Ancient gene variations at play today

Time and again we see evidence that genetic diversity is crucial for the success and survival of populations. In some cases, there are many different forms or versions, called alleles, of the same gene. Each does broadly the same job, but different variants of the gene might do that job in slightly different ways, or have additional functions alongside their main role. But, when it comes to humans, what no one knew was how far back in history those variations can be traced and what sorts of contributions they make to both health and ill-health. Now, as he explains to Chris Smith, by comparing genetic sequences from ancient human ancestors with biobank data, Omer Gokcumen, from the University at Buffalo, can tell us that in some cases these gene variants have been around for millions of years…

Omer - We thought that if we can find really, really old variation, that actually dates back not to our ancestors as modern humans, but to ancestors that are Homo erectus or even Homo habilis like millions of years ago. If there are variations that are being maintained this long, this cannot be a coincidence.

Chris - So something is artificially holding them in the population. So I mean, the example which we are always force fed when I went to medical school was "sickle cell anaemia". On the one hand, very bad for you. On the other hand, if you live in the right geography, it will protect you and perhaps save your life from malaria. So the price paid to keep that gene in circulation is defense against malaria.

Omer - That's a very similar idea. However, that is actually interesting because that particular allele is relatively recent because it's all about farming and human modification of the landscape that leads to these bigger populations facing the malaria problem. But I tried to identify similar types of alleles indeed, but much older.

Chris - How do you know that they're disease-linked first?

Omer - Okay. So we didn't actually know. We actually start from the scratch. We take all the variations in the human genome and ask which are the oldest ones. Then we actually go back and see if these really, really old ones happens to be associated with different diseases based on previous studies and previous databases. Yeah.

Chris - Did you delve into the Neanderthal genome then? Is that how you got back to the point where we've got a common ancestor with Neanderthal and so we're going back, you know, half a million years almost to find out what our genome is doing then?

Omer - That's exactly right. So we actually find these variations we share within Neanderthals because the ancestors to Neanderthals and modern humans happens to have those variations already in their genomes. This is one way to anchor in time when these variations evolved: it has to be more than 700,000 years old, and we find some of them with other methods, even older, like, you know, two, 3 million years old.

Chris - So you find these variations and you can find ones which appear to be conserved in the population. So what did you do next with that information?

Omer - The next step is to look if these variations are hitting important parts of the genome. We also look at these huge databases, including UK Biobank, this huge genomic database where people actually try to connect correlate genetic variation with known diseases and traits like height or diabetes or obesity. And then we actually looked if there's something particular about these ancient variations, and we find that there's indeed something particular they're different from the modern variants.

Chris - And your conclusion is what? That they've been held in the genome, held in the population, in the gene pool because under certain circumstances they confer some kind of survival advantage, which when things flip the other way, they might be deleterious, but they're there in hot large numbers because they saved the day previously?

Omer - That's exactly right. I think they saved the day sometimes multiple times, but occasionally they're bad for you. A good example would be this one deletion that we found that is 2 million years old. And if you have that deletion, you actually are much more prone to developing psoriasis. But at the same time, it seems that it is protecting you from skin infections, including potentially leprosy. So you have this like balance. If the infection rates are very high, then you don't really care about the autoimmunity that much because you don't wanna die from an infection. But if the you know, leprosy is not an issue, for example, then it becomes detrimental. You have other things like a growth hormone receptor happens to have two types. These two types goes all the way to one and a half million years. And it seems that one type is good in the times of starvation, but if there's starvation is not an issue, then it's not that great for you. So both types have kept in the population for a long time because on occasion we needed them.

Chris - How susceptible is your analysis to the problem of bottlenecking and, and founder effects, where if you get a very limited population for a while because times get really so bad, you might have an overrepresentation of the genes that help people under those circumstances, but won't you lose a whole heap of gene variants that might have helped them under other circumstances? So we're only seeing a sort of thin slice of the possibilities that were there or might have come since?

Omer - This is a wonderful question, and I think indeed we are finding only a smaller portion of what existed among our ancestors, and we may have lost amazing alleles that may have helped us because of all these bottlenecks and founder effects. But there's still, we find much one of, this is one of the findings of the paper, that we find these really old variants, much more than expected by chance, by these, you know, these bottlenecks and founder effects should have eliminated a lot of them. But they survived. The adaptation was strong enough that it was survived. So off the common variation, maybe 13% almost 15% of them are ancient according to our findings. And it is something that we all share.

Chris - To what extent are some of them convergences where because they're so useful, they've appeared, disappeared and reappeared again at different points in history?

Omer - Great question too! We eliminated all of the recurrent ones by mathematical and genomic techniques, like from our analysis.

Chris - So what really is your take home message then? I mean, apart from the fact that it, it does confirm to us what many suspected, but couldn't really explain very well, but now we've got some objective evidence for it. What's your take home message from, we've done this study and this is what we, we feel we now understand better?

Omer - I think that that if there's a single sentence that I have to utter it has to be that common variation affecting diseases are there for a reason that I don't think, like, you know, we can basically say, oh, like this is a disease allele and it's always bad. But I think from an evolutionary perspective, the things that remain in the population may actually have an adaptive reason for being here.