How the gene became medicine, atrocity, and identity's mirror

By Books (/journals/books)

In 60 seconds Mukherjee argues no scientific idea has done more to shape and endanger how humans understand themselves than the gene. Mendel proved heredity moves in discrete countable units, a finding ignored from 1866 until its 1900 rediscovery. Eugenics was a mainstream extension of Darwinian logic that sterilized roughly 60,000 Americans and informed Nazi policy. DNA's double helix and triplet code reframed heredity as readable, rewritable information rather than vague substance. The Human Genome Project found only about 20,000 to 25,000 human genes, far below the 100,000 many expected. CRISPR makes germline editing newly feasible and revives the old question of which human variations count as disease.

The big idea

Mukherjee argues that every gain in precision about what genes are also multiplied the power to manipulate them and the temptation to treat them as fate. His family's mental illness keeps the stakes personal: a gene is never only a molecule, but a claim about blame, risk, identity, and what parents owe their children.

The book in brief

Mukherjee begins with two uncles and a cousin marked by schizophrenia or bipolar disorder, and a father who half-expected the same fate. That bloodline frames the book's real question: what passes from parent to child, and how fixed is a person before birth. The history that follows is not a clean march of discovery. It is a story of knowledge repeatedly outrunning the culture ready to absorb it.

Mendel's pea crosses in the 1860s showed traits behaving as discrete packets rather than blended fluids, yet the work sat ignored until 1900. Darwin lacked that mechanism; his half-cousin Galton filled the gap by inventing eugenics in 1883 as deliberate human breeding. By the early twentieth century the idea was law. More than thirty U.S. states forced sterilizations; Buck v. Bell in 1927 blessed them; roughly 60,000 Americans were sterilized; Nazi Germany scaled the same logic into mass murder of the 'unfit.'

Laboratory genetics then made the unit physical. Morgan's fruit flies mapped genes onto chromosomes. Watson and Crick's 1953 double helix, built substantially on Franklin's X-ray data, showed how base pairing copies information. The triplet code of the late 1950s and 1960s turned genes into text. Recombinant DNA in the early 1970s let scientists cut and paste that text across species, yielding Genentech, bacterial human insulin by 1982, and the Asilomar attempt at self-regulation.

The Human Genome Project's 2003 draft catalogued three billion base pairs yet revealed only about 20,000 to 25,000 protein-coding genes, forcing attention onto regulation and environment. Jesse Gelsinger's 1999 death in a gene-therapy trial exposed how commercial and clinical enthusiasm can outrun safety. Epigenetics and twin studies then undercut hard determinism: shared sequence does not mean shared fate. CRISPR, refined around 2012, makes germline edits technically plausible and returns the book to its opening moral problem in actionable form.

Context

Siddhartha Mukherjee published the book in 2016 as a physician-writer tracing genetics from peas to gene editing while threading his Indian family's history of mental illness through the science. The timing mattered because CRISPR had just made embryo editing newly real, and public debate needed the deeper record of how genetic knowledge had already been used for medicine and for atrocity.

The key ideas

1

Bloodline as legitimate lens

Mukherjee treats his family's mental illness as a necessary way into heredity science, not memoir decoration.

UUncles and cousinDiagnosed with schizophrenia or bipolar disorder FFatherSpent his life half-expecting the same fate MMukherjeeUses the bloodline as the book's moral stake

Mukherjee opens on two uncles and a cousin diagnosed with schizophrenia or bipolar disorder, and a father who spent a lifetime half-waiting for the same fate in himself or his children. In the India of their lives, stigma was deep and clinical language thin, so families felt heredity as difference long before labs could name it.

He insists the history of the gene cannot be told as pure triumph. The concept has always been tangled with fear, blame, and the urge to explain suffering. Starting inside one bloodline commits the narrative to tracking not only what genes are, but what people have done with the idea of them.

That choice sets the book's moral register. Every later laboratory advance is measured against families who must live with risk numbers, identity stories, and the question of what is fixed before birth.

In practice: Mukherjee's mother calls his uncle Rajesh 'always a little different,' a family phrase that sensed hereditary pattern before diagnosis existed.

2

Heredity in discrete packets

Mendel showed traits travel as intact countable units that can hide in one generation and reappear in fixed ratios in the next.

1860sMendel counts pea trait ratios 1866Results published and ignored 1900Three botanists rediscover and name genes

In the 1860s an Augustinian monk in Brno crossbred peas for flower color and seed shape and counted offspring with obsessive care. Reigning theory expected parental traits to blend smoothly. The ratios refused to blend.

A purple-flowered plant crossed with a white-flowered one yielded purple offspring first, then white again in predictable fractions later. Something was carried intact, hidden then expressed. Mendel published in 1866 in an obscure journal and drew almost no notice. Darwin, who needed exactly such a mechanism, never read the paper.

Only in 1900, sixteen years after Mendel's death, did three botanists rediscover the work and name the units genes. Mukherjee treats the gap as a recurring pattern: genetic truths often exist in principle long before surrounding science or culture can absorb them.

In practice: Mendel's purple-by-white crosses hid white and then restored it in fixed next-generation ratios instead of producing pale blends.

3

Eugenics as respectable logic

Galton's eugenics was a direct extension of Darwinian selection thinking, which is why mainstream science and law adopted it so readily.

Darwin explained species change but left inheritance a black box. His half-cousin Francis Galton pushed the logic sideways: if nature could shape traits over millennia, humans could shape intelligence and character faster by selective breeding. In 1883 he coined eugenics, 'well-born,' and read talent in eminent Victorian family trees while discounting privilege and schooling that traveled with those same trees.

By the early twentieth century the idea was statute. More than thirty U.S. states authorized forced sterilization of people labeled feebleminded, criminal, or insane. In 1927 the Supreme Court upheld the practice in Buck v. Bell; Holmes wrote that three generations of imbeciles were enough. Roughly 60,000 Americans were sterilized. Nazi Germany studied the American statutes and built a vastly larger apparatus of sterilization and murder of those deemed genetically unfit.

Mukherjee's claim is stark. The danger was never only bad actors misusing pure science. The seduction of selection was present in the earliest respectable applications, and every later molecular triumph is meant to be read with that history still in mind.

In practice: Carrie Buck was sterilized after Buck v. Bell even though her supposed disability was never rigorously established.

4

Genes gain physical addresses

Morgan's fly room turned genes from statistical abstractions into locatable objects lined up on chromosomes.

1Spot mutantWhite-eyed fly appears among red-eyed stock 2Track sex linkTrait appears almost only in males 3Map orderSturtevant lines genes along chromosomes

Around 1910 at Columbia, Thomas Hunt Morgan bred millions of rapidly reproducing fruit flies. A white-eyed mutant appeared among red-eyed normals. Tracking showed the trait almost exclusive to males, tying its gene to the sex chromosome.

His student Alfred Sturtevant used linkage patterns to draw the first genetic map: genes in fixed linear order along chromosomes, like beads on a string. Crossing-over during reproduction could break linkages, explaining why paired traits sometimes separated.

Mendel's units had been ratios. Morgan's were addresses. Once genes had a physical home, the next unavoidable question was which molecule inside chromosomes actually carried the information.

In practice: A single white-eyed male fly let Morgan assign its gene to the sex chromosome by sex-linked inheritance.

5

Helix becomes writable code

The double helix and the triplet code together turned heredity into readable information with a physical copy rule.

1Shape1953 double helix shows base-pair copying 2TranscribeDNA copied into messenger RNA 3TranslateTriplet codons build protein

By the 1940s chromosomes were known to hold DNA and protein; most bet on protein as the complex carrier. Oswald Avery's 1944 bacterial work pointed to DNA alone. At Cambridge, Watson and Crick pursued DNA's shape, drawing critically on Rosalind Franklin's X-ray images shown to Watson without her consent—an ethical shadow Mukherjee records candidly. In 1953 they published two strands wound together, bases paired adenine-thymine and guanine-cytosine, so each strand could template its partner.

Structure still did not explain proteins. Through the late 1950s and 1960s, work associated with Marshall Nirenberg and others set the central dogma: DNA to messenger RNA to protein, with each three-base codon specifying an amino acid. Sixty-four codons cover twenty amino acids with redundancy, and the code is nearly universal across life.

A gene was no longer only a unit or an address. It was instructions in a four-letter alphabet. That pivot makes copying, splicing, sequencing, and editing thinkable; later engineering tools descend from form and function arriving together.

In practice: Franklin's diffraction images locked the helix dimensions, and codon tables later matched sixty-four triplets to twenty amino acids across species.

6

Genetics turns into engineering

Recombinant DNA in the early 1970s made genes movable between species and forced scientists to attempt self-regulation before disaster.

1972–73Boyer and Cohen develop recombinant splicing 1975Asilomar drafts voluntary safety rules 1976Genentech founded 1982Bacterial human insulin reaches patients

Herbert Boyer and Stanley Cohen learned to cut DNA with bacterial enzymes and splice foreign stretches into bacterial genomes, turning microbes into protein factories. Genes could now move by design, not only by descent.

Commerce followed fast. Boyer co-founded Genentech in 1976; by 1982 engineered bacteria produced human insulin, replacing animal extracts. The same methods raised fears of engineered pathogens and ecological accidents. In 1975, leading molecular biologists met at Asilomar to draft voluntary safety guidelines before full commercialization—an unusual attempt to police a tool ahead of public force.

Mukherjee marks the decade as the shift from description to industrial and medical technology. He also notes a pattern that returns with CRISPR: transformative tools arrive faster than ethical frames, and sincere self-policing has mixed success at containment.

In practice: Genentech's engineered bacteria were mass-producing human insulin by 1982, a direct clinical replacement for animal-pancreas insulin.

7

The genome's humbling headcount

Finishing the human sequence around 2003 found only about 20,000 to 25,000 genes, undercutting simple one-gene determinism.

Early expectation Roughly 100,000 human genes predicted

Sequence result About 20,000–25,000 protein-coding genes found

Launched in 1990 as a public international effort to read three billion base pairs, the project raced Celera Genomics under Craig Venter in the late 1990s. A joint announcement came in 2000; a fuller draft in 2003. The reference map aided disease-gene hunting and population variation studies.

The surprise was the count. Many had expected roughly 100,000 genes. Humans have about 20,000 to 25,000 protein-coding genes, not far above a roundworm's roughly 20,000, and fewer than some plants. Complexity could not be raw gene number.

Much of the genome is regulatory sequence controlling when and how strongly other genes fire. Mukherjee argues the result deflated naive determinism the project's early rhetoric had encouraged. Outcomes depend on interaction, overlap, regulation, and environment.

In practice: Human protein-coding gene counts landed near a roundworm's roughly 20,000 instead of the predicted 100,000.

8

When therapy outran caution

Jesse Gelsinger's 1999 death showed the lethal gap between understanding a genetic mechanism and delivering it safely.

Gene therapy aimed to insert corrected genes into patients' cells, often via modified viruses. In September 1999, eighteen-year-old Jesse Gelsinger joined a University of Pennsylvania trial for a diet-managed form of ornithine transcarbamylase deficiency. He volunteered partly to help sicker infants.

Days after the viral-vector dose, a massive immune response triggered multi-organ failure and death. Investigations found thin risk disclosure, financial conflicts among researchers, and underweighted animal warning signs. The field pulled back under tighter oversight.

Mukherjee reads the case as a recurring failure mode: science, investment, and clinical ambition can align to rush a treatment before risks—here, immune reaction to vectors—are mature. Later successes came only after slower, more incremental rebuilding of trust.

In practice: Gelsinger died from an immune storm against the viral delivery vector days after dosing in the 1999 Pennsylvania trial.

9

Same sequence, open fate

Identical genomes still yield different lives because environment, chance, and epigenetic switches shape expression.

Shared genome Identical DNA sequence in twin pairs

Divergent lives Different illness, personality, and trait outcomes

Identical twins share essentially the same DNA yet diverge in personality, disease, and some physical traits, especially with age or separate upbringing. One twin may develop schizophrenia while the other never does.

Mukherjee points to epigenetics: chemical marks such as methylation that turn genes on or off without rewriting sequence, triggered by diet, stress, trauma, and other exposures. Some marks can pass across a generation or two, blurring older nature-nurture lines.

He returns to his own family through this lens. Predisposition is real and measurable, yet most carriers of known risk variants never become ill. Onset hinges on unpredictable environmental combinations. Genes set ranges and probabilities, not sentences—the corrective the book has been building toward after its long arc from Mendelian units to full sequence.

In practice: Twin pairs sharing a genome still split on schizophrenia, showing sequence alone does not dictate onset.

10

Editing revives the old choice

CRISPR makes germline changes to human embryos technically plausible, forcing a collective decision about disease versus variation.

Around 2012, Jennifer Doudna, Emmanuelle Charpentier, and others adapted a bacterial immune system into CRISPR-Cas9, a cheap, fast way to cut DNA at chosen sites and rewrite sequence. Unlike earlier gene therapy that mainly added genes, CRISPR can correct existing ones and spread quickly through labs worldwide.

Near-term somatic uses—fixing sickle cell or cystic fibrosis only in the treated patient—look comparatively bounded. Germline edits to embryos or reproductive cells would pass to all descendants. That reopens, in working technical form, the eugenics question: who decides which traits are disorders to erase and which are human variation, especially when parents and clinics hold the tools rather than states alone.

Mukherjee argues CRISPR does not invent a new ethics problem. It collapses the distance between Galton's theories and instruments that function, returning the book to how much of a person is fixed and how much of that fixedness we can now choose.

In practice: Mukherjee contrasts somatic edits that end with the patient against germline edits that would alter every future descendant.

Key terms Gene The discrete hereditary unit Mendel counted, later found as a stretch of DNA that carries instructions. Eugenics Galton's program of improving humans by selective breeding, later written into sterilization law. Chromosome map Morgan and Sturtevant's linear order of genes along physical chromosomes. Double helix DNA's two paired strands whose base rules explain exact copying. Central dogma The flow DNA to RNA to protein via triplet codons. Recombinant DNA Cutting genes from one organism and splicing them into another to make living factories. Epigenetics Chemical marks that switch genes without changing sequence, often shaped by environment. CRISPR A bacterial defense turned into a precise, cheap tool for rewriting targeted DNA. Germline edit A genetic change to embryos or reproductive cells that inherits to future generations.

What to do

Book a genetic counselor's assessment of any family hereditary condition before trusting online calculators or family folklore. Read your workplace policy on genetic screening or embryo selection and note which choices go to individuals versus committees. When a headline claims a gene for a complex trait, ask what share of variation it explains and which environments were controlled. List the somatic versus germline boundary in any gene-editing story you follow this week. Write down one family health story you treat as destiny and mark what is known sequence risk versus unknown trigger.

Questions to think with Where should the line sit between treating a single-gene disease and editing traits that are only probabilistic risks? How should families talk about hereditary mental illness without turning predisposition into blame or prophecy? If eugenics grew from mainstream science once, what present-day incentives could repeat that pattern with CRISPR? What does a genome of roughly 20,000 genes change about everyday claims that 'it's genetic'? Who ought to decide when a variation is a disorder to correct rather than diversity to protect?

The other side

The narrative leans on a mostly Western, male canon and still centers Watson and Crick even while crediting Franklin more than older popular accounts; scholars of science and gender argue her role remains understated. Non-Western contributions and the global south's experience of genetic science get limited space despite the Indian family frame. Twin-study heritability figures the book uses are contested in behavioral genetics and sociology and easy for readers to overread. Published in 2016, the book predates He Jiankui's 2018 gene-edited babies and the commercial rise of polygenic embryo screening, so its closing speculation is already partly overtaken. Some geneticists also say the epigenetics chapter grants emerging multigenerational human claims more settled weight than the primary literature supports.

Who it's for

Readers who want genetics as a story with characters and moral weight, people weighing family hereditary illness, and anyone following gene editing, embryo screening, or genetic privacy who needs the deeper history behind the headlines.

Read the full summary in Library →