A four-hour yeast bioassay for the direct measure of estrogenic activity in wastewater without sample extraction, concentration, or sterilization

Abstract

The assay described here represents an improved yeast bioassay that provides a rapid yet sensitive screening method for EDCs with very little hands-on time and without the need for sample preparation. Traditional receptor-mediated reporter assays in yeast were performed twelve to twenty four hours after ligand addition, used colorimetric substrates, and, in many cases, required high, non-physiological concentrations of ligand. With the advent of new chemiluminescent substrates a ligand-induced signal can be detected within thirty minutes using high picomolar to low nanomolar concentrations of estrogen. As a result of the sensitivity (EC(50) for estradiol is approximately 0.7nM) and the very short assay time (2-4h) environmental water samples can typically be assayed directly without sterilization, extraction, and concentration. Thus, these assays represent rapid and sensitive approaches for determining the presence of contaminants in environmental samples. As proof of principle, we directly assayed wastewater influent and effluent taken from a wastewater treatment plant in the El Paso, TX area for the presence of estrogenic activity. The data obtained in the four-hour yeast bioassay directly correlated with GC-mass spectrometry analysis of these same water samples.

Figure 1

An illustration of the four-hour yeast bioassay protocol. A saturated overnight culture is diluted back to an O.D.₆₀₀ of 0.08 and allowed to reach log phase growth before the addition of sample. Wastewater samples that typically have higher levels of contaminants can be assayed directly by mixing with concentrated yeast medium and placed on the cells for 2 hours prior to substrate addition (left panel). The samples can be extracted and concentrated as with traditional yeast assays when ligand levels are below values needed for direct detection. However, in this assay, the cells only need to be treated with concentrated sample for 2 hours prior to substrate addition and, thus, the concentrated samples can still be assayed in a 4-hour time frame (right panel).

Figure 2

Four-hour yeast bioassay method comparisons. A) Representative 17β-estradiol dose-response curves in which the ligand solubilized in ethanol vehicle was added directly to the wells (open circles; Figure 1 right panel) or was diluted in water and treated the same as wastewater samples (closed squares, Figure 1 left panel) are shown. The resulting EC₅₀ values averaged from 3 independent assays±standard deviation were 1.45 × 10⁻¹⁰ M±0.1 for concentrated ligand and 1.51 × 10⁻¹⁰ M±0.1 for ligand diluted in water. B) Water samples taken from the same wastewater treatment plant were either assayed directly without any preparation or assayed after filtration to remove particulates and filter sterilization through a 0.2 μm filter. Effluent “Chemical” represents samples that were chemically treated prior to sample collection. Effluent “Biological” represents samples that were both chemically treated and biologically filtered prior to sample collection (see Materials and Methods for a detailed description of these treatment processes). Each data point is an average of three replicates with error bars representing standard deviation. The data obtained from the filtered samples were significantly reduced as compared to the non-filtered samples with a p value < 0.001. The inset is a 17β-estradiol positive control shown at 0 and 0.1 nM.

Figure 3

Dose-response curves for typical estrogenic wastewater contaminants using the four-hour yeast bioassay. Representative dose-response curves for the indicated ligands using the four-hour yeast bioassay are shown. The ligands used include 17β-estradiol (closed circle), 17α-ethynylestradiol (closed upside down triangle), estrone (closed upright triangle), nonylphenol (open square), and bisphenol A (open upside down triangle). All data points are averages of 3 independent replicates with error bars representing standard deviation.

Figure 4

A correlation between EEQ_GCMS and EEQ_yeast. A second order polynomial regression analysis reveals a positive correlation between the estrogenic activity determined by direct biological measurement (EEQyeast) and by chemical analysis (EEQ_GCMS). EEQ_GCMS is the calculated overall estrogenicity based on the concentration of individual compounds detected and their relative potencies. EEQ_yeast, is the measured estrogenicity determined by the yeast bioassay.