“How did they get THAT wrong?” mused Donagh Berry, Statistical geneticist at Teagasc in Ireland. “We asked producers for female DNA samples. These are bull samples!”
The physical differences between females and males are pretty clear. So, to paraphrase Donagh’s question, why this seemingly glaring error?
Let’s back up a bit…
The Irish Cattle Breeding Federation run a national program funded by the Irish government and the European Union to increase the productivity of the national beef herd by improving genetic merit. On average in Ireland, 83% of mated cows go into calf, which, in a herd of 1 million cows, means that 170,000 cows are walking around with no “output”. The benefit of simply increasing the pregnancy rate is putting more coin in producers’ pockets as well as improving environmental efficiency. The increase also means more meat on consumers’ plates.
Because using genomics tools can accelerate genetic gain, producers were asked to submit DNA samples of their cows and potential heifers in order to identify superior females to breed the next generation. Over 1 million samples were submitted.
Females have two X chromosomes; males have one X and one Y chromosome. That’s pretty simple. But as the DNA testing progressed, some anomalies were detected. A small number of samples were arriving with male DNA (i.e., one X and one Y chromosome) despite producers being instructed to submit samples from females. Hence Donagh’s question.
Producers had less faith in the technology, and more faith in the udders they saw in front of them. They sent in more samples from those animals. Wrong, wrong, wrong. The “heifers” still registered as male; and each side started digging in their heels. Donagh declared this mystery needed solving, and turned to his long-time collaborator, Gentec researcher Paul Stothard (UAlberta) for advice.
“I went to look at these heifers,” says Donagh. “They were definitely female. So I took yet more samples. It turns out that these animals are female on the outside but have the DNA of a male. In other words, they have the Y chromosome. Then I checked out the literature, and found that it was a known syndrome called Swyer, which exists in humans, too.”
Under necropsy, Donagh could see that the animals were female internally as well but infertile. Given the 680,000 genotypes generated at the time, and the eight anomalous results detected, Donagh and Paul deduced that the incidence of Swyer is 1 in 80,000 cows, about the same as in humans.
So what are the implications?
“Well, saving face is a big one,” says Donagh. “There’s always an undercurrent of skepticism about technology. An abnormal result like this puts the credibility of genomics into question. In Ireland, this could spread on social media, and the whole thing is in jeopardy.”
For producers, the implications are somewhat smaller—or 6-8 animals per million. A chromosome-counting technology is available for about $150 that detects karyotype abnormalities like Swyer, Edward syndrome, Turner syndrome and Down syndrome. Because these syndromes are so uncommon, Donagh doesn’t recommend this course of action.
But with current genomics technology that measures tens of thousands of pieces of DNA, Donagh and Paul discovered that seven of the eight cows were missing the SRY gene on the Y chromosome. (The eighth cow was an anomaly of an anomaly: she had the SRY gene. We don’t know why.)
“Detecting Swyer is as simple as adding DNA markers in the SRY gene on the genotyping platform already used at no additional cost. DNA markers on the rest of the Y chromosome already exist on these platforms” says Donagh. “A Swyer female will give you a signal for the Y chromosome but not the SRY gene on that chromosome because she doesn’t have it.”
Paul and Donagh have suggested DNA markers that companies can include on their platforms so that, when producers get their animals genotyped, they will get a result for Swyer as well. The group has already developed methodology to infer other chromosomal abnormalities, such as Turner Syndrome, again at no additional cost.
“We wanted to leverage the data to deliver more information,” says Donagh. “At the end of the day, producers are genotyping and getting information on parentage, abnormalities, congenital defects, etc. all from one sample. They get Swyer thrown in for free. You can argue that the incidence is low… but the cost is zero!”