Several factors may control trehalose and glycogen synthesis, like the glucose flux, the growth rate, the intracellular glucose-6-phosphate level and the glucose concentration in the medium. Here, the possible relation of these putative inducers to reserve carbohydrate accumulation was studied under well-defined growth conditions in nitrogen-limited continuous cultures. We showed that the amounts of accumulated trehalose and glycogen were regulated by the growth rate imposed on the culture, whereas other implicated inducers did not exhibit a correlation with reserve carbohydrate accumulation. Trehalose accumulation was induced at a dilution rate (D)≤0.10 h−1, whereas glycogen accumulation gradually increased at decreasing growth rates. The growth rate dependency of trehalose accumulation was supported by studies in cells overexpressing the G1-cyclin CLN3. The trehalose level appeared to be dependent on the duration of the G1 phase, as trehalose was only accumulated at a G1 phase duration of more than 5 h in both wild-type and CLN3-overexpressing cells. On the other hand, the glycogen level was reduced by CLN3 overexpression in a cell cycle-independent manner. A possible regulatory mechanism that links trehalose and glycogen accumulation to the growth rate is discussed.
Saccharomyces cerevisiae is able to accumulate trehalose and glycogen under appropriate conditions. The regulation of these reserve carbohydrates has been studied intensively during recent years (for review see ). The regulation has been demonstrated to occur on the transcriptional level, but in addition various enzymes have also been demonstrated to be involved, as well as several metabolites [2–5]. Of these latter glucose-6-P especially appeared to affect trehalose and glycogen accumulation [3,5]. The high number of different components involved in trehalose and glycogen accumulation suggests that different factors regulate the accumulation under different conditions. However, a common parameter under all conditions that lead to accumulation of trehalose and glycogen is a low growth rate or no growth at all. Therefore it is tempting to suggest that a common mechanism might exist that underlies regulation of both the growth rate of the cells and the accumulation of trehalose and glycogen. According to this idea, the accumulation of reserve carbohydrates would be strictly related to the growth rate or, in other words, to the duration of the cell cycle. In this context it is of interest that trehalose and glycogen accumulate during the G1 phase of the cell cycle . Furthermore, several possible pathways have been identified that may link regulation of trehalose and glycogen synthesis with progression through the cell cycle. The duration of the G1 phase of the cell cycle in response to nutrient availability is mainly dependent on the activity of the Cln3/Cdc28 kinase complex [7–9]. The availability of nutrients regulates the amount of Cln3 in the cell at both the transcriptional and the translational level. Addition of glucose or a nitrogen source to nutrient-starved cells rapidly increases the CLN3 level in the cell . At the translational level, the phosphatidylinositol kinase homologue Tor2 positively regulates the translation rate of Cln3 upon nutrient addition . The TOR signalling pathway functions upstream of Msn2 and Msn4, transcription factors that bind to stress response (STRE) elements in promoter sequences [11–14]. When the TOR pathway is deactivated, Msn2/4 become localised in the nucleus and activate STRE-controlled transcription of genes . The genes encoding trehalose synthase (TPS1) and glycogen synthase (GSY1 and GSY2) contain STRE elements in their promoter region [16,17]. Therefore, accumulation of carbohydrates and translational control of one of the main regulators of cell cycle progression, Cln3, seem to be under control of the same pathway. Although molecular studies suggest a link between the cell cycle and reserve carbohydrate accumulation, no direct evidence has been found that supports this possible correlation.
In this study, the correlation between trehalose and glycogen accumulation and the cell cycle was examined under well-defined growth conditions using nitrogen-limited continuous cultures. It was demonstrated that the growth rate applied to the culture relates to the amount of accumulated carbohydrates. Furthermore, we demonstrated that the amount of accumulated trehalose and glycogen is related to the duration of the G1 phase of the cell cycle.
2Materials and methods
2.1Yeast strains and plasmids
The haploid strain CEN-PK 113-7D (SUC2, MAL2-8c, MEL) was used as wild-type in all experiments. The CLN3 overexpression strain was obtained by insertion of PCR-amplified CLN3 into the multiple cloning site of yeast 2-μm plasmid pYEX-BX (Clontech, Palo Alto, CA, USA) and transformation into strain CEN-PK 113-5D (SUC2, MAL2-8c, MEL, ura3). The pYEX-BX plasmid contains the cupper-inducible CUP1-promoter, which has a low basal level of expression. No additional cupper was added in our experiments (YNB medium contains 0.14 μM Cu2+).
The −18 to +1773 fragment of CLN3 was amplified by PCR using a primer upstream of the CLN3 starting codon (5′-TATGGATCCTGATACGCTTTCTGTACGATG-3′) with a BamHI site (underlined) introduced and a primer downstream the CLN3 starting codon (′5-CTATGTCGACTTTGTCGTTTCAGCGAGTTTTC-3′) with a SalI site (underlined) introduced. The PCR product was digested with BamHI and SalI and inserted in the BamHI/SalI site of pYEX-BX. The plasmid construct was sequenced to justify that no mutations were generated by PCR.
2.2Fed-batch growth conditions
All experiments were performed at 30 °C in yeast nitrogen base without amino acids (6.7 g l−1 YNB, Difco) with galactose as carbon source. Synchronous fed-batch cultures were performed in YNB medium with a constant residual galactose concentration of 0.15 mM at a density of 1–2×107 cells ml−1. Galactose was continuously added at rates ranging from 12 fmol cell−1 h−1 to 20 fmol cell−1 h−1. The cell number and the external galactose concentration were monitored throughout the experiment. Galactose consumption rates were determined as described previously .
Centrifugal elutriation was performed essentially as described previously, with some modifications . The yeast strains were grown exponentially in YNB medium containing 1% galactose at 30 °C and 2×1010 cells were loaded in a 40-ml chamber of a Beckman J-6MI centrifuge (JE-5.0 rotor) at 30°C and 2000 rpm. Cells were cultivated in the elutriator chamber on YNB medium containing 1% galactose. Newborn daughter cells were washed out at a flow rate of 45 ml min−1 and collected on ice. The cell size was monitored with a Coulter Multisizer II, and the flow rate of the elutriation was adapted to maintain a constant cell size.
For continuous-culture experiments, cells were grown in a 2-l BiofloIII chemostat (New Brunswick Scientific, Nijmegen, The Netherlands) connected to a computer controller unit running on Advanced Fermentation Software (New Brunswick Scientific). Cells were inoculated in the medium as described previously , and a continuous feed was connected after batch growth overnight. The EGLI medium used in the continuous cultures was described previously , with NH4+ concentration adapted to 1.5 g l−1 for nitrogen limitation. In studies using varying dilution rates under nitrogen limitation, cells were grown at dilution rates (D) ranging from 0.07 up to 0.25 h−1 at a constant feed concentration of 200 mM glucose. During studies at constant dilution rates of 0.1 h−1 and 0.2 h−1, the glucose-feed concentrations ranged from 240 to 420 g l−1 (D=0.1 h−1) and from 140 to 340 g l−1 (D=0.2 h−1), respectively. Steady-state samples were taken as described previously . Additional samples for metabolite determination were rapidly frozen in cold methanol and isolated as described before .
2.5Analysis of sample parameters
Cell sizes and cell numbers were determined with an electrical particle counter (Coulter Multisizer II). Cell sizes were calculated by calibration with latex beads of known size. For determining budding percentages 200 cells were analysed microscopically. Dry weights (DW) were determined by spinning down 10 ml of culture volume in duplicate. Cell pellets were washed twice in water, transferred into pre-weighed bottles and dried for at least 12 h at 120°C before weight determination.
2.6Determination of trehalose and glycogen levels
Samples of 2 ml were centrifuged for 30 s at maximum speed (Eppendorf centrifuge). The cells were washed in ice-cold water, and trehalose and glycogen were extracted from the pellet as described previously .
External glucose and internal metabolite concentrations were determined essentially as described previously . In all metabolite determinations, enzyme and specific cofactors were added to the sample in a 100-mM imidazole, 10-mM MgCl2, pH 7.0 buffer.
3.1Trehalose and glycogen accumulation in synchronized cells grown in fed-batch cultures
The carbohydrates trehalose and glycogen accumulate in batch cultures when nutrients become depleted and cells enter the stationary phase . At the onset of trehalose and glycogen accumulation under these growth conditions the nutrient flux rapidly drops, which may trigger accumulation. The effects of the nutrient flux on trehalose and glycogen accumulation were studied by growing cells at different galactose consumption rates in fed-batch cultures. Synchronized cells were used to maintain a constant cell number and a constant galactose consumption rate per cell throughout the growth. As shown in Fig. 1, a biphasic correlation was observed between the galactose flux and the amount of carbohydrates accumulated. At galactose consumption rates of 20 fmol galactose cell−1 h−1 and higher, low amounts of trehalose and glycogen were accumulated, at maximal levels of 6 fmol (glucose) cell−1 trehalose and of 18 fmol (glucose) cell−1 glycogen. As the galactose consumption rate decreased from 20 to 14 fmol galactose cell−1 h−1, trehalose and glycogen accumulation rapidly increased to 38 and 57 fmol (glucose) cell−1, respectively. These results indicate that accumulation of carbohydrates is regulated by the nutrient flux applied to the cells.
As the carbon flux in the cell was decreased, the duration of the G1 phase of the cell cycle increased (Fig. 1). The minimal G1 phase duration of cells grown in batch culture at 55 fmol galactose cell−1 h−1 (1% galactose) was determined at 110 min. At decreasing galactose consumption rates, the G1 phase duration slowly increased to 160 min at 20 fmol galactose cell−1 h−1. As the galactose flux decreased further, the G1 phase duration rapidly elongated from 160 to 520 min at 14 fmol cell−1 h−1. Summarising, cells grown in fed-batch cultures at low carbon fluxes accumulated trehalose and glycogen concomitant with an increase in duration of the G1 phase.
3.2Effects of the carbon flux on trehalose and glycogen accumulation
In fed-batch cultures, the carbon flux and the growth rate of cells are always correlated, but in nitrogen-limited continuous cultures these parameters can be studied independently . The effects of both carbon flux and growth rate on trehalose and glycogen accumulation were therefore studied under these growth conditions. Cells were grown in nitrogen-limited continuous cultures at a constant growth rate and the glucose concentration in the feed medium was varied. At a constant dilution rate of 0.1 h−1 and feed concentrations from 240 to 420 mM glucose, the external glucose concentration increased from 2 to 17 mM glucose and the glucose flux increased from 4.1 to 7.7 mmol g−1 h−1 (Fig. 2A). On applying a dilution rate of 0.19 h−1 and feed concentrations from 140 to 340 mM glucose, the external glucose concentration increased from 1 to 97 mM glucose as the glucose flux increased from 8.0 to 12.6 mmol g−1 h−1 (Fig. 2B). Thus, under these growth conditions the glucose flux varied independently from the growth rate of the culture.
The effect of growth rate, glucose concentration and glucose flux on carbohydrate accumulation was examined by determining the amount of trehalose and glycogen in these nitrogen-limited continuous cultures. When grown at a growth rate of 0.1 h−1, the glucose flux increased from 4.1 to 7.7 mmol g−1 h−1, but the trehalose and glycogen accumulation remained at a constant high level of 5 and 110 μg (glucose) (mg DW)−1, respectively (Fig. 3A). This indicates that the glucose flux is not the trigger for carbohydrate accumulation under these growth conditions. At a higher growth rate of 0.19 h−1, both trehalose and glycogen were accumulated at low levels, although only the trehalose level remained constant. As the glucose flux increased from 8.0 to 12.6 mmol g−1 h−1, trehalose was accumulated at an amount of 0.8 μg (glucose) (mg DW)−1. Also at higher growth rates, the glucose flux did not correlate with the amount of trehalose accumulated, which indicates that the glucose flux does not regulate trehalose accumulation in these continuous cultures. At a growth rate of 0.19 h−1, the amount of accumulated glycogen decreased from 49 to 14 μg (glucose) (mg DW)−1 as the glucose flux increased from 8.0 to 12.6 mmol g−1 h−1 (Fig. 3B). This decrease in glycogen accumulation coincided with a substantial increase in the external glucose concentration from 1 to 97 mM glucose (Fig. 2A). In contrast to trehalose accumulation, intracellular glycogen levels may be influenced by the glucose flux or the external glucose concentration. Nevertheless our results indicate that the glucose flux is not the main regulator of glycogen accumulation, as constant levels of glycogen were accumulated at the low growth rate independent of the glucose flux.
Previous studies in batch cultures have also suggested that changes in glucose-6-phosphate levels regulate the trehalose and glycogen level. Upon depletion of glucose in the batch cultures, carbohydrates accumulate concomitant with a high level of glucose-6-phosphate and a drop in growth rate . In our continuous cultures, the intracellular glucose-6-phosphate concentration remained constant at approximately 7 μmol (mg DW)−1 at all growth conditions, whereas the accumulation of trehalose and glycogen decreased at increasing growth rates. Therefore, it appears that the glucose-6-phosphate level is not the main regulator of trehalose and glycogen accumulation in these experiments.
3.3Correlation between trehalose and glycogen accumulation and the growth rate
The results from both fed-batch cultures and nitrogen-limited continuous cultures demonstrated that increases in trehalose accumulation did not correlate with the carbon flux or intracellular glucose-6-phosphate levels. On the other hand, trehalose and glycogen accumulation appeared to be accompanied by a decrease in growth rates under all growth conditions. To further investigate this possible relation, we studied trehalose and glycogen accumulation as a function of the growth rate in nitrogen-limited continuous cultures (Fig. 4). Under these conditions the growth rate is determined by the addition of the growth-limiting substrate, while other variables are kept constant. As the concentration of the growth-limiting substrate in the culture is extremely low under all conditions, the flux is linearly related to the growth rate of the cells.
High amounts of trehalose were accumulated in cells grown at dilution rates from 0.07 h−1 to 0.095 h−1, at 10.5 to 8.8 μg (glucose)(mg DW)−1, respectively. At growth rates of D=0.125 h−1 and higher, low amounts of trehalose were accumulated [1.4 μg (glucose)(mg DW)−1]. Glycogen accumulation gradually decreased from 116 μg (glucose)(mg DW)−1 at D=0.07 h−1 to 42 μg (glucose)(mg DW)−1 at D=0.19 h−1. Whereas trehalose only accumulated at low growth rates, glycogen accumulation gradually decreased as the growth rate increased. These results clearly show that the content of accumulated carbohydrates can be regulated by the growth rate of the cells.
3.4Decreased trehalose and glycogen accumulation in CLN3-overexpressing cells
The results described above suggest a relationship between cell cycle duration and reserve carbohydrate production. To obtain further support for this suggestion, trehalose and glycogen accumulation in CLN3-overexpressing cells was compared to that in wild-type cells. Overexpression of CLN3 resulted in a shorter G1 phase and a reduction in cell size [22,23], and in an increase in the amount of daughter cells produced in asynchronous fed-batch cultures. After growth in a fed-batch culture at an initial galactose consumption rate of 15 fmol galactose cell−1 h−1 for 24 h, wild-type cells had formed 3.65×109 cells ml−1 from an initial population of 2.07×109 cells ml−1, which is a 1.76-fold increase in the number of cells. Overexpression of CLN3 resulted in the formation of 4.46×109 cells ml−1 from an initial culture of 1.38 ×109 cells ml−1, which is 3.23-fold increase. To study a possible relationship between G1 phase duration and accumulation of trehalose and glycogen, their levels were measured after growing the cells asynchronously for 24 h in the fed-batch culture. Trehalose was accumulated to an amount of 8.5 fmol (glucose) cell−1 in wild-type cells, and to an amount of 4.4 fmol (glucose) cell−1 in CLN3-overexpressing cells. Glycogen was accumulated to 7.0 fmol (glucose) cell−1 in wild-type cells, and accumulation decreased to an amount 3.8 fmol (glucose) cell−1 in cells overexpressing CLN3 (Fig. 5). These results indicate that overexpression of CLN3 results in a decrease in the G1 phase duration in cells growing asynchronously in fed-batch cultures, concomitantly with a decrease in trehalose and glycogen accumulation.
3.5Trehalose accumulation depends on duration of the G1 phase
As shown above, decreasing the duration of the cell cycle results in a reduction of trehalose and glycogen accumulation. To demonstrate that this phenomenon is dependent on the cell cycle duration, wild-type and CLN3-overexpressing cells were cultured in a nitrogen-limited continuous culture, so that the cell cycle duration of both strains was identical and determined solely by D.
Trehalose and glycogen levels were determined at growth rates from 0.07 h−1 to 0.15 h−1 and plotted against the length of G1 phase of CLN3-overexpressing and wild-type cells. The G1 phase duration was calculated from the budding percentage of the cells and the applied growth rate (Fig. 6A,B). The accumulated trehalose was below 2 μg (glucose)(mg DW)−1 at a G1 phase of 4.5 h and lower. As the G1 phase increased to more than 5 h, higher levels of trehalose were accumulated: up to 10.5 μg (glucose)(mg DW)−1 in wild-type cells with a G1 phase of 10 h and 8.1 μg (glucose)(mg DW)−1 in CLN3-overexpressing cells with a G1 phase of 8 h. These results indicate that trehalose accumulation is directly correlated with the duration of the G1 phase, both in wild-type and CLN3-overexpressing cells.
The glycogen levels showed a 2.5-fold decrease upon overexpression of CLN3 at the same G1 length as wild-type cells (Fig. 6B). At a short G1 phase of 2 h, wild-type cells accumulated 42 μg (glucose)(mg DW)−1, whereas CLN3-overexpressing cells accumulated 16 μg (glucose) (mg DW)−1. As the G1 phase was elongated to 8 h, wild-type and CLN3-overexpressing cells accumulated 105 and 38 μg (glucose)(mg DW)−1, respectively. These results indicate that glycogen accumulation is influenced by the length of the G1 phase but in addition also by CLN3, in contrast to accumulation of trehalose.
Several reports have described that accumulation of trehalose and glycogen coincides with a reduced growth rate of the cell [4,24]. However, different factors connected with low growth rates in batch culture have been implicated as regulators of reserve carbohydrate accumulation, like the level of glucose-6-phosphate, a decrease in the glycolytic flux, and depletion of external glucose [5,6].
By using nitrogen-limited continuous cultures, the influence of the individual factors on accumulation of trehalose and glycogen was studied at different growth rates. Here, we show that trehalose and glycogen levels can be regulated by altering the growth rate of wild-type cells in continuous cultures. Moreover, the glucose flux, the glucose-6-phosphate levels and the external glucose concentration do not correlate with the amount of reserve carbohydrates accumulated under different growth conditions. Although glucose-6-phosphate levels appear to control the accumulation of reserve carbohydrates in batch cultures [5,6], another factor seems to control the carbohydrate levels in our continuous cultures. Interestingly, trehalose accumulation is directly related to duration of the G1 phase and is induced at a G1 phase of 5 h and more. In agreement with this, recent experiments have shown that heat shock treatment of cells results in a decrease in growth rate , concomitant with accumulation of trehalose. This suggests that, in response to non-optimal environmental conditions in batch cultures, the growth rate is reduced and accumulation of trehalose is initiated. A similar relation between growth rate and reserve carbohydrate level is observed in our continuous cultures, as application of a lower growth rate results in an increase in the reserve carbohydrate level. Therefore, the growth rate of the culture can control the accumulation of reserve carbohydrates.
In nitrogen-limited continuous cultures, the growth rate is directly related to the nitrogen flux of the culture. Previous studies have described a significant increase in the trehalose and glycogen levels in response to nitrogen depletion (for review see ). Therefore, the nitrogen flux may also be involved in regulating the trehalose and glycogen levels in our experiments. However, treatment of cells with heat shock in batch cultures where the nitrogen source is still available also results in a reduction of the growth rate concomitant with the accumulation of reserve carbohydrates (data not shown). Furthermore, in nitrogen-limited continuous cultures, the nitrogen concentration is virtually zero at all dilution rates, whereas differences in the accumulation pattern of reserve carbohydrates are observed at the different dilution rates. Although changes in the nitrogen concentration may influence the level of carbohydrate accumulation, it seems not to be the primary determinant for trehalose and glycogen accumulation in the continuous culture experiments.
In contrast to the level of trehalose, glycogen levels in continuous cultures are influenced by overexpression of CLN3. The amount of Cln3 in the cell regulates the duration of the G1 phase and is therefore one of the main determinants of the growth rate of wild-type cells [10,26,27]. In wild-type cells, glycogen levels gradually increase as the growth rate decreases, whereas overexpression of CLN3 results in a decrease in the glycogen level, which was confirmed in the fed-batch experiments. These results indicate that the accumulation of glycogen is dependent on the amount of Cln3 in the cell. Previous studies have shown that Cln3 functions upstream of the SBF transcription factor, which regulates transcription of, amongst others, cyclins PCL1 and PCL2[22,28,29]. These cyclins can activate the CDK Pho85 at the end of the G1 phase of the cell cycle and initiate cell cycle progression [22,29]. Deletion of Pho85 results in hyperaccumulation of glycogen, whereas Pho85 activity is involved in phosphorylation of glycogen phosphorylase and glycogen degradation [30,31]. Possibly, overexpression of CLN3 results in an increase in glycogen degradation by induction of Pho85 activity.
The mechanism that couples the growth rate to the accumulation of carbohydrates has yet to be identified, but our results imply a link with cell cycle-regulatory pathways. A good candidate for such a regulatory function is the TOR pathway. This controls STRE-dependent expression by regulating the localisation of transcription factors Msn2 and Msn4 . The promoters of the genes encoding trehalose and glycogen synthetase contain several STRE elements that may therefore be negatively regulated by TOR activity [15–17]. Inhibition of TOR activity results in a reduction in translation of Cln3, cell cycle arrest early in the G1 phase, and accumulation of carbohydrates . Together with our results, this suggests that the TOR pathway may be involved in linking the growth rate to the amount of carbohydrate accumulation.
Recent observations further indicate that not only genes involved in trehalose and glycogen accumulation but also other STRE-controlled genes are regulated by the growth rate of the cell. The HXT5 gene, whose promoter contains two STRE-sequences, is expressed whenever trehalose is accumulated under conditions when cells have low growth rates . The growth rate-dependent behaviour of carbohydrate accumulation generally seems to coincide with the presence of STRE elements in the promoter of up-regulated genes. Genome-wide analysis of gene expression at different growth rates may reveal a possible general regulatory mechanism in which low growth rates induce expression of STRE-regulated genes.