| As a result of codon redundancy, a dipeptide can be encoded by as many as 36 different synonymous codon pairs. Some of these codon pairs are utilized more frequently than the others. This phenomenon is termed "codon pair bias". Previous results have shown that re-encoding viral proteins with less frequently used synonymous codon pairs (hence lower codon pair score) strongly attenuates poliovirus and influenza virus and that these attenuated viruses can be candidates for effective live vaccines. However, the molecular mechanism of codon pair bias is still unknown. We employ budding yeast as a model organism to study the mechanism. Two yeast genes, HIS3 and LYS2, were synthesized de novo so that a major portion of their original codon pairs are replaced by synonymous ones, to achieve a codon pair score of either much lower than (named de-optimized genes) or equal to the wild type gene (named scramble genes). These synthetic genes were then introduced into the yeast genome to replace their wild type counterparts. Since these two genes encode enzymes that are essential for amino acids biosynthesis, de-optimized genes caused severe growth defects on synthetic drop out medium. However, as a negative control, scramble genes had only minor if any effect on yeast growth, indicating that the growth defects were caused by using less favored codon pairs rather than simply changing of codon pairs. In addition, a mutation in the RNA binding domain of Hrp1, a protein involved in mRNA processing, was found to suppress the growth defects of the de-optimized lys2-1 (dlys2-1) gene. It was later found that a 3' cleavage and polyadenylation signal was randomly introduced into the dlys2-1 gene by de-optimization. It is possible that mutant alleles of Hrp1 recognize this site inefficiently, allowing some full-length dlys2-1 mRNA to escape and make functional protein. |
