Regulation of mitochondrial gene expression in the yeast, Saccharomyces cerevisiae

In order to elucidate the mechanisms governing mitochondrial (mt) transcription the yeast Saccharomyces cerevisiae, we examined the nuclear-mt stringent and upshift responses. In isonuclear rho+, 21S rRNA-containing rho-, and rho0 strains of yeast, transcription of nuclear 18S+26S rDNA sequences fell to 20% of control values within 30 minutes after downshift from 1X to 0.05X minimal media. Transcription of mt DNA also decreased to 20% of control values and at all time points tested, the kinetics of curtailment of mt rDNA transcription mirrored that for nuclear rDNA transcription. Identical results were obtained when the stringent response was elicited in the same strains via starvation for an amino acid. Both in vivo and in an in vitro transcription assay, the mt stringent response was identical between rho- and rho+ strains, indicating that control of the mt stringent response lies outside of mitochondria. When actively growing rho+ cells are upshifted from ethanol to glucose based media, transcript levels for nuclear 18S+26S rRNA increased 2-fold within 30 minutes. Surprisingly, mt transcription, which is repressed during fermentative growth on glucose, also transiently increased 1.8-fold. Thus, the mechanism which regulates mt transcription during fermentative growth is not immediately repressible. In an effort to begin defining the mechanism(s) controlling mt transcription we examined several RAS mutant strains of yeast. In S. cerevisiae adenosine 5'3'cyclic monophosphate (cAMP) levels are controlled via the RAS1 and RAS2 gene products. Via Northern blots, we analyzed transcript levels for the mt 21S rRNA and in vitro, the rate of production of mt transcripts in wild type and RAS mutants under various growth conditions. Our data showed that RAS1,ras2 cells show measurable amounts of cAMP and have normal mt transcript levels in glucose-based medium but contain negligible amounts of cAMP and only 5-10% of normal mt transcript levels during growth on ethanol-based medium; the rate of production of mt transcripts in vitro from RAS1,ras2 cells were significantly less than its wild-type parental strain when grown on ethanol-based medium. Mt 21S transcript levels in a ras1,RAS2^[val19] mutant, in which cAMP levels were shown to be constitutively 4-fold higher than in nonmutant cells, were concomitantly higher than in the parental strain, when both strains were grown in glucose-based medium where such transcripts are unneeded. Furthermore, the rate of production of mt transcripts in isolated organelles derived from our ras1,RAS2^[val19] mutant was also higher than that in the parental strain under the same experimental conditions. Addition of uM amounts of cAMP restored transcription in mitochondria isolated from nutritionally downshifted cells up to 75% of control organelles. When ras1,RAS2^[val19] cells were nutritionally downshifted, no mt stringent response was observed in vitro. Conversely, addition of cAMP to mitochondria from amino acid starved cells did not relieve mt transcriptional curtailment. Furthermore, the mt transcription rate in isolated organelles from the ras1,RAS2^[val19] mutant, which showed no stringent response during nutritional downshift, was as severely curtailed as its wild-type parental strain during amino acid starvation. When the regulatory subunit gene (BCY1) of the cytoplasmic cAMP dependent protein kinase was insertionally inactivated, endogenous mt cAMP dependent protein kinase activity was 2-3 fold higher in the bcy1 mutant than its parental, BCY1, strain. Additionally, no mt stringent response was observed in organelles derived from nutritionally downshifted cells in bcy1 mutants; a full mt stringent response was observed in bcy1 mutants during amino acid starvation. Measurement of mt cAMP dependent protein kinase activity in mitochondria derived from nutritionally downshifted yeast cells showed 2-3-fold lower mt cAMP dependent protein kinase activity compared to organelles derived from non-downshifted cells suggesting that a cAMP dependent phosphorylation event is critical for mt transcription. Taken together these data suggest that nuclear and mt rDNA transcription is coordinately regulated under some growth conditions. Further, the results indicate that neither mt transcription nor protein synthesis contribute to the overall nuclear-mt transcriptional regulatory system. Rather, mt transcription requires some minimal cellular level of cAMP for mt cAMP protein kinase activity. That cAMP does not restore mt transcription during starvation for a required amino acid strongly suggests that the target for phosphorylation in mitochondria is a cytoplasmically synthesized protein. This provides a mechanism by which one mt process is tightly linked to overall growth control, as well as a part explanation of how yeast adjust their mt metabolism to changes in environmental conditions.