2005 Progress Report: Systems Biology Modeling of Fathead Minnow Response to Endocrine Disruptors

EPA Grant Number: R831848
Title: Systems Biology Modeling of Fathead Minnow Response to Endocrine Disruptors
Investigators: Denslow, Nancy , Orlando, Edward F. , SepĂșlveda, Maria (Marisol) S. , Watanabe, Karen
Institution: University of Florida , Oregon Health & Sciences University , Saint Mary College
Current Institution: University of Florida , Florida Atlantic University - Boca Raton , Oregon Health & Sciences University , Purdue University
EPA Project Officer: Saint, Chris
Project Period: August 1, 2004 through July 31, 2007
Project Period Covered by this Report: August 1, 2004 through July 31, 2005
Project Amount: $722,851
RFA: Computational Toxicology and Endocrine Disruptors: Use of Systems Biology in Hazard Identification and Risk Assessment (2004) RFA Text |  Recipients Lists
Research Category: Economics and Decision Sciences , Computational Toxicology , Endocrine Disruptors , Health Effects , Health , Safer Chemicals


The objective of this research project is to develop a computational model to evaluate molecular and protein biomarkers in relation to reproductive dysfunction in fathead minnows (FHM) exposed to environmental estrogens.  The model incorporates a number of biochemical endpoints along the entire hypothalamic-pituitary-gonadal (HPG) axis, direct evaluation of physiological changes and reproductive endpoints, and the pharmacodynamics and kinetic distribution of the contaminants.  The overall hypotheses being tested are as follows:  (1) genomic and proteomic biomarkers will be diagnostic of the estrogenic effects of environmental estrogens; and (2) these biomarkers will provide a global understanding of mechanisms of action that will specifically relate to reproductive endpoints in FHM that are adversely affected by exposure to estrogenic compounds.  We further expect model compounds to have unique gene expression patterns that can predict exposure outcomes of other environmentally relevant compounds.  This project has been converted to a cooperative agreement with the U.S. Environmental Protection Agency (EPA) laboratories in Duluth, Cincinnati, and Athens.

Specific Aims

Specific Aim 1:  To determine and compare gene and protein expression profiles and physiological and reproductive endpoints for adult FHM exposed to a model estrogen 17 alpha-ethinylestradiol (EE2), androgen (17β-trenbolone), or their antagonists (ZM 189,154 and flutamide, respectively).

Specific Aim 2:  To predict gene expression patterns of two compounds (zearalenone and EE2) that are environmental estrogens.

Specific Aim 3:  To develop a computational modeling framework that integrates exposure concentration, gene expression, and proteomic profiles with physiological endpoints. 

Progress Summary:

Endocrine disrupting compounds (EDCs) can target the HPG axis at different levels of complexity.  These levels span from the molecular level to the whole organism.  Our work is centered on understanding how changes at the molecular and protein levels affect downstream physiological endpoints and how well these biomarkers predict adverse effects on reproduction.  There are many published studies that have shown a direct impact of EDCs on the reproductive success of fish and wildlife.  These compounds include certain organochlorine pesticides, industrial compounds, pharmaceuticals, plasticizers, surfactants, and metals.  These compounds have been shown to alter fertility, fecundity, egg hatchability, survival of young, sex ratio, and other reproductive parameters.  Although these endpoints have high ecological value, they do not point to specific mechanisms of action.  We are pursuing a “Systems Toxicology” approach in which we apply molecular tools to understand global mechanisms of action that are affected by chemical exposure.  We will relate these changes to reproductive endpoints using a novel physiology-based mathematical model.

Model Development

In Year 1 of the project, significant progress was made in the formulation of a physiologically based model of the HPG axis in male FHM.  Following the suggestions of our proposal reviewers, a modular approach to model development is being taken so that useful submodels will be developed in the process of constructing an integrative model representing multiple scales of biological organization (from molecular-level gene expression to physiological-level reproductive effects).  The first submodel to be developed is a physiologically based pharmacokinetic (PBPK) model that simulates the disposition of EE2 in male FHM.


Our physiologically based model is formulated on the concept of mass balances for each fish tissue “compartment” shown in Figure 1 (Nichols, et al., 1990).  Compartments are chosen to represent tissues that have a salient effect on the disposition of the compound(s) of interest.  Differential equations describe the accumulation, flows, and reactions (e.g., xenobiotic metabolism, ligand-receptor binding, protein induction, etc.) of each compound of interest.  The formulation of PBPK models falls into two regimes:  (1) perfusion-limited, or (2) diffusion-limited (Gibaldi and Perrier, 1982; Medinsky and Klaassen, 1996).  In the perfusion-limited case, the transport of the chemical to the tissue in the bloodstream is the rate-limiting factor in how the chemical distributes.  In the diffusion-limited case, the cell membrane permeability is much lower than the flow of chemical to the tissue through the bloodstream.  Because EDCs, in general, include hormone-mimics and chemicals, both transport regimes are possible (Watanabe, 2002); however, because estrogens and androgens are fat-soluble molecules that readily diffuse across the cell membrane, a perfusion-limited model formulation will be used that has the general form,

            .    (1)
Ci,j (mol/L) is the concentration of the compound i (e.g., EE2, 17b-estradiol, testosterone, vitellogenin) in the j tissue compartment, where j can be:  gill, richly perfused tissue, fat, kidney, poorly perfused tissue, brain, gonad, and liver.  CArti is the arterial blood concentration of compound i, and the concentration in venous blood, CVeni = Ci,jλi, where λi is the tissue to blood partition coefficient.  Vj is the tissue compartment volume (L); t is time (min); and Fj is volumetric blood flow entering and leaving the tissue compartment.  Rates of reactions (mol/min) that produce compound i are represented by Pi,j, and consumption reactions (e.g., xenobiotic metabolism) or elimination processes (e.g., urinary excretion) are denoted by Ri,j (mol/min).

The resulting set of coupled, ordinary differential equations then is integrated numerically to solve for the concentrations of the compound(s) as a function of time.  Model assumptions and a set of equations for EE2 will be provided upon request.  Such a model provides the framework for the biological system within which we can integrate (nest) biological processes occurring at smaller scales (e.g., cellular and molecular).

Figure 1.  Physiologically Based Pharmacokinetic Model of Male Fathead Minnows

Analysis of Baseline Data in FHM

As part of the process for developing a physiologically based model that integrates multiple biological scales, it is important to understand and calibrate the model for baseline (i.e., untreated or unexposed) conditions.  Data from the Ankley laboratory for male and female controls used in 11 chemical exposure experiments in recent years are being analyzed to quantify the biological variability that exists in male and female FHM.  Results from this study will be useful to calibrate our model, provide a database for use in probabilistic assessments, and for possible extrapolation to related fish species, ultimately to inform predictions of EDC effects on fish populations.  This work currently is being prepared for publication.

Exposure Paradigms

We initiated two exposure experiments using 48-hour static exposures to EE2.  In each experiment, 90 percent of the tank water was changed at 24 hours.  A spike indicates when the treatment compound is being added to the tanks.  The first experiment was set up to measure the distribution of two concentrations of EE2 (10 and 50 ng/L) among the endocrine active tissues, including brain, liver, gonad, and whole body.  Using gas chromatography/mass spectrometry, we also measured the actual concentrations of EE2 in the water at the start of the experiment, right before the 24-hour spike, and at the end of the experiment.

Table 1.  Measurement of Variation of EE2 Concentrations (ng/L)
During the 48-hour Experiment

Nominal Conc.

Actual Conc.
Spike 1

Actual Conc.
24 hours

Actual Conc.
Spike 2

Actual Conc.
48 hours











Experiments are still in progress to determine the concentrations of EE2 in the various tissues.  We are developing and validating a microscale extraction method for this purpose.

In the second experiment, the treatments were comprised of EE2 (5, 10, and 50 ng/L), the antiestrogen ZM189,154 (10, 50, and 100 µg/L), and a mixture of EE2 (50 ng/L) with three concentrations of ZM189,154.  In addition the experiment included vehicle control and no treatment control groups.  At 48 hours, tissues were removed and processed for various endpoints, including total RNA for microarrays, tissues for brain aromatase activity, and ex vivo incubation of gonads for steroidogenesis.

We have started to determine gene expression profiles for the brain, gonad, and liver.  We are using a 2,000 gene FHM microarray for this work.  Initial results for the liver are excellent and suggest that a group of 68 genes are up-regulated and a group of 128 genes are down-regulated by EE2 treatment at a statistical significance of P ≤ 0.05.  At the top of the list of the up-regulated genes, we found two vitellogenins (Vtg1 and Vtg3), the serpentine receptor, and hemaglutinin neuraminidase, whereas the most down-regulated genes included ribonuclease A, angiogenin precursor, and complement C3.  These gene changes will be verified by quantitative polymerase chain reaction (q-PCR), along with another 22 genes selected that have specific importance for the HPG axis.

21-Day Reproductive Assay – Purdue University

A total of 20 FHM breeding pairs (approximately 6 months of age) and approximately 250 juvenile fish (< 1 m of age) were obtained from Dr. Gary Ankley.  An indoor re-circulating culture system was established.  Water was aerated using a standard air pump.  The fish culture system, together with the biofilters and air pumps, were placed in an air-conditioned room of approximately 18 m2 (200 ft2).  The photo period was maintained at 18L:6D using a day-light control timer.  The fish culture system consisted of two 266 L (70 gallon) fiberglass holding tanks, two 114 L (30 gallon) brood-stock spawning glass aquaria, four growing glass aquaria:  three 152 L (40 gallon) tanks, and one 114 L (30 gallon) tank.  The system was acclimated for about 1 month prior to the introduction of fish.  During this time, several water quality parameters (pH, temperature, dissolved oxygen [DO], ammonia [NH3], nitrites [NO2], and nitrates [NO3]) were measured on a daily basis.  When these water quality parameters were maintained at conditions deemed to be safe for fish (i.e., pH: 5 - 7.5; temperature: 22 - 28°C; DO: 5 - 6 mg/L; NH3: < 1.0 mg/L NH2: close to 0 mg/L; and NO3: > 50 mg/L), the fish were introduced into the system.

Fish arrived at Purdue University in early March 2005.  Immediately after their arrival, adult fish were separated in pairs and allowed to breed in specially designed brood-stock tanks.  These consisted of two 114 L tanks separated in six identical chambers each.  Chambers were built using non-corrosive wire mesh (~ 1.5. mm mesh size).  Each chamber was stocked with a breeding pair and fitted with a spawning tile.  Spawning tiles were made out of 10 cm (4 inch) diameter polyvinyl chloride that was cut into pieces measuring 10 cm in length.  Adult pairs began spawning approximately 1 week after their arrival.  Eggs were collected from several pairs, pooled, and allowed to hatch in a plastic shallow tray (approximate dimensions 60 x 30 x 15 cm).  To date, we have successfully hatched and raised approximately 2,500 fish.

Immediately after hatching (FHM eggs develop into swimming fry in about 5 days), fry were moved to growing tanks.  These consisted of four glass aquaria:  three 152 L (40 gallon) tanks and one 114 L (30 gallon) tank.  At this stage, and during the first 2 to 3 weeks of age, fry are fed live prey (brine shrimp) at least three times/day.  We have a continuous source of live brine shrimp that we hatch using three hatchers.  After this time, fish are fed twice a day with frozen brine shrimp.  We have successfully raised 250 juveniles received from the EPA laboratory to adult stage (FHM reach maturity at about 4 - 6 months of age).  During the process, there was an extremely low mortality rate (< 5 % overall).

Purdue University has constructed a static-renewal exposure system that consists of thirty 9.5 L (2.5 gallon) glass aquaria supplied with air.  We also are in the process of constructing a flow-through exposure system that will consist of an enclosed aluminum-frame and glass sides unit, which will fit the same size and number of exposure tanks as described above, but that will allow for a continuous delivery of toxicants using a series of peristaltic pumps that are controlled with the aid of computer software.  This unit will be operational in April 2006.

For these exposures, we will measure different reproductive endpoints (such as spawning success, hatch rates, and fry survival) after exposing individual FHM (one pair/tank) to different EDCs.  This setup will allow for the exposure of up to five doses of different EDCs replicated six times.  Because our fish have just reached reproductive age, we have not begun testing for the effects of EDCs.  Between the months of August and September 2005, we tested the spawning ability of our adult fish by randomly selecting pairs from recently grown juvenile stock and looking at basic spawning parameters.  So far, 40 pairs have been selected that are spawning at comparable rates for our first experiment, which will consist of exposing adult FHM to EE2 (5, 10, and 50 ng/L) for 21 days.  This first experiment will be conducted in October 2005.

Antibody Development for FHM Follicle-Stimulating Hormone (FSH) and Luteinizing Hormone (LH)

We have initiated polyclonal antibody production for FHM FSH and LH to develop assays to measure these brain hormones directly in the blood.  The antibodies should be ready for evaluation in about 3 months.

Future Activities:

We will continue to obtain expression pattern profiles for EE2 and ZM189,154 in the brain, gonad, and liver and validate the changes in gene expression by q-PCR.  We also expect to complete exposures for trenbolone and the anti-androgen flutamide using a similar paradigm.  Additionally, we will focus our efforts on obtaining measures of changes in reproductive parameters for FHMs in a 21-day reproductive assay with males and females.  We also will measure the effects of exposure on brain aromatase activity and gonadal steroidogenesis.  Finally, we will complete our quantitative analyses of actual concentrations of contaminants in the various endocrine-sensitive tissues.  The values that we obtain for all of these measurements will be incorporated into our mathematical model.  These experiments will set the stage for examining the effects of a weak estrogen, zearalenone, which is used as a growth supplement in cattle.


Gibaldi M, Perrier D. Pharmacokinetics. New York, NY: Marcel Dekker, Inc., 1982.

Medinsky MA, Klaassen CD. Toxicokinetics. In: Klaassen CD, ed. Casarett and Doull’s Toxicology: The Basic Science of Poisons. New York, NY: McGraw-Hill, 1996;187-198.

Nichols JW, McKim JM, Andersen ME, Gargas ML, et al. A physiologically based toxicokinetic model for the uptake and disposition of waterborne organic chemicals in fish. Toxicology and Applied Pharmacology 1990;106(3):433-447.

Segel IH. Enzyme kinetics behavior and analysis of rapid equilibrium and steady-state enzyme systems. New York, NY: John Wiley & Sons, Inc., 1993.

Watanabe KH. Fundamentals of physiologically-based toxicokinetic models. In: Chyczewski L, Niklinski J, Pluygers E, eds. Endocrine Disrupters and Carcinogenic Risk Assessment. Amsterdam, Netherlands: IOS Press; 2002:271-280.

Journal Articles:

No journal articles submitted with this report: View all 23 publications for this project

Supplemental Keywords:

fathead minnows, gene arrays, reproductive assay, physiologically based model, environmental estrogens, protein biomarkers, estrogenic compounds, model compounds, gene expression, reproductive end points,, RFA, Health, Scientific Discipline, POLLUTANTS/TOXICS, Health Risk Assessment, Endocrine Disruptors - Environmental Exposure & Risk, Chemicals, endocrine disruptors, Risk Assessments, Biochemistry, Endocrine Disruptors - Human Health, Biology, bioindicator, fish, biomarkers, assays, animal model, EDCs, exposure studies, endocrine disrupting chemicals, sexual development, mechanistic screening, endocrine disrupting chemcials, animal models, human growth and development, toxicity, fathead minnow, estrogen response, invertebrates, invertebrate model, estrogen receptors, hormone production, androgen, assessment technology, ecological risk assessment model, estuarine crustaceans

Relevant Websites:


Progress and Final Reports:

Original Abstract
2006 Progress Report