This week I’m staying close to home with a post on genetics. Two days ago there was a paper published in Science (open access) that seems to suggest that we humans might be more genetically resilient than we previously thought.
When we discover a gene one of the ways we can try to find out what it does is to stop it functioning and then see what happens to the organism in question. We call this ‘knocking out’ and if we knock out gene x then the organism in question can be called a x knockout. It’s easy enough to do this in the various model organisms we use be they plant, fly, mouse or whatever else you are studying.
Generally, the results of a knockout will fit into 3 categories. You might see no effect at all implying that the gene in question doesn’t do anything particularly important or that there is redundancy built into the system. You might fail to be able to create a viable organism implying that the gene is essential for life in some way. Or, thirdly, you might obtain an organism that is altered somehow which you can go on to study for clues as to the function of the knocked out gene.
Whilst we can readily do this with plants and animals it is unethical and illegal to do so in humans and, so, we are left either to infer what we can from our models or to study naturally occurring knockouts in the human population. This latter option is what the group concerned went for.
They studied 3,222 British people of Pakistani origin to look for naturally occurring knockouts. The reason they used this particular cohort is that many Pakistanis seem to have a culture where marrying your first cousin isn’t seen as being a bad thing, indeed it is sometimes encouraged. The result of this is that long stretches of DNA in these individuals end up identical. Remember that we each have two pairs of 23 chromosomes inside us, one pair from mum and the other from dad. In most of us those two people are, relatively, unrelated and so each of the pairs can be quite different from the other. In inbred populations there isn’t that variety because the gene pool is shallower.
The usefulness of this is that to create a gene knockout then both pairs of the gene have to be disrupted, in people with homozygous stretches of identical DNA there is a greater likelihood of both pairs being so.
We already know that there are quite a few of our 22,000 or so genes that we can do without, perhaps as many as 800 and counting; not everything is vital. On average we each have about 20 genes that are knocked out and it doesn’t do us any harm at all. What the researchers wanted to find were the rarest (occurring in <1% of the population) variants that knockout a gene, known as loss of function mutations. If a mutation is rare this would normally be for two reasons, either it has arisen newly in the population and hasn’t had time to spread, or, it is harmful and is being actively selected against by natural selection. It is these mutations we want to study.
Having sequenced the exomes of all the participants, the small part of the genome that actually encodes for proteins (about 1.5% of the total), they identified 1,111 such deleterious variants, some of which appeared homozygously in the same individuals. By cross referencing with comprehensive health records of the participants they established that people with these rare gene knockouts were no more likely to have health problems than the general population.
There are some potential confounding factors for this study which the authors acknowledge. For example, it could be that the diseases caused by the gene knockouts only onset very late in life and so they may not have manifested yet in the cohort. Plus, there is a phenomenon in genetics known as reduced penetrance where, for reasons we don’t yet understand, even if you have a pathogenic mutation it only seems to manifest, or penetrate the phenotype, in a given proportion of cases.
Allowing for these potential shortcomings, though, what does this study tell us? Perhaps it is the case that far more genes than we previously thought are not vital to our health? Perhaps we have more redundancy and secondary systems than we thought we had? Probably it is a mixture of both of these.
There was a surprise in the data, however. One lady had a knockout of the PRDM9 gene which is involved in human meiotic recombination. In mice, knockouts of this gene are rendered infertile as the genome is rendered unstable. There is evidence that the mechanism is conserved across all eukaryotes and therefore is extremely important for life. Some animals, such as dogs, have evolved back up mechanisms which mean they are able to reproduce without PRDM9 but it is certainly vital in mice and it was thought to be so in humans, yet this lady has a child. Why she is still fertile we don’t yet know, but it was established that it isn’t the same mechanism dogs have.
This is very important because in my business, that of identifying mutations and then assigning a level of pathogenicity to them, we need to know what is essential and what is not. If there are back up systems we aren’t aware of then the chance of issuing a false positive result increases. This is especially important as we plough headlong into the new era of genetics hailed by the cheap and readily available ability to look at whole exomes of huge numbers of people. We are going to find more and more variants like this that we simply won’t know how to interpret. I am in full agreement with the authors when they call for more studies of this kind to give us the data we need to better understand our genome.