A jumping gene that also does a useful job is supposed to be rare. In a paper published in Science on September 24, 2026, a team led from Cornell University argues that one has been sitting in human neurons all along: BC200, a short noncoding RNA gene that primates recruited roughly 40 million years ago and that still appears able to copy itself into new places in the genome.

The evidence that got attention is strange. Two fragments of BC200 were found inside molluscum contagiosum virus (MCV), a poxvirus that infects only people and usually causes harmless skin bumps. A human gene, in other words, ended up in the DNA of a virus. This post walks through what the researchers did, which numbers hold up, and where the coverage of the paper contradicts itself.

One note on sourcing. The study is peer reviewed (Science, DOI 10.1126/science.aeh3441), but the paper is paywalled and I could not open it. Everything below comes from the Cornell press release and from three secondary reports, so quantities are attributed to those reports rather than to the paper’s own tables.

What BC200 is, and why movement is the surprise

BC200 is a gene whose product is an RNA, not a protein. It is expressed mostly in neurons, where the RNA is abundant in dendrites, and it was first described in the late 1980s. Its job is still poorly understood. The leading suggestion is that it helps regulate how messenger RNAs are translated near synapses, but according to the Cornell account that remains unconfirmed.

The gene did not appear from nothing. A progenitor was co-opted from a transposable element, one of the DNA or RNA sequences that can insert copies of themselves elsewhere in a genome. ScienceAlert’s summary says the ancestral element settled in a common ancestor of monkeys and apes somewhere between 35 and 55 million years ago, and that it took on a cellular role around 40 million years ago. Roughly half of human DNA is transposon-derived, but nearly all of it is inert, worn down by mutation until it can no longer move.

That is the usual fate of a domesticated transposon. Once a host starts depending on a sequence, selection tends to strip out its ability to hop, because a hopping copy can break something. Cedric Feschotte, the senior author, is quoted in the Cornell release saying that for BC200 evolution has not pulled those two properties apart. The authors call it the first human gene known to be both mobile and needed for normal function.

How the team spotted it in a virus

The first hit is older than the paper. According to the Cornell release, Cheng Sun searched for human transposon sequences inside viral genomes in 2010, while a postdoctoral researcher in the joint Feschotte and Ellen Pritham lab at the University of Texas at Arlington. He used specialized software against genome databases and found two BC200 insertions in MCV. The group recognized that this was interesting and, by their account, did not publish.

A Human Jumping Gene Turned Up Inside a Poxvirus - detail view

Years later Sun and his graduate student Pu Gao, the first author, picked the project back up. The release does not describe the 2026 experiments in detail, so the exact pipeline is not something I can report. What the secondary coverage does say is that the team screened the available poxvirus genomes and found no matching elements in the other poxviruses examined, including a close relative that infects horses. That negative result matters. It makes a one-off contamination or a generic poxvirus habit less likely, and it points at MCV specifically.

The two viral sequences are not intact copies of the gene. ScienceAlert reports that one contains the complete Alu-derived domain of BC200, while the other is shorter and truncated at both ends. Fragmentary insertions are normal for retrotransposition, which often produces 5-prime truncated copies.

The numbers: 908 people, eight insertions

The claim that BC200 is still active today rests on population data. According to the ScienceAlert report, the team examined sequencing data from 908 people and found eight BC200-derived insertions that differ between individuals. Their distribution is informative.

Two were widespread across five continental population groups. A variant present on every continent is usually old, because it had time to spread before populations separated. One insertion was largely restricted to people of African ancestry, which fits a variant that arose in or near the root of human diversity and was left behind by out-of-Africa migrations, or that arose later within African populations. Five were rare: one carried by two people, four each seen in a single person.

Singletons carry the most weight for the argument. A copy that shows up in exactly one person among 908 is recent on an evolutionary scale, because there has been no time for it to spread. It does not prove the copy arose in that person’s parents or in that person, a distinction the secondary report is careful to draw. Even so, five rare insertions out of eight is the pattern you expect from an element that is still occasionally producing new copies.

The primate comparison points the same way. The coverage says that every primate species examined with a BC200 gene has accumulated dozens to hundreds of lineage-specific insertions. If each of those is a copy of a gene present in the ancestor, BC200 has been a steady source of new insertions across roughly 40 million years, which is a long time for a domesticated element to keep its mobility.

How a gene ends up in a virus

BC200 has no machinery for copying itself. Its sequence is derived from an Alu element, and Alu elements are non-autonomous: they borrow the enzymes of another transposon, LINE-1. LINE-1 encodes a protein that can bind an RNA, nick DNA at a target, and use the RNA as a template to write a new DNA copy in place. Per the ScienceAlert report, LINE-1 is the most likely supplier here, and the abstract summarized on Lifeboat likewise attributes the viral insertions to LINE-1 activity.

A Human Jumping Gene Turned Up Inside a Poxvirus - working context

That raises a practical question: how would a retrotransposition event reach a poxvirus? Poxviruses replicate in the cytoplasm of the host cell, not the nucleus, and they carry their own DNA machinery. The authors’ suspicion, again from the Cornell release, is that the jumps happened in skin cells, the only cells MCV is known to infect. This is a suspicion, not a result. Nobody watched a transfer happen. The reconstruction is from DNA evidence after the fact, which is how most horizontal transfers are inferred.

Two further observations support a real biological interaction. BC200 expression went up in cultured human fibroblasts infected with the virus, and the viral BC200-derived sequences are themselves transcribed into RNA. What that viral RNA does is unknown. The researchers say they plan to test whether MCV uses it to manipulate host cells.

Where the reports disagree on timing

The three summaries I read do not agree on how old the viral insertions are. The Cornell Chronicle story says BC200 jumped into MCV twice, around 100,000 years ago, since the emergence of Homo sapiens. IFLScience says the two copies appear to be separate transfers, thousands of years apart, dated to at least 25,000 years ago and a minimum of 31,000 years ago. ScienceAlert gives no age estimate at all and says only that rare human insertions suggest very recent activity.

These are not necessarily in conflict. Minimum ages and a rough ‘since modern humans’ framing can both be true. But a reader should not take a single figure from this paper without the paper’s own dating method in hand, and the Cornell release does not explain how its number was derived. I would treat ‘at least 25,000 to 31,000 years’ as the safer phrasing until the full text is checked.

The context the authors give is more solid. Transposons are known to move into viruses: in the late 1980s a moth transposon was seen inserting into a baculovirus in cell culture, and in 2007 a Japanese group reported a snake-derived element inside a rodent poxvirus. IFLScience calls the rodent gene BC1 the closest parallel to BC200. The new claim is that this has now been caught in the human lineage within recent history.

What the study cannot show

Start with what is inferred. The transfer into MCV is a reconstruction from sequences. It has not been reproduced in a dish, and the location of the event is a suspicion.

A Human Jumping Gene Turned Up Inside a Poxvirus - supporting scene

Second, detection in a person is not the same as a new insertion in that person. A rare variant could have been inherited from a distant ancestor. Separating the two needs parent-offspring data, which a population panel of 908 genomes does not supply.

Third, the disease connections are speculation. BC200 is abnormally expressed in some tumors, including breast cancer, and is in excess in Alzheimer’s brains. The release suggests mobility might contribute, but whether BC200 jumps in cancer cells or creates mutations there is listed as an open question. BC200 is also present at low levels in eggs and sperm, which would permit heritable insertions in principle, but the sources do not show that any such event has occurred.

Finally, nothing here says MCV has become more dangerous. The virus is described as rarely serious, and the researchers present the work as a window into genetic flexibility, not as a clinical finding.

Frequently asked questions

What is a jumping gene?

A jumping gene, or transposable element, is a stretch of DNA that can copy or move itself to other places in a genome. Roughly half of human DNA derives from them, though most are now inactive.

Is BC200 dangerous?

Nothing in the sources reviewed says so. The link to cancer and Alzheimer’s disease rests on abnormal expression levels and is speculation about mobility, not a demonstrated cause.

Is this study peer reviewed?

Yes. It was published in Science on September 24, 2026. I could read only press releases and secondary coverage, not the paywalled paper.

Can the virus pass the gene back to people?

The sources do not address that. The finding is that human-derived sequences exist in the virus, not that the virus carries them into new hosts.

The tidy version of this story is that a human gene hitched a ride on a virus. The more interesting version is that BC200 may be a working neuronal gene and an active retrotransposition source at once, a combination genomes are supposed to resolve in favor of one or the other. If the 908-genome pattern survives scrutiny in larger panels, the next tests are concrete: sequence parents and children to catch an insertion as it happens, and ask whether the viral BC200 RNA does anything inside an infected skin cell. Until then, read the dates and the disease angles as hypotheses, and the discovery that a human neuronal gene sits inside a poxvirus as the part that is firmly documented.

Sources