The Gene

by Siddhartha Mukherjee

  But what actual genes were involved? Since the late 1990s, a host of novel DNA-sequencing methods—called massively parallel DNA sequencing or next-generation sequencing—have allowed geneticists to sequence hundred of millions of base pairs from any human genome. Massively parallel sequencing is an enormous scale-up of the standard sequencing method: the human genome is fragmented into tens of thousands of shards, these DNA fragments are sequenced at the same time—i.e., in parallel—and the genome is “reassembled” using computers to find the overlaps between the sequences. The method can be applied to sequence the entire genome (termed whole genome sequencing) or to chosen parts of the genome, such as the protein-encoding exons (termed exome sequencing).

  Massively parallel sequencing is especially effective for gene hunting when one closely related genome can be compared to another. If one member of a family has a disease, and all other members do not, then finding the gene becomes immeasurably simplified. Gene hunting devolves into a spot-the-odd-man-out game on a gigantic scale: by comparing the genetic sequences of all the closely related family members, a mutation that appears in the affected individual but not in the unaffected relatives can be found.

  The sporadic variant of schizophrenia posed a perfect test case for the power of this approach. In 2013, an enormous study identified 623 young men and women with schizophrenia whose parents and siblings were unaffected. Gene sequencing was performed on these families. Since most parts of the genome are shared in any given family, only the culprit genes fell out as different.I

  In 617 such cases, a culprit mutation was found in the child that was not present in either parent. On average, each child had only one mutation, although an occasional child had more. Nearly 80 percent of the mutations occurred in the chromosome derived from the father, and the father’s age was a prominent risk factor, suggesting that the mutations may occur during spermiogenesis, particularly in older males. Many of these mutations, predictably, involved genes that affect the synapses between nerves, or the development of the nervous system. Although hundreds of mutations occurred in hundreds of genes across the 617 cases, occasionally the same mutant gene was found in several independent families, thereby vastly strengthening the likelihood of its links to the disorder.II By definition, these mutations are sporadic or de novo—i.e., they occurred during the conception of the child. Sporadic schizophrenia is the consequence of alterations in neural development caused by the alterations of genes that specify the development of the nervous system. Strikingly, many of the genes found in this study have also been implicated in sporadic autism and bipolar disease.III

  But what about genes for familial schizophrenia? At first, you might imagine that finding genes for the familial variant would be easier. Schizophrenia that runs through families, like a saw blade cutting through generations, is more common to start with, and patients are easier to find and track. But counterintuitively, perhaps, identifying genes in complex familial diseases turns out to be much more difficult. Finding a gene that causes the sporadic or spontaneous variant of a disorder is like searching for a needle in haystack. You compare two genomes, trying to find small differences, and with enough data and computational power, such differences can generally be identified. But searching for multiple gene variants that cause a familial disease is like looking for a haystack in a haystack. Which parts of the “haystack”—i.e., which combinations of gene variants—increase the risk, and what parts are innocent bystanders? Parents and children naturally share parts of their genome, but which of those shared parts are relevant to the inherited disease? The first problem—“spot the outlier”—requires computational power. The second—“deconvolute the similarity”—demands conceptual subtlety.

  Despite these hurdles, geneticists have begun systematic hunts for such genes, using combinations of genetic techniques, including linkage analysis to map the culprit genes to their physical locations on chromosomes, large association studies to identify genes that correlate with the disease, and next-generation sequencing to identify the genes and mutations. Based on the analysis of genomes, we know that there are at least 108 genes (or rather genetic regions) associated with schizophrenia—although we know the identity of only a handful of these culprits.IV Notably, no single gene stands out as the sole driver of the risk. The contrast with breast cancer is revealing. There are certainly multiple genes implicated in hereditary breast cancer, but single genes, such as BRCA1, are powerful enough to drive the risk (even if we cannot predict when a woman with BRCA1 will get breast cancer, she has a 70–80 percent lifetime risk of developing breast cancer). Schizophrenia generally does not seem to have such strong single drivers or predictors of disease. “There are lots of small, common genetic effects, scattered across the genome . . . ,” one researcher said. “There are many different biological processes involved.”

  Familial schizophrenia (like normal human features such as intelligence and temperament) is thus highly heritable but only moderately inheritable. In other words, genes—hereditary determinants—are crucially important to the future development of the disorder. If you possess a particular combination of genes, the chance of developing the illness is extremely high: hence the striking concordance among identical twins. On the other hand, the inheritance of the disorder across generations is complex. Since genes are mixed and matched in every generation, the chance that you will inherit that exact permutation of variants from your father or mother is dramatically lower. In some families, perhaps, there are fewer gene variants, but with more potent effects—thereby explaining the recurrence of the disorder across generations. In other families, the genes may have weaker effects and require deeper modifiers and triggers—thereby explaining the infrequent inheritance. In yet other families, a single, highly penetrant gene is accidentally mutated in sperm or egg cells before conception, leading to the observed cases of sporadic schizophrenia.V

  Can we imagine a genetic test for schizophrenia? The first step would involve creating a compendium of all the genes involved—a gargantuan project for human genomics. But even such a compendium would be insufficient. Genetic studies clearly indicate that some mutations only act in concert with other mutations to cause the disease. We need to identify the combinations of genes that predict the actual risk.

  The next step is to contend with incomplete penetrance and variable expressivity. It is important to understand what “penetrance” and “expressivity” mean in these gene-sequencing studies. When you sequence the genome of a child with schizophrenia (or any genetic disease) and compare it to the genome of a normal sibling or parent, you are asking, “How are children diagnosed with schizophrenia genetically different from ‘normal’ children?” The question that you are not asking is the following: “If the mutated gene is present in a child, what are the chances that he or she will develop schizophrenia or bipolar disease?”

  The difference between the two questions is critical. Human genetics has become progressively adept at creating what one might describe as a “backward catalog”—a rearview mirror—of a genetic disorder: Knowing that a child has a syndrome, what are the genes that are mutated? But to estimate penetrance and expressivity, we also need to create a “forward catalog”: If a child has a mutant gene, what are the chances that he or she will develop the syndrome? Is every gene fully predictive of risk? Does the same gene variant or gene combination produce highly variable phenotypes in individuals—schizophrenia in one, bipolar disease in another, and a relatively mild variant of hypomania in a third? Do some combinations of variants require other mutations, or triggers, to push that risk over an edge?

  There’s a final twist to this puzzle of diagnosis—and to illustrate it, let me turn to a story. One night in 1946, a few months before his death, Rajesh came home from college with a riddle, a mathematical puzzle. The three younger brothers flung themselves at it, passing it back and forth like an arithmetic soccer ball. They were driven by the rivalry of siblings; the fragile pride of adolescence; the resilience of refugees;
the terror of failure in an unforgiving city. I imagine the three of them—twenty-one, sixteen, thirteen—splayed on three corners of the pinched room, each spinning fantastical solutions, each attacking the problem with his distinctive strategy. My father: grim, purposeful, bullheaded, methodical, but lacking inspiration. Jagu: unconventional, oblique, out-of-the-box, but with no discipline to guide him. Rajesh: thorough, inspired, disciplined, often arrogant.

  Night fell and the puzzle was still not solved. At about eleven at night, the brothers drifted off to sleep one by one. But Rajesh stayed up all night. He paced the room, scribbling solutions and alternatives. By dawn, he had finally cracked it. The next morning, he wrote the solution on four sheets of paper and left it by the feet of one of his brothers.

  This much of the story is imprinted in the myth and lore of my family. What happened next is not well-known. Years later, my father told me of the week of terror that followed that episode. Rajesh’s first sleepless night turned into a second sleepless night, then a third. The all-nighter had tipped him into a burst of fulminant mania. Or perhaps it was the mania that came first and spurred the all-night marathon of problem solving and the solution. In either case, he disappeared for the next few days and could not be found. His brother Ratan was dispatched to find him, and Rajesh had to be forced back home. My grandmother, hoping to nip future breakdowns in the bud, banned puzzles and games from the house (she would remain permanently suspicious of games all her life. As children, we lived with a stifling moratorium on games at home). For Rajesh, this was a portent of the future—the first of many such breakdowns to come.

  Abhed, my father had called heredity—“indivisible.” There is an old trope in popular culture of the “crazy genius,” a mind split between madness and brilliance, oscillating between the two states at the throw of a single switch. But Rajesh had no switch. There was no split or oscillation, no pendulum. The magic and the mania were perfectly contiguous—bordering kingdoms with no passports. They were part of the same whole, indivisible.

  “We of the craft are all crazy,” Lord Byron, the high priest of crazies, wrote. “Some are affected by gaiety, others by melancholy, but all are more or less touched.” Versions of this story have been told, over and over, with bipolar disease, with some variants of schizophrenia, and with rare cases of autism; all are “more or less touched.” It is tempting to romanticize psychotic illness, so let me emphasize that the men and women with these mental disorders experience paralyzing cognitive, social, and psychological disturbances that send gashes of devastation through their lives. But also indubitably, some patients with these syndromes possess exceptional and unusual abilities. The effervescence of bipolar disease has long been linked to extraordinary creativity; at times, the heightened creative impulse is manifest during the throes of mania.

  In Touched with Fire, an authoritative study of the link between madness and creativity, the psychologist-writer Kay Redfield Jamison compiled a list of those “more or less touched” that reads like the Who’s Who of cultural and artistic achievers: Byron (of course), van Gogh, Virginia Woolf, Sylvia Plath, Anne Sexton, Robert Lowell, Jack Kerouac—and on and on. That list can be extended to include scientists (Isaac Newton, John Nash), musicians (Mozart, Beethoven), and an entertainer who built an entire genre out of mania before succumbing to depression and suicide (Robin Williams). Hans Asperger, the psychologist who first described children with autism, called them “little professors” for good reason. Withdrawn, socially awkward, or even language-impaired children, barely functional in one “normal” world, might produce the most ethereal version of Satie’s Gymnopédies on the piano or calculate the factorial of eighteen in seven seconds.

  The point is this: if you cannot separate the phenotype of mental illness from creative impulses, then you cannot separate the genotype of mental illness and creative impulse. The genes that “cause” one (bipolar disease) will “cause” another (creative effervescence). This conundrum brings us to Victor McKusick’s understanding of illness—not as absolute disability but as a relative incongruence between a genotype and an environment. A child with a high-functioning form of autism may be impaired in this world, but might be hyperfunctional in another—one in which, say, the performance of complex arithmetic calculations, or the sorting of objects by the subtlest gradations of color, is a requirement for survival or success.

  What about that elusive genetic diagnosis for schizophrenia, then? Can we imagine a future in which we might eliminate schizophrenia from the human gene pool—by diagnosing fetuses using genetic tests and terminating such pregnancies? Not without acknowledging the aching uncertainties that remain unsolved. First, even though many variants of schizophrenia have been linked to mutations in single genes, hundreds of genes are involved—some known and some yet unknown. We do not know whether some combinations of genes are more pathogenic than others.

  Second, even if we could create a comprehensive catalog of all genes involved, the vast universe of unknown factors might still alter the precise nature of the risk. We do not know what the penetrance of any individual gene is, or what modifies the risk in a particular genotype.

  Finally, some of the genes identified in certain variants of schizophrenia or bipolar disease actually augment certain abilities. If the most pathological variants of a mental illness can be sifted out or discriminated from the high-functioning variants by genes or gene combinations alone, then we can hope for such a test. But it is much more likely that such a test will have inherent limits: most of the genes that cause disease in one circumstance might be the very genes that cause hyperfunctional creativity in another. As Edvard Munch put it, “[My troubles] are part of me and my art. They are indistinguishable from me, and [treatment] would destroy my art. I want to keep those sufferings.” These very “sufferings,” we might remind ourselves, were responsible for one of the most iconic images of the twentieth century—of a man so immersed in a psychotic era that he could only scream a psychotic response to it.

  The prospect of a genetic diagnosis for schizophrenia and bipolar disorder thus involves confronting fundamental questions about the nature of uncertainty, risk, and choice. We want to eliminate suffering, but we also want to “keep those sufferings.” It is easy to understand Susan Sontag’s formulation of illness as the “night-side of life.” That conception works for many forms of illness—but not all. The difficulty lies in defining where twilight ends or where daybreak begins. It does not help that the very definition of illness in one circumstance becomes the definition of exceptional ability in another. Night on one side of the globe is often day, resplendent and glorious, on a different continent.

  In the spring of 2013, I flew to San Diego to one of the most provocative meetings that I have ever attended. Entitled “The Future of Genomic Medicine,” the meeting was at the Scripps Institute in La Jolla, at a conference center overlooking the ocean. The site was a monument to modernism—blond wood, angular concrete, mullions of steel. The light on the water was blindingly glorious. Joggers with post-human bodies ran lankily across the boardwalk. The population geneticist David Goldstein spoke about “Sequencing Undiagnosed Conditions of Childhood,” an effort to extend massively parallel gene sequencing to undiagnosed childhood diseases. The physicist-turned-biologist Stephen Quake discussed the “Genomics of the Unborn,” the prospect of diagnosing every mutation in a growing fetus by sampling the scraps of fetal DNA that spill naturally into maternal blood.

  On the second morning of the conference, a fifteen-year-old girl—I’ll call her Erika—was wheeled onstage by her mother. Erika wore a lacy, white dress and had a scarf slung across her shoulders. She had a story to tell—of genes, identity, fate, choices, and diagnosis. Erika has a genetic condition that has caused a severe, progressive degenerative disease. The symptoms began when she was one and half years old—small twitches in her muscles. By four, the tremors had progressed furiously; she could hardly keep her muscles still. She would wake up twenty or thirty times every night, dre
nched in sweat and racked by unstoppable tremors. Sleep seemed to worsen the symptoms, so her parents took shifts to stay awake with her, trying to console her into a few minutes of rest every night.

  Clinicians suspected an unusual genetic syndrome, but all known genetic tests failed to diagnose the illness. Then in June 2011, Erika’s father was listening to NPR when he heard about a pair of twins in California, Alexis and Noah Beery, who also had a long history of muscle problems. The twins had undergone gene sequencing and ultimately been diagnosed with a rare new syndrome. Based on that genetic diagnosis, the supplementation of a chemical, 5-hydroxytryptamine, or 5-HT, had dramatically reduced the twins’ motor symptoms.

  Erika hoped for a similar outcome. In 2012, she was the first patient to join a clinical trial that would attempt to diagnose her illness by sequencing her genome. By the summer of 2012, the sequence was back: Erika had not one but two mutations in her genome. One, in a gene called ADCY5, altered nerve cells’ capacity to send signals to each other. The other was in a gene, DOCK3, that controls nerve signals that enable the coordinated movement of muscles. The combination of the two had precipitated the muscle-wasting and tremor-inducing syndrome. It was a genetic lunar eclipse—two rare syndromes superposed on each other, causing the rarest of rare illnesses.

  After Erika’s talk, as the audience spilled into the lobby outside the auditorium, I ran into Erika and her mother. Erika was utterly charming—modest, thoughtful, sober, mordantly funny. She seemed to carry the wisdom of a bone that has broken, repaired itself, and become stronger. She had written a book and was working on another. She maintained a blog, helped raise millions of dollars for research, and was, by far, among the most articulate, introspective teenagers that I have ever encountered. I asked her about her condition, and she spoke frankly of the anguish it had caused in her family. “Her biggest fear was that we wouldn’t find anything. Not knowing would be the worst thing,” her father once said.