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

1

Bloodline as legitimate lens

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

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.

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.

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.

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.

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.

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.

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.