemicals, St. Laurent, Canada) at a concentration of 8 g/L. The suspension was sonicated and vortexed, and was immediately administered to mice according to their mass. Mice received three doses of 100 mg ITR/kg body weight. Total volume administered did not exceed 0.57 mL. Stock solutions of 9.38 and 18.75 mg ZnCl₂/mL sterile saline (9 g/L) were prepared and immediately injected subcutaneously in mice according to their weight. Mice thus received doses of 20 and 40 mg Zn/kg body weight. Total volume injected did not exceed 0.2 mL. Control mice received injections of sterile saline (9 g/L) only.

Protein analysis.  Embryos, Reichert's membranes and yolk sacs were homogenized in sterile saline (9 g/L) with a Polytron homogenizer (Brinkman, PT 3000, Mississauga, Canada). Total protein content was measured by using the standardized Bradford Microprotein Assay (Bradford 1976), with the Bio-Rad Assay Kit II (Bio-Rad, Hercules, CA).

Zn analysis.  After nitric acid digestion of tissue samples, Zn concentrations were measured by flame atomic absorption spectrophotometry (Hitachi, Polarized Zeeman AAS, Z-8200 Missisauga, Canada) with a mixture of air and acetylene.

Metallothionein analysis.  Total MT concentration in embryos, Reichert's membranes and yolk sacs was measured by using an ELISA with an isolated IgG fraction of a rabbit antiserum against rat liver Cd-MT-2 polymer, as described by Chan et al. (1992).

Statistical analysis.  Embryonic and yolk sac MT of 24-h culture groups in Experiment 1 and pregnancy outcome indices in Experiment 5 were compared with a one-tailed, unpaired t test to identify significant differences between the treatment groups. Yolk sac Zn in Experiment 3, as well as fetal and placental weights in Experiment 5, were compared with a one-tailed, unpaired, nested t test to identify a significant difference between treatments, nested for dams. To control for different dam behaviors and metabolic responses to injection, nested ANOVA was used to nest dams within injection treatments in the statistical analyses of Experiment 2 and 3. Growth parameters (Experiments 4 and 6), embryonic MT and Zn (Experiments 3-6), yolk sac MT (Experiments 3, 4 and 6) and Zn (Experiments 4 and 6), as well as mean number of malformations per embryo (Experiments 4 and 6), were compared with a two-way ANOVA to identify the main effect of pretreatment and treatment, and their interaction on these variables. When the variances were unequal, data were log-transformed. When a significant main effect was observed, a least-significant difference multiple comparison test to identify group mean differences was used. Frequency of abnormal embryos (Experiments 4 and 6) and frequency of cleft palates and postpartum mortality (Experiment 5) were compared with chi-square statistics. Nested analyses were conducted with version 6.04 (1992) of the Statistical Analysis System software (SAS Institute, Cary, NC). For all other analyses, we used SPSS software for Windows version 5.0 (SPSS, Chicago, IL). A probability of P < 0.05 was accepted as the level of significance unless stated otherwise.

	RESULTS

Abstract Introduction Methods Results Discussion References

Experiment 1. Modulation of embryonic MT and tissue Zn concentrations after in vitro Zn and ITR treatment.  Exposure to an addition of 15 µmol/L Zn for 48 h resulted in a significant increase in embryonic MT but no change in Zn concentrations (Fig. 1A, B). The mean MT concentration in yolk sacs of the Zn-treated group was not significantly different than that of the control group (Fig. 1A). The Zn content of the yolk sac, however, was significantly higher in the saline-treated group than in the Zn-exposed embryos (Fig. 1B). Embryonic MT levels of the ITR-treated groups (102 ± 32.3 µg/g protein) was not significantly different than MT concentrations of control embryos (55 ± 35.2 µg/g protein).

View larger version (24K): [in this window] [in a new window]   Fig 1. Metallothionein (MT) concentrations (panel A) and Zn contents (panel B) in mice embryos and yolk sacs after exposure to saline or 15 µmol/L Zn for 48 h. Each bar represents the mean ± SEM, n = 4. Embryo or yolk sac means without a common superscript letter are significantly different (P < 0.05).

Experiment 2. Induction of embryonic MT by in vivo Zn treatment.  Injection of dams with 20 mg Zn/kg did not affect levels of MT in the embryos and the Reichert's membranes (Fig. 2A, B). The 40 mg Zn/kg dose, however, significantly induced MT in both embryos and Reichert's membranes compared with the saline-injected group (Fig. 2A, B). Decidual tissue of all three groups contained similar amounts of Zn, ranging from 0.700 to 0.776 ng/decidua. Indicative that the treatments were properly administered was the fact that maternal liver Zn concentrations reflected the injected doses (Fig. 2C). The highest dose of Zn led to significantly greater liver Zn concentrations compared with the saline-injected group.

View larger version (28K): [in this window] [in a new window]   Fig 2. Metallothionein (MT) concentrations in mouse embryos (panel A) and Reichert's membranes (panel B), and Zn concentrations in the livers (panel C). Tissues were collected 48 h after the injection of 0, 20 or 40 mg Zn/kg body weight at gestational day 6.5. Each bar represents the mean ± SEM, n = 4. Means in a panel without a common superscript letter are significantly different (P < 0.05).

Experiment 3. Comparison of the effects of embryo culture vs. in vivo Zn treatment on MT and Zn concentrations in embryos and yolk sacs.  As in Experiment 2, Zn injection of 40 mg/kg on GD 6.5 resulted in a significantly greater mean embryonic MT concentration on GD 8.5 (Zn + IV) compared with saline treated dams (S + IV) (Fig. 3A). Culturing of the Zn-treated embryos for 48 h (Zn + EC) did not further increase MT concentrations relative to those observed in the Zn + IV embryos. In contrast, saline-treated embryos cultured for 48 h (S + EC) had significantly higher MT concentrations than S + IV embryos, producing MT concentrations not different from those in the Zn + EC group (Fig. 3A). The S + IV embryos had significantly higher embryonic Zn concentrations than those in the S + EC group (Fig. 3C). The fact that embryonic Zn concentrations were not different in the Zn- and saline-injected groups suggests that transfer of the metal to the embryo may be highly regulated.

View larger version (33K): [in this window] [in a new window]   Fig 3. Metallothionein (MT) concentrations in embryos (panel A) and yolk sacs (panel B), and Zn contents in mouse embryos (panel C). Four mice were injected with saline on gestational day (GD) 6.5. Embryos from two mice were harvested after 96 h (S + IV). Embryos from the other two mice were harvested after 48 h and cultured for 48 h (S + EC). Four mice were injected with 40 mg Zn/kg body weight on GD 6.5. Embryos from two mice were harvested after 96 h (Zn + IV). Embryos from the other two mice were harvested after 48 h and cultured for 48 h (Zn + EC). Each bar represents the mean ± SEM, n = 4. Means in a panel without a common superscript letter are significantly different (P < 0.05).

MT concentrations in yolk sacs were about one hundredth those of embryos. Yolk sac concentrations of MT in the Zn + IV group were significantly higher than those of the S + IV-treated mice (Fig. 3B). Mean yolk sac MT was also significantly higher in the Zn + IV group than in the Zn + EC and S + EC groups. The S + EC and S + IV groups did not differ in yolk sac MT concentration. Yolk sac Zn concentrations for the S + EC (6.3 ± 0.3 nmol/yolk sac) and the Zn + EC groups (6.7 ± 0.4 nmol/yolk sac) were not different. Yolk sac Zn concentrations in the two in vivo groups could not be measured because of insufficient sample.

Experiment 4. Protection by Zn after in vitro ITR treatment.  Growth variables in terms of head diameter and somite number did not differ among the various treatments (Table 1). In contrast, the crown-rump length of the saline + ITR group was significantly lower than that of the Zn + saline group. The sum of morphological scores was significantly lower in both ITR-treated groups than in the Zn + saline group, whereas morphological scores for central nervous system development were significantly lower in the saline + ITR group than in the Zn + ITR treatment group. The score for flexion development was significantly lower in the saline + ITR group relative to the saline + saline and Zn + saline groups. The saline + ITR group also had lower morphological scores for forelimb development than the Zn + saline group.

The major embryonic malformations observed included central nervous system (CNS) defects (unfused neural tube, hindbrain, midbrain and/or forebrain), abnormal flexion, missing forelimb and enlarged pericardium. The saline + ITR treatment was associated with a higher mean number of defects per embryo compared with all other groups (Table 1). In terms of specific defects, Zn pretreatment before drug exposure was associated with a lower frequency of CNS defects (0%), and of abnormal flexion (10.5%), compared with the saline + ITR culture group (22 and 40% for frequencies of CNS defects and abnormal flexion, respectively).

The Zn + saline treatment resulted in a significantly higher embryonic MT concentration compared with the other treatments (Fig. 4). There were no significant differences in embryonic and yolk sac Zn concentrations among the treatment groups (data not shown). Yolk sac MT concentrations were 142 ± 37, 180 ± 53, 110 ± 44 and 161 ± 74 mg/g protein, for the saline + saline, Zn + saline, saline + ITR and Zn + ITR groups, respectively, and were not significantly different.

View larger version (18K): [in this window] [in a new window]   Fig 4. Metallothionein (MT) concentrations in embryos after saline or Zn pretreatment, followed by exposure to saline or isotretinoin (ITR). Experiment 4: on gestational day (GD) 8.5, embryos were cultured for 24 h with saline or 15 µM Zn. On GD 9.5, saline or 17 µmol ITR was added to the culture medium for the next 24 h. Experiment 6: dams were injected with saline or 40 and 20 mg Zn/kg body weight on GD 6.5 and 7.5, respectively. On GD 8.5, embryos were cultured with saline or 17 µmol ITR for 48 h. For both experiments, MT was measured on GD 10.5. Each bar represents the mean + SEM, n = 20. Means within each experiment without a common superscript letter are significantly different (P < 0.05).

Experiment 5. Protection by Zn after in vivo ITR treatment.  Pretreatment of dams with saline or Zn injections before ITR gavages did not affect pregnancy outcome, because both groups had comparable indices of maternal mass gain, implantation sites, live fetuses and resorptions (Table 2). Fetal (P < 0.001) and placental (P < 0.001) weights were greater, and the frequencies of cleft palate and postpartum mortality were significantly reduced when Zn was injected in dams before they were exposed to ITR (Table 2). Cleft palates were observed only in fetuses from mice who received saline injections before ITR treatments.

Experiment 6. Effect of in vivo Zn pretreatment on in vitro ITR teratogenicity.  Growth variables did not differ in the saline + saline and the Zn + saline groups (Table 3), whereas crownrump length, head diameter and somite number in the two ITR-treated groups were significantly lower than those of the control groups exposed to saline only. Similarly, the sum of scores as well as developmental scores for CNS, flexion and forelimb were significantly lower in the groups exposed to ITR compared with the saline control groups. In addition, Zn exposure in culture in the presence of ITR was associated with lower mean sum of morphological scores and a lower morphological developmental score for limb development relative to ITR exposure alone. The mean numbers of defects per embryo were significantly higher in both of the ITR-treated groups than in the control groups, regardless of a Zn pretreatment. These results indicate that considerable retardation of growth and morphological development resulted from in vitro drug exposure, regardless of in vivo Zn pretreatment.

Significantly lower embryonic MT concentrations were observed in the ITR-treated groups relative to the groups not exposed to ITR (Fig. 4). Embryonic Zn levels did not differ among treatment groups (data not shown).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

This study demonstrates that ITR teratogenicity, both in vivo and in embryo culture, was markedly diminished with Zn pretreatment. As reported previously (Kubow 1992), maternal treatment with ITR affected fetal and placental growth, increased postpartum fetal mortality and induced cleft palate. We have shown that Zn injections in dams before ITR exposure resulted in an improved fetal and placental growth, a reduced frequency of postpartum mortality and a complete inhibition of ITR-induced defects. Similar to previous observations (Ritchie and Webster 1991), exposure of cultured whole embryos to ITR produced dramatic effects on embryonic growth and development, including increases in the frequency of unfused midbrain, abnormal flexion, missing forelimb and enlarged pericardium. When embryos collected on GD 8.5 were pretreated with Zn for a 24-h culture period, embryonic growth and development were comparable to that seen in unexposed control groups. These findings indicate for the first time that Zn pretreatment can protect against the teratogenic action of exogenously administered agents.

The mechanism of Zn protection against the teratogenic action of ITR is not clear but may involve induction of embryonic MT. Although increased MT expression during embryonic development has been described, previous studies have reported MT induction primarily in its mRNA form (Andrews et al. 1991, De et al. 1990). MT induction, as detected in its protein form, has previously been measured only in GD 4 mouse embryos (Vidal and Hidalgo 1993) and in rat fetal liver on GD 16 (Charles-Shannon et al. 1981). The ability of GD 8-10 embryos to synthesize MT suggests that MT may protect against teratogenic insults during this period of high susceptibility to teratogenicity. The level of embryonic MT induction appeared to vary according to Zn dosage, the length of incubation time and the gestational day. The highest embryonic MT concentration was induced with in vitro Zn incubation for 48 h (Fig. 1A). Moreover, as shown in Experiment 3, cultured embryos from saline-treated dams had MT concentrations similar to those of the two Zn-treated groups (Fig. 3A), indicating that the process of embryo culture itself could induce MT, possibly due to stress (Oh et al. 1978).

As shown from the dose-response study (Experiment 2), relatively large in vivo doses of Zn were required to induce embryonic MT concentrations in vivo likely because of the blockage of Zn transfer to embryos by extra-embryonic tissues as suggested by other work (Taubeneck et al. 1994). Reis et al. (1988) also found that Zn concentrations in mouse conceptuses did not change during the first 10 d of gestation, despite Zn supplementation to dams before mating. In contrast to the in vivo findings, the effective dose of 15 µmol Zn/L, which induced an increase in embryonic MT concentrations and a protective effect on ITR teratogenicity in the cultured embryo studies, was surprisingly small. Because there were no extra-embryonic membranes to block embryonic Zn absorption in the cultures, most of the added Zn was available to the embryo and a small amount of Zn may have been needed to induce embryonic MT. These findings thus indicate that the mouse embryo is acutely sensitive to minute changes in zinc concentrations. In support of this contention, we observed an increase in embryonic MT concentrations after in vitro (Experiment 1) or in vivo (Experiment 4) dosing of zinc despite the lack of detectable changes in embryonic zinc concentrations. It is also possible that Zn accumulation in the embryo may be compartmentalized, and MT may be induced by an increase in Zn availability in some localized sites. To explore this latter possibility, immunochemistry studies are needed to investigate the localization of MT within the embryo.

Because reactive oxygen radical species may play a role in the in vitro and in vivo teratogenic effects of ITR (Davis et al. 1990, Kubow 1992), induction of embryonic MT by Zn may have offered protection against ITR teratogenicity by enhancing the scavenging capability of the embryos to scavenge free radicals. ITR, like other free radical-generating substances, may induce MT synthesis (Sato and Bremner 1993). However, we found that exposure to ITR decreased embryonic MT (Fig. 4). It is conceivable that induction of MT by ITR may have occurred at a time point beyond the 48-h period after ITR exposure when MT was measured in this experiment. Alternatively, the newly synthesized MT or the existing MT induced by Zn pretreatment may have been oxidized. The sequestration of free radicals by MT can cause thiolate oxidation (Fliss and Menard 1992), yielding oxidized MT that is not measurable by the ELISA used in the present study (Chan et al. 1992). The decrease in MT after ITR exposure may therefore indicate that MT via radical sequestration may be a possible mechanism for the protection of ITR toxicity. However, the protective effects of Zn may not necessarily involve MT because in vitro and in vivo supplementation of Zn can prevent excessive apoptotic cell death, which may not necessarily be mediated by MT (Sunderman 1995). The teratogenic action exerted by retinoids is mediated in part through induction of excessive programmed cell death, especially in areas involved in craniofacial and mesomelic chondrogenic development (Alles and Sulik 1989).

Our findings extend the proposed role of Zn in xenobiotic-induced teratogenicity suggested by Taubeneck et al. (1994). They proposed that a variety of agents exert teratogenic effects via an induction of maternal and placental MT. An induction of MT in extra-embryonic tissues could sequester Zn, making it unavailable for normal embryonic growth and development. In this paper, ITR exposure did not appear to induce Zn deficiency in the embryo. Moreover, concentrations of MT in the yolk sac tended to be reduced rather than enhanced by ITR treatment (Experiment 5). The lack of induction of MT in extra-embryonic tissues by ITR is consistent with the idea that ITR did not block embryonic Zn uptake. The protective action of Zn supplementation against in vivo and in vitro ITR teratogenicity, therefore, was unlikely to be mediated via an amelioration of a embryonic Zn deficiency. Yolk sac Zn content was unaffected in embryos exposed to ITR in culture, and morphological scores of yolk sac circulation, yolk sac diameter and allantois (data not shown) did not improve when ITR-treated embryos were exposed to supplemental Zn. Hence, the protective action of Zn on ITR teratogenicity did not appear to be mediated at the level of the yolk sac.

No significant protection against ITR was provided by maternal treatment with Zn before the explanting and culturing of the embryos. It is likely that a combination of factors may explain the observed lack of protection. The possibilities include the following: 1) less accessibility of Zn to the embryo in vivo where maternal transfer of the metal is highly regulated (Reis et al. 1988); 2) a lower degree of MT inducibility in embryos of dams treated with Zn in vivo earlier in gestation (GD 6.5 and 7.5); and 3) the protective effects offered by in vivo Zn pretreatment may be overridden by the higher teratogenic potency of the 48 h of in vitro ITR exposure.

In summary, we found that in vivo and in vitro Zn supplementation protects against ITR teratogenicity. Although the mechanisms involved require further investigation, an induction of embryonic MT could play a role because midgestational embryonic tissues induced an increase in MT concentration in response to Zn supplementation. The ability of Zn supplementation to induce embryonic MT could be exploited in investigating a protective role for MT against a variety of teratogenic insults, particularly those associated with an increase in oxidative stress.

	FOOTNOTES

Manuscript received 13 August 1997. Initial reviews completed 18 December 1997. Revision accepted 11 March 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Alles A. J., Sulik K. K. Retinoic-acid-induced limb-reduction defects: perturbation of zones of programmed cell death as a pathogenic mechanism. Teratology 1989; 40:163-171[Medline]
• Andrews G. K., Huet-Hudson Y. M., Paria B. C., McMaster M. T., De S. K., Dey S. K. Metallothionein gene expression and metal regulation during preimplantation mouse embryo development and MT mRNA during early development. Dev. Biol. 1991; 145:13-27[Medline]
• Bradford M. M. A rapid and sensitive method for quantitation of microgram quantities of protein utilising the principle of protein-dye binding. Anal. Biochem. 1976; 72:248-254[Medline]
• Brown N. A., Fabro S. Quantitation of rat embryonic development in vitro: a morphological scoring system. Teratology 1981; 24:65-78[Medline]
• Canadian Council on Animal Care (1984) Guide to the Care and Use of Experimental Animals, Vol. I and II. National Library of Canada, Ottawa, ON, Canada.
• Canadian Pharmaceutical Association (1994) Accutane. In: Compendium Of Pharmaceutical Products and Specialties, pp. 15-17. Canadian Pharmaceutical Association, Toronto, Ontario.
• Chan H. M., Cherian M. G. Protective roles of metallothionein and glutathione in hepatotoxicity of cadmium. Toxicology 1992; 72:281-290[Medline]
• Chan H. M., Pringle G., Cherian M. G. Heterogeneity of antibodies to metallothionein and development of a simple enzyme-linked immunosorbant assay. J. Biochem. Toxicol. 1992; 4:219-227
• Charles-Shannon V. L., Sasser L. B., Burbank D. K., Kelman B. J. The influence of zinc on the ontogeny of hepatic metallothionein in the fetal rat. Proc. Soc. Exp. Biol. Med. 1981; 168:56-61[Medline]
• Cherian, M. G. & Chan, H. M. (1993) Biological functions of metallothioneina review. In: Metallothionein III (Suzuki, K. T., Immura, N. & Kimura, M. eds.), Birkhauser Verlag, Basel, Switzerland. pp. 87-109.
• Creech-Kraft J., Juchau M. R. Correlations between conceptual concentrations of all-trans-retinoic acid and dysmorphogenesis after microinjections of all-trans-retinoic acid, 13-cis-retinoic acid, all-trans-RAG or retinol in cultured whole rat embryos. Drug Metab. Dispos. 1992; 20:218-225[Abstract]
• Davis W. L., Crawford L. A., Cooper O. J., Farmer G. R., Thomas D., Freeman B. L. Generation of radical oxygen species by neural crest cells treated in vitro with isotretinoin and 4-oxo-isotretinoin. J. Craniofac. Genet. Dev. Biol. 1990; 10:295-310[Medline]
• De S. K., Dey S. K., Andrews G. K. Cadmium teratogenicity and its relationship with gene expression in midgestation mouse embryo. Toxicology 1990; 64:89-104[Medline]
• Fliss H., Ménard M. Oxidant-induced mobilization of zinc from metallothionein. Arch. Biochem. Biophys. 1992; 293:195-199[Medline]
• Freeman B. A., Crapo J. D. Biology of disease-free radicals and tissue injury. Lab. Invest. 1982; 47:412-426[Medline]
• Greenstock C. L., Jinot C. P., Whitehouse R. P., Sargent M. D. DNA radiation damage and its modification by metallothionein. Free Radical Res. Commun. 1987; 2:233-239[Medline]
• Kagi J.H.R., Schaffer A. Biochemistry of metallothionein. Biochemistry 1988; 27:8509-8515[Medline]
• Kochhar D. M., Penner J. D. Developmental effects of isotretinoin and 4-oxo-isotretinoin: the role of metabolism in teratogenicity. Teratology 1987; 36:67-75[Medline]
• Kubow S. Inhibition of isotretinoin teratogenicity by acetylsalicylic acid pretreatment in mice. Teratology 1992; 45:55-63[Medline]
• Matsubara J. Alteration of radiosensitivity in metallothionein induced mice and a possible role of Zn-Cu-thionein in GSH-peroxidase system. Exp. Suppl. 1987; 52:603-612
• Mello-Filho A. C., Chubatsu L. S., Menehini R. V79 Chinese-hamster cells rendered resistant to high cadmium concentration also become resistant to oxidative stress. Biochem. J. 1988; 256:475-479[Medline]
• Ochi T. Effects of glutathione depletion and induction of metallothioneins on the cytotoxicity of an organic hyperoxide in culture mammalian cells. Toxicology 1988; 50:257-268[Medline]
• New D.A.T. Whole embryo culture and the study of mammalian embryos during organogenesis. Biol. Rev. 1978; 53:81-122 [Medline]
• Oh S. H., Deagen J. T., Whanger P. D., Weswig P. H. Biological function of metallothionein. V. Its induction in rats by various stresses. Am. J. Physiol. 1978; 234:E282-E285
• Reis B. L., Keen C. L., Lonnerdal B., Hurley L. S. Mineral composition and zinc metabolism in female mice of varying age and reproductive status. J. Nutr. 1988; 118:349-361
• Ritchie H., Webster W. S. Parameters determining isotretinoin teratogenicity in rat embryo culture. Teratology 1991; 43:71-81[Medline]
• Sato M., Bremner I. Oxygen free radicals and metallothionein. Free Radical Biol. & Med. 1993; 14:325-337[Medline]
• Sunderman F. W. Jr. The influence of zinc on apoptosis. Ann. Clin. Lab. Sci. 1995; 25:134-142[Abstract]
• Taubeneck M. W., Gaston G. P., Rogers J. M., Keen C. L. Altered maternal zinc metabolism following exposure to diverse developmental toxicants. Reprod. Toxicol. 1994; 8:25-40[Medline]
• Templeton D. M., Cherian M. G. Toxicological significance of metallothionein. Methods Enzymol. 1991; 205:11-24[Medline]
• Vidal F., Hidalgo J. Effect of zinc and copper on preimplantation mouse embryo development in vitro and metallothionein levels. Zygote 1993; 1:225-229[Medline]
• Yuschak M. M., Gautieri R. F. Teratogenicity of 13-cis-retinoic acid and phenobarbital sodium in vitro. Res. Commun. Chem. Pathol. Pharm. 1993; 82:259-278[Medline]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences