“Are you sure it’s not twins?”
The question came from a woman I’d never spoken to before, both of us standing on our pedals mid-class in a dark spin studio that smelled like clove deodorant and someone else’s protein shake. I was six months into carrying my second daughter, and my shirt had ridden up over a stomach that had stopped consulting my due date and started making its own decisions.
I laughed, mostly because crying into my water bottle felt like the worse option, and told her no, just one very ambitious baby. She looked unconvinced, and honestly, she wasn’t wrong to be amazed. She just had the wrong culprit. What was actually going on in there had nothing to do with twins and everything to do with a virus.
Roughly eight percent of the human genome is made of leftover retrovirus. Richard A. Stein, a scientist at New York University, put the figure plainly in a 2023 review of the research: “Approximately 8% of the human genome […] comprises sequences of viral origin that are known as human endogenous retroviral elements (HERVs).” Those sequences got in through the oldest trick a retrovirus has. It infects a cell, splices its own genetic code into the host’s DNA, and if that cell happens to be a sperm or an egg, the virus’s instructions get copied into every generation that follows. Not as an infection anymore. As inheritance, passed down the same way eye color is.
Most of that inherited viral code just sits there, mutated past usefulness, taking up space without doing anything. But every so often, evolution picked up a stolen gene and put it back to work, and the placenta is the clearest case of it happening. Somewhere in deep mammalian history, an ancestor caught a retrovirus whose outer coat protein was built for one job: fusing itself onto a cell wall so the virus could break in. That same fusing trick got repurposed to build the layer of tissue that lets a fetus and a parent’s bloodstream sit close enough to trade oxygen and nutrients without actually mixing.
The gene is called syncytin, and molecular biologist John McCoy, who worked on some of the earliest syncytin research at the Genetics Institute, described the discovery to NOVA in terms that still sound like they belong in a lab thriller rather than a pregnancy: “This was a bona fide retroviral envelope protein that had somehow been captured during evolution and been trained to operate in human biology.”
The not-mixing part matters more than it sounds. A fetus is, immunologically speaking, half foreign tissue, carrying genetic material from two people instead of one. Left unchecked, a parent’s immune system would treat that the way it treats anything else it doesn’t recognize. The same NOVA reporting on syncytin describes the layer it builds as a barrier that keeps maternal and fetal blood from mixing outright, since mixing them carries the risk of a fatal immune reaction. A protein whose original purpose was helping a virus sneak past a cell’s defenses ended up, generations later, helping a parent’s body tolerate the one intruder it’s supposed to love.
Say “every mammal” out loud and a biologist will want to add an asterisk. Marsupials build a much simpler placenta, and egg-laying mammals like the platypus skip the structure altogether. But among the placental mammals, meaning nearly everything most people picture when they picture a mammal, the pattern gets stranger the closer you look. Different lineages caught and repurposed their own separate viral genes at different points in evolutionary history. Rabbits ended up with one, primates with another, rodents with a third, each pulled from an unrelated viral infection and each landing on roughly the same solution.
A team led by French virologist Thierry Heidmann, writing up the rabbit version in the journal Retrovirology, called it evidence that “retroviral infections have resulted in the independent capture of genes that have been positively selected for a convergent physiological role.” Nobody engineered the placenta once and handed down a single blueprint. It got built, more than once, out of whatever viral spare parts happened to be lying around at the time.
I think about that spin class more than the moment probably deserved. The weight gain that made a stranger blurt out “twins” was real, and so was feeling self-conscious about a body that had, in a matter of months, stopped resembling the one I recognized. I knew it was temporary. Knowing that didn’t make it any less humbling to catch my own reflection mid-pedal and not quite recognize the shape looking back at me. I’m not a biologist, and nothing here is me pretending otherwise. I just find it strange and kind of wonderful that the discomfort had a backstory running tens of millions of years deep, to some ancestor’s infection that never fully cleared.
A less generous version of this story could read as unsettling, a virus quietly rewriting what a body is for. I don’t experience it that way. Every family I know of, on every continent I’ve lived on, runs on inheritances nobody asked permission for first: a grandmother’s temper, a father’s stubbornness, a language picked up before anyone could consent to learning it. A stolen viral gene from tens of millions of years ago just happens to be the inheritance sitting furthest back in line.
My daughter arrived in July, syncytin and all, the same way every one of her ancestors did going back further than anyone bothered naming. She has no idea her existence runs through a virus. Neither did I, honestly, until I went looking for it. It doesn’t make the recovery any faster or the newborn nights any shorter. It just means that somewhere underneath all of it, something ancient and uninvited is still, quietly, doing the one useful thing it ever learned how to do.
