Background, aim, and scope: Concentration monitoring as a basis for risk assessment is a valid approach only if there is an unambiguous relation between concentration and effect. In many cases, no such unambiguous relation exists, since various substances can exert the same effect with differing potencies. If some or all of these substances contributing to a biological effect are unknown, effect-related monitoring becomes indispensable. Endocrine-disrupting substances in water bodies, including the groundwater, are a prominent example of such a case. The aim of the investigations described here was to detect hormonally active substances in the groundwater downstream of obsolete landfills by using the E-screen assay and to possibly assign the biological effect to individual chemical compounds by means of instrumental analyses carried out in parallel.
Materials and methods: Grab samples of the groundwater were collected downstream from abandoned landfills and prepared by liquid/liquid extraction. The total estrogenic activity in these samples was determined in vitro by applying the E-screen assay. The human breast cancer cells (MCF-7) used in the E-screen proliferate in response to the presence of estrogenically active compounds. Expressed in concentration units of the reference substance 17beta-estradiol (E2), the test system allows the quantification of estrogenicity with a limit of detection (LOD) in the range of 0.1 ng/L. Aliquots of the samples were screened using gas chromatography/mass spectrometry (GC/MS) in order to quantify known estrogenically active substances and to identify unknown compounds. Estrogen-positive samples were extracted at different pH values, split into acidic, neutral, and basic fractions and analyzed by GC/MS, searching for individual components that display estrogenic activity.
Results and discussion: Estrogenic activity exceeding the LOD and the provisional benchmark of 0.5 ng E2/L was found at three out of seven abandoned waste disposal sites tested. The low concentrations of known xenoestrogens such as bisphenol-A, nonylphenols, or phthalic acid esters determined by GC/MS, however, were not sufficient to explain the detected activity. Neither natural nor synthetic hormones have caused the activity because these chemical structures are readily degradable and cannot persist in abandoned landfills for decades. The highest activity in the E-screen assay was found in the acidic fractions. Hydroxypolychlorinated biphenyls (PCBs), hydroxylated polycyclic aromatic hydrocarbons (PAHs) and hetero-PAHs, as well as alkylphenols could be identified as further compounds with possible hormonal activity.
Conclusions: Estrogenically active substances may occur in the groundwater below obsolete landfills, especially those that contain PCBs or waste from gasworks. These substances are not part of analytical programs routinely applied to contaminated sites and may therefore escape detection and assessment. Analyses using the E-screen assay and GC/MS in parallel have shown that the total estrogenic activity found in groundwater samples is to be ascribed to a multitude of individual compounds, some of which cannot be quantified due to lack of standard substances or assessed due to lack of a standardized procedure for determination of their estrogenic potency. By comparison with provisional guide values for estradiol (0.5 ng/L) and ethynylestradiol (0.3 ng/L), the damaging potential of the total estrogenic activity in groundwater samples can in fact be assessed, but specific remediation measures are impossible unless the hormonal activity can be attributed to individual chemical substances.
Recommendations and outlook: On the one hand, further analyses of samples taken from possible pollution sources should be conducted in order to characterize the extent of groundwater pollution with xenoestrogens. On the other hand, the most potent individual compounds should be identified according to their estrogenic potency. To this end, bioassay-directed fractionation and structure elucidation should be carried out with concentrated samples.