The Violinist's Thumb: And Other Lost Tales of Love, War, and Genius, as Written by Our Genetic Code

by Sam Kean

As probably its biggest benefit, the placenta allows a mammal mother to carry her living, growing children within her. As a result she can keep her children warm inside the womb and run away from danger with them, advantages that spawning-into-water and squatting-on-nest creatures lack. Live fetuses also have longer to gestate and develop energy-intensive organs like the brain; the placenta’s ability to pump bodily waste away helps the brain develop, too, since fetuses aren’t stewing in toxins. What’s more, because she invests so much energy in her developing fetus—not to mention the literal and intimate connection she feels because of the placenta—a mammal mom feels incentive to nurture and watch over her children, sometimes for years. (Or at least feels the need to nag them for years.) The length of this investment is rare among animals, and mammal children reciprocate by forming unusually strong bonds to their mothers. In one sense, then, the placenta, by enabling all this, made us mammals caring creatures.

  That makes it all the more creepy that the placenta, in all likelihood, evolved from our old friends the retroviruses. But from a biological standpoint, the connection makes sense. Clamping on to cells happens to be a talent of viruses: they fuse their “envelopes” (their outer skin) to a cell before injecting their genetic material into it. When a ball of embryonic cells swims into the uterus and anchors itself there, the embryo also fuses part of itself with the uterine cells, by using special fusion proteins. And the DNA that primates, mice, and other mammals use to make the fusion proteins appears to be plagiarized from genes that retroviruses use to attach and meld their envelopes. What’s more, the uterus of placental mammals draws heavily on other viruslike DNA to do its job, using a special jumping gene called mer20 to flick 1,500 genes on and off in uterine cells. With both organs, it seems we once again borrowed some handy genetic material from a parasite and adapted it to our own ends. As a bonus, the viral genes in the placenta even provide extra immunity, since the presence of retrovirus proteins (either by warning them off, or outcompeting them) discourages other microbes from circling the placenta.

  As another part of its immune function, the placenta filters out any cells that might try to invade the fetus, including cancer cells. Unfortunately, other aspects of the placenta make it downright attractive to cancer. The placenta produces growth hormones to promote the vigorous division of fetal cells, and some cancers thrive on these growth hormones, too. Furthermore, the placenta soaks up enormous amounts of blood and siphons off nutrients for the fetus. That means that blood cancers like leukemia can lurk inside the placenta and flourish. Cancers genetically programmed to metastasize, like the skin cancer melanoma, take to the blood as they slither around inside the body, and they find the placenta quite hospitable as well.

  In fact, melanoma is the most common cancer that mothers and fetuses get simultaneously. The first recorded simultaneous cancer, in 1866, in Germany, involved a roaming melanoma that randomly took root in the mother’s liver and the child’s knee. Both died within nine days. Another horrifying case claimed a twenty-eight-year-old Philadelphia woman, referred to only as “R. McC” by her doctors. It all started when Ms. McC got a brutal sunburn in April 1960. Shortly afterward a half-inch-long mole sprung up between her shoulder blades. It bled whenever she touched it. Doctors removed the mole, and no one thought about it again until May 1963, when she was a few weeks pregnant. During a checkup, doctors noticed a nodule beneath the skin on her stomach. By August the nodule had widened even faster than her belly, and other, painful nodules had sprung up. By January, lesions had spread to her limbs and face, and her doctors opened her up for a cesarean section. The boy inside appeared fine—a full seven pounds, thirteen ounces. But his mother’s abdomen was spotted with dozens of tumors, some of them black. Not surprisingly, the birth finished off what little strength she had. Within an hour, her pulse dropped to thirty-six beats per minute, and though her doctors resuscitated her, she died within weeks.

  And the McC boy? There was hope at first. Despite the widespread cancer, doctors saw no tumors in Ms. McC’s uterus or placenta—her points of contact with her son. And although he was sickly, a careful scan of every crevice and dimple revealed no suspicious-looking moles. But they couldn’t check inside him. Eleven days later tiny, dark blue spots began breaking out on the newborn’s skin. Things deteriorated quickly after that. The tumors expanded and multiplied, and killed him within seven weeks.

  Mayumi had leukemia, not melanoma, but otherwise her family in Chiba reprised the McC drama four decades later. In the hospital, Mayumi’s condition deteriorated day by day, her immune system weakened by three weeks of chemotherapy. She finally contracted a bacterial infection and came down with encephalitis, inflammation of the brain. Her body began to convulse and seize—a result of her brain panicking and misfiring—and her heart and lungs faltered, too. Despite intensive care, she died two days after contracting the infection.

  Even worse, in October 2006, nine months after burying his wife, Hideo had to return to the hospital with Emiko. The once-bouncing girl had fluid in her lungs and, more troublesome, a raw, fever-red mass disfiguring her right cheek and chin. On an MRI, this premature jowl looked enormous—as large as tiny Emiko’s brain. (Try expanding your cheek as far as it will go with air, and its size still wouldn’t be close.) Based on its location within the cheek, the Chiba doctors diagnosed sarcoma, cancer of the connective tissue. But with Mayumi in the back of their minds, they consulted experts in Tokyo and England and decided to screen the tumor’s DNA to see what they could find.

  They found a Philadelphia swap. And not just any Philadelphia swap. Again, this crossover takes place along two tremendously long introns, 68,000 letters long on one chromosome, 200,000 letters long on the other. (This chapter runs about 30,000 letters.) The two arms of the chromosomes could have crossed over at any one of thousands of different points. But the DNA in both Mayumi’s and Emiko’s cancer had crossed over at the same spot, down the same letter. This wasn’t chance. Despite lodging in Emiko’s cheek, the cancer basically was the same.

  But who gave cancer to whom? Scientists had never solved this mystery before; even the McC case was ambiguous, since the fatal tumors appeared only after the pregnancy started. Doctors pulled out the blood-spot card taken from Emiko at birth and determined that the cancer had been present even then. Further genetic testing revealed that Emiko’s normal (nontumor) cells did not show a Philadelphia swap. So Emiko had not inherited any predisposition to this cancer—it had sprung up sometime between conception and the delivery forty weeks later. What’s more, Emiko’s normal cells also showed, as expected, DNA from both her mother and father. But her cheek tumor cells contained no DNA from Hideo; they were pure Mayumi. This proved, indisputably, that Mayumi had given cancer to Emiko, not vice versa.

  Whatever sense of triumph the scientists might have felt, though, was muted. As so often happens in medical research, the most interesting cases spring from the most awful suffering. And in virtually every other historical case where a fetus and mother had cancer simultaneously, both had succumbed to it quickly, normally within a year. Mayumi was already gone, and as the doctors started the eleven-month-old Emiko on chemotherapy, they surely felt these dismal odds weighing on them.

  The geneticists on the case felt something different nagging them. The spread of the cancer here was essentially a transplant of cells from one person to another. If Emiko had gotten an organ from her mother or had tissue grafted onto her cheek, her body would have rejected it as foreign. Yet cancer, of all things, had taken root without triggering the placenta’s alarms or drawing the wrath of Emiko’s immune system. How? Scientists ultimately found the answer in a stretch of DNA far removed from the Philadelphia swap, an area called the MHC.

  Biologists back to Linnaeus’s time have found it a fascinating exercise to list all the traits that make mammals mammals. One place to start—it’s the origin of the term, from the Latin for breast, mamma—is nursing. In addition to providing nutrition, breast milk activates dozens of genes in s
uckling infants, mostly in the intestines, but also possibly in spots like the brain. Not to throw any more panic into expectant mothers, but it seems that artificial formula, however similar in carbs, fats, proteins, vitamins, and for all I know taste, just can’t goose a baby’s DNA the same way.

  Other notable traits of mammals include our hair (even whales and dolphins have a comb-over), our unique inner ear and jaw structure, and our odd habit of chewing food before swallowing (reptiles have no such manners). But on a microscopic level, one place to hunt for the origin of mammals is the MHC, the major histocompatibility complex. Nearly all vertebrates have an MHC, a set of genes that helps the immune system. But the MHC is particularly dear to mammals. It’s among the most gene-rich stretches of DNA we have, over one hundred genes packed into a small area. And similar to our intron/exon editing equipment, we have more sophisticated and more extensive MHCs than other creatures.* Some of those hundred genes have over a thousand different varieties in humans, offering a virtually unlimited number of combinations to inherit. Even close relatives can differ substantially in their MHC, and the differences among random people are a hundred times higher than those along most other stretches of DNA. Scientists sometimes say that humans are over 99 percent genetically identical. Not along their MHCs they aren’t.

  MHC proteins basically do two things. First, some of them grab a random sampling of molecules from inside a cell and put them on display on the cellular surface. This display lets other cells, especially “executioner” immune cells, know what’s going on inside the cell. If the executioner sees the MHC mounting nothing but normal molecules, it ignores the cell. If it sees abnormal material—fragments of bacteria, cancer proteins, other signs of malfeasance—it can attack. The diversity of the MHC helps mammals here because different MHC proteins fasten on to and raise the alarm against different dangers, so the more diversity in the mammalian MHC, the more things a creature can combat. And crucially, unlike with other traits, MHC genes don’t interfere with each other. Mendel identified the first dominant traits, cases where some versions of genes “win out” over others. With the MHC, all the genes work independently, and no one gene masks another. They cooperate; they codominate.

  As for its second, more philosophical function, the MHC allows our bodies to distinguish between self and nonself. While mounting protein fragments, MHC genes cause little beard hairs to sprout on the surface of every cell; and because each creature has a unique combination of MHC genes, this cellular beard hair will have a unique arrangement of colors and curls. Any nonself interlopers in the body (like cells from animals or other people) of course have their own MHC genes sprouting their own unique beards. Our immune system is so precise that it will recognize those beards as different, and—even if those cells betray no signs of diseases or parasites—marshal troops to kill the invaders.

  Destroying invaders is normally good. But one side effect of the MHC’s vigilance is that our bodies reject transplanted organs unless the recipients take drugs to suppress their immune systems. Sometimes even that doesn’t work. Transplanting organs from animals could help alleviate the world’s chronic shortage of organ donors, but animals have such bizarre (to us) MHCs that our bodies reject them instantly. We even destroy tissues and blood vessels around implanted animal organs, like retreating soldiers burning crops so that the enemy can’t use them for nourishment either. By absolutely paralyzing the immune system, doctors have kept people alive on baboon hearts and livers for a few weeks, but so far the MHC always wins out.

  For similar reasons, the MHC made things difficult for mammal evolution. By all rights, a mammal mother should attack the fetus inside her as a foreign growth, since half its DNA, MHC and otherwise, isn’t hers. Thankfully, the placenta mediates this conflict by restricting access to the fetus. Blood pools in the placenta, but no blood actually crosses through to the fetus, just nutrients. As a result, a baby like Emiko should remain perfectly, parasitically invisible to Mayumi’s immune cells, and Mayumi’s cells should never cross over into Emiko. Even if a few do slip through the placental gate, Emiko’s own immune system should recognize the foreign MHC and destroy them.

  But when scientists scrutinized the MHC of Mayumi’s cancerous blood cells, they discovered something that would be almost admirable in its cleverness, if it weren’t so sinister. In humans, the MHC is located on the shorter arm of chromosome six. The scientists noticed that this short arm in Mayumi’s cancer cells was even shorter than it should be—because the cells had deleted their MHC. Some unknown mutation had simply wiped it from their genes. This left them functionally invisible on the outside, so neither the placenta nor Emiko’s immune cells could classify or recognize them. She had no way to scrutinize them for evidence that they were foreign, much less that they harbored cancer.

  Overall, then, scientists could trace the invasion of Mayumi’s cancer to two causes: the Philadelphia swap that made them malignant, and the MHC mutation that made them invisible and allowed them to trespass and burrow into Emiko’s cheek. The odds of either thing happening were low; the odds of them happening in the same cells, at the same time, in a woman who happened to be pregnant, were astronomically low. But not zero. In fact, scientists now suspect that in most historical cases where mothers gave cancer to their fetuses, something similar disabled or compromised the MHC.

  If we follow the thread far enough, the MHC can help illuminate one more aspect of Hideo and Mayumi and Emiko’s story, a thread that runs back to our earliest days as mammals. A developing fetus has to conduct a whole orchestra of genes inside every cell, encouraging some DNA to play louder and hushing other sections up. Early on in the pregnancy, the most active genes are the ones that mammals inherited from our egg-laying, lizardlike ancestors. It’s a humbling experience to flip through a biology textbook and see how uncannily similar bird, lizard, fish, human, and other embryos appear during their early lives. We humans even have rudimentary gill slits and tails—honest-to-god atavisms from our animal past.

  After a few weeks, the fetus mutes the reptilian genes and turns on a coterie of genes unique to mammals, and pretty soon the fetus starts to resemble something you could imagine naming after your grandmother. Even at this stage, though, if the right genes are silenced or tweaked, atavisms (i.e., genetic throwbacks) can appear. Some people are born with the same extra nipples that barnyard sows have.* Most of these extra nipples poke through the “milk line” running vertically down the torso, but they can appear as far away as the sole of the foot. Other atavistic genes leave people with coats of hair sprouting all over their bodies, including their cheeks and foreheads. Scientists can even distinguish (if you’ll forgive the pejoratives) between “dog-faced” and “monkey-faced” coats, depending on the coarseness, color, and other qualities of the hair. Infants missing a snippet at the end of chromosome five develop cri-du-chat, or “cry of the cat” syndrome, so named for their caterwauling chirps and howls. Some children are also born with tails. These tails—usually centered above their buttocks—contain muscles and nerves and run to five inches long and an inch thick. Sometimes tails appear as side effects of recessive genetic disorders that cause widespread anatomical problems, but tails can appear idiosyncratically as well, in otherwise normal children. Pediatricians have reported that these boys and girls can curl their tails up like an elephant’s trunk, and that the tails contract involuntarily when children cough or sneeze.* Again, all fetuses have tails at six weeks old, but they usually retract after eight weeks as tail cells die and the body absorbs the excess tissue. Tails that persist probably arise from spontaneous mutations, but some children with tails do have betailed relatives. Most get the harmless appendage removed just after birth, but some don’t bother until adulthood.

  A hale and healthy baby boy born with a tail—a genetic throwback from our primate past. (Jan Bondeson, A Cabinet of Medical Curiosities, reproduced by permission)

  All of us have other atavisms dormant within us as well, just waiting for the right genet
ic signals to awaken them. In fact, there’s one genetic atavism that none of us escapes. About forty days after conception, inside the nasal cavity, humans develop a tube about 0.01 inches long, with a slit on either side. This incipient structure, the vomeronasal organ, is common among mammals, who use it to help map the world around them. It acts like an auxiliary nose, except that instead of smelling things that any sentient creature can sniff out (smoke, rotten food), the vomeronasal organ detects pheromones. Pheromones are veiled scents vaguely similar to hormones; but whereas hormones give our bodies internal instructions, pheromones give instructions (or at least winks and significant glances) to other members of our species.

  Because pheromones help guide social interactions, especially intimate encounters, shutting off the VNO in certain mammals can have awkward consequences. In 2007, scientists at Harvard University genetically rewired some female mice to disable their VNOs. When the mice were by themselves, not much changed—they acted normally. But when let loose on regular females, the altered mice treated them like the Romans did the Sabine women. They stormed and mounted the maidens, and despite lacking the right equipment, they began thrusting their hips back and forth. The bizarro females even groaned like men, emitting ultrasonic squeals that, until then, were heard only from male mice at climax.

  Humans rely less on scent than other mammals do; throughout our evolution we’ve lost or turned off six hundred common mammalian genes for smell. But that makes it all the more striking that our genes still build a VNO. Scientists have even detected nerves running from the fetal VNO to the brain, and have seen these nerves send signals back and forth. Yet for unknown reasons, despite going through the trouble of creating the organ and wiring it up, our bodies neglect this sixth sense, and after sixteen weeks it starts shriveling. By adulthood, it has retracted to the point that most scientists dispute whether humans even have a VNO, much less a functional one.