Translational regulation of mitochondrial gene expression by nuclear genes of Saccharomyces cerevisiae

Abstract

We describe several yeast nuclear mutations that specifically block expression of the mitochondrial genes encoding cytochrome c oxidase subunits II (COXII) and III (COXIII). These recessive mutations define positive regulators of mitochondrial gene expression that act at the level of translation. Mutations in the nuclear gene PET111 completely block accumulation of COXII, but the COXII mRNA is present in mutant cells at a level approximately one-third of that of the wild type. Mitochondrial suppressors of pet 111 mutations correspond to deletions in mtDNA that result in fusions between the cox II structural gene and other mitochondrial genes. The chimeric mRNAs encoded by these fusions are translated in pet 111 mutants; this translation leads to accumulation of functional COXII. The PET111 protein probably acts directly on cox II translation, because it is located in mitochondria. Translation of the mitochondrially coded mRNA for COXIII requires the action of at least three nuclear genes, PET 494, and a newly discovered gene, provisionally termed PET 55. Both the PET494 and PET54 proteins are located in mitochondria and therefore probably act directly on the mitochondrial translation system. Mutations in all three genes are suppressed in strains that contain chimeric cox III mRNAs with the 5'-untranslated leaders of other mitochondrial transcripts fused to the cox III coding sequence. The products of all three nuclear genes may form a complex and carry out a single function. A direct demonstration that the wild-type nuclear gene products act in the cox III 5'-leader has been obtained by showing that they are all required for translation of apocytochrome b from a novel mRNA consisting of the cox lIl 5'-leader attached to the cytochrome b coding sequence. The site (or sites) of action maps at least 172 bases upstream from the cox lll initiation codon in the 600 base cox III leader. Others have reported evidence which suggests that cox Ill translation is repressed by glucose. Consistently with the possibility that the nuclear genes described here may play a role in modulating mitochondrial gene expression, we have found that PET 494 expression is glucose-repressed.