the Shine-Dalgarno sequence
/ SHYNE dal-GAR-no /
A bacterium's ribosome cannot just scan an mRNA from its tip looking for a start codon — bacterial mRNAs often have no convenient end to begin at, and several genes can sit on one mRNA. So how does the ribosome find each gene's AUG? It looks for a short landmark a few bases upstream, the Shine-Dalgarno sequence, which acts like a signpost reading 'start codon just ahead'.
The Shine-Dalgarno sequence is a short purine-rich stretch, classically something like AGGAGG, sitting roughly five to ten bases before the start codon. Its trick is direct base-pairing: the very end of the ribosome's small-subunit RNA (the 16S ribosomal RNA) carries a complementary sequence, and the two zip together. This pairing parks the small subunit at exactly the right distance so that the start AUG falls into the P site, fixing the reading frame in one move. It is found in bacteria and archaea, and it is part of why bacterial translation can begin in the middle of a long mRNA, not only at its end.
This is one of the cleanest differences between bacterial and eukaryotic translation, and it has real consequences. Because several Shine-Dalgarno-plus-AUG units can sit along one bacterial mRNA, a single bacterial mRNA can encode several proteins (a polycistronic message) — something eukaryotic scanning generally cannot do. For genetic engineers, supplying a strong Shine-Dalgarno sequence is a standard way to make bacteria pump out lots of a desired protein. Eukaryotes have no Shine-Dalgarno sequence; they find the start codon by the entirely different cap-and-scan route.
A bacterial mRNA reading 5'-...AGGAGG-(7 bases)-AUG...-3' has its Shine-Dalgarno (AGGAGG) base-pair with the 16S rRNA's tail, positioning the ribosome so that the downstream AUG lands in the P site.
The SD sequence base-pairs with rRNA to position the start codon in the P site.
The Shine-Dalgarno sequence is a bacterial/archaeal feature. Eukaryotes lack it entirely and instead scan from the 5' cap, which is one reason bacterial mRNAs can be polycistronic and eukaryotic ones usually are not.