Co-translational protein folding in vivo
The hypothesis that translational pausing, resulting from groups of consecutive rare codons, facilitates co-translational protein folding in vivo was tested. The bifunctional TRP3 gene of S. cerevisiae encodes indoleglycerolphosphate synthase (IGPS) and anthranilate synthase component II (ASII). A group of ten consecutive rare codons within the IGPS domain was replaced with synonymous preferred codons and protein folding was monitored by measuring the AS and IGPS enzyme activities. Upon removal of the translational pause, there was a small, but statistically significant decrease in the AS and IGPS enzyme activities produced per TRP3 mRNA copy. This was unlikely to be due to decreased translation of the mRNA as a consequence of altered mRNA secondary structure. There was no apparent dosage effect on TRP3 expression and since the codon bias of the gene had been improved, translation might have been expected to increase. The amino acid sequence of the protein was unchanged, therefore the decreases in enzyme activities that were observed upon removal of the translational pause are consistent with an effect on protein folding. Analysis of the predicted secondary structure of the S. cerevisiae IGPS protein revealed a large insertion in the protein (relative to IGPS proteins from other organisms). It is possible that the translational pause has evolved to allow the protein synthesized before this region to begin to fold correctly before synthesis of the large insertion. Experiments carried out to investigate the effect of removal of the translational pause in chaperone mutant strains suggested that the SSA1, 2 and 4 and the SSB1 and 2 proteins are involved neither in the folding of the wild-type TRP3 protein, nor in rescuing mis-folded protein resulting from the removal of the translational pause.