Abstract
Cell size is proportional to growth rate. Thus, cells growing slowly in poor nutrients can be nearly half the size of cells growing rapidly in rich nutrients. The relationship between cell size and growth rate appears to hold across all orders of life, yet the underlying mechanisms are unknown. In budding yeast, most growth occurs during mitosis, and the proportional relationship between cell size and growth rate is therefore enforced primarily by modulating growth in mitosis. When growth is slow, the duration of mitosis is increased to allow more time for growth, yet the amount of growth required to complete mitosis is reduced, leading to birth of small daughter cells. Previous studies found that PP2A associated with the Rts1 regulatory subunit (PP2ARts1) works in a TORC2-dependent feedback loop that sets cell size and growth rate to match nutrient availability. However, it was unknown whether PP2ARts1 influences growth in mitosis. Here, we show that PP2ARts1 is required for the proportional relationship between cell size and growth rate during mitosis. Moreover, nutrients and PP2ARts1 influence the duration of mitosis, and thus the extent of growth in mitosis, via Wee1 and Pds1/securin, two conserved regulators of mitotic progression. Together, the data suggest a model in which the same global signals that set growth rate also set the critical amount of growth required for cell cycle progression, which would provide a simple mechanistic explanation for the proportional relationship between size and growth rate.
Introduction
A critical step in the evolution of life was attainment of the capacity for growth. Along with growth, the earliest cells must have evolved mechanisms for controlling growth if they were to survive and compete. Thus, early cells needed mechanisms to control the rate and extent of such processes as membrane growth and ribosome biogenesis, while also ensuring that the rates of each process are matched to each other and to the availability of building blocks derived from nutrients. Mechanisms that control growth ultimately determine the size and shape of a cell, and are responsible for the myriad sizes, shapes, and growth rates observed in cells spanning all orders of life.
Growth of budding yeast cells illustrates how mechanisms that control cell growth define the size and shape of a cell. Growth occurs in distinct phases during the cell cycle that are characterized by different rates and patterns of growth (Johnston et al., 1977; McCusker et al., 2007; Goranov et al., 2009; Ferrezuelo et al., 2012; Leitao and Kellogg, 2017). During G1 phase, growth occurs at a slow rate and occurs over the entire surface of the cell. At the end of G1 phase, growth of the mother cell ceases and growth becomes polarized as a daughter bud emerges. Entry into mitosis triggers a 2- to 3-fold increase in the rate of growth, as well as a switch to isotropic growth that occurs over the entire bud surface. Rapid growth continues throughout mitosis and accounts for most of the volume of a yeast cell (Leitao and Kellogg, 2017). Thus, a cell growing in rich nutrient conditions achieves greater than 60% of its volume during growth in mitosis, and only 20% of its volume during G1 phase. The volume achieved during G1 phase increases to approximately 40% under poor nutrient conditions. The distinct size and shape of a budding yeast cell is ultimately defined by mechanisms that control the location and extent of growth during each of these growth phases.
The size of budding yeast cells is further influenced by growth rate. For example, yeast cells growing slowly in poor nutrients are nearly half the size of cells in rich nutrients (Johnston et al., 1977; 1979). This observation illustrates a peculiar and poorly understood aspect of growth control: cell size is proportional to growth rate (Ferrezuelo et al., 2012; Leitao and Kellogg, 2017). The relationship holds when comparing cells growing under different nutrient conditions that support different growth rates, and when comparing cells that show different growth rates despite growing under identical nutrient conditions. Conversely, growth rate is proportional to cell size in yeast (Elliott and McLaughlin, 1978; Schmoller et al., 2015; Leitao and Kellogg, 2017). For example, the growth rate of the daughter bud is proportional to the size of the mother cell. A proportional relationship between cell size and growth rate appears to hold across all orders of life (Schaechter et al., 1958; Robertson, 1963; Hirsch and Han, 1969; Fantes and Nurse, 1977; Johnston et al., 1977).
Clues to a mechanistic basis for the relationship between cell size and growth rate in budding yeast have come from analysis of a form of protein phosphatase 2A that is bound to the Rts1 regulatory subunit (PP2ARts1). PP2ARts1 is required for normal control of cell growth and size (Artiles et al., 2009; Zapata et al., 2014). Thus, cells that lack PP2ARts1 are abnormally large and fail to reduce their size in poor nutrients. PP2ARts1 relays signals that influence a TORC2 signaling network that is required for normal control of cell size and growth rate (Lucena et al., 2017). The TORC2 network controls synthesis of ceramide lipids, which play roles in signaling. Genetic and pharmacological data suggest that modulation of ceramide synthesis by PP2ARts1 and TORC2 is required for nutrient modulation of cell size and growth rate in G1 phase. Cells that can not synthesize ceramides fail to increase their growth rate during G1 phase when shifted from poor to rich carbon, and they complete G1 phase at identical small sizes in rich or poor carbon. It is unknown whether signals from the PP2ARts1 network also control cell growth and size during the mitotic growth interval.
Here, we have investigated control of cell growth and size during growth in mitosis. Cell size at completion of mitosis is proportional to growth rate during mitosis (Leitao and Kellogg, 2017). Moreover, when growth rate is slowed by poor nutrients, the duration of mitosis is increased, which likely reflects mechanisms that ensure that completion of mitotic events occurs only when a sufficient amount of growth has occurred (Leitao and Kellogg, 2017). Together, these observations suggest that growth during mitosis is regulated. A proteome-wide mass spectrometry search for proteins controlled by PP2ARts1 identified numerous proteins implicated in control of mitotic progression and cell size, which suggested that PP2ARts1 could play an important role in control of cell growth and size in mitosis (Zapata et al., 2014). Thus, as a first step we tested whether PP2ARts1 influences cell growth and size in mitosis.
Results and Discussion
PP2ARts1 is required for the proportional relationship between cell size and growth rate
We first tested how inactivation of PP2ARts1 influences mitotic duration and daughter cell size. Previous studies have shown that mother cell size strongly influences daughter cell size and growth rate (Johnston et al., 1977; Schmoller et al., 2015; Leitao and Kellogg, 2017). Therefore, interpretation of results from rts1Δ cells would be complicated by the possibility that effects on bud growth and size are a secondary consequence of defects in mother cell size that arose in previous generations. To avoid this problem, we fused RTS1 to an auxin-inducible degron (rts1-AID), which allowed us to observe the immediate effects of inactivating PP2ARts1 (Nishimura et al., 2009). In the absence of auxin, rts1-AID did not cause obvious defects in cell size or cell cycle progression (Figures S1A,B). Addition of auxin caused rapid destruction of rts1-AID protein within 15–30 minutes (Figure S1C). Approximately 10% of the rts1-AID protein remained in the presence of auxin. In addition, rts1-AID cells growing in the presence of auxin at elevated temperatures formed colonies more rapidly than rts1Δ cells (Figure S1D). Together, these observations suggest that rts1-AID causes a partial loss of function when auxin is present. Nevertheless, we utilized rts1-AID because it allowed analysis of bud growth and mitotic duration without the complication of aberrant mother cell size.
We used a microscopy-based assay (Ferrezuelo et al., 2012; Leitao and Kellogg, 2017) to test how loss of rts1-AID affected growth during mitosis. Auxin was added to synchronized cells shortly after bud emergence. Destruction of rts1-AID caused a 3-fold increase in the average duration of metaphase and anaphase in both rich and poor carbon (Figure 1A; see Figure S2A for dot plots and p-values). Destruction of rts1-AID also caused a large increase in the variance of metaphase duration compared to wild type cells (Figure S2A).
We next analyzed daughter bud size at each mitotic transition. rts1-AID caused a large increase in daughter bud size at completion of each stage of mitosis in both rich and poor carbon (Figure 1B; see Figures S2C,D for dot plots and p-values). The variance in size at the end of metaphase was much larger in rts1-AID cells compared to wild type cells. Importantly, there was not a statistically significant difference in the size of rts1-AID daughter cells in rich or poor carbon at the end of metaphase (Figure S2C). By the end of anaphase, rts1-AID cells in rich carbon were slightly larger than their counterparts in poor carbon (Figure S2D).
Further analysis of the rts1-AID allele led to surprising insight into the role of PP2ARts1 in cell size control. Destruction of rts1-AID reduced the average rate of growth in mitosis by approximately 30% in both rich and poor carbon (Figure 1C). In normal cells, reduced growth rate causes reduced cell size; however, destruction of rts1-AID caused an increase in size. In addition, the correlation between growth rate and cell size was broken in rts1-AID cells (Figure 1D). Thus, buds with nearly identical growth rates completed mitosis at very different sizes. Conversely, cells that completed mitosis at identical sizes grew at very different rates during mitosis. A previous study found that inactivation of Ydj1, a member of the JJJ chaperone family, causes a similar discordance between cell size and growth rate at the end of G1 phase (Ferrezuelo et al., 2012). Proteome-wide mass spectrometry analysis of rts1Δ cells suggests that Ydj1 is controlled by PP2ARts1 (Zapata et al., 2014).
To summarize, a partial loss of PP2ARts1 function caused major defects in cell size control in mitosis, as well as a nearly complete loss of the proportional relationship between cell size and growth rate. These observations are consistent with our previous discovery that rts1Δ causes a nearly complete failure in modulation of cell size in response to changes in carbon source (Artiles et al., 2009).
As a next step, we sought to identify proteins that respond to PP2ARts1-dependent signals and control the duration of growth in mitosis. Over the long term, identification of these proteins and the signals that control their activity should yield insight into mechanisms that control cell growth and size.
Swe1 and Pds1 are required for normal control of cell size at completion of mitosis
We used a candidate approach to identify proteins that control the duration of growth in mitosis. To identify candidates, we used two criteria. First, we considered proteins that were previously found to control cell size and/or the duration of mitosis. Second, we considered potential targets of PP2ARts1-dependent regulation that were previously identified by proteome-wide mass spectrometry analysis of rts1Δ cells, or by phenotypic analysis of rts1Δ cells (Harvey et al., 2011; Zapata et al., 2014). Two candidates fulfilled both criteria: Swe1 and Pds1/securin.
Swe1 is the budding yeast homolog of the Wee1 kinase, which delays mitosis by phosphorylating and inhibiting mitotic Cdk1. Wee1 was originally found to control mitotic entry. However, more recent studies found that Wee1 family members also control the duration of mitosis in both yeast and human cells (Jin et al., 2008; Raspelli et al., 2011; Lianga et al., 2013; Toledo et al., 2015). Moreover, mass spectrometry identified the inhibitory site on Cdk1 targeted by Swe1 as the most strongly hyperphosphorylated site in rts1Δ cells (Zapata et al., 2014). The mass spectrometry also showed that Swe1 is hyperphosphorylated in rts1Δ cells on multiple sites that play a role in its activation, and analysis of Swe1 phosphorylation in rts1Δ cells by western blotting has shown that Swe1 persists in a partially hyperphosphorylated form that is thought to be the active form (Harvey et al., 2011). Finally, there is evidence that PP2ARts1 can also control removal of Cdk1 inhibitory phosphorylation independently of its role in controlling Swe1 phosphorylation (Kennedy et al., 2016). Together, these observations demonstrate that inhibitory phosphorylation of Cdk1 could play a role in prolonging mitosis when growth is slowed in poor carbon.
The second candidate, Pds1/securin, inhibits chromosome segregation by binding and inhibiting separase, a protease that cleaves cohesins that hold chromosomes together (Cohen-Fix and Koshland, 1997; Ciosk et al., 1998; Uhlmann et al., 1999). Exit from mitosis is triggered by activation of the anaphase-promoting complex (APC), which targets Pds1 for destruction. Pds1 and separase also control mitotic cyclin destruction, which indicates that they can control the duration of mitosis (Cohen-Fix and Koshland, 1999; Tinker-Kulberg and Morgan, 1999). In addition, DNA damage induces a mitotic arrest by triggering signals that lead to phosphorylation of Pds1, thereby protecting it from the APC (Yamamoto et al., 1996; Wang et al., 2001). Thus, there is a precedent for signals upstream of Pds1 controlling the duration of mitosis. Finally, proteome-wide mass spectrometry data suggest that Pds1 is hyperphosphorylated in rts1Δ cells (Zapata et al., 2014).
If Pds1 and Swe1 play roles in prolonging mitosis to allow growth to occur, then their loss should cause reduced cell size. Previous studies have shown that swe1Δ reduces cell size, but there has been no evidence that Pds1 is required for cell size control (Jorgensen et al., 2002; Harvey and Kellogg, 2003; Harvey et al., 2005; 2011). Coulter counter analysis of cell size revealed that pds1Δ caused a reduction in cell size similar to the reduction in cell size caused by swe1Δ (Figure 2A). When combined, swe1Δ and pds1Δ caused an additive decrease in cell size.
We next hypothesized that the prolonged duration of mitosis and increased cell size caused by loss of PP2ARts1 is due to hyperactivity of Swe1 and Pds1. To begin to test this, we first analyzed the effects of swe1Δ and pds1Δ on the size of rts1Δ cells. The abnormally large size of rts1Δ cells was eliminated by swe1Δ pds1Δ, which suggested that the increased duration of growth in mitosis in rts1-AID cells could be due entirely to hyperactive Pds1 and Swe1 (Figure 2B). pds1Δ alone did not reduce the size of rts1Δ cells. We suspect that this is the result of hyperactive Swe1 in rts1Δ cells, which would lead to decreased Cdk1 activity. Since Cdk1 is required for activation of the anaphase promoting complex, this could lead to a mitotic delay.
Inhibitory phosphorylation of Cdk1 is partially responsible for the increased duration of mitosis in poor carbon
We next investigated the role of Cdk1 inhibitory phosphorylation more closely. To test whether Swe1 plays a role in prolonging mitosis in poor carbon, we first tested whether Cdk1 inhibitory phosphorylation is prolonged in poor carbon. Cells growing in rich or poor carbon were released from a G1 arrest and Cdk1 inhibitory phosphorylation was analyzed with a phosphospecific antibody. Samples were also analyzed by immunofluorescence to determine the fraction of cells with metaphase spindles at each time point. Cdk1 inhibitory phosphorylation was prolonged in poor carbon and was correlated with the presence of metaphase spindles (Figure 3A).
We also analyzed Swe1 phosphorylation during the cell cycle in rich and poor carbon. Swe1 passes through multiple phosphorylation states during mitosis that can be detected via electrophoretic mobility shifts; attainment of a fully hyperphosphorylated state is correlated with inactivation of Swe1 (Sreenivasan and Kellogg, 1999; Harvey et al., 2005; 2011). In rich media, Swe1 reached full hyperphosphorylation and was degraded shortly thereafter (Figure 3A). In poor media, Swe1 was present throughout much of the prolonged mitosis. Moreover, Swe1 took longer to reach the fully hyperphosphorylated state, and it persisted in the partially hyperphosphorylated state that is thought to represent active Swe1 (Harvey et al., 2011). These observations suggest that signals that control Swe1 could prolong Cdk1 inhibitory phosphorylation in poor nutrients.
The role of Swe1 was further characterized by analyzing daughter cell growth and mitotic events in single cells. In rich carbon, swe1Δ caused a slight reduction in the average duration of metaphase, as previously described (Figure 3B; see Figure S2A for dot plots and p-values) (Lianga et al., 2013). In poor carbon, swe1Δ caused a greater reduction in metaphase duration, from 34 minutes to 29 minutes. Although swe1Δ reduced the duration of metaphase in poor carbon, it did not reduce it to the mitotic duration observed for wild type cells in rich carbon, which indicated that the mitotic delay caused by poor carbon is not due solely to inhibitory phosphorylation of Cdk1. Loss of Swe1 had little effect on the duration of anaphase, consistent with the observation that Cdk1 inhibitory phosphorylation is observed primarily during metaphase (Figure 3A).
In both rich and poor carbon, swe1Δ caused a reduction in growth rate (Figure 3C). This, combined with the reduced duration of metaphase, caused swe1Δ daughter buds to undergo mitotic transitions at a substantially reduced size in both rich and poor carbon (Figure 3D; see Figures S2C,D for dot plots and p-values). Together, the data demonstrate that Swe1 plays a role in the increased duration of metaphase in poor carbon and is required for normal control of daughter cell size at cytokinesis. The reduced growth rate of daughter buds in swe1Δ cells could be due to the reduced size of their mother cells.
We next defined the contribution of Cdk1 inhibitory phosphorylation to the mitotic delay observed in rts1-AID cells. Western blot assays confirmed that destruction of rts1-AID caused a prolonged mitotic delay in both rich and poor carbon (Figure S3A). The delay caused by rts1-AID in rich carbon was reduced, but not eliminated, by swe1Δ. We further discovered that rts1Δ caused a mitotic delay after release from a metaphase arrest. The delay was reduced by swe1Δ, but not fully eliminated (Figure S3B).
In single cell assays, the metaphase delays caused by rts1-AID in rich and poor carbon were reduced by swe1Δ, but not fully eliminated (Figure 3E; see Figure S2A for dot plots and p-values). The increased duration of anaphase caused by rts1-AID in rich and poor carbon was largely unaffected by swe1Δ. Although swe1Δ did not fully rescue the mitotic delays caused by rts1-AID, it caused rts1-AID cells to exit mitosis at sizes similar to those of swe1Δ cells (Figure 3F; see Figure S2 for dot plots and p-values). This was a combined result of the reduction in mitotic duration and decreased growth rate in rts1-AID swe1Δ cells relative to rts1-AID or swe1Δ cells.
Together, these observations demonstrate that nutrients and PP2ARts1 control mitotic duration and daughter cell size via a Swe1-dependent pathway, as well as by a Swe1-independent pathway. A previous study found that purified PP2ARts1 can not dephosphorylate
Swe1 in vitro, so it is likely that it controls Swe1 indirectly (Harvey et al., 2011).
Phosphorylation of Pds1 is controlled by nutrients and PP2ARts1
We next investigated the function and regulation of Pds1. Proteome-wide mass spectrometry analysis identified four serines in Pds1 that are hyperphosphorylated in rts1Δ cells (S185, S186, S212, S213) (Zapata et al., 2014). Two of the sites (S212,S213) were previously implicated in delaying mitosis in response to DNA damage (Wang et al., 2001). We hypothesized that hyperphosphorylation of Pds1 at these sites delays mitotic progression in poor carbon and in rts1Δ cells. To test this, we first used Phos-Tag gels to determine whether Pds1 undergoes hyperphosphorylation when cells are shifted to poor carbon. Hyperphosphorylated forms of Pds1 could be detected within 5 minutes of a shift to poor carbon (Figure 4A, lanes 1–3). Moreover, rts1Δ caused Pds1 to undergo more rapid and extensive hyperphosphorylation in response to poor carbon (Figure 4A, lanes 5–7). We also observed that rts1Δ caused Pds1 to undergo a slight shift to hyperphosphorylated forms in cells growing in rich carbon (Figure 4A, compare lanes 1 and 5). A mutant version of Pds1 that lacks the 4 sites controlled by PP2ARts1 (pds1–4A) largely failed to undergo hyperphosphorylation when shifted to poor carbon (Figure 4A, lanes 9–11). Thus, Pds1 undergoes rapid phosphorylation in response to carbon source, and PP2ARts1 is required for control of Pds1 phosphorylation.
Pds1 and Swe1 are required for nutrient modulation of the duration of metaphase
To test whether Pds1 contributes to the mitotic delay observed in swe1Δ cells in poor carbon, we analyzed the effects of pds1–4A on the duration of mitosis in swe1Δ cells. Metaphase in swe1Δ pds1–4A cells was shorter than metaphase in wild type cells or swe1Δ cells in both rich and poor carbon (Figure 4B; see Figure S2A for dot plots and p-values). In addition, there was no difference in the duration of metaphase between rich and poor carbon in the swe1Δ pds1–4A cells. Importantly, these data suggest that nutrient modulation of the duration of metaphase could be due solely to regulation of Swe1 and Pds1.
pds1–4A swe1Δ cells growing in poor carbon completed metaphase at a reduced volume compared to pds1–4A swe1Δ cells growing in rich carbon (Figure 4C; see Figure S2C for dot plots and p-values). This was entirely a consequence of reduced growth rate in poor carbon, since the duration of metaphase in pds1–4A swe1Δ cells was identical in rich and poor carbon. Thus, pds1–4A swe1Δ causes cell size at completion of metaphase to become a simple function of growth rate, as originally imagined in early models of cell size control (Hartwell and Unger, 1977).
In both rich and poor carbon, the duration of anaphase and the extent of growth during anaphase were increased in pds1–4A swe1Δ cells relative to swe1Δ cells. In addition, the duration of anaphase in the pds1–4A swe1Δ cells was modulated by nutrients. These observations suggest that the mechanisms that control the duration of anaphase in response to carbon source are largely independent of Swe1 and Pds1. The increased duration of anaphase in pds1–4A swe1Δ cells may reflect compensatory growth that occurs because the cells complete metaphase at an abnormally small size. Proteome-wide mass spectrometry identified numerous components of the mitotic exit network as potential targets of PP2ARts1-dependent regulation (Zapata et al., 2014). Thus, nutrient-dependent control of the mitotic exit network could account for the increased duration of growth in anaphase in poor nutrients.
The average size of pds1–4A swe1Δ buds at completion of anaphase in poor carbon was slightly larger than swe1Δ cells, which was due to a few very large outlier cells (see dot plots in Figure S2D). We noticed that some pds1–4A swe1Δ cells in poor carbon had more than two spindle pole bodies (Figure 4D). This suggested that they have multiple nuclei, which could be a consequence of a failure to undergo sufficient growth before nuclear division. Cells in which we could detect extra spindle poles, which constituted approximately 10% of the cells growing in poor carbon, were excluded from the analysis in Figures 4B and 4C. However, the large outlier cells in the pds1–4A swe1Δ data could be polyploid cells that did not have well separated spindle poles that would clearly identify them as having multiple nuclei. Since cell size scales with ploidy, the unusually large pds1–4A swe1Δ cells could correspond to polyploid cells.
Pds1 and Swe1 are not required for the proportional relationship between cell size and growth rate
A plot of cell size at completion of mitosis versus growth rate revealed that pds1–4A swe1Δ cells have a reduced size, yet show a proportional relationship between cell size and growth rate (Figure 5A). Moreover, analysis of pds1Δ swe1Δ cell size distributions showed that they are abnormally small, yet still show evidence of nutrient modulation of cell size (Figure 5B). In contrast, rts1Δ pds1Δ swe1Δ cells show little evidence of nutrient modulation of cell size (Figure 5C).
Together, the data show that Pds1 and Swe1, by virtue of their ability to increase the duration of growth in mitosis, strongly influence how nutrients modulate cell size. Thus, cells that lack Pds1 and Swe1 are abnormally small and fail to undergo an appropriate increase in size when shifted from poor to rich carbon. The data further suggest that Pds1 and Swe1 are controlled by PP2ARts1 and nutrient-dependent signals, consistent with a role in setting cell size in response to growth rate. Yet Pds1 and Swe1 are not strictly required for the proportional relationship between cell size and growth rate during growth in mitosis, whereas PP2ARts1 is required.
In previous work, we found that PP2ARts1 influences a TORC2 signaling network that is required for normal control of cell growth and size in G1 phase (Lucena et al., 2017). The TORC2 network includes a feedback loop in which TORC2 promotes production of ceramides, while ceramides generate signals that inhibit TORC2 signaling. An inhibitor of ceramide production causes a dose-dependent decrease in growth rate during G1 phase, as well as a decrease in cell size. Thus, manipulation of signals within the TORC2 network in cells that have functional PP2ARts1 causes strong effects on growth rate and cell size, yet the proportional relationship between cell size and growth rate appears to be maintained.
Here, we have extended these studies to show that PP2ARts1 strongly influences cell growth and size in mitosis. Importantly, we also show that PP2ARts1 is required for a proportional relationship between cell size and growth rate during mitosis. To explain the data, we suggest that PP2ARts1 relays signals within the TORC2 network that ensure a proportional relationship between cell size and growth rate. Cell growth is driven by complex signaling networks that control the rates of the diverse pathways that comprise cell growth. Since the rate of growth in mitosis is proportional to cell size, one might expect that the signals that drive growth are also proportional to cell size, and that specific mechanisms ensure that the signals that drive growth rate scale with size. In this case, a failure in a PP2ARts1-dependent mechanism that makes growth rate proportional to cell size should cause cells to grow at rates that are uncorrelated with size, leading to a loss of the correlation between cell size and growth rate. The correlation would breakdown in both directions – growth rate would no longer be proportional to cell size, and cell size would no longer be proportional to growth rate because cells of the same size would have different growth rates. The fact that PP2ARts1 influences signaling within the TORC2 feedback loop suggests that it is well-positioned to enforce a mechanistic link between cell size and growth rate (Lucena et al., 2017). Moreover, there is evidence that the level of signaling in the feedback loop is influenced by growth, which might be expected for a signaling network that ensures that cell size and growth rate are proportional (Clarke et al., 2017). For example, one way to enforce a proportional relationship between cell size and growth rate would be to have the events of growth generate feedback signals that modulate the signals that drive growth.
The data further suggest that the PP2ARts1-dependent signals that make cell size proportional to growth rate also control the activity of Pds1 and Swe1 to set the threshold amount of growth required for cell cycle progression. In this view, Pds1 and Swe1 would mediate PP2ARts1-dependent signals that delay cell cycle progression until growth reaches a threshold appropriate for the growth rate. Thus, Pds1 and Swe1 would be required for cells to reach a size that is appropriate for the growth rate, but they would not be required for the proportional relationship between cell size and growth rate.
A model in which the same global signals that set growth rate also set the critical amount of growth required for cell cycle progression would provide a simple mechanistic explanation for the proportional relationship between cell size and growth rate. Since cell size control likely evolved as an outcome of growth control, it would make sense that control of cell growth and size are mechanistically linked. Further analysis of the signals that connect Pds1 and Swe1 to PP2ARts1 and the TORC2 network should lead to important clues to how cell growth and size are controlled.
Materials and Methods
Yeast strains and media
The genotypes of the strains used in this study are listed in Table 1. All strains are in the W303 background (leu2–3,112 ura3–1 can1–100 ade2–1 his3–11,15 trp1–1 GAL+, ssd1-d2). Genetic alterations were carried out using one-step PCR-based integration at the endogenous locus (Longtine et al., 1998; Janke et al., 2004) or by genetic crossing.
pds1–4A mutants were created by replacing a 170bp fragment of PDS1 that contains S185, S186, S212 and S213 with the URA3 marker. The replacement was carried out in a PDS1–3xHA::TRP strain so that the final mutant version of PDS1 would be tagged with 3XHA. The URA3 marker gene was replaced by a mutant fragment of PDS1 in which all 4 sites were mutated to alanine, which was synthesized by overlap PCR. The mutant fragment was co-transformed with a LEU2 marked CEN vector (YCplac111) and LEU+ transformants were selected to enrich for transformation-competent cells. The LEU+ transformants were then replica plated to FOA to select for cells that lost the URA3 marker. Pre-selection for the LEU2 CEN vector dramatically reduced the background of spontaneous ura3 mutants. The resulting strain was backcrossed once to wild type and then to DK815 to obtain DK3320, which was transformed with SPC42-GFP::HIS to obtain DK3335.
For cell cycle time courses and analysis of cell size by Coulter counter, cells were grown in YP media (1% yeast extract, 2% peptone, 8ml/L adenine) supplemented with 2% dextrose (YPD), or with 2% glycerol and 2% ethanol (YPGE). For microscopy, cells were grown in complete synthetic media (CSM) supplemented with 2% dextrose (CSM-Dex) or 2% glycerol and 2% ethanol (CSM-G/E).
Microscopy and image analysis
Microscopy, image analysis, and statistical analysis of microscopy data were carried out as previously described (Leitao and Kellogg, 2017)
Cell cycle time courses, western blotting and Coulter counter analysis
Cell cycle time courses utilizing cells arrested in G1 phase were carried out as previously described. To arrest cells in mitosis, GAL1-CDC20 cells were grown overnight in YP media containing 2% raffinose and 2% galactose and arrested by washing into YP containing 2% raffinose. Cells were monitored until most cells had large buds. Cells were released from metaphase by re-addition of 2% galactose. SDS-polyacrylamide gel electrophoresis and western blotting were carried out as previously described. Blots were probed overnight at 4°C with affinity-purified rabbit polyclonal antibodies raised against Swe1 or with a phosphospecific antibody that recognizes Cdk1 phosphorylated at tyrosine 19 (Cell Signaling Technology, cat# 10A11). Blots were exposed to film or imaged using a ChemiDoc™ MP System (Bio-Rad). For quantification of rts1-AID degradation, band intensity was quantified using ImageLabTM. Destruction of Rts1-AID was initiated by addition of 1 mM auxin from a 50 mM stock made in 100% ethanol.
For PhosTag western blots, cells were lysed by bead-beating into sample buffer without phosphatase inhibitors. After cell lysis, samples were centrifuged for 1 min at 13,000 rpm at 4°C and quickly placed in a boiling water bath for 7 min. Samples were loaded into 10% SDS-polyacrylamide gels supplemented with 100 µM PhosTag and 200 μM MnCl2. To prepare PhosTag gels, the gel mixture was degassed for 5 min prior to addition of TEMED and polymerization was allowed to occur for 1–2 hours at room temperature followed by overnight at 4°C. Gels were run at 2.5 mA for 16 hrs until a 29-kD marker was at the bottom of the gel. The gel was incubated for 10 min in transfer buffer supplemented with 2 mM EDTA, followed by a second incubation without EDTA. Gels were transferred onto nitrocellulose via the Trans-Blot Turbo Transfer System (Bio-Rad). Blots were probed at room temperature with the 12CA5 anti-HA monoclonal antibody followed by HRP-conjugated anti-mouse secondary antibody. Secondary antibodies were detected via chemiluminescence using Quantum substrate (Advansta).
Analysis of cell size by Coulter counter was carried out as previously described (Leitao and Kellogg, 2017).
Acknowledgments
We thank Ben Abrams, Facilities Manager for the UCSC Life Sciences Microscopy Center for support and mentoring with all microscopy related techniques, and the Aldea lab for sharing BudJ: an ImageJ plugin to analyze images of budding yeast cells (http://www.ibmb.csic.es/home/maldea). R.L. was supported by a fellowship from the “Fundação para a Ciência e a Tecnologia” (FCT), with funds from “Programa Operacional Potencial Humano/ Fundo Social Europeu” (POPH/FSE), under the fellowship SFRH/BD/75004/2010. This work was supported by National Institutes of Health Grant R01-GM053959.