Many studies have used chemostats and gene expression microarrays to characterize the growth rate response of the budding yeast ( Saccharomyces cerevisiae) growing on glucose carbon source (Hayes et al., 2002; Pir et al., 2006; Regenberg et al., 2006; Castrillo et al., 2007; Brauer et al., 2008). These studies demonstrated a common growth rate response (GRR) in continuous exponentially growing cultures, both aerobic and anaerobic and limited by different natural nutrients as well as by auxotrophic requirements. However, in all studies the carbon source was glucose, which is highly preferred by S. cerevisiae and special in many ways (Zaman et al., 2008). Thus, it is not clear how much of the identified GRR is specific to growth on glucose (Zaman et al., 2009; Futcher, 2006) as a sole carbon source and how much of the GRR is general to growth and independent of the carbon source. In fact, Zaman et al. (2009) have suggested that much of the observed common growth rate response can be due to glucose. To explore whether the common growth rate response is still going to be present in cultures grown on non-fermentable carbon source, I grew S. cerevisiae continuous cultures on ethanol carbon source and measured physiological parameters, gene expression, and metabolites.
I found that the growth rate response of a large number of genes (about 1500) remains very similar on ethanol carbon source and I call this common growth rate response universal growth rate response. Genes with positive universal growth rate response include ribosomal and translation genes. Genes with negative universal growth rate response include autophagy, vacuolar and stress response genes. Remarkably, all genes having universal growth rate response are expressed periodically in the yeast metabolic cycle (YMC) (Tu et al., 2005). Genes whose expression levels increase with growth rate are expressed in YMC phase with high oxygen consumption while genes whose expression levels decrease with growth rate are expressed in YMC phase with low oxygen consumption. To understand better the relationship between the YMC and the growth rate response, I synchronized metabolically continuous cultures and quantified the relationship between the growth rate and the periods of YMC phases. The relative duration of the YMC phase with high oxygen consumption increases with growth rate, which can account quantitatively for the observed universal growth rate response. Furthermore, I measured a linear dependence between the periods of the YMC and the cell cycle, which suggests a switch from YMC to fermentation at growth rates too high for the YMC to ensure reductive period that is long enough for DNA replication.
In contrast to the universal growth rate response, the growth rate response of many other genes is carbon source and/or limitation specific. Some of the carbon source specific growth rate response genes are expected (such as the stronger induction of mitochondrial and ethanol utilization genes in ethanol carbon source compared to glucose) while other carbon source specific growth rate response genes are more surprising, such as genes related to generation of precursor metabolites and energy, microtubules and the cell–cycle. To characterize the underlying regulatory mechanisms behind the observed growth rate response, I identified transcription factors (TFs) likely to mediate the growth rate response and inferred their activities in different nutrients and growth rates using RCweb (Slavov, 2010). Based on the gene expression data, I inferred that some TFs have carbon source dependent activities (GCN4, HAP4, FHL1, YAP5) and even more TFs have growth rate dependent activities, including RAP1, GAT3, CBF1, MET4, INO4, HAP4. Interestingly, for most TFs the change in activity is not reflected in the level of the corresponding mRNA. In ethanol carbon source, I found very strong induction, positive growth rate response and differential usage of isoenzymes in pathways (such as gluconeogenesis, TCA, and ethanol utilization) whose metabolic fluxes are expected to increase with growth rate and to be higher in ethanol compared to glucose carbon source. These findings suggest that transcription likely plays a role in regulating those metabolic pathways, but not in regulating the activities of TFs.
To identify growth rate response differences between auxotrophs and prototrophs, I grew his and lys auxotrophs limited on their auxotrophic requirements at different growth rates. The gene expression data from these experiments indicate significantly weaker induction of autophagy genes in slowly growing auxotrophic cultures compared to prototrophic cultures growing at the same growth rate. From the growth rate experiments with his and lys auxotrophs as well as from batch experiments I discovered very wide distribution of cell sizes (3-5 fold difference in cell volumes) in cultures of auxotrophs starving for their auxotrophic requirement. Both the failure to induce autophagy and the poor control of cell-size are likely to contribute to the lower viability of starving auxotrophs. Based on analysis used by Slavov and Dawson (2009), I identified the genes whose combinatorial regulation is most different between auxotrophic and prototrophic cultures. These genes are likely to mediate glucose wasting by auxotrophs and I experimentally demonstrated that single deletions for of SFP1, CCC1 or CCP1 have highly significant effects on glucose wasting.