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

by Sam Kean

  All the while, beneath his incinerated skin, Yamaguchi’s DNA was nursing even graver injuries. The nuclear bomb at Hiroshima released (among other radioactivity) loads of supercharged x-rays called gamma rays. Like most radioactivity, these rays single out and selectively damage DNA, punching DNA and nearby water molecules and making electrons fly out like uppercut teeth. The sudden loss of electrons forms free radicals, highly reactive atoms that chew on chemical bonds. A chain reaction begins that cleaves DNA and sometimes snaps chromosomes into pieces.

  By the mid-1940s, scientists were starting to grasp why the shattering or disruption of DNA could wreak such ruin inside cells. First, scientists based in New York produced strong evidence that genes were made of DNA. This upended the persistent belief in protein inheritance. But as a second study revealed, DNA and proteins still shared a special relationship: DNA made proteins, with each DNA gene storing the recipe for one protein. Making proteins, in other words, was what genes did—that’s how genes created traits in the body.

  In conjunction, these two ideas explained the harm of radioactivity. Fracturing DNA disrupts genes; disrupting genes halts protein production; halting protein production kills cells. Scientists didn’t work this out instantly—the crucial “one gene/one protein” paper appeared just days before Hiroshima—but they knew enough to cringe at the thought of nuclear weapons. When Hermann Muller won his Nobel Prize in 1946, he prophesied to the New York Times that if atomic bomb survivors “could foresee the results 1,000 years from now… they might consider themselves more fortunate if the bomb had killed them.”

  Despite Muller’s pessimism, Yamaguchi did want to survive, badly, for his family. He’d had complicated feelings about the war—opposing it at first, supporting it once under way, then shading back toward opposition when Japan began to stumble, because he feared the island being overrun by enemies who might harm his wife and son. (If so, he’d contemplated giving them an overdose of sleeping pills to spare them.) In the hours after Hiroshima, he yearned to get back to them, so when he heard rumors about trains leaving the city, he sucked up his strength and resolved to find one.

  Hiroshima is a collection of islands, and Yamaguchi had to cross a river to reach the train station. All the bridges had collapsed or burned, so he steeled himself and began crossing an apocalyptic “bridge of corpses” clogging the river, crawling across melted legs and faces. But an uncrossable gap in the bridge forced him to turn back. Farther upstream, he found a railroad trestle with one steel beam intact, spanning fifty yards. He clambered up, crossed the iron tightrope, and descended. He pushed through the mob at the station and slumped into a train seat. Miraculously the train pulled out soon afterward—he was saved. The train would run all night, but he was finally headed home, to Nagasaki.

  A physicist stationed in Hiroshima might have pointed out that the gamma rays finished working over Yamaguchi’s DNA in a millionth of a billionth of a second. To a chemist, the most interesting part—how the free radicals gnawed through DNA—would have ceased after a millisecond. A cell biologist would have needed to wait maybe a few hours to study how cells patch up torn DNA. A doctor could have diagnosed radiation sickness—headaches, vomiting, internal bleeding, peeling skin, anemic blood—within a week. Geneticists needed the most patience. The genetic damage to the survivors didn’t surface for years, even decades. And in an eerie coincidence, scientists began to piece together how exactly genes function, and fail, during those very decades—as if providing a protracted running commentary on DNA devastation.

  However definitive in retrospect, experiments on DNA and proteins in the 1940s convinced only some scientists that DNA was the genetic medium. Better proof came in 1952, from virologists Alfred Hershey and Martha Chase. Viruses, they knew, hijacked cells by injecting genetic material. And because the viruses they studied consisted of only DNA and proteins, genes had to be one or the other. The duo determined which by tagging viruses with both radioactive sulfur and radioactive phosphorus, then turning them loose on cells. Proteins contain sulfur but no phosphorus, so if genes were proteins, radioactive sulfur should be present in cells postinfection. But when Hershey and Chase filtered out infected cells, only radioactive phosphorus remained: only DNA had been injected.

  Hershey and Chase published these results in April 1952, and they ended their paper by urging caution: “Further chemical inferences should not be drawn from the experiments presented.” Yeah, right. Every scientist in the world still working on protein heredity dumped his research down the sink and took up DNA. A furious race began to understand the structure of DNA, and just one year later, in April 1953, two gawky scientists at Cambridge University in England, Francis Crick and James Watson (a former student of Hermann Muller), made the term “double helix” legendary.

  Watson and Crick’s double helix was two loooooooong DNA strands wrapped around each other in a right-handed spiral. (Point your right thumb toward the ceiling; DNA twists upward along the counterclockwise curl of your fingers.) Each strand consisted of two backbones, and the backbones were held together by paired bases that fit together like puzzle pieces—angular A with T, curvaceous C with G. Watson and Crick’s big insight was that because of this complementary A-T and C-G base pairing, one strand of DNA can serve as a template for copying the other. So if one side reads CCGAGT, the other side must read GGCTCA. It’s such an easy system that cells can copy hundreds of DNA bases per second.

  However well hyped, though, the double helix revealed zero about how DNA genes actually made proteins—which is, after all, the important part. To understand this process, scientists had to scrutinize DNA’s chemical cousin, RNA. Though similar to DNA, RNA is single-stranded, and it substitutes the letter U (uracil) for T in its strands. Biochemists focused on RNA because its concentration would spike tantalizingly whenever cells started making proteins. But when they chased the RNA around the cell, it proved as elusive as an endangered bird; they caught only glimpses before it vanished. It took years of patient experiments to determine exactly what was going on here—exactly how cells transform strings of DNA letters into RNA instructions and RNA instructions into proteins.

  Cells first “transcribe” DNA into RNA. This process resembles the copying of DNA, in that one strand of DNA serves as a template. So the DNA string CCGAGT would become the RNA string GGCUCA (with U replacing T). Once constructed, this RNA string leaves the confines of the nucleus and chugs out to special protein-building apparatuses called ribosomes. Because it carries the message from one site to another, it’s called messenger RNA.

  The protein building, or translation, begins at the ribosomes. Once the messenger RNA arrives, the ribosome grabs it near the end and exposes just three letters of the string, a triplet. In our example, GGC would be exposed. At this point a second type of RNA, called transfer RNA, approaches. Each transfer RNA has two key parts: an amino acid trailing behind it (its cargo to transfer), and an RNA triplet sticking off its prow like a masthead. Various transfer RNAs might try to dock with the messenger RNA’s exposed triplet, but only one with complementary bases will stick. So with the triplet GGC, only a transfer RNA with CCG will stick. And when it does stick, the ribosome unloads its amino acid cargo.

  At this point the transfer RNA leaves, the messenger RNA shifts down three spots, and the process repeats. A different triplet is exposed, and a different transfer RNA with a different amino acid docks. This puts amino acid number two in place. Eventually, after many iterations, this process creates a string of amino acids—a protein. And because each RNA triplet leads to one and only one amino acid being added, information should (should) get translated perfectly from DNA to RNA to protein. This same process runs every living thing on earth. Inject the same DNA into guinea pigs, frogs, tulips, slime molds, yeast, U.S. congressmen, whatever, and you get identical amino acid chains. No wonder that in 1958 Francis Crick elevated the DNA → RNA → protein process into the “Central Dogma” of molecular biology.*

  Still, Crick’s dogma doesn
’t explain everything about protein construction. For one thing, notice that, with four DNA letters, sixty-four different triplets are possible (4 × 4 × 4 = 64). Yet all those triplets code for just twenty amino acids in our bodies. Why?

  A physicist named George Gamow founded the RNA Tie Club in 1954 in part to figure out this question. A physicist moonlighting in biology might sound odd—Gamow studied radioactivity and Big Bang theory by day—but other carpetbagging physicists like Richard Feynman joined the club as well. Not only did RNA offer an intellectual challenge, but many physicists felt appalled by their role in creating nuclear bombs. Physics seemed life destroying, biology life restoring. Overall, twenty-four physicists and biologists joined the Tie Club’s roster—one for each amino acid, plus four honorary inductees, for each DNA base. Watson and Crick joined (Watson as official club “Optimist,” Crick as “Pessimist”), and each member sported a four-dollar bespoke green wool tie with an RNA strand embroidered in gold silk, made by a haberdasher in Los Angeles. Club stationery read, “Do or die, or don’t try.”

  RNA Tie Club members sporting green wool ties with gold silk RNA embroidery. From left, Francis Crick, Alexander Rich, Leslie E. Orgel, James Watson. (Courtesy of Alexander Rich)

  Despite its collective intellectual horsepower, in one way the club ended up looking a little silly historically. Problems of perverse complexity often attract physicists, and certain physics-happy club members (including Crick, a physics Ph.D.) threw themselves into work on DNA and RNA before anyone realized how simple the DNA → RNA → proteins process was. They concentrated especially on how DNA stores its instructions, and for whatever reason they decided early on that DNA must conceal its instructions in an intricate code—a biological cryptogram. Nothing excites a boys’ club as much as coded messages, and like ten-year-olds with Cracker Jack decoder rings, Gamow, Crick, and others set out to break this cipher. They were soon scribbling away with pencil and paper at their desks, page after page piling up, their imaginations happily unfettered by doing experiments. They devised solutions clever enough to make Will Shortz smile—“diamond codes” and “triangle codes” and “comma codes” and many forgotten others. These were NSA-ready codes, codes with reversible messages, codes with error-correction mechanisms built in, codes that maximized storage density by using overlapping triplets. The RNA boys especially loved codes that used equivalent anagrams (so CAG = ACG = GCA, etc.). The approach was popular because when they eliminated all the combinatorial redundancies, the number of unique triplets was exactly twenty. In other words, they’d seemingly found a link between twenty and sixty-four—a reason nature just had to use twenty amino acids.

  In truth, this was so much numerology. Hard biochemical facts soon deflated the code breakers and proved there’s no profound reason DNA codes for twenty amino acids and not nineteen or twenty-one. Nor was there any profound reason (as some hoped) that a given triplet called for a given amino acid. The entire system was accidental, something frozen into cells billions of years ago and now too ingrained to replace—the QWERTY keyboard of biology. Moreover, RNA employs no fancy anagrams or error-correcting algorithms, and it doesn’t strive to maximize storage space, either. Our code is actually choking on wasteful redundancy: two, four, even six RNA triplets can represent the same amino acid.* A few biocryptographers later admitted feeling annoyed when they compared nature’s code to the best of the Tie Club’s codes. Evolution didn’t seem nearly as clever.

  Any disappointment soon faded, however. Solving the DNA/RNA code finally allowed scientists to integrate two separate realms of genetics, gene-as-information and gene-as-chemical, marrying Miescher with Mendel for the first time. And it actually turned out better in some ways that our DNA code is kludgy. Fancy codes have nice features, but the fancier a code gets, the more likely it will break down or sputter. And however crude, our code does one thing well: it keeps life going by minimizing the damage of mutations. It’s exactly that talent that Tsutoma Yamaguchi and so many others had to count on in August 1945.

  Ill and swooning, Yamaguchi arrived in Nagasaki early on August 8 and staggered home. (His family had assumed him lost; he convinced his wife he wasn’t a ghost by showing her his feet, since Japanese ghosts traditionally have none.) Yamaguchi rested that day, swimming in and out of consciousness, but obeyed an order the next day to report to Mitsubishi headquarters in Nagasaki.

  He arrived shortly before 11 a.m. Arms and face bandaged, he struggled to relate the magnitude of atomic warfare to his coworkers. But his boss, skeptical, interrupted to browbeat him, dismissing his story as a fable. “You’re an engineer,” he barked. “Calculate it. How could one bomb… destroy a whole city?” Famous last words. Just as this Nostradamus wrapped up, a white light swelled inside the room. Heat prickled Yamaguchi’s skin, and he hit the deck of the ship-engineering office.

  “I thought,” he later recalled, “the mushroom cloud followed me from Hiroshima.”

  Eighty thousand people died in Hiroshima, seventy thousand more in Nagasaki. Of the hundreds of thousands of surviving victims, evidence suggests that roughly 150 got caught near both cities on both days, and that a handful got caught within both blast zones, a circle of intense radiation around 1.5 miles wide. Some of these nijyuu hibakusha, double-exposure survivors, had stories to make stones weep. (One had burrowed into his wrecked home in Hiroshima, clawed out his wife’s blackened bones, and stacked them in a washbasin to return them to her parents in Nagasaki. He was trudging up the street to the parents’ house, washbasin under his arm, when the morning air again fell quiet and the sky was once again bleached white.) But of all the reported double victims, the Japanese government has recognized only one official nijyuu hibakusha, Tsutomu Yamaguchi.

  Shortly after the Nagasaki explosion, Yamaguchi left his shaken boss and office mates and climbed a watchtower on a nearby hill. Beneath another pall of dirty clouds, he watched his cratered-out hometown smolder, including his own house. A tarry radioactive rain began falling, and he struggled down the hill, fearing the worst. But he found his wife, Hisako, and young son, Katsutoshi, safe in an air-raid shelter.

  After the exhilaration of seeing them wore off, Yamaguchi felt even more ill than before. In fact, over the next week he did little but lie in the shelter and suffer like Job. His hair fell out. Boils erupted. He vomited incessantly. His face swelled. He lost hearing in one ear. His reburned skin flaked off, and beneath it his flesh glowed raw red “like whale meat” and pierced him with pain. And as badly as Yamaguchi and others suffered during those months, geneticists feared the long-term agony would be equally bad, as mutations slowly began surfacing.

  Scientists had known about mutations for a half century by then, but only work on the DNA → RNA → protein process by the Tie Club and others revealed exactly what these mutations consisted of. Most mutations involve typos, the random substitution of a wrong letter during DNA replication: CAG might become CCG, for instance. In “silent” mutations, no harm is done because of the DNA code’s redundancy: the before and after triplets call for the same amino acid, so the net effect is like mistyping grey for gray. But if CAG and CCG lead to different amino acids—a “missense” mutation—the mistake can change a protein’s shape and disable it.

  Even worse are “nonsense” mutations. When making proteins, cells will continue to translate RNA into amino acids until they encounter one of three “stop” triplets (e.g., UGA), which terminate the process. A nonsense mutation accidentally turns a normal triplet into one of these stop signs, which truncates the protein early and usually disfigures it. (Mutations can also undo stop signs, and the protein runs on and on.) The black mamba of mutations, the “frameshift” mutation, doesn’t involve typos. Instead a base disappears, or an extra base squeezes in. Because cells read RNA in consecutive groups of three, an insertion or deletion screws up not only that triplet but every triplet down the line, a cascading catastrophe.

  Cells usually correct simple typos right away, but if something goes wrong (a
nd it will), the flub can become permanently fixed in DNA. Every human being alive today was in fact born with dozens of mutations his parents lacked, and a few of those mutations would likely be lethal if we didn’t have two copies of every gene, one from each parent, so one can pick up the slack if the other malfunctions. Nevertheless all living organisms continue to accumulate mutations as they age. Smaller creatures that live at high temperatures are especially hard hit: heat on a molecular level is vigorous motion, and the more molecular motion, the more likely something will bump DNA’s elbow as it’s copying. Mammals are relatively hefty and maintain a constant body temperature, thankfully, but we do fall victim to other mutations. Wherever two T’s appear in a row in DNA, ultraviolet sunlight can fuse them together at an odd angle, which kinks DNA. These accidents can kill cells outright or simply irritate them. Virtually all animals (and plants) have special handyman enzymes to fix T-T kinks, but mammals lost them during evolution—which is why mammals sunburn.

  Besides spontaneous mutations, outside agents called mutagens can also injure DNA, and few mutagens do more damage than radioactivity. Again, radioactive gamma rays cause free radicals to form, which cleave the phosphate-sugar backbone of DNA. Scientists now know that if just one strand of the double helix snaps, cells can repair the damage easily, often within an hour. Cells have molecular scissors to snip out mangled DNA, and can run enzymes down the track of the undamaged strand and add the complementary A, C, G, or T at each point. The repair process is quick, simple, and accurate.

  Double-strand breaks, though rarer, cause direr problems. Many double breaks resemble hastily amputated limbs, with tattered flaps of single-stranded DNA hanging off both ends. Cells do have two near-twin copies of every chromosome, and if one has a double-strand break, cells can compare its ragged strands to the (hopefully undamaged) other chromosome and perform repairs. But this process is laborious, and if cells sense widespread damage that needs quick repairs, they’ll often just slap two hanging flaps together wherever a few bases line up (even if the rest don’t), and hastily fill in the missing letters. Wrong guesses here can introduce a dreaded frameshift mutation—and there are plenty of wrong guesses. Cells repairing double-strand breaks get things wrong roughly three thousand times more often than cells simply copying DNA.