In a genetic selection designed to isolate Escherichia coli mutations
that increase expression of the IS10 transposase gene (tnp), we unexpe
ctedly obtained viable mutants defective in translation initiation fac
tor 3 (IF3), Several lines of evidence led us to conclude that transpo
sase expression, per se, was not increased. Rather, these mutations ap
pear to increase expression of the tnp'-'lacZ gene fusions used in thi
s screen, by increasing translation initiation at downstream, atypical
initiation codons, To test this hypothesis we undertook a systematic
analysis of start codon requirements and measured the effects of IF3 m
utations on initiation from various start codons, Beginning with an ef
ficient translation initiation site, we varied the AUG start codon to
all possible codons that differed from AUG by one nucleotide. These po
tential start codons fall into distinct classes with regard to transla
tion efficiency in vivo: Class I codons (AUG, GUG, and UUG) support ef
ficient translation; Class IIA codons (CUG, AUU, AUG, AUA, and ACG) su
pport translation at levels only 1-3% that of AUG; and Class IIB codon
s (AGG and AAG) permit levels of translation too low for reliable quan
tification. Importantly, the IF3 mutations had no effect on translatio
n from Class I codons, but they increased translation from Class II co
dons 3-5-fold, and this same effect was seen in other gene contexts, T
herefore, IF3 is generally able to discriminate between efficient and
inefficient codons in vivo, consistent with earlier in vitro observati
ons. We discuss these observations as they relate to IF3 autoregulatio
n and the mechanism of IF3 function.