👤 Mike Tyers

Commitment to cell division at the end of G1 phase, termed Start in the budding yeast Saccharomyces cerevisiae, is strongly influenced by nutrient availability. To identify new dominant activators of Start that might operate under different nutrient conditions, we screened a genome-wide ORF overexpression library for genes that bypass a Start arrest caused by absence of the G1 cyclin Cln3 and the transcriptional activator Bck2. We recovered a hypothetical gene YLR053c, renamed NRS1 for Nitrogen-Responsive Start regulator 1, which encodes a poorly characterized 108 amino acid microprotein. Endogenous Nrs1 was nuclear-localized, restricted to poor nitrogen conditions, induced upon TORC1 inhibition, and cell cycle-regulated with a peak at Start. NRS1 interacted genetically with SWI4 and SWI6, which encode subunits of the main G1/S transcription factor complex SBF. Correspondingly, Nrs1 physically interacted with Swi4 and Swi6 and was localized to G1/S promoter DNA. Nrs1 exhibited inherent transactivation activity, and fusion of Nrs1 to the SBF inhibitor Whi5 was sufficient to suppress other Start defects. Nrs1 appears to be a recently evolved microprotein that rewires the G1/S transcriptional machinery under poor nitrogen conditions. Show less

Progression through the cell cycle requires the coordination of basal metabolism with the cell cycle and growth machinery. Repression of the sulfur gene network is mediated by the ubiquitin ligase SCF(Met30), which targets the transcription factor Met4p for degradation. Met30p is an essential protein in yeast. We have found that a met4Deltamet30Delta double mutant is viable, suggesting that the essential function of Met30p is to control Met4p. In support of this hypothesis, a Met4p mutant unable to activate transcription does not cause inviability in a met30Delta strain. Also, overexpression of an unregulated Met4p mutant is lethal in wild-type cells. Under non-permissive conditions, conditional met30Delta strains arrest as large, unbudded cells with 1N DNA content, at or shortly after the pheromone arrest point. met30Delta conditional mutants fail to accumulate CLN1 and CLN2, but not CLN3 mRNAs, even when CLN1 and CLN2 are expressed from strong heterologous promoters. One or more genes under the regulation of Met4p may delay the progression from G(1) into S phase through specific regulation of critical G(1) phase mRNAs. Show less

In most eukaryotes, commitment to cell division occurs in late G1 phase at an event called Start in the yeast Saccharomyces cerevisiae, and called the restriction point in mammalian cells. Start is triggered by the cyclin-dependent kinase Cdc28 and three rate-limiting activators, the G1 cyclins Cln1, Cln2 and Cln3. Cyclin accumulation in G1 is driven in part by the cell-cycle-regulated transcription of CLN1 and CLN2, which peaks at Start. CLN transcription is modulated by physiological signals that regulate G1 progression, but it is unclear whether Cln protein stability is cell-cycle-regulated. It has been suggested that once cells pass Start, Cln proteolysis is triggered by the mitotic cyclins Clb1, 2, 3 and 4. But here we show that G1 cyclins are unstable in G1 phase, and that Clb-Cdc28 activity is not needed fgr G1 cyclin turnover. Cln instability thus provides a means to couple Cln-Cdc28 activity to transcriptional regulation and protein synthetic rate in pre-Start G1 cells. Show less

Regulated cell cycle progression depends on the proper integration of growth control pathways with the basic cell cycle machinery. While many of the central molecules such as cyclins, CDKs, and CKIs are known, and many of the kinases and phosphatases that modify the CDKs have been identified, little is known about the additional layers of regulation that impinge upon these molecules. To identify new regulators of cell proliferation, we have selected for human and yeast cDNAs that when overexpressed were capable of specifically overcoming G1 arrest signals from the cell cycle branch of the mating pheromone pathway, while still maintaining the integrity of the transcriptional induction branch. We have identified 13 human CPR (cell cycle progression restoration) genes and 11 yeast OPY (overproduction-induced pheromone-resistant yeast) genes that specifically block the G1 arrest by mating pheromone. The CPR genes represent a variety of biochemical functions including a new cyclin, a tumor suppressor binding protein, chaperones, transcription factors, translation factors, RNA-binding proteins, as well as novel proteins. Several CPR genes require individual CLNs to promote pheromone resistance and those that require CLN3 increase the basal levels of Cln3 protein. Moreover, several of the yeast OPY genes have overlapping functions with the human CPR genes, indicating a possible conservation of roles. Show less

Wild-type cells of the budding yeast Saccharbmyces cerevisiae arrest in G1 upon nutrient exhaustion. Cell cycle arrest requires the WHI2 gene since whi2 mutants continue to divide and become abnormally small as nutrients are depleted. Here we show that CLN1 and CLN2 transcript levels in a whi2 strain are higher during exponential growth, and persist longer upon starvation, than in an isogenic wild-type strain. In contrast to CLN1 and CLN2, CLN3 levels declined only at very high cell density and were unaffected by the whi2 mutation. Elevated CLN expression is sufficient to explain the whi2 phenotype since ectopic expression of CLN1 in a nutrient-depleted culture caused cells to continue dividing and interfered with the acquisition of heat resistance. These observations show that, either directly or indirectly, Whi2 negatively regulates G1 cyclin expression. Interestingly extremely high levels of Cln1 induced filamentous growth upon nutrient deprivation, suggesting a direct connection between G1 cyclin activity and morphological responses to poor nutrient conditions. Show less

In budding yeast, cell division is initiated in late G1 phase once the Cdc28 cyclin-dependent kinase is activated by the G1 cyclins Cln1, Cln2, and Cln3. The extreme instability of the Cln proteins couples environmental signals, which regulate Cln synthesis, to cell division. We isolated Cdc53 as a Cln2-associated protein and show that Cdc53 is required for Cln2 instability and ubiquitination in vivo. The Cln2-Cdc53 interaction, Cln2 ubiquitination, and Cln2 instability all depend on phosphorylation of Cln2. Cdc53 also binds the E2 ubiquitin-conjugating enzyme, Cdc34. These findings suggest that Cdc53 is a component of a ubiquitin-protein ligase complex that targets phosphorylated G1 cyclins for degradation by the ubiquitin-proteasome pathway. Show less

In yeast, commitment to cell division (Start) is catalyzed by activation of the Cdc28 protein kinase in late G1 phase by the Cln1, Cln2, and Cln3 G1 cyclins. The Clns are essential, rate-limiting activators of Start because cells lacking Cln function (referred to as cln-) arrest at Start and because CLN dosage modulates the timing of Start. At or shortly after Start, the development of B-type cyclin Clb-Cdc28 kinase activity and initiation of DNA replication requires the destruction of p40SIC1, a specific inhibitor of the Clb-Cdc28 kinases. I report here that cln cells are rendered viable by deletion of SIC1. Conversely, in cln1 cln2 cells, which have low CLN activity, modest increases in SIC1 gene dosage cause inviability. Deletion of SIC1 does not cause a general bypass of Start since (cln-)sic1 cells remain sensitive to mating pheromone-induced arrest. Far1, a pheromone-activated inhibitor of Cln-Cdc28 kinases, is dispensable for arrest of (cln-)sic1 cells by pheromone, implying the existence of an alternate Far1-independent arrest pathway. These observations define a pheromone-sensitive activity able to catalyze Start only in the absence of p40SIC1. The existence of this activity means that the B-type cyclin inhibitor p40SIC1 imposes the requirement for Cln function at Start. Show less

In the yeast Saccharomyces cerevisiae, commitment to cell division (Start) requires growth to a critical cell size. The G1 cyclins Cln1, Cln2 and Cln3 activate the Cdc28 protein kinase and are rate-limiting activators of Start. When glucose is added to cells growing in a poor carbon source, the critical cell size required for Start is reset from a small to a large size. In yeast, glucose acts through Ras proteins to stimulate adenylyl cyclase, activating the three cyclic AMP-dependent protein kinases Tpk1, Tpk2 and Tpk3 (refs 8, 9). We find that stimulation of the Ras/cAMP pathway represses expression of CLN1, CLN2 and co-regulated genes, inhibiting Start. This helps explain the increase in critical size when cells are shifted from poor to rich medium. This connection between the molecules controlling growth (Ras/cAMP) and those controlling division (cyclins) helps explain how division is co-ordinated with growth. Show less

In the yeast Saccharomyces cerevisiae, the Cdc28 protein kinase controls commitment to cell division at Start, but no biologically relevant G1-phase substrates have been identified. We have studied the kinase complexes formed between Cdc28 and each of the G1 cyclins Cln1, Cln2, and Cln3. Each complex has a specific array of coprecipitated in vitro substrates. We identify one of these as Far1, a protein required for pheromone-induced arrest at Start. Treatment with alpha-factor induces a preferential association and/or phosphorylation of Far1 by the Cln1, Cln2, and Cln3 kinase complexes. This induced interaction depends upon the Fus3 protein kinase, a mitogen-activated protein kinase homolog that functions near the bottom of the alpha-factor signal transduction pathway. Thus, we trace a path through which a mitogen-activated protein kinase regulates a Cdc2 kinase. Show less

In the budding yeast Saccharomyces cerevisiae, the G1 cyclins Cln1, Cln2 and Cln3 regulate entry into the cell cycle (Start) by activating the Cdc28 protein kinase. We find that Cln3 is a much rarer protein than Cln1 or Cln2 and has a much weaker associated histone H1 kinase activity. Unlike Cln1 and Cln2, Cln3 is not significantly cell cycle regulated, nor is it down-regulated by mating pheromone-induced G1 arrest. An artificial burst of CLN3 expression early in G1 phase accelerates Start and rapidly induces at least five other cyclin genes (CLN1, CLN2, HCS26, ORFD and CLB5) and the cell cycle-specific transcription factor SWI4. In similar experiments, CLN1 is less efficient than CLN3 at activating Start. Strikingly, expression of HCS26, ORFD and CLB5 is dependent on CLN3 in a cln1 cln2 strain, possibly explaining why CLN3 is essential in the absence of CLN1 and CLN2. To explain the potent ability of Cln3 to activate Start, despite its apparently weak biochemical activity, we propose that Cln3 may be an upstream activator of the G1 cyclins which directly catalyze Start. Given the large number of known cyclins, such cyclin cascades may be a common theme in cell cycle control. Show less

In Saccharomyces cerevisiae, several of the proteins involved in the Start decision have been identified; these include the Cdc28 protein kinase and three cyclin-like proteins, Cln1, Cln2 and Cln3. We find that Cln3 is a very unstable, low abundance protein. In contrast, the truncated Cln3-1 protein is stable, suggesting that the PEST-rich C-terminal third of Cln3 is necessary for rapid turnover. Cln3 associates with Cdc28 to form an active kinase complex that phosphorylates Cln3 itself and a co-precipitated substrate of 45 kDa. The cdc34-2 allele, which encodes a defective ubiquitin conjugating enzyme, dramatically increases the kinase activity associated with Cln3, but does not affect the half-life of Cln3. The Cln--Cdc28 complex is inactivated by treatment with non-specific phosphatases; prolonged incubation with ATP restores kinase activity to the dephosphorylated kinase complex. It is thus possible that phosphate residues essential for Cln-Cdc28 kinase activity are added autocatalytically. The multiple post-translational controls on Cln3 activity may help Cln3 tether division to growth. Show less