The genetic code

DNA stores instructions as a sequence of four bases, but proteins are chains of twenty different amino acids. The genetic code is the translation table between them. It works by reading the sequence in non-overlapping groups of three — codons — each of which specifies one amino acid to be added to a growing protein chain. Four letters taken three at a time give 64 possible codons, more than enough for twenty amino acids plus "start" and "stop" punctuation, so the code is redundant: several codons can name the same amino acid. And it is very nearly universal — the same table runs in a bacterium, an oak, and a human, which is the strongest single piece of evidence that all life shares one ancestry.

The double helix showed how genes are *copied*; the code showed how they are *expressed* — how the inert sequence in a chromosome is turned into the working machinery of a cell. It is the missing link between genotype and phenotype, between the letters in the DNA and the organism they build. Knowing it means you can read a gene and predict the protein it makes, spot how a single misread letter causes a disease like sickle-cell anemia, and — eventually — write new instructions of your own.

The great theoretical worry was how to read a continuous string of letters without spaces. The empirical crack came from an unexpected, almost crude experiment. Marshall Nirenberg and Heinrich Matthaei at the U.S. National Institutes of Health built a "cell-free" system — the guts of *E. coli* cells that would still stitch amino acids into protein if fed RNA — and then fed it an artificial RNA made of a single repeated base. Uracil-only RNA produced protein made of a single repeated amino acid, phenylalanine. UUU meant phenylalanine. When Nirenberg presented the result in Moscow that August, he was an unknown; barely three dozen people came to his talk. Within five years, he, Har Gobind Khorana, and others had filled in the entire 64-codon table, and in 1968 Nirenberg shared the Nobel Prize.

The finished code is the reference key of molecular biology. It made it possible to read what any gene does, to engineer organisms to manufacture human proteins like insulin, and to design DNA sequences to order. Every downstream technology that treats a gene as text you can read and rewrite — genetic engineering, synthetic biology, the reading of whole genomes — depends on possessing this dictionary.