蛋白質合成中的遺傳調控機制

The review opens by splitting the genome's labour in two. Some genes — the structural genes — carry the sequence of a protein: in the lactose system of E. coli these are z, y and a, specifying β-galactosidase, a galactoside permease and a transacetylase. Other genes do nothing but govern: a regulator gene makes a diffusible product whose only function is to control whether the structural genes are read.

That regulatory product is named the repressor. In the absence of an inducer it keeps the structural genes silent — not by acting on each gene, but by binding a single controlling site on the DNA, the operator, that sits at one end of the cluster. Genetics had already pointed the way: operator-constitutive mutants escape control and express the genes whatever the conditions, and they act only on the genes physically next to them on the same chromosome.

The operator together with the block of structural genes it commands is proposed as a single unit of coordinated expression — christened the operon. Because the whole block is governed through one operator, its genes are switched on and off together; this is why the three lactose enzymes rise and fall in concert.

Induction, the paper argues, is simply the lifting of repression. The inducer is a small molecule that combines with the repressor and inactivates it; with the repressor disabled, the operator is freed and the operon is read. The logic is a double negative — the inducer represses the repressor — and it unifies enzyme induction with the mirror-image case of repression, where a small molecule instead activates a repressor to switch a pathway off.

For the scheme to work, the structural genes' instructions must reach the protein-making machinery quickly and transiently. Jacob and Monod propose an unstable intermediate — a short-lived 'messenger' RNA, copied from the structural genes and carried to the comparatively stable ribosomes, present only while the operon is on. (The messenger was demonstrated experimentally the same year.)

The review is candid that the chemical nature of the repressor is unknown, and that the model is built largely from genetics rather than from isolated molecules. It does not settle how the operator is recognised, nor prove that any of these objects could be purified — questions it hands, explicitly, to the experiments it provokes.