Translation in E. coli is usually initiated at an AUG codon, which encodes the amino acid methionine. In some cases, however, the start codon is GUG, which normally encodes valine. If GUG is used as the start codon, is a tRNA charged with methionine used or one charged with valine, and does the use of GUG as a start codon affect the efficiency of translation?

  • $\begingroup$ Below I think is the most extensive study so far quantitatively comparing all codons as a possible initiation codon in E. coli for sfGFP and Nano-Ogra-luc ("Nanoluc") proteins (published in 2016): "Measurements of translation initiation from all 64 codons in E. coli" Ariel Hecht, Jeff Glasgow, Paul R. Jaschke, Lukmaan Bawazer, Matthew S. Munson, Jennifer Cochran, Drew Endy, Marc Salit doi.org/10.1101/063800 doi.org/10.1093/nar/gkx070 $\endgroup$ Jan 21, 2019 at 17:08
  • $\begingroup$ There are two substantially different questions here. As this was posted 7 years ago (before I was involved) there may be little to be done. It would be better if they could be separated, and certainly others should be discouraged from doing likewise. $\endgroup$
    – David
    Jan 21, 2019 at 20:42

1 Answer 1


The NCBI translation table translates all alternative start sites as methionines. To my understanding, all translation is initiated by the fMet-tRNA. I don't know if there are any exceptions to this rule.

Regarding translation efficiency, I only found a 1985 paper in PNAS (Reddy et al, PNAS 82:5656-60), in which they compared the translation efficiency of adenilate cyclase's own UUG start vs GUG or AUG, obtaining a translation ratio 1:2:6 UUG:GUG:AUG, suggesting that AUG is the most efficient one, followed by GUG. Also, Romero and Garcia, FEMS microbiology letters 84:325-330 (1991) compared the efficiency of AUG vs AUC, AUA and AUU, showing a much lower efficiency for those codons, but they did not compare it to GUG.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .