Wolfgang Körner

Input/output balance of estrogenic active compounds in a major municipal sewage plant in Germany

24 h samples of untreated and treated wastewater were taken in parallel from a modern municipal sewage plant in southern Germany in March and June 1998. After solid phase extraction, total estrogenic activity was quantitatively measured with a miniaturized E-screen assay and the levels of nine estrogenic phenolic chemicals analyzed by HRGC/LRMS. 17Beta-estradiol equivalent concentrations (EEQ) were 58 and 70 ng/l in the influent and 6 ng/l in the effluent, indicating that the load of estrogenic activity of the wastewater was reduced by about 90% in the sewage plant. Less than 3% of the estrogenic activity was found in the sludge. 4-t-octylphenol, 4-nonylphenol, bisphenol A, 2-hydroxybiphenyl, and 4-chloro-3-methylphenol were detected in the untreated wastewater at levels from 0.13 to 3.6 microg/l. 4-t-octylphenol, 4-nonylphenol, and bisphenol A were present in the effluent at concentrations from 0.16 to 0.36 microg/l, 2-hydroxybiphenyl and 4-chloro-3-methylphenol were not detectable. The contribution of the quantified levels of phenolic xenoestrogens to total estrogenic activity in the sewage was 0.7-4.3%.

Phenolic xenoestrogens in surface water, sediments, and sewage sludge from Baden-Württemberg, south-west Germany.

Nine structurally different phenolic chemicals, which have been reported to mimic estrogen effects, were determined in various aquatic environmental compartments. Twenty-three water samples from five streams and rivers showed levels up to 458 ng/l for 4-nonylphenol (4NP), 189 ng/l for 4-t-octylphenol (4tOP), 272 ng/l for bisphenol A (BPA) and 47 ng/l for 2-hydroxybiphenyl (2OHBiP). Elevated levels of these compounds in a stream with a high load of effluents of sewage treatment plants (STPs), compared to a brook free of sewage, identified STPs as major sources. With a similar order, 4NP (10-259 micrograms/kg dry matter), 4tOP (< 0.5-8 micrograms/kg), BPA (< 0.5-15 micrograms/kg), and 2OHBiP (2-69 micrograms/kg) were also detected regularly in riverine sediment (n = 11). Levels in sewage sludge were one order of magnitude higher than in sediments. 4-Hydroxybiphenyl and 4-chloro-3-methylphenol were found predominantly in sludge and sediment in the lower ppb range.

Development of a sensitive E-screen assay for quantitative analysis of estrogenic activity in municipal sewage plant effluents

A simplified proliferation test with human estrogen receptor-positive MCF-7 breast cancer cells (E-screen assay) was optimized and validated for the sensitive quantitative determination of total estrogenic activity in effluent samples from municipal sewage plants. After solid phase extraction of 1 l sewage on either 0.2 g polystyrene copolymer (ENV+) or 1 g RP-C18 material and removal of the solvent, analysis of the extracts in the E-screen assay could be performed without any clean-up step. This was even possible with untreated sewage. Parallel extraction of four sewage samples on both different solid phase materials gave comparable quantitative results in the E-screen. A blank sample did not induce cell proliferation. As additive behaviour of the estrogenic response of single compounds was proven for two different mixtures each containing three xenoestrogens, total estrogenic activity in the sewage samples, expressed as 17 beta-estradiol equivalent concentration (EEQ), could be calculated comparing the EC50 values of the samples with those of the positive control 17 beta-estradiol. The detection limit of the E-screen method was 0.05 pmol EEQ/l (0.014 ng EEQ/l), the limit of quantification 0.25-0.5 pmol EEQ/l (0.07-0.14 ng EEQ/l). In total, extracts of nine effluent and one influent sample from five different municipal sewage plants in South Germany were analyzed in the E-screen. All samples strongly induced cell proliferation in a dose-dependent manner which was completely inhibited by coincubation with 5 nM of the estrogen receptor-antagonist ICI 182,780. The proliferative effect relative to the positive control 17 beta-estradiol (RPE) was between 30 and 101%. 17 beta-Estradiol equivalent concentrations were between 2.5 and 25 ng/l indicating a significant input of estrogenic substances via sewage treatment plants into rivers.

Validation and application of a rapid in vitro assay for assessing the estrogenic potency of halogenated phenolic chemicals

The E-Screen assay serves as an in vitro tool for the detection of estrogenic activity of chemicals and extracts of environmental samples. Based on the induction of proliferation in human estrogen receptor-positive MCF-7 breast cancer cells we could substantially simplify the assay. As one important step of validation we applied the modified assay for testing nine known xenoestrogens. We could confirm the results of other groups assuring the reproducibility of the E-Screen assay. The results provide evidence that the E-Screen assay is suitable for determination of estradiol equivalency factors (EEFs) for environmental estrogens to rank their estrogenic potency relative to the natural estrogen 17 beta-estradiol. Further, we used the optimized proliferation test to screen nine halogenated phenolic compounds for their possible estrogenic potency. Three widely applied chemicals expressed a weak receptor-mediated estrogenic activity: the flame retardant Tetrabromo-Bisphenol-A, the disinfectant 4-chloro-3-methylphenol, and the herbicide educt 4-chloro-2-methylphenol. Their estrogenic potencies were five to six orders of magnitude lower than that of 17 beta-estradiol.

Development and validation of a GC/MS method for determination of phenolic xenoestrogens in aquatic samples.

A simple and sensitive GC/MS method for the quantitative determination of the estrogenic phenolic compounds 4-nonylphenol, 4-t-octylphenol, bisphenol A, 3-t-butyl-4-hydroxyanisole, 2-t-butyl-4-methylphenol, 4-hydroxybiphenyl, 2-hydroxybiphenyl, 4-chloro-3-methylphenol, and 4-chloro-2-methylphenol in aquatic samples was developed. The method for assessing their occurrence in sewage, surface and drinking waters consists of solid phase extraction (SPE) using a polystyrene copolymer phase. After methylation of the extract HRGC/LRMS analysis was possible without any clean up, even in raw sewage samples. Limits of detection and determination were between <0.01 and 0.05 ng/l and 0.01 and 0.05 ng/l, respectively. Recoveries were above 70% with exception of 3-t-butyl-4-hydroxyanisole.

Steroid analysis and xenosteroid potentials in two small streams in southwest Germany

The presence of estrogenic substances in thewater of the small streams Körsch (Kö)and Krähenbach (Kr), Southwest Germany, wasdetermined by chemical and biological analysis.Because a large proportion of the Kö waternear its mouth consists of sewage treatmentplant (STPs) effluents, the impact of STPs onlevels of estrogens in surface water is anenvironmental issue of concern. In July 1996,water samples were taken from Kr and Kö(four sites) and tested in the E-Screen assaywith human MCF-7 breast cancer cells. AllKö samples induced estrogen-dependent cellproliferation resulting in 17β-estradiolequivalent concentrations (EEQ) between 3.3 and9.7 ng/L while the Kr water showed no effect.In 1998/99 eight samples from Kö (near itsmouth) and nine samples from Kr were collectedand tested in the E-Screen after solid phaseextraction. Some estrogenicity was detectablein three Kr samples but Kö samples had amedian EEQ of 3.1 ng/L (range: 1.2–42 ng/L).GC/MS analysis revealed differences in thelevels of 17β-estradiol and estronebetween the two streams. 17β-estradiolwas detectable in five Kö samples only (0.7–1.8 ng/L). Estrone was found in the Köfrom 2.5 to 38 ng/L (median: 7.6 ng/L) and inthe Kr between 0.8 and 22 ng/L (median: 1.7 ng/L). Analysis for nine phenolic xenoestrogensrevealed rather low levels for five compoundswhich occurred more frequently and in higherconcentrations in the Kö. After asemi-field exposure of adult male rainbow troutfor 4 weeks in Kö water, plasmavitellogenin levels were significantly highercompared to those levels detected in the sameanimals before exposure.

Tissue distribution after a single subcutaneous administration of 2,3,7,8-tetrabromodibenzo-P-dioxin in comparison with toxicokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin in female wistar rats

Tissue concentrations of 2,3,7,8-tetrabromodibenzo-p-dioxin (TBDD) and induction of ethoxyresorufin O-deethylase (EROD) were determined in female Wistar rats following a single subcutaneous (s.c.) injection of TBDD. Two sets of experiments were performed in order to study (a) the time course after a single s.c. administration of 600 ng TBDD/kg body wt up to 78 days, and (b) the dose-response seven days after a single s.c. injection of different doses of TBDD (3 to 3,000 ng/kg body wt). The results obtained on toxicokinetics and enzyme induction were compared with those following a single s.c. administration of 300 ng/kg body wt 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Three days after the injection, approximately 93% of TBDD and 90% of TCDD had been absorbed. Fourteen days after s.c. injection less than 1% of the administered dose of both substances remained at the injection site. Three days after a single s.c. injection of 600 ng TBDD/kg body wt and 300 ng TCDD/kg body wt, the maximum tissue concentrations in the liver amounted to (M +/- S.D.) 5.7 +/- 0.8 and 4.7 +/- 0.9 ng/g wet weight, respectively. In adipose tissue, the peak concentration was 3.2 +/- 0.2 ng/g wet weight for TBDD on day 14, and 0.8 +/- 0.1 ng/g for TCDD on day 7. Throughout the study, the concentration ratio in the TCDD-treated group was always at least twice as high as that in the TBDD-treated group. The elimination half-life (t1/2) of TBDD and of TCDD in the liver was 13.3 and 13.6 days, respectively. In the adipose tissue the t1/2 of TCDD was 24.5 days but no reliable t1/2 could be calculated for TBDD (t1/2 = 39.4 days with a 95% confidence interval of 25.9 to 82.4 days). Tissue content of TBDD and TCDD in liver and adipose tissue increased dose-dependently, and the linear regression in a double-logarithmic plot showed a straight line. Time course of the induction of hepatic EROD activity after treatment with 600 ng TBDD/kg body wt was almost identical with that observed following a single dose of 300 ng TCDD/kg body wt. The induction of hepatic EROD activity was linearly correlated in a double-logarithmic plot to the hepatic concentrations of the congeners (both TBDD and TCDD). The slopes of the dose-response curves after administration of TBDD and TCDD were almost parallel for tissue concentrations ranging from 0.1 to 30 ng/g wet weight.

Tissue concentrations and induction of a hepatic monooxygenase in male Wistar rats after repeated doses of defined polychlorinated dibenzo-p-dioxin and dibenzofuran (PCDDs and PCDFs) mixtures

Abstract. Two groups of male Wistar rats were treated 16 times (every 3rd day) subcutaneously with a defined mixture of polychlorinated dibenzo-p-dioxins (PCDDs) or of polychlorinated dibenzofurans (PCDFs). These mixtures contained no measurable amount of 2,3,7,8-TCDD. Each single dose was calculated to contain either 57 ng I-TEq (international 2,3,7,8-T4CDD toxicity equivalencies)/kg body weight of the PCDD mixture or 39 ng I-TEq/kg body weight of the PCDF mixture. Both mixtures contained a large excess of non-2,3,7,8-substituted congeners. The activities of ethoxyresorufin O-deethylase (EROD) in liver microsomes were correlated with the corresponding concentrations of PCDDs or PCDFs in hepatic tissue. Data were compared with results obtained after single injections of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-T4CDD). As expected, a complex kinetic situation resulted, because of the different tissue distributions and elimination half-lives of the various congeners: (1) 2,3,7,8-substituted PCDDs: the time course of the concentrations in liver and adipose tissue was similar for all congeners, the levels increased during the treatment period and decreased after treatment. Tissue concentrations of all 2,3,7,8-substituted PCDDs were considerably higher in liver than in adipose tissue. The liver/adipose tissue concentration ratios increased with the degree of chlorination. The ratio of 1,2,3,7,8-P5CDD was much lower than those of all other 2,3,7,8-substituted congeners. (2) 2,3,7,8-substituted PCDFs: 1,2,3,7,8-P5CDF was rapidly eliminated from liver and adipose tissue while 2,3,4,7,8-P5CDF largely persisted after the treatment period in both tissues. 2,3,7,8-T4CDF was eliminated even more rapidly than 1,2,3,7,8-P5CDF and could not be detected after treatment in both tissues. Time courses of the concentrations of 2,3,4,7,8-P5CDF, H6CDFs, H7CDFs and OCDF in liver and adipose tissue were similar: the levels of all congeners increased during the treatment period but no clear-cut decrease was observed within 34 days after the last treatment. Tissue concentrations of all 2,3,7,8-substituted PCDFs were higher in liver than in adipose tissue. The liver/adipose tissue concentration ratios increased with the degree of chlorination. The ratios of 2,3,7,8-T4CDF and 1,2,3,7,8-P5CDF were much lower than those of all other 2,3,7,8-substituted congeners. (3) non-2,3,7,8-substituted PCDDs and PCDFs: a number of non-2,3,7,8-substituted PCDD and PCDF congeners were found in both tissues in concentrations below 1 ng/g. In adipose tissue nearly all congeners were found during the treatment period showing a decrease after the treatment. In liver samples, many higher chlorinated PCDF congeners (with >4 chlorine atoms) could be detected. Most of those substituted in three of the four 2, 3, 7 and 8-positions persisted after treatment. In contrast, only one 1,4,6,9-substituted isomer of each PCDD homologue group was found during treatment with high recoveries after the third injection, but a rapid decline occurred already during the treatment period. (4) EROD activity: a good linear relationship (when using a double-log plot) between the EROD activities and the hepatic concentrations (ng I-TEq/g tissue) was found both in the PCDD-treated (r2=85.8%) and in the PCDF-treated group (r2=87.3%). A similar correlation (r2=95.6%) was observed in rats treated with 2,3,7,8-TCDD alone (concentration range in liver tissue: 0.2 to 9.7 ng/g wet weight). The concentration-response curves for both the PCDD and PCDF mixtures run parallel to the curve for 2,3,7,8-T4CDD. However, the inductive potency of the PCDD or PCDF mixture was approximately 3-fold or 4-fold lower, respectively, compared with the inductive potency of 2,3,7,8-T4CDD. Thus, the I-TE factors overestimated the potency of the mixtures in the concentration range tested and taking EROD induction as an end point.

Kinetics and inductive potency of 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (h7cdd) in rats

The kinetic properties and the inductive potency of 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (H7CDD) were studied in Wistar rats following subcutaneous (s.c.) injections. For assessing the dose-response, rats were treated with a single dose of 3. 10 or 30 microg H7CDD/kg body wt. Tissue concentrations and enzyme induction were measured 1, 2, and 3 weeks after treatment, and in the 30 microg/kg group additionally after 6, 20 and 57 weeks. Tissue concentrations increased dose-dependently from 3 to 30 microg/kg. Concentrations in liver were always higher than in adipose tissue, the concentration ratio: liver/adipose tissue varied between 32 and 67. The activity of (ethoxyresorufin O-deethylase) (EROD) in liver microsomes was clearly induced by H7CDD, reaching maximal induction three weeks after treatment. (3-fold at 3 microg/kg, 5-fold at 10 microg/kg and nearly 30-fold at 30 microg/kg). For assessing the time dependency, tissue levels and hepatic enzyme induction were monitored over a period of 57 weeks after a single s.c.-injection of 30 microg H7CDD/kg body wt. Hepatic concentrations of the congener remained rather constant from the 2nd to the 20th week after treatment (280 ng/g and 319 ng/g, respectively). In contrast, concentrations in adipose tissue and thymus increased 2-fold during this period, and 20 weeks after injection reached a maximum of 11 ng/g and 3 ng/g, respectively. Thereafter, the concentrations decreased and tissue levels of 91 ng/g (liver), 3 ng/g (adipose tissue) and 2 ng/g (thymus) were detected 57 weeks after treatment. The elimination half-life (t 1/2) calculated from our data was 140 days in liver and 130 days in adipose tissue. The reasonable explanation for the increase in tissue concentrations of H7CDD up to 20 weeks after treatment is the slow release of this congener from the subcutaneous injection site. Induction of hepatic EROD activity always closely followed changes in the hepatic concentrations of H7CDD, reaching a maximum 3 weeks after treatment and remaining at this level until the 20th week. Correlation analysis of hepatic H7CDD concentrations versus the extent of EROD induction indicated a linear relationship in a double-logarithmic plot. When compared with TCDD, the hepatic monooxygenase-inducing potency of H7CDD within the low dose range was found in the rat to be 170 to 440-times lower than that of TCDD. Measurement of 14C-caffeine demethylation, using a 14CO2 breath test, revealed a similar time course in vivo when compared with the microsomal EROD activity ex vivo.

Concentration-effect analyses with TCDD, H7CDD and OCDD in female Wistar rats

The induction of hepatic monooxygenases in female Wistar rats was measured after repeated subcutaneous injections of 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (H7CDD) and octachlorodibenzo-p-dioxin (OCDD), and the potency of these congeners to induce this hepatic enzyme activity was compared with that of TCDD. The biological effects were correlated with tissue concentrations of the congeners, and concentration-effect curves were established, allowing the assessment of a NOEC (no-observed-effect-concentration). A clear induction of ethoxyresorufin O-deethylase (EROD) activity in liver microsomes was observed after treatment with highly purified OCDD, but the inducing potency of this congener was clearly more than 10 times lower than that of H7CDD. When compared with TCDD our data suggest that the potency of H7CDD at the NOEL is 1700 of that of TCDD, whereas OCDD, at this biological level, seems to have an inducing potency of less than 110,000 of that of TCDD. Since it was found that the concentration-effect curves for the congeners do not run completely parallel, lower potency factors (when compared with TCDD) are calculated, namely: 100 for H7CDD, and 2,000 for OCDD. Our results explain some of the apparent discrepancies found in the literature.