Japanese medaka (Oryzias latipes): developmental model for the study of alcohol teratology


  • Xueqing Wang,

    1. National Center for Natural Product Research, Environmental Toxicology Research Program, Research Institute of Pharmaceutical Sciences, Department of Pharmacology, School of Pharmacy, University of Mississippi, University, Mississippi
    Search for more papers by this author
  • Erin Williams,

    1. National Center for Natural Product Research, Environmental Toxicology Research Program, Research Institute of Pharmaceutical Sciences, Department of Pharmacology, School of Pharmacy, University of Mississippi, University, Mississippi
    Search for more papers by this author
  • Mary L. Haasch,

    1. National Center for Natural Product Research, Environmental Toxicology Research Program, Research Institute of Pharmaceutical Sciences, Department of Pharmacology, School of Pharmacy, University of Mississippi, University, Mississippi
    Current affiliation:
    1. USEPA Mid-Continent Ecology Division, Molecular and Cellular Mechanisms Research Branch, NRC Research Associate Senior Scientist, Duluth, MN 55804
    Search for more papers by this author
  • Asok K. Dasmahapatra

    Corresponding author
    1. National Center for Natural Product Research, Environmental Toxicology Research Program, Research Institute of Pharmaceutical Sciences, Department of Pharmacology, School of Pharmacy, University of Mississippi, University, Mississippi
    • National Center for Natural Product Research, Environmental Toxicology Research Program, RIPS, School of Pharmacy, 313 Fraser Hall, PO Box 1848, University, MS 38677
    Search for more papers by this author


BACKGROUND: Animal models are necessary to investigate the mechanism of alcohol-induced birth defects. We have used Japanese medaka (Oryzias latipes) as a non-mammalian model to elucidate the molecular mechanism(s) of ethanol teratogenesis. METHODS: Medaka eggs, within 1 hr post-fertilization (hpf) were exposed to waterborne ethanol (0–1000 mM) in hatching solution for 48 hr. Embryo development was observed daily until 10 days post-fertilization (dpf). The concentration of embryonic ethanol was determined enzymatically. Cartilage and bones were stained by Alcian blue and calcein, respectively and skeletal and cardiovascular defects were assessed microscopically. Genetic gender of the embryos was determined by PCR. Levels of two isoenzymes of alcohol dehydrogenase (Adh) mRNAs were determined by semi-quantitative and real-time RT-PCR. RESULTS: The concentration of ethanol required to cause 50% mortality (LC50) in 10 dpf embryos was 568 mM, however, the embryo absorbed only 15–20% of the waterborne ethanol at all ethanol concentrations. The length of the lower jaw and calcification in tail fin cartilaginous structures were reduced by ethanol exposure. Active blood circulation was exhibited at 50+ hpf in embryos treated with 0–100 mM ethanol; active circulation was delayed and blood clots developed in embryos treated with 200–400 mM ethanol. The deleterious effects of ethanol were not gender-specific. Moreover, ethanol treatment was unable to alter the constitutive expression of either Adh5 or Adh8 mRNA in the medaka embryo. CONCLUSIONS: Preliminary results suggested that embryogenesis in medaka was significantly affected by ethanol exposure. Phenotypic features normally associated with ethanol exposure were similar to that observed in mammalian models of fetal alcohol syndrome. The results further indicated that medaka embryogenesis might be used as an alternative non-mammalian model for investigating specific alterations in gene expression as a means to understand the molecular mechanism(s) of ethanol-induced birth defects. Birth Defects Res (Part B) 77:29–39, 2006. © 2006 Wiley-Liss, Inc.


Ethanol is a teratogen and induces birth defects in humans. One of the most studied suite of birth defects due to maternal ethanol consumption during pregnancy is fetal alcohol syndrome (FAS) observed in some newborns of alcohol abusing women. A baby born with FAS is affected over an entire lifetime and this leads to increased health care costs (Lupton et al., 2004) with considerable significance in terms of economic and emotional cost to society. Although alcohol has been established as a teratogen and significant efforts have been made to educate women about the toxic effects of alcohol during pregnancy, the incidence of FAS remains unchanged (Cudd, 2005). Studies by the Centers for Disease Control and Prevention (CDC) showed that in the United States, FAS rates range from 0.2–1.5 per 1000 live births in different areas of the country (http://www.cdc.gov/ncbddd/fas/fasask.htm). Moreover, alcohol-related birth defects (ARBD) or alcohol-related neurodevelopmental disorders (ARND) are predominant in babies of alcoholic mothers. In the Western world, the combined rate of FAS and ARND are 9.1 per 1000 live births (Stromland and Pinazo-Duran, 2002). Recently, the term fetal alcohol spectrum disorder (FASD) was used to describe all these physical, mental, or behavioral disorders occurring due to prenatal ethanol exposure (Riley and McGee, 2005). Human studies of FAS are very limited because of ethical constraints. however, several non-human vertebrate and invertebrate animal models including C57BL/6J mice, Caenorhabditis elegans, Drosophila, and zebrafish (Danio rerio) have been utilized successfully and have provided major contributions to our understanding of FAS (Cudd, 2005; Sulik, 2005). The choice of animal model is very important in FAS studies, because prenatal alcohol exposure causes damage to the embryo by multiple mechanisms, depending on dose, pattern, and time of exposure (Cudd, 2005).

Fish models, particularly zebrafish and Japanese medaka, are currently emerging as an alternative to mammalian models, because of their easy availability, low maintenance cost, short life cycle, and their amenability to the study of gene function (Furutani-Seiki and Wittbrodt, 2004). Fertilization and embryonic development (fertilized eggs and larvae) in these fish is external (develop directly in the outside environment), eliminating maternal/placental fetal influences. Moreover, embryos are covered by a transparent egg chorion that allows direct observation of the development without interfering with the normal embryogenesis. Large-scale mutagenic screens in zebrafish and medaka have identified several orthologue genes that are related closely to those involved in human genetic diseases (Naruse et al., 2004). The response of zebrafish embryos to ethanol has been studied by many investigators and zebrafish have been recognized as a useful model for the study of alcohol-induced teratogenic effects (Bilotta et al., 2004; Carvan et al., 2004; Lockwood et al., 2004; Reimers et al., 2004a). In zebrafish, ethanol causes cyclopia, affects visual functions, induces pericardial edema and otolith defects, lowers heart rate, reduces eye diameter, induces axial malformations, delays development, and produces axial blistering (Bilotta et al., 2002, 2004; Reimers et al., 2004a). In addition, ethanol exposure has been shown to alter gene expression in the ventral aspects of the fore- and mid-brain (Blader and Strahle, 1998). Furthermore, in the zebrafish embryo, ethanol induces heat shock proteins (Krone et al., 1997; Lele et al., 1997), produces developmental abnormalities of the notochord and spinal cord, and malformation of the body trunk (Baumann and Sander, 1984). The effects of ethanol were found to be strain-dependent in zebrafish (Dlugos and Rabin, 2002). Interestingly, ethanol exposure in zebrafish has been shown to modulate aggression and other social behavior (Gerlai, 2003; Lockwood et al., 2004).

Although zebrafish have been established as a suitable model to study mechanisms of ethanol toxicity, development of another fish model for comparative studies provides increased confidence and opportunities for advancement. Comparisons between species are informative with regard to investigating gene function and genome evolution. Japanese medaka, the Far East cousin of zebrafish (Wittbrodt et al., 2002), is rapidly emerging as an important model fish with regard to developmental biology and toxicology; these two species (zebrafish and Japanese medaka) separated from a common ancestor about 140 million years ago (Wittbrodt et al., 2002). Moreover, the spectrum and phenotypes of induced mutations in medaka is often different from those found in zebrafish (Ishikawa, 2000; Loosli et al., 2000; Naruse et al., 2004). In addition to genome size advantages between medaka and zebrafish (800 Mb in medaka and 1700 Mb in zebrafish), medaka is the only known non-mammalian vertebrate where the gender determining gene Sry and DMY were identified and characterized (Matsuda, 2005). Therefore, medaka may be an ideal alternative non-mammalian animal model that provides another opportunity for investigating mechanisms of ethanol toxicity. The aim of the present study was to identify measurable phenotypic features induced by ethanol during embryogenesis of medaka. Our results indicate that medaka embryos are more resistant to ethanol than zebrafish and only partially absorbs the ethanol from the culture media. Further, as in other models, ethanol delays developmental processes and induces structural and functional abnormalities in several organ systems, the most significant of which are cardiovascular and craniofacial changes with parallels to human developmental disorders. Moreover, expression pattern of mRNAs of alcohol metabolizing enzymes remained unaltered by ethanol in Japanese medaka embryogenesis.


The Institutional Animal Care and Use Committee (IACUC) of the University of Mississippi (UM) approved all the experimental protocols.

Experimental Procedure

The maintenance and collection of fertilized embryos of medaka was carried out as previously described (Dasmahapatra et al., 2005). In brief, adult male and female medaka (3–4 months old: breeding) were maintained at 25°C in balanced salt solution (BSS, 17 mM NaCl, 0.4 mM KCl, 0.3 mM MgSO4, 0.3 mM CaCl2) with standard diet and a 16 L:8 D photoperiod. Fertilized embryos with intact chorion were collected in the morning (09:00) of the experimental day and maintained at one egg/ml hatching solution (17 mM NaCl, 0.4 mM KCl, 0.6 mM MgSO4, 0.36 mM CaCl2) with or without ethanol (0–1000 mM) in 48-well culture plates at 25°C with 16 L:8 D photoperiod. Each group consisted of 8–16 embryos and each experiment was repeated three to five times. Duration of ethanol treatment to the embryo was dependent on the nature of the experiment. Only viable and fertilized eggs were used in the experiment. The embryos were examined for developmental changes (cardiovasculature, blood clots, active circulation) under a phase contrast microscope (AO Scientific Instruments, Buffalo, NY) daily until 10 dpf with a 50% static renewal of the medium and the removal of dead embryos, if present. The embryonic development was classified after Iwamatsu (2004). Under these conditions, embryos generally began to hatch at 7 dpf and started normal feeding at 10 dpf. Due to the onset of hatching on 7 dpf, microscopic examination for developmental changes in some experiments was restricted from 0–6 dpf. Only hatched embryos at 10 dpf were used for cartilage and bone staining.

Embryonic Ethanol Determination

The ethanol concentration in the treated embryos was determined as previously described (Reimers et al., 2004a), with some modifications. Forty-eight fertilized medaka embryos (Iwamatsu stage 3–7) with intact chorion were exposed to 100 and 400 mM ethanol (24/group) for a maximum period of 48 hr and the ethanol concentration of the embryo was determined at five time points: 3, 8, 24, 30, and 48 hpf after ethanol exposure (100 mM and 400 mM). Embryos after removal from culture conditions were transferred to 1.5 ml microfuge tubes on ice and were washed with 1 ml cold 3.5% perchloric acid (PCA) twice to remove any residual alcohol from the outside of the chorion. Embryos were pooled (4–5 embryos for 100 mM and 3 embryos for 400 mM) and homogenized in 100 µl of 3.5% PCA followed by centrifugation at 12,000 × g for 15 min. The supernatants were saved and stored at 4°C in sealed tubes. The ethanol concentration of the supernatants was determined in a 96-well plate reader (HTS-700 Bio Assay Reader, Perkin-Elmer, Buckinghamshire, UK) at 340 nM by measuring NADH production at 37°C. The reaction mixture in 200 µl final volume contained 174 µl NAD (Sigma-Aldrich, St. Louis, MO) in 0.5 M Tris pH 8.8, 17.3 µl of yeast ADH (0.75 mg/ml) and 8.7 µl of either sample or standard ethanol solution. After 10 min preincubation at 37°C, production of NADH was measured. A blank reaction without ethanol was run simultaneously to correct for any substrate-independent generation of NADH. Embryonic ethanol concentration was determined by comparison to standard curves (0–2 mM). The experiment was repeated three times. The results were expressed as mM alcohol considering the average diameter of the egg is 1200 µm.

Cartilage Staining

Embryos at 10 dpf were used for cartilage staining in 0.1% Alcian blue (Sigma-Aldrich) as described for zebrafish by Carvan et al. (2004) with minor modifications. Hatchlings (5–10/group) were anesthetized in MS 222 (0.1% in hatching solution: Sigma-Aldrich) and fixed at 4°C for 24 hr in 4% paraformaldehyde (Sigma-Aldrich) in phosphate-buffered saline (PBS) with 0.1% Tween 20 (PBT). Fixed samples were washed twice in water and then treated for 10 min in 5% TCA. After brief washing in acid ethanol (0.37% HCl, 70% ethanol), embryos were stained in 0.1% Alcian blue for 1 hr. Excess stain was removed by washing in acid ethanol and samples were transferred to hydrogen peroxide (1%) for 12–24 hr for clearing. Another wash in Glycerol-KOH (50% glycerol and 0.25% KOH) was carried out before storing the stained samples at 4°C in 70% glycerol. Digital images of head skeletons, lower jaw including Meckel's cartilage (MC) of five to eight hatchlings in different orientations (dorsal, ventral, lateral) from various groups were captured on an Olympus B-Max 40 microscope with Optimus 6 image analysis software (Media Cybernetics, Silver Springs, MD).

Calcein Staining

Calcein (Sigma-Aldrich) is a fluorescent chromophore that specifically binds to calcium and fluorescently stains the calcified skeletal structure in live fish (Du et al., 2001). Calcein (0.2%) was dissolved in hatching solution and adjusted to pH 7.4 using 0.5 M NaOH. Control and ethanol-treated hatchlings (5–10/group) at 10 dpf were immersed in Calcein solution (1 embryo/ml) for 10 min and then washed in fresh hatching solution for another 30 min. The calcein-stained fish was anesthetized in MS 222 and mounted on a glass slide. Observations were carried out at a magnification of 4 × and 10 × by using a Nikon Eclipse E 600 microscope with a green fluorescence filter set (excitation filter, 450–490 nm; DM 500 nm; BA 515nm). Images of embryos in different orientations were captured with a Kodak MDS camera and saved in JPEG format. Composite images of the embryos were produced with Adobe PhotoShop 5.0 and saved in JPEG format.

Embryo Gender Identification

The gender of the embryo was determined by polymerase chain reaction (PCR) based on the method described by Kurauchi et al. (2005). Embryos treated with 400 mM ethanol for the first 2 days of development were used for genomic DNA extraction on 6 dpf. A total of 21 embryos was used for gender analysis. To each embryo 200 µl of 2 × CTAB reagent (2% hexadecyltrimethylammonium bromide in 100 mM Tris-HCl, pH 8.0, NaCl 1.4 M, EDTA 20 mM, and 0.2% β-mercaptoethanol; all from Sigma-Aldrich) was used for homogenization and followed by incubation at 60°C for 1 hr. An equal volume of chloroform (Sigma-Aldrich) was added to the homogenate, mixed and then centrifuged at 12,000 × g for 5 min at room temperature. The aqueous phase was removed to a clean 1.5 ml centrifuge tube and an equal volume of 2-propanol (Sigma-Aldrich) was added to the tube to precipitate genomic DNA. The mixture was centrifuged in 12,000 × g for 5 min at room temperature and the genomic DNA precipitate was saved and washed with 1 ml 70% alcohol. The genomic DNA was dissolved in 15 µl of nuclease free water and 1 µl of the extracted genomic DNA was used for PCR analysis. The reaction mixture for PCR in a 25 µl reaction volume contained 12.5 µl 2 × buffer (Promega, Madison, WI), 50 mM each of medaka dmy gene specific forward (5′-CCG GGT GCC CAA GTG CTC CCG CTG-3′) and reverse (5′-GAT CGT CCC TCC ACA GAG AAG AGA-3′) primers. Amplification reactions were carried out on a PTC-200 thermal cycler (MJ Research, Reno, NV) at 94°C for 2 min, one cycle, and then denaturation at 94°C for 30 sec, annealing at 60°C for 1 min and extension at 68°C for 2 min for 40 cycles with a final extension of 7 min for 1 cycle. The PCR products were visualized by 2% agarose gel electrophoresis containing 0.1% ethidium bromide. The primers amplified male-specific dmy gene and also the nongender-specific dmrtI gene resulting in two bands for male (1.0 and 1.2 kb) and one band for the female (1.2 kb) embryo (Kurauchi et al., 2005) (Table 1).

Table 1. Gender Ratio of Medaka Embryo Developmentally Exposed to 400 mM Ethanol
 With Active CirculationWithout Active Circulation
  1. a
    original image
  2. b

    A representative gel picture of the product(s) of dmy gene (partial) amplification by PCR. The female (♀) amplified a single band at 1200 bp and the male (♂) amplified two bands at 1200 and 1000 bp. The lanes marked B have no product due to the omission of template DNA in PCR reactions. The lane marked M represent DNA marker (100 bp ladder).

Male (n)86
Female (n)43

Semi-Quantitative and Quantitative Real-Time PCR

RNA from the pooled embryos (eight embryos/set) was extracted by Trizol reagent (Invitrogen, Carlsbad, CA) and treated with DNase I (Promega) to remove genomic DNA (Dasmahapatra et al., 2005). One microgram of DNA-free total RNA was used to synthesize cDNA using a first strand cDNA synthesis protocol (Invitrogen). The cDNA synthesis was made at 42°C for 1 hr in a 50 µl reaction volume containing 5 µl 10 × buffers, with 2.5 µl 10 mM dNTPs, 5 µl 25 mM MgCl2, 5 µl 10 mM DTT, and 200 U of superscript II RNA polymerase (Invitrogen). One microliter of synthesized cDNA in a 20 µl final reaction volume was used to determine Adh5 and Adh8 mRNA levels by semi-quantitative PCR (Dasmahapatra et al., 2000) or by real-time PCR described previously (Dasmahapatra et al., 2005) with some modifications. In both semi-quantitative and real-time PCR the same forward (Adh5: 5′-GTC ACA CAG ATG CCT ACA CTC-3′, Adh8: 5′-CAT TGC TGG ACG GAC CTG GAA G-3′) and reverse (Adh5: 5′-GCC CCG GCA ACT TTG CAG CCC-3′, Adh8: 5′-GTC GGG AAA CAC TCA GGA CTG-3′) primers for target genes were used. In semi-quantitative PCR, β-actin forward (5′-TTC AAC AGC CCT GCC ATG TA-3′) and reverse (5′-GCA GCT CAT AGC TCT TCT CCA GGG AG-3′) primers were used as an internal control. The reaction mixture in a final volume of 20 µl, contained 10 µl of 2 × master mix (Qiagen, Valencia, CA), 1 µl cDNA, 50 pM each of forward and reverse primers of target gene product, 50 pM of forward and reverse primers of internal standard (β-actin) and the remainder was nuclease free water. The amplification reaction was conducted in a PTC 200- thermal cycler (MJ Research) at 94°C for 2 min, followed by 30 cycles of denaturation at 94°C for 30 sec, annealing at 60°C for 1 min, extension at 72°C for 2 min with a final extension at 72°C for 7 min. After amplification, 5 µl of the each reaction product was separated in 2% agarose gel electrophoresis and the image of the gel was captured on a Versadoc Image analyzer (BioRad, Hercules, CA). The band intensity of the target and internal control was determined by image analysis soft ware (BioRad). The results were expressed as relative units to β-actin. In real-time PCR, SYBR green I (Sigma Aldrich) was used to label the double stranded DNA. Reactions were conducted in 20 µl reaction vol containing 10 µl of 2 × reaction buffer (Qiagen), 50 pM each of forward and reverse primer for the target mRNA (Adh5 or Adh8) and 1 µl of 4000 × -diluted SYBR green I (Sigma-Aldrich) in a real-time PCR thermal cycler (Opticon 2, MJ Research). The preparation of Adh5 and Adh8 standard curves was the same as described previously (Dasmahapatra et al., 2005). The reaction conditions were: initial denaturation at 94°C for 2 min, one cycle, followed by 40 cycles of denaturation at 94°C for 30 sec, annealing at 52°C for 1 min for Adh5 and 60°C for Adh8, extension at 72°C for 2 min, fluorescence data collection for 1 sec. To avoid error due to primer-dimer formation, a second set of fluorescence data was collected after incubating the samples for 1 sec at 78°C. A final extension of one cycle at 72°C for 10 min was made. The melting curve was constructed by plotting fluorescence data against temperature (65–95°C with an interval of 0.2°C). Each reaction was carried out in duplicate. The cycle threshold or C(t) line was set manually for the standard curve for each of the genes using the Opticon monitor software (MJ Research). This threshold was applied to all wells for consistent analysis of individual samples and standards for the experiments. Standards were run each time in every set of real-time PCR analysis. The results were expressed as mRNA copy number/ng of total RNA initially used for cDNA preparation. In this modified method of real-time RT-PCR the values of Adh mRNA copy numbers were 3- to 100-fold less than reported previously (Dasmahapatra et al., 2005). The exact cause of this difference has not been determined. It is possible that the differences may be due to the use of different ingredients like Taq polymerase and SYBR green I that may have been less sensitive compared to the commercial kit used previously. It is also possible that the adult breeding pairs used in this experiment were very young and possibly therefore, phenotypically different from the previous pairs. Despite differences in copy numbers, the nature of changes (Adh5 mRNA level remained unaltered throughout developmental period and Adh8 mRNA was significantly higher in 6-day embryos compared to 2-day embryos) were more or less similar to our previous studies (Dasmahapatra et al., 2005).

Statistical Analysis

The data were analyzed using one way ANOVA with post-hoc Tukey's multiple comparison test. The results are expressed as mean±SE with p<0.05 considered as significant.


Effects of Ethanol on Morphologic Features of Japanese Medaka

Embryonic development in medaka is comparatively slower than zebrafish under the experimental conditions of 25°C; 16 L:8 D. Medaka embryos began hatching at approximately ∼175 hpf (∼7.5 day) under these conditions compared to ∼48–72 hpf for zebrafish.

The calculated LC50 value for ethanol to cause 50% mortality in embryos at 10 dpf as determined from three separate independent experiments was 568 mM (3.2%) with 0–48 hpf constant exposure followed by clean hatching solution. Embryo survivability and hatching was severely affected at concentrations >600 mM with 100% mortality occurring at concentrations ≥800 mM (Fig. 1).

Figure 1.

Ethanol-mediated mortality in medaka embryos. Fertilized eggs of medaka within 1 hpf were exposed to ethanol (0–1000 mM) for the first 48 hr of development and the effect on mortality recorded at 10 dpf. Each group consisted of 8–16 embryos. The experiment was repeated three times. The LC50 was calculated to be 568 mM with an r2 of 0.9243 by log transformed data using non-linear regression (curve-fit) (GraphPad Prism). Each point represents the mean mortality percentage±SEM (n=3).

Embryos were examined for cardiovascular defects during development by microscopic observations. Ethanol dose-dependently affects cardiovascular development in medaka embryos. Embryos exposed to 400 mM ethanol showed blood clots in Blood Island, vitelline veins, embryonic body, eyes, and brain compared to controls (Fig. 2A–D). At 400 mM ethanol, onset of active circulation was significantly delayed and ∼50% of the embryos were unable to initiate active circulation; embryos exposed to lower concentrations of ethanol (0–300 mM) exhibited active circulation (Fig. 2E).

Figure 2.

Effects of ethanol on cardiovascular development in the Japanese medaka. A–D: Representative figures showing embryolethal effects of ethanol on cardiovascular morphology of medaka. Fertilized eggs of medaka were exposed to 400 mM (∼2.25%) ethanol for 48 hr. Photographs (40 ×) were taken at 96 hpf. Control (no ethanol) (A) with normal heart and three prominent cardinal veins. Ethanol-treated embryos developed tube heart (B), blood clots in the embryonic body (C), and the blood cells aggregated in cardinal veins (D). E: The time required to onset of active blood circulation in the embryos after ethanol treatment is presented in histograms. Each group contains 8–16 embryos. Each bar is the mean±SEM of three independent observations. *Values significantly different from control values. F: Embryos with the occurrence of blood clots in the body and the embryos able to initiate active circulation after removal of ethanol (400 mM) from the hatching solution. *Data significantly different from the corresponding embryos at 3 dpf. The control embryos (no ethanol) have normal circulation at 50 hpf and no observed blood clots. The values are the mean±SE of 60–80 embryos.

The extent of craniofacial and axial skeleton defects at 10 dpf were examined after staining with either Alcian blue or calcein. Embryos treated with ethanol (100–400 mM) had a malformed lower jaw (Fig. 3) compared to control animals as observed in Alcian blue staining (Fig. 3A,B). The length of the lower jaw was reduced significantly in ethanol-treated embryos compared to controls (Fig. 3C). In the tail, the developing hypural, epural, and perhypural cartilages were the only cartilaginous structures stained with Alcian blue (Fig. 3D). Calcification in these cartilages was inhibited by ethanol as shown with calcein staining (Fig. 3E,F). The urostyle was negative to Alcian blue and calcein at this stage of development (Fig. 3D,F).

Figure 3.

Effects of ethanol on lower jaw morphology and tail fin cartilage calcification. A,B: Representative figures showing embryolethal effects of ethanol on lower jaw (marked by an arrow) development of Japanese medaka embryos (40×). A: Control embryo. B: Embryos treated with 200 mM waterborne ethanol for first 48 hr of development. C: Effect of ethanol on the length of lower jaw. Embryos were stained with Alcian Blue and the lengths were measured from photomicrographs using Optimus image analysis software. The results are mean±SEM of five to eight individual observations. Bars with asterisks (*) indicate significant reduction in jaw length compared to controls. D–F: Effect of ethanol on axial skeleton structure and fin of tail region. Representative figures (40× for Alcian blue and 10 × for calcein) of Control (D,E) or 400 mM ethanol treated (F) medaka embryos stained either with Alcian Blue (D) or with Calcein (E,F) on 10 dpf. Two hypural (Hp), two parhypural (Ph), and one epural (Ep) cartilages were the only skeletal structures in the tail region positive for Alcian Blue staining (D). Ethanol treatment (400 mM) did not alter chondrification in these skeletons. Calcification was observed in parhypural, hypural and epural cartilages in control embryos (E) but was delayed/reduced in all the observed cartilages of the embryo treated with 400 mM ethanol (F). The urostyle (Ur) was negative for both Alcian blue and calcein staining.

Medaka Embryos Absorbed Only a Part of Waterborne Ethanol

The ethanol concentration in the embryo maintained a steady-state level throughout the exposure period; however, embryos exposed to 400 mM waterborne ethanol had a significantly higher but proportionate level of embryonic ethanol than those exposed to the 100 mM concentration. At 100 and 400 mM exogenous ethanol exposures, embryonic ethanol was approximately 15–20% of the waterborne ethanol concentration (Fig. 4).

Figure 4.

Embryonic ethanol determination of Japanese medaka embryo. The embryonic concentration after a 100 mM or 400 mM waterborne ethanol exposure was estimated using an ADH-dependent kinetic assay following Reimers et al. (2004a) at several developmental stages. Each experimental group consisted of 24 embryos. At each time point four to five embryos for the 100 mM group and three embryos from the 400 mM group were pooled for extraction of embryonic ethanol. The volume of the embryo was calculated on the basis of an average embryonic diameter 1200 µm. The results are expressed as the mean±SEM of three independent experiments.

Effect of Ethanol Is Not Gender-Specific

Results (Fig. 2E) indicate that ∼50% of embryos treated with 400 mM ethanol were able to start active circulation during development; however, the remaining ∼50% were unable to produce active circulation. To determine if this response was gender-specific, we determined the gender of the embryo at 6 dpf. The gender ratio of the embryos with active circulation was identical to the gender ratio of embryos without circulation (Table 1).

Effect of Ethanol on Expression of Adh5 and Adh8 mRNA in Japanese Medaka Embryos

Dasmahapatra et al. (2005) showed that the medaka embryo, like other fish species, expresses Adh mRNA during development. In medaka, Adh5 expression was constitutive throughout development whereas Adh8 mRNA expression was developmentally regulated (Dasmahapatra et al., 2005). To examine if ethanol treatment modulates the developmental expression pattern of Adh mRNAs, we measured Adh5 and Adh8 mRNA levels in medaka embryos by semi-quantitative and real-time RT-PCR at 2, 4, and 6 dpf after 100 and 400 mM ethanol treatment during the first 48 hr of development. Using either molecular technique, the results indicated that both Adh5 and Adh8 mRNA expression in controls and ethanol treated groups maintained the normal developmental rhythm (Adh5 expression was constitutive and Adh8 expression was increased over development) with no significant difference in mRNA levels between control and treatment groups in all 3 days examined (Fig. 5A–C, Table 2).

Figure 5.

Effect of ethanol on Adh5 (A) and Adh8 mRNA (B) expression in Japanese medaka embryo. mRNA expression was made from eight pooled embryos and analyzed by semi-quantitative RT-PCR. Each bar represents the mean±SEM of four to six separate experiments. *Value is significantly different from the corresponding embryo at 2 dpf. C: Representative gel electrophoresis picture of semi-quantitative PCR. Lanes 1–5 represent Adh8 and lanes 6–10 represent Adh5 mRNAs. The unmarked lane extreme left is the 100 bp DNA ladder.

Table 2. Effect of Ethanol on Adh5 and Adh8 mRNA Expression in Japanese Medaka
 Adh5 (copy number/ng RNA)Adh8 (copy number/ng RNA)
Day 2Day 4Day 6Day 2Day 4Day 6
  • a

    Results are mean ± SEM of three to eight separate observations.

  • *

    Values are significantly different from corresponding day 2 samples.

100 mM2722±4181524±4571900±24231.90±9.3262.79±27.3066.05±22.72
400 mM2332±4831976±9561745±80122.56±3.0365.04±16.5946.6±3.81*


It is evident from the present experimental results that medaka embryos exposed to ethanol early in development were able to produce several measurable phenotypic features associated with FAS, particularly in the cardiovascular system, the morphology of the lower jaw and tail cartilage calcification (that is also an indication of axial skeletal deformities) (Figs. 2, 3). These areas should be targeted for gene analysis.

Several animal models from non-human primates to round worms have been used to investigate alcohol's teratogenic action. Each model system has individual strengths and weaknesses, depending on the question being addressed (Cudd, 2005). Generally, simple animal models like fish have been used for identifying basic questions related to molecular biology and genetics. A criticism of the fish model has been that in comparison to mammalian models, very high concentrations of ethanol were required to produce significant morphologic changes (Cudd, 2005). The requirement of high concentrations of exogenous ethanol (400 mM) to induce cardiovascular malformation prompted us to determine the internal ethanol concentration of the medaka embryo during ethanol exposure (Fig. 4). The chorion represents a significant barrier to all exogenous chemicals and protects the embryo. It was likely that there would be a significant difference between exogenous ethanol added to the medium and the endogenous ethanol absorbed by the embryo. It was observed that medaka embryos exposed to 400 mM ethanol absorbed only a fraction (15–20%) of the waterborne ethanol and the embryonic ethanol concentration was maintained at a steady-state throughout the exposure period. In CD-1 mice, Blakley and Scott (1984) demonstrated that intraperitoneal injection of alcohol at the dose of 6 g/kg produced an embryonic concentration of 65.1 mM. Therefore, delivered ethanol in our model that induced FAS features were similar to the concentration used in mammalian models (Becker et al., 1996; Pitt and Carney, 1999). It is unclear why medaka embryo absorbs only a fraction from the surrounding environment and maintains a steady-state. In a similar experiment, Reimers et al. (2004a) demonstrated that zebrafish embryo absorbs ∼32% of the ethanol from the culture media. Therefore, a common feature for teleost embryos may be low chorionic membrane permeability with a yolk syncytial layer preventing movement of water and other materials into and out of the yolk compartment (Hagedorn et al., 2002).

Compared to mammalian models, the fish models offer unique advantages. Recently three laboratories independently documented measurable ethanol-induced phenotypic features in zebrafish that could be utilized further for analysis of ethanol-responsive genes (Bilotta et al., 2004; Carvan et al., 2004; Loucks and Carvan, 2004; Reimers et al., 2004a). As shown here and in previous studies, the Japanese medaka is an alternative non-mammalian model, in addition to zebrafish, to evaluate ethanol toxicity. Japanese medaka possesses unique combination of features that make it well suited, for studying several specialized phenotypic features including very easily observed features of the cardiovascular system (Kawamura et al., 2002). During embryogenesis, the medaka develops distinct yolk veins, serving as important target sites to investigate the effects of ethanol on the cardiovascular system. Another unique feature not found in mammalian models is the small size of the embryo that permits survival and continued development and observation for many days in the complete absence of blood circulation (Lambrechts and Carmeliet, 2004). Species-sensitivity differences can also be used as an advantage in investigating the mechanisms of teratogenesis. Apparently, medaka embryos were more resistant to ethanol than zebrafish embryos. The calculated LC50 for ethanol in medaka determined in the present experiment was 568 mM lpar;Fig. 1) compared to 338.5 mM in zebrafish (Reimers et al., 2004a) exposed to ethanol under similar conditions. Embryonic ethanol concentration is almost identical in these two fish species (15–20% of 568 mM [85.2–113.6 mM] for medaka and 32% of 338.5 [108.32 mM] for zebrafish). Therefore, Japanese medaka has all the necessary features for use as an alternative fish model or analyzing ethanol-sensitive genes expressed during embryogenesis.

The data obtained in this study indicated that medaka embryos responded to ethanol in a similar fashion as found in other animal models. Our observations were concentrated in two areas, cardiovascular and craniofacial features, highly relevant to fetal alcohol syndrome in humans. Ethanol exposure in medaka altered cardiovascular development structurally and functionally. At 400 mM ethanol exposure, the heart turns tubular nd blood clots initially in the Blood Island and then spreads into other organs of the embryo (Fig. 2). The onset of active blood circulation was significantly delayed and ∼50% of embryos were unable to start active circulation even after the removal of ethanol from the culture medium. The lower jaw of the embryo was shorter and was abnormally shaped (Fig. 3A–C). Similar to ethanol, chemicals like 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), α-naphthoflavone, resveratrol, and piperonyl butoxide (Kawamura and Yamashita, 2002) caused blood clotting and vascular damage with malformation of bones in medaka treated during embryogenesis. The effects of these chemicals, as demonstrated by the authors, were mediated through the aryl hydrocarbon receptor (AHR) pathway. In rainbow trout (Oncorhynchus mykiss), TCDD altered cardiovascular and craniofacial development (Hornung et al., 1999). In medaka, estrogen and retinoic acid inhibited yolk vein development, and caused blood clotting and bone defects (Kawamura et al., 2002; Hayashida et al., 2004). The cardiovascular and craniofacial cartilage developmental processes in fish have been shown to be highly sensitive to exogenous chemicals including ethanol and the effect may be due to cross talk between ethanol-affected pathways and AHR signaling pathways.

During early morphologic and functional development, the embryonic cardiovasculature may possibly be altered by any agent capable of inducing embryonic physiologic or oxidative stress. Ethanol is metabolized to acetaldehyde by the enzyme alcohol dehydrogenase (ADH) and acetaldehyde is further metabolized to acetate by the enzyme aldehyde dehydrogenase (ALDH). Fish embryos express both ADH and ALDH enzymes during development (Frankel, 1987; Danielsson et al., 1992, Funkenstein and Jakowlew, 1996; Dasmahapatra et al., 2001; Reimers et al., 2004b; Wang et al., 2005). During the metabolism of alcohol, superoxide ions are generated inducing embryonic oxidative stress. Moreover, acetaldehyde, the first metabolic product of alcohol, is itself highly toxic and accumulates in the heart (Espinet and Argiles, 1984). In zebrafish, acetaldehyde was found to be more lethal than ethanol (Reimers et al. 2004a). Therefore, the cardiovascular damage produced by ethanol in the medaka embryo may be an indirect effect due to the metabolism of ethanol.

Craniofacial cartilage and bone formation are dependent on the differentiation and migration of neural rest cells (NCC) during embryogenesis. Therefore, any modification in embryonic differentiation of NCC may have an effect on craniofacial cartilage and bone structure. Ethanol was demonstrated to alter NCC structure and function in mouse and chick (Chen et al., 2000; Debelak and Smith, 2000). In medaka, neural crestectomies on neurula stage embryos induced defects in craniofacial cartilage and bones (Langille and Hall, 1988). It is therefore possible that the malformed lower jaw observed in ethanol-exposed medaka was the result of ethanol-induced abnormal NCC morphogenesis. Moreover, it was observed that calcification in tail cartilage (hypural, epural, perhypural) of medaka embryos was altered/delayed by ethanol exposure (Fig. 3D,F). This indicates that ethanol targeted several axial skeletal structures other than the neurocranium. Taken together, the result suggests that embryonic ethanol exposure may alter the function of some cartilage and bone forming cells (osteocytes) and induce abnormality in osteogenesis.

Epidemiologically it has been demonstrated that the effect of ethanol is gender-specific with females being more sensitive than males to alcohol-induced diseases (i.e., liver diseases, cardiovascular diseases, brain damage) (Sato et al., 2001). The mechanism of such gender-specific effect of alcohol is unknown. In medaka, a variety of environmental chemicals like o,p′-DDT induces gender reversal of genetic males to a female phenotype (Stewart et al., 2000). High temperature (32°C) treatment also induces female–male gender reversal (Sato et al., 2005). As shown here on the basis of the PCR data, ethanol toxicity is not gender specific in Japanese medaka. The expression of dmy gene product (mRNA or protein) or gonad histology after ethanol treatment was not checked, however, the results allow us the tentative conclusion that in medaka, cardiovascular disorders induced by ethanol are not gender-specific.

Our results further indicated that ethanol even at the 400 mM concentration was unable to alter the expression pattern of Adh5 and Adh8 mRNA in the embryos (Fig. 5A,B, Table 2). The ability of fish embryos during ethanol-sensitive periods of development to metabolize ethanol to toxic metabolites has not been shown definitively. It is necessary to understand the response of the ethanol metabolizing enzymes during early life stages in normal and ethanol-influenced development. Mammalian studies have demonstrated that ethanol is oxidized to acetaldehyde by alcohol dehydrogenase (ADH), cytochrome P-4502E1 (CYP2E1), and catalase. ADH was detected in human fetal tissues at a mean gestational age of 11 weeks (Smith et al., 1971). Estonius et al. (1996) detected Class I ADH mRNA transcript as early as 18 weeks of gestation in human fetal lung, liver, and kidney, but not in the brain. In medaka, we have identified developmentally-expressed ADH enzyme mRNAs (Adh5 and Adh8). The expression of CYP2E1 or catalase enzymes during embryogenesis in medaka has not been investigated; however, CYP2E1-like activity in adult medaka was reported by Geter et al. (2003). In zebrafish CYP2E mRNA was detected by RT-PCR as early as 36 hpf (Reimers et al., 2004a). Taken together with our Adh mRNA data, it is probable that in the medaka embryo alcohol is metabolized primarily by ADH enzymes. The response of ADH enzyme gene(s) to ethanol is important in ethanol detoxification. In Drosophila melanogaster, Adh mRNA levels are increased in adult and larval stages upon exposure to ethanol, but no increase in ADH enzyme activity was observed (Malherbe et al., 2004). In postnatal rats developmentally exposed to ethanol, ADH enzyme activities in liver and intestine were significantly reduced after 3–4 weeks of development (Bhalla et al., 2005) but not immediately after birth or during first 2 weeks after birth. In medaka, cytosolic ADH enzyme activity as measured by enzymatic assay in the embryo is below the detection limit (data not shown). Therefore, we do not know the alteration of ADH enzyme activity after ethanol exposure. From our mRNA data it is possible that ADH enzyme remained unaltered after ethanol treatment and the embryo is able to metabolize ethanol through this pathway.

In conclusion, the current study has identified several measurable phenotypic features in medaka embryo useful for identifying genes related to ethanol-induced birth defects in humans. The well developed cardiovascular system during medaka embryogenesis with features not present in zebrafish and the species-specific sensitivity of medaka as compared to zebrafish provide unique advantages for the study of ethanol-induced teratogenesis.


This study was supported partially by the Office of Research and Sponsored Programs (small grant), a Chancellor's Partner Grant and the Environmental Toxicology Research Program of the University of Mississippi. We are thankful to S. Zhu, a graduate student of the Department of Pharmacology, University of Mississippi, for her considerable technical expertise and generous help in this study.