Genomics in environmental health research—opportunities and challenges
Department of Health and Human Services, National Institute of Environmental Health Sciences and The National Toxicology Program, National Institutes of Health, P.O. Box 12233, Research Triangle Park, NC 27709, USA
Abstract
Environmental health research impacts both environmental health regulatory policy and the practice of medicine. However, this area of medical research has not garnered public support and attention of medical researchers because of its emphasis on prevention and public health. Also, the pervasiveness of a scientific culture wedded to old problems and outdated technologies and models systems has not been helpful in generating enthusiasm for the field. While the emphasis on prevention is both laudable and appropriate, the adoption of cutting-edge technologies to exploit the new scientific opportunities, made possible by the nation’s investment in genomics, is essential if the discipline expects to be competitive with other highly deserving programs. The new ‘omics’ era of environmental health research, ushered in over the past decade, characterized by the linkage of genomics, proteomics and metabolomics to conventional toxicology and pathology databases, holds great promise for elucidating mechanisms of gene–environment interaction in human health and disease. These combined approaches will allow one to monitor multiple molecular events, pathways and interactive networks simultaneously—a requirement for elucidating toxic mechanisms. But, before embracing the ‘omics’ technologies as the ‘be all-end all;’ they need to be validated for their predictive capacities in large-scale multi-institutional studies, such as those described in this article.
Author Keywords: Toxicogenomics; ‘Omics’; Systems biology; Gene–environment interactions; Polymorphism; Susceptibility
1. Introduction
Presently, we lack adequate information to promulgate evidence-based environmental health regulatory policies and to prevent or cure most chronic diseases. This paucity of information has an enormous impact on the world’s economy both in terms of costs associated with health care and regulatory compliancy. In part, this situation exists because of our emphasis on genetics as the primary cause of human illness. While this still may turn out to be the case; nonetheless, it now appears that the development of most chronic diseases is the result of complex interactions involving genes, environmental factors, behavior and random stochastic events associated with endogenous metabolism. Age and stage of development are also important factors in disease or toxicity development, as they influence the timing of gene expression and susceptibility to environmental toxicants.
For years, the environment was considered to be a minor player in the etiology of human illness; in part, because only radiation, synthetic chemicals and industrial by-products were included in the definition. But, now that the definition of the environment has been expanded to include the diet, behavior and other social and cultural factors, the tide is shifting in favor of gene–environment interactions. Another contributing factor is that the promise of the human-genome project remains largely unfulfilled in terms of discovery of disease genes and development of new pharmaceuticals to prevent or treat diseases. In fact, studies are now being reported that ‘blow away’ the myth that ‘bad genes’ are responsible for the majority of human morbidity and mortality. For example, recent studies show that no more than one-third of the cancer burden can be attributed to the action of genes alone (Verkasala et al., 1999 and Lichtenstein et al., 2000), only 15% of Parkinson’s disease ( Tanner et al., 1999), and about a third of autoimmune diseases ( Powell et al., 1999). And, a recent study reported that 90% of individuals with severe heart disease have at least one or more of four classic risk factors captured in the current definition of the environment ( Khat et al., 2003). Because of these and other studies, it is now generally accepted that without more informative and cost-effective methods for assessing and predicting risk resulting from environmental exposures, we will not be able to prevent or cure most chronic diseases, and the costs associated with health care and environmental regulatory compliancy will continue to escalate.
Altering the current state-of-affairs will require investments in research to (i) identify the genetic, environmental, behavioral and age-or-stage-of-development-related risk factors; (ii) development of strategies to nullify or reduce the causative risk factor(s), and (iii) translation of new scientific knowledge and technologies into public health and medical practice. The achievement of these objectives will require research investments in five areas: (i) improved understanding of the factors that contribute to differences in susceptibility to environmental exposures; (ii) better understanding of differences and similarities in the biology of human versus rodent models to validate the current practice of extrapolation in human risk assessment; (iii) development of better technologies to quantify exposure to reduce reliance on the use of indirect surrogates; (iv) development of high-throughput, low-cost and more informative testing strategies to assess the intrinsic toxicity of drugs and environmental xenobiotics, and (v) elucidation of mechanisms, pathways and metabolic networks involved in the development of toxicity induced by chemical, physical or biological agents to which humans are exposed.
Until recently, untangling the ‘Gordian Knot’ of complex interactions between genes, the environment and behavior was beyond our capacity, technologically. But, thanks to investments in the human genome and other related projects, we now have the tools, databases and animal models to begin to untangle such complex biological problems.
2. Genetic susceptibility and toxicity
Over the years, biological systems have evolved buffering or protective mechanisms to prevent damage from environmental exposures. Unfortunately, most of the synthetic chemicals to which humans are now exposed have only been recently introduced into the environment; hence, they have not co-evolved with the human genome and their protein products. Therefore, protective or buffering mechanisms have not had adequate time to evolve. It is now well-established by research in toxicology and pharmacogenetics that polymorphisms (e.g. single base changes, insertions or deletions) exist in most, if not all, of the genes involved in biological processes such as the uptake, metabolism and excretion of drugs and environmental xenobiotics, DNA repair, cell cycle control, and membrane signaling; and that such genetic variations can modify an individual’s risk for disease by altering the amount or activity of the respective protein products (Nebert et al., 1996 and Costa et al., 2003). For example, the genes coding for most of the P450s are structurally and functionally polymorphic ( Nebert and McKinnon, 1994), and estimates are that 56% of existing pharmaceuticals are metabolized by this family of enzymes ( Ingelman et al., 1999); and the relative distribution of the variant alleles differs markedly between ethnic groups. Such variations are likely to be very important factors in determination of clinical efficacy and safety of a variety of pharmaceuticals, and the likelihood of adverse health outcomes resulting from environmental exposures.
With the reference sequences of the human and mouse genomes now completed, many researchers have now turned their attention to studying variation in specific gene sequences, and the possible consequence of such change on the phenotype of the host, with respect to predisposition to diseases and adverse health outcomes from environmental exposures. The approximately 0.1% variation reported for the human genome means that millions of differences exist among the 3.2 billion nucleotide base pairs. The first systematic, national effort to discover such polymorphisms and investigate their health effects was initiated by the National Institute of Environmental Health Sciences (NIEHS), with the announcement of the Environmental Genome Project (EGP) in 1997 (Kaiser, 1997; Brown and Hartwell, 1998; Guengerich, 1998 and Olden and Wilson, 2000). The discovery and functional characterization of polymorphisms, in DNA samples derived from 90 different individuals, is being conducted using a pre-selected set of 544 candidate genes. The candidate genes were selected by input from the scientific community on the basis of well-established genetic and biochemical studies which indicate that they play a major role in regulating risk posed by drugs and environmental xenobiotics. The benefit of the candidate gene approach, versus the non-discriminant discovery of single nucleotide polymorphisms (SNPs) in the entire genome, is that much information is already known about the structure and function of the specific protein products. To date, more than 20,000 SNPs have been discovered in 217 of the candidate genes involved in cell cycle control and DNA repair, including more than 1000 in coding regions ( Olden, 2004). The newly discovered SNPs can be found at the web site (http://www.sciencedirect.com/science?_ob=RedirectURL&_method=externObjLink&_locator=url&_cdi=5175&_plusSign=%2B&_targetURL=http%253A%252F%252Fwww.genome%252Futah.edu%252Fgensnps) established for this project. Also, to assist potential users, NIEHS has provided a graphical user interface, extensive gene annotation, and a suite of bioinformatics tools for functional analysis. Biochemical and genetic studies are now being done to determine which of the polymorphic alleles are functionally important and, eventually, population-based studies will be conducted to determine which are associated with specific diseases.
Since most human genes have counterparts in the mouse genome, one can build on the longstanding use of mice as surrogate models for humans in toxicology studies by constructing variations in specific genes, using knock-out and knock-in technologies, similar to those discovered in the candidate genes of the EGP. Such genetically-modified animals can be used to probe for specific health outcomes following exposure to environmental agents, based on our knowledge of gene function. So, to assist with the functional analysis in animal models, NIEHS developed the Comparative Mouse Genomics Centers Consortium in 2001. Investigators supported by this program are to construct mice with allelic variations in coding and regulatory sequences of homologous genes, and characterize them with respect to phenotype; and, upon request, provide them to investigators in the scientific community.
In addition to the development of mouse models with the polymorphic variations discovered in the EGP, NIEHS is developing a contract to obtain haplotype maps of 15 mouse strains commonly used as surrogate models in toxicology research. Such genetic variation maps, in combination with the polymorphism database derived from the EGP, should be useful resources for studies of gene–environment interactions. The discovery of the haplotype maps of several mouse strains is important because only one strain (i.e., the B6) was used in sequencing the genome, and significant variation exists between strains with respect to susceptibility to carcinogens. Also, environmental health regulators usually use data generated on the toxic effects of chemicals in a single strain to guide risk assessment decisions. This risk assessment approach is problematic because inbred laboratory animals are not representative of the genetic variability found in human populations. For example, we do not know how genetic background influences environmental-response genes. The generation of haplotype maps of various mouse strains will be useful in sorting out the potential influence of modifier genes.
3. Toxicogenomics
A major problem in assessing the contribution of the environment to human health and disease is the paucity of toxicity data on the majority of the high-production volume chemicals introduced into the environment in the last half of the 20th century. Furthermore, the toxicity studies that have been performed are, in most cases, less than satisfactory for use in assessing risk to human health. Also, the existing test systems are too costly and time-consuming to evaluate more than only a small fraction of the chemicals in need of toxicity assessment. But, perhaps the most serious concern about existing test systems is that they require the use of animals as model systems. Finally, good biomarkers of early molecular alterations that lead to toxicity and chronic disease, resulting from environmental exposures, do not exist. The early detection of such interactions involving genes, gene-products and environmental factors could lead to early detection and development of effective prevention strategies and elucidation of mechanisms of pathogenesis.
The development of toxicity or disease phenotype is a complex process involving many interactive molecules and pathways organized into networks. Therefore, current approaches based on pathological examination and biochemical analysis of a few proteins in a single pathway are not robust sufficiently enough to elucidate toxic mechanisms. To elucidate mechanisms of toxicity and pathogenesis of disease, one needs to know which genes, proteins and metabolites are involved and have insight about timing of their expression, quantity, activity and flux.
Because of these issues, a serious bottleneck exists in drug development and environmental health risk assessment is fraught with problems that leave decision-makers in the difficult or untenable position of regulating exposure to environmental agents without adequate information to be certain that environmental policies are protective of human health, or that they are even necessary. Fortunately, recent studies suggest that early interactions between genes, gene-products and environmental factors can be detected and monitored by analysis of gene and protein expression arrays (Neuvaysir et al., 1999; Afshari, 2002; Burchiel et al., 2001; Bushel et al., 2001; Olden, 2002 and Waters et al., 2003), possibly by using single cell organisms or in vitro tissue culture systems. Fundamentally, metabolic mechanisms are highly conserved, so by determining how single cell or tissues adapt to biochemical derangement from specific environmental exposures can be instructive in predicting toxicity or an adverse health outcome in humans. This new molecular approach to studying gene–environment interactions is referred to as toxicogenomics; it combines genomics, proteomics, metabolomics and informatics with conventional toxicology and pathology to analyze effects on thousands of genes and proteins simultaneously. This so-called systems toxicology approach is more relevant for studying complex events (e.g. the development of toxicity and disease) that result from interactive networks involving multiple genes, proteins and small molecules. This genome-wide screening approach will allow for the discovery of unique biomarkers and molecular mechanisms characteristic of specific diseases or dysfunctions; and will lead to the development and validation of test systems that are less costly, time-consuming and dependent on the use of animals, and that are more informative and relevant to human risk assessment. Also, by comparing the expression levels of RNA, protein and metabolites, one can determine whether specific polymorphisms are likely to increase or decrease the health risk to drugs or environmental xenobiotics because of alterations in the amount or activity of specific gene-products.
To promote the development and application of technologies for monitoring gene, protein, and metabolite expression, in response to specific environmental stressors, NIEHS developed the National Center for Toxicogenomics (NCT), a consortium consisting of six research institutions. Members of the consortium are now in the process of validation and standardization of methodologies and development of new and improved computational and analytical methods to analyze gene and protein structure and function and the production of metabolites.
Investigating gene, protein and metabolite expression, in response to specific environmental exposures on a genome-wide scale, is in order now that the genomes of humans and other vertebrate species have been sequenced, and various polymorphism discovery projects are underway. Whereas, information is rapidly accumulating about sequencing and mapping of genes, much less is known about their expression in individual cells and tissues and under various environmental conditions. By manipulating the environment, one can get cells or tissues to express specific genes, proteins and metabolites in response to the stress. The patterns of expression can be expected to vary in response to exposure to different environmental toxicants or stressors, due to disruption of different metabolic pathways or networks. Oligonucleotide and cDNA arrays are being used for analysis of relative expression and the microbead method for quantification of RNA transcripts, respectively.
4. Summary
Traditionally, toxicologists and environmental health scientists have relied on pathology, biochemistry, genetics and cell biology approaches to study one or a few genes or proteins in a single pathway. These approaches are too costly, time-consuming and woefully inadequate for analyzing more than a few events simultaneously. Furthermore, they typically do not address issues related to susceptibility, low dose exposures, and possible interactions between drugs and environmental xenobiotics due to co-exposure. Over the years, there has been a tendency to modify existing test systems to meet ever increasing requirements because investigators have not been very receptive of innovation in toxicity testing. These practices have increased the burden on both government and industry in terms of costs, animal use, and time required for safety assessment.
The grand goal of toxicology is to characterize the entire set of genes and proteins that are affected when humans are exposed to drugs or environmental xenobiotics. What has long been needed to achieve this objective is a systems biology approach for monitoring dynamic changes in activity and quantity of molecular constituents of cells and tissues in animals with known genetic backgrounds. This is especially true for assessment of drugs and environmental xenobiotics that exert their effects through multiple mechanisms, depending on dose, timing and duration of exposure, and cell or tissue type. Using ‘omics’ technologies, environmental health scientists can conduct large-scale studies of the effects of toxicants on gene expression at the mRNA and protein levels, while, simultaneously, monitoring metabolite profiles in body fluids and tissues to gain insight into the activity state of all relevant genes and gene-products. The approaches described here will identify biomarkers of toxicity, that is, molecules that can be directly associated with specific adverse events. Toxicogenomics data generated in multiple species will enhance our understanding of complex interactions between genes, the environment and human health, and can improve our ability to use laboratory animals to predict effects of chemicals in people.
While ‘omics’ technologies offer considerable promise, their potential usefulness must await the development of a knowledge base that will enable discrimination between adaptive or pharmacological responses and toxicological effects. Also, the various methodologies have to be validated before they will have routine application in toxicity assessment.
References
Afshari, C.A., 2002. Perspective: microarray technology, seeing more than spots. Endocrinology 14, pp. 1983–1989.
Brown, P.O. and Hartwell, L., 1998. Genomics and human diseases—variations on variation. Nat. Genet. 18, pp. 91–93.
Burchiel, S.W., Knoll, C.M., Davis, J.W. et al., 2001. Analysis of genetic and epigenetic mechanisms of toxicity: potential roles of toxicogenomics and proteomics in toxicology. Toxicol. Sci. 59, pp. 193–195.
Bushel, P.R., Hamadeh, H., Bennett, L. and Sieber, S., 2001. MAPS: a microarray project system for gene expression experiment information and data variation. Bioinformatics 17, pp. 564–565.
Costa, L.G., Cole, T.B., Jarvik, G.P. and Furlong, C.E., 2003. Functional genomics of the paraoxonase (PON1) polymorphisms: effects on pesticide sensitivity, cardiovascular disease, and drug metabolism. Annu. Rev. Med. 54, pp. 371–392.
Guengerich, F.P., 1998. The environmental genome project: functional analysis of polymorphisms. Environ. Health Perspect. 106, pp. 365–368.
Ingelman, M.S., Oscarson, M. and McLellan, R., 1999. Polymorphic human cytochrome P450 enzymes: an opportunity for individualized drug treatment. Trends Pharmacol. Sci. 20, pp. 342–348.
Kaiser, J., 1997. Environment institute lays plans for gene hunt. Science 278, pp. 569–570.
Khat, U.M., Khat, M.B., Bayzer, C.T. et al., 2003. Prevalence of conventional risk factors in patients with coronary heart disease. J. Am. Med. Assoc. 290, pp. 899–904.
Lichtenstein, P., Niels, V.H., Verkasala, P.K., Iliadou, A., Kapria, J. et al., 2000. Environmental and heritable factors in the causation of cancer. Analysis of cohorts of twins from Sweden, Denmark, and Finland. New Eng. J. Med. 343, pp. 78–85.
Nebert, D.W., McKinnon, R.A. and Puga, A., 1996. Human drug-metabolizing enzyme polymorphisms: effects on risk of toxicity and cancer. DNA Cell Biol. 15, pp. 273–280.
Nebert, D.W. and McKinnon, R.A., 1994. Cytochrome P450: evolution of functional diversity. Prog. Liver Dis. 12, pp. 63–97.
Neuvaysir, E.F., Bittner, M., Trent, J. et al., 1999. Microarrays and toxicology: the advent of toxicogenomics. Mol. Carcinog. 24, pp. 153–159.
Olden, K. and Wilson, S.H., 2000. Environmental health and genomics: visions and implications. Nat. Rev. Genet. 1, pp. 149–153.
Olden, K., 2004. Use of ‘Omic" approaches in unraveling mechanisms of gene environment interactions, Current Genomics 5, 1–6.
Olden, K., 2002. New opportunities in toxicology in the post-genome era. Drug Discovery Today 7, pp. 273–276.
Powell, J.J., Vande Water, J. and Gerschwin, M.E., 1999. Evidence for the role of environmental agents in the initiation or progression of autoimmune conditions. Environ. Health Pespect. 107 Suppl. 5, pp. 667–672.
Tanner, C.M., Ottman, R., Goldman, S.M. et al., 1999. Parkinson’s disease in twins: an etiologic study. J. Am. Med. Assoc. 281, pp. 341–346.
Verkasala, P.K., Kapria, J., Kaskenvua, M. and Pukkala, E., 1999. Genetic predisposition, environment and cancer incidence: a national twin study in Finland. Int. J. Cancer 83, pp. 743–749.
Waters, M., Boorman, G., Bushel, P., Cunningham, M. et al., 2003. Systems toxicology and the chemical effects on biological systems (CEBS) knowledge base. Environ. Health Perspect. Toxicogenom. 111, pp. 15–28.
Corresponding author. Tel.: +1-919-541-3201; fax: +1-919-541-2260.