French, Russian, and German researchers thawed sediment from nine sites of Siberian and Kamchatkan permafrost and revived thirteen previously undescribed viruses, the oldest of them from ground beneath a Yakutian lake bed dated to more than 48,500 years, and the reason every one of the thirteen can infect nothing but amoebae is a decision the team made before the first sample went into the dish

Thirteen viruses that have not circulated on Earth since before the last Ice Age’s coldest stretch are now, technically, alive again. A team led by Jean-Michel Claverie and Chantal Abergel at Aix-Marseille University, working with Russian colleagues from the Melnikov Permafrost Institute and other institutions, and German permafrost scientists from the Alfred Wegener Institute, pulled the samples from nine locations across Siberian and Kamchatkan permafrost and coaxed the ancient particles back into active replication in a lab dish. The oldest, a giant virus the team named Pandoravirus yedoma, came from sediment more than 16 meters beneath a thermokarst lake at Yukechi Alas in central Yakutia, radiocarbon-dated to over 48,500 years old. The strange part is not any of that. It’s that every single one of the thirteen, across five separate viral families, turned out capable of infecting exactly one thing: amoebae.

That is not a coincidence or a discovery about what these viruses are inherently limited to — it’s the result of a rule the researchers built into their methods before they ever opened a sample.

Nine sites, one very old lake bed, and a five-family haul

The study, published in the journal Viruses in February 2023, describes sampling seven ancient Siberian permafrost cores, a Lena River sediment sample, and one modern surface soil sample from Kamchatka as a comparison point. Out of that set, the team isolated 13 new viral strains spanning five clades: pandoraviruses, cedratviruses, megaviruses, pacmanviruses, and a new strain of pithovirus, the family first revived by the same lab back in 2014. Beyond the 48,500-year Pandoravirus yedoma strain, other samples dated to more than 27,000 and 28,600 years old, meaning the youngest viruses in the batch had still been dormant since well before agriculture existed anywhere on the planet.

Getting a virus that old to replicate again sounds harder than it is, mechanically. Permafrost is an unusually good preservative: constant sub-zero temperatures, no light, and low oxygen slow decay to something close to a standstill. The genuine difficulty was deciding what, exactly, the revived particles should be allowed to infect — not reviving them in the first place.

Why the answer was decided in advance

Every one of the 13 viruses was isolated using a single host organism: Acanthamoeba castellanii, a free-living amoeba common in soil and freshwater that happens to be an excellent, and epidemiologically inert, lab culture system for giant viruses. The paper’s own framing makes clear this was the point, not an accident: as the authors put it, the viruses recovered “belong to five different clades infecting Acanthamoeba spp.” precisely because of “the host specificity imposed by our protocol.” In plain terms, the team only exposed each sample to amoeba cultures, so the only viruses that could be detected at all were ones already capable of infecting amoebae. Anything in the sediment that might have targeted a mammal, a bird, or a human cell simply had no host present to reveal itself.

Claverie has been explicit about why that constraint exists. “Our protocol is to put amoeba cultures in the lab in contact with various samples, in the hope that they will contain viruses capable of infecting amoebas,” he told Live Science, adding a flatter, more direct version of the same point: “We will never risk isolating a virus eventually capable of infecting modern mammals.” The paper itself makes the comparison explicit, noting that the biohazard from reviving amoeba-only viruses is “totally negligible” next to the far riskier work of hunting for pathogens directly inside the frozen remains of mammoths and other ancient mammals, which is why that specific line of research is confined to a biosafety level 4 facility in Novosibirsk rather than an ordinary university lab.

Every choice the team made before opening a single sample was a choice about what they were willing to risk being wrong about.

The part that actually worries the researchers

That caution isn’t really about this batch of 13 viruses. It’s about what else permafrost is holding, and how much more of it is on its way to thawing. Claverie has made this argument before: back in late 2022, months before Pandoravirus yedoma’s record-setting age was even published, he told Newsweek, “If amoeba viruses can survive in these conditions, there is no reason for other viruses not to survive as well” — the same underlying logic the yedoma finding would go on to reinforce.

The paper’s own conclusion frames the risk in terms of exposure rather than certainty: nobody can currently estimate how long a thawed ancient virus would stay infectious once it hit open air, or how likely it would be to encounter a suitable host in that window. What has changed, and what makes the question worth asking at all, is the denominator. Arctic permafrost is thawing faster as the region warms, and industrial development, mining, and new settlements are sending more people into exactly the areas where 48,500-year-old ground is being exposed for the first time since before modern humans reached most of the planet. The amoebae were never really the point. They were the deliberately narrow lens the researchers chose so that the first question, can something this old still replicate at all, could be answered without opening a second, much harder question before anyone was ready to.