This lung specimen from 1912 had been sitting in the basement of the Berlin Museum of Medical History at the Charité. (Düx etal / bioRxiv)

Sometime in late 2018, Sébastien Calvignac-Spencer descended into the basement of a Berlin medical history museum, a place so teeming with old medical specimens that each organ gets its own storage room. He headed for the lungs. Among the 400 or 500 jarred specimens that had accumulated over decades, he and his colleagues ended up finding one particular lung—preserved chemically in formalin—with clues to the origin of measles. The lung was from a girl who died of pneumonia after a measles infection in 1912.

Now more than a century after her death, the team led by Calvignac-Spencer, a virologist at the Robert Koch Institute, has managed to sequence the measles virus in the girl’s lungs. It is by far the oldest measles genome ever assembled. And by comparing that 1912 measles virus to more modern strains and to related animal viruses, the team pushed back the earliest likely date that measles could have emerged in humans to 400 B.C., 1,500 years earlier than estimated with previous genetic evidence. (A preprint of the paper, which has not yet been peer-reviewed, is available on bioRxiv.)

“With viruses, we don’t have fossils,” says Calvignac-Spencer. But in recent years, scientists have started rummaging through old pathology collections and using computational models to trace the otherwise lost history of viruses. Calvignac-Spencer thought he might do the same for measles, which is why he got in touch with the Berlin Museum of Medical History at the Charité. The museum was eager to collaborate and to find new uses for its extensive pathology collection.

Pathology samples are usually preserved in formalin, a chemical solution that halts decay by stopping all biological processes but also “glues” RNA to other molecules. (Viruses can use RNA or DNA as their genetic code; measles is an RNA virus.) To extract just measles RNA, the team would have to break those links to other molecules. The easiest way is with heat. “You boil the sample,” explains Calvignac-Spencer. The team boiled the lung sample at 98 degrees Celsius for 15 minutes. With that, they were able to extract RNA from the measles virus and reconstruct its genome.

To reconstruct an accurate history of measles, however, you need not one but many measles genomes, the older the better. So Calvignac-Spencer and his colleagues went searching for others. In a reference lab, they found another isolate from Czechoslovakia in 1960, which they also sequenced. They scoured the literature for the published genomes of 127 other more modern measles isolates. Finally, they compared all the measles genomes with those of its closest cousins: rinderpest, a since-eradicated virus that once devastated cattle herds, and peste des petits ruminants, a virus that still infects goats and sheep. Measles was probably an animal virus that spilled over into humans.

By modeling how these viruses would have evolved over time, the team concluded that measles and rinderpest likely diverged 2,400 years ago. Sometime after that—though this study cannot resolve exactly how long after—measles evolved to exclusively infect humans.

This divergence date would fit what many measles researchers have long suspected about the virus and the appearance of big, dense human cities. Measles is exceptionally contagious, but anyone who survives it becomes immune. In a small community, everyone will quickly catch measles and either die or develop immunity; the virus runs out of victims to infect. But in a large, dense city, where susceptible babies are being born all the time, measles can sustain its chain of transmission, says Ottar Bjornstad, an epidemiologist at Penn State. Around 2,300 years ago, cities in North Africa, India, China, and Europe began growing to hundreds of thousands of people—large enough, presumably, to begin sustaining measles. This may be the story of how measles came to infect humans.

“You can really have a glimpse of the past,” says Sophie Gryseels, a virologist at KU Leuven, who has used similar techniques to study the history of HIV from a 1966 formalin-fixed sample. (Gryseels was not involved in the measles study, though she currently works in the lab of Philippe Lemey, one of the paper’s co-authors.) The difference is that HIV spilled over to humans only about a century ago, and scientists have studied many related viruses that infect nonhuman primates. Our understanding of measles’ relatives is not nearly as detailed, and its spillover event was much more ancient.

And of course, extrapolating to events so far in the past, with so few inputs, is hard. The study makes use of the most advanced models available, but the data are limited. “Essentially you’re taking two points very, very close to you and trying to sight a target a really, really long [way] away,” says Michael Baron, a virologist and rinderpest expert at the Pirbright Institute, which specializes in diseases of livestock.

Meanwhile, newly sequenced rinderpest genomes might elucidate the evolutionary relationship between the two viruses, too. Since rinderpest’s eradication in 2011, Baron has also been sequencing rinderpest genomes as part of a “sequence and destroy” campaign. (Pirbright destroyed its final samples last June.) The campaign yielded some 50 additional examples of rinderpest genomes, and a paper is under review. Baron says rinderpest researchers have also gotten interested in finding formalin-fixed cattle tissue that might be hiding in an archive somewhere. Rinderpest research is strictly controlled to prevent the virus from escaping into the wild; formalin, conveniently, destroys the viruses so they are no longer infectious but still retain RNA. It’s a way to study actual rinderpest viruses but safely.

Measles, rinderpest, and PPRV are all part of a larger group of highly contagious viruses called morbilliviruses that infect animals ranging from dogs to seals to dolphins. The conventional wisdom says rinderpest most likely spilled over to humans and evolved to become measles, the same way that bird flu can sometimes infect humans. But Edward Holmes, a virologist at the University of Sydney, says the story may ultimately be more complicated. “We are sampling more of the animal world and we’re finding so many more animal viruses out there,” he says, “We’ve sampled just a tiny fraction.” Perhaps other animals are also reservoirs for measles-like viruses out in the wild. Perhaps these viruses have jumped from species to species in ways we haven’t understood. Since rinderpest has been eradicated, cattle have started to get PPRV, the morbillivirus specialized for sheep and goats. And experts have suggested that skipping vaccinations could put humans at risk for not just measles but related morbilliviruses that normally infect other animals.  

Calvignac-Spencer is still working his way through the Berlin Museum of Medical History at the Charité’s lung collection. His team informally cataloged the several hundred specimens in the lung room, and they’ve also found cases of the Spanish flu, which killed 50 to 100 million people in a pandemic starting in 1918. Yet more viral genomes may be hidden in the museum basement.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.