In Vivo Visual Reporter System for Detection of Estrogen-Like

David E. Hinton , Ron C. Hardman , Seth W. Kullman , Jerry M. (Mac) Law , Michael C. Schmale , Ronald B. Walter , Richard N. Winn , Jeffrey A. Yoder...
6 downloads 0 Views 474KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 2762-2768

In Vivo Visual Reporter System for Detection of Estrogen-Like Substances by Transgenic Medaka KANTA KURAUCHI,† YOSHITSUGU NAKAGUCHI,‡ MAKIKO TSUTSUMI,§ HIROSI HORI,§ RYO KURIHARA,⊥ SHINYA HASHIMOTO,| RYOKO OHNUMA,| YOSHIKAZU YAMAMOTO,| SUMIKO MATSUOKA,| SHIN’ICHIRO KAWAI,| TAKASHI HIRATA,† AND M A S A T O K I N O S H I T A * ,† Division of Applied Biosciences, Graduate School of Agriculture, Kyoto University, Sakyo, Kyoto 606-8502, Japan, Department of Microbiology, Graduate School of Medicine, Kyoto University, Sakyo, Kyoto 606-8501, Japan, Division of Biological Sciences, Graduate School of Science, Nagoya University, Nagoya 464-8601, Japan, Institute for Environmental Sciences, University of Shizuoka 52-1 Yada, Shizuoka 422-8526, Japan, and Department of Human Environmental Sciences, Kobe College, Okadayama 4-1, Nishinomiya, Hyogo 622-8505, Japan

Detection of endocrine disrupting chemicals, in particular, environmental estrogens with living organisms, has many advantages if compared to chemical analysis. The screening of novel pollutants with meaningful endpoints, the integration of uptake, bioconcentration, and excretion as well as the evaluation of endocrine disrupting effects with respect to toxicity require in vivo biotests for estrogenlike substances (ELSs). Critical disadvantages of whole organism biotests are their low sensitivity and the need for laborious and time-consuming work. To overcome these problems, we have developed a transgenic medaka strain harboring the green fluorescence protein (GFP) gene driven by choriogenin H gene regulatory elements. Choriogenin H is an egg envelope protein induced by estrogens in the liver. With yolk sac larvae of this strain, GFP induction in liver was observed 24 h after onset of aqueous exposure to 0.63 nM 17β-estradiol (E2), 0.34 nM ethynylestradiol, or 14.8 nM estrone. Furthermore, concentrated sewage treatment effluent induced GFP expression. Comparison of E2 equivalents estimated by GFP-induction in transgenic medaka, a YES assay, and GC/MS showed detection limits in the same order of magnitude. These results indicated that the sensitivity of the transgenic medaka strain was sufficient for application as an alternative model in monitoring environmental water samples for ELSs.

* Corresponding author phone: +81-75-753-6213; fax: +81-75753-6223; e-mail: [email protected]. † Graduate School of Agriculture, Kyoto University. ‡ Graduate School of Medicine, Kyoto University. § Nagoya University. ⊥ University of Shizuoka. | Kobe College. 2762

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005

Introduction Water pollution by estrogen-like substances (ELSs) of anthropogenic origin may interfere with the endocrine system of vertebrates and is considered a serious problem for aquatic and terrestrial vertebrates including man (reviewed in refs 1 and 2). Therefore, accurate and prompt detection of ELSs is a primary task to identify water with contamination and to produce waste cleanup methods that will reduce contamination of surface waters by ELSs. Chemical analysis is a widely used and sensitive method for the detection of ELSs in water samples. However, this type of analysis is characterized by certain drawbacks: it does not assess the bioavailability, the effect of complex mixtures of ELSs, toxicokinetics, or metabolic conversion of the original compounds to products with increased or reduced estrogenic properties. Furthermore, chemical analysis is not suitable for detection and searching of unknown estrogenic compounds because it requires known chemicals as standard controls. Different biotests using whole organisms have already been developed for the detection of ELSs (reviewed in refs 3-5): (1) morphological and histological analysis of gonads from wildlife and laboratory animals is used to characterize estrogenic effects on sex and gonadal differentiation (4, 6). (2) Detection of biomarkers, for example, the level of choriogenin mRNA and vitellogenin protein, is quantified in male liver or blood samples, respectively (7-10). The major disadvantage of these biotests is their demand for laborious experimental setup and time-consuming work. To overcome the disadvantages of the previous in vivo tests, in vitro assays were developed (reviewed in ref 5). Most of these assays depend on the binding activity of chemicals to the estrogen receptor. However, effects of metabolic conversion and/or bioconcentration are not considered and may therefore fail to correlate in vitro effects to effects on whole organisms. Another problem is the differential affinity of certain estrogen receptors for the compounds of interest. Furthermore, in vitro assays may tend to overestimate the potency of test chemicals (4). A transgenic zebrafish has been developed that utilizes an estrogen receptor-mediated luciferase gene induction system (ER-CALUC) to analyze estrogenic effects (11-13). This zebrafish model was able to detect estrogenic effects of various test chemicals during an exposure period of only 3 days. The detection limit of this zebrafish assay was 300 pM E2 (12). In the ER-CALUC assay, it is necessary to sacrifice the fish prior to analysis of luciferase activity. Therefore, it is impossible to perform time course experiments for estrogenic activity in the same set of fish. Produced in response to estrogens in the liver of medaka are vitellogenin and precursors of egg envelope proteins, choriogenin L and choriogenin H (ChgH) (14-17). Recently, Ueno et al. have developed a transgenic medaka harboring the green fluorescence protein (GFP) gene driven by a regulatory region of the choriogenin L gene. This transgenic strain expresses fluorescence of GFP in liver in response to stimulation by exogenous estrogen and may be useful to assess environmental estrogenic substances (16). Independently, we have established an easy detection and monitoring system for ELSs using a transgenic medaka with a choriogenin H regulatory region for the same model organism. In our study, we have focused on the quantification of the GFP fluorescence in 24 h short-term exposure experiments with yolk sac larvae. For the development of the transgenic strain, we have chosen the GFP as reporter gene since it can easily 10.1021/es0486465 CCC: $30.25

 2005 American Chemical Society Published on Web 03/05/2005

TABLE 1. Primer Sequences Used for the Cloning of the chgH Regulatory Sequence (mchr) and for the Determination of Sex in Yolk Sac Larvae (DMY) mchrF2: 5′-CTGAGCGGCCGCACCGACAGAAGCTCCAG-3′ mchrF3: 5′-ACCCTCAAGACCCTAAGGTT-3′ mchrF4: 5′-CCTGGTACCACACGTTATTTTGTGATAACC-3′ mchrR2: 5′-CTGAGAATTCTCACTGTTCAGATGGTTCGT-3′ mchrR3: 5′-TCCAGTGCCTTGCCATGGT-3′ mchrR4: 5′-GACAAGCTTCCTGAGGCTTCGGCTGTGGAT-3′ mchrR6: 5′-ATGGAATTCGCTGGTACTACTGCTGGTACA-3′ DMY-F01: 5′-CCGGGTGCCCAAGTGCTCCCGCTG-3′ DMY-R01: 5′-GATCGTCCCTCCACAGAGAAGAGA-3′

be detected with a fluorescence microscope in transparent fish larvae (18). The purpose of this paper is to detail our findings related to the establishment of the transgenic strain and its application for analysis of estrogenic substances and environmental samples (sewage treatment effluent) in water.

Materials and Methods Construction of pChgH-GFP. To obtain the partial cDNA sequence of the medaka chgH gene, total RNA was extracted from the liver of a sexually matured female of the inbred strain Hd-rR. RT-PCR was performed with the chgH specific primers designed by means of published DNA sequence information (15). Using the RT-PCR product as a probe, a bacterial artificial chromosome (BAC) library of the Hd-rR strain constructed by Matsuda et al. (19) was screened for the chgH gene. With the DNA of the positive BAC clone, part of the exon-intron structure was identified using primers for inverse-PCR: in case of cloning the 3′ flanking sequence of the chgH gene, DNA from the positive BAC clone was partially digested with HaeIII and self-ligated with T4 DNA ligase (Invitorogen). PCR was performed with the specific primers mchrF2 and mchrR2 using circular BAC clone DNA as a template. The longest PCR product (about 800 bp) was cloned into the pGEMeasy vector (PROMEGA, WI). On the basis of the sequence of the cloned fragment, we designed new primers (mchrF2 and mchrR6) that amplified about 800 bp downstream of the stop codon (TGA) of chgH. The resulting PCR product (accession no. AB115181) was inserted into the NotI and EcoRI restriction emzyme site of the pEGFP vector (Clontech). For the cloning of the 5′ flanking sequence, the procedures were almost the same as for the previous 3′ flanking sequence with the following modifications: extracted DNA from the positive BAC clone was digested with AflIII, and inverse-PCR was performed with primers specific for the 5′ flanking sequence of chgH (mchrF3 and mchrR3). The PCR product (about 2 kbp) was cloned into pGEMeasy, and the DNA sequence was determined. With primers mchrF4 and mchrR4, a 2 kbp region upstream of the translation initiation site was amplified using DNA from the BAC clone as template. The resulting PCR fragment (accession no. AB115180) was inserted into the KpnI and NcoI restriction emzyme sites of the pEGFP vector in which the 3′ flanking sequence was already inserted. The resulting plasmid was named pChgH-GFP. The DNA sequences of primers used for the cloning of the chgH regulatory elements are summarized in Table 1. Establishment of ChgH-GFP Transgenic Medaka Strain. The ChgH-GFP plasmid was injected into fertilized eggs of d-rR strain before the first cleavage as described by Kinoshita et al. (20). The d-rR strain is an out-bred strain that shows different body color between male (orange) and female (white). Out of 62 injected embryos, 25 were raised until sexual maturation. Five individuals (F0 generation) showed the expression of the GFP transgene in their germ cells. These

transgenic F0 individuals were mated with nontransgenic d-rR to establish F3 heterozygous transgenic fish. Exposure Conditions. Exposure to estrogenic chemicals was performed at 28 °C using a 14 h light/10 h dark cycle. Yolk sac larvae were incubated in 0.2 mL/individual of the test medium in 96-well plates. Fish were not fed during the exposure time. Eight fish were tested for each concentration of the chemical or sewage treatment effluent. Stock solutions of estrogenic chemicals (17β-estradiol: E2; estrone: E1; ethynylestradiol: EE2; and nonylphenol: NP) in ethanol were added to glass beakers. After the evaporation of ethanol, water was added to give a nominal concentration of each chemical (21). The actual concentration of E2 in the stock solution was determined by gas chromatography-mass spectrometry (GC/MS). The nominal concentration of 184 nM (50 µg/L) of the E2 stock solution corresponded to an actual concentration of 317 nM (86.3 µg/L). All chemical compounds were purchased from Wako (Tokyo, Japan) or Nacalai Tesque (Kyoto, Japan). In the environmental water samples, the water treated by an activated sludge (sewage treatment effluent) was collected from a sewage treatment plant in a Hyogo prefecture in Japan. Twenty liters of this sewage treatment effluent was extracted and concentrated 2000× in ethanol using a Sep-PakC18 cartridge (Sep-Pak plus long body type, Cat No. 023635, Waters, MA; the volume of this cartridge was 2 mL) as described previously (22). The yield in extraction using this cartridge was 90%. The concentrated sewage treatment effluent in ethanol was added to the test medium to give final concentrates from 2.5 × to 40× . The final solvent concentrations of samples and controls were 1% ethanol. Identification of the Gender of Yolk Sac Larvae. After the GFP expression was measured, DNA was extracted from the whole body of yolk sac larvae as described previously (23). The gender of yolk sac larvae was determined by PCR analysis of the male-specific dmy gene. The primers also amplified the nonsex-specific dmrt1 gene resulting in two bands for male (1.0 and 1.2 kbp) and one band for the female (1.2 kbp). Measurement of GFP Fluorescence. For the analysis of GFP fluorescence, fish were anesthetized with ice water and were rapidly observed with a fluorescence stereoscopic microscope (MZFLIII, Leica Microscopy Systems, Heerbrugg, Switzerland) or BX50WI (OLYMPUS, Tokyo, Japan) equipped with GFP filter sets. Fluorescent images were recorded with a color digital cooled charge-coupled device (CCD) camera (VB-7010, KEYENCE, Osaka, Japan). The CCD camera system used in this study calculates exposure time to automatically give a constant light intensity. The GFP amount in the liver was measured by the reciprocal calculated exposure time. Various concentrations of the standard GFP protein (rEGFP: Clontech #8365-1) solution were poured into a glass capillary (1.5 mm in diameter). Then, 9.45 nL (0.09 mm × 0.07 mm × 1.5 mm) of the GFP solution was observed with a fluorescent microscope, and the calculated exposure time was measured. The reciprocal calculated exposure time correlated with the amount of GFP. The relationship between GFP amount and reciprocal calculated exposure time could be described by the following equation:

G = RCET/8.39 - 0.024 (correlation coefficient ) 0.9637) where G is the amount of GFP in pg, and RCET is the reciprocal calculated exposure time (see Figure 1). In Vitro Bioassay and GC/MS Analysis. Estrogen-like substances in the sewage treatment effluent were estimated with the yeast estrogen screen (YES assay), enzyme-linked immunosorbent assays (ELISAs), and GC/MS analysis. Estrone (E1) ELISA kit (Takeda, Osaka Japan) or the 17β-estradiol VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2763

initiation site of the chgH gene could not be exactly determined, a putative TATA-box sequence was identified at the -33 position from the 5′ end of the cDNA reported by Murata et al. (15). An estrogen responsive element (AAAGGTCTCTGTGACCTGG) was detected 276 nts upstream from the translation initiation site.

FIGURE 1. Relationship between the intensity of GFP fluorescence and the reciprocal calculated exposure time of the CCD camera. The CCD camera automatically calculates an exposure time to obtain images with constant light intensity. The calculated exposure time was measured for various concentrations of standard GFP solutions. The reciprocal calculated exposure time was shown to linearly increase with increasing amount of GFP (see Materials and Methods for details). (E2) ELISA kit (Takeda, Osaka, Japan) were used to determine steroidal estrogens according to the manufacturer’s instructions. The YES assay and GC/MS analysis were carried out as described previously (22, 25). Statistics. The data were expressed as mean ( standard deviation (SD). Significant differences were determined by the Mann-Whitney test using Free JSTAT 8.2 (for Windows) software.

Results Analysis of the Regulatory Region of the Choriogenin H Gene. We identified a sequence of about 2000 nts upstream of the translation initiation site (ATG) of the chgH gene. The sequence was almost identical (1516 of 1528 nt) to that reported by Ueno et al. (16). Although the transcription

GFP Expression in Transgenic Medaka Strains. We obtained five transgenic strains harboring the ChgH-GFP reporter construct. These strains were derived from five different F0 individuals. Both male and female yolk sac larvae of all strains showed the induction of the GFP expression in liver when exposed to 5.1 nM E2 (data not shown). However, the intensity of the fluorescence induction by E2 varied among the different strains. We selected one strain, denoted as C-line, which showed a relatively strong fluorescence when compared to all other strains. The C-line was then selected for use in all subsequent analyses. In this strain, embryos of the 35-somite stage expressed a weak green fluorescence on their body surface without estrogen stimulation. In later stages, this fluorescence tends to fade and was hardly detectable before hatch (data not shown). Without stimulation by estrogens, no green fluorescence was observed at least until 14 days post hatching. In later stages, pigmentation in the abdominal cavity increased, and it was difficult to observe background or induced fluorescence in inner organs. At around 1 dph (days post hatching), the exposure to E2 stimulation induced GFP expression exclusively in the liver of both males and females (Figure 2). No significant difference in induced GFP amount was observed between males and females exposed to 3.2 nM E2 (Table 2). Therefore, we used mixed populations of both male and female yolk sac larvae for further experiments.

FIGURE 2. Induction of green fluorescence in ChgH-GFP transgenic yolk sac larvae of the medaka. Fluorescence was observed after exposure of 5.1 nM E2 for 24 h. The pictures show images of a male (A-C) and a female fish (D-F) with green fluorescence exclusively in the liver. The gender of the two yolk sac larvae was detected by the amplification of the male specific gene dmy and the nonsex specific gene dmrt1 using PCR and genomic DNA as a template. Males show two bands (1.2 and 1.0 kbp; left column in panel G). Female show the dmrt1 band only (1.2 kbp; right column in panel G). Therefore, fish 1 and 2 were identified as male and female, respectively. Panels A and D are bright light and panels C and F are fluorescence images. Panels B and E were obtained by the fusing of bright light and fluorescent images. Yellow spots in fluorescence images are caused by auto-fluorescence of pigment cells. Each bar represents 0.2 mm. 2764

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005

TABLE 2. Comparison of GFP Induction in Transgenic Male and Female Yolk Sac Larvae Exposed to 5.1 nM 17-β-Estradiol (E2) for 24 ha gender

n

GFP (pg)

male female

11 9

1.1 ( 0.6 1.3 ( 0.6

a Green fluorescence in liver was observed with a fluorescent microscope, and GFP content was determined by image analysis (see Figure 1 and Materials and Methods for details). n: number of tested yolk sac larvae. Data represent means ( SD. No statistically significant difference was detected between males and females.

Concentration-Response Analysis of GFP Amount in Yolk Sac Larvae Exposed to E2. To develop a simple and easy method to detect estrogenic substances in water

samples, we used yolk sac larvae since they exhibited a low body pigmentation. The transparent body allows us to observe the fluorescence in inner organs of intact living fish. Zero and 1 dph fish were exposed to various concentrations of E2 for 24 h (Figure 3). GFP fluorescence in liver of individual fish was induced at 0.63-10.2 nM E2 with increasing intensity at higher concentrations. At 0.63 nM, the fluorescence was observed only in some parts of the liver (Figure 3D). The area of fluorescence in liver varied among individuals (data not shown), and a few yolk sac larvae showed no fluorescence. At concentrations of 1.27 nM and above, a strong intensity of green fluorescence was observed in all individuals (Figure 3E-H). We developed a convenient method to quantify the amount of GFP in the liver of exposed fish. With the CCD camera used (details in Materials and Methods), it was

FIGURE 3. Concentration-dependent induction of green florescence in transgenic medaka yolk sac larvae exposed to 17-β estradiol (E2). Transgenic yolk sac larvae were exposed to various concentrations of E2 for 24 h, and the GFP content was calculated by image analysis (see Figure 1 and Materials and Methods for details). Panel A: bright light image. Panels B-H: fluorescent images obtained from 0, 0.32, 0.63, 1.27, 2.54, 5.08, and 10.2 nM E2 exposure. The graph (I) shows the quantitative analysis of GFP content obtained from eight individuals in each E2 concentration. Data represent means ( SD. Asterisks indicate statistically significant differences (P < 0.01) to controls. Bar is equal to 0.2 mm. VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2765

TABLE 3. Lowest Observed Effect Concentration (LOEC) and E2 Equivalent Valuea chemicals

LOEC (nM)

E2 eq

E2 E1 EE2 NP

0.63 14.79 0.34 4500

1 0.043 1.87 0.00014

a LOEC was defined as the lowest concentration of each chemical that induced a statistically significant increase of GFP amount. E2 equivalent value (E2 eq) was calculated by the ratio of [LOEC of E2]/ [LOEC of specific chemical].

FIGURE 4. Induction of GFP in ChgH-GFP transgenic medaka yolk sac larvae exposed to various concentrations of estrone (E1), ethynylestradiol (EE2), and nonylphenol (NP). Green fluorescence in liver was observed with a fluorescent microscope, and GFP content was determined by image analysis (see Figure 1 and Materials and Methods for details). Eight fish were exposed to each chemical. Data represent means ( SD. Asterisks indicate statistically significant differences (P < 0.01) to controls. possible to calculate the photo exposure time necessary to obtain pictures with the same overall light intensity. We observed that the reciprocal calculated exposure time correlated linearly with the GFP amount by analyzing droplets of equal volumes but with different GFP concentrations. By calculating a regression curve, we were able to transform the reciprocal calculated exposure times to the corresponding GFP amount in the range of 0-24 pg of GFP (Figure 1 and see Materials and Methods). On the basis of this quantification of fluorescence, GFP amount in fish exposed to E2 was observed to increase in an E2-concentration-dependent manner (Figure 3I). Although a high degree of individual variation in the GFP content was noticed, statistically significant differences in induction for the average GFP amount was obtained for 0.63 nM E2 and above. Induction of GFP by Exposure to E1, EE2, and NP. We also investigated the ability of other estrogenic chemicals to induce GFP fluorescence in the ChgH-GFP transgenic medaka. As shown in Figure 4, E1, E2, and NP showed an increase of GFP amount in a concentration-dependent manner similar to E2 (Figure 3). In the case of E1, a statistically significant difference for the average GFP amount was observed at 14.79 nM and above. EE2 showed a statistically significant increase of the GFP amount at 0.34 nM and above. For NP, a statistically significant difference was detected at 4.50 µM. Microscopical observation of yolk sac larvae exposed to 9.10 µM NP revealed that some fish showed an increase in liver GFP fluorescence, although some showed faint GFP fluorescence in the liver. In higher concentrations of NP than 9.10 µM, all fish were dead (data not shown). These findings suggested that exposure to 9.10 µM NP was harmful and critical to the survival of medaka and that depressed fish might show faint GFP fluorescence in liver. Because of a high degree of variation in fluorescence intensity, no statistically significant induction was obtained for this concentration. 2766

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005

We defined the lowest observed effective concentration (LOEC) for estrogenic effects as the lowest concentration at which a statistically significant induction of GFP amount was observed when compared to controls. The LOECs of E2, E1, EE2, and NP were estimated as 0.63, 14.79, 0.34, and 4500 nM, respectively (Table 3). Thus, the estrogenic potency (calculated as an E2 equivalent) of E1, EE2, and NP were 0.042, 1.9, and 0.00014, respectively (Table 3). Induction of GFP by Exposure to Sewage Treatment Effluent. To evaluate the suitability of the ChgH-GFP transgenic medaka strain for the detection of estrogens in environmental water samples, we exposed yolk sac larvae to various dilutions of a concentrated sewage treatment effluent. A statistically significant induction of GFP amount was observed when yolk sac larvae were exposed to sewage treatment effluent concentrated 5-40× (Figure 5). The intensity increased with higher concentrations showing a maximum GFP amount at effluent concentrated 20× (Figure 5D). At exposure to 40× concentrated effluent, the yolk sac larvae showed abnormal swimming behavior (they bent backward and curved quickly), indicating toxic effects of the highly concentrated samples. The E2 equivalent of the original effluent was estimated as 0.126 nM. By GC/MS analysis, the concentrations of the estrogenic compounds E1, E2, 4-NP, 4-octylphenol (4-OP), or bisphenol A (BPA) in the sewage treatment effluent were measured as 0.70 nM, 0.086 nM, 6.91 nM, 7.3 pM, or 0.5 nM, respectively (Table 4). EE2 was not detectable with GC/MC analysis. The concentrations of E2 and E1 together correspond to an E2 equivalent of 0.119 nM. Similar results were obtained if the concentrations of steroidal estrogens were determined by ELISA (0.692 nM for E1 and 0.043 nM for E2). The E2 equivalent concentration based on ELISA measurements for E1 and E2 was calculated as 0.0731 nM. The E2 equivalent of sewage treatment effluent determined with the YES assay was 0.126 nM.

Discussion In this study, we established a short term in vivo assay for the detection of estrogenic activity in environmental water samples using a transgenic medaka. The resultant convenient method enabled quantification of estrogenic activity by transgenic expression of GFP in yolk sac larvae exposed for only 24 h. We used a transgenic medaka harboring the GFP gene under the control of regulatory regions of the choriogenin H gene. Choriogenin H is an egg envelope protein induced in the liver of medaka by endogenous and exogenous estrogen stimulation (7, 15). Similar to results with the vitellogenin gene, estrogenic substances activate the expression of chgH by binding to the estrogen receptor, and the ligand-receptor complex then binds to a so-called estrogen-responsiveelement (ERE) in the promoter region and enhances the transcription of the gene. In contrast to other reporter proteins such as luciferase (11-13, 26), GFP can be detected without any sample preparation step by excitation to light

FIGURE 5. Detection of estrogenic activity in sewage treatment effluents using transgenic medaka yolk sac larvae. Yolk sac larvae were exposed to concentrates of sewage treatment effluents for 24 h. Panels A-C represent fluorescent images of yolk sac larvae exposed to 2.5, 5, and 10 times the concentrates of sewage water. Bar is equal to 0.2 mm. Panel D shows the GFP content calculated by image analysis (means ( SD). Values on the x-axis represent the degree by which the sewage treatment water was concentrated. Asterisks indicate significant differences among the means (P < 0.01) when compared with controls (the solvent concentration was 1% ethanol in all samples).

TABLE 4. Identification of Estrogenic Substances in a Sewage Treatment Effluent by the Transgenic Medaka (TGM), Yeast Estrogen Screen (YES), ELISA, and GC/MS Analysisa methods chemicals EE2 (nM) E2 (nM) E1 (nM) NP (nM) 4-0P (nM) BPA (nM) E2 eq (nM)

TGM

YES

ELISA 0.043 0.70

0.126

0.126

0.0731

GC/MS < 0.087 0.70 6.91 0.000035 0.0022 0.119

a

TG and YES assays cannot detect individual chemical substances but detect total estrogenic activity. To compare all methods, measurements of ELISA and GC/MS have been transformed to E2 equivalent values (E2 eq). E2 eq were calculated as detailed in the Materials and Methods. Because of the low E2 eq, NP, 4-OP, and BPA were not included in the calculation of E2 eq.