Longevity of U cells of differentiated yeast colonies grown on respiratory medium depends on active glycolysis

Abstract

Colonies of Saccharomyces cerevisiae laboratory strains pass through specific developmental phases when growing on solid respiratory medium. During entry into the so-called alkali phase, in which ammonia signaling is initiated, 2 prominent cell types are formed within the colonies: U cells in upper colony regions, which have a longevity phenotype and activate the expression of a large number of metabolic genes, and L cells in lower regions, which die more quickly and exhibit a starvation phenotype. Here, we performed a detailed analysis of the activities of enzymes of central carbon metabolism in lysates of both cell types and determined several fermentation end products, showing that previously reported expression differences are reflected in the different enzymatic capabilities of each cell type. Hence, U cells, despite being grown on respiratory medium, behave as fermenting cells, whereas L cells rely on respiratory metabolism and possess active gluconeogenesis. Using a spectrum of different inhibitors, we showed that glycolysis is essential for the formation, and particularly, the survival of U cells. We also showed that β-1,3-glucans that are released from the cell walls of L cells are the most likely source of carbohydrates for U cells.

Abbreviations

2DG

2-deoxy-D-glucose

2PE-CM

2-photon excitation confocal microscopy

AEBSF

4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride

IAM

iodoacetamide

BKP

bromcresol purple

DTT

dithiothreitol

DIC

differential interference contrast

DMSO

dimethyl sulfoxide

GFP

green fluorescent protein

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

SD

standard deviation

TCA

tricarboxylic acid cycle

TORC1

target of rapamycin complex 1

Introduction

When growing on complete respiratory agar medium, yeast colonies pass through specific developmental phases that are characterized by changes in the pH of surrounding medium from acidic to alkali and vice versa. In the alkali phase, ammonia is released and functions as a signal¹ involved in several processes, including synchronization of colony development,² activation of metabolic reprogramming³ and cell differentiation.^4-8 Two prominent differently localized cell subpopulations are formed during colony transition from the acidic to alkali developmental phase: U cells in the upper layers that are stress-resistant cells with a longevity phenotype and L cells in the lower layers that die more quickly and are sensitive to stresses.⁶ A comparison of U and L cell transcriptomes implied significant differences between these 2 cell types in their metabolic pathways and plasma membrane transporters and regulators and indicated a flow of nutrients and waste products between the U and L cell populations of colonies that resembled the Cori and glutamine/NH₃ cycles⁶ that have been described between cells of solid tumors and other tissues of tumor-affected mammalian organisms.⁹ A comparison of U and L cells from 14-day-old giant colonies⁶ with U_m and L_m of 3- to 4-day-old microcolonies,⁵ both occurring in the alkali developmental phase, enabled processes specific to ammonia signaling to be distinguished from processes mostly related to colony aging.⁵ Hence, most of the processes specific to U/U_m cells, such as decreased respiration, production of specific proteins (e.g., Ato proteins), typical morphological features, and TORC1 activation, are related to ammonia signaling. On the other hand, several specific features of L cells, particularly those related to lower viability and stress sensitivity, are more related to aging of these cells.^4,10

Here, we show that prominent expression differences in metabolic enzymes in U and L cells are reflected in the activities of enzymes belonging to particular metabolic pathways, such as glycolysis in U cells and tricarboxylic acid (TCA) and glyoxylate cycles and gluconeogenesis in L cells. We also show that active glycolysis is essential for the formation and viability of U cells and that the degradation of cell wall material, particularly in L cells, could be a source of carbohydrates for U cells.

Results

Differences in central carbon metabolism between U and L cells from differentiated colonies

Previous transcriptomic analyses have indicated metabolic differences between U and L colonial cells, including upregulated expression of genes involved in glycolysis and the pentose-phosphate pathway in U cells and of those involved in oxidative phosphorylation and gluconeogenesis in L cells.⁶ A more detailed comparison of complete L and U cell transcriptomic data⁶ with published microarray datasets of yeast cultures undergoing diauxic shift¹¹ and cultures grown on different carbon sources¹² revealed that the transcriptome of U cells is similar to those of pre-diauxic-shift cells and cells grown on fermentable carbon sources, while L cells are transcriptionally similar to post-diauxic-shift cells and cells grown on non-fermentable carbon sources (see Supplementary data and Fig. S1). With the aim of determining the actual enzymatic equipment of both cell types, we measured the activities of most of the enzymes that are involved in central carbon metabolism in lysates of U and L cells harvested from 14-day-old colonies. Enzymes of the TCA cycle and enzymes involved in the glyoxylate cycle (isocitrate lyase and malate synthase), as well as fructose-1,6-bisphosphatase (an enzyme specific to gluconeogenesis), were all significantly more active in L cells than in U cells (Fig. 1; Table 1). The activities of both NAD- (glutamate degrading) and NADP-dependent (glutamate synthesising) glutamate dehydrogenases connecting the TCA cycle to nitrogen metabolism were also higher in L cells. On the contrary, higher activities of most of the studied glycolytic enzymes were detected in U cells (Fig. 1; Table 1). One of the highest differences in glycolytic activity was found for phosphofructokinase, a glycolysis-specific enzyme that is not involved in gluconeogenesis. The activity of pyruvate decarboxylase, which is a key enzyme in directing pyruvate to fermentation rather than to respiration, and the activities of the pentose-phosphate shunt enzymes glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase were also increased in U cells. Hexokinase activity was the only activity related to glycolysis that was enhanced in L cells.

Figure 1.

Activities of enzymes of central carbon metabolism in (U) and (L) cells from 14-day-old colonies. In green are enzymes that are more active in U cells; the values represent U/L-activity ratios from Table 1. In red are enzymes that are more active in L cells; the values represent L/U-activity ratios from Table 1.

Table 1.

Enzymatic activities in U and L cells

Pathway	Enzyme	Activity in U cells1	Activity in L cells1
Glycolysis	Hexokinase*	3 412±758	4 495±564
	Glucose phosphate isomerase*	2 300±259	2 082±157
	Phosphofructokinase*	156±15	121±12
	Aldolase*	205±34	163±26
	Phosphoglycerate kinase	2 588±357	2 247±334
	Enolase*	1 008±79	865±33
	Pyruvate kinase	223±38	224±30
Fermentation	Pyruvate decarboxylase**	2 981±438	1 407±121
	Alcohol dehydrogenase	4 329±288	4 118±517
	Aldehyde dehydrogenase (NAD)*	23.5±1 .0	11.1±3 .2
	Aldehyde dehydrogenase (NADP)*	68.7±6 .8	95.3±4 .9
Gluconeogenesis	Fructose-1.6-bisphosphatase*	23.5±4 .1	32.8±9 .1
Pentose-phosphate shunt	Glucose-6-P dehydrogenase*	275±48	236±28
	6-phosphogluconate dehydrogenase*	396±34	312±19
TCA cycle	Citrate synthase**	23.3±2 .5	39.6±5 .6
	Aconitase***	11.1±0 .8	17.6±0 .9
	Isocitrate dehydrogenase (NAD)***	73±9	111±13
	Isocitrate dehydrogenase (NADP)***	188±34	243±41
	Fumarase***	31.0±7 .1	59.4±4 .8
	Malate dehydrogenase**	402±52	653±75
Glyoxylate bypass	Isocitrate lyase***	234±33	356±42
	Malate synthase***	197±16	482±76
Other	2-oxoglutarate dehydrogenase (NAD)*	71.1±17 .5	128±15
	2-oxoglutarate dehydrogenase (NADP)*	46.1±6 .1	79.1±12 .5

^*t-test P-value <0.05.

^**t-test P-value <0.01.

^***t-test P-value <0.001.

¹Enzymatic activities are shown as mmol of product produced (or substrate consumed) per min per 1 mg of protein in cell lysate. Activities ± SD are shown (n≥3).

Concomitantly with increased TCA enzyme activities, L cells harvested from colonies also consumed more oxygen than U cells⁶ (Fig. 2A). Additionally, an assay using the tetrazolium dye MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) pointed to a higher respiratory capacity of L cells (Fig. 2A). However, this assay is not strictly specific to respiratory activity, as it can detect the activities of NAD(P)-dependent oxidoreductases.¹³

Figure 2.

Fermentative and respiratory capacities of cells from developing colonies. (A) Comparison of capacities of U and L cells from 14-day-old colonies. (B) Timeline of development of capacities of colonial cells. Exponential cultures from liquid synthetic defined medium with glucose (SD-Glc) or glycerol-ethanol (SD-GE) are included for comparison. The mean is shown; n=3. Error bars, SD; t-test P-values are indicated with asterisks.

Production of fermentation end products by U and L cells

We also analyzed the capacities of U and L cells separated from 14-day-old colonies to produce the end products of fermentation, such as ethanol, acetate and CO_2, when incubated in liquid YPD under hypoxic conditions. Ethanol, acetate and CO₂ production was higher by 35%, 281% and 82%, respectively, in U cells than in L cells (Fig. 2A), showing that U cells possessed significantly higher fermentative capacity.

Timeline of development of fermentative versus respiratory capacities of U and L cells

The above data pointed to a higher fermentative capacity of U cells and higher respiratory capacity of L cells when taken from fully differentiated colonies. We aimed to follow the progression of these metabolic properties within specifically localized cells during colony development. We harvested cells from 2- to 16-day-old colonies and measured the abilities of these cells to produce ethanol, acetate and CO₂, and also to respire. As typical U and L cells gradually start to develop after 8–10 d of colony development,⁶ younger colonies (2–6 d old) were analyzed as one population. Exponential cultures from liquid synthetic defined media with glucose or glycerol/ethanol were used as the reference of cells with fully fermentative and fully respirative metabolism, respectively. As shown in Figure 2B, until day 8, cells maintained a relatively high capacity to produce both ethanol and CO₂. Later, the capacities of both CO₂ and ethanol production started to decrease in L cells but remained unchanged in U cells. Acetate production capacity gradually decreased in young colonies and was then maintained at a low level in L cells from day 8. During the same period (8–16 days), U cells increased their capacity to produce acetate back to a value typical of young colonies. The differences in production of fermentative products between U and L cells are similar to or even higher than the differences between fully fermentative and fully respiratory reference cells, which confirms that observed differences between U and L cells are physiologically relevant. Lower absolute values observed in some cases for colonial cells very probably reflect the fact that the colonial cells are chronologically older and thus generally less active than exponential reference cells. Respiratory capacity also dropped during early developmental phases (colonies 3–6 d old) and later only partially decreased in L cells; it decreased much faster in U cells after day 8, reaching about half of the typical value for L cells.

Effect of inhibitors of glycolysis and respiration on U and L cell formation and viability

To estimate the importance of glycolysis and respiration in the formation and viability of U and L cells, we applied different metabolic inhibitors to developing colonies of the strain BY-pTEF-GFP. The inhibitors were added in different concentrations and at different colony development times. At 2–7 d after inhibitor application, the morphology and vitality of U and L cells were analyzed by 2PE confocal microscopy (2PE-CM) of colony cross-sections. Both the morphology of cells based on GFP fluorescence and the presence of dead/weakened cells stainable by bromcresol purple (BKP) were monitored.

To inhibit glycolysis, we applied either 2-deoxy-D-glucose (2DG), a non-competitive inhibitor of hexokinase¹⁴ and competitive inhibitor of phosphoglucomutase¹⁵ or iodoacetamide (IAM), an irreversible inhibitor of glyceraldehyde-3-phosphate dehydrogenase¹⁶ near the colonies. As shown in Figure 3A–B, both of these inhibitors significantly affected the morphology and viability of U cells when applied during an interval of 7–14 d of colony development. While typical large U cells with small vacuoles are clearly visible in control colonies, smaller cells with large vacuoles are visible in the upper parts of colonies treated with 2DG or IAM. A similar effect on cell morphology was observed after the treatment of upper cells that were separated from 11- and 14-day-old colonies with 2DG (Fig. 3C). In addition, the viability of upper cells in situ in colonies was dramatically decreased, as most of the cells became stainable with BKP dye. On the other hand, the inhibitors of glycolysis were observed to have almost no effect on L cell morphology and viability.

Figure 3.

Effect of inhibitors of glycolysis on morphology and viability of (U) and (L) cells. (A) 2PE-CM of vertical cross-sections of 14-day-old-colonies treated for 7 d with inhibitor (20x objective). (B) Enlarged image (63x objective) showing differences in cell morphology and viability between control and 2DG-treated colonies. (C) Effect of 2DG treatment (1 day) on morphology of U cells separated from 11- and 14-day-old colonies. White/black arrows indicate examples of vacuolated cells.

To analyze the importance of respiration, we tested the effects of a spectrum of inhibitors of mitochondrial function. We tested the effects of different concentrations of antimycin A and thenoyltrifluoroacetone, inhibitors of the electron transport chain, of oligomycin A and triethyltin, inhibitors of ATP synthase, and of arsenite, a Krebs cycle inhibitor.^17-21 Although higher concentrations of some of these inhibitors significantly repressed the growth of cells in the margin regions of the colonies (data not shown), which agrees with the fact that the colonies were grown on glycerol-ethanol complete respiratory medium (GMA), no significant effect of these inhibitors on the morphology and/or viability of L cells (as well as U cells) was observed when applied to colonies that were 7–14 d old (not shown).

Potential sources of sugars to drive glycolysis in U cells

The above data showed that glycolysis is important for the development and viability of U cells. This finding raised an important question: what are the sources of sugars in U cells in 14-day-old colonies growing on respiratory GMA? We hypothesized that nutrients for U cells could be provided by the second major colony cell subpopulation of L cells.⁶ However, we were unable to detect free glucose directly in extracellular extracts of colonies or after the incubation of cells separated from colonies in water (data not shown). We therefore measured the total amount of extracellular carbohydrates released by U and L cells separated from 14-day-old colonies when incubated in water. As shown in Figure 4A, L cells released more than twice the amount of total carbohydrates released by U cells, although U cells contained 51% more carbohydrates per unit of biomass than L cells (Fig. 4B). To answer the question of whether both of these cell types are able to consume these extracellular carbohydrates, we incubated a mixture of equal amounts of U and L cells. In this instance, the amount of extracellular carbohydrates was approximately 27% lower than the amount calculated to be produced by separated U and L cells. This result clearly shows that carbohydrates released by L cells are immediately consumed by U cells. An additional experiment in which both cell types were individually incubated with 0.1% glucose found nearly double the rate of glucose consumption by U cells (Fig. 4C), which proved the U cells' ability to consume extracellular sugars more effectively than L cells.

Figure 4.

Flow of carbohydrates in (U) and (L) cells isolated from 14-day-old colonies. (A) Levels of total carbohydrates, β-1,3-glucans, trehalose and glycogen were measured in supernatants after 16 h incubations of equal amounts of separated U or L cells or cell mixtures (U+L) (n=5). Note that in the U+L cell mixtures, only half of the U- and half of the L-cell biomass is present. Theoretical values of the U+L mixture (U+L-CALC) are calculated as 1/2 of the U value + 1/2 of the L value and represent the anticipated results if U cells did not consume carbohydrates produced by L cells. (B) Total carbohydrates in the cell biomass (n=4). (C) Rate of glucose consumption by cells (n=4). Error bars, SD; t-test P-values are shown.

Potential sources of extracellular carbohydrates include hexoses and storage carbohydrates (glycogen and trehalose) released from the cells and free oligosaccharides or even monosaccharides coming from the degradation of cell walls. As no free glucose was detected, we measured extracellular trehalose and glycogen produced by U and L cells when they were incubated for 16 h in water. We only found a very low amount of these compounds in extracellular liquid (less than 2% of total released carbohydrates) (Fig. 4A). In contrast, β-1,3-glucans were a substantial part of the total carbohydrates that were released by separated L and U cells (approximately 60%); however, in absolute values, U cells released only about one third of the amount of β-1,3-glucans that were released by L cells. In the U+L mixed cell population, the amount of β-1,3-glucans was 32% lower than the theoretical value calculated from production by separated U and L cells.

Discussion

The cell population of an S. cerevisiae colony differentiates during its development into 2 major, specifically localized subpopulations of U and L cells. In addition to other specific features, differences in expression of genes involved in various metabolic pathways⁶ pointed to fermentative/respiro-fermentative metabolism of U cells and respiratory features of L cells. The enzymatic activities that we determined confirmed this hypothesis because all of the measured enzymes of the TCA cycle and of the glyoxylate shunt were at least 1.5 times as active in L cells as in U cells. In contrast, the activity of pyruvate decarboxylase, a crucial enzyme in fermentation, was 2.1x higher in U cells than in L cells. U cells also possessed generally higher activity (by 10–30%) in their glycolytic enzymes than L cells. It has been shown in fed-batch yeast cultures that a mere difference of 10–20% in the activities of glycolytic enzymes could lead to fold differences in the fermentative capacities of these cultures.^22-24 Therefore, we considered 10–30% differences in glycolytic enzyme activities between U and L cells to be physiologically relevant and to contribute to a higher fermentative capacity in U cells. The higher production of fermentation end products by U cells further supported this conclusion.

Both glutamate-synthesising NADP-dependent and glutamate-degrading NAD-dependent glutamate dehydrogenases are more active in L cells extracts, indicating glutamate degradation and resynthesis in L cells. However, actual metabolic flow in vivo reflects not only the enzymatic capabilities of the cells (as measured in vitro) but also the actual concentrations of substrates and products participating in metabolic flux regulation within the cells. Moreover, under conditions of high NH₄⁺, NAD-dependent glutamate dehydrogenase can in vivo function in the direction toward glutamate synthesis²⁵ and higher activity of both glutamate dehydrogenase forms could control the balance between 2-oxoglutarate and glutamate in respiratory active cells,²⁶ such as L cells. As we have shown previously that U cells release NH₄⁺ and uptake glutamine while L cells presumably release glutamine and uptake NH₄⁺ in a metabolic cycle resembling glutamine-NH₄⁺ cycle in the mammalian organism,^6,9 we presume that in L cells the overall glutamate dehydrogenase reaction runs toward glutamate and glutamine synthesis.

All of these data show that even when colonies are grown on a non-fermentable carbon source and differentiate into U and L cell subpopulations at a time when most cells have already stopped growing and most nutrients have been spent, the subpopulation of U cells is engaging in active fermentative/respiro-fermentative metabolism. In addition, the higher respiratory capacity of L cells compared to U cells is related to a significant decrease in the respiratory capacity of U cells rather than to increased respiration in L cells compared to their younger progenitors. These data are in accordance with findings that the formation of U cells is linked to active metabolic reprogramming of U cell ancestors during ammonia signaling, which is accompanied by the activation of different pro-growth features in metabolically active U cells, while L cells gradually attenuate some of the features of L cell progenitors.^4,5

Glycolytic inhibitors 2-deoxyglucose and iodoacetamide, inhibiting glycolytic enzymes hexokinase and phosphoglucomutase (2DG) and glyceraldehyde-3-phosphate dehydrogenase (IAM), completely disrupted the formation of typical U cells and decreased the viability of already formed U cells. These inhibitors, however, can have quite pleiotropic effect on living cells. For example, 2DG has been shown to interfere with cell wall synthesis and to cause ATP depletion irrespective of whether ATP comes from glycolysis or respiratory metabolism.²⁷ IAM, on the other hand, is a thiol-reactive compound inhibiting enzymes with a cysteine in their active site and causing depletion of reduced glutathione.²⁸ The fact that the treatment of colonies with inhibitors that affect different glycolytic enzymes gave the same results (significantly affecting U cells without any effect on L cells) suggests that the block in glycolysis is the cause for the observed effect. This conclusion is supported by previous observation that U cells of colonies of the strain pfk2Δ, lacking the gene for phosphofructokinase, decrease their viability.⁶ These data imply that active glycolysis is a crucial pathway in U cells and that its absence damages the properties of U cells and their longevity phenotype. Glycolysis has to be activated no later than on the 7^th day of colony development (i.e., when U and L cell diversification is initiated). Nevertheless, the fact that we were not able to observe any significant effect of inhibitors of respiration to L cells does not necessarily mean that respiration is not important in these cells. The testing of respiration inhibitors was complicated by the fact that increases in the concentrations of these inhibitors significantly reduced overall colony growth on respiratory GMA medium at all time-points tested. Thus, we cannot exclude that the certain level of inhibition of respiration at some time-point of colony development affects specifically the formation and/or survival of L and/or U cells.

The effect of fermentative and respiratory metabolism on yeast longevity in general is not straightforward. Several studies of liquid cultivations in glucose medium have suggested that shifting metabolism from fermentation to respiration increases longevity.^29-32 However, our results showed an inverse behavior of cells in colonies grown on complete glycerol-ethanol medium. Here, fermentative/respiro-fermentative metabolism was important for the longevity of U cells, while a higher respiratory capacity did not provide L cells with better survival during long-term aging. The induction of genes for glycolysis and respiration in the upper and lower cells of colonies, respectively, was also shown in glucose-grown colonies³³ when glucose had probably been spent. A benefit of induced glycolysis and reduced respiration similar to U cells was found in tor1Δ, sch9Δ and ras1Δ strains grown in liquid glucose medium.³⁴ However, colonial U cells with a longevity phenotype have an active TORC1 pathway and glycolysis and therefore more resemble tumor cells in mammals than “classical” stationary-phase yeast cells. In contrast to glycolysis, increased gluconeogenesis has been indicated as one of the hallmarks of aging in yeast, and its reduction leads to prolonged life-span.^35,36 In accordance, L cells that exhibit a higher rate of dying⁶ also exhibit increased activity of the key gluconeogenic enzyme fructose-1,6-bisphosphatase.

Fermentative metabolism induced in U cells suggests that these cells metabolize sugars despite growing on respiratory medium. The finding of a higher rate of carbohydrate release in L cells compared to U cells (although L cells have less carbohydrates per unit of biomass) pointed to L cells being a source of sugars for U cells. The finding of a higher amount of β-1,3-glucan being released into extracellular fluid by L cells compared to U cells, together with the observed decrease in extracellular β-1,3-glucan in the U+L mixture compared to the value calculated from the amounts released by separated U and L cells, implies that the cell walls of L cells could be a significant source of carbohydrates that are subsequently consumed by U cells and possibly metabolized as a source of fermentable sugars. This conclusion is also supported by a previous finding that L cells strongly express a number of genes for putative and confirmed endoglucanases (EGT2, SCW11, DSE2, DSE4)³⁷ and also by the fact that SUN-family proteins related to cell wall glucanases were identified in the intercellular spaces of colonies.³⁸ Importantly, we also show here, that, in contrast to U cells, L cells themselves are not able to effectively consume the carbohydrates that they release. We can hypothesize that either L cells acquire these features as part of their developmental program or that U cells somehow push L cells to induce these features and prevent L cells from re-consuming released carbohydrates or other nutrients. The existence of such a mechanism would explain why L cells exhibit a starvation phenotype⁶ while U cells do not, despite their living in similar environments and having similar resources available. Any nutrients that appear in the colony environment are then preferentially utilized by the population of U cells, which leads to a better prospective for long-term survival. Interestingly, as revealed by transcriptome analysis, glucose-limited chemostat cultures maintain a fermentative metabolism when fed extremely low quantities of glucose.³⁹ These data indicate that even a low supply of sugars, which are immediately metabolized, could induce fermentative metabolism in slowly growing or even non-growing cells, such as U cells.

Materials and Methods

Yeast strains, cultivations and inhibitor treatments

The BY4742 strain was used in all experiments. Isogenic BY-pTEF-GFP⁵ was used for confocal microscopy and inhibitor treatment to visualize differences in cell morphology and viability between U and L cells. Colonies were grown on GMA medium (2% agar, 1% yeast extract, 1% ethanol, with or without 100 mg/l of the pH indicator bromcresol purple). Synthetic defined liquid media with either 2% glucose (SD-Glc) or 3% glycerol/1% ethanol (SD-GE) were composed of 0.5% (NH₄)₂SO₄, 0.1% KH₂PO₄, 0.05% MgSO₄, 0.1% Wickerham's solution, 50 mg/l histidine, lysine and leucine and 20 mg/l of uracil. For inhibitor studies, 5–50 μl aliquots of stock solutions of individual inhibitors (0.5 M 2-deoxy-D-glucose (2DG) in water, 0.4 M iodoacetamide (IAM) in DMSO, 4 mM oligomycin in DMSO, 20 mM antimycin A in DMSO, 0.8 mM triethyltin in water, 50 mM sodium arsenite in water or 2.4 mM thenoyltrifluoroacetone in water) were placed into a well in agar adjacent to a colony. Control colonies were treated with a particular solvent (DMSO or water) in the same way. Colonies at 7, 10 and 14 d of development were treated for 2–5 d with inhibitors. In Figure 3, cross-sections of colonies treated with 10 μl of 2DG or 2 μl of IAM analyzed by 2PE-CM are shown. U cells (10 mg/ml) isolated from differentiated colonies were treated with 20 mM 2DG for 24 h.

Isolation of U and L cells

U and L cells were isolated from colonies by centrifugation in a density gradient as previously described.⁶ To avoid metabolic changes that may be induced by sucrose, we used a linear 10–35% gradient of sorbitol instead of sucrose for gradient centrifugation.⁵

Respiration and fermentation rate measurement

For oxygen consumption measurement, isolated U and L cells were resuspended in distilled water to a concentration of 100 mg/ml. The respiration of 50 μl of cell suspension in 1 ml of 5% ethanol was determined as the rate of oxygen consumption and was measured with an oxygen meter equipped with an MT200 Mitocell (Strathkelvin Instruments). For MTT assay, in which yellow MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) is reduced to the water-insoluble purple product of formazan inside of cells, 25 or 50 mg of biomass was resuspended in 5 ml of 50 mM succinate as a substrate and 0.5 mM MTT and 0.1 mM phenazine ethosulphate were added. After 1 hour at 30°C, the cells were centrifuged, and the formazan product of MTT reduction was extracted by vortexing them in 1 ml of DMSO. The A₅₇₀ of the extracts was measured, and the formazan concentration was calculated using an extinction coefficient of 13,000 M⁻¹cm⁻¹. Fermentation rate was measured either as the production of ethanol, acetate or CO₂ by 10 mg of cells in 1 ml of nitrogen-saturated YPD supplemented with 250 μg/ml of cycloheximide. Cycloheximide was added to the YPD medium to prevent potential changes in U and L cells caused by incubation with glucose. Cell suspensions were incubated in completely filled, tightly sealed vials to ensure almost anaerobic conditions. After 3 hours in 30°C and cell removal, produced metabolites were assayed with an Ethanol Assay Kit and an Acetic Acid Assay Kit (Megazyme). CO₂ production was monitored in a syringe with 10 mg of cells in 1 ml of YPD with cycloheximide (any air bubbles in the medium were carefully removed), which was connected to a vertically placed 1-ml glass pipette with a 100 μl air bubble trapped in its lower part by a column of water. CO₂ that is produced in the syringe increases the volume of the air bubble, which can be measured on the pipette scale.

Carbohydrate measurement

Total carbohydrates released by U and L cells and carbohydrate content of the biomass were measured by a phenol-sulfuric acid method.⁴⁰ Briefly, 15 mg of isolated U or L cells in 30 μl of water were incubated at 30°C. After 16 hours, 150 μl of water was added, the cells were removed (14.000 g, 5 min), 70 μl of the supernatant was mixed with 210 μl of concentrated H₂SO₄ and 45μl of 5% phenol solution in water, and the solution was incubated at 90°C for 5 min. The absorbance of the reaction product was measured at 490 nm. The concentration of carbohydrates was calculated using glucose as a standard. For the detection of carbohydrates released by the U+L cell mixture, 15 μl suspensions of both cell types were mixed and incubated in the same way. The theoretical amount in the U+L mixture (U+L-CALC) was calculated as the sum of the carbohydrate amount determined for separated U and L cells according to U and L cell biomass in the U+L mixture. To determine the total carbohydrates in U and L cells, 100 μg of cell biomass in 70 μl of water was used. To determine glycogen, trehalose and β-1,3-glucan content, 25μl of supernatant after a 16-hour incubation with cells was mixed with 25 μl of 0.2 M sodium acetate buffer (pH 4.5) and 5 μl of either 2 U/ml amyloglycosidase (Sigma), 0.1 U/ml trehalase (Megazyme) or 0.2 U/ml β-1,3-D-glucanase (Sigma). A 20 hour incubation at 37°C was used for the trehalase and β-1,3-glucanase reactions, and a 20 hour 57°C incubation was used for the amyloglycosidase reaction. The amount of liberated glucose was determined with a GO glucose assay kit (Sigma). For glucose uptake experiments, 20 mg of isolated U and L cells were incubated in 1 ml of 0.1% glucose in 0.1x PBS. Aliquots of 50 μl were taken after 0, 15, 30, 45 and 60 min of incubation, and the remaining glucose was measured using a GO glucose assay kit (Sigma). The rate of glucose consumption was calculated from a slope obtained by linear regression.

Crude cell extract preparation and enzyme assays

Isolated U and L cells were stored at −80°C until use. Cells were resuspended and disrupted in extraction buffer (20 mM HEPES, pH 7.1; 1 mM DTT; 100 mM KCl; 40 μl/ml protease inhibitor cocktail (Roche); and 1 mM 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF)) by glass beads using FastPrep (Thermo). Cell lysate was cleared by 10 min of centrifugation (9000 g, 4°C). Enzyme assays were conducted with a protein concentration of 2 mg/ml at 25°C. The experimental details are in Supplemental material.

Enzyme assays with cell extracts (diluted to a protein concentration of 2 μg/μl) were conducted according to the cited references (see supplementary methods for details) and included the following: hexokinase (EC 2.7.1.1),⁴¹ phosphoglucose isomerase (EC 5.3.1.9),⁴² phosphofructokinase (EC 2.7.1.11),⁴³ fructose-1,6-diphosphate aldolase (EC 4.1.2.13),⁴⁴ 3-phosphoglycerate kinase (EC 2.7.2.3),⁴⁵ enolase (EC 4.2.1.11),⁴⁶ pyruvate kinase (EC 2.7.1.40),⁴⁷ pyruvate decarboxylase (EC 4.1.1.1), alcohol dehydrogenase (EC 1.1.1.1),⁴⁸ K⁺-dependent aldehyde dehydrogenase (EC 1.2.1.5),⁴⁹ Mg²⁺-dependent aldehyde dehydrogenase ⁵⁰ glucose-6-phosphate dehydrogenase (EC 1.1.1.49),⁴⁸ 6-phopshogluconate dehydrogenase (EC 1.1.1.44),⁵¹ fructose-1,6-bisphosphatase (EC 3.1.3.11),⁵² citrate synthase (EC 4.1.3.7),⁵³ aconitase (EC 4.2.1.3),⁵⁴ NAD-dependent isocitrate dehydrogenase (EC 1.1.1.41),⁵⁵ NADP-dependent isocitrate dehydrogenase (EC 1.1.1.42),⁵⁶ fumarase (EC 4.2.1.2),⁵⁷ malate dehydrogenase (EC 1.1.1.37),⁵⁸ isocitrate lyase (EC 4.1.3.1), malate synthase (EC 4.1.3.2),⁵⁹ NAD-dependent glutamate dehydrogenase (EC 1.4.1.3) and NADP-dependent glutamate dehydrogenase (EC 1.4.1.4).⁶⁰ Two different amounts of crude extract were measured for each reaction to verify the linearity of the assay. At least 3 independent cell isolations were measured.

Microscopy

2PE-CM of colonies was performed according to Vachová et al. (2009). In brief, colonies were embedded in low-gelling agarose and cut vertically down the middle. The cut surface of the colonies were placed on a coverslip, and colony side views were obtained by 2PE-CM. Images were acquired with a true confocal scanner microscope (SP2 AOBS MP; Leica) fitted with a mode-locked laser (Ti:Sapphire Chameleon Ultra; Coherent Inc.) for 2-photon excitation and 20×/0.70 and 63×/1.20 water immersion plan apochromat objectives. An excitation wavelength of 920 nm was used, and emission bandwidths were set to 480–595 nm for GFP and 600–750 nm for BKP fluorescence. Cells harvested from colonies were observed under a Leica DMR fluorescent microscope using a DIC and HCX PL fluotar 100×/1.3 oil objective.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.

Funding

This work was supported by the Grant Agency of the Czech Republic (15–08225S), Charles University in Prague (UNCE 204013), RVO 61388971 and by the project ‘BIOCEV – Biotechnology and Biomedicine Center of the Academy of Sciences and Charles University’ (CZ.1.05/1.1.00/02.0109) from the European Regional Development Fund and by the Ministry of Education, Youth and Sports of the Czech Republic.