Any geneticist who has discussed genes with friends and family knows that there are a lot of misconceptions floating around out there. This is understandable – genetics involves some tricky concepts, and sometimes we use confusing linguistic shortcuts to talk about genes without using jargon. Sometimes scientists get confused as well (although that’s a topic for another day).
One of the big misconceptions is that there are genes ‘for’ specific traits — you’ll often hear that we have a gene ‘for’ X, X being some phenotypic trait. (If X is not a trait, but some sort of molecular player, than the language is correct, e.g. we do in fact have a gene — actually multiple genes — for the enzyme alcohol dehydrogenase.) This language is often rightly condemned as being misleading, because there is no one-to-one correspondence between genes and traits: traits are the product of multiple genes, while any given gene will contribute to multiple traits.
But you only need to tweak the ‘gene for X‘ language slightly to get at a correct and important concept in genetics: variation in a single gene is often responsible for important differences in X (in a particular population). This is usually what we mean when we say there is a ‘gene for X‘, but this clarification is rarely noted when people knock the phrase.
That’s not surprising, because genetic variation has been almost completely ignored by large numbers of biologist, including me. During graduate school I worked in yeast, and when I thought about genes ‘for’ something I was thinking about which genes contributed to a particular pathway. A gene ‘for’ X meant that, if you knocked out the gene, then pathway X would be defective. We weren’t asking questions about variation, and so it never came up on my radar. I had a minor epiphany when I joined a lab that was interested in knowing what genes carry the genetic variation responsible for major variation in X, and not just what genes are required for X. Those often are not the genes with the biggest knockout phenotypes.
In my Pacific Standard column this week, I take up another ignored concept – gene by environment interactions. The public understands that both genes and environment contribute to who we are. But how does this play out in real situations?
Too often, we talk about the genetic component and the environmental component as if they acted independently. However, any geneticist will tell you that gene by environment interactions are everywhere — though usually they are extremely hard to measure.
A study out this month in PNAS takes an interesting approach to detecting gene by environment interactions in the human obesity epidemic. Obesity clearly has an environmental factor, and yet there is a very common genetic variant in a gene called FTO that is linked with an increased obesity risk. What is the relationship between these genetic risk factors and environmental ones?
The authors took several decades of Body Mass Index (BMI) data from thousands of subjects in the Framingham Offspring study, and analyzed them by genotype. They were testing the hypothesis that some sort of global environmental changes over the past half century have altered the effect of the risky FTO allele — in other words, there was a gene by environment interaction.
Their study population allowed them to control for the well-known effect of age on BMI, and so they were able to look at the effect of birth cohorts on the risk posed by the risky FTO allel. What they found was that people who were born after 1942 and were heterozygous for the risky FTO allele were more likely to be obese than those hets born before 1942. Furthermore, they did not see such an effect for those who were homozygous for the non-risky FTO allele.
In other words, there is a clear gene by environment interaction: the genetic obesity risk posed by this one allele has changed with time, presumably as a consequence of our changing lifestyles. The study population has some limitations that the authors acknowledge, and the effect obviously needs to be tested in other populations. But it’s a neat result, a clever illustration of an important concept in human genetics that we often forget about.