PxER
-------
                                            EPA 100/B-04/002
                                              December 2004
  U.S. Environmental Protection Agency
Potential Implications of Genomics for Regulatory
    and Risk Assessment Applications at EPA
   Prepared for the U.S. Environmental Protection Agency
         by members of the Genomics Task Force,
        a workgroup of EPA's Science Policy Council
                 Science Policy Council
          U.S. Environmental Protection Agency
                 Washington, DC 20460
                                      /'Y~y Becycled/Becyclable
                                          Printed with vegetable-based ink on
                                          paper that contains a minimum of
                                          50% post-consumer fiber content
                                          processed chlorine free.

-------
Page ii                   Genomics Task Force White Paper

                               DISCLAIMER

      This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication and distribution. Mention of trade names or
commercial products does not constitute endorsement of recommendation for use.

-------
                    Genomics Task Force White Paper
                       Page iii
                    Genomics Task Force
     A Workgroup of the Science Policy Council
                         Task Force Co-Chairs
        William Benson
           NHEERL
Office of Research and Development
     Kerry Dearfield
Office of the Science Advisor
                          Task Force Members
      Michael Brody, OCFO
     Anne Fairbrother, ORD
     Kathryn Gallagher, OSA
        Jafral Hasan, OW
     Lee Hofmann, OSWER
        Jack Jones, ORD
Rebecca Klaper, AAAS Fellow, ORD
       David Lattier, ORD
      Susan Lundquist, OEI
    Nancy McCarroll, OPPTS
    Elizabeth Mendez, OPPTS
   Gregory Miller, OPEI
     Ines Pagan, ORD
   Maria Pimentel, OAR
   Julian Preston, ORD
   Philip Sayre, OPPTS
    Rita Schoeny, OW
   Jennifer Seed, OPPTS
Bobbye Smith, Region 9 RSL
    Anita Street,  ORD
  Richard Troast, OSWER
               Task Force Leads for the Science Policy Council
        Lawrence Reiter
           NHEERL
Office of Research and Development
       Vanessa Vu
  Science Advisory Board
 Office of the Administrator

-------
^	Genomics Task Force White Paper                    Page v

                          TABLE OF CONTENTS

FOREWORD	Page vii

ACKNOWLEDGMENTS	  Page viii

ACRONYMS	Page ix

EXECUTIVE SUMMARY	 Page 1

I. Introduction	Page 3
      A.  Background 	,	.	 Page 3
      B.  Emerging Impacts of Genomics Technologies	Page 4
      C.  Overview of Genomic Science	Page 5
      D.  Purpose and Intent of this Document	Page 7

II. Regulatory Applications	  Page 10
      A.  Prioritization of Contaminants (Chemicals and Microbes) and Contaminated Sites
              	Page 10
             1. Introduction	  Page 10
             2. Regulatory and Voluntary Activities Potentially Affected by Genomics
                   Information	Page 11
                   a. Representative Activities 	  Page 11
                   b. Additional Activities	Page 11
      B.  Monitoring	Page 12
             1. Introduction	  Page 12
             2. Regulatory Activities Potentially Affected by Genomics Information .  Page 13
                   a. Representative Activities 	Page 13
                   b. Additional Activities	  Page 14
      C.  Reporting Provisions	  Page 14
             1. Introduction	Page 14
                   a. Reporting on Adverse Effects of Commercialized Chemicals and
                          Pesticides	Page 14
                   b. Toxics Release Inventory Program	Page 15
             2. Regulatory Activities Potentially Affected by Genomics Information .  Page 15

HI.  Risk Assessment	  Page 17
      A.  Introduction	Page 17
      B.  Mode of Action	  Page 17
             1. Overview	Page 17
             2. Risk Assessment Activities Potentially Affected by Genomics Information
                    	Page 18

-------
Page vi                     Genomics Task Force White Paper

                    a. Representative Activities ,	Page 18
                           i.  Improving Hazard Identification 	Page 18
                           ii. Improving Dose-Response Assessment	Page 19
                           iii. Improving Extrapolations 	Page 20
                           iv. Improving Exposure Assessment	Page 21
      C. Susceptible Populations and Sensitive Life Stages	Page 22
              1. Overview	Page 22
                    a. Susceptible Human Populations and Sensitive Life Stages  ...  Page 22
                    b. Susceptible Wildlife Populations and Sensitive Life Stages  ..  Page 23
              2. Risk Assessment Activities Potentially Affected by Genomics Information
                     	Page 24
                    a. Susceptible Human Populations and Sensitive Life Stages  ...  Page 24
                           i.  Representative Activities	,	,	Page 24
                    b. Susceptible Wildlife Populations and Sensitive Life Stages  ..  Page 26
                           i.  Representative Activities	  Page 26
      D. Mixtures	Page 27
              1. Overview	Page 27
              2. Risk Assessment Activities Potentially Affected by Genomics Information
                     	Page 28
                    a. Representative Activities	Page 29
                    b. Additional Activities	Page 29

IV. Research Needs and Activities	Page 30
      A. Introduction 	Page 30
      B. Research Needs and Activities for Regulatory Applications	Page 31
              1. Prioritization	Page 31
              2. Monitoring	Page 32
      C. Research Needs and Activities for Risk Assessment	Page 34
              1. Mode of Action	Page 34
              2. Susceptible Populations and Life Stages	  Page 35
              3. Mixtures	Page 37

V.  Challenges and Recommendations  	  Page 39
       A. Research Challenges	Page 40
              1. Linking Genomics Information to Adverse Outcomes	Page 40
              2. Interpretation of Genomics Information for Risk Assessment	Page 40
              3. Recommendations to Address Research Challenges 	Page 41
       B. Technical Development Challenges	Page 42
              1. Framework for Analysis and Acceptance Criteria for Genomics Information
                     			Page 42
              2. Recommendations to Address Technical Development Challenges...  Page 42
       C. Capacity/Human Capital Challenges	  Page 43
              1. Applying Strategic Hiring Practices to Recruit Individuals Who Possess

-------
                         Genomics Task Force White Paper                  Page vii

                   "Genomics Core Competencies"	Page 43
             2. Training EPA Risk Assessors and Managers to Interpret and Understand
                   Genomics Data in the Context of a Risk Assessment	Page 43
             3. Recommendations to Address Capacity/Human Capital Challenges  .. Page 43
      D. Summary Recommendation	Page 44

REFERENCES	 Page 46

Appendix A: Interim Policy on Genomics	Page A-l

Appendix B: Details on Additional Activities Potentially Affected by Genomics Information
       	 Page B-l
      I.  Prioritization	Page B-l
             A. Program Offices	Page B-l
                   OPPT - Premanufacture Notices	 Page B-l
                   OPPT - Voluntary Children's Chemical Evaluation Program ... Page B-l
                   OPPTS/OSCP - Endocrine Disrupters Screening Program	Page B-l
                   OPP - Prioritization of Pesticides and Inert Chemicals	Page B-l
                   OW - Prioritizing Streams or Wetlands for Study or Clean Up .. Page B-l
                   OAR - Hazardous Air Pollutants	 Page B-l
                   bSWER - Superfund	Page B-l
                   ORD - Research Planning 	Page B-2
             B. Regions, States, and Tribes 	Page B-2
                   Resource Prioritization for Site Remediations and Chemical Evaluations
                           	 Page B-2
      n. Monitoring 	 Page B-2
             A. Program Offices	 Page B-2
                   OPP - Exposure Monitoring	Page B-2
                   OAR/Office of Air Quality Planning and Standards (OAQPS) - Stationary
                         Source Monitoring	Page B-2
                   OSWER - Superfund Monitoring  	 Page B-2
                   OSWER/OSW - RCRA-Required Monitoring		Page B-2
                   OEI, ORD - Bioindicator Development	Page B-3
             B. Regions, States, and Tribes 	 Page B-3
                   State and Local Beach Closures	 Page B-3
                   Air Quality Monitoring Program	Page B-3
                   Endocrine-Disruptor Monitoring	Page B-3
                   Regional Pesticide Program Decisions 	Page B-3
      III. Risk Assessment	 Page B-4
             A. Program Offices	Page B-4
                   OSWER, OAR, OW - Site Assessments/Monitoring	 Page B-4
             B. Regions, States, and Tribes	 Page B-4
                   Site Assessments/Monitoring	 Page B-4

-------
Page viii	Genotnics Task Force White Paper
Appendix C; Glossary	Page C-l
       Glossary References	 Page C-5

-------
                           Genomics Task Force White Paper                    Pageix

                                  FOREWORD

       Genomics information has great potential to enhance assessment of risks to human health
and the environment and will have significant implications for EPA's risk assessment practice
and regulatory decision making.  EPA's Interim Policy on Genomics (2002) states that while
genomics data may be considered in the decision making process at this time, these data alone
are insufficient as a basis for decisions, and will be considered for assessment purposes on a
case-by-case basis only.

       Following the release of the Interim Policy, at the request of the Science Policy Council
(SPC), a cross-Agency Genomics Task Force was formed.  The Task Force was charged with
examining the broader implications genomics is likely to have on EPA programs and policies,
and with developing scenarios to describe various circumstances under which EPA might receive
these data and the resulting implications for EPA policies and programs. Thus, the purpose of
this document is to present exemplary applications and resultant implications of the use of
genomics technologies in EPA practice for the consideration of Agency managers. Although, as
the Interim Policy notes, understanding genomic responses with respect to adverse ecological
and/or human health outcomes is far from established, it is important for managers to begin to
consider the likely future impacts of genomics technologies on their programs.

       It is clear that genomics technologies have great potential to enhance assessment of risks
to human health and the environment, and the Agency must be proactive in preparing itself to
address the oncoming challenges associated with interpreting and applying genomics
information. It is essential for EPA to continue to collaborate with other federal agencies,
academia, the regulated community, public interest groups, and other stakeholders in this
endeavor in order to benefit from ongoing advances in genomics in the  wider scientific and
regulatory communities.

       I want to acknowledge and thank the Genomics Task Force for their efforts in assembling
this document. I particularly appreciate the efforts of the SPC co-chairs, Larry Reiter and
Vanessa Vu, and the Task Force co-chairs, Kerry Dearfield and Bill Benson in leading the Task
Force.

       It is with great pleasure that I present the Potential Implications of Genomics for
Regulatory and Risk Assessment Applications at EPA,

                                        Paul Oilman, Ph.D.
                                        EPA Science Advisor
                                        Chair, Science Policy Council

-------
Page x                   Genomics Task Force White Paper

                         ACKNOWLEDGMENTS
      The Genomics Task Force would like to acknowledge the Science Policy Council and its
Steering Committee for their recommendations and contributions to this document. In particular,
we would like to acknowledge the following individuals who provided comments through the
internal peer review process: Roland Hemmett (Region 2), Carl Mazza (OAR), Michael
Firestone (OCHP), Claudia Walters (ORD), Marian Olsen (Region 2), and Craig Annear (OGC).
We would also like to acknowledge program office and regional staff who provided comments
on the document to Task Force members: including Nicole Paquette (OEI), Stephen Devito
(OEI), Edward Bender (OSA), and Karen Hammerstrom (ORD). We also appreciate the
comments from the external peer reviewers, whose comments and suggestions enhanced the final
document

-------
                          Genomics Task Force White Paper
Page xi
                                 ACRONYMS

ATSDR      Agency for Toxic Substances and Disease Registry
BBDR       Biologically-Based Dose Response
BST         Bacteriological Source Tracking
CAA         Clean Air Act
CCL         Contaminant Candidate List
CERCLA     Comprehensive Environmental Response, Compensation and Liability Act
CIIT         Chemical Industry Institute of Technology
CWA        Clean Water Act
CYP         Cytochrome P-450
DMA         Deoxyribonucleic Acid
EDC         Endocrine Disrupting Chemical
EDSP        Endocrine Disrupters Screening Program
EPA         Environmental Protection Agency
ELSI         Ethical, Legal, and Social Implications
EPCRA      Emergency Planning and Community Right-to-Know Act
EUP         Experimental Use Permit
FDA         Food and Drug Administration
FlFRA       Federal Insecticide, Fungicide and Rodenticide Act
FQPA        Food Quality Protection Act
GM          Genetically Modified
HAPs        Hazardous Air Pollutants
HPV         High Production  Volume
ICCVAM     Interagency Coordinating Committee on the Validation of Alternative Test
             Methods
ILSI         International Life Sciences Institute
IRIS         Integrated Risk Information System
LOAEL      Lowest Observed Adverse Effect Level
MO A        Mode of Action
MRA         Microbial Risk Assessment
MST         Microbial Source Tracking
NHEERL     National Health and Environmental Effects Research Laboratory
NIEHS       National Institute of Environmental Health Sciences
NPDES      National Pollutant Discharge Elimination System
NOAEL      No Observed Adverse Effect Level
OAQPS      Office of Air Quality Planning and Standards
OAR         Office of Air and Radiation
OCFO       Office of the Chief Financial Officer
OCHP       Office of Children's Health Protection
OECD       Organization for  Economic Cooperation and Development
OEI          Office of Environmental Information

-------
Page xii
Genomics Task Force White Paper
OGWDW    Office of Ground Water and Drinking Water
OPEI        Office of Policy, Economics and Innovation
OPP         Office of Pesticide Programs
OPPT        Office of Pollution Prevention and Toxics
OPPTS       Office of Prevention, Pesticides and Toxic Substances
OKD        Office of Research and Development
OSA        Office of the Science Advisor
OSWER      Office of Solid Waste and Emergency Response
OW         Office of Water
PBPK        Physiologically-Based Pharmacokinetic
PCR         Polymerase Chain Reaction
PMN        Premanufacture Notice
POD        Point of Departure
QSAR       Quantitative Structure Activity Relationship
RCRA       Resource Conservation and Recovery Act
RED        Reregistration Eligibility Decision
RfC         Inhalation Reference Concentration
RfD         Oral Reference Dose
RNA        Ribonucleic Acid
RT-PCR      Reverse-Transcription Polymerase Chain Reaction
SAR        Structure Activity Relationship
SDWA       Safe Drinking Water Act
SIDS        Screening Information Data Set
SNP         Single Nucleotide Polymorphism
SPC         Science Policy Council
TMDL       Total Maximum Daily Load
TRI         Toxics Release Inventory
TSCA       Toxic Substances Control Act
UF          Uncertainty Factor
VCCEP      Voluntary Children's Chemical Evaluation Program

-------
	Genomics Task Force White Paper	Page 1

                           EXECUTIVE SUMMARY

       Advances in genomics will have significant implications for risk assessment practice and
regulatory decision making. The Environmental Protection Agency's (EPA's) Interim Policy on
Genomics (Appendix A), issued in 2002, appropriately acknowledges that genomics technologies
have the potential to improve our understanding of an organism's response to stressors (USEPA,
2002a). The Interim Policy describes          	
genomics as the study of all the genes of a     [~~~     "'T~  ""   .   ,
      .        ,  TV*,.,   r,xT*      4.         'Genomics analysis is the study of all the genes
cell or tissue, at the DNA, mRNA, or protein      f   ,,    .,      .,,  r,XTA ,     ,   ^
                                  v          of a cell or tissue, at the DNA (genotype),
                                             mRNA (transcriptome), or protein (proteome)
                                             level." Interim Policy on Genomics
level. This policy states that while genomics
data may be considered in the decision
making process at this time, these data alone
are insufficient as a basis for decisions. EPA
will consider genomics information for
assessment purposes on a case-by-case basis only. Following release of the Interim Policy, EPA
held internal discussions regarding the potential of genomics approaches to improve our
understanding of the effects of environmental stressors on cells. It was concluded that genomics
information may lead to the development of predictive biomarkers of effect, thereby allowing for
the identification of potentially sensitive populations and earlier predictions of adverse outcomes
and, ultimately, leading to better intervention strategies.  Enhancing understanding of the
molecular mechanisms of toxicity may greatly reduce the uncertainty of extrapolations used in
the current risk assessment process. Further, genomics technologies may enhance the
development of more sensitive and cost-effective methods for toxicity screens and tests and may
ultimately lead to the reduction, refinement, or replacement of more complex  and costly standard
tests for human and wildlife species.

       Following these internal discussions, at the request of EPA's Science Policy Council
(SPC), a Genomics Task Force was formed. The Task Force was charged with examining the
broader implications genomics is likely to have for EPA programs and policies, to attempt to
gain further understanding of the appropriate usage of these data and the potential consequences
of their use, as well as to identify possible infrastructure needs. The Task Force was also charged
with developing scenarios to describe various circumstances under which EPA might receive
these data.  The resulting document is intended to present implications of the  use of genomics
technologies in EPA practice for the consideration of Agency managers. It is  the intent of the
Genomics Task Force  to initiate a discussion of the scientific issues regarding the incorporation
of genomics information into human  health and ecological risk assessments and of how these
data will likely affect regulatory policy and decision making in  the future. Although, as the
Interim Policy notes, understanding genomic responses with respect to adverse ecological and/or
human health outcomes is far from established, it is important for managers to begin to consider
the likely future impacts of genomies technologies on their programs. Four areas have been
identified as those very likely to be influenced by the generation of genomics  information within
EPA and the submission of such information to EPA: (1) prioritization of contaminants and

-------
Page 2                     Genomics Task Force White Paper

contaminated sites, (2) monitoring, (3) reporting provisions, and (4) risk assessment. This
document briefly addresses ongoing research within the Agency for each of these four areas and
identifies remaining research needs.  It should be noted that genomics will not fundamentally
alter the risk assessment process, but is expected to serve as a new, more powerful tool for
evaluating the exposure to and effects of environmental stressors.

       The Task Force identified several overarching challenges associated with genomics that
fall into three categories: research, technical development, and capacity. These challenges are
defined as critical needs for the Agency to strengthen its capability to use genomics information
in a meaningful way, and to enable the Agency to address potential regulatory applications that
are likely to arise with respect to genomics, such as those outlined in this paper. For research, the
critical needs are identified as (1) linking genomics information to adverse outcomes; and (2)
interpreting genomics information for risk and hazard assessment. It is important to note that
significant research by EPA and other agencies and researchers will be necessary to fully
understand and apply genomics technologies to human health and ecological risk assessment.
One critical need in the area of technical development was identified as the need to establish a
framework for analysis and acceptance criteria for genomics information for scientific and
regulatory purposes (including data quality standards based on genomic assay performance).
Two critical needs were identified with respect to capacity, including human capital: (1)
applying strategic hiring practices to recruit individuals who possess "genomics core
competencies" essential for crucial areas of research, analysis, systems biology, bioinformatics,
and risk assessment; and (2) training EPA risk assessors and managers to interpret and
understand genomics data in the context of a risk assessment.

       Though advances  in genomics and chemical development will present the Agency with
new challenges, it is likely that genomics approaches will greatly assist in advancing EPA's risk
assessment and regulatory policy and decision making processes. The Agency must be proactive
in identifying, developing, and standardizing applicable genomics approaches. Additionally,
many scientific, policy, ethical, and legal concerns are developing along with the emergence of
this science and will need to be addressed.  The Genomics Task Force recommends that EPA
begin taking steps to address the identified research, technical development,  and capacity
challenges in order to strengthen its capability to effectively use genomics information in the
future. Recommendations for initial steps to address these challenges are presented in the final
section of this paper. It is essential for EPA to continue to collaborate with other federal
agencies, academia, the regulated community, public interest groups, and other stakeholders in
this endeavor in order to benefit from ongoing advances in genomics in the wider scientific and
regulatory communities.

-------
                           Genomics Task Force White Paper                      Page 3

I. Introduction

A.  Background

       The mapping of the genomes of diverse animal, plant, and mierobial species, and related
technologies are already significantly affecting research across all areas of the life sciences and
will continue to do so for decades to come.  The current understanding of biological systems is
rapidly changing in ways previously unimagined, and novel applications of this technology are
already being commercialized.  These scientific and technological advances have spurred many
federal agencies to consider the far-reaching implications for policy, regulation, and society as a
whole.

       On June 25, 2002, EPA released the Interim Policy on Genomics (Appendix A, USEPA,
2002a) communicating its initial approach to using genomics information in risk assessment and
decision making.  The policy describes genomics as the study of all the genes of a cell or tissue,
at the DNA (genotype), mRNA (transcriptome), or protein (proteome) level.  The Interim Policy
notes that, while genomics offers the opportunity to understand how an organism responds at the
gene expression level to stressors in the environment, understanding such molecular events with
respect to adverse ecological and/or human health outcomes is far from established. It concludes
that while genomics data may be considered in the decision making process at this time, these
data alone are insufficient as a basis for decisions. Therefore, EPA will consider genomics
information for assessment purposes on a case-by-case basis only.

       Following release of the Interim Policy, EPA held internal discussions to consider the
potential genomics technologies have to improve our understanding of the effects of
environmental stressors on cells. It was concluded that genomics information may lead to the
development of predictive biomarkers of effect, thereby allowing for the identification of
potentially sensitive populations and earlier predictions of adverse outcomes and, ultimately,
leading to better intervention strategies.  Enhancing understanding of the molecular mechanisms
of toxicity could greatly reduce the uncertainty of extrapolations used in the current risk
assessment process. The potential results may be the development of more sensitive and cost-
effective methods for toxicity screens and tests and significant reductions in, or eventual
elimination of, conventional animal testing.

       Following these discussions, at the request of the Science Policy Council (SPC), a
Genomics Task Force was formed.  The  Task Force was charged with the task of examining the
broader implications genomics is likely to have on EPA programs and policies,  to attempt to gain
further understanding of the appropriate  usage of these data and the potential consequences of
their use, as well as to identify possible infrastructure needs. The Task Force was also charged
with developing scenarios to describe various circumstances under which EPA might receive
these data and the resulting implications (e.g., interpretation, relevance, evaluation, analytical,
and research needs) for EPA policies and programs.

-------
Page 4                     Genomics Task Force White Paper

B.  Emerging Impacts of Genomics Technologies

       While these are new technologies and most are not as yet ready for application in risk
assessment and decision making, it is important for Agency managers to begin to consider the
likely future impacts of genomics technologies on their programs.  It should be noted that
genomics will not fundamentally alter the risk assessment process, but is expected to serve as a
more powerful tool for evaluating the exposure to and effects of environmental stressors and will
offer a means to simultaneously examine a number of response pathways.  EPA and other
regulatory agencies are beginning to address the use of genomics data for various risk assessment
applications, including the need to establish a link between genomic alterations and adverse
outcomes of regulatory concern (Klaper et al., 2003). EPA must soon develop an explicit
prescriptive strategy for accepting "omics" data submissions because such information (Genter et
al., 2002) has already been referenced in a submission for a pesticide reregistration.  Given the
rapidly evolving nature of genomics technologies, care must be taken to develop an acceptable
scheme to simplify and refine the risk-related information and to distinguish it from the large
amount of complex scientific and statistical data available.  This strategy must remain dynamic in
anticipation of continuing technical evolution  at the molecular levels (e.g., DNA, RNA, and
protein).  Furthermore, bioinformatic approaches for data acquisition and analysis, including
technologies designed to store and analyze the profusion of data generated from microarray
analysis, must be considered in parallel with the data-generating methods. Additionally, many
scientific, policy, ethical, and legal concerns are developing along with the emergence of this
science and will need to be addressed.

       The Interim Policy on Genomics provides guidance concerning how and when genomics
information should be used to assess the risks  of environmental contaminants under the various
regulatory programs implemented by the Agency at the present time. The standardization of
experimental  design and data analysis for genomics is important for the utility of genomics
information in future risk assessment and regulatory decisions. Such standardization will
enhance the reproducibility of results obtained and the reliability of conclusions drawn from
these data.  Furthermore, EPA should consider the development of data quality standards based
on performance of microarrays, as well as other genomics technologies. This in turn will help to
ensure the integrity of EPA's approach to assessing the genomics information submitted to the
Agency.

       As noted above, genomics issues have already arisen in environmental decision making.
For example, a pesticide registrant has cited a published genomic article (Genter et al., 2002) in
their data package submission for product registration to EPA's Office of Pesticide Programs.
The citation was referenced to propose an alternative mode of action that would affect human
health assessment conclusions. Additional, similar submissions may soon be made by other
pesticide registrants.

-------
                           Genomics Task Force White Paper                     Page 5

       There are a number of other regulatory areas where genomics information will start
having an impact. For example, a research consortium including State of California regulatory
agencies, public utilities, and EPA recently participated in a study comparing the performance of
various genomics-based methods designed to identify the source of fecal material in ambient
waters in an approach called microbial source tracking (Griffith et al., 2003). These methods are
being evaluated to assist dischargers in complying with Clean Water Act (CWA) requirements to
develop Total Maximum Daily Loads (TMDLs) for water bodies that are listed as impaired due
to the presence of fecal coliforms. This work will also address the issue of beach closures;
current microbial methods require several days to complete and do not distinguish between
bacteria from humans and other sources such as  sea gulls or seals. In another application, the
State of California, as part of an ongoing ambient water quality monitoring program, is initiating
an effort to evaluate surface waters for the presence of estrogenic endocrine effects using a
reverse-transcription polymerase chain reaction (RT-PCR) assay for vitellogenin gene expression
in livers of exposed male rainbow trout.  If results show that some surface waters exhibit
estrogenic effects, the Regional Water Quality Controls Boards in California, which issue
National Pollutant Discharge Elimination System (NPDES) permits and perform ambient water
quality monitoring, may begin to consider including this bioassay in their monitoring program for
wastewater treatment facilities even though it is  not yet an approved "EPA method."
Additionally, one group of tribes in Northern California and Southern  Washington proposes to
use a series of molecular-biology-based assays to assess exposure to  hormonally active
compounds using either a multiplex RT-PCR approach or a multigene  array.  The information
could ultimately be used to establish Tribal Water Quality Standards.

       These examples indicate the emerging need to make proactive policy decisions and to
develop processes to address  how genomics data will be used in Agency decision making.

C.  Overview of Genomic Science

       Genomics tools provide the observer with a means to examine changes in gene
expression and protein and metabolite profiles within the cells of any organism, in contrast to
older methods of analyses which restrict
observers to looking only at whole organism
effects or changes in single biochemical
pathways. Genomics tools can provide          significant implications for risk assessment and
j,.,^,.,.,,     j  i                   regulatory decision making.
detailed data about the underlying                      J             
Rapid advances in genomics may have
biochemical mechanisms of disease or           ^^^^^^^^^^^^^^^^^^^^^^m
toxicity (i.e., disease etiology), sensitive
measures of exposures to chemicals, new approaches to detecting effects of such exposures, and
methods for predicting genetic predispositions that may lead to disease or higher sensitivity to
particular stressors in the environment.

-------
Page 6                     Genomics Task Force White Paper

       As a means of introduction to genomics and its potential impact on regulatory decision
making, it is important to understand the basic principles behind the technology. Only about 1-
2% of the human DNA actually codes for RNA message that can be translated into a protein.
This 1-2% is considered the theoretical functional genome. Any particular cell type (i.e., from
various organs or species) will have its own practical functional genome, which is a subset of the
entire functional genome that encodes the proteins actually functioning in that cell.  The
functional genome for any cell type can be assessed by measuring its messenger RNA (mRNA)
profile. The mRNA copies the necessary portion of the cell's DNA code and takes the
information to the place in the cell where proteins are manufactured. Thus, the assessment of
mRNA profiles is called functional genomics. Such profiles are constructed using muroarrays
that contain all (or a sampling) of a cell's functional genome. Hybridization of the mRNA that is
being actively produced by the cell to these microarrays demonstrates which genes are currently
active in that cell.  Within the 98-99% of DNA not coding for RNA message is information that
affects the activity of the functional genome by influencing where and when genes are active in
an organism. Thus both coding and noncoding DNA are important in organismal function and
response to perturbations. The oft-repeated statement that no two humans are alike  (with the
exception of identical twins) is valid at the genomics level as well.  There is a wide range of
DNA among individuals, even within the same family.  Some of these differences arise
spontaneously (but rarely) as mutations. Others are more frequent  and represent very small DNA
alterations that might or might not affect gene function; these are called single nucleotide
polymorphisms (SNPs). While measurement of SNPs is not difficult, the need remains to
associate these mutations or polymorphisms with specific genetic traits or cellular activities that
could lead to adverse health outcomes.

       The  study of a cell's protein composition is calledproteomics. Currently, it is possible to
analyze only a fraction of a cell's proteins, but rapid advances in this field should allow more
complete profiling in the near future. Another discipline of biology analyzes biofluids and
tissues to determine the profiles of endogenous metabolites present under normal conditions or
when the organism has been affected by factors such as exposure to environmental chemicals.
This type of whole cell analysis is called metabonomics (or metabolic profiling). In order to
understand how a cell functions under normal or stressed circumstances, it is necessary to
characterize the proteins that are manufactured by the cell,  as well  as endogenous metabolites.
This facilitates an understanding of global metabolism and how proteins interact along cellular
activity pathways.  This approach describes the area of systems biology, in which the cell, tissue,
or organism is considered as a complete, albeit complex, system.

       For the purposes of this document it is  	
important to note that all of these so-called     I  _. .       ".    ""        ~"
    ^  .   i        u     j *              Bioinformatics is data acquisition and
 omic technologies can be used to compare            .      ,   ,  .   ,       ,
f   ^.   i          f  r,XT A\   j    *           processing technologies designed to store
functional genomes (mRNA) and proteomic        ,   ,    .            _, <
   ,   , ,      .     r-i          i  11   j    and analyze data generated from genomic
and metabonomic profiles in normal cells and       i
tissues with responses in stressed cells and          ^

-------
                           Genomics Task Force White Paper                     Page 7

tissues such as those exposed to environmental agents. Analysis of the large data sets generated
for these type of analyses requires the development of new bioinformatic and computational
tools. An integrated analysis and understanding of biological systems and their responses to
perturbation, from genes to adverse effects, and the capacity to collect and evaluate data
supportive of such a view would be expected to greatly enhance the risk assessment process and,
thus, aid in formulating regulatory policy and making regulatory decisions.

       As the Agency considers the significance of the current state of genomics technologies, it
is critical to realize that these technologies continue to advance at very rapid rates. Some of the
technology "laws" that have been developed to describe this advance include Monsanto's Law,
"the amount of useful genetic information doubles every 18-24 months," and Dawkin's Law,
"the cost of sequencing DNA base pairs halves every 27 months."  As an example, a commercial
producer of gene chips has reported that the information content of their chips has been growing
exponentially from 16,000 cDNA probes per chip in 1994 to over 500,000 in 2002. While
commercial enterprises have recently developed arrays capable of handling large portions of the
human genome, the cost of such microarray technology is still high enough that its use in
toxicity-based approaches is relatively limited. However, these costs are falling rapidly as the
broad utility of the technology becomes apparent.  In this regard, it is estimated that in
approximately two years, costs will decline to where clinical use will be feasible (e.g., assist in
selecting treatments, intervene with disease before overt symptoms occur, offer genetically
personalized nutrition  and lifestyle advice, customize drug prescriptions), and the resulting
economies  of scale will contribute to a continuing decline in costs to the research community
(Personal communication, from interviews  by Robert Olson, Research Director, Institute for
Alternative Futures, September 2003).

D. Purpose and Intent of this Document

       This is an unprecedented time in the history of science because of the rapid development
of genomics and associated technologies. Genomics technologies are becoming highly
sophisticated  and have great potential for contributing to the assessment and management of
environmental risks. The challenge lies in
understanding how this information is likely    I _,          . ,. .        '.
   ,             .              .              The purpose of this document is to present
to  change current Agency approaches to         >?**,     f           A
       0          o   j ft-                  implications or the use 01 genomics and
                                             associated technologies in Agency decisions.
human and ecological risk assessment and
decision making.  Although EPA recognizes
the inherent issues currently associated with
genomics studies in general and microarray
experiments in particular and that these issues will need to be addressed before this technology
can be fully accepted in risk assessment, the Agency also recognizes that genomics information
will likely become an integral part of risk analysis in the future. The purpose of this document,
consequently, is to present implications of the use of genomics technologies in Agency practice.

-------
Page 8                     Genomics Task Force White Paper

It is also the intent of the Genomics Task Force to inform, invite discussion, and shed light on
issues that will need to be addressed now or in the near future.

       A key charge for the Task Force was identifying various exemplary circumstances under
which EPA might receive genomics data and the resulting implications for EPA policies and
programs. Sections n and ffl outline these circumstances or applications.

       Section II identifies examples of regulatory applications in which genomics will likely
affect regulatory decision making;

       a)     Prioritization of Contaminants and Contaminated Sites

       b)     Monitoring

       c)     Reporting Provisions

       Section ffl addresses areas where genomics will likely have applications for risk
assessment practices.  The risk assessment applications will also serve as tools for regulatory
applications and decision making. For each of the regulatory and risk assessment applications,
select representative activities are presented to illustrate the application.  Additional activities are
identified and described in Appendix B.

       Section IV identifies genomics research needs and provides an overview of current EPA
genomics research that may aid in addressing the regulatory and risk assessment applications
outlined in Sections II and EL

       Section V describes  three categories of challenges EPA faces in  applying genomics
information to risk assessment and decision making and provides recommendations for
addressing these challenges.

       a)     Research

              1)     Linking genomics information to adverse outcomes

             2)     Interpreting genomics information for risk assessment

       b)     Technical Development

              1)     Establishing a technical framework for analysis and acceptance criteria for
                     genomics information (including data quality standards based on genomic
                     assay performance)

-------
                    Genomics Task Force White Paper                     Page 9

c)     Capacity/Human Capital

       1)     Applying strategic hiring practices to recruit individuals who possess
             genomics core competencies essential for crucial areas of research,
             analysis, systems biology, bioinformatics, and risk assessment

       2)     Training EPA risk assessors and managers to interpret and understand
             genomics data in the context of risk assessment

-------
Page 10                    Genomics Task Force White Paper

II. Regulatory Applications

A.  Prioritization of Contaminants (Chemicals and Microbes) and
Contaminated Sites

1. Introduction

       There are over 80,000 chemicals currently listed in the Toxic Substances Control Act
(TSCA) Inventory (Personal communication, Dr. Henry Lau, USEPA, March 2004); a portion of
these chemicals may no longer be in commerce. Sufficient information to allow a thorough
evaluation of risk exists for only a fraction of
these chemicals since most of them have not
undergone extensive toxicological testing.
Nevertheless, EPA program  and regional
offices need to make a variety of decisions
about these chemicals. These decisions may
include prioritization of the chemical(s) for
Genomics, combined with modern computing
and information technologies, will advance
predictive toxicology and improve the
efficiency and reliability of prioritization and
risk assessments within the Agency.
further evaluation or a decision that no further    ^^^^^^^^^^^^^^^^^^^B^^B
research is needed. A variety of approaches
has been developed to assist in prioritization decisions. In the current approach, chemical
prioritization may be determined by several factors including production volume, exposure
information, persistence, chemical class, analysis of structural analogues, and consideration of
more formal structure-activity relationships (SARs). However, all these attributes have
limitations, and a better knowledge-based approach is needed.

       As an example, microorganisms have the potential to be spread through drinking water
supplies and distribution systems. The Office of Water's 1998 final Contaminant Candidate List
(CCL) comprises 60 contaminants and contaminant classes, including 10 mierobial contaminants
and groups of related microorganisms. Computer model results or expert judgment are currently
used for CCL hazard estimation and prioritization activities.

       Thus, there is a large number of stressors that the Agency must prioritize for further
evaluation. Currently, there is no rapid, comprehensive method for prioritizing which chemicals
or microbes should be tested based on the potential for toxicity, and it is recognized that it is not
possible to test all stressors. Genomics technologies hold the promise of providing more
mechanistic, molecular-based data for risk-based prioritization of these stressors. In addition,
these technologies are likely to offer more efficient, potentially high throughput, and low cost
alternatives to the tests EPA currently relies on for prioritization.  However, there is currently
little scientific consensus concerning which tests would be most appropriate for the Agency's
different prioritization needs. Which model system(s) would one employ to test chemicals and
what endpoint(s) would one look for are important questions to ask when using genomics
information for prioritization.

-------
                           Genomics Task Force White Paper                   Page 11

2. Regulatory and Voluntary Activities Potentially Affected by Genomics Information

       a. Representative Activities

Office of Pollution Prevention and Toxics (OPPT) - High Production Volume (HPV)
Challenge: Genomics data could be applied to the voluntary HPV screening process. For
example, specific gene expression data could be used to predict the relevance of an endpoint
evaluated in an animal model Screening Information Data Set (SIDS) test to an adverse health
response in humans. Further, these types of genomics data could, in the future, potentially
supplement or supplant animal testing needed to complete the SIDS data set. However, given the
timeframe of the HPV program, genomics data will unlikely supplant part of the SIDS dataset in
the near term. More likely for the near term is the use of genomics data to validate category
groupings in HPV and possibly future high  volume chemical screening processes.

Office of Water - Contaminant Candidate List: Genomics data generated for CCL chemicals
may be able to supplement computer model results or expert judgment in hazard estimation and
prioritization activities. For example, computerized analysis and the growing use of automated
polymerase chain reaction (PCR) techniques have allowed tremendous gains in the study of
microbial genomics, as well as of whole organisms.  A number of microorganism genomes have
already been studied, many of which are associated with waterborne disease. Genomics
databases may play a role in prioritizing pathogens based  on the availability of virulence genes of
concern and their corresponding gene products.

       b. Additional Activities

       Genomics may also have regulatory implications for prioritization in other program
offices as well as in regions, states, and tribes. Further details on the following activities are
found in Appendix B.

       Program Offices
       *     Office of Pollution Prevention and Toxics  (OPPT):  Premanufacture Notices
             (PMNs), Voluntary Children's  Chemical Evaluation Program (VCCEP),
             Endocrine Disrupters Screening Program (EDSP)
             Office of Pesticide Programs (OPP): Pesticides, fiierts, EDSP
            Office of Water (OW): Prioritizing stream or wetlands for study or cleanup
            Office of Air and Radiation (OAR): Hazardous air  pollutants (HAPs)
            Office of Solid Waste and Emergency Response (OSWER):  Superfund sites
            Office of Research and Development (ORD): Future research on chemicals

       Regions, States, and Tribes
            Site remediations and chemical evaluations

-------
Monitoring and compliance applications may
be some of the quickest Agency uses of
genomics-based information.
Page 12                    Genomics Task Force White Paper

B.  Monitoring

1. Introduction

       The term "monitoring" in the present context refers to any activity by which
environmental samples are taken and used for
regulatory or prioritization decisions and for
developing environmental status and trends
information. Many programs have either site-
specific or media-specific data requests that
are used to make regulatory decisions,
monitor compliance, and/or to prioritize the
use of EPA's human and economic resources.  In addition, EPA is charged with determining the
state of the environment, as well as assessing the status and trends of ecological condition. In
many instances, entities other than EPA generate the data EPA uses (e.g., other federal agencies,
states, tribes, the regulated community, and interested stakeholders in volunteer monitoring
programs). These data are generated through various headquarters and regional programs
through contracts, grants, and cooperative agreements. In fact, a large portion of EPA regions'
budgets are directed to programs (e.g., in states and tribes) that generate information that could
fall into the category of monitoring data or information.

       EPA obtains, requests, and receives many types of environmental data for both
assessment and compliance purposes, including but not limited to the following:  chemical and
physical analyses of air, water, soil, and sediment; toxicity testing of various environmental
media or chemicals; plant, animal, and human  tissue residues of various chemicals or their
breakdown products; community structure analyses (e.g., fish and/or invertebrate IBIs [index of
biotic integrity], algal  and plant community structure, invasive species evaluations in terrestrial
and aquatic ecosystems); and microbial community and pathogenic microorganism analyses of
air, water, soil, and sediment.

        Many  of these types of environmental  data could be generated using genomics-based
techniques, and some  applications are already being tested. One example is in the area of
microbial source tracking to determine the sources of fecal contamination that may be causing
impairment of a water body resulting in a beach closure.  Several state agencies and public
utilities are evaluating molecular-biology-based and genomics-based techniques to determine
whether these  approaches can distinguish among fecal sources in order to develop TMDLs  for
impaired water bodies (Griffith et al., 2003). A second example is the area of site clean up in
which changes in microbial community response to a stressor such as an oil spill may be
characterized using these techniques. One genomic approach to evaluating changes in microbial
community is to use total DNA, representing all of the microbial community, rather than the one-
to-two percent of the microbes that can be cultured.  This genomic information could be used to
differentiate and evaluate the feasibility among remedial alternatives, such as active remediation

-------
                           Genomics Task Force White Paper                   Page 13

(i.e., adding nutrients or microbial cultures) versus monitored natural attenuation. The
Department of Energy is exploring how this type of mierobial community "fingerprinting" can be
used to distinguish the conditions that promote effective bioremediation of petroleum-
contaminated soils or sediments (http://www.sc.doe.gov/ober/ERSD/ersd nabir.html). A third
example is in the area of development of multi-gene arrays of model animals (e.g., "fish-on-a-
chip"). Future toxicity testing for compliance with discharge requirements could involve using
gene chips for species such as fathead minnow to determine whether a water sample resulted in a
toxic pattern of response in exposed fish.

      Thus, there is a wide range of monitoring information that the Agency considers and a
wide range of potential applications of genomics technologies. The cost and time required to
collect and analyze the large number of conventional environmental samples needed to make
sound regulatory decisions and to evaluate environmental status is enormous. Genomics
technologies may ultimately yield rapid, efficient, and cost-effective methods for environmental
monitoring.

2. Regulatory Activities Potentially Affected by Genomics Information

      a.  Representative Activities

Office of Water (OW)/Office of Ground Water and Drinking Water (OGWDW): OGWDW
anticipates that monitoring for chemicals or microbial pathogens will use genomics-based data in
five to ten years for the following purposes:

           Compliance monitoring by federal, state, and tribal agencies to determine if
             surface waters  meet designated uses standards under CWA (TMDLs, Criteria,
             Standards)

           Monitoring finished or source drinking waters for contaminants

           Real-time monitoring of ambient surface water, e.g., for beach or shellfish bed
             closures (Beaches Environmental Assessment and Coastal Health [BEACH] Act,
             Standards)

           Monitoring classes of compounds based on biological  activity or mode of action
             (e.g., cholinesterase inhibition)

           Developing occurrence data as a basis for Safe Drinking Water Act (SDWA)
             regulation or listing on CCL

           Developing future drinking water regulations (6 Year Review)

-------
Page14                    Genomics Task Force White Paper

Regions, States, and Tribes - State NPDES permits:  The State of California, as part of an
ongoing ambient water quality monitoring program, is initiating an effort to evaluate surface
waters for the presence of estrogenic endocrine effects, using a RT-PCR assay for vitellogenin
gene expression in livers of exposed male rainbow trout. If results show that some surface
waters exhibit estrogenic effects, the Regional Water Quality Controls Boards in California,
which issue NPDES permits as well as perform ambient water quality monitoring, may consider
including this type of bioassay into the monitoring program for  waste water treatment facilities
even though it is not yet an "EPA method." Currently, NPDES  permits contain only chemical
and toxicity-based limits, require chemical analyses and toxicity testing of effluents to
demonstrate compliance with permit limits, and relate waste loads to watershed TMDLs.
Genomics technologies could provide new and more sensitive monitoring tools to develop
discharge limits for NPDES  permits.

       b. Additional Activities

       Genomics may also have regulatory implications for monitoring in other program offices,
as well as in regions, states, and tribes. Further details on the following activities are found in
Appendix B.

       Program Offices
             OPP:  Pesticide monitoring for registrations and reregistrations
       *      OAR: Stationary source monitoring
             OSWER:  Superfund and Resource Conservation and Recovery Act (RCRA)-
             required monitoring
             Office of Environmental Information (OEI) and  ORD:  Biomarker development

       Regions, States, and Tribes
             State and local beach closure and TMDL issues associated with pathogens
             State and local air quality monitoring
             Tribal issues  (e.g., monitoring for endocrine disrupters)
             Regional pesticide program inspections

C. Reporting Provisions

1. Introduction

       a. Reporting on Adverse Effects of Commercialized  Chemicals and Pesticides

       Reporting of certain  adverse effects/risks for industrial chemicals and pesticides already
on the market is mandated under both the Toxic Substances Control Act (TSCA) and the Federal
Insecticide, Fungicide, and Rodentitide Act (FIFRA). TSCA section 8(e) requires that "[a]ny
person who manufactures, processes, or distributes in commerce a chemical substance or mixture

-------
                           Genomics Task Force White Paper                    Page 15

and who obtains information which reasonably supports the conclusion that such substance or
mixture presents a substantial risk of injury to health or the environment shall immediately
inform [EPA] of such information" (15 U.S.C. 2607(e)).  FIFRA section 6(a)(2) states "If at any
time after the registration of a pesticide the registrant has additional factual information regarding
unreasonable adverse effects on the environment of the pesticide, the registrant shall submit such
information to the Administrator." There is a need to interpret how these TSCA and FIFRA
provisions apply to genomics data.  There are already certain types of conventional tests whose
data are not considered to present indication of substantial risk to health or the environment and
are not required by the Agency as stand-alone submissions. As the predictability and validity of
genomics methods increase, EPA may need to re-evaluate its stance on these reporting
provisions. Because these provisions address the reporting of adverse effects, the issue of what
genomic changes mean in terms of adversity must be addressed before reporting for genomic
responses may be required.  This issue may be best approached in a multi-stakeholder process to
ensure scientific consensus around the understanding of adverse effects based on genomics data.

       b. Toxics Release Inventory Program

       The Toxics Release Inventory (TRI) database was established under the Emergency
Planning and Community Right-to-Know (EPCRA) Act of 1986. Section 313 of EPCRA
requires certain  industrial facilities to annually report information on toxic chemical releases and
other waste management activities to EPA and the states  to inform communities of chemical
hazards in their area.

       The statutory chemical listing/delisting criteria  of EPCRA section 313 (d)(2) are
primarily based on hazard, not on risk.  The emphasis of EPA's hazard assessment is on a
chemical's inherent toxicity rather than the potential risks from exposure to the chemical.  A
chemical may be added to the TRI list if (a) the chemical is known to cause, or can reasonably be
anticipated to cause, significant adverse acute human health effects at concentration levels that
are reasonably likely to exist beyond facility site boundaries as a  result of continuous, or
frequently recurring, releases; (b) the chemical is known to cause or can reasonably be
anticipated to cause, in humans, cancer or teratogenic effects or serious or irreversible
reproductive dysfunctions, neurological disorders, heritable genetic mutations, or other chronic
health effects; or (c) if the chemical is known to cause, or can reasonably be anticipated to cause,
because of its toxicity, its toxicity and persistence in the environment, or its toxicity and tendency
to bioaccumulate in the environment, a significant adverse effect on the environment of sufficient
seriousness, in the judgement of the Administrator, to warrant reporting.

2. Regulatory Activities Potentially Affected by Genomics Information

       In order  for genomics technologies to have an effect on reporting provisions, the issue of
linking genomics changes to adverse effects or response pathways needs to be addressed.  Once
genomie changes are linked to adverse effects, the Agency will need to make decisions regarding
whether genomic changes apply to reporting provisions.

-------
Page 16                    Genomics Task Force White Paper

       Genomics technologies could affect reporting requirements under TSCA 8(a) and MFRA
6(a)(2) if genomic changes detected are linked with substantial risks or adverse effects. If
genomics data do, in the future, become a reporting requirement, this could also affect the
number of reports received under these reporting provisions and the resources required to
evaluate the reports.

       If there is a linkage to adverse effects in humans or on the environment, genomics data
may be considered in the hazard assessment when determining whether or not a chemical meets
the TRI chemical listing/delisting criteria. The Interim Policy on Genomics allows such
information to be used in the overall assessment on a case-by-case basis, but genomics
information alone currently cannot be used to determine hazard at this time. Practical  application
of genomics-derived information will improve the quality of hazard assessments conducted by
EPA, including those conducted by EPA's TRI Program.

-------
Genomics information will not fundamentally
alter risk assessment practices, but genomics
will provide powerful new tools and insights to
use for risk assessment.
                           Genomics Task Force White Paper                    Page 17

III.  Risk Assessment

A. Introduction

       Genomics technologies present an opportunity to greatly enhance Agency risk
assessments. Specifically, genomics          	
technologies are likely to contribute
significantly to improvements in defining a
chemical's mode of action, evaluating effects
on susceptible populations and life stages, and
assessing exposure to and effects of chemical
mixtures, as outlined below. For example, in
collaboration with the International Life Sciences Institute (ILSI), EPA has developed a
framework for microbial risk assessment (MRA), and OW is working to expand the framework
into a full MRA protocol which may include consideration of genomics data.  OW is also
working with EPA's Risk Assessment Forum to develop MRA guidelines. It is important that
any change, refinement, or addition to a risk assessment practice be vetted by interested
stakeholders and ultimately peer reviewed to ensure scientific consensus around the practice.

B. Mode of Action (MOA)

1. Overview

       The term "mode of action" (MOA) is defined in the Draft Guidelines for Carcinogen Risk
Assessment (USEPA, 2003a) as  a sequence of key events and processes, starting with the
interaction of an agent with a cell, proceeding through operational and anatomical changes, and
resulting in an adverse outcome. A "key event" is an empirically observable precursor step that
is in itself a necessary element of the MOA or is a biomarker for such an element. Genomics
technologies can be used to better understand the MOA of a chemical agent and, thus, can lead to
advances in human and ecological risk assessments of chemicals. As genomics information
contributes to our understanding of MOAs, the validity of using this information as an indicator
of both adverse  effects and exposure is enhanced.

       Genomics data may allow the development of gene, protein, or metabolite profiles that
can advance the screening of individual chemicals and allow faster and more accurate
categorization into defined classes according to their MOA.  There are many examples of
possible modes  of carcinogenic action, such as mutagenicity, mitogenesis, inhibition of cell
death, cytotoxicity with regenerative cell proliferation, and immune suppression.  MOAs have
been identified for other adverse outcomes, both human health and ecological  (e.g., cytotoxicity,
endocrine disruption, loss of homeostasis). Such approaches will significantly enhance the
Agency's ability to harmonize risk assessment approaches for different outcomes for which the
development of a list of common MOAs is essential.

-------
Page 18                    Genomics Task Force White Paper

       Understanding the MOA of environmental agents that induce toxic effects other than
cancer or that induce carcinogenicity in animal models should facilitate the assessment of the
relevance of these findings in protecting human health  and safeguarding the environment. An
important issue for extrapolation of responses in animal models to humans or environmental
endpoints is to establish whether the MOA in the test species is relevant in the target species.

2. Risk Assessment Activities Potentially Affected by Genomics Information

       Many program and regional offices need to make judgments about chemicals with little or
no data available. Others have extensive datasets, but still struggle with accuracy and precision
as well as extrapolations to nontested species or scenarios. Genomics approaches are envisioned
to provide improvements in (a) hazard identification, (b) dose-response assessment, (c)
extrapolations, and (d) exposure assessment.

       a. Representative Activities

       i. Improving Hazard Identification

Evaluating chemicals for genotoxic or other MOAs. New approaches using tissue microarrays
can enhance throughput and the linking of genomic and cellular outcomes. Further, combining
the findings of gene expression studies with data from chemical exposures of genetically altered
animal models (e.g., knockout or null mice) is a powerful tool to link specific genes to specific
detrimental outcomes.  Such data will allow the development of gene profile "fingerprints" of
genomic characteristics for specific MOAs. The development of genomic "fingerprints" will
provide a rapid screening method to categorize chemicals with unknown MOAs for both human
health and ecological assessments.

Predicting or Defining Metabolic Pathways. The chemical evaluation process includes
consideration of the parent compound and its potentially active metabolites. Genomics
approaches, particularly at the proteome level, will aid in the characterization of metabolic
pathways and the identification of  lexicologically active metabolites. Computational toxicology
approaches will further enhance the prediction of metabolic pathways and metabolites for
chemicals that have not been investigated experimentally and potentially will reduce the use of
test animals and the cost of data generated to support risk assessments. Metabolic pathways and
the genes associated with those pathways need to be linked to adverse effects of concern.

Replacement of standard toxicity tests in regulatory batteries with rapid, pathway-specific
response tests.  It has been envisioned that relevant genes and gene products for specific
toxicities such as genotoxicity can be formatted on arrays to provide a more comprehensive
analysis than currently available assays (Aardema and MacGregor, 2002). Because many of the
lexicological testing procedures and  strategies required by EPA have remained largely
unchanged for 20 years, it is reasonable to assume that many of the current assay systems used
may be replaced by more sensitive, rapid, and predictive genomic assays able to identify specific

-------
                           Genomics Task Force White Paper                    Page 19

pathways of response. Acceptance of these genomic protocols for both human health and
ecological assessfnents will lead to time and cost savings and may also lead to more accurate risk
assessments.

       ii.  Improving Dose-Response Assessment

Linear versus nonlinear. The Agency's traditional approach to cancer risk assessment for
agents that are known mutagens and carcinogens employs a linear,  low dose extrapolation to
quantify possible human cancer risks. The underlying premise for this linear default is that
electrophilic compounds are presumed to form single DNA modifications in single cells that
could potentially lead to cancer. Due largely to the successful use of genetic toxicological testing
schemes for screening, however, the number of new genotoxic carcinogens entering the
environment is likely to be small. The recent Final Draft Guidelines for Carcinogen Risk
Assessment addresses the issue of nongenotoxic carcinogens and encourages the use of
mechanistic data to identify whether a nonlinear extrapolation is appropriate for nongenotoxic
carcinogens.  For this purpose, biomarkers of response for genotoxic carcinogens are available at
least at the cellular response level if the general cancer MOA is known. The challenge will be  to
develop biomarkers of response that can be used for predicting specific outcomes for
nongenotoxic chemicals (Bartosiewicz et al., 2001 a).  This goal can be realized through gene
expression pattern recognition that parallels histological changes in tissues and the eventual
progress to tumor formation. An example would be the CUT (Chemical Industry Institute of
Technology) Centers for Health Research's efforts to  identify genes associated with peroxisome
proliferators,  such as the PPAR-cc (peroxisome proliferator-activated receptor alpha) that are
linked to alterations in mouse hepatocellular growth following peroxisome proliferator exposure.

       The regulatory impact of genomics on possible nonlinear extrapolation for nongenotoxic
carcinogens is significant.  Within the Agency, nongenotoxic carcinogens without plausible
MOA data are currently subjected to the same linear low dose extrapolation applied to genotoxic
carcinogens.  It is a reasonable assumption that a collaboration among industry and EPA
scientists will occur in the area of MOA-based cancer risk assessment to ascertain if a nonlinear
low dose extrapolation is appropriate for nongenotoxic carcinogens. The same MOA approach
can be used to help more clearly discern dose-response relationships for chemicals that affect
other health endpoints as well.

Lowering of Points of Departure (PODs) based on genomic responses. Genomics technologies
have the potential to affect dose-response analyses for nonlinear assessments of adverse
toxicological  outcomes.  In traditional toxicology, doses used to determine adverse effects are
generally high to ensure that tissue level or whole animal toxic responses are demonstrated. This
permits the selection of toxicity endpoints and establishment of doses at which no adverse effects
are seen (NOAEL) and the lowest doses at which an adverse effect is seen (LOAEL). Thus, most
toxic substances currently are regulated on frank toxicity rather than on a molecular level
response, and the association between a molecular level change and an adverse outcome has only
rarely been established. A few  substances are regulated based on biochemical changes with

-------
Page 20                    Genomics Task Force White Paper

known relationships to adverse outcomes such as cholinesterase-inhibiting pesticides and lead.
Organophosphate pesticides are regulated at NOAELs which are generally much lower than
many chemical classes because the endpoint, cholinesterase inhibition, is determined
biochemically, and inhibition can generally be detected well below the levels showing overt
toxicity.  In contrast, many fungicides or herbicides have relatively high NOAELs because clear
pathological alterations only occur in animals at high doses. Regardless of the chemical class or
use, however, with the advent of molecular technologies including genomics, chemically-induced
changes in gene expression are likely to result in the identification of simple, sensitive, and
relevant biomarkers of effect that can be used in dose-response studies to more readily identify
effects in the low dose range (i.e., below doses causing frank pathology) for humans and wildlife
species.

       If EPA chooses to establish regulatory limits (e.g., NOAELs) based on changes in gene
expression (e.g., identifying upstream precursor effects or markers of adverse effects), the POD
used to set the regulatory limit could be higher or lower in both the human health and ecological
arenas. EPA needs to examine whether a lower effect level based on a molecular effect is "safer"
than a level based on the currently used toxic  effect. How this will affect risk assessment
practices will need to be addressed via a multi-stakeholder process and peer review.

       ill. Improving Extrapolations

High to low dose extrapolations; Route-to-route extrapolations. Reduction of uncertainty is
one of the primary ways to improve the risk assessment process. Reduction in uncertainty in
dose-response assessments can be enhanced by the use of predictive models  such as
Physiologically-Based Pharamacokinetic (PBPK) and Biologically-Based Dose Response
(BBDR) models. These models can provide better methods for calculating dose metrics (e.g.,
target tissue doses) that are more flexible and relevant for extrapolation across exposure routes,
between  species, and from high to low doses. The potential of molecular indicators to define the
shape of the dose-response relationship at low exposures suggests the possibility that alteration or
elimination of some uncertainty factors (UFs) may be justified; data-derived factors based on
genomies information may be determined and applied in the future. Similarly, molecular-based
pharmaeokinetic data that describes the distribution of biologically effective doses of active
ingredients to target organs via other portals of entry offers the possibility of reducing
uncertainties associated with route-to-route extrapolations.

Interspecies extrapolations.

       Relevance to humans. Further improvements in human health assessments can be
       realized through the use of genomics data that support an evaluation of whether or not
       MOAs determined in test animals are  similar and feasible in humans (i.e.,  whether the
       target genes are conserved and operative across species). Genomics data that show little
       or no similarity in key genes or patterns of gene expression between humans and rodents
       would indicate interspecies differences and support a possible conclusion of non-

-------
                           Genomics Task Force White Paper                    Page 21

       relevance to humans. Conversely, data showing good agreement in key genes or
       expression patterns between humans and rodents would provide higher confidence in the
       relevance of the findings to human health. Similarly, interspecies comparison of
       pharmacodynamic responses enhanced by the use of genomics data could be used to
       define toxicological pathways in a quantitative sense. This information could be
       compared across species by the choice of appropriate molecular markers. Gene
       expression profiling is one approach that looks promising for linking cellular responses to
       a specific environmental chemical or mixture in laboratory animals to responses in
       humans or human cells in vitro,

              Key biological systems have fundamental genomics processes, some of which, if
       altered, are universally deleterious. Demonstration of a common interspecies genomic
       response linked to an adverse effect and evaluation of the dose-response relationships in
       the lower animal (e.g., invertebrates) and humans could permit the extrapolation of
       genomic responses in lower order animals to adverse effects in humans. The increasing
       development of genomic information in lower organisms may provide a means to
       evaluate potential effects in humans that will extend the use of lower organisms beyond
       current mutagenicity testing.

       Relevance to wildlife species of concern. In ecological risk assessment, it is necessary
       to extrapolate results from a very limited set of test species to a wide range (potentially
       hundreds to thousands) of species present in the environment.  The development of
       reliable methods for extrapolating toxicity information from test species to those that are
       of concern but cannot be directly tested is necessary.  As in human health assessments, an
       important issue is determining  whether the MOA in the test species is feasible for other
       species present in the ecosystem. Use of genomics tools for the development of
       quantifiable pharmacodynamic models and applicable molecular markers will also
       significantly enhance species-species extrapolations and reduce the current reliance on the
       application of uncertainty factors.

       iv. Improving Exposure Assessment

       Genomics technologies are likely to lead to the development of simple, sensitive, and
informative biomarkers of exposure that can be used in exposure assessments, particularly in the
evaluation of potential occupational exposures for human health assessments and for
environmental exposures for both human health and ecological risk assessments. Current
methods rely on residue analyses or modeled scenarios  and a few well-documented biomarkers of
exposure (e.g., CYP1A, cholinesterase, delta-aminolevulinie acid dehydratase, metallothionein).
Molecular techniques such as the use of microarrays or RT-PCR are already providing tools for
documenting actual exposures to humans and ecological species of concern using identified
biomarkers for which there is a  good understanding of the relationship of the level of biomarker
to the level of exposure. In the near future, genomics may aid in the identification of new
biomarkers that can identify exposure to more stressors or more MOAs. and potentially enhance

-------
Page 22                    Genomics Task Force White Paper

the quantitation of these. When pharmacodynamic and MOA studies become sufficiently robust
to relate exposure endpoints to whole organism adverse effects, risk assessment predictions will
become significantly more accurate.

C. Susceptible Populations and Sensitive Life Stages

1. Overview

       a.  Susceptible Human Populations and Sensitive Life Stages

       Genomics and related technologies offer a tremendous opportunity to define and identify
people with enhanced susceptibility to many environmental contaminants. The human genome
consists of 30,000 or  so genes that build
cellular structures, control the cell cycle,
execute metabolic functions, and mediate the
information flow within and between cells.
Small differences in gene sequence, known as
single nucleotide polymorphisms (SNPs), can    contaminants.
Genomics technologies offer a tremendous
opportunity to define and identify people with
enhanced susceptibility to many environmental
have a dramatic or inconsequential effect on      ^^^^^^^^^^^^^^^^^^^^^M^M
protein function and activity depending on the
particular polymorphism. Genomics technologies have the potential to yield information about
the distribution of SNPs within the human population and their potential effects on genes that are
responsive to various environmental contaminants. The interaction of genetic variants with
environmental conditions can affect individual susceptibility to a variety of diseases such as
cancer, diabetes, and heart disease and can promote sensitivity to disease from exposure (Bishop
at al, 2001).

       Delineation of the frequency of occurrence of these polymorphisms within racial or ethnic
groups may raise ethical, legal, and social implications in the area of environmental justice
(Marchant, 2002). For example, genomics technologies might result in further categorization of
individuals, and consideration must be given to how these newly identified at-risk groups will be
included in environmental policy. Ethical practices by which genomics studies are conducted on
human test subjects in specific ethnic or racial communities will also need to be carefully
considered. The Agency will need to be proactive to ensure that these issues are handled in a just
manner.

       For a susceptible human population, an increased risk of illness could result from
exposure to environmental chemicals or microbial pathogens at any age. In contrast, an exposure
during a susceptible life stage could result in higher risk during a specific portion of an
individual's lifetime or could influence an outcome at another life stage and the potential adverse
effect incurred may be irreversible. For example, an exposure during early childhood
development might yield a specific form of cancer that would not have been induced if the

-------
                            Genomics Task Force White Paper                     Page 23

exposure occurred later in life. Similarly, an exposure during a sensitive life stage could result in
an increased probability of disease later in life.  The primary difference between susceptible
populations and susceptible life stages is that a susceptible population's toxic exposure generally
will yield an adverse outcome regardless of the age at which the exposure occurred.
Nevertheless, many of the ethical, legal, and social implications (ELSI) that apply to susceptible
populations will also apply to susceptible life stages. As a consequence, the Agency will need to
develop a policy regarding the collection and use of human genomics information from
individuals to provide safeguards regarding privacy.

       The National Institute of Environmental Health Sciences (NIEHS) has funded and
supported research to identify gene variations that affect susceptibility to environmental agents.
Approximately 500 genes were identified as part of the Environmental Genome Project of 1997.
These genes affect metabolism, DNA repair, cell cycle control, receptors, and immune function.
Although scientific progress has been made in understanding genetic variations and susceptibility
to toxic chemicals and pathogens, research efforts have yielded inconsistent results. In spite of
this, researchers continue to characterize genetic variations of susceptibility and provide insights
about individuals  who are more or less susceptible to disease from exposure to toxic substances
(Marchant, 2002).

       b. Susceptible Wildlife Populations and Sensitive Life Stages

       The term "susceptible population" is most frequently used in reference to at-risk human
populations; however, the term can also be used to describe wildlife populations that may be at
risk due to exposure to environmental
contaminants.  Certain taxonomic subgroups
of plants and animals may be more
susceptible than others although this varies
with contaminant  and life stage. Currently,
EPA examines species sensitivity through
Genomics technologies offer a powerful tool to
examine toxicological responses across species
for prediction of sensitivities or tolerances in
untested organisms.
toxicity testing of environmental
contaminants on representatives from various
classes of organisms. Genomics technologies offer a powerful tool with which to examine
toxicological responses across species for prediction of sensitivities or tolerances in untested
organisms. In addition, genomics technologies will allow for the examination of long-term
ecosystem health and the potential irreversibility of toxic effects. Examining the genetic
diversity of organisms inhabiting a given ecosystem would allow risk assessors to determine if
exposure to environmental contaminants might cause an evolutionary change in ecosystem
structure or function.

       Threatened and endangered species represent a subset of all species faced with the
possibility of extinction and are afforded extra protection under the Endangered Species Act.  A
species, however, may be sensitive to exposure to an environmental contaminant without being
endangered. Comparisons of the sensitivities of endangered and common fish test species to

-------
Page 24                    Genomics Task Force White Paper

toxicants have been made. This research indicated that the sensitivities of the endangered species
tested were generally not greatly different from that of the more common species (Sappington, et
al., 2001). However, because the population size of endangered species is, by definition, quite
small, reduced genetic variation may result in reduced tolerance to multiple stressors such as
combinations of contaminants and climate stress (Porter et al., 1984).  Alternatively, continuous
exposure of a widespread, susceptible species to an environmental contaminant might result in
the species becoming endangered.

       While human behaviors such as over-hunting or habitat destruction are the main stressors
that threaten species survival, contamination of an ecosystem with a pesticide, industrial
chemical, or pathogen can also harm ecosystem health. Genomics technologies will provide
significant insight into the biochemical mechanisms by which environmental contaminants  might
adversely affect certain species and may provide insights into which species are more likely to be
headed down the path towards extinction.  Cross-species extrapolations of sensitivity coupled
with the ability to measure contaminant-induced reduction in genetic diversity within specific
populations will provide valuable input to population viability models.

2. Risk Assessment Activities Potentially Affected by Genomics Information

       a. Susceptible Human Populations and Sensitive Life Stages

       EPA anticipates genomics research will be used to assess hazards and risks of chemicals
and pathogens to specific human populations.  Currently, the Agency does not generally take into
account genetic factors when assessing the risks posed by chemical or biological substances
although life stage and, to some extent, gender are considered. Additionally, the Agency rarely
considers the genetic predisposition of a specific individual, race, or ethnic background when
determining the toxic effects of chemicals.  The Food Quality Protection Act (FQPA) of 1996
does, however, direct the Agency to determine the potential of increased susceptibility of infants
and children from exposure to toxic substances such as pesticides.

       i. Representative Activities

OPPTS, OW, OAR, ORD, OSWER. In any situation where a "safe" exposure level is
designated (FIFRA, Clean Water Act, Safe Drinking Water Act, Clean Air Act, CERCLA), the
identification of a susceptible population or life stage has the potential to force a change in  the
health standards employed. For example, the health standards used during hazardous waste
remediation projects are often first reviewed and published in the Integrated Risk Information
System (IRIS) database maintained by ORD.  One of the goals of the IRIS database is to provide
oral Reference Doses (RfDs) and inhalation Reference Concentrations (RfCs).  Both the RfD and
RfC are based on the assumption that nonlinear dose-response curves exist for certain toxic
effects such as cellular necrosis (USEPA, 2002b). In general, these values provide an estimate of
a daily exposure to the human population (including sensitive subgroups) that is likely to be
without an appreciable risk of deleterious effects during a lifetime (USEPA,  1999). Genomics

-------
	Genomics Task Force White Paper	Page 25

technologies will provide a powerful tool for the identification of sensitive subgroups and allow
specific decisions to be made that address specific sensitivities. As the use of genomics
technologies becomes widespread, the number of susceptible populations identified will likely
grow in number. Consequently, the Agency will need to be vigilant when revisiting health
standards in all media to identify and protect susceptible populations.

       The impact of genomics technologies on EPA's understanding of life stage sensitivity
also will become an important issue. For example, different human age groups may express
varying levels of some metabolic enzymes (Hakkola et al., 1998). Enzyme over- or under-
expression could play a part in determining the severity of a toxic exposure. Additionally, the
rapid growth taking place early in life is largely dependent on gene-environment interactions,
Perturbation of this interaction through toxic exposure has the potential to significantly influence
development.  Genomics will provide tools to identify life stages that need separate assessments
based on their unique susceptibilities,

OPPTS. The extent  of enzyme activation is partially responsible for determining the severity of
response to a chemical exposure. In addition, individuals may respond to metabolic stimuli to
varying degrees depending on their genetic composition. Ethanol exposure, for example, is
known to induce the metabolic  enzyme CYP2E1; however, the amount and activity of CYP2E1
produced may vary among individuals (Haber et al., 2002; Snawder and Lipscomb, 2000). If
genomics technologies are successful in identifying populations susceptible to specific pesticides
or industrial chemicals, product labeling will probably be necessary. For example, labels might
include warnings for  particular populations known to exhibit higher frequencies of an at-risk
genetic polymorphism. The pharmaceutical industry already includes warnings to susceptible
populations on drug labels. The Agency has the ability to follow similar practices for pesticides
because Section 3 of  FIFRA provides EPA the authority to regulate labels (USEPA, 2003c).
OSWER, Regions, States, and Tribes. Genomics data might be used in the future to help
identify susceptible populations or life stages when assessing risks at hazardous waste sites or for
remediation of contaminated areas such as Superfund sites or other scenarios in which relatively
small, identifiable populations are exposed, Genomics data may provide information on
potential exposure patterns and also might be useful in developing site-specific remediation
goals. If, for example, a genomics study was to identify a susceptible population at risk due to
exposure to a contaminant at a Superfund site through a correlation of genomic analysis of local
populations and measured or expected exposure levels, the Agency might choose to reduce the
RfD/RfC value and propose more strict remediation measures. This, of course, presupposes an
established linkage of the genomic endpoint and an adverse effect. Use of new genomics tools
could, however, limit the extent of remediation measures by more accurately predicting the
potential for exposure of the sensitive population.  Thus, genomics tools may play a key role in
determining intensity and extent of clean up practices and have large implications for time and
cost of such procedures.

-------
Page 26                     Genomics Task Force White Paper

       b. Susceptible Wildlife Populations and Sensitive Life Stages

       EPA currently does not use genomics technologies in ecological assessments or criteria
development.  However, numerous potential applications for the identification of sensitive
species, populations, or life stages may become available in the near term. Longer-term
applications are also under development.

       i. Representative Activities

OPPTS, OSWEE, OW, OAR, ORD. Due to time and resource limitations, as well as ethical
considerations, all species cannot be tested for responses to contaminants either for site-specific
mitigation needs, for product registration, nor for criteria development.  Therefore, the
development of reliable methods for extrapolating toxicity information from tested species to
those that are of concern but cannot be directly tested is necessary. This need is particularly
acute for chemicals that may target sensitive life stages (e.g., metamorphosis in amphibians) or
vulnerable species (those with small population sizes,or that may have greater sensitivities to
particular chemicals). Genomics technologies will provide the potential for extrapolating
between test species and sensitive  wildlife species or life stages in a rapid, cost-effective manner.

       Because of the smaller population sizes of endangered species, it may be possible to
identify biomarkers  of effect can be evaluated non-invasively using genomics techniques (e.g.,
metabonomic analysis of urine or feces). Additionally, genomic techniques could potentially be
used to evaluate the differences between populations of species, to determine whether they are
unique species or are geographically separated populations of the same species. This could affect
the way potential impacts on threatened or endangered species  are managed.

       The most immediate application is likely to be with aquatic organisms and  the application
to quantitative structure activity relationships (QSARs) or computational toxicology. This will
be particularly useful in the Premanufacture Notice (PMN) process of OPPTS, but may find
applications in other offices as well (e.g., Office of Water). The next application likely will be in
the development of methods for screening chemicals that are potential endocrine disrupters in
aquatic ecosystems.

       While the protection of individual plants and animals from clinical disease caused by
xenobiotic compounds and pathogens is an achievable goal, how chemicals combine with other
environmental stressors to change the genetic properties of a population over time  is less clear. If
genetic diversity is reduced or if particular genes are suppressed or expressed at abnormal rates, it
is possible that (1) the population may become less fit over time, (2) response to additional
stressors may not be adequate; and (3) the reduced diversity may lead to a bottleneck and/or
eventual extinction. Because genetic diversity is the fundamental basis for adaptation and
evolution, it is increasingly being recognized as an important endpoint for risk assessment.  To
date, the methodology has not been available to address this  issue, but newly emerging genomics
techniques will allow such assessments in the future.

-------
	Genomics Task Force White Paper                    Page 27

OSWER, Regions, States, and Tribes.  The current Agency practice in ecological risk
assessment and clean up of contaminated sites (Superfund, Brownfields) is to focus on the most
sensitive species when determining the effects of a chemical contaminant on an ecosystem.
Genomics will allow risk assessors a greater ability to focus the ecological risk assessment on the
mechanistic level.  Although the Agency does not usually take into account genetic factors when
assessing the ecological risks posed by chemical or biological substances, genomics data gained
from the examination of a single species  might prove useful in preventing harm to other species
with similar genetic characteristics.

D. Mixtures

1. Overview

       Human health and ecological risk assessments of toxic substances often are incomplete
because most toxicological screening is performed for single chemicals. However, human and
wildlife exposures to chemicals are rarely limited to a single chemical, but instead are usually to
complex mixtures of chemicals during a        	
lifetime. Environmental exposures from
point and nonpoint sources, irrespective of
medium, generally occur as simultaneous or
sequential exposures to multiple chemicals.
In addition to environmental exposures, the
majority of the human population engages in
intentional exposure to a variety of
pharmacologically active chemical
compounds such as those in recreational  drugs (alcohol and tobacco), medicinal products, and
foods and is inadvertently exposed to other chemicals, such as those in vehicle exhaust, drinking
water, indoor air, and workplace environments.

       Chemical mixtures in the environment are,  in general, a complex group of active and inert
parent compounds, transformation products, and/or residues of which composition is
qualitatively and quantitatively not fully known (Feron and Groten, 2002). Because mixtures
change with time and distance from the original release site due to the differential fate and
transport of their components (Pohl et al., 1997), regulations established on toxicological data for
the original mixture may have little bearing on the  actual exposures resulting from a release.  It is
estimated that approximately 275 million tons of hazardous waste are produced annually in the
United States, and more than 2000 distinctive toxicants in site-specific media have been
identified by EPA, with hazardous substances in given mixtures numbering in hundreds (Suk et
al., 2002).

       Specific environmental chemicals have been demonstrated to adversely affect the health
of humans and wildlife.  These concerns  are amplified by the awareness that exposure to
Genomics technologies may aid in the
identification of unique patterns of gene
expression in the tissues of aquatic and
terrestrial organisms, and human cell-based
models, induced by exposure to multiple
environmental stressors.

-------
Page 28                    Genomics Task Force White Paper

chemicals often occurs in mixtures of chemicals that might exhibit complex interactions. The
various types of toxicological interaction associated with complex chemical mixtures can be
sorted into three reference categories: greater-than-additive (synergism, potentiation), additive
(the sum is equal to the parts, no apparent interaction), and less-than-additive (antagonism,
inhibition, or masking).  Of particular importance is whether a mixture of components, each of
which is present at concentrations below the level of concern, may be hazardous due to
additivity, specific interactions, or both (Hertzberg and MacDonell, 2002). Synergistic toxicity
resulting from co-exposure to pesticides has been observed, and greater than expected toxicity
has been noted for pesticide mixtures of certain cholinesterase inhibiting insecticides and some
fungicides.

       The Agency has been directed by FQPA to consider the combined effects on human
health that can result from exposure to toxic substances that share a common mechanism of
toxicity (e.g., organophosphates). This cumulative risk assessment approach is based on an
evaluation of the potential for people to be exposed to more than one member of a group of
chemicals at a time and considers exposures from food, drinking water, and residential sources.
It is important to note that for any group of toxic  substances with a common mechanism of
action, agents within that group that have low toxicity but the potential for high exposure can
present a risk similar to a toxic substance with higher toxicity and lower exposure potential. It
seems likely that the hazard,  dose-response, and exposure assessment components of the
cumulative risk assessment process will be greatly improved by the elucidation of mechanisms or
modes of action made possible by genomics data. While genomics information will be useful in
demonstrating whether two or more compounds share a MOA, there may be a need to establish a
logical framework for applying genomics information to the interpretation of the effects of
mixtures.

       The Agency for Toxic Substances and Disease Registry (ATSDR) of the Department for
Health and Human Services and EPA, in addition to other Federal research and regulatory
agencies, support in vitro and in vivo research to  further understand the chemical
characterization, molecular mechanisms of action, and toxicity of chemical mixtures and their
relationship with human health effects and other  biological systems. Genomics can provide the
tools for accomplishing this work that has to date been very expensive and required a large
commitment to animal testing.

2. Risk Assessment Activities Potentially Affected by Genomics Information

       Genomics technologies will aid in the identification of unique patterns of gene expression
in aquatic and terrestrial organisms and human cell-based models induced by exposure to
multiple environmental stressors. Several studies already have described tissue-specific
transcriptional patterns and have begun to address the concept of "fingerprinting" for chemical
mixtures in laboratory animals (Bartosiewicz et al., 2001a, 2001b) and in specific cell lines
(Mumtaz et al., 2002).

-------
                           Genomics Task Force White Paper                    Page 29

       a.  Representative Activities

OPPTS. There are several specific applications of genomics technologies that may improve risk
assessment of mixtures in the future.

            The toxicity of a test mixture could be evaluated through effects on genomic
             biomarkers that have been linked to adverse effects.

            Genomics technologies could be applied to the evaluation of constructed test
             mixtures to examine how chemicals may interact (additivity, synergism,
             antagonism).

            Genomic diagnostic indicators may help identify components of chemical
             mixtures with unidentified constituents through the use of a genomic "fingerprint"
             database.  For example, genomics technologies could be applied to evaluating the
             differences in gene expression between a mixture of known  chemical constituents
             and a test mixture containing unknown chemical constituents.  Differences in
             response to the two mixtures could  then be evaluated  and attributed to unknowns
             in the test mixture. This approach could be applied to exposure assessments.

       b.  Additional Activities

       Genomics may have additional regulatory implications for evaluating chemical mixtures
for other offices, as well as regions, states, and tribes. Further details on the following activities
are found in Appendix B.

       Program Offices
            OSWER, OAR, OW:  identifying unknowns in mixtures  at contaminated sites and
             in media, or during monitoring

       Regions, States, and Tribes
            Identifying unknowns in mixtures at contaminated sites during monitoring

-------
Page 30                    Genomics Task Force White Paper

IV.  Research Needs and Activities

A.  Introduction

       The Genomics Task Force identified a number of research needs that should be addressed
in order to fully utilize genomics data in decision making, including:

       a)      the link between molecular indicator, exposure, and adverse outcome needs to be
              established

       b)      the genomic dose-response curve needs to be delineated

       c)      inter- and intraspecies variations in response need to be quantified

       d)      detection limits and variability for the genomic indicator need to be established

       e)      the normal variability of gene expression, protein, and metabolite profiles needs
              be understood to evaluate any changes induced by stressors

       f)      baseline genomic responses of species to stressors other than chemical challenges
              need to be developed in establishing the utility of specific genomic indicators as
              markers of responses

       g)      the complexity of the toxicological data bases that are likely to be developed by
              EPA's ORD will require new computational approaches for their analysis

       It should be recognized though that there should not be a "higher standard" for use of
genomics data in decision making than for other more traditional types of data typically used.

       Within ORD, much research activity is directed toward the identification of gene
expression changes in cells and, to some extent, in tissues in response to environmental chemical
exposure.  To date, the majority of this research involves the assessment of changes at the
transcriptional level using mRNA microarrays rather than at the translational level using
proteomic approaches.  At this relatively early stage of the use of molecular profiling techniques,
the majority of the effort in Agency research is to establish reproducibility and consistency of
data for single molecular profiles for a time, concentration, species, cell type, or tissue set of
parameters.  The functional aspects of genomics (i.e., relating gene expression changes to cellular
perturbations) will be addressed in the next phase of the research program. The overall goal of
ORD's overarching Computational Toxicology Research Program is to use emerging
technologies to improve quantitative risk assessment. Readers are encouraged to review the
Framework (USEPA, 2003b) for additional details on ORD's Computational Toxicology
activities at: http://www.epa.gov/comptox/publications/comptoxframework06 02 04.pdf.

-------
                           Genomics Task Force White Paper                    Page 31

       The following section outlines genomics research needs and the current or planned
Agency genomics research activities. In addition, it delineates how the Agency plans to use the
data generated to address the types of regulatory and risk assessment activities described in
Sections II and III.

B.  Research Needs and Activities for Regulatory Applications

1. Prioritization Research Needs and Activities

       The overall aim of EPA in prioritization efforts is to establish which chemicals (or
chemical classes) and microbial contaminants warrant a more rigorous scientific investigation
leading up to a risk assessment. To this end, the gaps in the Agency's genomics initiative are
more in what is actually being conducted, rather than in what is planned. For example, to date
there has been a limited research effort with wildlife species because of the paucity of genome
sequence and genetics information for the majority of such species. Attempts are being made to
help rectify this, in part through a collaboration with the Department of Energy's Joint Genome
Institute to provide cDNA libraries of the fathead minnow and a frog species (e.g., Xenopus
tropicalis, the Western clawed frog). Similar efforts with other species will be needed so that a
set of sentinel species can be made available for environmental assessment.

       Research is needed to develop a functional approach to establishing a linkage between
genomic changes at the mRNA level and protein changes and cellular and tissue changes. This
type of approach can  provide information necessary for eventually developing a systems biology
approach for defining pathways to  disease or other adverse outcomes. These research needs will
require a strategic hiring process so that the necessary expertise is available within the Agency.
Collaborations will also be required to address the complex issues associated with a systems
biology approach. Analysis of the  large data sets generated through genomic assays will require
the development of bioinformatic and computational methods. In addition, enhanced QSAR
methods are required because the judicious use of QSAR approaches can greatly reduce the
reliance on experimental approaches for establishing chemical priorities for additional research.

       The initial steps in developing approaches for prioritizing chemicals for additional
research that could lead to the development of a risk assessment are described in the Agency's
Computational Toxicology initiative. The overall approach requires the development of
toxicological pathways for candidate chemicals such that the key events leading to specific
adverse outcomes can be identified. Chemicals can be  designated as requiring further research if
they are predicted to initiate the key events for a particular adverse outcome.

       The Agency's Computational Toxicology initiative stems from an FY02 Congressional
mandate to explore alternatives to the use of animals in toxicological studies. To address this
mandate, ORD is using endocrine disrupting chemicals (EDCs) as model compounds in research
that includes in silica, in vitro, and in vivo approaches as proof-of-concept for the overall

-------
Page 32                    Genomics Task Force White Paper

approach. The objectives of this effort are to determine the feasibility of using genomics and
computational toxicology to facilitate the prioritization of chemicals for screening. Another goal
is to reduce the need for some in vivo assays while providing a greater breadth of coverage of
endocrine alterations and a better predictiveness of potential adverse health outcomes.

       Similar approaches are being pursued for other chemical classes and other adverse
outcomes (e.g., disinfection by-products and cancer, and eonazoles and cancer, reproductive,
developmental, and neurological effects). The aim is to establish a priority for chemicals that
require further study for development of risk assessments.

       Another example of ongoing research in support of prioritization involves determining
the effects of contaminants on aquatic animals using protein profiling. This will result in the
ability to rapidly screen chemicals based on specific modes of action. Protein profiles can be
used as specific biomarkers of effects of bioactive compounds. The sheepshead minnow is being
used as a model species for proof-of-concept studies.

       A necessary component for the development of toxicological pathways in support of
prioritization (and for MOA and risk assessment approaches) is the establishment of standardized
approaches for conducting genomic and proteomic studies including data acquisition, data
storage, and bioinformatics approaches. Efforts to achieve these goals are underway within the
Agency.

2.  Monitoring Research Needs and Activities

       The research needs to support EPA's monitoring programs cover a broad range of issues,
including environmental monitoring and public health monitoring, specifically in the areas of
epidemiology and molecular epidemiology. It is likely that genomics technologies will prove
productive in each of these fields. The following examples provide an indication of the types of
research needs that the Agency faces in an expanded monitoring program and highlight Agency
activities that may help to address these needs.

       If environmental and human health assessments are to employ biological indicators,
reliable and informative markers of exposure, dose, and response must be selected. The more
that is known about the mechanisms involved in the pathway from exposure to adverse outcome
or response, the more readily informative biomarkers can be identified. The research need,
therefore, is to develop these mechanistic data, select proposed informative biomarkers, and
utilize these in field conditions or in molecular epidemiological studies.  Validation of the
biomarkers can be achieved through human or environmental health assessments, to establish
how accurately the various biomarkers predicted outcomes. This is a massive effort that will
require considerable collaboration within the Agency and outside the Agency.

       EPA is conducting research to address the need for environmental biomarkers. For
example, Agency scientists are conducting a comparison of the sensitivity of cellular indicators

-------
^	Genomics Task Force White Paper                    Page 33

of genetic damage in model stream fish using controlled laboratory exposures and subsequent
field validation. These indicators of cellular damage are likely to have an application in
ecological monitoring projects. Another example of ongoing research in this area involves the
development of methods for measuring the induction of the vitellogenin gene during water
monitoring studies.  This indicator provides information on exposure of an organism to an
endocrine disrupting chemical. The next step is to conduct research that can help establish the
relationship between measured exposure and adverse effect. Such information is required prior
to the use of the method in a regulatory setting.

       Similarly, for microbial source tracking, the relationship between existing indicators (e.g.,
total coliforms, enterococci, etc.) and genomics-based indicators must be established. The
relationship between occurrence and disease response in humans from human and non-human
sources of pathogens, especially bacteria, must be defined. These types of information will be
used by state and local agency decision makers to determine which methods are "acceptable,"
This is a  significant step because most agencies have limited resources and are often reluctant to
change to new technologies because of the associated high capital and human resource
investments that must be made. The Agency is evaluating ways to apply DNA-based technology
to detect  and track fecal contamination to its source in complex environmental matrices,
including recreational  and drinking water sources. A microarray method to identify potential
waterborne pathogens is also under development by the Agency. The Office of Water is
currently supporting research investigating the effects of specific gene combinations that are
associated with waterborne pathogen virulence. These projects could be applied to ambient and
drinking  water monitoring.

       The development of genetic markers and/or proteomic markers of plant responses to
herbicides and other xenobiotics is needed by the Regions to enhance monitoring capability for
assessing effects of spray drift and determining which plant groups are most likely to be at risk
from xenobiotics.  Researchers are currently studying proteomic responses of plants to high
potency,  low-dose herbicides as a method of monitoring exposure.  Further, markers for
monitoring gene transgression from genetically modified (GM) crops to non-crop plants as well
as for detecting contamination of non-GM seed shipments or foodstuffs with GM material, are
being developed for use by the Regions.

       Thus, the aim of a genomics research program in support of the Agency's monitoring
efforts is  to develop informative biomarkers of response for the assessment of the impact of
environmental exposures on human and environmental health. In parallel, such bioindicators can
be used to assess the impact of regulatory actions on human and environmental health.

-------
Page 34                    Genomics Task Force White Paper

C,  Research Needs and Activities for Risk Assessment

1. Mode of Action Research Needs and Activities

       There are numerous issues associated with MO As that require additional research before
genomics technologies can be fully utilized in risk assessments. An overriding issue that affects
more than just the MOA is the need to relate changes in gene expression to adverse effects.  To
establish the linkages between genomic changes and adverse outcomes, reliable data sets for gene
expression at the RNA and protein levels are required.  These data need to include a range of
sample times and exposure concentrations and be repeatable.  A parallel need is the development
of expertise in analyzing these data so that profiles of responses at the molecular level can be
produced and linked to specific chemicals or mixtures. This type of correlation then has to be
extended to adverse outcomes at the organ, tissue, and whole animal level.  This approach could
be applied in both human health and ecological risk assessments.

       Metabolic pathways for chemicals need to be defined, and the active metabolites that
cause cellular responses need to be identified. The use of mRNA or protein arrays would
enhance the Agency's ability to address this issue in a timely fashion. The initial requirement is
to establish a set of profiles that are in toto representative for each known metabolic pathway.

       Genomics-based approaches, including proteomic and metabonomic tests, will need to be
developed to reduce, refine, or replace more complex and costly standard tests for human and
wildlife species. Public pressure to reduce reliance on animal testing, particularly for
toxicological studies, will continue to increase, making this a relatively high Agency priority.
The overriding research need is the development of molecular profiles for in vitro cellular
models and for a suite of animal species exposed to chemicals.  The aim is to identify key
components of MOAs from such profiles for comparison with similar profiles in humans and
wildlife species. The long-term goal is to determine  whether molecular profiles can be used to
evaluate risk levels for chemicals with little toxicological information and for nontested species.
A high priority, long-term research goal of the Interagency Coordinating Committee on the
Validation of Alternative Test Methods (ICCVAM) is to investigate the utility of genomics  for
the assessment of acute toxicity, especially for the prediction of NOAELs and LOAELs. EPA's
membership in ICCVAM will promote collaboration with other federal agencies to achieve  this
goal.

       The following are descriptions of some of the research activities that are underway in
ORD that address MOA.  The list is not exhaustive, but does provide an idea of the breadth of
the activities under the MOA umbrella. A significant fraction of the genomics research currently
ongoing in the Agency is directed towards identifying MOAs for a range of chemicals for a
number of different adverse outcomes, including cancer, endocrine disruption, reproductive and
developmental effects, and neurotoxicity. These research activities are briefly presented to show
their range.

-------
                           Genomics Task Force White Paper                    Page 35

       In an attempt to use MOAs in the harmonization of risk assessment approaches, a
research team is comparing the effects of a group of conazole pesticides in different tissues and
for different endpoints. The aim is to establish if a common MOA is able to explain the range of
different endpoints and the specificity of tissue responses,

       Genomics technologies are also being used to characterize the MOA of selected drinking
water disinfection by-products for use in BBDR models.  In the same studies, genomics tools are
being applied to develop markers of response that will provide information for predicting adverse
outcomes at low doses.

       Gene array technologies are being used to identify biomarkers that will  be informative of
responses specific to the human testis. The MOA that is being developed for rodents will be
used to establish whether responses to particular chemicals have relevance to humans.  Other
studies are developing various biomarkers of response for environmental monitoring with
relevance to humans.

       Additional efforts are underway to  establish if readily available cells in  humans, such as
peripheral lymphocytes or buccal cells, can be used as predictors of adverse responses in tissues
that are targets for adverse outcomes such  as cancer, and reproductive and developmental effects.
Initial approaches involve the use of microarrays to study gene expression patterns in
lymphocytes and in germ cells using appropriate animal models and selected chemical  stressors.
In a similar way, genomics approaches are being used to establish whether markers of
susceptibility identified in readily available cells in humans, such as peripheral lymphocytes or
skin fibroblasts, can predict sensitivity to adverse health outcomes. These initial gene expression
studies will be expanded to include protein changes and functional associations with exposures.
Currently, these types of approaches are in the early stages of development within the Agency.

       As a part of the current effort in genomics, several groups are attempting to identify
informative biomarkers of response in laboratory animals as well as in sentinel species for use in
ecological assessments. It is proposed that MOA is a viable way of conducting interspecies
comparisons of outcome.

       The current and proposed Agency  genomics research directed towards enhancing our
knowledge of the various MOAs whereby  chemicals can induce adverse outcomes is focused on
identifying key events along toxicological  pathways from exposure to response. The
identification of key events will not only aid risk assessment approaches for single chemicals, but
will enhance efforts to harmonize risk assessment, to predict responses to chemical mixtures, and
to identify susceptible populations.

2. Susceptible Populations and Life Stages Research Needs and Activities

       Before the issue of how to incorporate susceptible populations into human health or
ecological risk assessment can be addressed through EPA policy, the methods for identifying

-------
Page 36                    Genomics Task Force White Paper

susceptible populations must first be developed along with quantitative methods for assessing the
magnitude of the sensitivity. To accomplish this, the Agency needs to develop knowledge of the
MOA for a chemical(s) of concern as well as the prevalence of this MOA in the population. For
example, this will require establishing the frequency of SNPs and their effects within human
populations as part of the identification of a susceptible population. An important proviso to
these types of studies is that ethical, social, and legal issues need to be addressed before starting
such work.

       There is a significant research effort in ORD to address the issue of children as a
susceptible lifestage. Specific focus regards the induction of diseases in children and the effects
of early life exposures on the development of adult diseases. The aims are to determine the
magnitude of any sensitivities and the underlying mechanisms that might account for increased
sensitivity. The genomics research component is directed towards developing informative
biomarkers of response that can eventually be used in animal model systems to predict adverse
outcomes from specific exposure scenarios and in human epidemiological studies.  These
informative biomarkers can also encompass specific genetic markers such as SNPs.

       Genomic methods, including proteomic approaches, may also be useful in more
accurately estimating exposures of individuals to contaminants in the environment, thereby
identifying susceptible populations at the exposure level. These emerging technologies could
lead to the development of personal dosimeters for a wide range of chemicals such that exposure
would be assessed at the individual level.  Similarly, genomic-level biomarkers (e.g., enhanced
personal microarray technologies) could provide a real-time, high throughput method for
screening potentially exposed individuals for incipient effects. While these approaches are
technologically feasible, the Agency has no definitive plans to develop research programs along
these lines in the near-term.

        Sensitive fish and/or wildlife species might serve as early indicators of overall ecosystem
health and as sentinels for risks to human health. In cases where chemical or pathogen
contamination reduces species fitness, genomics technologies  could be used to examine the
genetic makeup of a species in order, to determine the biochemical mechanism of the adverse
effect(s).  Like humans, plants and animals possess genetic polymorphisms that code for multiple
metabolic enzyme variants. In addition, levels and forms of the same enzyme (e.g., the
cytochrome P450 family of enzymes) vary between species and between life stages within
species.  Thus, as the genomes of species are sequenced, genomics can be used to identify the
most sensitive species and sensitive life stages. This will significantly enhance our ability to set
scientifically defensible water quality criteria or sediment and soil protection values under the
Clean Water and Safe Drinking Water Acts.

       Similarly, understanding genetic-based differences among plants and wildlife species in
terms of the MOAs of chemicals is a fundamental step towards understanding which one(s) will
be responsive. In the long term, this will enable more accurate cross-species extrapolations and
will significantly reduce the need for animal testing.  The Computational Toxicology initiative in

-------
                            Genomtcs Task Force White Paper                    Page 37

ORD is directed towards utilizing genomies to identify toxieity pathways. The current focus is
on humans and fish. In the future, genomic-based approaches need to be developed for other
wildlife species, as well as for aquatic and terrestrial plant species.

       ORD is currently developing tools to incorporate genomies technologies into population
dynamics models to enhance their predictive and explanatory power for assessing risks to
populations of wildlife and aquatic life. Genetic processes include the distribution and dynamics
of neutral and fitness-linked genetic markers. Depending upon their sophistication and data
requirements, the resulting population models can be used in screening to definitive tiers in the
ecological risk assessment process. In addition, a genetic dissection of the mechanisms of
resistance to anthropogenic contaminants is underway in zebrafish and fathead minnows. Both
of these research efforts will  yield information regarding the sensitivities of various fish species
and will likely be helpful in projecting the potential impacts of environmental contaminants on
ecosystem health.

       Similar approaches need to be considered for human populations. Here the need is to
establish the overall effect of environmental exposures on human health. This will require
knowledge of susceptible populations in terms of both the frequency and magnitude of
sensitivity. The use of genomies to aid in the development of informative bioindicators for this
effort is essential.

       In the context of assessing the impact of chemical exposures to overall  human and
ecological health, the influence of susceptible populations is of critical importance.  The needs
are to consider the roles of both life stages and genetic variation in the etiology of susceptibility.
While there is ongoing research addressing these issues, it is currently relatively limited.

3. Mixtures Research Needs and Activities

       Exposures of human and wildlife populations to environmental contaminants generally
involve complex mixtures of chemicals, rarely individual chemicals. Although there have been
some efforts to address responses to both simple and complex mixtures, much  of the past and
current research of the Agency has addressed the risk from exposures to single chemicals.
Clearly, addressing the overall toxicological responses to mixtures is a complex problem that
may require approaches different from those used for single chemicals.  Given the charge to the
Agency to increase its focus on research into the effects of mixtures, it is important to        how
genomies techniques might aid in meeting this need.

       A range of research is needed to assess the risks of chemical mixtures.  For example,  it is
necessary to determine if the MO A approach discussed above can be used to determine whether
a mixture can induce a qualitatively different set of key events than any of the individual
chemicals constituting the mixture. The next step would be to determine whether there are
quantitative differences in the induction of key events by mixtures as compared to the individual
chemicals and to use genomic measures to assess the magnitude of these definitive key events.

-------
Page 38                    Genomics Task Force White Paper

Because both human health and ecological risk assessment could benefit from a genomics
approach, discussion is underway concerning how to incorporate this type of research into the
Computational Toxicology initiative.

       The research needs for mixtures overlap considerably with those of prioritization, MOA,
and susceptible populations. Thus, mixtures assessment is an issue that will need to be addressed
in concert with these other Agency priority regulatory needs.

-------
Though not detailed in this paper,
bioinformatics (including data analysis and
interpretation) is one of the biggest challenges
to the effective use of genomics information.
                           Genomics Task Force White Paper                    Page 39

V.  Challenges and Recommendations

       As noted throughout this document, advances in genomics have significant implications
for risk assessment practice and regulatory decision making. The use of genomics technologies
generates a large volume of data and the field of bioinformatics is evolving rapidly to meet data
analysis needs.  The Agency's Interim Policy   	
on Genomics (USEPA, 2002a) appropriately
acknowledges that genomics technologies
have the potential to improve our
understanding of an organisms response to
stressors. This information, in turn, can lead
to the development of predictive biomarkers
of effect and allow the identification of potentially sensitive populations and earlier predictions
of adverse outcomes.  Early detections can be converted into more effective intervention
strategies. Genomics technologies will also enhance the understanding of the molecular
mechanisms of toxicity.  This will significantly reduce the uncertainty of extrapolations used in
the risk assessment process. The result will be the development of more sensitive and cost-
effective methods for toxicity screens and tests. Although, as the Interim Policy states,
understanding genomic responses with respect to adverse  ecological and/or human health
outcomes is far from established, it is important for managers to begin to consider the likely
future impacts of genomics technologies on their programs.

       Chemical production is highest in the Organization for Economic Cooperation and
Development (OECD) countries and that growth is fastest in specialty chemicals and the life
science sectors (OECD, 2001).  Moreover, innovation in new chemical development and
manufacturing practices is extremely high due to advances in combinatorial chemistry,
nanotechnology, and biotechnology. These changes have  raised concerns about the Agency's
ability to sustain its current approaches  to prioritization, monitoring, and risk assessment
activities.

       This paper outlined the potential of genomics technologies to improve and refine the
current approach to regulatory applications and risk assessment and identified genomics as a
means to alleviate the above concerns.  There are, however, a number of challenges that must be
overcome in order for genomics technologies to be fully applied to regulatory decision making.
In this regard, the Genomics Task Force identified three categories of overarching scientific and
resource challenges: research, technical development, and capacity. To address the regulatory
and risk assessment applications outlined in this paper and to most effectively use genomics
information, the Agency needs to address these challenges.

-------
Page 40                    Genomics Task Force White Paper

A.  Research Challenges

1. Linking Genomics Information to Adverse Outcomes

       Linking genomics changes to adverse outcomes represents a significant research
challenge that must be addressed before genomics data can provide information essential to the
support of risk assessment and regulatory decision making. Additionally, establishing a
quantitative relationship between gene expression changes and adverse responses will provide
essential  information. As noted throughout this paper, changes in gene expression at the mRNA
and protein levels need to be related to cellular effects and, ultimately, to adverse outcomes. In
many ways, the detection of gene expression changes is the easiest part of a genomic assessment.
However, in a risk assessment framework, it is necessary to link a function to genes whose
expression is altered. Gene expression changes that encompass defense mechanisms, which may
be adaptive or beneficial and bear no causal relationship with the development of pathologies,
must be separated from those that damage key cell functions (e.g., cell cycle control, structural
integrity of proteins, control of free radicals, or loss of homeostasis and DNA repair
mechanisms).  Combining the findings of gene expression studies with data from in vivo
chemical exposure of genetically altered animal models (e.g., knockout or null mice) is a
powerful way to link specific genes to specific detrimental outcomes. Simpler whole organ
systems may also offer powerful means to link genomic response to adverse effect. Key
biological systems have fundamental genomic processes, some of which, if altered, are
universally deleterious.

2. Interpretation of Genomics Information for Risk Assessment

        Genomics information can be very relevant, and at times critical, to Agency risk
assessments  by providing mechanistically oriented insight into the hazard identification, dose-
response, and exposure portions of risk assessments. This paper outlined specific areas where
genomics data may aid in risk assessment including MOA assessment, identification of
individual and population susceptibility, application of biomarkers of exposure, evaluation of
effects of mixtures, understanding gene-environmental interactions, and application of a systems-
wide examination of responses  to stressors.

       As a  major example  of how genomics information can provide insight for risk
assessment, the mode of action of a  stressor has been discussed. Mechanism  or mode of action
information can help identify potential hazards and help interpret dose-responses and
extrapolations. Genomics technologies can be used to better understand the MOA of a chemical
agent, and thus lead to advances in human and ecological risk assessments of chemicals. As
genomics information contributes to our understanding of MOAs, the validity of using genomics
information is  in turn enhanced as an indicator of both adverse effects and exposure.

-------
                           GenomicsTask Force White Paper                    Page 41

       Providing links between genomics changes and phenotypic changes at the cell and tissue
levels requires the use of a number of rapidly evolving cellular and molecular techniques (e.g.,
immunocytochemistry, gene silencing) and bioinformatic technologies. New approaches using
tissue microarrays will enhance throughput and the linking of genomic and cellular outcomes.
However, approaches to unraveling a profile of gene expression linked to a significant
toxicological event present a number of challenges (Fielden and Zacharewski, 2001) in part
because of the magnitude of the data sets developed and the potential variations in the level of
expression for a single parameter (e.g., expression of a single gene). Analysis of the large data
sets generated via genomics assays will require the development of new bioinformatic and
computational tools. An integrated analysis and understanding of biological systems and their
responses to perturbation, from genes to adverse effects, and the capacity to collect and evaluate
data supportive of such a view, would be expected to greatly enhance the risk assessment
process, and thus aid in formulating regulatory policy and making regulatory decisions,

       Understanding the MOA of environmental agents that induce toxic effects other than
cancer or induce earcinogenieity in animal models should facilitate the assessment of the
relevance of these findings in protecting human health and safe guarding the environment. An
important issue for extrapolation of responses in animal models to humans or environmental
endpoints is to establish whether the MOA in the test species is relevant in the target species. A
range of different types of data can be used to establish a MOA, but the endpoints for cross-
species extrapolation generally are more limited.  Such approaches will aid in addressing EPA's
challenge to harmonize risk assessment approaches for different outcomes.

3. Recommendations to Address Research Challenges

       In order to contribute to the development of linkages between genomic changes and
adverse outcomes and to the interpretation of genomics information for risk assessment, the
Agency should aggressively support and build its own genomics research through the ORD
Computational Toxicology initiative and support external  research through competitive grants
and contracts.  Research plans and timing should be guided by developments in the genomics
field.  Through appropriate direction of its research, EPA can support important regulatory
applications that are more likely to arise in the  near future. These include priority-setting
activities such as high  throughput screening of chemicals,  and monitoring activities such as
source tracking of pollutants and pathogens in water.  EPA should also encourage industry efforts
to conduct genomics research necessary for application in  risk assessment, when developing its
data.  It is critical that the Agency coordinate research with other agencies and institutions to link
genomic changes and adverse outcomes

-------
Page 42                    Genomics Task Force White Paper

B.  Technical Development Challenges

1. Framework for Analysis and Acceptance Criteria for Genomics Information

       EPA acknowledges that genomics technologies will eventually contribute to risk
assessment through a better understanding of mechanisms of chemical toxicity, dose-response
relationships, identification of susceptible populations, and estimates of uncertainty factors.
However, to date, EPA has had limited access to relevant genomics data to begin examining its
potential influence. Even without specific cases relevant to the Agency, it is clear that a plan is
needed to develop methods for incorporating these types of information into the decision making
process.

       EPA and other regulatory agencies recognize the requirement to develop acceptance
criteria for genomics data. The Food and Drug Administration's (FDA) Center for Drag
Evaluation  and Research is in the process of identifying which pharmacogenomic data developed
by companies for the evaluation of human drugs will be required by the FDA due to the
regulatory implications of the data (FDA, 2003). The FDA is consulting with its Federal
Advisory Committee and stakeholders to develop its policies and guidance.  The efforts at the
FDA are aimed at developing a framework for submission, storage, analysis, and regulatory
review of genomics data. This is driven, in part, by the high level of use of genomics
information for human drug development and evaluation.

       The use of genomics information for the analysis of risks of industrial chemicals,
agrochemicals, and microbial contaminants is not as widespread at the current  time. However, as
the understanding of genomics data and its relevance and applicability to chemical and microbial
risk analysis increases, EPA will need to continue developing its own technical framework for
the consideration of genomics information for scientific and regulatory purposes.  The full range
of ethical and legal issues (Marchant, 2003) will also have to be considered as  genomics is
incorporated into the Agency's risk assessment process.

2. Recommendations to Address Technical  Development Challenges

       The Genomics Task Force recommends that the Agency charge a workgroup with
developing a technical framework for analysis and acceptance criteria for genomics information
for scientific and regulatory purposes. This framework should build upon EPA's Interim Policy
on Genomics.  Issues that need to be considered in developing such a framework include
consideration of the performance of assays across genomic platforms (e.g., reproducibility,
sensitivity) and the criteria for accepting genomics data for use in a risk assessment (e.g., assay
validity, biologically meaningful response).  It is essential for the Agency to continue to engage
other federal agencies, such as the FDA, NIEHS, and Department of Energy, as well as other
stakeholders, including industry, academia, and public interest groups when developing this
framework. Such a framework, once established, can be used by the EPA program offices to

-------
                           Genomics Task Force White Paper                    Page 43

determine the applicability of specific genomics information to the evaluation of chemical risks
under various statutes. The Interim Policy of considering genomics data on a case-by-case basis
should continue to be applied until the technical framework is completed,

C.  Capacity/Human Capital Challenges

1. Applying Strategic Hiring Practices to Recruit Individuals Who Possess "Genomics
Core Competencies"

       An important undertaking will be to identify the skills needed to establish "genomics core
competencies" and to apply strategic hiring practices to recruit individuals who possess these
skills.  It will be essential to have technical specialists in genomics on staff in the crucial areas of
research, analysis, systems biology, bioinformaties, and risk assessment to enhance the Agency's
expertise in genomics and related technologies.

2. Training EPA Risk Assessors and Managers to Interpret and Understand Genomics
Data in the Context of a Risk Assessment

       It will be essential to train current EPA risk assessors so that they will be prepared to
interpret and apply genomics data in the context of a risk assessment, including consideration of
genomics data uncertainties. Risk assessors must be able to communicate both the underlying
science and the interpretative tools and models used to develop the risk assessment to risk
managers and stakeholders.  Along similar lines, it will be important to provide training to risk
managers regarding the use of genomics information in risk assessments and the strengths and
limitations of such data.

       A related concern is the capacity of regions, states, tribes, and local agencies to
implement genomics tools and to evaluate genomics data, particularly with respect to their
responsibilities under delegated programs. They will need resources, technical support, and
targeted training to bring the scientists and managers within their organizations to a point where
they will be able to effectively use genomics tools in their regulatory decision making, especially
with respect to risk characterizations. Regional, state, tribal and local agencies' use of genomics
tools will require both capital investment for analytical equipment, plus ongoing expenses for
disposables such as the microarrays and associated supplies.

3. Recommendations to Address Capacity/Human Capital Challenges

       EPA programs and regions should apply strategic hiring practices to recruit individuals
who possess genomics skills and should consider existing guidance from other agencies, such as
the Centers for Disease Control, regarding recommended "genomics core competencies."

-------
Page 44                    Genomics Task Force White Paper	___^_

       The Genomics Task Force also recommends that the Agency convene a workgroup tasked
with developing training modules for the interpretation and application of genomics data for risk
assessments for both risk assessors and risk managers.  It would be useful to develop and initiate
training in the near future to prepare risk assessors and risk managers, because genomics issues
have begun to arise in environmental decision making. The initial training could address basic
genomics concepts, technologies and potential applications including consideration of the basic
steps necessary to interpret and apply genomics data to a risk assessment. Regions 9 and 10, in
collaboration with ORD, have already developed and conducted basic genomics training on a
pilot basis. Development of Agency-wide training materials could build upon these efforts and
those of external sources.  It is expected that the training would need to be revised and expanded
as our understanding of genomics improves over time. The training could also be offered on a
limited basis to tribes, states, and local governments to assist in the development of their
capability to effectively use genomics tools in regulatory decision making. Finally, the Genomics
Task Force recommends that the development of training tools and workshops be conducted in
collaboration with other agencies and institutions.

       It is recognized that it may be difficult to allocate additional resources (e.g., funds,
people, and training programs) required to efficiently incorporate and effectively understand
these new types of genomics data.  It is important for EPA not to lose focus on these important
needs and ensure that adequate funds and people are brought to bear on this need.

D.  Summary Recommendation

       The Genomics Task Force recommends that EPA begin taking steps to address the
identified research, technical development, and capacity challenges to strengthen its capability to
effectively use genomics information (see table below for summary).
                          Recommendations to Meet Challenges
 Research
        a)     Encourage work linking genomics changes to adverse effects
        b)     Build EPA research capacity (e.g., Computational Toxicology program)
        c)     Consider genomics capabilities in developing research plans
        d)     Encourage others (e.g., industry) to conduct research applicable to risk
               assessment
 Technical Development
        a)     Form technical framework workgroup(s), to develop:
               1)     acceptance criteria (e.g., data quality, experimental design)
               2)     analysis guidance
               3)     performance standards
        b)     Engage all stakeholders

-------
                           Genomics Task Force White Paper	Page 45
                          Recommendations to Meet Challenges
 Capacity/Human Capital
        a)     Apply strategic hiring practices to recruit talented individuals with genomics
               skills
        b)     Develop training tools, modules, etc. for different audiences (e.g., risk
               assessors, risk managers, non-technical users)
        c)     Collaborate with others in workshops
        d)     Work via budget process so adequate resources are available for these needs
       It is essential for the Agency to continue to collaborate with other federal agencies,
academia, the regulated community, public interest groups, and other stakeholders in this
endeavor, in order to benefit from ongoing advances in genomics in the wider scientific and
regulatory communities. Such collaborations can include development of training tools,
workshops to ensure interaction of active parties, development of common approaches for use,
storage, acceptance, and interpretation of genomics related information, and leveraging research
resources to increase our research capabilities.

-------
Page 46                    Genomics Task Force White Paper

                                 REFERENCES

Aardema, MJ., MacGregor, J.T. 2002. Toxicology and genetic toxicology in the new era of
"toxicogenomics": impact of "-omics" technologies. Mutation Research 499(1): 13-25.

Bartosiewicz, MJ., Jenkins, D., Penn, S., Emery, J., Buckpitt, A. 2001a. Unique gene expression
patterns in liver and kidney associated with exposure to chemical toxicants. Journal of
Pharmacology and Experimental Therapeutics 297:895-905.

Bartosiewicz, M,, Penn, S., Buckpitt, A. 2001b. Applications of gene arrays in environmental
toxicology: fingerprints of gene regulation associated with cadmium chloride, benzo(a)pyrene,
and trichloroethylene. Environmental Health Perspectives, 109:71-74.

Bishop, W.E., Clarke, D.P., Travis, C.C. 2001. The genomic revolution: what does it mean for
risk assessment? Risk Analysis 21(6):985-987.

Feron, V.J., Groten, J.P. 2002. Toxicological evaluation of chemical mixtures. Food and
Chemical Toxicology 40:825-839.

Fielden, M.R., Zacharewski, T.R. 2001. Challenges and limitations of gene expression profiling
in mechanistic and predictive toxicology. Toxicological Sciences 60:6-10.

Center, M.B., Burman, D.M., Soundarapandian, V., Ebert, C.L., Aronow, B.J. 2002. Genomic
analysis of alachlor-induced oncogenesis in rat olfactory mucosa. Physiological Genomics 12:35-
45.

Griffith, J.F., Weisberg, S.B., McGee, C.D. 2003. Evaluation of mierobial source tracking
methods using mixed fecal sources in aqueous test samples. Journal of Water and Health
Haber, L.T., Maier, A., Gentry, P. R., Clewell, H.J., Dourson, M.L. 2002. Genetic
polymorphisms in assessing interindividual variability in delivered dose. Regulatory Toxicology
and Pharmacology 35:177-197.

Hakkola, J., Pelkonen, O., Pasanen, M., Raunio, H. 1998. Xenobiotic-metabolizing cytochrome
P450 enzymes in the human feto-placental unit: role in intrauterine toxicity. Critical Reviews in
Toxicology 28(1): 35-72.

Hertzberg, R.C., MacDonell, M.M. 2002. Synergy and other ineffective mixture risk definitions.
Science of the Total Environment 288:31-42.

-------
                           Genomics Task Force White Paper                    Page 47

Klaper, R., Euling, S., Frederick, R., Sonawane, R. 2003, Current use and future needs of
genomics in ecological and human health risk assessment. Internal Workshop Report, May 8,
Alexandria, VA. EPA/600/R-04/039.

Marchant, G.E. 2002. Genomics and environmental regulations: scenarios and implications.
Report under Foresight and Governance Project at the Woodrow Wilson Center for Scholars. 29
pp. http://www.environmentalfutures.org/Images/marchan twhitepaper.pdf

Marchant, G.E. 2003. Genomics and toxic substances: Part I - Toxicogenomics. Environmental
Law Reporter 33:10071-10093.

Mumtaz, M.M., Tully, D.B., El-Masri, H.A., De Rosa, C.T. 2002. Gene induction studies and
toxieity of chemical mixtures. Environmental Health Perspectives 110  (Supplement 6):947-956.

Organization for Economic Cooperation and Development. 2001. The Chemicals Industry. In:
Environmental Outlook, OECD Publications, Paris, pp. 224-225.

Pohl, H.R., Hansen, H., Chou, C.H.S.J. 1997. Public health guidance values for chemical
mixtures: current practice and future directions. Regulatory Toxicology and Pharmacology
26:322-329.

Porter, W.P., Hinsdill, R.D., Fairbrother, A., Olson, L.J., Jaeger, I, Yuill, T.M., Bisgaard, S.,
Hunter, W.G., Nolan. K. 1984. Toxicant-disease-environment interactions associated with
suppression of immune system, growth and reproduction. Science 224:1014-1017.

Sappington, L.C., Mayer, F.L., Dwyer, F.J., Buckler, D.R., Jones, J.R.,  Ellersieck, M.R. 2001.
Contaminant sensitivity of threatend and endangered fishes compared to standard surrogate
species. Environmental Toxicology and Chemistry 20(12):2869-2876.

Simpson, J., Santo Domingo, J.W., Reasoner, D.J. 2002. Microbial source tracking: state of the
science. Environmental Science and Technology 36(24):5279-5288.

Snawder, J.E., Lipscomb, J.C. 2000. Interindividual variance of cytochrome P450 forms in
human hepatic microsomes: correlation of individual forms with xenobiotic metabolism and
implications in risk assessment. Regulatory Toxicology and Pharmacology 32:200-209.

Suk, W.A., Olden, K., Yang, R.S.H. 2002. Chemical mixtures research: significance and future
perspectives. Environmental Health Perspectives 110 (Supplement 6):891-892,

U.S. Department of Energy, Office of Science, Biological and Environmental Research Division.
http://www.sc.doe.gov/ober/ERSD/ersdnabir.html

-------
Page 48                   Genomics Task Force White Paper

U.S. Environmental Protection Agency. 1999. Glossary of Integrated Risk Information System.
http://www.epa.gov/iris/gloss8.htm

U.S. Environmental Protection Agency, Science Policy Council. 2002a. Interim Policy on
Genomies. http://www.epa.gov/OSP/spc/genomics.pdf

U.S. Environmental Protection Agency. 2002b. A Review of the Reference Dose and Reference
Concentration Processes. EPA/630/P-02/002F.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=SS365

U.S. Environmental Protection Agency. 2003a. Draft Final Guidelines for Carcinogen Risk
Assessment. EPA/630/P-03/001A. http:www.epa.gov/ncea/raf/cancer2003.htm

U.S. Environmental Protection Agency. 2003b. A Framework for a Computational Toxicology
Research Program in ORD. EPA/600/R-03/065.
http://www.epa.gov/comptox/publications/comptoxframeworlc06 02 04.pdf

U.S. Environmental Protection Agency. 2003c. Pesticide product label system (PPLS).
http://www.epa.gov/pesticides/pestlabels/

U.S. Food and Drug Administration. 2003. Guidance for Industry: Pharmacogenomic Data
Submissions, Draft Guidance.
http://www.fda.gov/OHRMS/DOCKETS/98fr/2003d-0497-gdl0001.pdf

-------
	Genomics Task Force White Paper	Page A-l

                 Appendix A: Interim Policy on Genomics

                               Science Policy Council

                                    June 25,2002

Background
       The Environmental Protection Agency (EPA) is aware of the rapidly advancing field of
genomics since completion of the initial sequencing of the human genome. EPA expects that
genomics data may be received as  supporting information for various assessment and regulatory
purposes, e.g., identifying an environmental stressor's mode or mechanism of action. Genomic
research tools now permit the study of gene and protein expression changes in various organisms
and their cells, or tissues, with specificity to the level of molecular function. While genomics
offers the opportunity to understand how an organism responds at the gene expression level to
stressors in the environment, understanding of such molecular events with respect to adverse
human or ecological health outcomes is far from established. Understanding these relationships
becomes increasingly more complex as the number of sequenced genomes for an ever increasing
variety of organisms becomes available. Thus, there is a need for the Agency to provide an
interim policy on the current interpretation and utility of genomics data in the context of risk
assessment and risk management and the implications this has for EPA's infrastructure needs.
       Genomics approaches have the long term promise to  aid in the understanding of an
organism's response to stressors and to guide the selection of informative bioindicators for
monitoring the impact of stressors on human and ecological health. Thus, EPA believes that
genomics will have an enormous impact on our ability to assess the risk from exposure to
stressors and ultimately to improve our risk assessments.

Science Policy Council's Interim Position
       EPA believes that genomic data and analyses will significantly impact many areas of
scientific research and human and  ecological health assessments. Genomics data may allow EPA
to enhance its assessments and better inform the decision-making process. Accordingly, EPA
must understand how to develop and use the research tools made possible from genomics and
understand the appropriate uses of genomics data to inform Agency decisions.  For EPA, the
term "genomics" encompasses a broader scope of scientific inquiry and associated technologies
than when genomics was initially considered. A genome is the sum total of all an individual
organism's genes. Thus, genomics is the study of all the genes of a cell, or tissue, at the DNA
(genotype), mRNA (transcriptome), or protein (proteome) levels.
       Genomics methodologies are expected to provide valuable insights for evaluating how
environmental stressors affect cellular/tissue functions and how changes in gene expression may
relate to adverse effects. However, the relationships between changes in gene expression and
adverse effects are unclear at this time and may likely be difficult to elucidate.  Nonetheless, EPA
believes that some of these changes may prove to be predictive of subsequent adverse effects.
Changes in gene expression can be informative when a weight-of-evidence approach for human

-------
Page A-2                    Genomics Task Force White Paper

and ecological health assessments is performed, particularly when used to explore the possible
link between exposure, mechanism(s) of action, and adverse effects. In addition, genornics
information may be useful to EPA in setting priorities, in ranking of chemicals for further testing,
and in supporting possible regulatory actions. While genornics data may be considered in
decision-making at this time, these data alone are insufficient as a basis for decisions. For
assessment purposes, EPA will consider genornics information on a case-by-case basis. Before
such information can be accepted and used, agency review will be needed to determine adequacy
regarding the quality, representativeness, and reproducibility of the  data.
       EPA believes that genornics will ultimately improve the quality of information used in the
risk assessment process. For example, genornics shows promise to identify variability and
susceptibilities in individuals from exposed populations or among different species. It will also
likely provide  a better understanding of the mechanism or mode of action of a stressor and thus
assist in predictive toxicology, in the screening of stressors, and in the design of monitoring
activities and exposure studies.  Application of genomics methodologies may help reduce or
eliminate traditional types of toxicity testing as well as improve the scientific rationale for when
such testing is needed. Genomic analysis also holds promise to evaluate the cumulative impacts
resulting from the interplay of factors such as genetic diversity, health status, and life stage in
responding to exposure(s) to multiple stressors.
       EPA encourages further research on methods development,  methods evaluation, and data
collection to address existing gaps in knowledge concerning the consequences of genomic
changes. EPA's goal is to develop knowledge that will ultimately reduce the uncertainties in the
assessment of  hazard, exposure, and risk from stressors. In parallel with data generation, there is
an equal need for developing information technologies, for research on the analysis of data, and
for applications of genomics data in computational toxicology. As EPA gains experience in
applying genomics information and refines its understanding of the use of such information, it
will develop guidance to explain how genomics data can be better utilized in informing decision-
making and related ethical, legal, and social  implications. EPA is working with other Federal,
state, and Tribal organizations, as well as with academic, international, and industry groups to
facilitate scientifically sound applications of genomics data. In addition, EPA will continue to
build partnerships and communicate with all interested stakeholders as an essential component of
the Agency's future activities in genomics.

-------
                          Genomics Task Force White Paper                  Page B-1

Appendix B: Details on Additional Activities Potentially Affected by
                            Genomics Information

I. Prioritization

A.  Program Offices

OPPT - Premanufacture Notices (PMN): Genomics data may be useful for evaluating PMNs.
Genomics data generated for PMNs may be able to supplement, or potentially supplant, computer
model results or expert judgment in hazard estimation and prioritization activities.

OPPT - Voluntary Children's Chemical Evaluation Program (VCCEP): Genomics data may
be useful in the VCCEP program. These data could support, or potentially supplant, current
toxicity tests.

OPPTS/OSCP - Endocrine Disrupters Screening Program (EDSP): OSCP has been
developing a tiered approach for testing under the EDSP. Currently, prioritization is primarily
based upon exposure information, but it is anticipated that chemical prioritization will be greatly
facilitated by the use of high throughput screening. Genomics could be used in the high
throughput screening process.

OPP - Prioritization of Pesticides and Inert Chemicals: Genomics data could be useful for
prioritizing pesticide products for testing procedures.  For example, data demonstrating that an
agent does not elicit differential gene expression that is predictive of toxicologically relevant
responses may indicate that a pesticide is non-toxic in a particular test species.  The chemical
might then be slated to receive expedited review under the reduced risk chemicals program, and
waiver requests for associated standard toxicity tests might be considered. Conversely, data
indicating that a pesticide produces an altered gene expression profile for a toxicological pathway
that is relevant for an adverse outcome may potentially signal an alert.  The pesticide may then be
assigned to an evaluation pathway involving a second level of genomics and standard toxicity
testing.

OW - Prioritizing Streams or Wetlands for Study or Clean Up:  Genomics data could be
applied to prioritizing streams and wetlands for additional study or clean up activities.

OAR - Hazardous Air Pollutants:  Genomics data may be useful in prioritizing hazardous air
pollutants for chemical testing.

OSWER - Superfund:  Superfund site prioritization  may be enhanced through the use of
genomics data.

-------
Page B-2                  Genomics Task Force White Paper

ORD - Research Planning: Genomics data may provide useful information to ORD researchers
and managers in prioritizing chemicals for future research.

B.  Regions, States, and Tribes

Resource Prioritization for Site Remediations and Chemical Evaluations: Genomics
technologies could provide regions, states, and tribes with a fast, relatively low cost method to
prioritize their areas of focus and deployment of resources for delegated program site
remediations and chemical evaluations.

II.  Monitoring

A.  Program Offices

OPP - Exposure Monitoring: OPP uses exposure monitoring data for new chemicals generated
via Experimental Use Permits (EUPs) in registration decisions, and exposure monitoring data for
currently registered chemicals for reregistration eligibility decisions (REDs). For these EUPs and
REDs, genomics data could be used to track the movement of pesticides off-site via spray drift or
into ground or surface water. Biological monitoring data for human and wildlife exposures and
potential effects could contribute to regulatory decisions, and genomics technologies could
provide information about occupational exposures and wildlife incident data for reregistration
decisions.

OAR/Office of Air Quality Planning and Standards (OAQPS) - Stationary Source
Monitoring:  Genomics technologies could contribute to stationary source monitoring conducted
under the Clean Air Act (CAA).

OSWER - Superfund Monitoring: Genomics technologies could be applied to monitor
movement of contaminants off-site from Superfund sites prior to remedial actions. In evaluating
contaminated sites, near term benefits could be derived by targeting toxic chemical remediation
using biomarkers present in lower organisms (e.g., metallothionein expression).  Bioavailability,
a key parameter in determining the toxicity of a chemical, can be effectively determined using
genomics tools; thus greater precision could be achieved in remedial activities. That is, the truly
hazardous compounds could be identified and removed with more precision, and other materials
need not be disturbed unnecessarily. Further, monitoring the operation and maintenance of
remedial actions and residual contaminants at Superfund sites that have undergone cleanup could
be enhanced through genomics technologies.

OSWER/OSW - RCRA-Required Monitoring:  Genomics could contribute to post-clean up
monitoring activities conducted under RCRA.

-------
                           Genomics Task Force White Paper                  Page B-3

OEI, ORD - Bioindicator Development:  OEI and ORD expect to use genomics approaches to
develop a selection of informative bioindicators for monitoring the exposures and effects of
stressors on human and ecological health.

B.  Regions, States, and Tribes

State and Local Beach Closures - TMDL Issues Associated with Pathogens: A possible
near-term scenario is the use of genomics technologies to detect microbial pathogens and to
determine their sources (so-called microbial or bacteriological source tracking, MST or BST).
The Southern California Coastal Water Research Program, a State of California research agency
that receives partial funding from waste water dischargers, has sponsored the first round of
research addressing the feasability of molecular-based MST techniques.  Several Regions are
working with this group and with ORD scientists to identify the "best performers" among the
various MST techniques and to produce guidance on the effective use of use these methods. This
guidance could be used by State and local agencies to make decisions about testing ambient
surface waters for beach closures and establishing pathogen/bacterial TMDLs. Recently,
however, EPA researchers evaluating MST methods have concluded  these methods will require
further development before they can be widely applied (Simpson et al., 2002).

Air Quality Monitoring Program: EPA's ambient air quality monitoring program for criteria
pollutants is carried out by State and local agencies. Genomics approaches could be used in the
state and local air monitoring programs.

Endocrine-Disrupter Monitoring: Another example of near-term decision making using
genomics data comes from a group of tribes in Northern California and Southern Washington
that is interested in the potential impact of exposure to pharmaceuticals and personal care
products and the potential for endocrine disrupter effects from municipal waste treatment
facilities and/or cattle grazing on the health of wild salmon populations that are part of tribal
cultural and economic resources. The Tribe proposes to use a series of molecular-biology-based
assays for exposure to hormonally active compounds, either a multiplex RT-PCR approach or a
multigene array. The information could be used to develop NPDES permit limits and establish
Tribal Water Quality Standards.

Regional Pesticide Program Decisions: Another use of genomics data may be in the area of
the  regions' pesticides programs. If genomics data show that a particular unregistered pesticide
is of concern in certain populations of humans, animals', or plants, the region could work with the
state to assign a higher fine to an incident of use and/or offering for sale if the violation was in an
area where such a  sensitive population existed. Alternatively, the "harm value" might be lower if
the  non-target groups in an area were not considered to be part of a sensitive population.

-------
Page B-4                  Genomics Task Force White Paper

III. Risk Assessment - Mixtures

A.  Program Offices

OSWER, OAR, OW - Site Assessments/Monitoring:  Genomics technologies could be used as
diagnostic indicators to help identify components of chemical mixtures with unidentified
constituents. This approach could be used in initial assessments of contaminated sites or in
general monitoring applications.

B.  Regions, States, and Tribes

Site Assessments/Monitoring: Regions, states, and tribes could apply genomics diagnostic
indicators to identify components of chemical mixtures with unidentified constituents for initial
assessments of contaminated sites or in general monitoring applications.

-------
                           Genomics Task Force White Paper                   PageC-1

                             Appendix C: Glossary

Allele: an alternative form of a gene or any other segment of a chromosome1.

Bioinformatics: the analysis of biological information using computers and statistical
techniques; the science of developing and utilizing computer databases and algorithms to
accelerate and enhance biological research1.

Biomarker: a molecular indicator of a specific biological property; a biochemical feature or
facet that can be used to measure the progress of disease or the effects of treatment1.

Biotechnology: the set of biological techniques developed through basic research and now
applied to research and product development. In particular, biotechnology refers to the use by
industry of recombinant DNA, cell fusion, and new bioprocessing techniques2.

Complementary DNA (cDNA): DNA made from a messenger RNA (mRNA) template. The
single-stranded form of cDNA is often used as a probe in physical mapping1.

Computational Toxicology - Comp Tox:  the application of mathematical and computer
models and molecular biological approaches to improve the Agency's prioritization of data
requirements and risk assessments3.

Deoxyribonucleie acid (DNA): the substance of heredity; a large molecule that carries the
genetic information that cells need to replicate and to produce proteins4. The nucleic acid that
constitutes the genetic material of all cellular organisms and DNA viruses. The genetic
information is used in the synthesis of ribonucleic acids (RNAs) from DNA templates
(transcription) and in  the synthesis of proteins from messenger RNA (mRNA) templates
(translation).

Expressed sequence  tag: a unique  stretch of DNA within a coding region of a gene that is
useful for identifying  full-length genes and serves as a landmark for mapping1.

Gene:  the fundamental physical and functional unit  of heredity. A gene is an ordered sequence of
nucleotides located in a particular position on a particular chromosome that encodes a specific
functional product (i.e., a protein or RNA molecule)2.

Gene chip technology: development of cDNA microarrays from a large number of genes; used
to monitor and measure changes in gene expression  for each gene represented on the chip2.

Gene expression: process by which a gene's coded information is converted into the structures
present and operating in the cell.  Expressed genes include those that are transcribed into mRNA

-------
Page C-2                   Genomics Task Force White Paper

and then translated into protein and those that are transcribed into RNA but not translated into
protein (e.g., transfer and ribosomal RNAs)2.

Genetics: the study of inheritance patterns of specific traits2.

Genetic testing: analyzing an individual's genetic material to determine predisposition to a
particular health condition or to confirm a diagnosis of genetic disease2.

Genomics: the study of all the genes of a cell or tissue, at the DNA (genotype), mRNA
(transcriptome), or protein (proteome) level5.

Genome: all the genetic material in the chromosomes of a particular organism; its size is
generally given as its total number of base pairs2.

Genotype: the genetic composition of an organism or a group of organisms; a group or class of
organisms having the same genetic constitution1.

Hazard Assessment: the process of determining whether exposure to an agent can cause an
increase in the incidence of a particular adverse health effect (e.g., cancer, birth defect) and
whether the adverse health effect is likely to occur in humans6.

In Silico: literally "within silicon "; refers to modeling research conducted with computers
only7.

In Vitro: literally, "in glass," i.e., in a test tube or in the laboratory; the opposite of in vivo (in a
living organism)1.

In Vivo: in a living organism, as opposed to in vitro (in the laboratory)1.

Knockout: inactivation of specific genes. Knockouts are often created in laboratory organisms
such as yeast or mice so that scientists can study the knockout organism as a model for a
particular disease1.

Mapping:  charting the location of genes on chromosomes1.

Mass spectrometry: a method used to determine the masses of atoms or molecules in which an
electrical charge is placed on the molecule and the resulting ions are separated by their mass to
charge ratio1.

Messenger RNA (mRNA): a type of RNA that reflects the exact nucleotide sequence of the
genetically active DNA. mRNA carries the "message" of the DNA to the cytoplasm of cells
where protein is made in amino acid sequences specified by the mRNA1.

-------
                           Genomics Task Force White Paper                   Page C-3

Metabonomics: the evaluation of tissues and biological fluids for changes in metabolite levels
that result from toxicant-induced exposure1.

Microarray: a tool used to sift through and analyze the information contained within a genome.
A microarray consists of different nucleic acid probes that are chemically attached to a substrate,
which can be a microchip, a glass slide, or a microsphere-sized bead1.

Northern blot:  a technique used to separate and identify pieces of RNA1.

Nucleotide: a subunit of DNA or RNA. To form a DNA or RNA molecule, thousands of
nucleotides are joined in a long chain1.

"Omics": term including genomics, proteomics, metabonomics (some differentiate this term
from metabolomics), transcriptomics, and associated bioinformatics8.
          <

Phenotype: the observable physical or biochemical traits of an organism as determined by
genetics and the environment; the expression of a given trait based on phenotype; an individual
or group of organisms with a particular phenotype1.

Polymorphism: the quality or character of occurring in several different forms1.

Proteome: all of the proteins produced by a given species just as the genome is the totality of
the genetic information possessed by that species1.

Proteomics:  study of the full set of proteins encoded by a genome2.

Risk Assessment:  a qualitative or quantitative evaluation of the risk posed to human health and
the environment by the  actual or potential presence of pollutants9.

RNA: a chemical found in the nucleus and cytoplasm of cells; it plays an important role in
protein synthesis and other chemical activities of the cell. The structure of RNA is similar to that
of DNA. There  are several classes of RNA molecules, including messenger RNA, transfer RNA,
ribosomal RNA, and other small RNAs, each serving a different purpose2.

Signal transduction pathway:  the course by which a signal from outside a cell is converted to a
functional change within the cell1.

Single nucleotide polymorphism (SNP): a change in which a single base in the DNA differs
from the usual base at that position1.

-------
Page C-4                   Genomics Task Force White Paper

Susceptibility: the increased likelihood of an adverse effect, often discussed in terms of
relationship to a factor that can be used to describe a human subpopulation (e.g. life stage,
demographic feature, or genetic characteristic)6.

Susceptible Subgroups: may refer to life stages, for example, children or the elderly, or to other
segments of the population, for example, asthmatics or the immune-compromised, but are likely
to be somewhat chemical-specific and may not be consistently defined in all cases6.

Systems Biology:  a holistic approach to the study of biology with the objective of
simultaneously monitoring all biological processes operating as an integrated system10.

Systems Toxicology: the study of perturbation of organisms by chemicals and stressors,
monitoring changes in molecular expression and conventional toxicological parameters, and
iteratively integrating biological response data to describe the functioning organism11.

Throughput:  output or production, as of a computer program or a biological assay, over a
period of time1.

Toxicity:  deleterious or adverse  biological effects elicited by a chemical, physical, or biological
agent6.

Toxicology: the study of harmful interactions between chemical, physical, or biological agents
and biological systems6.

Toxicogenomics:  the study of how genomes respond to environmental  stressors or toxicants.
Combines genome-wide mRNA expression profiling with protein expression patterns using
bioinformatics to understand the role of gene-environment interactions in disease and
dysfunction2.

Transgenic: having genetic material (DNA) from another species1.

-------
                          Genomics Task Force White Paper                  Page C-5

                                Glossary References

1. U.S. Department of Health and Human Services, Public Health Service, National Institutes of
Health, National Institute of Environmental Health Sciences, National Center for
Toxicogenomics, Glossary. http://www.niehs.nih.gov/nct/glQssary.htm

2. U.S. Department of Energy, Oak Ridge National Laboratory, Human Genome Project
Information, Genome Glossary. http://www.oml.gov/sci/techresources/Human Genome/glossary

3. U.S. Environmental Protection Agency. A Framework for a Computational Toxicology
Research Program in ORD. EPA/600/R-03/065.
http://www.epa.gov/nheerl/comptoxframework/comptoxframeworkfinaldraft7  17  03.pdf

4. U.S. Department of Health and Human Services, Public Health Service, National Institutes of
Health, National Cancer Institute, Understanding Gene Testing, Glossary.
http://www.accessexcellence.org/AE/AEPC/NIH/gene27.html

5. U.S. Environmental Protection Agency, Science Policy Council. 2002a. Interim Policy on
Genomics. http://www.epa.gov/OSP/spc/genomics.pdf

6. U.S. Environmental Protection Agency. 1999. Glossary of Integrated Risk Information
System, http://www.epa.gov/iris/gloss8.htm

7. U.S. Department of Health and Human Services, Public Health Service, National Institutes of
Health, National Institute of General Medical Sciences, The Chemistry of Health Glossary
http://www.nigms.nih.gov/news/science_ed/chemhealth/glossary.html

8. Henry, C.J., Phillips, R., Carpanini F., Gorton, J.C., Craig, K., Igarashi, K., Leboeuf, R.,
Marchant, G., Osborn, K., Pennie, W.D., Smith, L.L., Teta, M.J., Vu, V. 2002. Use of genomics
in toxicology and epidemiology: findings and recommendations of a workshop. Environmental
Health Perspectives 110(10):1047-1050,

9. U.S. Environmental Protection Agency, Risk Assessment Web Page.
http://www.&pa.gov/ebtpages/enviriskassessment.html

10. Sumner, L.W., Mendes, P., Dixon, R.A. 2003. Plant metabolomics: large-scale
phytochemistry in the functional genomics era. Phytochemistry 62:817-836.

11. Waters, M., Boorman, G., Bushel, P., Cunningham, M., Irwin R., Merrick, A., Olden, K.,
Paules, R., Selkirk, I., Stasiewicz, S., Weis, B., Van Houten, B., Walker, N., Tennant, R. 2003.
Systems toxicology and the Chemicals Effects in Biological Systems (CEBS) knowledge base.
Environmental Health Perspectives Toxicogenomics 111(6): 811-824.

-------
SEPA
      United States
      Environmental Protection
      Agency

      Science Policy Council
      US Environmental Protection Agency
      Washington, DC 20460
      Official Business
      Penalty for Private Use
      $300

      EPA/iOO/B-04/002 "
      December 2004
PRESORTED STANDARl
 POSTAGE & FEES PAID
       EPA
   PERMIT No. G-35
                                        Recycled/ Recyclable
                                        Printed with vegetable-based ink on
                                        paper that contains a minimum of
                                        50% post-consumer fiber content
                                        processed chlorine free

-------