Twin Studies are a very useful method for finding out what component of a certain trait is made up from a person’s genes and how much from their environment. They’re a deceptively simple way of finding out something that would otherwise seem like an uncrackable problem, the downside is that you generally need a lot of twins to hand.
They work like this. There are two types of twins out there: monozygotic twins (MZ) and dizygotic twins (DZ). MZ twins are identical twins that were the product of one act of conception between one egg and one sperm that, for some reason, ended up splitting into two genetically identical embryos not long after conception. DZ twins are the result of more than one egg being released at a time during ovulation and both of them being fertilised by different sperm. Although they’re developing at the same time in the same womb they are no more related to each other than any other pair of siblings, they are non-identical twins.
Now, because they’re twins, they tend to live in the same place at the same point in time, they get treated similarly, eat the same food and so, at least whilst they’re kids, it’s not too unreasonable to assume that their shared environment was all but the same. Over the course of their lives, as these tend to be very long term studies, the twins are regularly checked to see if they have developed any particular traits. Normally we’re looking out for diseases and disorders. Some twins are adopted, though, and don’t grow up in the same environment. If one half of a pair of MZ twins is adopted and goes to live in a compeltely different part of the country with a completely different family then we stand to learn an awful lot; differences between the two twins are likely to be a product of their environment as opposed to their genetics.
If there is a strong genetic component to whatever trait is being studied then you would expect the MZ twins to have a higher incidence of it than the DZ ones. There are various fancy bits of maths you can use to calculate your result, such as Falconer’s Formula, but ultimately it is possible to come up with a number that falls between 0 and 1. The closer something is to 1 then the more genetics can be said to be involved in that trait. If it is nearer to 0 then the environment is the more important factor. Most, of course, are a mixture of the two.
Things like type 1 diabetes, obesity and certain cancers have very high heritability scores, high being around 0.7-0.8. Some; like leukaemia, lung cancer and type 2 diabetes; have very low scores ranging from 0.01-0.25.
When a trait has a very high heritbility it normally isn’t too difficult to find at least some of the genetic factors involved with it. Not all of them, by any means, but at least the major players. The bulk of diseases that have genetic components tend to be caused by multiple different genes, all of them contributing different amounts to the overall picture, we call them polygenic disorders.
The reason I mention all this is that there has been a very obvious and very frustrating exception to this: schizophrenia. It is a disease that has one of the highest heritability scores out there, in the region of 0.8, and yet, for decades, the genetic community has completely failed to come up with even a sniff of a modus operandi for this debilitating psychiatric disorder that affects 1% of people.
We may, however, have just changed that. A paper published in Nature last week proposes the first ever genetic mechanism that may go some way to explaining how the symptoms develop.
Looking at the genomes of more than 60,000 people they found a small association with an area known as the Major Histocompatibility Complex. This is a complex of genes involved in our immune systems. The researchers found that there was a gene called Complement Component 4 (C4) where the risk for schizophrenia changed depending upon which version of the gene was present. In people who had type A there was an increase in risk from 1% to 1.27%.
This is a very small increase in risk but this makes sense to me. If the risk was large we would have already found it. The lead and last author in this study are from the Department of Genetics at the Broad Institute in Boston, this is where the ExAC database team works, a database of 60,706 genomes that have been compiled to give researchers the power to look for these small effects. I suspect, then, that the ExAC database was the source of their revelation. I went to a lecture by the creator of ExAC, Daniel MacArthur, a couple of weeks ago and it was very exciting. They have impressive plans for ExAC in the coming year, not least of which is a doubling of its size.
Anyway, to get back on track, we now have a candidate gene to investigate. The C4 gene is involved in cutting the connections between the cells in our brain. This is a very normal process that we all go through starting at age 5 or so and continuing through childhood. In those first few years our brains are making about 1000 new synaptic connections every second. One of the jobs of C4 is to break some of these connections in a process known as synaptic pruning. People with the type A version of C4 undergo more pruning than those with other types.
This finding fits in nicely with existing knowledge about schizophrenia. It traditionally onsets during adolescence and at post mortem schizophrenic brains have been seen to have different architecture which could be explained by the lack of proper pruning. It also seems reasonable that too many broken connections in the brain could explain the subjects inability to distiguish beween real and imagined voices and some ofthe other symptoms. The C4 gene seems like a reasonable lead.
We need to be cautious, though; there is not even a suggestion yet that this might lead to some kind of treatment or drug that may help. All we have at this point is an interesting feature at which to aim future research. In all likelihood we are still at least two decades from any kind of drug treatment if, indeed, one ever will exist. Given that the pruning process starts in very young children long before schizophrenia could ever be diagnosed you would have to take a gamble on treating someone who might, at most, merely be at greater risk of developing schizophrenia as opposed to waiting for them to become symptomatic; a course that is fraught with risk and ethical concerns.
Another consideration is that schizophrenia is certainly a polygenic disorder. The C4 gene is our first hint and may well be the largest single genetic contributor to the disorder. But then that would leave potentially hundreds of other genes all of which contribute just a very small amount individually but, together, account for the bulk of it. Pinning them all down is the stuff of dreams at this point, decades away at least.
As I embark on my masters in Genomic Medicine it seems that these are the sorts of problems that will increasingly come under scrutiny. As our data stores increase so does our power to find these small effects. The genetic revolution may yet arrive, but it’s going to be a long time coming.