Genes Are Overrated
When explaining our differences, the social sciences usually distinguish between heredity and environment – nature versus nurture. That is, our traits may be innate (something in our genes), or they may be learned (something that has been shaped by our experiences). For example, if you had been adopted by a purple-eyed family in Bulgaria as an infant, that wouldn’t change your own eye colour. But it would mean that you would be speaking Bulgarian today, not English. That’s because eye colour is genetically determined while language is environmentally determined.
Though this clear distinction works for a few traits, it is somewhat artificial. The vast majority of our traits are due to both genetics and environment. Take your personality. You have some natural inclinations – maybe you are a little temperamental or shy, for example. But it can get much better (or worse) depending on what kind of upbringing you had and what environment you find yourself in.
In the same way, we would expect our health and longevity to be affected by both genes and environment. If we want to learn about ageing and find ways to combat it, we should try to untangle the contribution of each.
The most commonly used method for studying genes versus environment is twin studies. Here, scientists take advantage of a gift from nature: the fact that identical twins have the same DNA. They are like genetic clones. You see, normally, after a sperm cell fertilises an egg, the fertilised egg will develop into a single person. However, sometimes, there can be a split during early cellular divisions. When this happens, the fertilised egg ends up becoming two people instead of one – both of them made from the same genetic blueprint.
By contrast, fraternal twins do not have the same DNA. They come from two different eggs that have each been fertilised by different sperm cells. As a result, fraternal twins are no more closely related than normal siblings, sharing fifty per cent of their DNA.
This key difference between identical and fraternal twins can be used to examine how important genes are when it comes to various traits.
Both sets of twins grow up in similar environments, but they are not equally related, because identical twins share twice as much DNA as fraternal twins. If the identical twins are more similar than the fraternal twins when it comes to a particular trait, that is a sign that genes are important for that trait.
An interesting example of a twin study is the Minnesota Twin Study, which followed identical and fraternal twins who were adopted by different families and therefore grew up apart. The researchers expected the identical twins to differ a lot when raised apart, but were surprised by just how similarly they turned out. If you met these people, you’d probably have guessed they grew up together even though they had never met.
Nancy Segal, one of the researchers behind the study, has used the identical twins James Lewis and Jim Springer as an example. The two first met when they were in their forties, but until then they had lived strangely similar lives: they regularly went on vacation to the same beach in Florida; they both bit their nails, drove light blue Chevrolets, suffered from similar types of headaches; and had both worked part-time at a sheriff’s office and at McDonald’s. One twin named his son James Alan, while the other named his son James Allan. The similarities actually extended to the absurd. Both twins first married women named Linda, then divorced the Lindas and later married women named Betty. Finally, one twin divorced his Betty, so maybe the other Betty should be worried.
Now, of course, the name of your wife isn’t coded in your genes. But the two brothers are evidence of just how much our genetics can influence our traits. So how about our lifespans?
One of the most prominent studies of twins and longevity was made on Danish twins born between 1870 and 1900. Here, researchers found a so-called ‘heritability’ for longevity of 0.26 for men and 0.23 for women. Similar results have been found in other studies: 0.25 among Amish people, 0.15 in the state of Utah and 0.33 in Sweden. The exact number isn’t that important. The important thing is that the heritability is low: closer to 0 than 1.
Heritability is a somewhat technical concept, but you can understand it like this: if the heritability of a trait is 1, that means that all the differences between individuals are due to their genes. For instance, if the heritability of height was 1 and one person is taller than another, it means that the height difference is solely due to genetic differences between the two. If the heritability of height was 0, the difference would be solely due to the environment. So when the heritability for lifespan is 0.15–0.33, it shows that a large part of the variation in lifespan is due to something other than our genes.
Researchers are still doing twin studies, but they have also begun employing new study designs to untangle genes and environment. For instance, Google-owned Calico (California Life Company), has conducted a study in collaboration with Ancestry.com, which hosts more than 100 million family trees. These family trees include enormous amounts of data on the lifespans of different families, which can, of course, be analysed.
The result of the study confirmed the low heritability of longevity. That is, while your genes are highly influential on many traits, they don’t matter much to how long you live.
Actually, the Calico researchers discovered that genes might be even less important than suggested by twin studies. They found that married couples – who generally are not related – have more similar lifespans than opposite-sex siblings. And, overall, there is a correlation between the lifespans of a given family and those who marry into it. This might be of some comfort if your mother-in-law has moved in and refuses to give up the ghost.
The similar lifespans of spouses probably stem from the fact that we tend to marry people who are somewhat similar to ourselves. Obviously, we don’t know the life expectancy of our future partners in advance, but they are likely to be people with whom we share interests in things like diet and exercise (or lack thereof), and who have similar levels of wealth, and similar physical traits.
The point of this detail in the story is that the correlation between spouses makes it look like longevity is more genetically determined than it actually is. When researchers adjust for the effect that we marry people similar to ourselves, the heritability for longevity drops below 0.1. In other words, your lifespan is not very genetically determined at all. That’s good news if you want to do something about how long you’re going to live.
Heritability in history
All the studies on the heritability of lifespan are done on deceased people who were obviously born in a very different time to you and me. This might influence the results.
Height is a good analogy. In the past, your adult height was much more conditional on your environment – social class – than it is now. If you were born wealthy, you got plenty to eat, including lots of protein. If you were born poor, you’d most likely subsist on a monotonous diet and might even experience periods of hunger, all while living in crowded conditions conducive to the spread of disease. These differences meant wealthy people used to be taller than poor people, due not to their genes but to their upbringing. Today that is no longer the case. In most developed countries, even poor people get enough to eat, along with sufficient protein and childhood vaccinations. That means everyone gets the chance to grow as tall as their genetics allow. Thus, today, your adult height is much more genetically determined than it would have been in the past. Maybe the same thing will happen with longevity: the more that everyone has access to optimal conditions for a long life, the more important genetics might become.
People tend to think that if something is genetic, it is set in stone. But you should know that genes are neither magic nor fate. They’re just the recipes for proteins. A genetic difference between you and me could mean that you produce a little more or a little less of a given protein, or that your version of the protein is shaped slightly differently from mine. These differences sometimes lead to variation in our traits, but they are not due to magic, just proteins.
If we learn how genetics shape differences between people, we can find ways to mimic the effect using drugs or technology. For instance, genes have an impact on your likelihood of developing poor eyesight, but today we’ve invented glasses, contact lenses and laser eye surgery. Eventually, we’ll develop technology to make it completely irrelevant whether you’re genetically inclined to near-sightedness – perhaps by mimicking the genetic mechanisms that protect some people from developing poor eyesight.
The same is true for the genetics of lifespan. While we’ve learned that genes have a limited impact on how long we live, their impact is not zero either. This means we can get clues to the secrets behind a long life from the genetics of long-lived people. Upon cracking these secrets, we can then design drugs to mimic the effect for the rest of us so that everyone will be able to reap the benefits.
Imagine, for example, that we discover you have a mutation in the fictional gene GENE1. At the same time, we find that you and other people who have this mutation have an increased likelihood of living a long life. When we research the mutation, we might discover that it makes you produce slightly less GENE1 protein than normal. Then, all we need to do is find a way to mimic that in the rest of us, for instance by breaking down the GENE1 protein or by using drugs that inhibit it from being produced in the first place.
To be fair, real-life biology is a bit messier than my simple theory here. The problem is that we have around 21,000 genes. Back in the day, it was normal to say stuff like ‘the gene for height’ or ‘the gene for obesity’. However, today we know genetics are vastly more complicated. Most of our traits are not determined by single genes, but influenced by thousands of different genes at once. For the most part, each gene – or genetic variant – only has a small impact. That means you’ll have to sum up all these small effects if you want to predict something about a person. Fortunately, this is something we can do, by way of studies called Genome-Wide Association Studies (or GWAS). The statistics behind these studies are pretty complex, but the concept itself is simple. In a GWAS, scientists use the genomes of thousands of people in an effort to find correlations between specific genetic variants and certain traits. For example, imagine that you identify a genetic variant that is found in all blue-eyed people but not in brown-eyed people. That could be a sign that the genetic variant has something to do with eye colour. If we already know that this gene has been linked to pigment production or eye development in previous studies, the case strengthens.
Once scientists have identified tons of these small correlations, they sum them up into what is called a polygenetic risk score using statistics. Let’s take a crude example: imagine that we are a pair of under-stimulated researchers who want to investigate the genes behind restlessness. We conduct a GWAS on a lot of people, and in this case we find out that the differences in restlessness are due to a thousand different genetic variants.
Then we look at you and me. In this case, we use a simple model: if a genetic variant makes a person more restless, we say +1; and if it does the opposite, we say -1. When we add all the thousand genetic variants together, I get a risk score for restlessness of +600, while you get a score of -200. In other words, I’d better get a move on with writing this book. And then you can sit back on the couch and read it.
The scientists who carry out actual GWAS for lifespan are still far from understanding the genetics of a long life. But they have uncovered some interesting genetic mechanisms that we can use as clues.
First, there’s a clear connection to the immune system. Many genetic variants that help people live longer play some kind of role in our defence against infections.
Second, there’s a connection to metabolism and growth. For example, there are genetic variants in a gene with the comprehensible name Forkhead Box O3 (FOXO3) which are correlated to living a long life. FOXO3 has a lot of jobs but one is involvement in hormonal signalling by the growth-promoting and metabolism-influencing hormones insulin and IGF-1.
Third, there’s a connection to genetic variants for age-
related diseases. That is, while some of the genetic variants influencing lifespan affect the ageing process itself, others affect your risk of getting an age-related disease once you’ve grown old. The most prominent of these genetic variants is in a gene called Apolipoprotein E (APOE). APOE helps transport fats, vitamins and cholesterol from the lymphatic system back into the bloodstream. But nature is fond of recycling, so it also plays a role in the nervous system and regulation of the immune system. For reasons that are still not entirely clear, APOE is a major modulator of the risk of getting Alzheimer’s disease. There are three variants of the APOE gene among humans: ε2, ε3 and ε4. Most people have two versions of the ‘normal’ ε3 variant (one from each parent). But twenty to thirty per cent of people have one normal ε3 variant and one ε4 variant. This increases the risk of developing Alzheimer’s disease. An unlucky two per cent have two ε4 variants, and these people have many times the normal risk of Alzheimer’s disease.
* * *
Generally, GWAS are most suitable for identifying the effects of genetic variants that are found in a large number of people. If a genetic variant is too rare, its effect can fly under the radar. That doesn’t mean that rare genetic variants are unimportant for health and longevity – in fact, there’s reason to believe the opposite. Fortunately, rare genetic variants with exciting effects are occasionally discovered in other circumstances.
To learn about one such case, we’ll have to take a detour to the small town of Berne, Indiana. At first glance, Berne looks like most other cities in the American Midwest – a gridded street layout, big houses with nice lawns and surrounded by fields as far as the eye can see. But meet the inhabitants, and you’ll notice something is different from your average Midwesterner. Many inhabitants of Berne are dressed in modest, old-fashioned clothes and travel in horse-drawn carriages. If you get close enough to listen in on their conversations, you won’t hear English but a dialect of German.
These people are Amish: a tight-knit group who practise a particular form of Christianity. Their way of life is based around hard work, modesty and eschewing most modern technology. Originally, the Amish came to North America from Germany and Switzerland in the eighteenth and nineteenth century. This is evident in that they still call every non-Amish person ‘the English’. But the Amish of Europe are long gone, and now they are only found in the New World.
A hundred years ago, there were only about 5,000 Amish in the entire United States. But by the turn of the millennium, there were 166,000 and now there are more than 330,000. That’s not because it’s suddenly become fashionable to be Amish. In fact, it’s very uncommon for outsiders to join. Instead, the Amish increase their numbers by having lots of children. As a result, the Amish of Berne mostly descend from a small group of families who moved to Indiana from Ohio in the nineteenth century. Without knowing it, one of these migrants carried a unique genetic mutation. If the person had married into the broader American population, their descendants would have spread widely, and we would probably never have discovered it. But because the person was Amish, many of their descendants are right here in Berne. In fact, some Bernese have inherited the mutation from both their parents because they are descended from the original carrier on both sides of their family tree.
The mutation in question is located in a gene that normally makes the protein PAI-1. It’s a so-called loss-of-function mutation: a mutation that causes a gene to stop working. Someone who has inherited a single mutated version of the gene will produce roughly fifty per cent less PAI-1 than normal. And someone who has inherited the mutant variant from both parents will not make PAI-1 at all.
The reason we know about this genetic variant today is because of research from Northwestern University in Evanston, Illinois. Here, researchers have shown that increased levels of PAI-1 accelerates the ageing process in mice. Meanwhile, lowering PAI-1 is protective. Can you see where this is going?
The Amish people of Berne who carry the special PAI-1 mutation have genetically low levels of PAI-1: a genetic gift from one of their ancestors. If lower PAI-1 levels slow down ageing in mice, could it do the same in people?
The researchers set out to investigate by comparing the mutation carriers with those Amish carrying normal versions of PAI-1. Because the Amish community is tight-knit, the researchers could use family trees to go back in time and figure out who must have been carrying the mutation.
They found that the people carrying the PAI-1 mutation had indeed lived longer lives than ‘normal’ Amish people. That’s an exciting hint that PAI-1 could affect people and mice similarly.
As we’ve previously discussed, the next step is to transfer this genetic gift to the rest of us. Of course, more studies are needed to confirm the effect and understand it better. But biotechnology companies are actually already working to create drugs that inhibit PAI-1. While we wait for that, we might wonder a little bit about why PAI-1 is accelerating the ageing process.
One suggestion is that PAI-1 plays an important role in something called cellular senescence. This is a special condition that some cells enter as we age, in which they hover between life and death. Call them zombie cells. The zombie cells lose their ability to divide, as well as most of their normal functions. However, for some reason they stick around and start spewing out a cocktail of molecules. These molecules – one of which is PAI-1 – can damage tissue and seem to accelerate the ageing process. So we can add ‘zombie cells’ to our list of biological phenomena genetically predicted to play a role in ageing.