There’s a lot of emphasis on genome-wide association studies nowadays, but the genomic contribution to disease — or any other phenotype, really — is only about half the story. The problem is that GWAS provide a fairly straightforward measurement, while it’s a monumental task to figure out whether and how much a host of environmental factors contribute to a phenotype: there’s the questionable applicability of animal models, variable individual response, a truly massive number of potential daily exposures, and the technical difficulties of testing tissue samples for particular compounds. But epigenetics has the potential to change all that.
Particular epigenetic changes might be able to serve as useful biomarkers that incorporate environmental effects, such as exposure to endocrine disruptors, or famine. It’s my new favorite idea, and I’m sure it’s not new to everyone. Now, it’s wishful thinking until we understand how epigenetic marks actually work. But maybe it’ll help placate folks like Jonathan Latham and GeneWatch UK, who’ve been loudly complaining about the amount of research money spent on GWAS, the Human Genome Project, and various pharmacogenomic efforts.
There’s an interesting but brief discussion of this epigenetic biomarker subject in the recent “Minireview: Epigenetics of Obesity and Diabetes in Humans” by Howard Slomo and colleagues at Yeshiva University’s Albert Einstein College of Medicine. In January’s Endocrinology, they phrase it this way:
For assessment of the cumulative effects of environmental exposures and prediction of individual disease susceptibility, epigenetic-based assays offer significant advantages as biomarkers if the challenges of study design, validation, and interpretation can be overcome.
To be clear, we’re not talking about cataloging individual environmental contributions, but a diversity of environmental contributions, as well as the effects of an individual’s genes, and interactions between an individual and the environment. But that lack of specificity is not necessarily a weakness — and it’s probably something that further epigenetic understanding will clear up.
Such a study might involve testing a subject group’s tissue samples for methylation in particular genomic regions, measuring the expression of particular genes, measuring levels of endocrine disrupting compounds, and recording cancer prevalence. And I think it has the potential to put the study of environmental contribution on an equal footing with GWAS. As it is, GWAS have a big advantage, as Slomo, et. al point out:
Although many of the loci identified [by GWAS] are low-penetrant genes with low relative risk that may not be clinically relevant, results from GWAS are generally more highly regarded then environmental exposure studies revealing similar or even slightly higher relative risks. This has been attributed to the lower error rate of genotyping techniques and the low reproducibility of environmental exposure assessment in human populations (45).
I don’t think critics like Latham fully acknowledge these drawbacks. With the traditional study of environmental contributions, you’ve got a much more complicated — and therefore expensive — undertaking. But they do hold the potential for equally important results.
This is easy to understand with a quick look at this 2010 PLoS paper, “An Environment-Wide Association Study (EWAS) on Type 2 Diabetes Mellitus,” by Stanford University School of Medicine researcher Chirag Patel and colleagues. They compared diabetes status and 266 environmental factors in people who participated in the CDC’s National Health and Nutrition Examination Survey. (Depending on the factor, they studied this diabetes association in 507 to 3318 people.) The conclusion:
Despite difficulty in ascertaining causality, the potential for novel factors of large effect [to be] associated with T2D justif[ies] the use of EWAS to create hypotheses regarding the broad contribution of the environment to disease. Even in this study based on prior collected epidemiological measures, environmental factors can be found with effect sizes comparable to the best loci yet found by GWAS. [Emphasis mine.]
One of the big problems, of course, is that there are far more environmental factors than have ever been measured by all the epidemiological studies of heaven and earth. But it’s possible that epigenomic studies could account for some of these unmeasured contributions.
I have no way of really comparing the amount of grant money devoted to traditional genetic association studies, GWAS, and environmental effect studies — not without devoting all my time to it, anyway. But as a rough gauge of the size of the pie, the NIH has awarded at least $2,306,222,186 since 2007 to studies that mention “genome-wide association” or “GWAS” in their titles and abstracts, according to the agency’s RePORTER database. (I eliminated known sub-studies.)
Now that’s walking-around money. With deeper understanding and new, powerful tools, I think we’ll get a much clearer understanding of environmental effects — and maybe figure out a little of that “lost heritability” along the way.
[The trippy picture of filthy lucre is "Money tunnel" by Flickr user Ramberg MediaImages. Used here under a Creative Commons license.]
Slomko, H., Heo, H., & Einstein, F. (2012). Minireview: Epigenetics of Obesity and Diabetes in Humans Endocrinology DOI: 10.1210/en.2011-1759