Very few people suspected the seals.
Kirsten Bos from the University of Tubingen certainly didn’t when she and her colleagues started studying three Peruvian skeletons. They were just trying to understand the history of tuberculosis—a disease that has affected humans for millennia and still kills millions every year. Team member Jane Buikstra excavated three early victims from a site in southern Peru; their warped spines and ribs showed unmistakeable signs of the disease.
Even though the bones were around 1,000 years old, the team still managed to extract DNA from them. These included sequences belonging to Mycobacterium tuberculosis, the bacterium that causes tuberculosis. The researchers calculated that these ancient sequences last shared a common ancestor with modern M.tuberculosis strains 6,000 years ago.
That was the first big surprise. The general opinion among scientists who study tuberculosis is that it’s an ancient disease that started infecting humans back when we all lived in Africa—after all, that’s where the strains are at their most diverse. As we spread around the globe, this pernicious partner hitched a ride and co-evolved with us. Genetic studies support this view. A big one, published just last year, estimated that all human tuberculosis strains evolved from a common ancestor that lived 70,000 years ago, before the great expansion out of Africa. But the new results suggest that this ancestral microbe was just 6,000 years old!
It’s not just the discrepancy that’s baffling. By 6,000 years ago, humans had already spread around the world, including all over the Americas. The land bridge that connected Asia and North America had long since flooded. And it would be several millennia before any Europeans sailed across the Atlantic. So if tuberculosis originated in Africa, how did it get into South America?
The team considered the possibilities. Maybe some animal rafted across the ocean, taking the bacteria with it? Maybe some bird flew it over? Or maybe, one of them suggested, seals carried it across. They’re long-distance swimmers, they’re infected with a relative of M.tuberculosis called M.pinnipedii, and people often kill and eat them. But, come on. Seals? Seriously? “We had a good laugh over that,” says Bos. “It seemed so silly.”
Still, it was worth testing. The team compared the genomes of many species of tuberculosis bacteria from a variety of animals—humans, cows, chimpanzees, goats, seals, and more. And they found that the closest relatives of the Peruvian strains weren’t the ones that infect today’s humans… but the ones from seals. Seals! (No, not SEALs. Or Seal. Seals.)
“We couldn’t believe that was what the data was showing, but it was pretty clear,” says Bos. “I got the data and sent a text message to Johannes Krause [the senior author], which just said: Arf!”
“This is a triumph of technical and analytical approaches, and It also delivers a wonderfully unexpected result. It’s great science!” says Mark Pallen from the University of Warwick.
Here’s what Bos thinks might have happened: M.tuberculosis evolved in Africa and could have made it into coastal populations of seals (actually, probably sea-lions). That’s reasonable—these microbes are really good at hopping between mammals, as the furious debate around British badger culls attests to. It adapted to the seals, producing the lineage we know as M.pinnipedii. It then spread throughout the southern hemisphere in its new hosts.
Eventually, some of these sick animals were killed by humans living in coastal Peru. We know that many of these groups used seal hides in funeral rituals. They ate seal meat as a regular part of their menu. They could have caught tuberculosis through these practices. If this sounds implausible, note that seals in zoos have passed M.pinnipedii to people before. And some archaeologists have actually speculated that coastal people who hunted and maybe even farmed seals might have caught tuberculosis from them.
The team’s discovery may help to explain some uncertainty around the smudgy history of tuberculosis in the Americas. Scientists used to think that European colonists brought the disease over, since strains that currently circulate in the New World are closely related to European ones. But once they started finding very old skeletons with signs of infection, they knew this couldn’t be right. And in 1994, one team recovered M.tuberculosis DNA from a thousand-year-old Peruvian mummy. The microbe was clearly in the Americas long before Europeans were.
Could seals have been responsible for this early foothold? “It would be quite a brave extrapolation to make at this stage,” says Terry Brown from the University of Manchester. It’s entirely possible that the seals are red herrings, and some other animal that the team didn’t include in their analysis brought tuberculosis to Peru. After all, they only looked at the genomes of 14 animal strains. “They are just scratching the surface of mycobacterium diversity,” says Hendrik Poinar from McMaster University. “There could be plenty of strains from other animals that will fall closer than seals.” The seal story is plausible, but that doesn’t mean it’s right.
Even if seals were involved, it’s unclear how often they passed tuberculosis to people, or what happened afterwards. Their strains could have jumped from person to person and swept the Americas. Or they could have infected those three unfortunate Peruvians and no one else. “[This could have been] just a one-off zoonotic episode, restricted in time and space, leaving the majority of Pre-Columbia tuberculosis in the Americas unexplained,” says Pallen.
Bos agrees. “These three might just have eaten sick seals, got the infection and died, without transmitting it to their peers.” To show human-to-human transmission, the team would need to find similar strains of M.tuberculosis in skeletons from inland archaeological sites, where people didn’t have direct contact with seals. They’re working on that.
Meanwhile, Brown adds that transmission-by-seal isn’t actually the most important bit of the study. He’s more captivated by the suggestion that tuberculosis is just 6,000 years old, rather than 70,000 as previously suggested.
“These dates are worked out by measuring the amount of genetic diversity among all known strains of TB bacteria, and then using a molecular clock – based on the rate at which genetic changes occur during evolution – to work out how much time was needed for all that diversity to evolve,” explains Brown. “To do this, the molecular clock has to be calibrated—we need to know how rapidly the genetic changes accumulated in the past.”
The earlier study calibrated their clock using imprecise figures, based on estimates of when humans spread through the world. Bos’s team (which actually includes six authors from the previous work) calibrated their clock using one of their skeletons. Thanks to carbon-dating, they knew that it was between 1,000 and 1,200 years old. They could work out how much the bacteria have changed since then, and how much time they needed to evolve before. Hence: 6,000 years.
If that estimate is right, it would totally refute the idea that tuberculosis evolved when we were still confined to Africa, and diversified with us as we colonised the world. Instead, it arose when that worldwide spread was already mostly complete.
Of course, they could be wrong. Pallen says that the study doesn’t explain why another group found signs of tuberculosis in a 17,000-year-old bison from North America. Brown adds, “They had to make certain assumptions about the way in which tuberculosis bacteria evolve, and those assumptions might not be entirely secure. We definitely need more ancient Mycobacterium genome sequences, for example from Europe or Asia, and from different time periods, to check this result.”
“The study of ancient DNA [will] continue to contribute significantly to filling gaps in our knowledge of tuberculosis, a devastating disease today that still kills many thousands of people each year,” says Charlotte Roberts from Durham University.
Reference: Bos, Harkins, Herbig, Coscolla, Weber, Comas, Forrest, Bryant, Harris, Schuenemann, Campbell, Majander, Wilbur, Guichon, Wolfe Steadman, Collins Cook, Niemann, Behr, Zumarraga, Bastida, Huson, Niesell, Young, Parkhill, Buikstra, Gagneux, Stone & Krause. 2014. Pre-Columbian mycobacterial genomes reveal seals as a source of New World human tuberculosis. Nature http://dx.doi.org/doi:10.1038/nature13591