In the fast-paced life of a bacterium, the ability to manufacture proteins quickly and efficiently is crucial. In these organisms, mRNAs—the templates for building proteins—have a string of bases near the start called the Shine-Dalgarno (SD) sequence. This motif increases the rate at which translation is initiated. Some results suggest that the presence of SD sequences further into an mRNA, in the coding region, actually slow the elongation rate—but work by a few other groups does not support this claim.

In the November issue of G3, Yang et al. reason that if SD sequences do slow elongation, there should be fewer of the sequences than would be expected (given codon use biases) in the coding regions of bacterial mRNAs. Using data from many species of bacteria, they found that not only are SD sequences rarer in coding regions, but they are also even more depleted in the genes that are most expressed—and the effect was greatest in the bacteria with the shortest generation times.

In addition to slowing elongation, there are a few other possible explanations for the dearth of SD sequences in coding regions. The sequences may be mistaken by ribosomes for translation start sites, resulting in truncated proteins, or they may cause ribosomes to slip up and result in a frameshift. It’s also possible that the ribosomes get stuck on the SD sequences, leading to a reduced pool of ribosomes available to start translation.

If SD sequences are so detrimental in this context, it might seem strange that the coding regions of bacterial mRNAs would have any SD sequences at all. But natural selection typically keeps bacterial genomes small—a trim genome allows bacteria to replicate quickly and potentially beat out its competitors—so it’s possible that the initiation sites for a gene might need to overlap with the end of the prior gene in the operon. And despite being potentially harmful, random mutations ensure that some of these misplaced sequences will arise by chance.

An important implication of this work is that the presence or absence of SD sequences within coding regions is a significant modulator of translation efficiency. So far, much research has been dedicated to codon usage as the primary contributor to translation speed, since codons calling for rarer tRNAs might take longer to be translated. But according to these researchers’ results, SD sequences in coding regions also have a significant effect. This implies that when designing artificial genes to insert into bacteria, researchers might want to avoid including SD sequences if they hope to get the highest protein yield. Thinking like a bacterium may not sound like a good strategy for a scientist, but after being molded by billions of years of competition with their fellow microbes, these organisms do seem to have learned a trick or two.


Yang, C.; Hockenberry, A.; Jewett, M.; Amaral, L. Depletion of Shine-Dalgarno Sequences within Bacterial Coding Regions Is Expression Dependent.
G3, 6(11), 3467–3474.

Nicole Haloupek is a freelance science writer and a recent graduate of UC Berkeley's molecular and cell biology PhD program.

View all posts by Nicole Haloupek »