Identical twins have early genetic differences

A new study has shown that identical twins… are not actually identical! Researchers already knew that they had certain differences, because everyone’s DNA mutates throughout their lives - that’s why some people get cancer. But those kinds of mutations seem to be restricted to the body, and don’t get passed on through sperm or eggs. Now though, human genetics company deCODE in Iceland have used the genomes of nearly 400 pairs of twins to show that they often have a few differences that they actually pass on to their kids - a result that could have implications for other parts of genetics research. Phil Sansom heard from Kari Stefansson, head of deCODE...

Kari - There are certain differences in the genome of identical twins. And the reason that is important is that we have always assumed, when we are trying to separate the facts of the environment from the fact of genetics, that the genomes of identical twins are identical. So finding these mutations changes, a little bit, the way in which we can use identical twins for the purpose of separating environmental effects from genetic effects.

Phil - How is that possible, that identical twins are not identical? Because I thought they come from the same egg, don't they?

Kari - They come from the same fertilised egg. But when an egg is fertilised, it divides several times before there is a beginning of the formation of twin. And the cell divisions can lead to what is called replication error; when a cell divides, the genome in the cell is duplicated. And even though this process of duplication is a very good one, it's not flawless. And we call the flaws 'mutations'. So there is an opportunity for mutations to take place, and what we document in this paper is that they indeed do take place and they have a significant impact on differences between genomes in the twins.

Phil - When are they taking place? How big is the embryo at this point?

Kari - It is very small - around 16 to 20 cells or so.

Phil - And how can you tell that this is going on? Were you looking at these embryos at this early stage yourself?

Kari - No, the way in which we can do this is by sequencing the whole genome of the twins, of their parents, of the children, and of their spouses. We can infer when the mutations take place. If a mutation happens very early in the embryo, you will find that mutation not only in the children - because they have made it into the cells that lead to sperms in the male or eggs in the female - they will also end up in the body. And when a mutation is found both in the body and in the children, we know that mutation happened before they're differentiated. So it helps you to time it, and the timing is all-important when it comes to development.

Phil - They must also be happening only in one twin as well, right, for you to get the difference?

Kari - Yeah, because the cell that leads to one of the twins may not lead to the other. In some instances, both of the twins are developed from the descendants of the same cell; sometimes they're descendants of different cells. And if they are formed from descendants of different cells, then they will harbour different mutations.

Phil - Is it totally random which one of those you get?

Kari - I think it is very close to being random.

Phil - Now how big do these differences actually get?

Kari - These differences are not big. On average you will find about five mutations of this short that differ between identical twins, but there can be more than that.

Phil - So what does that mean for twin studies in genetics? You alluded to this earlier, because geneticists use twins all the time to try and separate out the genetics from the environment for a particular trait, or disease, or something.

Kari - We should definitely demand that people sequence the whole genomes of both of the twins before we ascribe differences between them to the environment, because indeed these mutations that are different between them could be the reason that they are different.

Phil - Does it affect the results of twin studies that people have done in the past?

Kari - It should caution people, but it should only affect differences that are rare. For example, when one twin develops a disease and the other does not, if the disease is very rare then we should definitely re-explore whether that is explained by a mutation and not just environmental effect.

Phil - Finally, I just want to ask: do you know if these differences between twins ever translate into something visible, like you might get two identical twins with something visibly different between them?

Kari - I am absolutely certain that occasionally it does. Because occasionally you'll see differences between identical twins that will not be explained by the environment.

Phil - It's pretty amazing, I think, for people to imagine that identical twins are not actually identical!

Kari - Yeah. Monozygotic, but not identicals.

Is this information a blow to the types of experiments known as “twin studies”? Geneticist Robert Plomin says no. He runs one of the largest such experiments, called the Twins Early Development Study, and he explained to Phil how they normally work...

Robert - A twin study is a design that's been around for over a hundred years, that uses a natural experiment. About 1% of all births are twins, and a third of those are what we call identical or monozygotic twins. They're clones of one another because they're derived from the same fertilised eggs; s,o when an egg and sperm get together and produce a zygote, that zygote sometimes splits in the first few days of life. The other type of twins - two thirds of all twin births - are non-identical twins or fraternal twins. Like any brother and sister they're 50% similar genetically. So the essence of the twin design is that: if a trait - say musical ability - is influenced by inherited DNA differences, you have to predict that identical twins are more similar than non-identical twins. And the extent to which that's true is an index of what we call heritability - how much of a difference does genetic differences make in the trait that you're studying.

Phil - So you're separating genes from environment. Nature from nurture, basically.

Robert - Yes, that's the idea. We've known forever that things run in families, but the question is, do they run in families for reasons of nature - that is inherited DNA sequences, genetics - or nurture - that is systematic effects of the family environment?

Phil - Can you talk me through maybe an example of a twin study so that I can picture how one might work?

Robert - Yeah. There are literally thousands of twin studies that have been reported throughout the life sciences. If you take a trait like musical ability - for which only recently do we have any twin studies at all, because it's kind of hard to measure musical ability - what we do is we get a large group of twins. We then correlate the identical twin pairs, and you say, "how similar are they?" Their correlation might be, say, 0.5; whereas non-identical twins, if they correlated at 0.5 for musical ability, that would mean there's no genetic influence. If the identical twin correlation is 0.5 and the fraternal twin correlation is 0.25, that suggests genetic influences account for about 50% of the differences that we observe in musical ability. But as always our studies are limited to the samples we described, so we're only describing a particular population at a particular time.

Phil -
Now Kari Stefansson in his paper is seemingly saying that identical twins are, in fact, not identical. Is that a shock to you?

Robert - No, I think we've known that for a long time. His study is great in terms of having a large sample and very systematically looking at this, because he actually sequences all 3 billion base pairs of DNA. But the twin method depends on inherited DNA differences. That's what we call genetic. It's a very narrow definition of genetic, because there are many biological and even DNA factors that are not included in that definition; they would be called environmental. In Kari's study particularly, what I'm afraid people will take away from it is the idea, "well, the twin method's no good because identical twins have differences." But in fact, we've known that we all get mutations as we go through life - cancers are often due to mutations - but these are not inherited DNA differences. And we know from lots of studies using a technology called SNP chips, and they focus on inherited DNA differences. They don't show any differences between identical twins. So identical twins don't differ in terms of the DNA differences they inherit from their parents.

Phil - You're saying that for the purpose of twin studies, these genes are not inherited from the twins' parents; they're eventually passed on, but that still doesn't disrupt the way the studies work.

Robert - Yeah. And it's always a problem in science when a word like 'inherited' or 'transmitted' is used in such different ways. But in genetics, what we're talking about is: the twins we're talking about as the offspring. That is, did they inherit these DNA differences, or are they spontaneous mutations. Inherited DNA differences is a very narrow definition, but it's a precise one, and it helps us understand things like Kari's finding.

Phil - So are twin studies safe?

Robert - Yeah, I think without doubt!