Environmental Engineering Reference
In-Depth Information
experiments, cells were plated on 96- or 384-well plates, and after 30-min treatment followed
by ixation, stained with DRAQ5 (BioStatus Limited) for 15 min. Cells were either immediately
imaged on the Perkin Elmer OPERA imaging system, or kept in phosphate-buffered saline
at 4°C for later imaging. An algorithm was customized using the Acapella image analysis
software development kit (Perkin Elmer) to automatically segment both the nucleus and cyto-
plasm of each cell in the digital micrographs. The algorithm also calculated the mean GFP-GR
or GFP-AR intensity in both compartments, and translocation was measured as a ratio of
these intensities. Each value was further normalized to the value for the control (dimethyl
sulfoxide, DMSO) sample. A graph demonstrating the concentration-dependent GFP-GR
translocation to the nucleus in response to synthetic glucocorticoid (dexamethasone), human
glucocorticoid (hydrocortisone), and rodent glucocorticoid (corticosterone) is shown in Figure
35.3c. Conident in the sensitivity of this assay, we tested >100 water samples collected from 14
states in the United States for glucocorticoid and androgen activities. We discovered a previ-
ously unrecognized glucocorticoid activity in 27% of the samples (Figure 35.4) and androgen
activity in 35% of the samples (Figure 35.5). Some of these samples were discrete “grab” water
samples (circles) and the others were collected by polar organic chemical integrative samplers
(POCIS) deployed in the water at the collection sites over a period of time (triangles).
Contamination of water sources with androgen/anti-androgen activity was previously dem-
onstrated in the Netherlands (van der Linden et al., 2013), United Kingdom (Kirk et al., 2002),
New Zealand (Leusch et al., 2006), Denmark (Kusk et al., 2011), and China (Chang et al., 2009a).
Contamination of water sources with glucocorticoids and its negative effects on ish health
were also reported (Chang et al., 2007, 2009a; Schriks et al., 2010; Kugathas and Sumpter, 2011).
The major advantages of the translocation assay over other previously described assays
are that they are rapid, quantitative, inexpensive, and compatible with high throughput
screening. Receptor translocation is completed in 20-30 min, and multiple cell lines or
multiple color-tagged receptors in the same cell line can be tested and utilized simul-
taneously. In contrast to the chemical methods, this assay does not reveal the chemical
structure of the active contaminants. Instead, it detects biologically active EDCs capable of
interacting with a speciic nuclear receptor (in this case, GR or AR). To identify a speciic
or multiple chemical structures responsible for these effects, we tested one sample (SS97)
positive in the translocation assay by chemical detection methods to perform “forensic
chemistry.” Sample SS97 tested positive in the GFP-GR translocation assay (Figure 35.6a)
and was also capable of inducing transcriptional response from a GR-regulated gene,
Per1
.
To identify the active glucocorticoid contaminant(s), we analyzed SS97 fractions of the HPLC
retention on a C18 column (Chang et al., 2009a) for GFP-GR translocation (data not shown).
Positive fractions were further analyzed by ultra-performance liquid chromatography/MS
and GC/MS (Mansilha et al., 2010). Resulting mass spectra were searched extensively in
the National Institute of Standards and Technology/Environmental Protection Agency/
National Institutes of Health Mass Spectral Library, and in the Wiley Mass Spectra Database
of Androgens, Estrogens, and other Steroids 2010. In spite of this detailed analysis, we found
no evidence of any known glucocorticoids in this sample. However, comparison of the mass
spectra of chromatographic peaks 1-3 (Figure 35.6c) with standard spectra from the Atomic
Emission Spectroscopy (AES) 2010 database suggested similarities to known androstane-
class compounds that are potentially capable of activating AR. One of these compounds,
androst-4-en-3,6-dione (peak 2), was synthesized (Hunter and Priest, 2006) and its transloca-
tion activity tested in cell lines expressing GFP-GR or GFP-AR. We conirmed that androst-4-
en-3,6-dione induced GFP-AR (Figure 35.6d) but not GFP-GR translocation (data not shown).
Some of the EDCs are resistant to a complete biodegradation and persist in the eflu-
ents of the WWTPs (Chang et al., 2009b). Thus, current wastewater treatment methods
