Toxic masking and synergistic modulation of the estrogenic activity of chemical mixtures in a yeast estrogen screen (YES)

Background, aim and scope

Estrogenic and non-estrogenic chemicals typically co-occur in the environment. Interference by non-estrogenic chemicals may confound the assessment of the actual estrogenic activity of complex environmental samples. The aim of the present study was to investigate whether, in which way and how seriously the estrogenic activity of single estrogens and the observed and predicted joint action of estrogenic mixtures is influenced by toxic masking and synergistic modulation caused by non-estrogenic chemical confounders.

Materials and methods

The yeast estrogen screen (YES) was adapted so that toxicity and estrogenicity could be quantified simultaneously in one experimental run. Mercury, two organic solvents (dimethyl sulfoxide (DMSO) and 2,4-dinitroaniline), a surfactant (LAS-12) and the antibiotic cycloheximide were selected as toxic but non-estrogenic test chemicals. The confounding impact of selected concentrations of these toxicants on the estrogenic activity of the hormone 17ß-estradiol was determined by co-incubation experiments. In a second step, the impact of toxic masking and synergistic modulation on the predictability of the joint action of 17ß-estradiol, estrone and estriol mixtures by concentration addition was analysed.

Results

Each of the non-estrogenic chemicals reduced the apparent estrogenicity of both single estrogens and their mixtures if applied at high, toxic concentrations. Besides this common pattern, a highly substance- and concentration-dependent impact of the non-estrogenic toxicants was observable. The activity of 17ß-estradiol was still reduced in the presence of only low or non-toxic concentrations of 2,4-dinitroaniline and cycloheximide, which was not the case for mercury and DMSO. A clear synergistic modulation, i.e. an enhanced estrogenic activity, was induced by the presence of slightly toxic concentrations of LAS-12. The joint estrogenic activity of the mixture of estrogens was affected by toxic masking and synergistic modulation in direct proportion to the single estrogens, which allowed for an adequate adaptation of concentration addition and thus unaffected predictability of the joint estrogenicity in the presence of non-estrogenic confounders.

Discussion

The modified YES proved to be a reliable system for the simultaneous quantification of yeast toxicity and estrogen receptor activation. Experimental results substantiate the available evidence for toxic masking as a relevant phenomenon in estrogenicity assessment of complex environmental samples. Synergistic modulation of estrogenic activity by non-estrogenic confounders might be of lower importance. The concept of concentration addition is discussed as a valuable tool for estrogenicity assessment of complex mixtures, with deviations of the measured joint estrogenicity from predictions indicating the need for refined analyses.

Conclusions

Two major challenges are to be considered simultaneously for a reliable analysis of the estrogenic activity of complex mixtures: the identification of known and suspected estrogenic compounds in the sample as well as the substance- and effect-level-dependent confounding impact of non-estrogenic toxicants.

Recommendations and perspectives

The application of screening assays such as the YES to complex mixtures should be accompanied by measures that safeguard against false negative results which may be caused by non-estrogenic but toxic confounders. Simultaneous assessments of estrogenicity and toxicity are generally advisable.