Grantee Research Project Results
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Research
Science to Achieve Results (STAR) Program
CLOSED - FOR REFERENCES PURPOSES ONLY
Computational Toxicology Research Centers: in vitro and in silico Models Of Developmental Toxicity Pathways
This is the initial announcement of this funding opportunity.
Funding Opportunity Number: EPA-G2008-STAR-W1
Catalog of Federal Domestic Assistance (CFDA) Number: 66.509
Solicitation Opening Date: November 4, 2008
Solicitation Closing Date: January 29, 2009, 4:00 pm Eastern Time
Eligibility Contact: William Stelz (stelz.william@epa.gov); phone: 202-343-9802
Electronic Submissions: Ron Josephson (Josephson.Ron@epa.gov); phone: 202-343-9643
Technical Contact: Deborah Segal (segal.deborah@epa.gov); phone: 202-343-9797
Access Standard STAR Forms
Research awarded under previous solicitations
SUMMARY OF PROGRAM REQUIREMENTS
Synopsis of Program:
The U.S. Environmental Protection Agency (EPA), as part of its Science to Achieve Results (STAR) program, is seeking applications proposing to develop in vitro and in silico (computational) models for developmental toxicity pathways. The STAR program is issuing this request for applications (RFA) for research that in conjunction with in vivo data, will seek to integrate in vitro biochemical and cellular response data with computational models and theoretical or applied mathematic techniques. The research conducted under this RFA will facilitate the development of a predictive capacity for estimating outcomes or risk associated with particular toxicity processes as a result of developmental exposure to environmental pollutants and toxicants. Predictive computational modeling of core processes that drive development, including patterning, morphogenesis, selective growth and cell differentiation, and the detailed understanding of biological pathways that regulate these processes have the potential to address environmental and human health factors with broad scientific or economic impacts.
The goals of the computational research effort supported by the U.S. EPA are to develop the use of computational approaches to provide tools for quantitative risk assessment and develop more efficient strategies for prioritizing chemicals for screening and testing. Through the support of the computational toxicology initiative, EPA’s STAR program will fund research that addresses data gaps in human health risk assessment and will strengthen the ability of predictive scientific data to guide future scientific research, policy, and decisions.
To support the development of predictive models and simulations, the Center will be funded for up to 4 years. The Center should be comprised of multiple scientists with different backgrounds and capabilities, from a single or a variety of institutions, working collaboratively as a team. Each Center must foster the professional development of junior faculty and facilitate the training of students or postdoctoral fellows in the application of computational systems biology. Proposals must have an ultimate focus on risk assessment and improving the use of biological data in quantitative models of developmental toxicity. The proposed research must be consistent with the strategic objectives of the computational toxicology program: (1) improve understanding of the linkages in the continuum between the source of a chemical in the environment and adverse health and/or ecological outcomes; (2) provide predictive models for screening and testing; and (3) improve quantitative risk assessment.
This is the initial announcement for this year’s program. Although not anticipated, should modifications of this announcement be necessary, they will be posted.
Award Information:
Anticipated Type of Award: Grant or Cooperative Agreement
Estimated Number of Awards: Approximately 1 award
Anticipated Funding Amount: Approximately $3.2 million total for all awards
Potential Funding per Award: Up to a total of $3,200,000, including direct and indirect costs, with a maximum duration of 4 years. Cost-sharing is not required. Proposals with budgets exceeding the total award limits will not be considered.
Eligibility Information:
Public nonprofit institutions/organizations (includes public institutions of higher education and hospitals) and private nonprofit institutions/organizations (includes private institutions of higher education and hospitals) located in the U.S., state and local governments, Federally Recognized Indian Tribal Governments, and U.S. territories or possessions are eligible to apply. See full announcement for more details.
Application Materials:
To apply under this solicitation, use the application package available at Grants.gov (for further submission information see Section IV.E. “Submission Instructions and other Submission Requirements”). The necessary forms for submitting a STAR application will be found on the National Center for Environmental Research (NCER) web site, https://www.epa.gov/research-grants/funding-opportunities-how-apply-and-required-forms. If your organization is not currently registered with Grants.gov, you need to allow approximately one week to complete the registration process. This registration, and electronic submission of your application, must be performed by an authorized representative of your organization.
If you do not have the technical capability to utilize the Grants.gov application submission process for this solicitation, call 1-800-490-9194 or send a webmail message to https://www.epa.gov/research-grants/forms/contact-us-about-research-grants at least 15 calendar working days before the submission deadline to assure timely receipt of alternate submission instructions. In your message provide the funding opportunity number and title of the program, specify that you are requesting alternate submission instructions, and provide a telephone number, fax number, and an email address, if available. Alternate instructions will be e-mailed whenever possible. Any applications submitted through alternate submission methods must comply with all the provisions of this RFA, including Section IV, and be received by the solicitation closing date identified above.
Agency Contacts:
Eligibility Contact: William Stelz (stelz.william@epa.gov); phone: 202-343-9802
Electronic Submissions: Ron Josephson (Josephson.Ron@epa.gov); phone: 202-343-9643
Technical Contact: Deborah Segal (segal.deborah@epa.gov); phone: 202-343-9797
I. FUNDING OPPORTUNITY DESCRIPTION
A. Introduction
Computational modeling of biological systems at different scales is gaining momentum as a tool in cell biology and disease studies. Advancements in the ability to implement and develop highly improved computer-based approaches to modeling biological systems are key elements in facilitating the development of a predictive capacity for estimating outcomes or risk associated with exposure of organisms during their development to environmental toxicants. Therefore, we seek to fund research in developmental toxicology, computer science and systems biology that address the environmental problems and research needs facing the scientific community by applying sophisticated computing techniques and resources to in silico multi-scale modeling applications at the molecular, cellular, organ, and system-wide levels.
In assessing risk associated with exposure to a chemical or other environmental stressor, a number of scientific uncertainties exist along a “source-to-adverse outcome” continuum. This pathway from exposure to biological effect ranges from: (1) the presence of the chemical in the environment, (2) the uptake and distribution of the chemical in the organism or environment, (3) the presence of the active chemical at a systemic target site, (4) and the series of biological events that lead to the manifestation of an adverse human health or ecological outcome that can be used for risk assessment. The “Human Health Research Strategy” (https://www.epa.gov/nheerl/humanhealth/HHRS_final_web.pdf (67 pp 1.5 MB)) developed by the U.S. Environmental Protection Agency’s (EPA) Office of Research and Development (ORD) describes the scientific uncertainties and some of the multidisciplinary approaches that are needed to build linkages between exposure, dose and effects. ORD’s research program in Computational Toxicology (https://www.epa.gov/comptox/comptox_framework.html) supports the use of emerging technologies to improve risk assessment and reduce uncertainties in this source-to-adverse outcome continuum. The first strategic objective of the Computational Toxicology Initiative is to develop improved linkages across the source-to-outcome continuum, including the areas of chemical transformation and metabolism, better diagnostic/prognostic molecular markers, improved dose metrics, characterization of toxicity pathways, metabonomics, systems biology approaches, modeling frameworks, and uncertainty analysis. The second strategic objective is to provide improved predictive models for hazard identification, including the areas of Quantitative Structure Activity Relationships (QSARs) and other computational approaches, improved pollution prevention strategies, and high-throughput (HTP) screening approaches. The third strategic objective is to apply computational toxicology to enhance quantitative risk assessment in the areas of dose-response assessment, cross-species extrapolation, and chemical mixtures.
The EPA is committed to the development and application of novel technologies, derived from computational chemistry, molecular biology, systems biology, and bioinformatics in toxicological risk assessment and seeks to fund research that will develop computational models and simulations that use a broad range of biological information, including genomics, proteomics and cellular data. Further development of advanced, mechanistic models will allow scientists to create new hypotheses as well as analyze the ever-expanding mass of biological data. New in vitro and in silico toxicological methods are needed to provide solutions that can be applied to the HTP screening and evaluation of large numbers of chemicals, filling data gaps for hazard and risk assessment, and rapidly prioritizing substances of high concern. Therefore, this STAR Computational Toxicology RFA seeks to fund a Research Center that will generate new in vitro data and in silico models of developmental processes that will advance scientific understanding of developmental toxicities and predictive ability in vivo. This RFA seeks applicants who propose to integrate in vitro biochemical and cellular response data with computational models, and theoretical or applied mathematic techniques, to facilitate the development of a predictive capacity for estimating outcomes or risk associated with developmental exposure to environmental toxicants.
B. Background
The mission of the U.S. EPA is to protect human health and the environment from harmful effects that may be caused by exposure to pollutants in the air, water, soil, and food. Protecting human health and the environment carries the challenge of assessing the risk that is posed by tens of thousands of chemicals. The large number of chemicals present in our surroundings has made it impossible to evaluate every chemical with the most rigorous testing strategies; consequently, standard toxicity tests have been limited to only a relatively small number of chemicals. For example, in over 30 years since the enactment of the Toxic Substances Control Act of 1976 the EPA has acquired data on about 200 of approximately 80,000 chemicals in the U.S. market. REACH (Registration, Evaluation, and Authorization of Chemicals) legislation passed by the European Parliament in 2006 will require toxicity evaluations of up to 30,000 chemicals in a short time (11-year time frame for registering 30,000 chemicals under REACH, http://ec.europa.eu/enterprise/reach/index_en.htm). Information from toxicity tests conducted on laboratory animals to evaluate the potential for chemicals to cause cancer, birth defects, and other adverse health outcomes have traditionally served as part of the basis for public health and regulatory decisions regarding the toxic health risks posed by chemicals found at low levels in the environment. Current animal test methods that have evolved over the last half-century require a number of assumptions and extrapolations to translate in vivo toxicity data and exposure data into predictions of human health effects. In the most basic form of extrapolation of a critical effect in a laboratory animal study to the human situation, little or no information about the metabolism of the chemical or potential mode of action is involved in estimating the safe level of exposure. Furthermore, whole animal testing studies are time consuming and costly.
In recognition of the need for innovative approaches to toxicity testing the U.S. EPA commissioned the National Research Council (NRC) to develop a long-range vision and a strategy to advance toxicity testing in the 21st century. The NRC strategy [1] builds from recent advances in the genomic sciences, including bioinformatics and computational biology, coupled with technological advances in HTP and high-content screening (HCS) assays for cells and tissues. These methodologies have a potential to fundamentally change the way chemicals are tested for risks that they may pose to humans and wildlife populations. Implementation of this strategy requires a collaborative network of investigators and organizations focused on HTP and HCS assays, utilizing novel in vitro and in silico (computational) toxicological methods. These methods will be utilized to provide solutions for screening and assessment of large numbers of chemicals to which humans are potentially exposed, and for assessing the risk of environmental chemicals at different exposure scenarios, dose levels, susceptible individuals and sensitive life-stages for substances of high concern [2].
In November 2005, the U.S. EPA funded two STAR Environmental Bioinformatics Research Centers to foster research focused on the design, development, and application of computer systems and software enabling scientists to explore HTP data from gene expression microarray experiments, mass spectrometry-based peptide and protein identification experiments, and various quantitative measures on metabolic state and metabolites. In April 2008, the U.S. EPA funded a third STAR Computational Toxicology Research Center, to stimulate the development of predictive environmental and biomedical computer-based simulations and models that will span from mechanistic to discovery-based efforts.
This STAR Computational Toxicology RFA seeks to fund a new Research Center that will generate in vitro data and computational models of developmental processes that will further the predictive ability and understanding of developmental toxicities. The research funded under this RFA should promote the interface of data generation and model development. Models developed should utilize data from high-information content (HTP, HCS) assays, and where possible incorporate data, methods, and approaches from EPA STAR Computational Toxicology initiative grantees, complementary research programs in EPA’s intramural laboratories, other U.S. government laboratories, laboratories supported by the European Union’s Work Program ENV 2009.3.3.1.1, or other academic or industry research laboratories from the broader scientific community.
Through its application, computational toxicology can provide a novel framework for in silico modeling and simulation to validate and predict key aspects of biology that are difficult to analyze experimentally due to cost, scale or complexity. Simulation models of biological systems should ideally contain sufficient detail to reconstruct normal functions and predict major disease states. As these models become more widely developed, simulation modeling will aid in the targeting of current knowledge gaps, as well as in providing solutions for filling them through providing a research tool for selecting critical factors from multiple simulations for actual experimental design. Utilizing network models of biological systems, scientists may gain a better understanding of how cells sense their environment and respond to environmental stimuli. In turn, this understanding can help unravel complex relationships across biological systems and support a scientifically sound process of projecting human health risks posed by chemical compounds. Through the application of today's powerful computational research platforms to the genomic sciences, the scientific community is now able to begin to identify, analyze, and compare the fundamental biological components and processes that result from exposures to environmental toxicants with their predicted impact on human health.
The envisaged long term end product will be in silico models that describe signatures of biological activity and environmental response for various organ systems undergoing development, and ‘virtual tissue’ models that draw from new data, database repositories, bioinformatics and computational resources to build:
- knowledgebases to extract, organize and store data, facts and concepts about developing tissues and structures;
- classification of ‘toxicity pathways’ through which chemical perturbation at a molecular level invokes perturbations in embryogenesis and development;
- multi-scale models to help users analyze key events during chemical toxicity and understand the progression of altered events from cells to phenotype (developmental defects).
The overall goal of the computational research effort supported by the U.S. EPA’s STAR program is to develop the use of computational approaches to provide tools for quantitative risk assessment and more efficient strategies for prioritizing chemicals for screening and testing. Relevant embryological systems might include, for example, vertebrate segmentation, neural tube morphogenesis and neural crest migration, early eye development, craniofacial and branchial arch development, morphogenesis and differentiation of the urogenital tract, and cardiac development. Other tissues and systems may be appropriate study models. Proposals addressing the use of simpler organismal models (e.g., Echinoderm, Drosophila) must be accompanied by coherent approaches to translate data to human developmental affects and health outcomes.
The specific Strategic Goal and Objective from the EPA’s Strategic Plan that relate to this solicitation are:
Goal 4: Healthy Communities and Ecosystems, Objective 4.4: Enhance Science and Research
The EPA’s Strategic Plan can be found at https://www.epa.gov/ocfo/plan/2006/entire_report.pdf (PDF) (184 pp, 11.56 MB)
C. Authority and Regulations
The authority for this RFA and resulting awards is contained in the Toxic Substances Control Act, Section 10, 15 U.S.C. 2609, and the Federal Insecticide, Fungicide, and Rodenticide Act, Section 20, 7 U.S.C. 136r.
For research with an international aspect, the above statutes are supplemented, as appropriate, by the National Environmental Policy Act, Section 102(2)(F).
Applicable regulations include: 40 CFR Part 30 (Uniform Administrative Requirements for Grants and Agreements with Institutions of Higher Education, Hospitals, and Other Non-Profit Organizations), 40 CFR Part 31 (Uniform Administrative Requirements for Grants and Cooperative Agreements to State and Local Governments) and 40 CFR Part 40 (Research and Demonstration Grants). Applicable OMB Circulars include: OMB Circular A-21 (Cost Principles for Educational Institutions) relocated to 2 CFR Part 220, OMB Circular A-87 (Cost Principles for State, Local and Indian Tribal Governments) relocated to 2 CFR Part 225, OMB Circular A-102 (Grants and Cooperative Agreements With State and Local Governments), OMB Circular A-110 (Uniform Administrative Requirements for Grants and Other Agreements with Institutions of Higher Education, Hospitals and Other Non-Profit Organizations) relocated to 2 CFR Part 215, and OMB Circular A-122, (Cost Principles for Non-Profit Organizations) relocated to 2 CFR Part 230.
D. Specific Research Areas of Interest/Expected Outputs and Outcomes
Note to applicant: The term “output” means an environmental activity or effort, and associated work products, related to a specific environmental goal(s), (e.g., testing a new methodology), that will be produced or developed over a period of time under the agreement. The term “outcome” means the result, effect, or consequence that will occur from the above activit(ies) that is related to an environmental, behavioral, or health-related objective.
Introduction
The U.S. Environmental Protection Agency (EPA), as part of its Science to Achieve Results (STAR) program, is seeking applications proposing to develop in vitro data and in silico models of core developmental processes. Developmental processes of interest include patterning, morphogenesis, selective growth and cell differentiation, as well as projects that will provide detailed understanding of biological pathways that regulate these processes during normal embryogenesis and in response to environmental toxicants. The STAR program is issuing this request for applications (RFA) for research that will seek to apply data from state-of-the-art HTP and/or HCS to embryonic tissues and systems developing in vivo or in vitro. In order to facilitate the development of a predictive capacity for estimating outcomes or risk associated with toxicity processes as a result of developmental exposure to pollutants and toxicants, these data must be amenable to the application of high-performance computing and theoretical mathematical techniques. The research sponsored through this RFA must seek to better understand how developing tissues react to chemical exposures, and to improve the ability of biomedical research to address the increasingly complex array of exposure profiles in human health risk assessments. Because of the inherent complexity of real biological systems, computational models of virtual tissues based qualitatively on actual experimental data are necessary in order to produce validated predictive systems for quantitative risk assessment. Through this RFA, EPA seeks to fund centers to conduct research in developmental biology and developmental toxicology that will bridge the interface of in vitro data generation and in silico model development to answer critical biological questions related to toxicity pathways important to human development, hereafter referred to as ‘developmental toxicity pathways.’
The outcome of this model development research has the potential to increase the understanding and ability to predict the biological pathways that regulate processes during normal embryogenesis and in response to environmental toxicants. Because of the inherent complexity of real biological systems, development and analysis of computational models based directly on experimental data are necessary to achieve this understanding.
The output of this research will include the creation of simulation models for predicting toxicity pathways, mechanisms, and health impacts as a result of environmental exposures. This research effort will generate in vitro data and computational models of developmental processes that will further the predictive ability and understanding of developmental toxicities.
Proposals are expected to include multiple investigators working in collaboration. A proposed center must have a Center Director who will have the responsibility of assuring that resources are utilized in an optimal way, and that the efforts of each team are focused on their proposed research project. The Center Director will also be responsible for reports and reviews of the teams’ research progress. Centers are obligated to engage the expertise of a minimum of two multidisciplinary teams focused on basic and applied research projects needed to fill human health risk assessment knowledge gaps. Each team should have its own unique research focus, denoted as an Investigational Area, and be balanced in such a manner to make the most effective use of available resources in accomplishing its proposed objectives.
Each Center must encompass a minimum of two research teams, with each team’s Investigational Area focus detailed in a Research Plan (for guidance see “Research Plan” in section IV.B.7.a.). To be responsive to the requirements of this RFA, a Center is to be comprised of a minimum of two Investigational Areas, each with its own research team (2-5 proposed Investigational Areas are expected for each Center). For each Investigational Area proposed, successful applications will address at least one of the topics listed under Priorities of Model Development in Section I.D. of this RFA.
Research grant applications driven by high-information content assays and integrative biological approaches and their translation to mammalian embryogenesis and human developmental toxicity in principle, design, analysis, or validation will be given highest priority during peer review.
Embryogenesis and developmental toxicity
Embryogenesis is fundamental to all biological systems. The study of embryology entails questions such as what processes determine anatomical structures (morphogenesis) and tissues (differentiation) and the mechanisms through which these processes are controlled by the genome in interaction with the environment. Modern research has yielded extensive knowledge about the molecular control of embryogenesis and the conservation of cell signaling involved in cell-cell communication and interactions during development and beyond [3]. This knowledge may eventually hold the scientific basis for predicting the health effects arising from prenatal and early life exposure to environmental chemicals.
For humans, developmental toxicity refers to adverse effects produced prior to conception or during pregnancy and postnatally to the time of sexual maturation. The standard procedure for assessing the safety of chemicals to the developing organism involves the use of pregnant laboratory animals exposed throughout the period of major organogenesis, and subsequent examination of the offspring for growth and morphology. EPA’s guidelines for developmental toxicity risk assessment (FRL-4038-3, December 5, 1991) are recorded in the Federal Register 56(234):63798-63826 and updated in a 1998 workshop (SAP Report No. 99-01C, January 22, 1999). The Office of Prevention, Pesticides and Toxic Substances (OPPTS) has developed a Health Effects Test Guidelines for Prenatal Developmental Toxicity Study (OPPTS 870.3700) through a process of harmonization that blended the testing guidance and requirements that existed in the Office of Pollution Prevention and Toxics (OPPT) and appeared in Title 40, Chapter I, Subchapter R of the Code of Federal Regulations (CFR), the Office of Pesticide Programs (OPP) which appeared in publications of the National Technical Information Service (NTIS) and the guidelines published by the Organization for Economic Cooperation and Development (OECD) [https://www.epa.gov/opptsfrs/publications/OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/Series/870-3700.pdf (11 pp, 29 K)]
The limiting factors in traditional animal studies imposed by the number of chemicals that can be reasonably tested, and the need to reduce uncertainties that give rise during the extrapolation process of high-dose to low-dose and animal-to-humans, have motivated the development of more innovative approaches in biomedicine. HTP data generation and in silico modeling, combined with and computational analysis techniques, has the potential to integrate various sources of information quantitatively in risk assessment, as well as to improve the understanding of regulatory mechanisms leading to developmental defects. The capacity to accurately predict dose-dependent toxicity using in vitro and in silico models would greatly aid the risk assessment process by allowing chemical prioritization for animal testing. However, the complexity and diversity of spatio-temporal changes underlying developmental toxicity make incorporating biological details into risk assessment a difficult task. This is partly due to the existence of many interrelated molecular interactions, maternal physiology, and tracking the complicated non-linear response to exposures linked to critical biological processes that may vary by dose, genotype, chemical, stage, or tissue, and the resulting manifestation of structural or functional defects following prenatal exposure.
The NRC report [1] highlights the need for understanding cellular-response networks that control cellular function, mediate cell-cell communication and signaling, and orchestrate adaptive responses at different levels of chemical exposure. The term ‘toxicity pathways’ has been used by the NRC to describe the key cellular response pathways that, when sufficiently perturbed, are expected to result in adverse health effects. Similarly, understanding developmental defects at a systems-level requires elucidating how ‘developmental toxicity pathways’ are integrated with the genomic control of conserved cell signaling pathways that orchestrate morphogenesis and differentiation [4]. This requires substantial investment in research that can model, and ultimately predict, biological outcomes from a myriad of interrelated data and associative relationships covering the exposure-disease continuum.
Developmental toxicity pathways
The potential of an environmental chemical to cause adverse effects in the developing embryo or fetus is an important consideration in any health risk assessment. Many human health risk assessments have selected developmental endpoints (rather than cancer) for a critical effect, as demonstrated through a sampling of toxicological reviews in the U.S. EPA National Center for Environmental Assessment’s Integrated Risk Information System assessment database, which showed that 28% of cases had critical effects selected for childhood or gender susceptibilities. These endpoints included altered birth weight, skeletal variations, neurodevelopmental defects, altered maternal or postnatal weight gains in reproductive organs, and testicular lesions. In most cases the underlying developmental toxicity pathways remain largely uncharacterized. Therefore, in order to help fill scientific knowledge gaps, this RFA seeks to apply computational toxicology resources toward research in developmental toxicity.
Successful modeling of developing systems should describe different cell states (e.g., proliferation, differentiation, migration, apoptosis, shape, adhesivity, matrix remodeling), as well as heterogeneities in cellular competence and metabolic demands that unfold by the hour. A key challenge for science and technology development is to integrate heterogeneous molecular and cellular data as biological networks, to model developing systems with these networks, and then use these networks to analyze critical steps during toxicant exposure. In silico models developed must initially provide a qualitative prediction, which as research progresses, can eventually be translated into a quantitative model. An example of such a two-stage process would be first understanding the form of the dose-response curve, followed by connecting to the underlying model how the flow of molecular regulatory information is distributed across developmental toxicity pathways and intercellular signaling networks.
Modeling complex coordinated cellular behaviors and the many interacting factors that may influence higher-level functions requires information about basic entities at a lower level of the system [5]. Therefore, the scientific output of this RFA includes the collection of these molecular and cellular data for in silico modeling using potentially different kinds of in vitro research model systems. Two examples of in vitro systems that can be used for modeling vertebrate developmental processes are embryonic stem (ES) cell lines, or ES-like induced pluripotent stem cells, [6] and zebrafish embryos, [7] as these systems are free from direct maternal influence. Free living zebrafish embryos hold promise for the rapid screening of chemicals based on the potential of toxicants to directly perturb developmental processes [7]. The small eggs of zebrafish fit comfortably into microtiter plates thus enabling robotic handling and analyses of the developing embryo; furthermore, the ease of acquiring reporter fish, knock-out fish, or morpholino-induced ‘morphant’ phenotypes [8] for assessing developmental toxicity pathways is an advantage of using zebrafish over other small fish species. Because zebrafish embryos are transparent the sequence of development may be directly observed without disturbing the embryo and consequently followed through precisely timed stages as the embryo advances from fertilized egg to a swimming larva. This progression occurs rapidly, in just 5 days, and recapitulates many of the same anatomical features, morphogenetic processes, and cell signaling pathways used by the early human embryo. As such, zebrafish embryos provide a powerful alternative to mammalian animal models in research aimed to document and classify the potential developmental toxicity of environmental chemicals and to model these dynamic processes in silico.
Pluripotent ES cells have been considered for their potential in evaluating specific pathways and comparing observations to traditional animal bioassay models for identifying potential teratogens. HTP cellular screens of pluripotent ES cells can be used to study non-directed multiple endpoints, differentiation outcomes, directed linear-differentiation along various pathways to specific cell/tissue types, cellular plasticity and dedifferentiation, as well as developmental signaling pathways. Murine ES cell models accurately predict strong teratogenic pharmaceutical compounds but incompletely resolve non-teratogenic from weakly teratogenic compounds [6]. The capacity to identify biological signatures of toxicity in these in vitro models of development, and to map data from high-information content assays in normal embryogenesis, addresses the need for computer-based models that can be utilized to integrate these data into advanced models and simulations. These models and simulations will improve the understanding of critical steps in developmental toxicity and the ability to predict adverse effects and differentiate them from adaptive responses to chemical injury.
Computer-based models
A challenge for computational systems biology is to build useful multi-scale models that can be used to systematically investigate any or all interactions between the complex variables that are linked with abnormal developmental outcomes [4]. These interactions are potentially influenced by genetic and environmental signals, with the net outcome being the emergent properties associated with normal or abnormal collective cell behavior. Computational modeling of ‘virtual tissues’ may be useful to predict organ injury due to chemical exposure by: simulating the dynamics and characteristics of exposure and dose, the kinetics of perturbed molecular pathways, their linkage(s) with processes leading to altered cell states, and integration of the molecular and cellular responses into a predictive model. By placing emphasis on the normal biology of the system and its key regulatory components, virtual tissues represent a significant opportunity to understand the linkage between developmental toxicity pathways and developmental phenotype. Products from the proposals awarded under this RFA may provide valuable information that will advance the field of computational toxicology and apply this logic to predicting effects that would be difficult or costly to derive by traditional means [5]. Development of virtual tissue models require newly generated data collected across phylogenetic systems to fill the data gaps identified within the iterative process and test the predictive nature of virtual tissue models with real and simulated data.
To be responsive to this RFA studies must include pathways that are fundamentally reliant upon cell signaling (e.g., cell proliferation, apoptosis, adhesion and migration), intermediary metabolism (e.g., glycolysis, oxygen utilization, fatty acid biosynthesis), cell-specific functions (e.g., extracellular matrix remodeling, trans-differentiation, contractility and motility), and other categories to ensure that predictions are broadly applicable. Such efforts will also help answer how well in vitro experimental systems represent the full range of diverse cells present in the human embryo, how variability observed in the human population can modify quantitative predictions of in vivo dose-response, how exposure conditions influence outcomes, and how well the virtual tissue models represent the underlying biology of the system.
Priorities of Model Development
Areas of interest that will be supported by the program and are relevant to human health risk assessment, are models of a variety of types of networks including cell-cell signaling, signal transduction pathways; biochemical networks; developmental gene regulatory networks; metabolic networks; intracellular dynamics; cell structural dynamics; cell communication and adhesivity/motility. Areas of interest that represent the types of research that will be supported by the program are given below. Successful proposals must address at least one of these areas of interest listed below, and in doing so they must also highlight the direct relationship between the computational research and the biological applications for developmental processes and toxicities:
- Molecules to cells: Cellular and molecular data could be used to build models of developmental toxicity pathways that integrate biological processes across time and scale (e.g., molecular networks to cell behavior). For example, an integrative model of the embryonic transcriptome would track the ontogeny of specific RNAs or microRNAs and the protein products expressed in developing systems and correlate these measurements with measures of individual cellular behaviors at critical times following chemical insult. Functional analysis of the system through antisense (morpholino) or microRNA based approaches can serve to validate computational models.
- Cells to tissues: The ability to analyze complex cellular changes that follow chemical insult and reconstruct tissue dysfunction or dysmorphogenesis can be enhanced with various cell-based assays. For example, functional assessments of cellular differentiation have been facilitated using molecular reporters such as green fluorescent protein (GFP) coupled to promoter constructs to indicate expression of that phenotypic marker and, coupled with cell sorting techniques may enrich specific cell types for subsequent cellular and molecular analysis. Confocal microscopy, digital imaging or HCS techniques can provide high resolution analysis of specific intracellular parameters in real-time. Other technologies such as cell labeling in tissues with ‘quantum dots’, which are nanometer-sized semiconductor particles that can be used for multiplex fluorescence labeling, or Laser Capture Microdissection under optimal conditions extends the ability to characterize molecular phenotypes at the cellular scale.
- Full system modeling: Many computational models have been developed to cover a simple pathway (metabolic, regulatory, or signaling); however critical steps in developmental toxicity likely follow more complex, downstream consequences of pathway-level perturbation. To address these interlacing biological networks in embryonic systems, it would be useful to integrate high-information content data from molecular and cellular assays into computational models for morphogenesis and differentiation. Validating such methods could draw on techniques used in the development of multi-scale models, or may propose to develop reduced complexity representations of individual networks.
Model Evaluation
For every area of interest proposed for model development, the proposal must address each of the following aspects of model evaluation:
- Interplay between real and simulated data: Whereas it is expected that each center would generate new experimental data, it would be useful that each center also has access to, or deploys, computational (simulated) data to parameterize and/or validate the models being developed. This may include developing informatics tools for integrating, organizing, managing and providing open access to disparate biological experimental and reference biological data (not database development). It may also include developing visualization tools, because multiscale models are likely to be very complex.
- Model parameterization/validation: A key issue with using predictive models is the problem of determining parameters for the models. Parameters such as rate constants, reaction orders, and initial conditions that are well-known for metabolic (flux-driven) processes may not be as complete for regulatory (signal-driven) pathways and morphogenetic gradients. Improved methods are needed to determine key modeling parameters from real and computational data for developmental toxicity pathways.
- Model stability: Many models of individual cell function can produce the same output of collective cell behavior. Models may have different topology, or identical topology but different parameters, and still yield close to identical dynamic behavior. This is an issue related to model parameterization, and warrants further investigation due to its potential to impact the plausibility of an embryogenic or morphogenetic model.
- Research on characterization of the uncertainty of biologically based models and of models based on default assumptions in developmental toxicity: This would provide better tools for understanding in a rigorous and reproducible manner how to rank model uncertainty for models incorporating different amounts of biological and toxicological information into a developmental risk assessment. In other words, given two or more models for a toxicant, the preferred model is the one that is less uncertain (e.g., the model in whose predictions we have more confidence) and most plausible given existing knowledge.
Standard Model Formats, Computational Approaches, and the use of Existing datasets
Data and model exchange: The center must be prepared to exchange data and models in the standard computer language formats, such as systems biology mark-up language (SBML). Proposals should indicate existing and proposed sources of external data, including those available through currently funded EPA STAR Computational Toxicology initiative grantees, laboratories supported by the European Union’s Work Program ENV.2009.3.3.1.1, complementary research programs in EPA’s intramural laboratories, other U.S. government laboratories, or other academic or industry research laboratories from the broader scientific community.
In addition to the use of community developed model exchange formats, a strategy must be outlined for re-using existing public domain software, or plans for public release of software modules, e.g. bioconductor, bioperl, biolisp.
Modeling and simulation of biomolecular systems is very computationally intensive; therefore applicants should detail allocations of computer time and identification of computational support.
Organization of the Centers
Proposed projects are expected to be collaborative in design, and engage multiple investigators as part of a coordinated research team within an institution, or research network across institutions. This approach is designed to engage the expertise of researchers from a variety of disciplines to be focused on basic and applied research projects needed to fill human health risk assessment knowledge gaps. Research centers should consist of computational and biomedical scientists from public nonprofit institutions/organizations (this includes public institutions of higher education and hospitals), private nonprofit institutions/organizations, and national laboratories. The teams should be balanced in such a manner to make the most effective use of the resources in accomplishing their proposed objectives.
Proposed centers will have a Center Director who will have the responsibility to make certain that resources are utilized in an optimal way and the efforts of the team remain focused on the proposed research project. The Center Director will also be responsible for reports and reviews of the team's research progress.
- National Research Council, Committee on Toxicity and Assessment of Environmental Agents (2007) Testing in the Twenty-first Century: A Vision and a Strategy. National Academies Press (http://www.nap.edu/catalog/11970.html), 2007
- Collins FS, Gray GM and Bucher JR (2008) Transforming environmental health protection. Science 319: 906-907
- National Research Council (2000) Scientific Frontiers in Developmental Toxicology and Risk Assessment. National Academy Press (Washington, D.C.)
- Knudsen T.B. and Kavlock R.J. (2008) Comparative Bioinformatics and Computational Toxicology. In: Developmental Toxicology 3rd Edition (eds: B Abbott and D Hansen), Taylor and Francis, Target Organ Toxicology Series, Chapter 12, pp 311-360.
- Thorne BC, Bailey AM, DeSimone DW and Peirce SM (2007) Agent-based modeling of multicell morphogenic processes during development. Birth Def Res (Part C) 81: 344-353
- Chapin, R., Stedman, D., Paquette, J., Streck, R., Kumpf, S., and Deng, S. Struggles for equivalence: in vitro developmental toxicity model evolution in pharmaceuticals in 2006. Toxicol. in vitro 21, 1545-1551, 2007
- Yang L, Kemadjou JR, Zinsmeister C, Bauer M, Legradi J, Muller F, Pankratz M, Jakel J and Strahle U (2007) Transcriptional profiling reveals barcode-like toxicogenomics responses in the zebrafish embryo. Genome Biology 8: R227.1-R227.17
- Knowlton MN, Li T, Ren Y, Bill BR, Ellis LBM and Ekker SC (2008) A PATO-compliant zebrafish screening database (MODB): management of morpholino knockdown screen information. BMC Bioinformatics 9:7
F. Special Requirements
Agency policy prevents EPA technical staff and managers from providing individual applicants with information that may create an unfair competitive advantage. Consequently, EPA employees will not review, comment, advise, and/or provide technical assistance to applicants preparing applications in response to EPA RFAs, nor will they endorse an application or discuss in any manner how the Agency will apply the published evaluation criteria for this competition.
Multiple Investigator applications may be submitted as: (1) a single Lead Principal Investigator (PI) application with Co-PI(s) or (2) a Multiple PI application (with a single Contact PI). If you choose to submit a Multiple PI application, you must follow the specific instructions provided in Sections IV. and V. of this RFA. For further information, please see the EPA Implementation Plan for Policy on Multiple Principal Investigators (http://rbm.nih.gov/toolkit.htm).
Groups of two or more eligible applicants may choose to form a consortium and submit a single application for this assistance agreement. The application must identify which organization will be the recipient of the assistance agreement and which organizations(s) will be subawardees of the recipient.
The application must include a plan (see “Data Plan” in section IV.B.7.c.) to make available to the public all data generated from observations, analyses, or model development (primary data) and any secondary (or existing) data used under an agreement awarded from this RFA. The data must be available in a format and with documentation such that they may be used by others in the scientific community.
Each application must address the following:
- Computational Capabilities –
- Describe the computational and network resources available as well as the specific types of statistical and bioinformatic approaches that exist within the organization and how such approaches can be applied to the selected Investigative Areas.
- Highlight what new developments in computational toxicology analysis, database development, and other areas of de novo programming are proposed, as well as areas where existing computational and database resources will be utilized.
- Each Investigational Area (with a minimum of 2) should be completely described according to the APPLICATION AND SUBMISSION INFORMATION section of this RFA (see “Research Plan” in section IV.B.7.a.). Each Area is permitted a 15-page description. Individual project descriptions must explain how the Investigational Area fits into the overall Center program and relates to other projects in the proposal.
- The Center must have a Center Director.
- The Center must have an Administrative Core Unit.
- The application must include a "Center Integration Plan."
- The Center must submit a Communication and Public Outreach Strategy.
- This proposed research topic is complimentary with that of the European Union’s Work Program ENV 2009, which represents a parallel call that under section "Cooperation", Theme 6 Environment (including climate change), establishes a research grant entitled “Screening methods for assessing the toxicological and eco-toxicological properties of chemicals” (http://cordis.europa.eu/fp7/dc/index.cfm). The US-EPA strongly encourages applicants to take advantage of datasets developed by this and other academic, government, or industry research laboratories from the broader scientific community. Therefore, in conducting its research, the Center must demonstrate a willingness to use data generated by EPA and other institutions or organizations as the bases for their development and application of computational models. It is not expected for the Center to be solely responsible for the generation of new biological data, but rather focused on advancing the science of developmental toxicity risk assessment and benefiting from the use of existing databases or data available through collaborators such as with the European Union, Environmental Protection Agency, National Institute of Environmental Health Sciences, Department of Energy, Centers for Disease Control, Food and Drug Administration, or other government and academic institutions.
- New Investigators – To attract new investigators into the application of computer science and mathematics research to biological systems, the Center must partially support at least one newly recruited Center scientist. Up to $70,000 per year, direct cost, may be used for each newly required Center scientist to provide up to 75 percent salary support, technical support, equipment, and supplies. To the extent possible, the types of individuals sought and their expected roles should be described in the application if specific individuals have not been identified.
- Postdoctoral training - the Center must support a minimum of two postdoctoral scientists (at least one of whom is to be newly recruited) and provide a research environment in leading-edge informatics/computational tools or approaches, as well as a structured mentoring and career development opportunity (see “Center Description” in section IV.B.5.).
- Grantees Meetings – Centers must budget for an annual grantees meeting that should be attended by the Center director and the appropriate project leads. It can be assumed that all meetings will be held at the U.S. EPA research facility in RTP, NC. Annual meetings will be open to the public and will bring together all researchers currently supported through STAR computational toxicology efforts as well as U.S. EPA scientists.
It is anticipated that a total of approximately $3.2 million will be awarded under this announcement, depending on the availability of funds and quality of applications received. The EPA anticipates funding approximately 1 award under this RFA. Requests for amounts in excess of a total of $3,200,000, including direct and indirect costs, will not be considered. The total project period requested in an application submitted for this RFA may not exceed 4 years. The EPA reserves the right to reject all applications and make no awards, or make fewer awards than anticipated, under this RFA. The EPA reserves the right to make additional awards under this announcement, consistent with Agency policy, if additional funding becomes available after the original selections are made. Any additional selections for awards will be made no later than six months after the original selection decisions.
EPA may award a grant or cooperative agreement under this announcement.
Under a grant, EPA scientists and engineers are not permitted to be substantially involved in the execution of the research. However, EPA encourages interaction between its own laboratory scientists and grant Principal Investigators after the award of an EPA grant for the sole purpose of exchanging information in research areas of common interest that may add value to their respective research activities. This interaction must be incidental to achieving the goals of the research under a grant. Interaction that is “incidental” does not involve resource commitments.
Where appropriate, based on consideration of the nature of the proposed project relative to the EPA’s intramural research program and available resources, the EPA may award cooperative agreements under this announcement. When addressing a research question/problem of common interest, collaborations between scientists and the institution’s principal investigators are permitted under a cooperative agreement. These collaborations may include data and information exchange, providing technical input to experimental design and theoretical development, coordinating extramural research with in-house activities, the refinement of valuation endpoints, and joint authorship of journal articles on these activities. Proposals may not identify EPA cooperators or interactions; specific interactions between EPA’s investigators and those of the prospective recipient for cooperative agreements will be negotiated at the time of award.
A. Eligible Applicants
Public nonprofit institutions/organizations (includes public institutions of higher education and hospitals) and private nonprofit institutions/organizations (includes private institutions of higher education and hospitals) located in the U.S., state and local governments, Federally Recognized Indian Tribal Governments, and U.S. territories or possessions are eligible to apply. Profit-making firms are not eligible to receive assistance agreements from the EPA under this program.
Eligible nonprofit organizations include any organizations that meet the definition of nonprofit in OMB Circular A-122, located at 2 CFR Part 230. However, nonprofit organizations described in Section 501(c) (4) of the Internal Revenue Code that lobby are not eligible to apply.
National laboratories funded by Federal Agencies (Federally-Funded Research and Development Centers, “FFRDCs”) may not apply. FFRDC employees may cooperate or collaborate with eligible applicants within the limits imposed by applicable legislation and regulations. They may participate in planning, conducting, and analyzing the research directed by the applicant, but may not direct projects on behalf of the applicant organization. The institution, organization, or governance receiving the award may provide funds through its assistance agreement from the EPA to an FFRDC for research personnel, supplies, equipment, and other expenses directly related to the research. However, salaries for permanent FFRDC employees may not be provided through this mechanism.
Federal Agencies may not apply. Federal employees are not eligible to serve in a principal leadership role on an assistance agreement, and may not receive salaries or augment their Agency’s appropriations in other ways through awards made under this program.
The applicant institution may enter into an agreement with a Federal Agency to purchase or utilize unique supplies or services unavailable in the private sector. Examples are purchase of satellite data, census data tapes, chemical reference standards, analyses, or use of instrumentation or other facilities not available elsewhere. A written justification for federal involvement must be included in the application. In addition, an appropriate form of assurance that documents the commitment, such as a letter of intent from the Federal Agency involved, should be included.
Potential applicants who are uncertain of their eligibility should contact William Stelz (stelz.william@epa.gov) in NCER, phone (202) 343-9802.
B. Cost-Sharing
Institutional cost-sharing is not required.
C. Other
Applications must substantially comply with the application submission instructions and requirements set forth in Section IV of this announcement or they will be rejected. In addition, where a page limitation is expressed in Section IV with respect to parts of the application, pages in excess of the page limit will not be revie
The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.