INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Eukaryotic cells regulate proliferative growth in response to a variety of external signals. This is not only important to the survival of individual cells but also a critical part of the coordinated processes involved in the growth and differentiation of multicellular organisms. One approach to understanding the regulation of proliferation in eukaryotic cells has been to use the budding yeast Saccharomyces cerevisiae as a model system. This strategy is based on the observation that many processes found in yeast are conserved in larger, more complex eukaryotes and takes advantage of the genetic techniques available with yeast. Proliferation in S. cerevisiae is regulated in response to two known types of environmental signals, mating pheromones and nutrients (21). The details of the mating pheromone pathway are being rapidly pursued, but much less is known about how nutrients regulate progress through the cell cycle.

Although yeast cells grow in mass at widely varying rates in different media, they are able to maintain almost a constant cell size. This means that the rate of progress through the cell cycle must be adjusted to match changes in the rate of growth in mass. Much of this regulation of the cell cycle occurs by controlling the length of the G₁ phase of the cell cycle (14). Specifically, yeast cells modulate passage through a point in late G₁ referred to as Start in response to changes in their growth media. In this way, cells growing slowly on poor medium are delayed in passing from G₁ into S, allowing the cells more time to grow to the proper size before budding. Cells growing rapidly on rich medium spend only a short time in G₁.

Progress through Start is dependent on the kinase activity of the protein encoded by CDC28 (22). The Cdc28 kinase is itself regulated by a set of proteins called cyclins (19). It is therefore reasonable to expect some sort of connection between nutrient signals and the cyclin-Cdc28 kinase system. Of the G₁ cyclin genes, CLN3 appears to play a distinct role in controlling passage through Start, and it has been proposed to be an initiator of a cascade of G₁ cyclin expression (25, 26). Manipulation of CLN3 expression has been shown to alter G₁ length: loss of CLN3 delays Start, producing larger cells. Increasing CLN3 expression, either by increasing transcript levels or by mutations that produce a hyperstable Cln3 protein, shortens G₁ (5, 27). Recent work has shown that Cln3 protein levels are regulated by the growth medium, with high levels of Cln3 protein found in log-phase cells growing in rich medium and low levels found in cells growing slowly in the oxidative phase of growth following the diauxic shift (10). The finding that levels of Cln3-Cdc28 activity correlate with G₁ length in media of varying quality suggests a model in which Cln3 protein activity serves to regulate G₁ length to adjust for changes in growth rate in different media.

The decrease in Cln3 protein levels in poor medium is in part due to postranscriptional effects, involving translational regulation as well as the Ras-cyclic AMP (cAMP) pathway (9, 20). The decrease in Cln3 protein levels also clearly involves a substantial decrease in CLN3 mRNA levels (10, 13). This is consistent with a system in which multiple inputs can regulate Cln3.

In this report, we describe a set of experiments intended to identify what types of nutrient signals regulate CLN3 transcription. We find that both fermentable carbon sources and nitrogen are required for maximum CLN3 mRNA expression. In contrast, limitation for sulfur, phosphorus, or cAMP, while arresting growth, does not decrease CLN3 transcript levels. We find that regulation of CLN3 transcription by glucose is sensitive to very low levels of glucose and is distinct from the processes that regulate glucose repression.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Yeast strains and media. The yeast strains used in this study were TW38 (isogenic to HR125 except Ura⁺; MATa leu2-3 leu2-112 trp1-1 his3-532 his4), TC41-1 (12) (isogenic to HR125 except cyr1; MATa leu2-3 leu2-112 ura3-52 trp1-1 his3-532 his4 cyr::URA3 cam), and DS10 (1) (wild type; MATa his3-11,15 leu2-3 leu2-112 lys1 lys2 ura3-52 trp1). Cells were grown in liquid culture with vigorous shaking at 30°C in rich medium (yeast extract-peptone [YEP]) containing either glucose (YEPD; 1% yeast extract, 2% Bacto Peptone, 2% glucose) or ethanol (YEPEtOH; same as YEPD but with 2% ethanol in place of glucose) or in synthetic complete medium (SC) containing yeast nitrogen base (6.7 g/liter; Difco) supplemented with adenine, uracil, amino acids, and 2% glucose (24) except as noted. When necessary, cells were transferred to different media by spinning in a Beckman J6 centrifuge for 5 min at 3,000 rpm. For experiments involving nitrogen limitation, prototrophic yeast cells were grown in synthetic medium prepared with yeast nitrogen base without ammonium sulfate (Difco) and with no supplemental amino acids. For experiments involving sulfate limitation, the synthetic medium was made with chloride salts substituted for sulfate and methionine was omitted from the amino acid supplements. For low-phosphate medium, potassium chloride was substituted for potassium phosphate. Log-phase cells were growing rapidly at an optical density at 660 nm (OD₆₆₀) of 1 or less.

Preparation of RNA and RNA blots. RNA was prepared as previously described (5a). The RNA samples (15 µg/lane) were separated by formaldehyde agarose gel electrophoresis and transferred to a GeneScreen Plus membrane as instructed by the manufacturer (New England Nuclear). To ensure uniform loading and transfer of RNA, ethidium bromide was added to the samples prior to loading of the gel. Following blotting, ethidium-stained rRNA was visualized on the blots by UV illumination and photographed. The blots were probed with a 1-kb SacI/EcoRI fragment from CLN1, a 1-kb SacI/XhoI fragment from CLN2, a 1-kb EcoRI fragment from CLN3, a 1.4-kb BstXI/HindIII fragment from UBI4, a 0.5-kb EcoRI/PstI fragment from SSA3, or a 1.7-kb fragment from GAC1 as indicated. A 0.6-kb SacI fragment was used to probe for U2 RNA as a loading and transfer control. Probes were radiolabeled with ³²P by the random primer method to a specific activity of 10⁹ cpm/mg. Probes were used in quantities in excess over the RNA being measured.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Carbon source regulation of CLN3 mRNA levels. Cells growing in log phase on glucose have substantially more CLN3 message than cells growing oxidatively after the diauxic shift (10). To compare CLN3 mRNA levels in log phase, postdiauxic growth, and stationary phase, we collected samples from a YEPD culture for RNA preparation at day 1 (log), days 2 and 3 (OD of 4 to 6, postdiauxic), and day 4 (OD of 10, stationary phase). The major change in CLN3 message levels occurred between day 1 and day 2, when CLN3 mRNA levels fell approximately sevenfold as cells moved from fermentative to oxidative growth. A small additional decrease of approximately twofold occurred between the postdiauxic phase and stationary phase (Fig. 1).

View larger version (33K): [in this window] [in a new window]   FIG. 1.   CLN3 expression decreases as cells end fermentative growth on glucose. (A) Wild-type cells (TW38) were grown in YEPD with shaking at 30°C, and samples were collected for RNA preparation and blotting at log phase (day 1, OD of 0.7), day 2 (OD of 4, postdiauxic), day 3 (OD of 6, postdiauxic), and day 4 (OD of 10, stationary phase). RNA blots were probed with a CLN3 probe and analyzed with a phosphorimager as described in Materials and Methods. (B) CLN3 mRNA levels normalized to the U2 loading controls and plotted graphically. The scale is arbitrary.

View larger version (42K): [in this window] [in a new window]   FIG. 2.   CLN3 mRNA levels are decreased in cells growing on nonfermentable carbon sources. Wild-type cells (DS10) were grown in YEPD to mid-log phase and transferred to YEP containing either 2% glucose, 2% galactose, 2% glycerol and 2% lactate, or 2% ethanol at an OD of 0.1. Samples were collected at the indicated times for density measurements (B) and, as they passed OD = 0.5, for RNA preparation and Northern blotting with a CLN3 probe (A). (C) CLN3 mRNA levels normalized to the U2 loading controls and plotted graphically. The scale is arbitrary.

View larger version (42K): [in this window] [in a new window]   FIG. 3.   Glucose analogs fail to stimulate CLN3 expression. (A) Wild-type cells (TW38) were grown in YEPD until they had exhausted glucose from the medium (36 h after inoculation, OD of 6), and then either D-glucose (G) or a nonmetabolizable glucose analog (2-deoxyglucose [2-deoxy]) was added to produce a final concentration of 2%. Control cells received an equivalent volume of water. Cells were incubated for the indicated time (minutes) and then collected for RNA blotting with a CLN3 probe as described in Materials and Methods. All lanes are from the same exposure of a single blot. Some irrelevant intervening lanes have been removed. (B) Cells were treated as described above with either D-glucose, L-glucose, or 6-deoxyglucose and then incubated with the added sugar for 30 min.

View larger version (75K): [in this window] [in a new window]   FIG. 4.   Glucose regulation of CLN3 expression is independent of cAMP. Cells carrying a CYR1 deletion (TC41-1) were grown to mid-log phase (OD of 1), and cAMP was then removed by washing the cells in SC (2% glucose) with no cAMP and resuspending them in this cAMP-free medium at an OD of 1. After 1 h, a sample was removed and the remaining cells were transferred to glucose-free SC with no cAMP at an OD of 1. After incubating without glucose for 4 h, another sample was collected and the remaining cells were again transferred into fresh SC (2% glucose) with no cAMP at an OD of 1 and incubated for an additional 30 min. The samples were used for RNA preparation and RNA blotting with a CLN3 probe. Lane A, RNA from cells after the 1-h incubation in fresh SC (2% glucose) without cAMP; lane B, RNA from cells after 4 h of starvation for glucose; lane C, RNA from glucose-starved cells after 30 min of reincubation in SC (2% glucose) without cAMP.

Glucose repression and CLN3. We wanted to determine whether the signals produced by glucose that inhibit transcription of glucose-repressible genes might be in some way also involved in the induction of transcription by glucose. We found that glucose could induce CLN3 expression at levels that were lower than those required for glucose repression. When cells were transferred to YEP medium (containing no glucose), CLN3 mRNA levels fell very rapidly. At the same time, transfer to YEP caused the induction of a number of stress response genes that are repressed by glucose (Fig. 5). In contrast, transfer to YEP with 0.1% glucose, a level of glucose that allows derepression of many glucose-repressed transcripts, did not decrease CLN3 transcription even though the move produced an immediate derepression of UBI4 and GAC1. CLN3 transcripts remained at high levels for over an hour until glucose was depleted from the medium. We also examined cells carrying null mutations in genes that play a role in glucose repression, HXK2, MIG1, SNF1, and SNF4 (15), to determine whether genes involved in transmitting signals needed for repression of transcription might also be involved in glucose induction of transcription. None of these glucose mutations appeared to interfere with CLN3 message levels (not shown). These results indicate that the processes mediated by these genes are not necessary for glucose induction of CLN3 transcription. However, the signals generated by glucose that lead to repression of transcription are poorly understood. We therefore cannot rule out the possibility that common upstream signaling components, or events, produce both glucose induction of CLN3 and glucose repression of other genes.

View larger version (71K): [in this window] [in a new window]   FIG. 5.   CLN3 mRNA levels remain high at nonrepressing levels of glucose. Wild-type cells (TW38) were grown in SC medium (2% glucose) to mid-log phase (OD660 of 1), centrifuged, and resuspended at the same density in SC containing either no glucose, 0.1% glucose, or 2% glucose as a control. The cells were incubated for the indicated times (minutes) at 30°C before being harvested for RNA preparation and Northern blotting with the indicated probes.

Effects of nitrogen, sulfur, and phosphorus limitation on CLN3 expression. In addition to regulating the cell cycle in response to glucose depletion, yeast cells also arrest growth in G₁ in response to limitation for nitrogen, sulfur, or phosphorus. It therefore seemed possible that these nutrients also affect CLN3 mRNA levels. To test this, we passaged wild-type yeast cells through medium lacking either nitrogen, sulfur, or phosphorus until the cells stopped growing. We confirmed that the cells were indeed limited for the nutrient in question by demonstrating that the cells would resume growth in response to addition of that nutrient. We then determined whether CLN3 expression changes in these cells as they return to proliferative growth in response to fresh nutrients.

View larger version (31K): [in this window] [in a new window]   FIG. 6.   CLN3 message levels are decreased by nitrogen starvation. Prototrophic wild-type yeast cells (S288C) were passaged through synthetic medium without nitrogen until nitrogen availability limited growth. These cells were then transferred at a final OD660 of 0.2 to synthetic medium without nitrogen and incubated overnight with shaking at 30°C. At time zero, the nitrogen-starved cells were divided and received either water as a control (filled squares in panel A; lanes N in panel B), the amino acid supplements found in SC medium (filled triangles and lanes +aa), or 5 mg of ammonium chloride per ml (open squares and lanes +N). Samples were collected for OD readings (A) and RNA preparation (B) at the indicated times. RNA blots were probed for CLN1, CLN2, or CLN3 as described in Materials and Methods.

View larger version (36K): [in this window] [in a new window]   FIG. 7.   CLN3 message levels are not decreased by starvation for sulfur. Wild-type cells (TW38) were passaged through sulfate-deficient medium until growth became limited by lack of sulfur. The cells were then transferred to sulfate-deficient medium to a final OD660 of 0.2 and grown overnight with shaking at 30°C (closed squares in panel A; lanes S in panel B). For comparison and as a control to demonstrate sulfur limitation, cells were also transferred to synthetic medium with sulfate added. These cells grew overnight until they had depleted the glucose from the medium (open circles and lane G). At time zero, sodium sulfate was added to a portion of the sulfate-limited culture to a final concentration of 1 mg/ml (open squares and lanes +S), and samples were collected for RNA blotting with a CLN1, CLN2, or CLN3 probe (B) or to monitor culture density at the indicated times (A).

View larger version (34K): [in this window] [in a new window]   FIG. 8.   CLN3 message levels are not decreased by starvation for phosphorus. Wild-type cells (TW38) were passaged through phosphate-deficient medium until growth became limited by lack of phosphorus. The cells were then transferred to phosphate-deficient medium at a final OD660 of 0.2 and grown overnight with shaking at 30°C (closed squares in panel A; lane P in panel B). For comparison and as a control to demonstrate phosphorus limitation, cells were also transferred to synthetic medium with phosphate added. These cells grew overnight until they had depleted the glucose from the medium (open circles and lane G). At time zero, potassium phosphate was added to a portion of the phosphate-limited culture to a final concentration of 1 mg/ml (open squares and lane +P), and samples were collected for RNA blotting with a CLN1, CLN2, or CLN3 probe (B) or to monitor culture density at the indicated times (A).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Nutrients and the cell cycle. The need to coordinate the rate of cellular growth in mass and volume with the rate of progress through the cell division cycle has been recognized for many years (11). Because nutrient availability places major constraints on the rate of cellular growth, nutrients must in some way also regulate progression through the cell cycle. The fact that yeast cells grow at vastly different rates on different media with only modest changes in cell size tells us that such coordination indeed takes place. Most of the normal regulation of the cell cycle in response to differing growth rates occurs at the G₁-to-S boundary. Measurements of cell cycle progression show that the G₁ phase of the cell cycle lengthens as cells grow more slowly in poor medium and contracts as cells grow more rapidly in rich medium (14). It should be noted that nutrient regulation of G₁, at least in mother cells, is independent of cell size. Nutrients affect the cell cycle length of mother cells almost as much as daughter cells (10, 16), and both for mother cells and for daughter cells that are above the normal size for bud emergence, size is not correlated with G₁ length. In these populations of cells, larger cells are no more likely than smaller cells to pass Start (16, 17, 30).

Role of CLN3. The fact that nutrients can regulate the length of G₁ suggests that nutrients in some way influence the activity of the cyclin-dependent Cdc28 kinase that is thought to regulate Start. Recent reports have shown that Cln3 protein levels are increased in cells growing in rich medium (10), a process that involves both transcriptional and posttranscriptional regulation (9, 10, 20). These results, taken together with those of past experiments showing that manipulation of CLN3 expression can affect the timing of Start, support a model in which nutrients that allow rapid growth in size also increase CLN3 expression to promote rapid progress through the cell cycle.

Regulation of CLN3 and the cell cycle as part of a global regulatory process. As cells deplete glucose from the medium and enter the postdiauxic phase of growth, both mRNA synthesis and protein synthesis dramatically fall (2, 3, 8). This has been termed a global change in transcription, but that term must be carefully qualified. Although transcription of most messages decreases, transcription of many other messages, most notably those that are repressed by glucose, actually increases (15, 29). In addition, the levels of some messages seem to remain constant. This finding suggests that the activity of RNA polymerase II is not simply shut off but rather is turned down for one class of messages and redirected toward others. The finding that so many messages are affected by this process lends itself to the idea that it is not specific, but again caution is in order. Glucose affects so many aspects of yeast metabolism and growth that one might well expect that a proper and specific response to changes in carbon source would indeed involve regulation on a very wide scale. Much of the observed decrease in transcription in response to glucose depletion is consistent with keeping metabolic demand in line with the availability of nutrients and energy. Although it is plausible that all or most of this global change in transcription serves such a purpose, we simply do not know how many of these changes have functional significance. Slowing down cell division is one obvious response to decreased nutrient availability.

Evidence for other mechanisms coupling nutrients and the cell cycle. While our results point to a role for the transcriptional regulation of CLN3 in controlling G₁ length in response to changing nutrients, several other results indicate that other mechanisms also play important roles in this process. Glucose and other fermentable sugars have long been known to regulate intracellular cAMP (6), which in turn has been shown to be necessary for progress through Start (18). We have reported that cAMP plays a positive role in transcription of Start-specific genes, most notably CLN1 and CLN2 (13), and more recently we have found that cAMP positively regulates Cln3 protein levels at the posttranscriptional level (9). Polymenis and Schmidt have recently shown that rich medium increases the translation of Cln3 (20). Beyond that, there is evidence for transcriptional regulation of CDC28 and BCK2, genes that are also involved in regulating the timing of Start (10). While not lending themselves to a simple model, these results illustrate the potential for multiple levels of input for linking nutrient signals with regulation of cell cycle progression.