She Has Her Mother's Laugh: The Powers, Perversions, and Potential of Heredity (25 page)

But which family? No one had preserved any tissues from the Romanovs from which scientists could isolate DNA in order to make a precise match. Living relatives would have to stand in for them.

To make the match, Gill took advantage of a special set of genes that lie beyond our chromosomes. They lurk inside mitochondria, the pouches where our cells generate fuel. Each mitochondrion carries thirty-seven genes of its own, which encode proteins essential for its tasks. Mitochondria also divide on their own, making new copies of their own DNA without any meiosis.

What makes mitochondrial DNA especially attractive to geneticists is the way in which it is passed down from generation to generation. Both eggs and sperm contain mitochondria. But if a sperm manages to make contact with an egg, it produces
enzymes that shred its own mitochondrial DNA. The mother's mitochondria, and only the mother's mitochondria, becomes the mitochondria of her child.

This quirk means that mitochondrial DNA can act as a record of our maternal ancestry. Meiosis scrambles chromosomes from one generation to the next. But we inherit a precise replica of our mother's mitochondrial DNA. What's more, your mother got her mitochondrial DNA from your
grandmother, who got it from your great-grandmother, and so on back through more generations than even the most stubborn child can ask about. Each time mitochondria duplicate their DNA, there is a minuscule chance that it will mutate. That new mutation will be inherited down the maternal line in future generations. If a woman's female descendant picks up a second mutation, the mitochondrial DNA will now get passed down with both distinctive mutations. Relatives can be joined by this mitochondrial record of their shared ancestry.

Like the Habsburgs before them, the Russian Tsars were tightly bound by marriage to the other royal families of Europe. Tsarina Alexandra, for example, was the daughter of Princess Alice of England who in turn was the daughter of Queen Victoria. Tsarina Alexandra thus inherited Queen Victoria's mitochondrial DNA. And Alexandra passed it on in turn to the Romanov princes and princesses.

Looking over royal pedigrees, Gill realized that there was someone alive who also inherited Victoria's mitochondrial DNA: Prince Philip, the Duke of Edinburgh and the husband of Queen Elizabeth II. (Prince Philip is the great-great-grandson of Queen Victoria, through a line of female ancestors.) Gill contacted Philip, who agreed to provide his DNA for the research.

Gill found that
Philip's mitochondrial DNA matched the genetic material in the remains of one of the Yekaterinburg adults, along with all the children. This result indicated that the adult was Alexandra. The remains of the other adult in the pit had a different sequence of mitochondrial DNA. Gill found that it matched genetic material from a relative of Tsar Nicholas.

When Gill and his colleagues published their results in 1994, it seemed to many observers that they had tied the Romanov mystery up in an especially neat bow. In 1998, the skeletons were interred in the Cathedral of Saints Peter and Paul in St. Petersburg. Yet, even after the burial,
some skeptics questioned the identity of the bones. They raised the possibility that someone else's DNA had contaminated the equipment used to study the bones. If Avdonin really had found three of the five Romanov children, then what had become of the other two? The skeptics speculated that the bones in the shallow pit belonged to relatives. Given how many Russian
aristocrats were slaughtered at the time, it seemed like a plausible alternative explanation.

Archaeologists continued to study the area where Avdonin had found the shallow pit, and in 2007, more bones turned up, 230 feet from the original grave. Russian and American anthropologists inspected forty-four bone fragments and teeth from the second site and concluded that they came from at least two individuals. The shape of the remains indicated that some belonged to a girl in her late teens, while the others probably belonged to a boy between twelve and fifteen years old. The silver fillings in their teeth indicated they were aristocrats.

Once again, Gill examined the remains, this time working with researchers from the US Armed Forces DNA Identification Laboratory. They also extracted more DNA from the bones originally found by Avdonin. Once again, DNA from the original five skeletons showed they were parents and children. And the two new skeletons belonged to the family as well. At last all seven of the Romanovs were reunited, through a genetic genealogy.

As scientists learned how to analyze larger pieces of DNA, they could uncover more variants joining people together in lineages. They could go beyond close cousins and join people sharing common ancestors who lived thousands of years ago.

According to the Bible, Aaron became the first Jewish priest some 3,300 years ago, and the designation was passed down from fathers to sons since then. Today, many people with surnames like Cohen and Kahn believe themselves to be descendants of those priests, known as Cohanim. In the 1990s, Michael Hammer, a geneticist at the University of Arizona, set out to search for evidence of the Cohanim by studying the Y chromosome, which fathers pass down to sons. Because the X and Y chromosomes do not cross over like other chromosomes, the Y behaves like a male version of mitochondrial DNA, staying nearly identical from one generation to the next.

Hammer and his colleagues tested the story of the Cohanim by getting
cheek swabs from 188 Jewish men, 68 of whom had been told by their
parents that they belonged to the priestly line. The scientists extracted DNA from their cells and examined mutation-rich regions in the Y chromosomes. Hammer and his colleagues found a single mutation in significantly higher numbers among the self-identified Cohanim than in other Jewish men. The Cohanim, they concluded, inherited their Y chromosome from a common male ancestor.

In later years, Hammer and his colleagues examined Y chromosomes from more men—both Jews and gentiles. A lot of the Jewish men turned out, once more, to share a close common male ancestor. But others had different mutations that pointed back to other men in the past. In 2009, when
Hammer and his colleagues published their new research, they proposed that the priestly line started out just as the ancient stories said. But once the Cohanim tradition emerged, other Jewish men, with different Y chromosomes, somehow became priests as well.

At first, reading DNA was such an expensive undertaking that only experts such as Gill and Hammer could do it, and only for scientific research. But the cost dropped so quickly that it became economical for Hammer and others to launch companies that could provide genetic genealogy on demand. People simply spat into a tube that they mailed to the companies, which extracted the DNA inside and compared it to the growing databases of genetic variations in humans. The companies started off looking at a few regions of mitochondrial DNA and the Y chromosome, reporting back about where on Earth a customer's combination of mutations was most common. In later years, they cast a wider net, scanning genetic markers across all the chromosomes. Taken together, the markers made up less than a thousandth of the entire human genome. But they varied so much from person to person that they could reveal clues about people's ancestry or even link them, like the Romanovs, to their relatives.

Some customers took the tests in the hope of finding cousins or connecting themselves back to famous ancestors. Others hoped to reach back across the oceans to places where they came from. Europeans searched for their Viking ancestors. African Americans could leap beyond slavery's void. In 2016, a remake of
Roots
aired on the History Channel. LeVar
Burton, who had played Kunta Kinte in the original version, now served as an executive producer. To promote the new version of
Roots
, Burton took a DNA test from 23andMe, along with Malachi Kirby, the actor playing Kinte in the remake.

“I've always felt there was
a piece of me missing,” Burton said in a video. As he studied his results on an iPad, he looked deeply moved. “Who I am did not just begin here,” Burton said. “To have the proof in my hands is just powerful.”

I had heard much the same thing from other people, both friends who got themselves tested and people who have read some of the articles I've written about genetics. One reader told me how she had spent years using historical records to trace her ancestry to Jamaica and Ghana. She then told me that a genetic test showed she descended from two peoples in particular, the Akan and Guan. “In my family, I am tall and my siblings and parents are short,” she told me. “It was when I received DNA and folk history that showed a blending of Akan and Guan ancestors that I understood the appearance of tall and short in our family.”

I wondered what I might find in my own DNA. Was I carrying a molecular version of
History of the Goodspeed Family
? Would a Zimmer version hold more surprises? I wondered what collection of genetic variants I had inherited from those ancestors, how they had influenced my fate. I thought back to the visit Grace and I had had with a genetics counselor when we had floundered our way through a family history. At the time, the first human genome project was still under way, at a cost of $3 billion. Fifteen years later, my friends were giving ancestry tests as birthday presents. In the fifteen years since my last visit with a genetics counselor, the gene BRCA1 had become a medical celebrity. Certain mutations to the gene drastically raise the odds that a woman will develop breast and ovarian cancer. Those mutations are especially common in Ashkenazi people. I knew of at least one woman on my father's side of the family who had had breast cancer. A friend of mine who had a BRCA1 mutation was dying of breast cancer at age forty-eight. Had my daughters inherited that fate from me? If I found out, how would I tell them?

These questions were humming in my head when an e-mail turned up in my inbox. A geneticist named Robert Green invited me to a meeting. “This should be an extraordinary and very select learning experience if you can make it,” he promised me.

The meeting would be on the future of genomes in medicine. Green and other scientists would give talks about how they were using genomes in their research today, and how they might use them in the future. People coming to the meeting could opt to pay to have their genomes sequenced. For $2,700, a gene-sequencing company called Illumina would determine all 3.2 billion base pairs in a person's DNA. Clinical geneticists would then look at the variants people had, searching for mutations linked to 1,200 diseases—some familiar, such as lung cancer, and others obscure, such as cherubism. (Don't be fooled: Cherubism doesn't make you look like an angel. It fills your jaw with cysts.)

I signed up. I was attracted not by the prospect of a medical report; I wanted to get my hands on the raw data itself and find some scientists to help me explore it. Before Green's invitation, I could only guess about my DNA from the stories my relatives told. Now I'd be able to read my genetic heredity, down to the letter.

R
OBERT
G
REEN
and I stood inches apart. His eyes scanned across my face, from ear to ear, from forehead to chin.

“What I'm doing,” he murmured, “is looking for any facial features that would suggest an underlying genetic illness.” He looked me over as if I were a horse he was thinking about buying. “The shape of your eyes, whether your ears are low set or not,” he said. “The complexity of your ears.”

Getting my genome was turning out to be a lot more complicated than I had expected. I could not simply spit into a tube and mail it off to a company like 23andMe. In 2007, 23andMe began providing reports on DNA directly to consumers. For $999, they would identify the variants at half a million sites in a person's genome, analyze them for clues to their ancestry, and even supply a report about how the variants influenced risks for disorders ranging from diabetes to Alzheimer's disease. Their service was a profound leap from conventional genetic tests. They had to be approved by the FDA and ordered by doctors. Now 23andMe was delivering information straight to customers. In 2013, the FDA told 23andMe to stop selling unvalidated tests or face the consequences. In response, the company cut back their reports to ancestry and nothing more.

Other companies, such as Illumina, took notice. To get my genome from them, I would have to get a doctor to order it for me as a medical test.
Green, who had originally invited me to get my genome sequenced, also agreed to sign for the test. First, however, he would put me through a thorough, old-fashioned genetic exam—the kind that Lionel Penrose might have given in the 1950s.

“Future clinicians may judge this to be unnecessarily cautious,” Green told me. “But there is no standard for how we do whole genome sequencing. So this is how I've decided to do it.”

I had taken the train to Boston and made my way to Brigham and Women's Hospital for the exam. I first sat down with a genetic counselor named Sheila Sutti, who took out a form entitled “Family History.” She began asking about my relatives. As we spoke, she filled the page with circles and squares, slashing some of them with the diagonal of death. She noted allergies and surgeries. Question marks recorded the many times I shrugged my shoulders in ignorance. Sutti drew a network of symptoms and uncertainty. When I looked at the form, I could not see any signal of heredity.

Green arrived just as Sutti was finishing up. A looming, silent medical student trailed him. Green peered at my face through his narrow frameless glasses. He was taking advantage of the fact that genes play many different roles in our bodies. A hereditary disease that causes hidden damage to the nervous system may also disrupt the development of the face, leaving behind clues that a geneticist can spot with the naked eye. Green then asked me to walk back and forth from wall to wall. He crossed the arms of his white lab coat as he looked down at my feet, sizing up my gait.

Green told me these conventional exams didn't reveal any signs that required a test for a specific disease. He signed the request for my genome, and Sutti led me to another wing of the hospital, where a phlebotomist slid a needle into my arm. I watched blood glide like scarlet motor oil out of my arm and into three tubes.

The tubes were shipped across the country to San Diego, where Illumina's technicians cracked open my white blood cells and pulled out my DNA. They blasted the molecules with ultrasound, shattering them into fragments, and then made many copies of each one. Adding chemicals to the fragments, they were able to determine their sequence.

Now they had to assemble these fragments together like the pieces of a jigsaw puzzle. Just as a puzzle solver can use the picture on the box lid as a guide, the Illumina team consulted a reference human genome to figure out where each of my fragments had come from. Some fragments were too enigmatic to locate, but overall, Illumina was able to rebuild over 90 percent of my genome.

From one person to the next, human genomes are mostly identical. But in a genome stretching over three billion base pairs, the tiny fraction of DNA that varies adds up to millions of differences. Most of these variations are harmless. But some can give rise to a disorder such as PKU. Others raise the risk of more common conditions like cancer or depression. Illumina's clinical geneticists searched my own collection of variants for any especially worrying ones. A few weeks after my visit to Brigham and Women's, Sutti called me with the results.

“The reason we're doing this over the phone and not in person is that we didn't find anything of clinical importance,” she said. “You had a very benign report, Carl.”

Sutti told me that I didn't have any dominant mutations known to cause diseases with just a single copy. Nor had I inherited two copies of a dangerous recessive mutation. I did find out a few useful things about my health, though. The sequencing revealed variants that could affect the way I respond to certain medicines. If I ever get hepatitis, I know I shouldn't get treated with a combination of interferon and ribavirin.

And, like all humans, I'm also a carrier. That is, I carry single copies of recessive variants. If my children inherited the same variants from both me and Grace, they might develop genetic diseases. In the early 2000s, when Grace and I became parents, DNA sequencing technology was far too crude for me to get a full catalog of my carrier variants. The best we could hope for was to have our daughters tested for a few diseases, such as PKU.

It turned out I'm a carrier for two genetic disorders I never heard of: one called mannose-binding lectin protein deficiency, the other familial Mediterranean fever. I had to do a little research to understand this particular inheritance. I learned that mannose-binding lectin protein deficiency weakens the immune system, leaving babies to develop disorders such as
diarrhea and meningitis. Familial Mediterranean fever, the result of mutations to a gene called MEFV, causes people to suffer from painful bouts of inflammation in their abdomen, lungs, and joints.

I don't know which of my parents I inherited those mutations from, but I'd bet that I got my faulty MEFV gene from my father. It is most common among people of Armenian, Arab, Turkish, or Jewish descent. It's far rarer in other ethnic groups, like the Irish—from whom Grace descends. It would be extraordinarily unlikely that she would have a faulty MEFV gene, too. At worst, my daughters are carriers like me.

And that was all. After more than a century of advances in genetics, I got a glimpse at my genome, something that had been impossible until recently, and there wasn't much for Sutti and me to talk about. A week after our phone call, I took the train to Boston to attend the “Understand Your Genome” meeting. At lunch, an Illumina representative logged me into a secure web page that elegantly displayed my results. I could compare my own genome to the reference genome, displayed as two rows of colored letters. Where my DNA differed, the colors were brightly mismatched. Along with disease-related variants, Illumina revealed a few more associated with physical traits. They meant little to me. “Your odds of developing male pattern baldness are increased if you are Caucasian,” Illumina told me. You could call me Caucasian, but I have a thick thatch of hair. “Your muscle fibers are built for power,” the website lied.

The whole experience was charming but dull. I certainly didn't want the excitement that comes from discovering you have cherubism. But getting to see my own genome shouldn't have been boring. I was pretty sure that if I could dig deeper—or, rather, if I could enlist the help of some scientists to dig deeper—I'd be able to learn much more about heredity.

After a few weeks of wrangling and paperwork, I managed to get all the raw data from Illumina. It showed up at my door one January afternoon in a white cardboard box. Inside was a shroud of green bubble wrap, inside of which was a kidney-shaped black pouch, inside of which was a slim brushed-metal hard drive. It contained seventy gigabytes of data—the equivalent of more than four hundred high-definition movies.

To make sense of that data, I took my genome on the road. On one trip,
I drove down I-95 to the Yale campus and walked up Science Hill to reach the office of Mark Gerstein. Gerstein's office was heaped with scientific bric-a-brac: Galileo thermometers, Klein-bottle coffee mugs, blinking lights that feed off the electric current in your skin. Gerstein's conversation was packed as well, pinging so quickly between genomes and cloud computing and open-access scientific publishing that I sometimes had to look back at my notebook to remember the question I had just asked him.

The idea of telling me about my own genome intrigued Gerstein to no end. Over the course of his career, he has analyzed thousands of genomes—he helped lead a study called the 1000 Genomes Project, for starters—but he'd almost never looked straight in the face of the person from whom one of those genomes came. As I handed him the drive so that he could copy my genome to his computer, he confessed to a vicarious thrill.

“I'd never have the courage to do this—I'm just too timid,” he said, laughing. “I'm a worrier. Every time there would be a new finding, I'd look in my genome to see if I had it.”

While Gerstein and his team got to work, I went to the New York Genome Center, where a group of scientists were building a genealogical database they called DNA.Land. They created a website where anyone could upload their genetic data for scientific research. In exchange, they would analyze people's DNA and share whatever genealogical clues they could find.

My brother, Ben, had gotten his DNA sequenced by a company called Ancestry.com—not his whole genome, of course, but 682,549 genetic markers. I asked him to upload his file to DNA.Land in order to compare our genes.

Thanks to meiosis, Ben and I are not genetically identical. Our genomes are made up of different selections from our parents' chromosomes. Yet, despite our differences, we still have many long stretches of identical DNA in common.
DNA.Land could confidently recognize 112 identical segments, each one stretching 100 million bases or more. While we are far
from clones, there's no one on Earth who's more genetically similar to me than my brother.

If I were to compare myself to one of my first cousins, I'd find fewer identical segments. We share a pair of grandparents, but they also inherited some DNA from their two other grandparents. The segments we share are also smaller, because there have been two generations between us and our grandparents for meiosis to chop our inherited chromosomes into more pieces.

DNA.Land found 45 other people among its 46,675 volunteers who had enough stretches of identical DNA to suggest they might be my cousins. It was also possible they were not closely related at all, our identical DNA a persistent legacy of ancestors who lived centuries ago. I looked at the names of these possible kin and recognized none of their names. Ben was inspired to do some digging and discovered that one of them—a possible fourth cousin named
Elias Gottesman—had a harrowing story.

As a child, he had been sent to Auschwitz with his family, and there the camp's doctor, Josef Mengele, did experiments on him and his twin brother, Jeno. Mengele was especially taken with twins because he believed he could discover the genetic roots of diseases by examining them—sometimes even dissecting them alive. By the end of the war, Gottesman lost his entire family and even lost his name. Only decades later, as an old man in Israel, did he begin searching for them again. A genetic match to cousins in the United States revealed his birth name; his cousins even sent him a picture of his lost parents.

The DNA Gottesman and I inherited from an ancestor might mean we were close kin. But I didn't contact him, or any of the other possible matches that DNA.Land sent me. A genetic connection did not join our lives together. In fact, if I were to compare my genome to those of my fourth cousins, I'd find that I don't even share any DNA with some of them. That may sound impossible, but only because in modern Western culture we've made the mistake of equating DNA with kinship. That's not actually how heredity works.

The more distantly a cousin is related to you, the more generations back you have to go to find your common ancestors. It also means that over those
generations, the DNA from those ancestors got cut into ever smaller pieces and was mixed with the DNA from ancestors you and your cousin do not share. It's purely a matter of chance which copy of a DNA segment ends up in an egg or a sperm. And so, in time, one ancestor's genes may disappear altogether. In 2014,
Graham Coop, a geneticist at the University of California, Davis, determined that if you brought together 100 pairs of third cousins, one of those pairs would share no identical segments of DNA at all. If you brought together 100 pairs of fourth cousins, 25 would lack this genetic connection.

The same holds true for our ancestors. If I were to compare my genome to those of my grandparents, I'd be able to find large chunks of identical DNA from all four of them, each totaling roughly 25 percent of my genome. In the next generation back, I have eight great-grandparents, contributing more chunks—but smaller ones. With every generation back, my number of ancestors doubles. Roger Goodspeed is among 1,024 ancestors of mine ten generations back. But according to Coop's estimates, I inherited only
628 chunks of DNA from that entire generation. There's only so much room in my genome, and so a lot of their DNA did not finish the journey from my tenth-generation ancestry to me. For any particular ancestor from Roger Goodspeed's generation, there's a 46 percent chance I didn't inherit any of their DNA. I grew up imagining Roger Goodspeed as some kind of American Adam to my family, bestowing Goodspeed genes on all his descendants. But it's pretty much a coin toss whether I have any of his DNA at all. And even if I did, Coop's calculations show I'd be able to trace only about 0.3 percent of my DNA to him.