2007 Progress Report: Development of Receptor- to Population-Level Analytical Tools for Assessing Endocrine Disruptor Exposure in Wastewater-Impacted Estuarine SystemsEPA Grant Number: R832737
Title: Development of Receptor- to Population-Level Analytical Tools for Assessing Endocrine Disruptor Exposure in Wastewater-Impacted Estuarine Systems
Investigators: Ferguson, P. Lee , Chandler, G. Thomas
Institution: University of South Carolina at Columbia
EPA Project Officer: McOliver, Cynthia
Project Period: January 1, 2006 through December 31, 2009 (Extended to December 31, 2010)
Project Period Covered by this Report: January 1, 2007 through December 31,2007
Project Amount: $526,028
RFA: Exposure Measurement Tools for Endocrine Disrupting Chemicals in Mixtures (2005) RFA Text | Recipients Lists
Research Category: Endocrine Disruptors , Health , Safer Chemicals
The research supported by this STAR grant is primarily aimed at developing and validating novel analytical tools for identifying and quantifying receptor-active endocrine disrupting chemicals in complex environmental samples. In addition, we are linking these new bioanalytical tools to sensitive in-vivo bioassays for assessing the potential for exposure of aquatic organisms to wastewater-derived endocrine disrupting chemicals. Specific objectives of the project are to:
- Develop nuclear hormone receptor-affinity extraction techniques as tools for isolating endocrine disrupting chemicals (EDCs) from complex wastewater mixtures.
- Apply these methods in combination with high performance mass spectrometry for activity-directed analysis of EDCs in wastewater and estuarine receiving waters on the SC coast.
- Utilize sensitive vertebrate (zebrafish) and invertebrate (copepod) EDC-exposure laboratory bioassays to link exposure measurements (above) to biological effects.
- Apply novel biomolecular endpoints to assess EDC exposure in field populations of sensitive meiobenthic invertebrates in wastewater-impacted estuarine environments.
Initial work on specific objective 1 of this project has focused on the cloning, expression, and purification of human estrogen receptor ligand binding domain (ER-LBD) and human androgen receptor ligand binding domain (AR-LBD) in bacterial vectors for use in developing receptor-affinity extraction methods suitable for isolation of receptor-active xeno-ligands from complex wastewater environments. This work has been necessary, as it was found to be impractical to use commercially-available, recombinant nuclear hormone receptors (e.g. ERα and ERβ from Invitrogen) as affinity media for capturing EDCs from complex media (as proposed in the original work plan). Specifically, these receptors, while they have very high affinity for the xenoestrogens of interest, were found to be very susceptible to environmental conditions, losing ligand-binding activity rapidly upon addition to complex wastewater extracts. We have found that recombinant ER ligand binding domain (as produced and purified in our laboratory) is much more stable and exhibits nearly identical ligand binding activity to the full-length receptor when tested in vitro (similar to findings of other researchers).
Figure 1. Purified recombinant ERα-LBD in fusion with his6-thioredoxin (blue circle in lane 2).
To date, we have successfully expressed and purified active ER-LBD (in fusion with his6 and thioredoxin) in ~10 mg L-1 yield and > 95% purity from bacterial expression vectors (Figure 1). The receptor thus purified is active to bind [3H]-17β-estradiol in solution and has been used in subsequent receptor-affinity extraction experiments (below). Expression and purification of AR-LBD has been pursued using a GST-fusion protein construct in bacterial expression host. This purification has been problematic due to poor solubility of the overexpressed protein in lysis buffers. We are currently attempting to purify this receptor using denaturing conditions with the goal of re-folding and solubilizing the protein during size-exclusion chromatographic purification.
The initially proposed strategy for performing receptor-affinity extraction of endocrine disruptors from wastewaters and surface waters was to bind the purified, recombinant receptors covalently to activated glass beads in order to create a solid-phase affinity support over which EDC-containing extracts could be purified. This strategy had worked well in our previous work with immunoaffinity extraction methods, but first experiments using purified ER-LBD indicated that dramatic loss of ligand-binding activity resulted from covalent attachment of the protein to controlled-pore glass beads. An alternate strategy was employed by which purified ER-LBD (containing a his6 motif engineered into the expression vector) was added to wastewater extracts and allowed to bind xenoestrogens. The receptor (still bound to ligands) was then isolated by immobilized metal affinity chromatography using an FPLC, and the bound ER-LBD/xenoestrogen complexes were eluted using an imidazole step gradient. HPLC-MS/MS analysis of natural and synthetic steroid estrogens was then performed on the eluent. The new analytical strategy is illustrated below in Figure 2.
Figure 2. Analytical strategy for isolation and analysis of xenoestrogens from wastewater using receptor-affinity extraction coupled with HPLC-MS/MS analysis
We have thus far successfully applied the purified ER-LBD to isolation of ER ligands from treated wastewater by the receptor-affinity extraction methods described above. In combination with mass spectrometric (HPLC-MS/MS and HPLC-QTOF-MS) methods, natural estrogens (including estrone and 17β-estradiol) have been identified and quantified in ER-affinity extracts of several wastewater samples obtained from coastal and freshwater discharge regimes within South Carolina. Specifically, we have collected and analyzed biologically-treated wastewater from Kiawah Island wastewater treatment plant, which generates approximately 1.1 million gallons per day (MGD) of treated municipal wastewater, the majority of which is applied as irrigation water to adjacent golf courses. Following the methods outlined in Figure 2, we have utilized a nonspecific solid-phase extraction method to generate organic contaminant extracts from Kiawah Island WWTP effluent and receiving waters and have applied estrogen receptor-affinity extraction to isolate component xenoestrogens. The ER-LBD remained active in the wastewater extract, and [3H]-17β-estradiol tracer added to the wastewater extract was co-purified with the protein by immobilized metal affinity chromatography (Figure 3). After elution from the FPLC system, the receptor-affinity extract was assayed for the presence of the female sex hormones 17β-estradiol (E2) and estrone (E1) as well as the synthetic contraceptive ethynylestradiol (EE2) by HPLC-MS/MS. In keeping with the quality-assurance practices stated in the proposed plan of work, these determinations were made using published analytical methods and all samples were amended with stable-isotope-labeled (deuterated) surrogate standards prior to chemical- and receptor affinity extractions.
Figure 3. Purification of [3H]-17β-estradiol from Kiawah Island, SC wastewater effluent using estrogen receptor-affinity extraction. The receptor and radiotracer coelute (fractions 21-23) after extraction from the wastewater onto a nickel-agarose column and subsequent elution using an imidazole step-gradient.
Representative results of EE2, E2 and E1 analyses are illustrated in Figure 4. EE2 was not detected in any wastewater extract analyzed thus far. Estrogens (E1 and E2) and their deuterated surrogate standard analogs were detected in Kiawah wastewater extracts isolated by receptor affinity extraction, illustrating the suitability of this approach for isolating receptor-active endocrine disruptors from highly complex environmental sample matrices at trace levels. Quantitative analyses of these estrogen compounds in the same wastewaters were performed in parallel using HPLC-MS/MS without receptor-affinity extraction. Concentrations of E1 and E2 measured in these samples ranged from <1 - 30 ng L-1, with surrogate standard extraction efficiencies varying from 30 – 100 % (depending on the sample matrix and solid-phase adsorbent used). Overall, results from this initial work (addressing specific objectives 1 & 2 from the proposed research) are highly promising and indicate that the receptor-affinity extraction methods being developed will prove very useful for identifying and quantifying novel EDC compounds in wastewater-impacted aquatic environments of interest to EPA.
Figure 4. Detection of estrogens and their deuterated surrogate analogs in ER-LBD-containing fractions from Figure 3 (A). No estrogens were detected in a "control" fraction from the FPLC purification, which did not contain ER-LBD (B).
In parallel to the studies above, we have applied sensitive, EDC-specific molecular bioassays to determine the effects of ecdysteroid-receptor active compound exposure on the harpacticoid copepod Amphiascus tenuiremus, in support of objectives 3 and 4. Specifically, we have performed microplate-based, full life-cycle screening assays (ASTM E 2317-04) using the insect growth-regulating pesticide tebufenozide and the fungicide fenarimol. No biologically significant effects were observed on A. tenuiremis growth, development, and reproduction following exposure to 0.005, 0.05, and 0.5 mg/L fenarimol, however exposure did cause a significant decrease in copepod ecdysteroid titers for both male and female copepods at all doses (Figure 5) , suggesting that fenarimol either inhibits ecdysteroid biosynthesis or stimulates an acceleration of hormone metabolism. Tebufenozide exposure accelerated naupliar development, increased male abundance, and reduced female offspring production at 0.05, 0.5, and 2-mg/l exposure, respectively; however, no biologically-significant alterations of ecdysteroid titers were measured in copepods under laboratory exposure of copepods to tebufenozide.
Figure 5. Ecdysteroid titers in adult virgin Amphiascus tenuiremis males and females after being reared from nauplii in fenarimol. Error bars = ± SD. * denotes treated organisms were significantly different from the controls.
Future work will focus on extending the receptor-affinity extraction methodology to include the human androgen receptor (expression and purification in progress) and the mysid ecdysone-receptor/ultraspiracle heterodimer (expression plasmid received). Method validation activities will include standard-addition experiments with known or suspected EDCs in wastewater extracts as well as targeted quantitative analysis of these compounds in wastewater and receiving waters on the SC coast. Identification of novel receptor-active EDCs will be facilitated by the application of high-resolution HPLC-QTOF-MS to receptor-affinity extracts from these samples. In addition, we are extending our suite of EDC bioassays to include a sensitive vertebrate xenoestrogen assay (vitellogenesis and sex-ratios in medakafish embryos) for testing of EDCs identified in wastewater using the receptor-affinity extraction approach outlined above.
There have been no significant changes in the number or category of key personnel involved in the project. Dr. Ferguson is primarily responsible for project management, and oversees work on the development of receptor-affinity extraction methods. Dr. Chandler is responsible for managing work on the invertebrate bioassay method development. Both of the co-PI’s are involved in data interpretation and experimental design, and they meet at least weekly to discuss the project and progress toward meeting goals and milestones. Dr. Paul Thompson is serving an advisory role as specified in the grant proposal – his role is primarily in advising on protein purification methods for generating active nuclear hormone receptors for use in receptor-affinity extractions. The following junior personnel are supported on the project:
Boyd Pritchard (M.S. candidate, Dept. of Chemistry & Biochemistry) – anticipated graduation date: December 2007
Lauren Kimberley (Ph.D. student, Dept. of Chemistry & Biochemistry) – anticipated graduation date: August 2009
Shosaku Kashiwada, Ph.D. (Research Assistant Professor, Dept. of Environmental Health Sciences) – partially supported by this project to develop medaka and copepod bioassay methods
QA/QC procedures for all experiments conducted within the scope of this project follow the outline presented in the initial grant proposal. In general, data interpretation and presentation of results is based upon rigorous statistical analysis of all original data and experimental design is supported/justified by previously published literature whenever possible. The comparability of multiple datasets has been tested prior to statistical comparisons using Shapiro-Wilk's test of normality, and Bartlett's test of sphericity for homogeneous variance. In cases where data were found to be non-normal or non-homogeneous (heteroscedastic), then appropriate data transformations have been performed for correction (e.g. Log10, angle arcsine sq.root, reciprocal, etc.). Statistical analyses utilize a variety of statistical procedures: Differences among treatments and replicate blocking variables are determined by SAS GLM ANOVA, Tukey's or Dunnett's multiple comparison tests, orthogonal contrasts, and, where necessary, non-parametric Mann-Whitney or Kruskal-Wallace analyses. Multivariate analyses utilize orthogonal and non-orthogonal principal components and factor analyses, and stepwise multiple regressions where appropriate to show strengths of association. A minimum significance level of p < 0.05 has been used in all hypotheses tests.
For all analytical determinations, appropriate calibration curves are generated and method/sample blanks and positive controls are incorporated into the analytical procedure. Mass spectrometers are externally and internally mass calibrated using standard reference masses (e.g. NaTFA clusters for electrospray Q-TOF MS). For elemental composition analysis by time-of-flight mass spectrometry, mass measurement accuracy is maintained within stated instrumental limits (5 ppm RMS, internally calibrated) and validated using authentic standards. Detection limits for all analytes are established as the concentration giving an instrument response equivalent to triple the standard deviation of sample blank analyses. All analytical procedures follow EPA established protocols or previously published methods when possible and are tested for extraction efficiency and total recovery. Stable isotope-labeled internal and surrogate standards have chosen for quantitative reference where available in order to maximize precision of the analyses in HPLC-MS/MS. Precision is maintained within ASTM guidelines wherever available. Sample handling, identification, and transport follow chain-of-custody guidelines. Wastewater, irrigation runoff, and surface water sampling utilize muffle-furnace baked (450° C) glassware, and sample holding times before analyses/bioassays do not exceed 24 h at 4°C. All samples that are transported lab-to-lab are sealed on ice paks for periods not to exceed 24 h. All contaminant chemical analyses utilize standard QA/QC procedures including analysis of blanks, matrix spikes, duplicates and periodic matrix standard addition experiments.