The most exciting recent development in human genetics research has been the ability to perform large-scale systematic studies of genetic variation in thousands of people. These genome-wide association studies (GWASs) have revolutionised our understanding of many different complex diseases.
But despite these advances, we are still only able to explain a small fraction of the heritability of many health conditions. In a study that my lab has just published in Science, we show that a person’s attributes could be strongly influenced by genetic variation in an unexpected part of the genome that has been overlooked in previous studies.
The environmental factors that play a role alongside genetics in determining a person’s attributes are also present in the womb. When offspring are in the womb, what their mothers experience environmentally (including diet, stress, smoking) has the potential to influence an offspring’s attributes when they become adults. This “developmental programming” is understood to be a large contributor to the obesity epidemic seen today.
A key player in this process is epigenetics. Epigenetics are modifications that sit outside the genome and determine which bits of DNA to make more active or inactive. One such modification involves tagging DNA with compounds called methyl groups. Methyl groups determine whether genes are expressed (switched on) or not. Liver cells and kidney cells are genetically identical apart from their epigenetic marks. It has been proposed that in response to a poor environment in the womb, an offspring’s epigenetic profile will change.
In our study, we compared the offspring of pregnant mice when given a low-protein diet (8% protein) and a normal diet (20% protein). After they were weaned, all offspring were given a normal diet. We then looked at the difference in the offspring’s DNA methylation, comparing those mice whose mothers had a low-protein diet to those whose mothers had a normal diet.
Looking in the wrong place
Initially, we found nothing, so that was a big surprise, but then we looked at the ribosomal DNA (rDNA) data and found huge epigenetic differences. Ribosomal DNA is the genetic material that forms ribosomes – the protein-building machines within the cell.
When cells are stressed – for example when nutrient levels are low – they alter protein production as a survival strategy. In the mice whose mothers were fed a low-protein diet, we found that they had methylated rDNA. This slowed the expression of their rDNA and resulted in smaller offspring – as much as 25% lighter.
These epigenetic effects occur in a critical developmental window while the offspring is in the womb but is a permanent effect that remains into adulthood. So a mother’s low-protein diet while pregnant is likely to have more severe consequence on the offspring’s epigenetic state and weight than an offspring’s own diet after it has been weaned.
Looking beyond the epigenetic markers, when we looked at the basic genetic sequence of the rDNA, we found an even bigger surprise. Even though all the mice in the study were bred to be genetically identical, we found that the rDNA between the individual mice was not genetically identical – and that, even within an individual mouse, different copies of rDNA were genetically distinct. So there is huge variation in rDNA which is also playing a big role in determining the attributes of offspring.
In any given genome, there are many copies of rDNA, and we found that not all copies of the rDNA were responding the same way epigenetically. Only one type of rDNA – the “A-variant” – appeared to undergo methylation and affect weight. This means that the epigenetic response of a given mouse is determined by the genetic variation of their rDNA – those who have more A-variant rDNA end up being smaller.
Heritability (how much the risk of a disease is explained by genetic factors) of type 2 diabetes has been estimated to be between 25% and 80% in different studies. However, only about 20% of the heritability of type 2 diabetes has been explained by genome studies of people with the disease.
The fact that genetic variation of ribosomal DNA seems to have such a strong influence suggests that GWASs in humans could be missing a key part of the puzzle, as so far they have only looked at the single copy part of people’s genomes. Genetic and epigenetic analysis of rDNA in humans could yield very important insights into a variety of human disease.