Inbreeding Might be Less Dangerous than Anecdotes

The following is written by MP a Doctor of Biogolgy recently graduated from Harvard University - Department of Organismic and Evolutionary Biology. In the wake of Eight Belles’ tragic Derby finish, there has been an effort to uncover the smoking gun. Was it the whip, was it the drugs, was it the track, was it the stress of racing with the boys? Was it the eerie parallel to the Democratic primary, brought on by Hillary Clinton’s endorsement (she finished second to a less experienced contender, Big Brown)? Or was it all the inbreeding? It’s on this last question that I’d like to offer my two cents. In doing so, I’d like to also clear up what inbreeding is exactly and why anybody cares about it. First, here is a crash course, or a refresher, in genetics. If you’re cozy with Mendel, skip to the next paragraph on inbreeding. Every sperm and every egg contain every gene in the genome (minus some special exceptions which we can ignore for the moment). But every sperm and every egg contain only one copy of every gene. When the sperm and the egg meet, fertilization occurs, and you get the beginnings of an individual that will have every gene in the genome in duplicate with one copy coming from the egg (Mom) and one copy coming from the sperm (Dad). So, if we pick one of the ~25,000 genes in the horse genome – let’s say a gene involved in ankle development – we would find a single copy of it in a sperm, a single copy of it in an egg, and two copies of it in every cell of a grown horse (except, of course, in that horse’s eggs or sperm, which will have just one copy). Every time a grown male horse produces sperm, it’s a random pick of its two copies of the ankle development gene that gets packaged into the sperm (the same applies to females and their eggs). One may hear that we only pass on half of our genes to an offspring. This doesn’t mean that they only get 12,500 from us; rather, it means that we only pass on one copy of the two that we have for all 25,000 genes to a given offspring. That other copy of each gene does not get passed to that offspring. It can however, be passed to subsequent offspring, which is one reason why brothers and sisters don’t look exactly alike. It’s also the reason we must talk about inbreeding as a probability. Inbreeding occurs when there is a shared ancestor between an individual’s mother and an individual’s father. This means that this shared ancestor can contribute genes by descent via the mother and the father of an individual. And so, an inbred individual can have two replicate copies of a gene for ankle development that both come from that shared ancestor. Why might this be problematic? It seems as though there’s nothing wrong with the numbers. This inbred individual will still have all 25,000 genes in duplicate copy. What’s the problem if some of those duplicates are identical? The problem arises if that shared ancestor had a bad copy of the gene. Mutations can occur in genes that render them useless. Genes code for proteins but when they get certain kinds of mutations, they may no longer code for anything at all and the individual can end up deficient for that protein. However, having one functional version of the gene is typically enough to make all the protein that the body requires. So, this shared ancestor may have suffered no ill effects of its one bum copy because it may have had one functional copy that compensated. But a descendant that is inbred to this ancestor has the possibility of inheriting the bad copy in duplicate. This is the risk of inbreeding. (This is also the allure of inbreeding, as it’s a way to double up on good mutations from a shared ancestor. But rare mutations are on average harmful.) The farther back in the pedigree these shared ancestors are, though, the less likely it is to inherit both copies from that ancestor. We can quantify this chance of inheriting the same allele twice from a shared common ancestor. For instance, let’s say that one of your two maternal great-grandfathers is also one of your paternal great-grandfathers (in other words, your parents are first cousins). The chance of inheriting the same copy of a gene twice from that great-grandfather is 1/32. Here is the intuitive explanation. There are 8 individuals in your great grandparental generation: 4 on your mother’s side and 4 on your father’s side. Each of them has two copies of each gene. So, first, there are 8 copies of the gene on your mother’s side in that distant generation and second, there are 8 copies on your father’s side. Now, one of those first 8 had to find it’s way into the egg that made you and one of those second 8 had to find its way into the sperm that made you. Two of the first 8 belong to that shared ancestor. So, there is a 2/8 chance that one of his genes made it to you via the egg. Then there is a 1/8 chance that that exact same version of this gene made it to you via the sperm. We multiply these probabilities and get 1/32. That’s the risk of inheriting the same copy of a gene twice from a shared great grandparent. We call this value the inbreeding coefficient. One common misunderstanding about inbreeding is that it somehow gets “passed on” from generation to generation. This is not necessarily true. For instance, let’s suppose your mother is inbred; she’s the offspring of a brother-sister mating, say. Your mother’s inbreeding coefficient is 1/4. What is your inbreeding coefficient? The answer is impossible to give because I haven’t specified who your father is. If your father is unrelated to your mother then you are not inbred and your inbreeding coefficient is zero. If your parents aren’t relatives, then you avoid the risk of inheriting the same copy of a gene from both of them. It doesn’t matter in that case if your mother is inbred or not. Inbreeding is not a trait like hair color or temperament that is heritable and passed down to successive generations – at least not necessarily. If the same ancestors keep appearing on both sides of the pedigree as generations proceed, however, then inbreeding will persist and may even increase. I’ve calculated the inbreeding coefficient for Eight Belles, at least to a rough approximation, using the pedigree from pedigreequery.com. I’ve only gone back five generations and I’ve assumed that the ancestors that she shares on both her mother’s side and father’s side are themselves not inbred, which will downwardly bias my estimate. In Eight Belles’ pedigree, Mr. Prospector is both her maternal great-grandfather and her paternal great-great-grandfather (4 X 3). Raise a Native is both her maternal and paternal great-great-great-grandfather (5 X 5). And finally, Native Dancer is both her maternal great-great-great-grandfather and her paternal great-great-great-great-grandfather (6 X 5). These are the relevant ancestral relationships. This produces an inbreeding coefficient of 19/1024 (or ~0.019). Is this high for a Derby horse? The average inbreeding coefficient for the twenty horses in the 2008 Derby was ~0.012. Eight Belles was the fourth most inbred horse in that race. The highest inbreeding coefficient, 0.051, belonged to the winner, Big Brown. Only three horses (Z Fortune, Recapturetheglory, and Bob Black Jack) had an inbreeding coefficient of zero. Fourteen of the horses in the Derby were inbred to Native Dancer or one of his direct descendants (Mr. Prospector, Northern Dancer, or Raise a Native). Only one horse (Gayego) had no trace of Native Dancer in the first five generations of his pedigree. We could ask two questions: 1.) Is inbreeding generally associated with unsoundness? and 2.) Is inbreeding to those three stallions (Mr. Prospector, Raise a Native, and Native Dancer), in particular, associated with bad ankles, as some have claimed? Good evidence would come in the form of a statistical association between higher inbreeding coefficients and higher unsoundness. What we have at the moment is an abundance of anecdotal evidence, something that is too readily accepted as truth in racehorse breeding circles, and not enough careful, quantitative evidence. Eight Belles has a high inbreeding coefficient and she broke down – “therefore, it must have been the inbreeding,” say the anecdotalists. But other horses break down, too, and we can’t always chalk that up to their inbreeding. Barbaro comes to mind. His inbreeding coefficient was a rather miniscule 1/256 (0.004). “But, look - Barbaro also had Native Dancer blood in his pedigree,” is what an anecdotalist might cry. Of course Barbaro did. Native Dancer had 247 offspring! And Raise a Native had 700+ offspring and Mr. Prospector had 1000+ offspring. There are so many horses in the current generation who have these three males in their pedigrees that we should expect a certain number of them to break down just by chance alone. In fact, in this year’s Derby, there was a 95% chance that if one of the horses broke down, it would have been a descendant of Native Dancer. We are almost as certain to find Native Dancer blood in the pedigree of a broken down horse as we are to find that the broken down horse had two eyes, ate grass, and whinnied. Another question that has been raised is the following: Is the observed increase (*I’m not sure if this increase has been quantified carefully or if it’s just another anecdote) in breakdowns over the last few decades due to this sweep of Native Dancer’s genes into the population? More generally, might we be sweeping genes for unsoundness into the population by racing horses only for a single season and then cashing them in for stud if they perform well? These two changes in breeding might very well pose a risk for the breed. But by the same token there might be other changes in the practice of racing that have taken place since the 1950’s along with this sweep of Native Dancer’s genes and the breeding-driven economics of the industry that have lead to an observed increase in unsoundness. Teasing apart this correlation would require something better than anecdotal evidence. My purpose here has been to explain what inbreeding is, explain why it’s a concern, and to show how anecdotal evidence doesn’t get you very far. I enjoy the odd anecdote about racehorse breeding as much as the next guy. It’s one of the charms of horse racing. Further, anecdotes are often verified when subjected to formal, careful analyses, pointing to the instincts that close followers of horse racing hone. But anecdotal evidence is potentially dangerous if it distracts us from sorting out the real cause of these breakdowns. -MP