Review

Journal of Exposure Science and Environmental Epidemiology (2011) 21, 5–9; doi:10.1038/jes.2010.50; published online 17 November 2010

Implications of the exposome for exposure science

Stephen M Rappaporta,b

aSchool of Public Health, University of California, Berkeley, California, USA

Correspondence: Professor SM Rappaport, School of Public Health, University of California, MC 7360, Berkeley, CA 94720, USA. Tel.: +1 510 642 4355; Fax: +1 510 642 5817; E-mail: srappaport@berkeley.edu

bProfessor SM Rappaport was the 2010 recipient of the Jerome J. Wesolowski award for sustained and outstanding contributions to the knowledge and practice of human exposure assessment.

Top

Abstract

During the 1920s, the forerunners of exposure science collaborated with health professionals to investigate the causes of occupational diseases. With the birth of U.S. regulatory agencies in the 1970s, interest in the environmental origins of human diseases waned, and exposure scientists focused instead upon levels of selected contaminants in air and water. In fact, toxic chemicals enter the body not only from exogenous sources (air, water, diet, drugs, and radiation) but also from endogenous processes, including inflammation, lipid peroxidation, oxidative stress, existing diseases, infections, and gut flora. Thus, even though current evidence suggests that non-genetic factors contribute about 90% of the risks of chronic diseases, we have not explored the vast majority of human exposures that might initiate disease processes. The concept of the exposome, representing the totality of exposures received by a person during life, encompasses all sources of toxicants and, therefore, offers scientists an agnostic approach for investigating the environmental causes of chronic diseases. In this context, it is appropriate to regard the “environment” as the body's internal chemical environment and to define “exposures” as levels of biologically active chemicals in this internal environment. To explore the exposome, it makes sense to employ a top-down approach based upon biomonitoring (e.g. blood sampling) rather than a bottom-up approach that samples air, water, food, and so on. Because sources and levels of exposure change over time, exposomes can be constructed by analyzing toxicants in blood specimens obtained during critical stages of life. Initial investigations could use archived blood from prospective cohort studies to measure important classes of toxic chemicals, notably, reactive electrophiles, metals, metabolic products, hormone-like substances, and persistent organic compounds. The exposome offers health scientists an avenue for integrating research that is currently fractured along lines related to particular diseases and risk factors, and can thereby promote discovery of the key exposures responsible for chronic diseases. By embracing the exposome as its operational paradigm, exposure science can play a major role in discovering and mitigating these exposures.

Keywords:

exposome; internal chemical environment; disease risks

Top

Exposure science and the regulatory agenda

Many of the endeavors we associate with exposure science originated in the first half of the 20th century when health scientists and engineers collaborated to construct exposure–response relationships for occupational diseases. Methods for measuring airborne dusts and chemicals can be traced to surveys of mines and factories during the 1920s (see, e.g., (Greenburg and Smith, 1922; Greenburg, 1926)), and personal sampling was first applied in workplaces in 1960 (Sherwood and Greenhalgh, 1960). Early investigators sometimes employed both air and biological measurements to study tissue damage and toxicokinetics (see, e.g., (Smyth and Smyth, 1928; Roach, 1953, 1966)). Studies of urban air pollutants were initiated in the 1950s, and were extended to contaminants of water and indoor air in the 1970s. By the 1980s, practitioners of exposure assessment had largely become segregated according to occupational versus ambient sources of exposure. This division was motivated, in part, by U.S. laws from the 1970s that established the Occupational Safety and Health Administration (OSHA) to set and enforce exposure limits in the workplace and the Environmental Protection Agency (EPA) to assess risks and regulate contaminants in air and water.

With primary emphasis on chemicals that were OSHA and EPA priorities, the motivation for exposure science shifted away from human health and towards risk assessment and compliance. Tellingly, only 8% of recent papers published in the Journal of Exposure Science and Environmental Epidemiology (JESEE) involved studies of health effects (these statistics were kindly provided by Dana Boyd Barr, editor of JESEE, based upon the 491 papers accepted for publication from 2006 to 2010), and these focused primarily on chemicals of regulatory importance. As the vast majority of exposures involve chemicals that are not being scrutinized by OSHA and the EPA, this regulatory imperative implicitly discourages discovery of exposures that may be responsible for major chronic diseases (more on this shortly). Furthermore, because non-compliance with OSHA and EPA standards can lead to legal sanctions against economic interests, measurement of personal exposures has arguably been discouraged (Rappaport, 1984). Certainly, EPA standards do not mandate personal sampling and the numbers of measurements in workplaces have actually declined since about 1980, despite major technologic advances (Cherrie, 2003; Rappaport, 2009). Instead of measuring exposures, exposure scientists increasingly predict exposure levels from deterministic or probabilistic models based on observational data, spatiotemporal determinants, and/or stationary sampling of ambient air and water. This movement away from empirical measurements of exposures is also reflected by recent JESEE statistics, wherein fully 62% of all papers reported no exposure data (based upon the 491 papers accepted for publication from 2006 to 2010).

Returning to the major causes of human diseases, it is becoming apparent that genetic variability is not the major contributor. Studies of twins indicate that cancer risks attributable to genetic factors are probably about 10% (Lichtenstein et al., 2000), and results from more than 400 genome-wide association studies (GWAS) indicate that the heritability of degenerative diseases is also typically about 10% (Manolio et al., 2009; Hindorff et al., 2010). Thus, it seems that non-genetic, that is, ‘‘environmental’’, factors are typically responsible for about 90% of chronic disease risks. But what are the important environmental exposures? The simple answer is that we have only a sketchy understanding of the culprits because information about exposures is gleaned almost exclusively from indirect and uncertain questionnaire data. Using such data, the estimated proportion of all human diseases attributable to environmental pollutants (the sum of outdoor air pollution; indoor air pollution from solid fuel use, lead; water, sanitation and hygiene) plus occupational exposures is about 7–10% (Rodgers et al., 2004; Saracci and Vineis, 2007), roughly the same as attributable to genetic factors. Thus, traditional environmental pollutants appear to explain a small, but significant, part of the 90% of disease risks attributable to non-genetic factors. On the other hand, it has been estimated that modifying broad environmental descriptors (diet, overweight, inactivity and smoking) reduces risks of stroke, colon cancer, coronary heart disease and type 2 diabetes by 70–90% (Willett, 2002). The message for exposure science should be clear — if the field is to have a more prominent role in discovering and preventing the causes of chronic diseases, it must broaden its coverage of exposures beyond those of regulatory interest.

Top

The internal chemical environment

To reestablish human health as the primary focus of exposure science, we should recognize that the vast majority of toxic chemicals affect critical targets inside the body and, therefore, that the relevant ‘‘environment’’ is the body's internal chemical environment (Rappaport and Smith, 2010). Such an emphasis on the internal chemical environment would have three important consequences. First, it would promote the concept of dose that has been central to the related disciplines of pharmacology, toxicology and radiation biology (Gibaldi and Perrier, 1982; Ehrenberg et al., 1983). Second, it would encompass chemicals not only from air and water, but also from diet, smoking, drugs, radiation and important endogenous processes, including inflammation, stress, lipid peroxidation, infections, gut bacteria and pre-existing diseases (see Figure 1). This is crucial because blood levels of toxic chemicals from dietary and endogenous sources are generally much greater than those from polluted air and water (Ames, 1983; Dalle-Donne et al., 2006; Liebler, 2008), and important risk factors like psychological stress probably cause diseases by generating reactive chemicals and hormones inside the body (Epel et al., 2004). And finally, focusing on the internal chemical environment would encourage biological monitoring for assessing exposures. This is important because biomonitoring is virtually non-existent in U.S. workplaces and has been applied sparingly by the EPA. Indeed, only 16% of recent papers in JESEE reported biomonitoring data (based upon the 491 papers accepted for publication from 2006 to 2010). (Interestingly, many of these papers were authored by investigators at the Centers for Disease Control and Prevention (CDC), a U.S. governmental agency that has no regulatory authority).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Environmental exposures to chemicals arise from both external and internal sources. The exposome represents the combined exposures from all sources that reach the internal chemical environment. Note that radiation, stress, infections, behavior and lifestyle factors affect the internal chemical environment due to inflammation, oxidative stress, hormone production, and so on.

Full figure and legend (98K)

Although focusing upon the internal chemical environment would allow exposure scientists to discover causes of chronic diseases, it would not preclude them from hypothesis-driven avenues of inquiry. On the contrary, after identifying exposures that are likely contributors to disease risks, a host of questions would emerge concerning the identities, origins and, ultimately, the reduction of these exposures. To the extent that these newfound exposures arise from chemicals in air and water, monitoring of sources and risk assessment/management would proceed much as before. But more importantly, when exposures point to non-traditional sources, exposure scientists can contribute methods for characterizing and quantifying exposures that would supplant inaccurate and imprecise questionnaire-based approaches.

Top

The exposome

In contrasting the rich science spawned by the human genome project with the fragmented efforts to characterize environmental exposures, Christopher Wild defined the ‘‘exposome’’, representing the totality of exposures from conception onwards, as a quantity of critical interest for understanding the environmental causes of disease (Wild, 2005). As diseases are caused by chemicals that affect DNA and proteins inside the body, the exposome represents all biologically active chemicals inside a person during life. This creates complications because, although a person's genome is fixed at conception, his or her internal chemical environment varies throughout life because of changes in external and internal sources of exposure. Given such variability, characterizing the exposome requires longitudinal sampling, particularly during critical life stages, such as fetal development, early childhood, puberty and the reproductive years. That is, an exposome of a person would be re-constructed from a series of environmental or biological samples in almost the same manner that a collection of key scenes in a 30-s movie trailer captures the essence of a 2-h film. Exposome variability also extends across populations residing in different places with varying diets and lifestyles. Such differences explain why immigrant populations adopt disease patterns of their host countries and why distinct regions of the United States have different rates of the same diseases (Willett, 2002).

Exposure scientists can adopt either ‘‘bottom-up’’ or ‘‘top-down’’ strategies for characterizing exposomes, (Rappaport and Smith, 2010). As shown in Figure 2 (left), a bottom-up strategy would involve sampling of air, water and dietary sources of exposure, followed by quantitation of chemicals in these samples. Although this approach has the advantage of tying exposures directly to their immediate sources (e.g., air), it would require enormous effort, and would miss essential features of the internal chemical environment caused by gender, stress and endogenous processes. In contrast, a top-down strategy (Figure 2, right) would employ biological monitoring of subjects, most appropriately by blood sampling. This approach would cover both exogenous and endogenous chemicals in the internal chemical environment, but would not pinpoint particular sources. A top-down characterization of the exposome would utilize repeated blood samples collected during the life of a person, and could be tested for proof of concept with archived blood or serum from prospective cohort studies (Collins, 2004; Ollier et al., 2005). As future analytical and information technologies merge, it should be possible to acquire snapshots of exposomes during routine medical visits, consistent with the notion that medicine will become increasingly personalized, predictive and preventive (Hood et al., 2004; Collins, 2010).

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Two strategies for characterizing exposomes. With bottom-up exposomics (left) chemicals would be measured in air, water and food. This would identify potentially important exogenous exposures and their sources, but would miss endogenous exposures. With top-down exposomics (right) chemicals would be measured in blood. This would identify all potentially important exposures, but would provide no information about sources.

Full figure and legend (145K)

Top

Moving forward

As it is not currently feasible to measure all chemicals in the blood, exposure scientists should focus initially on classes of substances that are known to be biologically active, especially reactive electrophiles, hormone-like substances, persistent organic compounds, metals and metabolites (Rappaport and Smith, 2010). The CDC has made a good start in this context, having developed and applied assays to detect 200–300 candidate exposures in small quantities of blood or urine, including many metals, hormone-like substances and persistent organic compounds (CDC, 2009). In fact, Patel et al. (2010) recently used publically accessible CDC data from the National Health and Nutrition Examination Survey (NHANES) to investigate possible health effects of 266 candidate exposures. Strong associations were reported between the risk of type 2 diabetes and exposures to heptachlor epoxide, γ-tocophperol, β-carotenes and polychlorinated biphenyls, with effect sizes comparable to the strongest loci ever reported in GWAS. Although such candidate exposures offer tantalizing glimpses of the exposome and its potential importance, they are biased in favor of particular chemicals and classes from exogenous sources. Thus, it is important to apply agnostic approaches to detect a larger cross-section of toxic chemicals arising from all sources (Rappaport and Smith, 2010). Fortunately, some ‘‘omics’’ methods offer unbiased means for characterizing exposures to several important classes of toxicants, namely, metals (metallomics) (Mounicou et al., 2009), small metabolic products (metabolomics) (Dunn et al., 2008) and reactive electrophiles (adductomics) (Rubino et al., 2009; Li et al., in press).

It might seem ambitious to harness the exposome concept as a means of discovering the most health-impairing exposures that plague mankind. But the hurdles to making such discoveries are more conceptual than technical, and we need look no further than the human genome project for inspiration (Collins et al., 2003). At the beginning of the genome project, around 1990, DNA sequencing relied upon automated slab gels that operated at 50kb per day per sequencer. Today, short-read DNA sequencers operate at 100 million kilobases per day, a 2 million-fold increase over the slab technology (Stratton et al., 2009). To effectively characterize exposomes, we will need to measure a few thousand chemicals in repeated blood or serum samples archived from case and control subjects, nested within large cohort studies. As noted earlier, prospective cohort studies are currently being organized and could archive the repeated blood samples needed for such analyses (Collins, 2004; Ollier et al., 2005). Also, with proper financing, teams of scientists and engineers could harness state-of-the-art mass spectrometry, gene and protein chips and microfluidic lab-on-a-chip systems to satisfy the analytical requirements for performing thousands of assays. And once the analytical platforms have been developed, economies of scale should follow naturally, as they did from the genome project (Collins et al., 2003). As exposome technologies would also encourage therapeutic interventions and personalized medicine, they could well motivate government–private partnerships to develop and commercialize them.

Finally, it is worth pondering the milieu for exposome research as well as the training that future exposure scientists will require to be important contributors. As the internal chemical environment drives disease processes, a set of common chemical stressors is likely to arise from multiple sources and probably contributes to different disease endpoints. Thus, we should integrate research that is now fractured along lines related to particular diseases (cancer, cardiovascular disease, diabetes, etc.) and environmental risk factors (infections, air and water, occupation, nutrition and psychosocial). Failure to do so will continue to deprive science of the teamwork and serendipity that often leads to major breakthroughs and also dooms us to miss the general causes and mechanisms of chronic diseases. Exposure science is well positioned to have a key role in this research if the field can embrace the internal chemical environment as its major focal point. As current graduate training programs emphasize external sources, primarily air and water pollution, and put a premium on regulatory concerns, this will require a substantial change of direction. In particular, students of exposure science will need better grounding in biological systems, analytical chemistry and biomonitoring. Future exposure scientists must collaborate with epidemiologists, chemists, molecular and systems biologists, and bioinformaticists to acquire and interpret exposome data from appropriate biospecimens. As epidemiologists, who seek to discover the major causes of disease, would benefit greatly from unbiased measures of environmental exposure (Vineis et al., 2009), their support in recalibrating the mission of exposure science would be most welcome.

The author appreciates the provocative nature of these remarks and hopes this commentary will stimulate constructive discussions both within the exposure science community and across disciplines that seek to discover the causes of chronic diseases.

Addendum

Responding to the need for better tools to characterize environmental exposures, the National Institute for Environmental Health Sciences (NIEHS) requested that the National Academy of Sciences convene a workshop to discuss the exposome and its implications for understanding the causes of human diseases. The workshop (The Exposome: A Powerful Approach for Evaluating Environmental Exposures and their Influences on Human Disease, February 25–26, 2010, Washington, D.C., USA), brought together epidemiologists, exposure scientists, analytical chemists, molecular biologists and other scientists from academia, government, non-governmental organizations and the private sector. It motivated lively discussions regarding possible technologies and resources that might be used to develop and validate exposomic methods (available at: http://www.dels.nas.edu/envirohealth/exposome.shtml). Although the author of this commentary helped to organize, and participated in, the workshop, these remarks reflect solely his views.

Top

Conflict of interest

The author declares no conflict of interest.

Top

References

  1. Ames B.N. Dietary carcinogens and anticarcinogens. Oxygen radicals and degenerative diseases. Science 1983: 221(4617): 1256–1264. | Article | PubMed | ISI | ChemPort |
  2. CDC. Fourth National Report on Human Exposure to Environmental Chemicals. National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA, 2009.
  3. Cherrie J.W. The beginning of the science underpinning occupational hygiene. Ann Occup Hyg 2003: 47(3): 179–185. | Article | PubMed
  4. Collins F. Has the revolution arrived? Nature 2010: 464(7289): 674–675. | Article | PubMed | ChemPort |
  5. Collins F.S., Morgan M., and Patrinos A. The Human Genome Project: lessons from large-scale biology. Science 2003: 300(5617): 286–290. | Article | PubMed | ISI | ChemPort |
  6. Collins F.S. The case for a US prospective cohort study of genes and environment. Nature 2004: 429(6990): 475–477. | Article | PubMed | ISI | ChemPort |
  7. Dalle-Donne I., Rossi R., Colombo R., Giustarini D., and Milzani A. Biomarkers of oxidative damage in human disease. Clin Chem 2006: 52(4): 601–623. | Article | PubMed
  8. Dunn W.B., Broadhurst D., Ellis D.I., Brown M., Halsall A., and O’Hagan S., et al. A GC-TOF-MS study of the stability of serum and urine metabolomes during the UK Biobank sample collection and preparation protocols. Int J Epidemiol 2008: 37(Suppl 1): i23–i30. | Article | PubMed
  9. Ehrenberg L., Moustacchi E., and Osterman Golkar S. Dosimetry of genotoxic agents and dose-response relationships of their effects. Mut Res 1983: 123: 121–182.
  10. Epel E.S., Blackburn E.H., Lin J., Dhabhar F.S., Adler N.E., and Morrow J.D., et al. Accelerated telomere shortening in response to life stress. Proc Natl Acad Sci USA 2004: 101(49): 17312–17315. | Article | PubMed | ChemPort |
  11. Gibaldi M., and Perrier D. Pharmacokinetics, Second Edition, Revised and Expanded. Marcel Dekker, New York, 1982.
  12. Greenburg L. Benzol poisoning as an industrial hazard. US Pub Health Repts 1926: 41: 1516–1539.
  13. Greenburg L.W., and Smith G.W. A new instrument for sampling aerial dust. US Bur Mines Rept Invest 1922, 2392.
  14. Hindorff L.A., Junkins H.A., Hall P.N., Mehta J.P., and Manolio T.A. A Catalog of Published Genome-Wide Association Studies. www.genome.gov/gwastudiesAccessed Sep. 10, 2010.
  15. Hood L., Heath J.R., Phelps M.E., and Lin B. Systems biology and new technologies enable predictive and preventative medicine. Science 2004: 306(5696): 640–643. | Article | PubMed | ISI | ChemPort |
  16. Li H., Grigoryan H., Funk W.E., Lu S.S., Rose S., and Williams E.R., et al. Adductomics of Human Serum Albumin by Fixed-Step Selected Reaction Monitoring. Mol Cell Proteomics, in press.
  17. Lichtenstein P., Holm N.V., Verkasalo P.K., Iliadou A., Kaprio J., and Koskenvuo M., et al. Environmental and heritable factors in the causation of cancer--analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 2000: 343(2): 78–85. | Article | PubMed | ISI | ChemPort |
  18. Liebler D.C. Protein damage by reactive electrophiles: targets and consequences. Chem Res Toxicol 2008: 21(1): 117–128. | Article | PubMed
  19. Manolio T.A., Collins F.S., Cox N.J., Goldstein D.B., Hindorff L.A., and Hunter D.J., et al. Finding the missing heritability of complex diseases. Nature 2009: 461(7265): 747–753. | Article | PubMed | ChemPort |
  20. Mounicou S., Szpunar J., and Lobinski R. Metallomics: the concept and methodology. Chem Soc Rev 2009: 38(4): 1119–1138. | Article | PubMed
  21. Ollier W., Sprosen T., and Peakman T. UK Biobank: from concept to reality. Pharmacogenomics 2005: 6(6): 639–646. | Article | PubMed
  22. Patel C.J., Bhattacharya J., and Butte A.J. An Environment-Wide Association study (EWAS) on type 2 diabetes mellitus. PLoS One 2010: 5(5): e10746. | Article | PubMed
  23. Rappaport S.M. The rules of the game: an analysis of OSHA's enforcement strategy. Am J Ind Med 1984: 6(4): 291–303. | Article | PubMed
  24. Rappaport S.M. Assessing workplace exposures: turning to the past for guidance. Occup Environ Med 2009: 66(7): 429–430. | Article | PubMed
  25. Rappaport S.M., and Smith M.T. Environment and disease risks. Science 2010: 330: 460–461. | Article | PubMed
  26. Roach S.A. A method of relating the incidence of pneumoconiosis to airborne dust exposure. Brit J Ind Med 1953: 10: 220–226.
  27. Roach S.A. A more rational basis for air-sampling programs. Am Ind Hyg Assoc J 1966: 27: 1–12. | PubMed |
  28. Rodgers A., Ezzati M., Vander Hoorn S., Lopez A.D., Lin R.B., and Murray C.J. Distribution of major health risks: findings from the Global Burden of Disease study. PLoS Med 2004: 1(1): e27. | Article | PubMed
  29. Rubino F.M., Pitton M., Di Fabio D., and Colombi A. Toward an “omic” physiopathology of reactive chemicals: Thirty years of mass spectrometric study of the protein adducts with endogenous and xenobiotic compounds. Mass Spectrom Rev 2009: 28(5): 725–784. | Article | PubMed
  30. Saracci R., and Vineis P. Disease proportions attributable to environment. Environ Health 2007: 6: 38. | Article | PubMed
  31. Sherwood R.J., and Greenhalgh D.M.S. A personal air sampler. Ann Occup Hyg 1960: 2: 127–132. | PubMed |
  32. Smyth H.F., and Smyth Jr H.F. Spray painting hazards as determined by the Pennsylvania and the National Safety Council surveys. J Ind Hyg 1928: 10(6): 163–214.
  33. Stratton M.R., Campbell P.J., and Futreal P.A. The cancer genome. Nature 2009: 458(7239): 719–724. | Article | PubMed | ChemPort |
  34. Vineis P., Khan A.E., Vlaanderen J., and Vermeulen R. The impact of new research technologies on our understanding of environmental causes of disease: the concept of clinical vulnerability. Environ Health 2009: 8: 54. | Article | PubMed
  35. Wild C.P. Complementing the genome with an “exposome”: the outstanding challenge of environmental exposure measurement in molecular epidemiology. Cancer Epidemiol Biomarkers Prev 2005: 14(8): 1847–1850. | Article | PubMed | ChemPort |
  36. Willett W.C. Balancing life-style and genomics research for disease prevention. Science 2002: 296(5568): 695–698. | Article | PubMed | ISI | ChemPort |
Top

Acknowledgements

I am grateful for the assistance of Dana Boyd Barr in compiling the JESEE statistics mentioned in this commentary. I also appreciate helpful discussions with Martyn Smith, Christopher Wild, Thomas McKone, and Tina Bahadori. The author acknowledges support through Grant U54ES016115 from the National Institute for Environmental Health Sciences of the National Institutes of Health.