Suzi Gage - Linking genes to lifestyle

Suzi Gage and her colleagues at Bristol University are unpicking how lifestyle factors interact with genes to increase the risk of diseases.
13 February 2016

Interview with 

Suzi Gage, University of Bristol


Kat - With all this talk of engineering the genome, it's important to remember that we really don't understand much at all about how our genes work and what they do. Every year scientists find hundreds if not thousands of associations between tiny variations in our DNA and the risk of all kinds of diseases and disorders from cancer and heart disease to diabetes and depression through Genome Wide Association Studies, usually known as GWAS. Unpicking exactly how these genetic variants work is a complex task, and a new paper from Suzi Gage and her colleagues at Bristol University has just thrown some more complexity into the mix - the impact of genes on lifestyle factors. I asked her to explain what she's found.

Suzi - The sort of rationale behind this paper is just to get a better handle on what the results of genome wide association studies actually mean, what they can actually tell us. Because it's all well and good knowing that certain genetic variants are associated with a disease or with a phenotype of any kind. But until you actually can sort of delve a bit deeper and work out what that association actually means that's only of have limited interest really.

Kat - So, we hear these stories in the papers. Ten new genes for cancer found, hundred new genes for autism. Is that this kind of study that's throwing up these genes and we don't really know what they are and what they do?

Suzi - Yeah, absolutely. So, in some cases, you do know what a particular genetic variant does in terms of coding for a protein or something like that where you can actually test it. But more often than not actually, we don't know at the moment and so, the question is, how do we know what these variants are actually doing? Is it some sort of biological effect that they're having on the disorder - on cancer for example - or is it actually that they're having an effect on something else and that something else might be having the effect on cancer?

Kat - So, unpack that a bit. What do you mean by that? What examples would apply here?

Suzi - The reason that we became interested in this is that we noticed that this genetic variant that is quite robustly associated with how heavy a smoker you are, if you are a smoker, this genetic variant was actually seen in a genome wide association study of lung cancer. So, that could mean that there's just some sort of shared genetic architecture - so this gene might have an effect on both smoking and your risk of lung cancer. But what seems a far more parsimonious explanation of this is actually that the reason that this genotype or this genetic variant is coming out of the lung cancer GWAS is because smoking causes lung cancer. So genetic variants that predict smoking are going to be seen in GWASs of lung cancer.

Kat - Basically, if someone has this genetic variation and they are a smoker, they're going to want to smoke loads and that's what's going to be more likely to give them lung cancer.

Suzi - Exactly and that increases their risk of lung cancer. Absolutely, yes.

Kat - So, are there any other genetic variations that might be working in a similar way?

Suzi - Quite possibly, yes. At the moment, the sort of main thing holding us back is actually, we don't know the specific genotypes that are associated with these kind of modifiable exposures. There's a good example in the alcohol literature. There's a genetic variant that quite strongly predicts alcohol consumption. It's not very common in European populations but a bit more common in samples of east Asian ancestry. If you look in these populations then genome wide association studies of high blood pressure identified this particular alcohol-related genotype. But it's not seen in GWASs of high blood pressure in European populations which is pretty strong or certainly fairly strong evidence that it might be alcohol causing high blood pressure. And that's why you only see this particular genotype in these GWASs in these specific populations.

Kat - Do you think that there might be other ones out there? We've done cigarettes and alcohol. Any other vices that might be linked to conditions?

Suzi - Well, absolutely. Quite a lot of substance use is known to be heritable-  so for example, cannabis use. But as yet, we haven't identified the specific genetic variants that are associated with cannabis use. But as and when these variants were identified, it would be perfectly plausible that we might see these genetic variants coming out of GWASs for diseases. So, if there is a causal link between cannabis and schizophrenia. We identified this cannabis risk-increasing genetic variants. Then if there's truly a causal association between cannabis and schizophrenia, you might well expect that these genetic variants would come out in GWASs of schizophrenia.

Kat - So, what do we do with this information? I always find it interesting with these genetic studies whether you sort of say, "Oh well, you've got the gene variation that makes you more likely to be a stoner. Well, what do we do?" Or you're going to smoke more so, bad luck? Where do we take this work?

Suzi - In terms of what these studies might actually tell us, when we're looking at genome wide association of diseases, if a genetic variant comes out of one of these studies, that's actually telling us that there's a modifiable risk factor that's increasing the risk of this disease then that's a much easier place to target and intervention at than a biological mechanism.

Kat - So, it's easier to help someone stop smoking say, than it is to find and develop a drug that targets a molecular pathway that might reduce their lung cancer risk.

Suzi - Absolutely, yeah.

Kat - But don't we already know that people should give up smoking?

Suzi - Well, I mean, that is a very good point. I think being able to sort of say a bit more definitively than what we're seeing are causal associations here can only help in terms of getting that message across.

Kat - There's also a lot of talk about the nature versus nurture or nature with nurture and I guess what you are unpicking here is how our genes interact with the things that we do, the things in our lifestyle. It feels like this is really first steps into that world - that we still don't really understand much about how our lifestyle and environment affects our genes and what happens to us?

Suzi - Yeah. I think that's the key point here that we have to be very cautious when we kind of interpret these studies because we're really just at the very beginning of being able to understand what these genetic associations might actually mean and how we can then harness them to improve people's health and improve our own understanding of humans, our behaviour, and our health.

Kat - Bristol University's Suzi Gage, and her paper was published this week in the journal PLoS Genetics. 


Add a comment