GENE Amsterdam Interview Session 6: Dorret Boomsma

Interviewed by: Jana Hirzinger and Anaïs Thijssen

Looking back, what inspired you to focus on twin genetics? Was it something you set out to do from the beginning?

No, not right from the beginning. As a student, I applied for a bursary that allowed you to do whatever you wanted for one year. I had taken a course from a twin researcher at the VU and thought it was a fascinating topic that I would like to study in the US for a year. He advised me to apply in Minneapolis and Boulder. Minneapolis turned me down, but Boulder accepted my application. Robert Plomin was there at the time, and I had a great introduction to quantitative genetics by John DeFries. It was the methodological work that made this field of huge interest to me. When I came back to Amsterdam, it was at least as difficult as it is nowadays to get a position in academia. So I actually started my career as a statistician at the Department of Dentistry, and then I moved on to a PhD position on cardiovascular risk factors in twin families. Many years later, I was the president of the Behaviour Genetics Association and had to give the speech at the annual meeting in Minneapolis. And I still had my rejection letter, so it was a fantastic moment to show that I had returned as president of the Behaviour Genetics Association to the university that turned me down. The funniest thing was that afterwards, many people from the university wrote me emails saying it wasn’t them who rejected my application.

What are some of the most memorable or surprising discoveries you’ve made about twins or twin genetics over your career? 

Already early on in our work, within the first 10 years of results based on the Netherlands Twin Register, we saw that at very early ages (before 7 years), the influence of the genome on behavioural problems and psychopathology in children is very large. I believe we would have predicted that the influence of genetics on early life traits is quite small and that when children develop genes become much more influential, which is what we see for IQ, for example, but which is not at all what we see for other traits. Traits like attention problems or aggression in very young children are highly heritable. In the early days, this was one of the more surprising outcomes.

How do you look back on your career?

It has been and still is a very happy journey. When I started working in human genetics, we almost exclusively did statistical modelling using twin and family data. Then came the development of multivariate models, as initiated by Lindon Eaves and Nick Martin, which was a huge progress in statistical software. And then, at the point in time where one might start to wonder, well, am I going to spend my life estimating heritability for yet another trait? Along comes the possibility of doing linkage studies for quantitative traits. In hindsight, we recognize that those studies were hugely underpowered because we did not have a good feeling about the large degree of polygenicity that is responsible for the heritability in most traits. After that, we moved on to the somewhat miserable period of candidate gene studies, and then finally, there were technological breakthroughs that meant we could do large-scale SNP typing and GWAS for about every trait. It has been fun to see the field develop this much. What the next breakthrough will be is difficult to predict. We should be careful not to think that with the discovery of SNPs, we now know everything. 

Where do you see the future of twin studies going?

I think the enormous value of (monozygotic) twins will remain. Take a trait like schizophrenia, for example, which is a highly heritable disorder with estimates of up to 80% of the variation on the liability scale. However, if you look at the concordance of schizophrenia in monozygotic twins, the upper limit seems to be somewhere between 40-50%. So why are two people who share the same genetic variants not more alike in their phenotype? This is a very important finding to keep in mind for personalized medicine. If you cannot predict from the characteristics of your monozygotic twin whether you will also become a schizophrenia patient, then you can also not predict this longitudinally in the life of a single individual. There is still a lot going on that we cannot and probably will never be able to predict, so the hype around genomic prediction is not really warranted.

And then there is the exciting new development of what the aetiology of twinning itself might mean. The Netherlands Twin Register was established with two large grants that aimed to resolve the aetiology of twinning. For thirty years, this was a spectacularly unsuccessful enterprise. Only when GWAS became available, we found the first sets of genes for having dizygotic twins as a trait in mothers of twins. Interestingly, a polygenic score that predicts dizygotic twinning is also predictive for female infertility, greatly expanding the value of discovering those genes. Finally, we found evidence for the hypothesis that there is a continuum that goes from multiple ovulation on the one extreme to no ovulation on the other extreme, with everything in between almost varying on a quantitative scale.

Have you studied any twin phenomena that are not well known to the public, but that you think deserve more attention or research?

I would love to talk about a phenomenon called chimerism. It means that, in addition to your DNA, you’re also carrying the DNA of another person. This was initially discovered in dizygotic twins and thought to be extremely rare. In the 1990s, we showed that chimerism is actually much more common. For some time, this paper was a Sleeping Beauty. No one picked it up. Then, one Friday afternoon, the New York Times called me because the paper had been supplied as evidence in a court case involving a cyclist accused of blood doping. He claimed the blood was from his twin brother, arguing he was chimeric because of a vanishing twin. Of course, that was disproven easily, and he was found guilty in the end. Chimerism also became much more of a focus because it turned out that people are not only chimeric with their twin brothers or sisters, but all women who have been pregnant are chimeric with their offspring, and vice versa. 

Follow-up question: Are there any specific questions regarding chimerism that you would like to have answered?

I would love to extend the research into chimerism because it is related to many health outcomes, both favourably and unfavourably. Autoimmune disorders, especially those that have a higher prevalence in women and that become apparent in their 30s when women often have their first child, seem to be linked to chimerism. But chimerism also seems to protect against developing cancer. I would like to assess the degree of chimerism in a much larger group with more advanced methods, for example, developing technology to determine chimerism in males. The next step would be to find out who the donor of the chimeric material is.

What do you think the future holds for you?
In the Netherlands, the rule of retirement at 67 is strictly enforced. But if there is a department that invites you to stay on, you can work and supervise your PhD students for another five years. I feel extremely happy in this department of Complex Trait Genetics, where I have gotten the opportunity to continue to work. I have no idea what will happen after five years. Maybe five years is enough.

Picture by B. Bronshoff

Share This