The ability to exit and re-enter the cell division cycle in response to altered nutrient conditions is of paramount importance for cellular viability and cooperation. When starved, diploid yeast cells enter the sexual cycle by sporulation, while haploid cells arrest in a stationary phase. Cells in these two quiescent stages share many properties; they do not proliferate, their metabolic, transcriptional and translational rates are low, and they are surrounded by a thickened cell wall that provides resistance to harsh environments . Consistently, their modes of re-entry into active division share many features, including initiation by a fermentable carbon source, acquisition of increased sensitivity to cell wall degrading enzymes and heat, and they rely on similar gene expression programs as described here and previously [11, 24]. It has been reported for stationary phase cells that RNA Polymerase II is positioned upstream of genes rapidly expressed upon addition of nutrients and glucose , and we find that many of the same genes are highly responsive during onset of spore germination. This indicates that a similar positioning may exist in spores, although further experimental evidence remains to be provided.
Further, we identified a common early and transient peak of transcription, including genes encoding transporters and transcriptional regulators that presumably are required for setting up proliferation after spore dormancy . We also see that glucose repression is quickly turned on and that mRNAs for gluconeogenesis, the TCA cycle and the carboxylate cycle are rapidly and strongly down-regulated both during germination and stationary phase exit (Hap 2–4, Rox1; Figure 5) [11, 24]. The identification of these subprograms also during germination advances our understanding of the germination program and further bridges the differences between re-entry from the different quiescent stages.
There are also discrepancies between dormancy exit of spores and stationary cells. In our hands, germination is associated with down-regulation of Gcr1/Gcr2-regulated glycolytic mRNAs, which does not occur during quickening from stationary phase. This is counterintuitive as resumption of growth in glucose containing medium requires glycolysis. We can only speculate about the possible purpose of this. To germinate quickly is crucial in the competition with other micro-organisms for nutrients and colonization of the micro environment; perhaps the spores contain high levels of these glycolytic mRNAs for translation immediately upon germination signals as mRNAs already present in the dormant spores most likely is the first choice for translation upon germination initiation . It is also possible that the mRNAs of glycolytic genes, if bound to ribosomes or accumulated in P-bodies, are protected from degradation in the dormant spore and therefore enriched. Whether these expression changes have any impact on the glycolytic protein levels during germinating still needs to be determined. We also see that Hsf1 activation of genes associated with protein folding and refolding is more prominent during germination than stationary phase exit. The dormant spore contains high levels of trehalose that serves a protective role for proteins [5, 26, 27]. This trehalose is rapidly mobilised during germination and the Hsf1 mediated response may be required to protect or refold proteins during this transition [28, 29]. Consistently, trehalose accumulates at high temperatures [29, 30], while protein levels of chaperone Hsp104 increase only upon heat stress relief (Youlian Goulev, personal communication). Stationary phase cells also accumulate trehalose , and we note that genes implicated in protein (re)folding are up-regulated also during exit from stationary phase. The absence here of a significant peak of Hsf1 activation could reflect that these cells accumulate relatively little trehalose, are less dehydrated or that the first time point examined, six minutes into the quickening process, is too late to capture the Hsf1 peak. Hence, most apparent discrepancies between the two quiescent states may be quantitative rather than qualitative.
We further characterise the spore quickening by examining the effect of re-entry of media on the gene expression program. Rich growth medium with glucose has proved to be a more powerful germinant than glucose alone when it comes to acquisition of zymolyase sensitivity and trehalose breakdown [10, 28]. This holds true also at the level of gene expression; although the transcriptional outputs are qualitatively very similar, YPD induces a stronger, faster and/or more synchronised response than pure glucose. These results support the idea that spores sense not only glucose but also the presence of nutrients and growth factors in the environment. In particular, the presence of amino acids appears essential for the rapid and massive induction of ribosomal genes that we observe exclusively in YPD . This implicates the TOR pathway in the quickening process but it remains unclear if glucose is the initial trigger that enables other nutrients to be sensed, or if energy and other nutrients are sensed simultaneously. We note that in glucose-treated spores, TF activity and genes for amino acid biosynthesis are induced within half an hour, clearly showing that the limitation of amino acids is sensed early during the germination process. The ribosomal protein genes are induced within a similar time window, suggesting that the delay in ribosomal up-regulation reflects the delayed amino acid supply in glucose The TFs for amino acid biogenesis displayed rapid and sustained induction also during YPD-induced stationary phase exit (Gcn4 and Leu3; Figure 5), possibly reflecting differences between the auxotrophic BY4741 used for stationary phase exit and the prototrophic Y55 strain used in this study. However, this is not reflected in a delayed induction of the ribosomal protein genes showing that they are regulated by distinct mechanisms and/or cues.
Finally, we resolve the gene expression profile in temporally distinct subprograms that are orchestrated by interconnected TFs. The data set we present here has a higher initial time resolution than previous studies, more stringent evaluation criteria, less noise and hence more coherent clusters, and benefit from the increased resolution gained by two-dimensional clustering. This allows us to analyse the temporal and compositional organisation of the germination gene expression program. As shown in Figure 4, the candidate TF network driving the quickening is highly interconnected and the temporal organisation of the response reflects the interconnections in this network and the activities of interconnected TFs are highly synchronous. In particular, we see an early peak of the Sok2-Yap6-Phd1 cluster, which has been implicated in the metabolic and respiratory oscillations . This is paralleled by a strong activity in Hsf1 as discussed above. The Sok2-Yap6-Phd1 peak is followed by rapid and sustained induction of ribosomal genes in YPD (Fhl1, Sfp1, Rap1) and biosynthetic genes (Gcn4, Bas1, Leu3, Dal81) in glucose where ribosomal protein genes do not peak until after full induction of the biosynthetic program (Figure 4). Interestingly, the Hsf1 peak is sustained until the ribosomal induction, which may suggest that its down-regulation is dependent on amino acid and/or protein synthesis. This delayed response in glucose is also reflected in a slower mating response, again potentially reflecting a need for protein synthesis in the acquisition of sexual (haploid) identity. Finally, germination in glucose leads to a second biosynthetic peak (Met31-Met32-Met4-Cbf1) that is absent in YPD. Also this cluster has been implicated in metabolic oscillations and constitutes a key regulator of the biosynthetic program peaking in the oxidative phase . The absence of this peak in YPD could be due to the abundant amino acids in rich media or the entry into a different subprogram: Mating. Taken together, the transition from quiescence to growth primarily includes metabolic changes and key regulators of oscillatory metabolism appear to strongly influence also germination, which may reflect a highly synchronised return to active metabolism. We find the lack of germination specific targets noteworthy and conclude that this highly complex program is primarily a metabolic adaptation that adapts its subprogram composition to the exact environmental conditions of the quickening spores.