Ecosystems Research Division

QSAR Evaluation of ER Binding Affinity of Chemicals and Metabolites. R. Serafimova, H. Aladjov, R. C. Kolanczyk, P. K. Schmieder, Y. Akahori, W. J. Jones, AND O. Mekenyan.  Society of Environmental Toxicology and Chemistry Annual Meeting, Baltimore, MD, November 13 - 17, 2005.

Chemicals in commerce are assessed for a variety of potential adverse effects. As governments around the globe strive to meet the challenge of assessing chemicals as endocrine disruptors, the need for hypothesis-driven strategies to prioritize chemicals for testing has risen to the forefront. As part of a larger research effort, this study describes quantitative structure-activity relationships (QSAR) to predict chemical ER binding using an iterative process of strategic selection of chemicals for testing, in vitro data generation, in silico model development, etc., to facilitate model evaluation and refinement for user-specified levels of predictive certainty. A multi-dimensional formulation of a COmmon REactivity PAttern (COREPA) modeling approach has been used to investigate chemical binding to the human estrogen receptor (hER).  A training set of 656 chemicals includes 500 steroid and 156 environmental chemicals (CERI) to further explore hER-structure interactions. Analysis of reactivity patterns based on the distance between nucleophilic sites resulted in identification of distinct interaction types: a steroid-like A-B interaction described by frontier orbital energies and distance between nucleophilic sites with specific charge requirements; an A-C interaction where steric effects are combined with electronic interactions to modulate binding; and mixed A-B-C. Chemicals are grouped by interaction type, then COREPA models are developed for specific relative binding affinity ranges of >10%, 10 to 0.1%, and 0.1 to 0.001%.  The derived models for each interaction type and affinity range incorporate specific interatomic distances and a COREPA classification node using < 2 discriminating parameters. The interaction types become less distinct in the lowest activity range for each chemical of each type.  A battery of models is presented for pre-screening ER binding affinity of parent chemicals, or simulated metabolites (i.e., interfacing of  toxic effects predictions with a liver tissue metabolism simulator (TIMES) as described in a companion presentation).

Determination of Perfluorinated Chemicals (PFCs) in Soils, Sediment and Other Matrices.  J. J. Ellington, J. W. Washington, M. J. Strynar, J. J. Evans, T. Jenkins, and W. M. Henderson. Society of Toxicology and Chemistry Annual Meeting, Baltimore, MD, November 13 - 17, 2005.

Perfluorinated chemicals (PFCs) serve either as the building-block monomers for commercial fluorotelomer-based polymers (FBPs) or the monomers may be applied directly to products to impart fire-protection or non-stick properties (e.g., carpets). Soils that receive land application of treated wastewater from carpet manufacturing plants were sampled for perfluorooctanoic acid (PFOA) and other PFCs that are potential degradation products of FBPs. Soils from other potentially impacted sites and soils remote from exposure to PFCs (except by atmospheric deposition) were analyzed to determine ambient levels of PFCs as well.  Large hysteric effects in sorption can impact interpretation of spike-and-recovery quality-assurance practices.  To determine the extent of soil-sorption hysteresis for PFOA, dried and sieved remote soil was fortified with PFOA, which is considered to be the stable degradant end-point for many PFCs, and the soil rehydrated to typical moisture conditions. Triplicate samples were extracted at 3-week intervals over three months. We report PFC concentrations measured in selected soils and recovery of PFOA spiked in soil through time.

Increased Endocrine Activity of Xenobiotic Chemicals as Mediated By Metabolic Activation.  R. C. Kolanczyk, M. A. Tapper, B. Nelson, V. Wehinger, J. S. Denny, D. W. Kuehl, B. R. Sheedy, C. S. Mazur, J. F. Kenneke, W. J. Jones, AND P. K. Schmieder.  Society of Environmental Toxicology and Chemistry Annual Meeting, Baltimore, MD, November 13 - 17, 2005.

EPA is faced with long lists of chemicals that need to be assessed for hazard. This research is part of a larger effort to develop in vitro assays and QSARs applicable to untested chemicals on EPA inventories through study of estrogen receptor (ER) binding and estrogen mediated gene expression in fish. The current effort investigates metabolic activation of chemicals resulting in increased estrogenicity.  Phenolphthalin (PLIN) was shown not to bind trout ER in a competitive binding assay but vitellogenin expression was induced in trout liver slices exposed to 10^-4 and 10^-3.7  M PLIN. Phenolphthalein (PLEIN), a metabolite of PLIN, was subsequently determined to be formed when slices were exposed to PLIN.  PLEIN binds rtER with a relative binding affinity (RBA) of  0.020%. Slices exposed to PLEIN expressed vitellogenin mRNA at 10^-4.3 ,10^-4 , and 10^-3.7  M, with no detectable PLIN present. Thus, vitellogenin expression noted in PLIN slice exposures was explained by metabolism to PLEIN in trout liver slices. A second model chemical, 4,4’-diaminodiphenylmethane (MDA) was not shown to bind rtER, but did induce vitellogenin mRNA production in tissue slices at 10^-4.3 ,10^-4 , and 10^-3.7 M in amounts nearly equal to reference estradiol induction, thus indicating metabolic activation of MDA. A series of experiments was performed to identify potential metabolites responsible for the observed increase in activity. Nitroso-MDA, hydroxyl-MDA, azo-MDA, and azoxy-MDA were not observed. Acetylated-MDA was observed and tested in both ER-binding and tissue slice vitellogenin induction assays. Comparisons of Phase I metabolic activation in trout and rat liver microsomes is also presented to elucidate cross species similarities and to enhance development of predictive models.

Design of Biodegradation Experiments for Fluorotelomer-Based Polymers. J. W. Washington, J. J. Ellington, T. M. Jenkins, J. J. Evans, and W. M. Henderson. Society of Toxicology and Chemistry Annual Meeting, Baltimore, MD, November 13 - 17, 2005.

Fluorotelomer-based polymers (FBPs) are used in a wide variety of consumer products and are widely distributed throughout society.  Accordingly, there is great interest in whether and how fast these materials might degrade in various environmental settings.  Known and expected qualities of FBPs and fluorinated monomers, as well as extraction and analytical considerations, contribute challenges in logical design of FBP-biodegradation studies to avoid artifacts.  Microbial enzyme induction for degradation of FBPs, should it occur, could be long relative to typical periods for biodegradation experiments; consequently, it might be advisable to collect and test acclimated as well as pristine media.  A useful quality of FBPs is the chemical stability of the polymers and, therefore, half-lives are anticipated to be long relative to typical laboratory experiments.  Failure to account for this stability in experimental design can affect results by: i) yielding unsatisfactory resolution of reaction rate to quantitate degradation constants; and ii) ingrowth of potential degradation products could be insufficient to determine whether the products are from leaching of residual monomers in the FBP or microbially mediated breakdown of the FBP structure.  Consequently, sufficient microcosms, resources and time should be allotted to allow extended incubations.  The compatibility of potential perfluorinated degradation products with glass, Teflon and other common experimental materials is uncertain with respect to irreversible sorption, diffusivity and extraneous sources of perfluorinated compounds; consequently, experimental materials must be tested for compatibility with potential degradation products.  Maintenance of the air-tight integrity of reaction vessels is critical so that volatile potential degradation products will not diffuse from the containers over long experiments; this is especially relevant considering possible constraints on acceptable materials for sealing containers.  Consequently, integrity of microcosm containers must be confirmed experimentally.  We report: 1) how these and similar factors have affected our experimental design; and 2) on our early experimental efforts.

Metabolism and Metabolic Activation of Chemicals: in-Silico Simulation. O.G. Mekenyan, S. D. Dimitrov, T. S. Pavlov, W. J. Jones, and P. K. Schmieder. 3rd International Symposium on Computational Methods in Toxicology and Pharmacology Integrating Internet Resources, Shanghai, China, October 29 - November 01, 2005.

The role of metabolism in prioritizing chemicals according to their potential adverse health effects is extremely important because innocuous parents can be transformed into toxic metabolites.  This work presents the TIssue MEtabolism Simulator (TIMES) platform for simulating metabolism. TIMES uses a heuristic algorithm to generate plausible metabolic maps from a comprehensive library of biotransformations and abiotic reactions with estimates for system-specific transformation probabilities.  An advantage of TIMES is the use of probabilities of occurrence to prioritize individual transformations thus defining the order of their implementation within the simulator.  This capability allows better control over the propagation of metabolic maps, and prioritization of metabolites according to their amount, reactivity, solubility, toxicity, etc. Thus a major disadvantage in previous attempts to model metabolism is reduced, namely the prediction of large numbers of possible but not probable metabolites. The reliability of simulated pathways, metabolites, and maps was assessed based upon the agreement with observed metabolism data. Hence, as with metabolites, competing metabolic pathways can be prioritized by their probability of occurrence and reliability. The reliability estimates also facilitate the strategic selection of chemicals for testing to most effectively expand the domain of the simulator.  A procedure for improving the reliability of metabolism simulators is introduced.  The ability of TIMES to interface the prediction of chemical metabolism and toxicity due to metabolic activation is an important advantage of the method.  It allows prioritization of chemicals based on predicted toxicity of metabolites. Applicability of the method to prioritize chemicals as potential skin sensitizers or mutagens resulting from their metabolic activation is illustrated.

Characterizing Spatial and Temporal Dynamics: Development of a Grid-Based Watershed Mercury Loading Model.  T. Dai, R. B. Ambrose Jr., K. Alvi, T. A. Wool, H. Manguerra, M. Chokshi, H. Yang, and S. R. Kraemer. 2005 Watershed Management Conference, Williamsburg, VA, July 20-23, 2005.

A distributed grid-based watershed mercury loading model has been developed to characterize spatial and temporal dynamics of mercury from both point and non-point sources. The model simulates flow, sediment transport, and mercury dynamics on a daily time step across a diverse landscape. The model is composed of six major components: (1) an ArcGIS interface for processing spatial input data; (2) a basic hydrological module; (3) a sediment transport module; (4) a mercury transport and transformation module; (5) a spreadsheet-based model post-processor; and (6) links to other models such as WASP and WhAEM 2000 developed by the U.S. Environmental Protection Agency (U.S. EPA). The model fully uses the grid processing capacity of the latest ArcGIS technology. The water balance, sediment generation and transport, and mercury dynamics are calculated for every grid within a watershed. Water and pollutants are routed daily throughout the watershed based on a unique and flexible algorithm that characterizes a watershed into many runoff travel-time zones. The mercury transport and transformation module simulates the following key processes: (1) mercury input from atmospheric deposition; (2) mercury assimilation and accumulation in forest canopy and release from forest litter; (3) mercury input from bedrock weathering; (4) mercury transformation in soils; (5) mercury transformation in lakes and wetlands including reduction and net methylation; (6) mercury transport through sediment and runoff; and (7) mercury transport in stream channels. By using the grid-based technology, flow and mercury dynamics can be examined at any of several points in the watershed. The model is capable of supporting large-scale watershed modeling with high-resolution raster datasets and will be used in mercury research projects sponsored by U.S. EPA. The model is programmed in Visual Basic and requires two ArcGIS (version 9.0) components—ArcView 9 and the Spatial Analyst extension.

Toxicity-Based Identification of Drinking Water Disinfection By-Products Using LC/MS and LC/MS/MS.  S. D. Richardson. First International Workshop on Liquid Chromatography-Tandem Mass Spectrometry for Screening and Trace Level Quantitation in Environmental and Food Samples, Barcelona, Spain, June 20 - 21, 2005.

Drinking water disinfection by-products (DBPs) are of concern because some epidemiologic studies have shown that some DBPs are associated with cancer or adverse reproductive/developmental effects in human populations, and other studies have shown that certain DBPs cause similar health effects in laboratory animals.  As a result, the U.S. EPA has regulated several DBPs.  However, more than 500 DBPs have been reported in drinking water for which little or no occurrence and health data exist.  In addition, there is almost no information on high molecular weight DBPs (>1000 Da), which are indicated to comprise >50% of the total organic halide (TOX) from chlorinated drinking water.  None of these high molecular weight halogenated by-products have ever been precisely identified, and there is no information on potential toxicity of this high molecular weight fraction.  There is also very little information regarding highly polar DBPs, which are missed by most commonly used analytical techniques.  The goal of this research is to use a bio-assay directed approach to focus identification work on the most toxicologically important DBPs.  To this end, drinking water is being collected from full-scale treatment plants that use chlorine, ozone, chlorine dioxide, and chloramines as disinfectants, and this drinking water is being fractionated according to polarity (through the use of different polarity XAD resins, preparatory liquid chromatography columns, and solvent gradients) and molecular size (through the use of ultrafiltration and size exclusion chromatography).  Molecular weight fractions to be studied include <1000, 1000-3000, 3000-5000, 5000-10,000, and >10,000 Da.  The bioassays used include a mammalian cell genotoxicity/cytotoxicity assay and a transgenic medaka fish genotoxicity assay.  Gas chromatography (GC) with high resolution mass spectrometry (MS) is being used to identify the lower molecular weight, non-polar drinking water DBPs, and liquid chromatography (LC)/MS and LC/MS/MS is being used to identify high molecular weight and highly polar DBPs.

Measurement and Toxicity of Iodo-Acid Disinfection By-Products in Chloraminated Drinking Water.  S. D. Richardson, J. J. Ellington, F. G. Crumley, J. J. Evans, M. J. Plewa, and E. D. Wagner.  American Society for Mass Spectrometry Conference, San Antonio, TX, June 05 - 09, 2005.

As part of a recent Nationwide Disinfection By-Product (DBP) Occurrence Study, iodo-acids were identified for the first time as DBPs in drinking water disinfected with chloramines.  The iodo-acids identified included iodoacetic acid, bromoiodoacetic acid, (E)-3-bromo-3-iodo-propenoic acid,  (Z)-3-bromo-3-iodo-propenoic acid,  and (E)-2-iodo-3-methylbutenedioic acid.  These identifications have since been confirmed using authentic chemical standards, both commercial (for iodoacetic acid) and synthesized (for the other four iodo-acids).  There is concern about these new iodo-acid DBPs because mammalian cell cytotoxicity and genotoxicity studies have revealed that iodoacetic acid is highly cytotoxic and genotoxic, with a genotoxicity potency 2X higher than bromoacetic acid, the most genotoxic of the regulated haloacetic acids.  Also, many drinking water treatment plants in the United States have switched from chlorine to chloramines for treatment.  New evidence indicates that the formation of iodinated DBPs will be higher in chloraminated drinking water than in chlorinated drinking water.  The goal of this work was to develop an analytical method to quantify these five iodo-acids in drinking water, measure their occurrence in several drinking waters treated with chloramination, determine whether they are maximized in waters treated with chloramines only (compared to chlorine and chlorine-chloramines), and investigate the mammalian cell cytotoxicity and genotoxicity of the four synthesized iodo-acids (beyond iodoacetic acid).

Coupled Free and Dissolved Phase Transport: New Simulation Capabilities and Parameter Inversion.  M. Tonkin, J. W. Weaver, C. Zheng, and C. Muffels.  National Ground Water Association, MTBE and Perchlorate Conference, San Francisco, CA, May 26 - 27, 2005.

The vadose zone free-phase simulation capabilities of the US EPA Hydrocarbon Spill Screening Model (HSSM) (Weaver et al., 1994) have been linked with the 3-D multi-species dissolved-phase contaminant transport simulator MT3DMS (Zheng and Wang, 1999).  The linkage provides the means to include releases from multiple leaking underground storage tanks in a numerical simulation of aquifer contamination.  The HSSM simulates the release of fuel, its movement to the water table, transient movement of a lens-shaped body of fuel, and dissolution of contaminants into the aquifer.    The combination of the codes has been accomplished through (a) the development of a new time-varying source package for MT3DMS, referred to as the Hydrocarbon Spill Source Interface Package (HSS Package), and (b) the compilation of the vadose zone modules of HSSM as a Dynamic Link Library (DLL).  This paper describes the technical approach for linking the HSSM and MT3DMS programs, and presents an illustrative example simulation in which MT3DMS correctly executes the new HSS Package for multiple leaking underground storage tanks and then accurately solves the mass-transport equation.  The capabilities of the linked programs are then demonstrated in the context of a real-world groundwater flow-and-transport model application where model parameters are calibrated to field data using PEST and utility programs developed to facilitate the use of the new HSS Package with PEST. The linked programs greatly increase the capability of including realistic aspects of hydrocarbon spill-related sources of groundwater contaminants in contaminant transport models.

Mammalian Metabolism and Distribution of Perfluorooctyl Ethanol (8-2 Telomer Alcohol) and Its Oxidation Metabolites. W. M. Henderson, S. E. Duirk, E. J. Weber, and M. A. Smith.  Society of Toxicology Meeting, New Orleans, LA, March 06 - 10, 2005.

Perfluorinated compounds have been shown to be globally distributed, bioaccumulative, persistent and potentially toxic.  It has been hypothesized that many precursor fluorinated compounds, including the telomer alcohols, degrade or metabolize to the common metabolite PFOA.  Currently, little information is available on the mammalian metabolism of the telomer alcohols.  Both alpha and beta oxidation pathways are proposed, similar to fatty acid oxidation, resulting in the formation of perfluorononanoic acid (PFNA) and perfluorooctanoic acid (PFOA), respectively.    An animal model is needed to fully elucidate the mechanisms of metabolism and translocation as it relates to humans.  The objective of this study was to investigate the distribution and metabolism of the telomer alcohol in mice.  Male CD-1 mice received a gavage dose of 20 mg 8-2 telomer alcohol/kg BW in a propylene glycol/water (1:1) vehicle.  Animals were serially sacrificed at 1, 3, 6, 12, and 24 hours post-treatment.  Serum was collected via cardiac puncture and visceral organs excised for analysis of fluorochemicals.  After extraction, the volatile compounds were analyzed by negative chemical ionization GC/MS.  The non-volatile metabolites, after ion pair extraction, were derivatized with diazomethane and analyzed similarly with EI-GC/MS.  In liver, the telomer alcohol was seen at peak concentrations at 1 hour and in serum was quantifiable only at 1 and 3 hours post-treatment.  The telomer alcohol temporally decreased becoming non-detectable by 12 hours in both liver and serum.  In serum the most abundant metabolite was PFOA with a concentration of 186-212 ng/mL at 6 hours post-treatment and decreased by 22% at 12 and 24 hours.  In liver, PFOA is also the major metabolite reaching peak concentrations (38-44 ng/g) at 12 hours post-treatment.  The highest concentration of PFNA coincides with peak PFOA concentrations and the metabolites decrease at 24 hours.  In conclusion, the appearance of PFNA in the liver suggests that alpha oxidation occurs although it does not appear to be the predominant metabolism pathway.  However, these results suggest the telomer alcohol is metabolized by both alpha and beta oxidation pathways.   To gain further understanding of the metabolism pathway, analysis of the brain, kidney, spleen, gonads and fat is ongoing.

Models for Submarine Outfall Validation and Prediction Uncertainties.  W. E. Frick.  5th International Exhibition and Conference on Environmental Technology, Athens, Greece, February 03 - 05, 2005.

The U.S. Environmental Protection Agency's Center for Exposure Assessment Modeling (CEAM) in Athens, Georgia develops environmental exposure models, including plume models, and provides technical assistance to model users. The mixing zone and far-field dilution modeling application, Visual Plumes (VP), is a recent addition to the public‑domain models available on the center's web page. Windows‑based, VP adapts, modifies, and enhances the earlier DOS‑based PLUMES with a new interface, models, and capabilities. VP is a platform for competing mixing zone models intended to facilitate outfall design and assessment. By facilitating verification and direct inter‑model comparison, it is also intended to encourage the continued improvement of plume theory and models. Empirical or physics-based, environmental models eventually exhibit limitations and uncertainties as users apply or adapt them to new and challenging applications. It is important to continue to identify and remove them. This address reports on some efforts to verify and validate dilution models, including those found in Visual Plumes. This is done in the context of problem experience: a range of problems, including different pollutants such as bacteria; scales, including near-field and far-field effects; temporal extent, including analyses of extended time-series features; and adaptation, particularly the development of a new beach bacteria forecasting tool. The latter is intended to reduce beach closures due to bacterial contamination. The presentation concludes with the discussion of a design application featuring a planned outfall in the Aegean.

The U. S. Environmental Protection Agency Visual Plumes Modeling Software.  W.E. Frick.  Presented at Technical University of Crete Lecture, Crete, Greece, January 31, 2005.

The U.S. Environmental Protection Agency's Center for Exposure Assessment Modeling (CEAM) at the Ecosystems Research Division in Athens, Georgia develops environmental exposure models, including plume models, and provides technical assistance to model users. The mixing zone and far-field modeling application, Visual Plumes (VP), is a recent addition to the public‑domain models available on the center's web page. Windows‑based, VP adapts, modifies, and enhances the earlier DOS‑based PLUMES with a new interface, models, and capabilities. VP is a public domain platform for mixing zone models intended to facilitate outfall design and assessment, as well as to encourage the continued improvement of plume theory and models by facilitating verification and inter‑model comparison. Some examples are presented to illustrate VP's new capabilities. One example demonstrates its ability, for reasonably simple, one‑dimensional estuaries, to estimate background concentrations due to tidal re‑circulation of previously contaminated receiving water. This capability depends on the optional linkage to time‑series input files that enables VP to simulate mixing zone and far‑field parameters for long periods. A second example describes the new bacterial decay models for estimating depth changes in first‑order decay rates based on environmental stresses, including solar insolation, salinity, and temperature. Other examples are also briefly considered, including assessment of the nascent density phenomenon, thermal discharges, and tidal plumes. The lecture concludes with a design application featuring remote outfalls for medium sized municipalities.

Decision tool for riparian ecosystem managment in the Mid-Atlantic Highlands.  J.M. Johnston, R. Parmar, K. Wolfe, P. Kinder, J. Newland, T. DeMoss, and D. Kemlage.  Ecological Society of America Annual Meeting,  Portland, OR, August 1-6, 2004.
In the Canaan Valley Highlands of the Mid-Atlantic, riparian zone restoration has been identified as a critical watershed management practice not only for the ecosystem services provided but also for the potential socioeconomic growth from environmental investment and job creation. The Canaan Valley Institute (CVI) has partnered with the US Environmental Protection Agency to develop a watershed decision tool that evaluates riparian management scenarios via changes in instream habitat quality and aquatic community productivity and stability. The decision tool incorporates expert knowledge, models of system/biotic response and supporting data into a single interface for trade-off evaluation and scenario development. Fisheries endpoints considered include indicator species as well as sport fishes.

Interactive habitat models for Mid-Atlantic Highland stream fishes.  B. Rashleigh, R. Parmar, and J. Johnston.  Ecological Society of America Annual Meeting, Portland, OR, August 1-6, 2004.
In most wadeable streams of the Mid-Atlantic Highland region of the eastern United States, habitat alteration resulting from development in the watershed is the primary stressor for fish.  Models that predict the presence of stream fish species based on habitat characteristics can be useful in watershed management.  We developed such a model for thirteen of the most common Mid-Atlantic Highland stream fishes using stepwise multiple logistic regression and seventeen instream habitat variables.  The predictive ability of the models (one per species) ranged from 64-87%.  In general, the most suitable habitats for these species are characterized by: lower gradient, because this region is mountainous and the very high gradient streams provide less stable habitat; greater stream depth; and lower percentages of fine substrate and undercut banks, which are expected to occur in less disturbed areas.   We have incorporated our findings into an interactive software tool that outputs changes in the probability of occurrence for these fish species under various habitat management scenarios.

Development and validation of a mechanistic ground sprayer model.  M.E. Teske, N.J. Ewing, N.B. Birchfield, and S.L. Bird.  2004 ASAE/CSAE Annual International Meeting, Ottawa, Canada, August 1-4, 2004.
In the last ten years the Spray Drift Task Force (SDTF), U.S. Environmental Protection Agency (EPA), USDA Agricultural Research Service, and USDA Forest Service cooperated in the refinement and evaluation of a mechanistically-based aerial spray model (contained within AGDISP and AgDRIFT).  This model was validated with an extensive set of aerial field trials conducted by the SDTF, and is now used by the EPA for pesticide risk assessments.  Building on previous work, the authors have developed a companion model for ground-based pesticide applications that extends the capabilities of the empirical ground sprayer model contained in AgDRIFT.  Because the AgDRIFT ground sprayer model is empirically based on a data set with a limited range of ground sprayer operational conditions, current AgDRIFT ground sprayer predictions are limited to estimating downwind deposition from one of four application scenarios (two release heights and two atomization spectra).  With the development of a new mechanistic ground sprayer model, users will have more flexibility in selecting release height, nozzle type, tank mix properties, crosswind, temperature and relative humidity, number of swaths, etc.  This paper will present the algorithms of the complete mechanistic model, and compare model predictions with downwind deposition measurements from previously conducted ground sprayer field studies.