Non-targeted detection and differentiation of agonists versus antagonists, directly in bioprofiles of everyday products.

Xenoestrogens exert antiandrogenic effects on the human androgen receptor. In the analytical field, such antagonists block the detection of testosterone and falsify results obtained by sum parameter assays. Currently, such agonistic versus antagonistic effects are not differentiated in complex mixtures. Oppositely acting hormonal effects present in products of everyday use can only be differentiated after tedious fractionation and isolation of the individual compounds along with subjection of each fraction/compound to the status quo bioassay testing. However, such long-lasting procedures are not suited for routine. Hence, we developed a fast bioanalytical tool that figures out agonists versus antagonists directly in complex mixtures. Exemplarily, 8 cosmetics and 15 thermal papers were analyzed. The determined antagonistic potentials of active compounds found were comparable to the ones of known antagonists (in reference shown for bisphenol A, 4-n-nonylphenol and four parabens). Relevant biological/chromatographic parameters such as cell viability, culture conditions, dose response curves, limits of biological detection/quantification and working range (shown for testosterone, dihydrotestosterone, nandrolone and trenbolone) were investigated to obtain the best sensitivity of the biological detection. The developed and validated method was newly termed reversed phase high-performance thin-layer chromatography planar yeast ant-/agonistic androgen screen (RP-HPTLC-pYAAS bioassay). Results were also compared with the RP-HPTLC-Aliivibrio fischeri bioassay (applied on RP plates for the first time). As proof-of-concept, the transfer to another bioassay (RP-HPTLC-pYES) was successfully demonstrated, analogously termed RP-HPTLC-pYAES bioassay detecting anti-/estrogens (exemplarily shown for evaluation of 4 pharmaceuticals used in breast cancer treatment). The new imaging concept provides (1) detection and differentiation of individual agonistic versus antagonistic effects in the bioprofiles, (2) bioanalytical quantification of their activity potential by scanning densitometry and (3) characterization of unknown bioactive compound zones by hyphenation to high-resolution mass spectrometry. Depending on the hormonal bioassay, 15 samples were analyzed in parallel within 5 h or 6 h (calculated as 20 or 24 min per sample). For the first time, piezoelectric spraying of the yeast cells was successfully demonstrated for the planar yeast-based bioassays.