Codon reading and translational error. Reading of the glutamine and lysine codons during protein synthesis in vitro

The reading of glutamine and lysine codons during protein synthesis in vitro has been investigated using an MS2-RNA-programed system derived from Escherichia coli. Under conditions when either glutaminyl-tRNA1Gln (s2UUG) or glutaminyl-tRNA2Gln (CUG) was the only source of glutamine for protein synthesis both tRNAs were able to read the glutamine codons CAA and CAG as indicated by the incorporation of labeled glutamine into the pertinent coat protein tryptic peptides. On the other hand, when the two glutamine tRNAs competed for the codon CAA the reading efficiency of the anticodon s2UUG, which reads the codon according to the wobble rules, was almost 40 times higher than that of the competing anticodon CUG, which reads the codon by "two out of three," i.e. it cannot form a regular base pair with the third codon position. In reading the codon CAG the anticodon CUG was approximately eight times more efficient than the anticodon s2UUG. The lysyl-tRNA1Lys (CUU) could not alone sustain any detectable coat protein synthesis in the MS2 system indicating that there was no significant reading of the lysine codon AAA. This conclusion is supported by the outcome of experiments where lysyl-tRNA1Lys (CUU) and lysyl-tRNA2Lys (s2UUU) competed for the codon AAA. The reading efficiency of the anticodon CUU was less than 1% of that of the competing s2UUU which represents the limit of resolution of our experimental system. When the two lysine tRNAs competed for the codon AAG the anticodon CUU was about four times more efficient than s2UUU. These results are discussed in the context of the two out of three hypothesis, which attempts to relate the frequency of such reading to the hydrogen bonding properties of the codon nucleotides.