one in Ashkenazi Jews and another in Hispanics in New Mexico that can be traced to descendants of 16th-century Spanish colonists. These mutations usually happen in one of three genes individually named CCM1, CCM2, and CCM3. Any of them can make the walls of certain blood vessels unusually weak. In the brain, these weakened blood vessels can grow into mulberry-shaped malformations full of blood. CCMs are rare, but they do sometimes form in people with no family history of them; patients with an inherited CCM mutation, however, can have multiple malformations at once. Any single CCM may stay silent, but having so many increases the odds of at least one leaking or bursting. And they tend to start doing so when people are between the ages of 20 and 50.

The Bowlin family’s malformation can indeed be explained by a mutation—a newly discovered one. Theirs is a deletion in the gene CCM2, and this mutation is dominant, meaning it does not skip generations and has a 50 percent chance of being passed on. Marchuk’s lab didn’t just find it in Bettina’s donated tissue, though; the scientists also found it in seven other unrelated patients all at the same time in 2007. How odd, Marchuk remembers thinking, to see the identical deletion eight times in a row. But, he says, “we couldn’t see any common ethnic heritage, and those eight families, we couldn’t connect them in any way ourselves.” The deletion happens to occur in a natural recombination hot spot—where DNA gets cut and pasted back together—so he figured it might just have happened independently in those families.

Then things got stranger. More families with the exact same deletion started showing up after the Angioma Alliance began a genetic-testing program for people with suspected CCM mutations. In the first year alone, a quarter of participants tested had the same CCM2 deletion. The affected families lived across the U.S., with the exception of the Northeast, but they were concentrated in the South and Midwest. (The Bowlins are from Mississippi.) “About two years in, I could say that if you live in Oklahoma, and I’m about to test you, I can pretty much predict which mutation you’re going to have. Or if you live in Mississippi, Alabama, Louisiana—same thing,” says Connie Lee, Angioma Alliance’s president. The distribution clearly wasn’t random, which suggested the deletion wasn’t showing up by chance. The families probably were related. They just didn’t yet know how.

In search of a connection, the families and Lee created a Facebook group to share stories and names. They scoured census, birth, marriage, death, and other public records. They took AncestryDNA tests. A professional genealogist even came on board and looked for records in person at the Family History Library, in Salt Lake City. In late 2018, a new family joined whose ancestry allowed the group to link up two family trees. That was a crucial clue.

Over the next few months, the families managed to trace their rare mutation back 250 years to a single North Carolina couple born in the 1760s: Matthew Malachi Rushing and Sarah Mae Harrell. Group members couldn’t test the long-dead couple’s DNA directly, of course, but they could infer from the pattern of inheritance in living descendants. Bettina, Margaret, and Susan are the couple’s great-great-great-great-grandchildren. Their migraines, their seizures, their tingling and weakness in the limbs—it all goes back to the DNA of this one couple in colonial America.

The Duke researchers have now studied 27 families with this CCM2 deletion. Whenever Carol Gallione, a researcher in Marchuk’s lab, analyzed the CCM2 gene from a patient in one of these families, she found the same sequence. “They were all interchangeable,” she says, which is what you would expect if they all came from the same ancestor. A recent preprint, which has not yet been peer-reviewed, describes how scientists and patients together traced the history of this mutation.

The precise origin of the mutation likely goes back further than Rushing or Harrell, as not all of the families today with the CCM2 deletion are directly descended from the couple. A few have also found common ancestors in 1600s Virginia. But the mutation can’t have gone back too much further, because it seems to have originated in America. Geneticists in the U.K., Europe, and Brazil have looked for but not found the deletion. “It certainly looks like it’s American-specific,” says Jonathan Berg, a geneticist at the University of Dundee, in Scotland, who was not involved with this study but has worked with Marchuk in the past. As the mutation spread across the U.S., it followed the migrations that shaped larger American history. It began on the East Coast, before spreading south to the Gulf Coast and west to Oklahoma and then California. The families saw in the mutation’s dispersal the imprint of old family stories passed down over the years. “Yes, this is when the Mississippi land grants happened,” Lee remembers hearing the families say. “And yes, my father was one of the people who was following where the lumber harvests were happening … going back and forth along the coast of the Gulf.” The gene’s history is the families’ history is America’s history.

Evolution has not selected out the CCM2 deletion over the years because its consequences tend to manifest only in adulthood, when patients have already had children. So the mutation has persisted for centuries. Some of the family members with the CCM2 deletion were so distantly related, Marchuk says, that they shared little DNA beyond the mutation. This sliver of DNA connected them all, its consequences lurking in their brains.

When Bettina learned she needed brain surgery in 2004, she sat in her car and prayed for some kind of purpose. The ensuing research from the tissue she donated after that surgery has been exactly that, she says: “the answer to my prayers.” With genetic testing, patients can find out if they’re at risk for these  brain malformations and get them monitored. Multiple medications are now in clinical trials for controlling the condition.

But mysteries about these mutations remain too. A CCM mutation does not necessarily guarantee a lifetime of brain bleeds. Different families might have inherited the same mutation, “yet we see striking differences,” says Helen Kim, the director of the UC San Francisco Center for Cerebrovascular Research, who studies the CCM mutation predominant in New Mexico. (Kim was not involved with Marchuk’s study, but she is on the Angioma Alliance’s scientific board.) Genes may not be enough; environmental factors might also play some role.

This pattern is clear even in the Bowlin family. Jerry, the father, is now 82 years old. A recent MRI for an unrelated issue found malformations in his brain. But, he says, “to my knowledge, I’ve never had one symptom.” He’s had headaches, of course, like everyone, but nothing like the migraines, seizures, or brain bleeds experienced by his daughters. The story of this CCM2 deletion is a story about the power of genes, but also their limits.