Blood Matters

by Masha Gessen

  The scientists try to be cautious, but the specter of the Khazars haunts Jewish population genetics. “I cannot argue with somebody who is telling me, Look, we know that the Khazarian king lived in Eastern Europe,” Behar tried to explain to me, pointing out that this was not his argument exactly but just an argument he would be hardpressed to refute. “I mean, I’m telling you what people told me. We know that the royal family converted to Judaism, okay? If I was the king, I wouldn’t want to become an ordinary Israelite, I would want to become a Levite or something. So is it possible that our day’s Levites are the great-great-great-sons of one Khazarian nobleman or king who converted to Judaism? It’s an explanation. We can’t do much to approach the genetics, because we don’t have anyone who says, ‘I’m the descendant of Khazarians’ today. It’s possible? Yes. Can I prove it? No. Should we be cautious? Yes. We are not coming out with the title, ‘The Levites Are Descendants of Khazarians.’ We don’t think it’s appropriate.”

  In fact, a couple of villages populated with people who call themselves Khazars exist in Lithuania and in the Crimea. They do not, however, practice Judaism, and they have never agreed to be tested genetically. But if they did, and they did not show genetic similarities to the Levites or other Ashkenazim, this would do nothing to prove or disprove the Khazar “event” theories. So the scientists try to keep their Khazar theories to themselves. Only Ariella Oppenheim was either brave or careless enough to put the possible Khazar explanation in print in a scientific paper. Well, she and a group of Harvard doctors who reported, in the mid-1990s, that Jewish and Iranian patients with a ghastly skin disease called pemphigus vulgaris had similar genetic traits.

  ***

  DNA evidence is tricky that way: It seems that, outside the courtroom, it can serve only to corroborate, never to disprove. “I talked to one historian,” Oppenheim told me, “who said she was so disappointed because they thought we were going to reveal something completely new that they never thought about, never knew about, and here we were going to prove it by genetics. But we did not, and I feel very good about it. If we found something that was not in recorded history, I would not feel quite good about it, quite uncertain about it.” Her team’s work showed, in essence, that Jews are Jews and are pretty much who we say we are. Nice, but not a historical sensation.

  Doron Behar and Karl Skorecki’s 2006 study showing that half of all Ashkenazim are the descendants of four women was neither a historical nor a scientific sensation. We all know—or we claim to know—that all humans have a single common mother. We also know that any two people in the world have common ancestry—a forefather and a foremother whom they share. That is the science. History is also fairly straightforward: Ashkenazi Jews were a small group that settled in the Rhine Valley around the seventh or eighth century. The small group grew to around eight and a half million at the turn of the twentieth century and more than ten million before World War II, in which about six million were killed, and doubled again by the turn of the twenty-first century, reaching about eight million people. With this kind of dramatic expansion—and with what we already knew about Jews, by and large, being Jews, it should not have surprised anyone that so many Ashkenazim could be shown to have come from so few.

  The scientists took eleven thousand blood samples and used them to isolate genetic signatures passed from mother to daughter to daughter (the mitochondrial DNA, unlike the Y chromosome, is passed on to offspring of both sexes, but a son’s children will inherit his wife’s mitochondrial DNA, so this effectively is a woman-to-woman thing). They compared the DNA of Jewish women to samples from what geneticists call the “host population”—the titular nations of the European countries in which Ashkenazi Jews have lived. It turned out that roughly three and a half million people, or 40 percent of Ashkenazi Jews now living in Europe, Israel, and North America, were the descendants of four women who had lived one thousand to fifteen hundred years ago, give or take a thousand.

  “I had discussions with my coauthors about where we were going to send this paper,” Karl Skorecki told me. “I was saying, ‘This is parochial, this will not be of great interest to the larger scientific community, not so much to the larger community.’ They were saying, ‘This is a finding which is absolutely stunning!’ I was saying, ‘It’s an interesting finding, but I don’t think it’s startling or stunning.’” Skorecki had used the same word—parochial—when he described the question posed by his Cohanim paper, which remains one of the most noticed genetic studies of our time. And he was wrong this time too. The study, published in the American Journal of Human Genetics, was one of those very rare scientific papers that make it to ABC News, MSNBC, and USA Today.

  Why? “It’s not an abstract concept,” suggested Skorecki. “It’s not a theoretical woman: These were real living, breathing women.”

  That was an obvious exaggeration. They had not lived or breathed in a thousand years, give or take as much. But Skorecki had a point. Knowledge of how many women there were, and when, gave scientists minimal grounds for theorizing who the women were. “It sounds like such a family had to have some sort of economic advantage,” said Behar. Previous studies attempting to link mitochondrial DNA to physical attributes that would have given the women a selective advantage—health, beauty, and extraordinary fertility would be three such traits—had failed. So the edge had to be socioeconomic. Perhaps our great-great-great... grandmothers were rich and this was why their daughters survived to marry well and have lots of daughters of their own (in which case the genetic-drift theory and the selective-advantage theory are, in fact, casually reconciled). Perhaps, like the Orthodox Jews of today, they followed a tradition of having many, many children. “It must have been something in the beginning,” said Behar. The beginning, he theorized, would have been enough: After a few (three, six, maybe ten) generations, the lineages would have gained enough of a foothold in the population to continue to propagate.

  So, many of us had a wealthy, nurturing great-great-great... grandmother. All she left us was her DNA signature. After the study came out, a New York Times reporter sent away for a DNA cheek-swab test kit from a commercial company (Skorecki’s original coauthor, Michael Hammer, and Behar both work with the company) and published a breathless article titled “Love You, K2a2a, Whoever You Are.” K2a2a was one of the four genetic signatures the researchers identified.

  There is a thin line between theoretical knowledge and not-so-theoretical knowledge. In 2001 the Oxford geneticist Bryan Sykes published a bestselling book called The Seven Daughters of Eve, claiming that all modern Europeans are descendant from seven women (other researchers later commented it may have been more like a dozen). Sykes also decided to name the alleged seven daughters, or mothers. The one whose signature began with K he chose to call Katrine. Thanks to Behar and Skorecki’s research, she had actually acquired enough distinguishable features for us to be able to say Katrine was an unlikely name. A Hanna or a Sara she may have been, or a Rakhel. Perhaps that is the magic trick of studies like these: Somehow, they make it all more specific, so the abstract foremother of all yields to the specific four mothers of us.

  The question is why we are so taken with this kind of specificity. Why do small numbers—four women fifteen hundred years ago, as opposed to a dozen women tens of thousands of years ago—hold such sway? “I don’t know what is it in us humans that makes us want to learn about similarities,” Doron Behar said, trying to answer my question why, being a medical doctor and the father of three small children, he chose to pursue a second, challenging, wildly time-consuming, and often tedious career as a geneticist. “I don’t know what makes us dig in the ground to look for the ancient—I don’t know. There is no easy answer for this, because I can’t really say that if you will dig in the ground and find a cave then you will get this, this, and this benefit.” Humans are driven to look for continuity. On some level, it may be part of our eternal search for immortality. No one feels fully mortal as long as his mother is alive. No one dies fully as long
as his children go on living. And perhaps if our great-great-great... grandmother’s DNA is maintained in our flesh and blood, that makes her, suddenly, not quite so dead, and us not quite so doomed.

  ***

  “I know that genealogy is the second most popular hobby in the United States,” mentioned Behar. This is one of those frequently repeated factoids that seem to bear no relationship to any statistics, polls, or other objective measures (I could find no such citation, in any case). Behar may be forgiven for his unfounded assertion. He was desperately looking for a way to explain the obvious to me, that people have always found this sort of thing interesting, and they always will. Americans are by no means unique in their yearning to explore their roots: Icelanders, for example, make it a national pastime, often maintaining meticulous family records going back many centuries. What may distinguish Americans from the world’s other potential genealogists, however, is their frequently woeful ignorance of their roots, which goes hand in hand with the potential for surprises to be found along the family tree. Americans constitute the perfect market for genetic ancestry tests, which became commercially available in the early 2000s. Within the first three or four years of their existence, these companies drew a couple hundred thousand clients.

  One of the pioneers in this business was a company called DNAPrint, one of a slew of biotech start-ups that began big around 2000 and soon fell on hard times. What saved DNAPrint was that its researchers stumbled upon the idea, and the technology, for performing genetic ancestry tests. DNAPrint is in Sarasota, Florida, a city of low houses huddled close to the road, mailboxes specifying “back” and “front” units, of pickup trucks and grazing cows—many, many cows grazing everywhere—and otherwise a conspicuous absence of human activity outside the islands that are the city’s tourist centers. At first glance, the nonfarm economy is limited to restaurants and proud white resort hotels. The building that houses DNAPrint is a modest low office structure along a wide dusty road, with nothing to the left and nothing to the right of the parking lot. But I saw a Jaguar in that parking lot as I drove up. Biotech money, I thought.

  “I don’t know,” chief research scientist Matt Thomas told me. “It certainly does not belong to anyone who works here.” In the first quarter of 2006 DNAPrint Genomics would post losses of nearly a million dollars. At one point during our conversation Matt called the company “Tony’s DNA shop in Florida”—Tony being Anthony Frudakis, a molecular biologist who had found some venture money in Florida and moved there to start a company in 1999. The idea was to locate genes that influence or determine the way people respond to various medications, which would have made the company invaluable to the health-care system and to humankind in general. In 2000—the year DNAPrint went public—by the Food and Drug Administration’s estimate, 2.4 million patients had adverse reactions to medication and more than 125,000 people had died as a result, making prescription-drug reaction a leading cause of death in the United States. Common chemotherapy drugs can lead to heart failure; anesthesia can kill unexpectedly. Florida, with its aging, relatively affluent, and consequently much-medicated population, seemed like a good place to do this sort of research.

  Frudakis and his team decided to begin their pharmacogenetic research with eye color. The idea was to conduct what is called a proof-of-principle study: Eye color, they thought, was like drug response in that it was a trait affected by more than one factor, but all of these factors were genetic. After he explained this to me, Matt Thomas said something that sounded almost impossibly obvious: “And what we quickly saw with the eye color stuff is, we had a number of samples from different populations around the world. And some of them—genetic locations and markers we were looking at, we saw strong segregation among the populations. So basically a particular marker, an allele or a version of the marker, was very prevalent among, let’s say, Europeans, and less so among Africans, or vice versa.” The scientists started talking with an associate professor at Penn State University, who was just getting into molecular anthropology, which is essentially measuring the differences between populations at the molecular level. “And this kind of opened up some new avenues for us, some new interests of ‘maybe the best way to understand drug response and the eye color project is to understand the world’s populations better.’ And that led to some of the developments in our ancestry testing. The idea behind the ancestry testing is, we have another way of stratifying a population.”

  In other words, by looking at gene sequences on their computer screens, Frudakis’s team arrived at the same conclusion as many physicians across the United States who had started asking patients about their ethnic backgrounds. The physicians did this because, for example, an Ashkenazi Jew who had lost a mother to premenopausal breast cancer—someone like me—had become a clear candidate for genetic testing. The Florida scientists thought that maybe figuring out that a particular drug response was more characteristic of one ethnic group or another would ease the search for genes that determined the reaction. In both cases, this is what Matt Thomas called “stratifying a population.” If that sounds perfectly obvious, it was not: It would be another three years before the African American Heart Failure Trial made headlines across the United States. The result of that trial, which was halted early because of excellent preliminary results, was medication certified for use in African Americans at risk for heart failure.

  The African American Heart Failure Trial enrolled over a thousand people who self-identified as African Americans. The only term less specific might be “human.” People of African descent come from the most varied gene pool on the planet: This is what makes geneticists think the human race originated on that continent. In African Americans, that gene pool was altered further by the admixture with the “host population,” which generally includes people from everywhere else. What is most striking is that, despite the imprecision of this particular race classification, statistics on the effect of, say, heart disease show a striking difference between African Americans and Caucasian Americans.

  Presumably, the different effect of both heart disease and medication on people of different races is the work of a set of gene variations that, when identified, will eliminate the need to classify patients by race. That will happen, no doubt, but it will take a while. In the meantime, researchers could at least attempt to break people up into smaller groups and try to answer such questions as whether people of North African extraction should be classified as African Americans for the purposes of these studies and these treatments. Someone from Morocco, Tunisia, or Libya would probably look like a Caucasian to most Americans and might, if living in the United States, self-identify as anything at all. But how should such a patient be treated by medical doctors? If DNAPrint could tell a drug manufacturer that much, it would make itself indispensable.

  Somewhere along this thought process—probably when DNAPrint ran out of cash—someone had the idea of using the results of this research to create a test, commercially offered to people curious about their ethnic background. DNAPrint teamed up with Mark Shriver, the molecular anthropologist at Penn State, who was starting to use gene sequencing to try to develop molecular measurements of the things anthropologists measure: things that make us who we are. I visited Shriver’s lab and talked to a group of young female graduate students who showed me how they did their research based on the material at hand: several thousand eighteen-to-twenty-year-olds who were willing to have themselves studied for ethnic markers. The technology was apparently simple: Shriver’s lab had collected DNA samples from various parts of the world. “Parts of the world” were fairly broadly defined: Sub-Saharan Africa, for example, was represented by Nigeria and the Congo, while East Asians were a combination of Han Chinese and Japanese, a few Koreans and a few Vietnamese. They compared the samples, isolating bits of gene sequence that seemed to differentiate one group from others, and designated these parts of the genome markers of a particular population. Then they started testing people, giving them results that would inform them that t
hey were, say, 80 percent European, 10 percent African, 9 percent East Asian, and 1 percent Native American. The 1 percent would most likely be what researchers call “noise in the test,” meaning either a mistake or the result of imprecise classification, while the percentages of African and East Asian might well surprise the subject.

  Amy Bigham, the senior graduate student in the lab, showed me around. DNA-testing equipment tends to fall into two categories: things that look like printers and things that look like toasters. There were also some unusual pieces of machinery, like a three-dimensional camera, a contraption of many lenses and lightbulbs on movable metal arms that allowed the researchers to photograph a subject’s head from all sides at once. The digital image could then be used to take all sorts of measurements of facial and other features. Amy cheerfully told me that in addition to taking subjects’ DNA samples and personal histories—their families’ ethnic backgrounds and their self-identification—they also used to take measurements of their skulls. Then she showed me poster boards hung all along the hallway. Subjects’ pictures were mounted on the boards, next to sheets with their self-reported data. Flipping up the sheets exposed the subjects’ test results, which were presented in the form of a pyramid, with the bulk of the subject’s ethnic makeup at the bottom and what might be statistical noise at the top. The boards posted in the hallway featured subjects whose results could be curious: someone who looked white but was largely of African descent, or someone who turned out to carry markers from most of the world. It all struck me as borderline distasteful, like the mention of the skull measurements. An image from a Nazi propaganda film showing the cranial measurements of a Jew immediately flashed through my mind.

  “You stopped that because it was kind of creepy?” I asked.

  “No.” Amy smiled, clearly confused. “Because our craniometer broke.”