Transfer in the Yeast Saccharomyces cerevisiae - NCBI

Plant Physiol. (1992) 100, 769-777 0032-0889/92/100/0769/09/$01.00/0

Received for publication March 2, 1992 Accepted June 1, 1992

Iron Reduction and Trans-Plasma Membrane Electron Transfer in the Yeast Saccharomyces cerevisiae' Emmanuel Lesuisse2 and Pierre Labbe* Laboratoire de Biochimie des Porphyrines, Institut 1. Monod, Tour 43, Universite Paris 7, 2 Place Jussieu, 75251 Paris Cedex 05, France ABSTRACT The ferri-reductase activity of whole cells of Saccharomyces cerevisiae (washed free from the growth medium) was markedly increased 3 to 6 h after transferring the cells from a complete growth medium (preculture) to an iron-deficient growth medium (culture). This increase was prevented by the presence of iron, copper, excess oxygen, or other oxidative agents in the culture medium. The cells with increased ferri-reductase activity had a higher reduced glutathione content and a higher capacity to expose exofacial sulfhydryl groups. Plasma membranes purified from those cells exhibited a higher reduced nicotinamide adenine phosphate (NADPH)-dependent ferri-reductase specific activity. However, the intracellular levels of NADPH, NADH, and certain organic acids of the tricarboxylic acids cycle were unchanged, and the activity of NADPH-generating enzymes was not increased. Addition of Fe(lll)-EDTA to iron-deprived and iron-rich cells in resting suspension resulted in a decrease in intracellular reduced glutathione in the case of iron-deprived cells and in an increase in organic acids and a sudden oxidation of NADH in both types of cells. The depolarizing effect of Fe3" was more pronounced in iron-rich cells. The metabolic pathways that may be involved in regulating the trans-plasma membrane electron transfer in yeast are discussed.

The biochemical and physiological basis of the iron-stress response in plants remains poorly understood despite a profuse literature (for a review, see ref. 1). The increased reducing activity of root cells under iron deprivation has been related either to increased concentrations of intracellular reduced pyridine nucleotides (25) or to an increase in the catalytic activity of the trans-plasma membrane electron carrier system (3). Either NADPH (25, 27) or NADH (3) has been proposed as the electron donor for extracellular ferric reduction. However, the metabolic links between the electron flow through the plasmalemma and the iron status of the cells are still unknown. Crane et al. (5) suggested that NADH is the electron donor for the reduction of extracellular ferricyanide in S. cerevisiae because this reduction was stimulated by ethanol and inhibited by pyrazole, an inhibitor of alcohol dehydrogenase activity. A similar proposal was put forward by Yamashoji and Kajimoto (31), who found that a decrease in intracellular NADH concentration occured upon addition of ferricyanide to resting cells. More recently (17), we showed that the increased ferri-reductase activity of iron-deprived cells was accompanied by a significant increase in the activity of cytosolic glucose-6-phosphate dehydrogenase, the first enzyme of the hexose monophosphate pathway, which is assumed to fumish most of the cellular NADPH. We also showed that the NADPH-dependent ferri-reductase activity was increased 2-fold in plasma membranes isolated from iron-deficient cells, whereas the NADH-dependent activity was not. An active protein fraction using NADPH but not NADH as electron donor and containing a weakly bound FMN,3 but no detectable heme, was purified 100-fold with respect to plasma membranes (15). These results suggest that the large increase in ferri-reductase activity of iron-deficient cells could result from increases in both the catalytic activity of the transmembrane redox system (NADPH-dependent) and the glucose flux through the hexose monophosphate pathway, generating NADPH. The present article describes some properties of the ferrireductase activity of whole yeast cells of normal and mutant strains (deficient in glucose-6-phosphate-dehydrogenase or heme) grown under iron-rich and iron-deficient conditions. Other enzymic activities and the levels of some metabolites

The reductive pathway of iron assimilation in plants and microorganisms involves reduction of ferric iron, either at the cell surface via a trans-plasma membrane redox system, or in the extracellular medium via excreted reducing compounds. Iron is then taken up by the cells as soluble Fe2" ions (1, 8, 21). We have previously shown that yeast cells, Saccharomyces cerevisiae, exhibit a ferri-reductase activity that is dramatically increased in response to iron limitation. We provided both physiological and genetic evidence that the reductase activity was required for the uptake of iron by the cells from several ferric chelates (15-17). Heme-deficient mutants, which were shown to be also deficient in ferri-reductase activity, are unable to take up iron efficiently from ferric citrate (16). Our results were recently confirmed by the work of Dancis et al. (6), who also isolated a ferri-reductase-deficient mutant. This work was supported by grants from the Centre National de la Recherche Scientifique, Universite Paris VII, and the Ministere de la Recherche et de l'Enseignement Superieur (MRT 510069). E.L. held a Federation of European Biochemical Societies fellowship. 2 Present address: Unite de Genetique, Universite Catholique de Louvain, Place Croix du Sud 4, B-1348 Louvain-La-Neuve, Belgium. '

3Abbreviations: FMN, flavin mononucleotide, ALA, 5-aminolevulinate; DTNB, 5,5'-dithio-bis(2-nitrobenzoate); DMP, 2-(dimethylaminostyryl)- 1 -ethylpyridinium. 769

770

LESUISSE AND LABBE

suspected of playing a role in iron reduction and/or acquisition by S. cerevisiae were also measured. These data were correlated with the physiological changes that occur upon addition of Fe3" to these cells. MATERIALS AND METHODS Yeast Strains, Growth Conditions, General Experimental Design Three strains of Saccharomyces cerevisiae were used: strain G204 is deficient in ALA synthetase activity and is heme deficient unless ALA is added to the growth medium, which restores the wild-type phenotype of the parent strain FL200 (28); strain T8A is deficient in glucose-6-phosphate dehydrogenase activity (18). The experimental design involved three steps: (a) the cells were precultured for 40 h in various conditions (as specified in the text) and harvested by centrifugation. They were washed twice with 2% EDTA (disodium salt, pH 7) and the pellets were stored overnight at 40C. (b) The stored cells were transferred to the culture media at 1 g wet weight L1` and allowed to grow in various conditions (essentially in iron-deficient or iron-rich conditions; other conditions are specified in the text). Cells were collected from the culture media at intervals (3, 6, 9 h) and washed as described above. (c) The washed cells were suspended in a buffer, and only these resting cells washed free from the culture medium were used for experiments (within the same

day). Unless specified, the preculture medium contained (for 1 L of water): 30 g of glucose, 6.7 g of yeast nitrogen base without amino acids (Difco), 1 g of yeast extract, 50 mg of histidine, 50 mg of uracil, 50 mg of adenine, and either 50 mg of ALA (G204 supplemented with ALA, T8A) or 2 mL of Tween 80 (polyoxyethylene sorbitan mono-oleate) plus 30 of mg ergosterol (G204). Iron-rich and iron-deficient culture media had the same compositions as the preculture medium, but iron-rich medium was supplemented with 180 ,M Fe(III)citrate (1:20); iron was removed from the medium with 3hydroxyquinoline as previously described (16). This procedure also removed Zn2+ and Cu2+ from the medium. Unless specified, the iron-deficient medium was supplemented with 50 mg L-l ferrozine. The cells were precultured and cultured at 300C in Erlenmeyer flasks with magnetic stirring. The buffer used for suspending the cells in resting conditions was either 50 mm citrate (trisodium salt, pH 6.5) or 50 mm Mops, pH 7.

Cell Fractionation and Enzymic Assays The ferri-reductase activity of whole cells and purified plasma membranes was measured as previously described (15, 17) with either Fe(III)-citrate, Fe(III)-EDTA, or ferrioxamine B (180-360 /tM) and bathophenanthroline sulfonate (1 mM) as the Fe2+-trapping reagent. Plasma membrane purification and H+-ATPase assays in vitro were performed as described by Dufour et al. (7). Glucose-6-phosphate dehydrogenase and isocitrate dehydrogenase were assayed by following the increase in absorbance at 340 nm (300C) in cuvettes containing either glucose-6-phosphate or isocitrate (1 mM), NADP+ (1 mM), MgCl2 (1 mM), and 50 ,tg of protein

Plant Physiol. Vol. 100, 1992

mL-' from the cytosolic fraction (as defined previously in ref. 17) in Tris buffer (50 mm, pH 7.5).

Flavin and Reduced Pyridine Nucleotide Contents of the Cells Total flavins were quantified as lumiflavin after hot water extraction of the cells and photolysis as described by Yagi (30). NADH and NADPH were quantified with glycerophosphate dehydrogenase and glutamate dehydrogenase systems according to Klingenberg (14). The cells were incubated at 300C at 100 mg mL-' with magnetic stirring in 50 mm Mops buffer (pH 7). One-milliliter aliquots were withdrawn at intervals and added to 1 mL of 0.5 M KOH in 50% (v/v) ethanol in an ice-salt mixture at -210C. The samples were shaken for 5 min in a 900C water bath, cooled to 0°C, and allowed to stand for 5 min; the pH of the extracts was brought to 7.8 with 0.5 M triethanolamine-HCl, 0.4 M KH2PO4, and 0.1 M K2HPO4, and the samples were allowed to stand for 10 min. The samples were centrifuged and the supernatants used immediately for NADH and NADPH measurements. Cellular Organic Acids and Thiols The cells were preincubated and incubated as described for the assay of NAD(P)H. For organic acids and the total low molecular weight thiols, 1-mL aliquots of the yeast suspension were added to 1 mL of 1 M HCl04 and allowed to stand for 30 min at room temperature and centrifuged. One milliliter of supernatant was added to 0.3 mL of 1 M K2HPO4 and the pH was brought to 7.5 with 5 M KOH; potassium perchlorate was allowed to sediment and the supernatant was used to quantify isocitrate, citrate, and malate with the specific kits from Boehringer. Total thiol groups were assayed by adding 0.7 mL of the neutralized sample to 0.3 mL of 0.2 mm DTNB. After 5 min, the absorbance was read at 412 nm (e = 13.6 mm-' cm-'). Exofacial sulfhydryl groups were assayed by a method modified from Famaey and Whitehouse (9). Aliquots (0.7 mL) of the 100 mg mL-' cell suspension were added to 0.3 mL of 0.2 mm DTNB and allowed to stand at room temperature for 5 min. The samples were then quickly centrifuged and the absorbance of supematants was read at 412 nm.

Membrane Potential The fluorescent reagent DMP was used to probe the changes in the membrane potential of the cells, as described by van de Mortel et al. (29). Washed cells were suspended at 20 mg wet weight mL-1 in 20 mm Mops buffer, pH 7. DMP was added to a final concentration of 1.4 MM and the fluorescence emission at 565 nm was followed using an exciting wavelength of 470 nm. All results refer to wet weight of yeast cells. RESULTS

Reductase Activity of Whole Cells and Isolated Plasma Membranes: Effects of the Culture Parameters Yeast cells harvested after growth in iron-deficient conditions had a 5- to 20-fold greater capacity to reduce extracel-

TRANS-PLASMA MEMBRANE ELECTRON TRANSFER IN SACCHAROMYCES CEREVISIAE

lular ferric iron than cells grown in complete or iron-supplemented medium. The reduction process of the cells in resting suspension was strongly stimulated by the addition of glucose. The effect was biphasic (Fig. 1) with a slight increase in the rate of iron reduction immediately after glucose addition, and the maximum reduction rate was reached after a lag period of 2 to 3 min. Iron-deficient cells reached a plateau after 5 to 10 min, even in the presence of excess substrates. Replacing glucose by ethanol as energy source produced only a small, rapid stimulatory effect (Fig. 1). The ferri-reductase activity of whole cells in batch culture depends on the initial iron concentration in the growth medium and on the cell growth phase; even in iron-rich medium, a peak of ferri-reductase activity occurs in the late exponential growth phase (17). We therefore tried to improve the experimental conditions so that the ferri-reductase activity of the cells would be induced only in iron-deficient medium. These conditions were obtained by inoculating cells from a 40-h preculture (very low cell ferri-reductase activity) at 1 g wet weight L-1 in the culture media. The cell ferri-reductase activity was induced very early in iron-deficient conditions, but remained very low in iron-rich media. The maximum activity occurred after 3 to 6 h in iron-deficient medium, depending on the culture conditions (volume, agitation, aeration) (Fig. 2). At the peak of maximum activity, there was no difference in the growth of cells cultured in iron-rich and cells cultured in iron-deficient media (about 1.5 generations in both cases). Nor was there any difference in either the oxidative or fermentative activities (02 uptake, CO2 release) (data not shown). As previously shown (16), the heme-deficient G204 strain

5 I-,

E E +2.5 cm

Li

TIME (min) Figure 1. Time course of ferric iron reduction by an iron-deficient cell suspension. The cells were incubated at 30°C with 360 gM Fe(IlIl)-citrate and 1 mm bathophenantroline disulfonate in 50 mM citrate buffer, pH 6.5. At the time indicated (arrow), 5% (w/v) glucose (a) or 2% (v/v) ethanol (b) was added.

771

A

41 . ._

4W

X ._

U.

Iu

_m

ql E lUICM t

a

LL

TIME (h)

Figure 2. Evolution of the ferri-reductase activity of cells of different strains during growth. Inset, Dependence of the initial ferric iron concentration in the culture medium (closed symbols) or in the preculture medium (open symbols) on maximal cell ferri-reductase activity. Cells of the strain T8A (0, 0), G204 (A, A), or G204 supplemented with ALA (U, C) were precultured and cultured as described in the experimental section. Aliquots were withdrawn at intervals during growth in iron-rich (closed symbols) or in irondeficient (open symbols) medium, and the ferri-reductase activity of the washed resting cells was measured with 360 ,M Fe(lll)-citrate as substrate (means from 15 experiments). Inset, Different amounts of iron (as Fe[lIl]-citrate or 55Fe[ll]-citrate) were added to irondeficient culture (0) or preculture (0) media. Cells from a single standard preculture medium were inoculated in the different culture media (0), whereas cells from the different preculture media were inoculated in iron-deficient media (0). The cells (strain G204 supplemented with ALA) were harvested after 6 h in culture, and the ferri-reductase activity of the washed cells was measured. The iron content of the cells was determined in some cases as 10, 65, 320, 740, and 850 nmol (g wet weight)-1, respectively, for cells precultured in the presence of 0.1, 1, 10, 100, and 500 $M 55Fe(lIl)citrate (from a single experiment).

lacked inducible ferri-reductase activity unless ALA was added to the growth medium (Fig. 2). The strain T8A, a glucose-6-phosphate dehydrogenase-deficient mutant, exhibited a normal, inducible ferri-reductase activity (Fig. 2). Thus, at least in this strain, there was no possibility of the hexose monophosphate pathway participating directly in generating the reducing equivalents for extracellular ferric reduction. It cannot be excluded, however, that this pathway may play a role in iron reduction by wild-type strains and be replaced by other NADPH-generating activities in the absence of glucose-6-phosphate dehydrogenase. The cell ferri-reductase activity decreased faster as a function of initial iron concentration in the growth medium than as a function of cell initial iron content (Fig. 2). The presence of Cu2" in the growth medium strongly inhibited cell ferri-

LESUISSE AND LABBE

772

reductase activity (50% inhibition at 10 nM Cu2"), whereas Zn2+ was stimulatory, with a maximum around 1 uM (Fig. 3). Other cations such as Co2", Cr3+ (Fig. 3), or Al3" (data not shown) had inhibitory effects at higher concentrations. The cell ferri-reductase activity was increased by 40 and 30%, respectively when 1 mm of the iron-trapping reagent, ferrozine, and 1 mm of the copper-trapping reagent bathocuproin, sulfonate, were added to the growth medium. Interestingly, Cu2+ was reduced in vivo by the cells as well as Fe3+ (data not shown). The presence of 50 uM menadione in the growth medium inhibited cell ferri-reductase by 80%, and continuous air flushing during culture inhibited the activity by 65%. The effect of oxygen was complex since anaerobiosis during culture was also inhibitory (by more than 95%). If the genetic context 'heme+' was essential for the ferri-reductase activity, the presence of a functional mitochondrial respiratory chain was not required because a p0 strain exhibited an activity similar to that of the parental strain (data not shown). Thus, the cell ferri-reductase activity is not simply related to the iron status of the cells. Specific factors (related to the metabolism of iron, copper, oxygen, and heme) and nonspecific factors (presence of extracellular oxidizing promoting agents) probably contribute to the regulation of the ferri-reductase activity of the cells, maybe via modification of their redox status. The ferri-reductase activity was also studied in vitro using isolated plasma membranes. With either Fe(III)-EDTA or

I-

0

U.

.001

.01

.1

1

10

100

1000

CATION CONCENTRATION (uM) Figure 3. Effect of different cations in the growth medium on cell ferri-reductase activity. Iron-deficient media were supplemented with various concentrations of Fe(ill)-citrate (A), CUSO4 (0), ZnSO4 (0), CoCI2 (U), or CrCI3 (0) and inoculated at 1 mg wet weight mL-? with the strain FL200. The ferri-reductase activity of the washed cells was determined after 5 h of growth with Fe(1II)-citrate (360 !M) as substrate (results are means of two experiments).


ferrioxamine B as electron acceptor, the apparent Km for NADPH and NADH was 2 and 30 Mm, respectively. With NADPH (50 Mm) as electron source and ferrioxamine B (180 MM) as electron acceptor, the rate of electron transfer catalyzed by plasma membranes isolated from iron-rich T8A cells was 2.6 nmol min-' (mg protein)-', whereas it was 6.3 nmol min-' (mg protein)-' for iron-deficient T8A cells. The rates were only 0.3 (iron-rich) and 0.4 (iron-poor) nmol min-' (mg protein)-' when NADPH was replaced by NADH. With Fe(III)-EDTA as electron acceptor, these values were 4.3 (ironrich) and 15.7 nmol min-' (mg protein)-' (iron-poor) with NADPH and 2.7 (iron-rich) and 6.4 nmol min-' (mg protein)-' (iron-poor) with NADH. These results indicate that the increased reducing activity of cells transferred to irondeficient medium could result, at least partly, from an increase in the catalytic capacity of the transmembrane electron transport system, as suggested for plants (3). NADPH-Generating Enzymes

Glucose-6-phosphate dehydrogenase, the first enzyme of the hexose monophosphate pathway, and the cytosolic isocitrate dehydrogenase (NADP+-linked) are generally considered to be the key enzymes in the intracellular generation of NADPH (2). As shown in Table I, neither was increased after 6 h of iron deprivation. The glucose-6-phosphate dehydrogenase activity was not detectable in cells of the strain T8A; in strain G204, this activity became significantly higher after 9 h of iron deprivation, at which time the ferri-reductase activity of the cells was decreasing (Table I and Fig. 2). Therefore, the peak of glucose-6-phosphate dehydrogenase activity that was previously found to correlate well with the peak of ferri-reductase activity in batch culture (late exponential growth phase) (17) probably has no decisive role in generating the increased reducing capacity of the cells. The isocitrate dehydrogenase activity of both strains in both culture conditions (+ or -Fe) was very similar and exhibited only minor variations as a function of time (Table I).

Flavins and Reduced Pyridine Nucleotides As for many microorganisms, the iron status of S. cerevisiae affected the flavin content of the cells. The amount of total flavin in G204 cells (supplemented with ALA) in stationary growth phase was 12.7 ± 1.5 ug (g wet weight)-' when grown in iron-deficient conditions and 6.8 ± 0.8 Mug (g wet weight)-' when grown in iron-rich conditions (mean ± SE from three experiments). Because the plasma membrane NADPH-dependent ferri-reductase is a flavoprotein with FMN as cofactor (15), we examined the possibility that the ferri-reductase activity was regulated by the flavin content of the cells as a function of their iron status. Negative results were obtained with both strains tested. Cells of the strain G204 (supplemented with ALA) had a total flavin content of 8.6 ± 0.6 and 7.8 ± 1.1 Mug (g wet weight)-', respectively, after 6 h in iron-deficient or iron-rich culture media. The T8A strain cells had flavin contents of 6.6 ± 0.7 and 6.3 ± 0.4 Mg (g wet weight)-' (mean ± SE from three experiments). A significant increase in the flavin content of iron-deprived cells appeared only after 9 h (data not shown).


773

Table I. Isocitrate Dehydrogenase Activity (a) and Glucose-6-Phosphate Dehydrogenase Activity (b) Associated with Cytosolic Extracts of Different Strains Cultured in Iron-Rich (+Fe) or in Iron-Deficient Medium (-Fe) The results are expressed in ,umol h-1 (mg protein)-' and are either from a single experiment or from three experiments (mean ± SE). Growth time (t) is the number of hours after transferring cells from the preculture medium to the culture medium (iron-rich [+Fe] or iron-deficient [-Fe] medium). Culture Conditions and Growth Time

t=0

t=3h,+Fe t = 3 h, -Fe t=6h,+Fe t=6h,-Fe t = 9 h, +Fe t= 9 h, -Fe a nd, Not detectable.

Strain G204

T8A a

b

a

6±1

nda

5±0.5

nd

4 1 3.6±1.1 3.4±0.9 5 1.7 4±0.8

nd nd nd nd nd

1.1 2.2 1.7 1.6 1.5 1.5 1.7

The changes in intracellular NADH and NADPH were measured in iron-deprived and iron-rich cells in resting suspensions to which glucose and then Fe3" (as Fe[III]-EDTA) were successively added. Typical progress curves are shown in Figure 4 for iron-deficient cells of the T8A strain. In all strains tested (T8A and G204 with or without ALA) and in both growth conditions (+ or -Fe), exogenous glucose addition produced a rapid increase in NADH and, to a lesser extent, in NADPH (Fig. 4). The intracellular levels of NADH and NADPH after glucose addition did not differ significantly between strains or between growth conditions (+ or -Fe) (data not shown). Addition of Fe(III)-EDTA was followed by a sudden oxidation of NADH (Fig. 4). However, a similar decrease in NADH concentration following Fe(III)-EDTA addition occurred in iron-deficient cells, in iron-rich cells, and even in cells showing virtually no reductase activity (G204; data not shown). The addition of Fe(III)-EDTA to T8A cells and G204 cells supplied with ALA resulted in a progressive decrease in NADPH concentration (Fig. 4). This rate of NADPH depletion was low (maximum 10 nmol min-' [g wet weight]-') and not proportional to the reducing activity of the cells (data not shown). From these results, we cannot conclude that either NADH or NADPH was used as electron donor for extracellular ferric reduction because only the NAD(P)H pool size was measured and not the flux through the pool. Rather, our results show that the increased reducing capacity of iron-deficient cells cannot be simply related to an increased level of the intracellular NAD(P)H pool, although an increased cell reducing activity may be accompanied by an increased turnover of NAD(P)H/NAD(P)+. The lag period (2-3 min) between glucose addition and the maximum stimulation of the ferrireductase activity supports this conclusion (compare the effect of glucose in Figs. 1 and 4).

Organic Acids As suggested for plants (24), the increased reducing capacity of iron-deficient cells could result from an increased flux of carbon through citrate, leading to increased NADPH pro-

b 12.6 13 22.8 15.7 13.1 13.5 10.3

G204 cultured with ALA a

6±1 5.7±0.9 4.8 ± 1.2 3.9±0.7 4.1 ±0.7 4.3 ± 0.5 3.7±0.3

b 19.6±2 16.3±0.2 12.5 ± 1.6 11.4±0.9 14.5±1.3 12.8 ± 2 23.6 ± 1.5

0.75f

I

4 9

0 Oa5s

E

z

z

0.25-

*

0

10

15

TIME (min) Figure 4. Intracellular NADH (0, 0) and NADPH (O, *) in irondeficient resting cells. Cells of the strain T8A were cultured for 6 h in iron-deficient medium and incubated in resting conditions as described in the experimental section. Glucose (5% W/V) was added at t = 3 min (first arrow); Fe(lIl)-EDTA (360 jAM) was added to some assays (0, *) at t = 7 min (second arrow) (from a single representative experiment).

LESUISSE AND LABBE

774

duction via mitochondrial aconitase and cytosolic NADP+linked isocitrate dehydrogenase. Figure 5 shows the changes in the intracellular concentrations of citrate, malate, and isocitrate in iron-deprived and iron-rich cells in resting suspension. Addition of glucose resulted in a decrease of citrate and malate concentrations. This decrease could result from the reduction of NAD+ (Fig. 4), leading to inhibition of the tricarboxylic acid cycle (NADH cannot easily be reoxidized by oxygen because in our experimental conditions, anaerobiosis was reached during the 1st min of preincubation). Addition of Fe(III)-EDTA resulted in a rapid increase in citrate (about 200 nmol min-' [g wet weight]-') and malate concentrations (about 120 nmol min-' [g wet weight]-'), whereas isocitrate increased much more slowly (about 20 nmol min-' [g wet weight]-'). Since no difference could be detected between iron-deprived and iron-rich cells (Fig. 5), there is no way of deciding whether the increase in organic acid produc-

1.5

E

0.51

1I

2

4

6

1T

10

TIME (min) Figure 5. Intracellular citric acid (0, *), malic acid (O, *), and isocitric acid (V, V) in iron-deficient (open symbols) and iron-rich resting cells (closed symbols). Cells of the strain G204 (supplemented with ALA) were cultured 6 h in iron-deficient or in ironrich medium and incubated in resting conditions as described in the experimental section. Glucose (5% w/v) and Fe(IlIl)-EDTA (360 /tM) were added at t = 3 min and t = 7 min, respectively (first and second arrow) (results are either means ± SE from three experiments or means from two experiments).


tion upon iron addition actually plays a role in the ferric iron reduction process. The increase should probably be regarded as a consequence of the NADH reoxidation following the addition of ferric iron (see Fig. 4).

Exofacial and Intracellular Thiols The redox status of a cell is closely related to the redox status of cellular thiol-disulfide redox buffers (essentially glutathione-GSSG; ref. 13). Thiol-disulfide exchange could also probably proceed from intracellular glutathione to extracellular plasma membrane proteins in some conditions (20). The possible involvement of exofacial sulfhydryl groups in extracellular ferric reduction by iron-rich and iron-deprived cells was investigated using the membrane-impermeant Ellman's reagent (DTNB). The results for the T8A strain are given in Figure 6 (similar results were obtained with the strain G204 supplemented with ALA; data not shown). In contrast to iron-rich cells, iron-deprived cells responded to exogenous glucose by a sudden increase in surface thiol groups accessible to DTNB. The subsequent addition of ferric iron (Fe[III]-EDTA) resulted in a drop of these exposed thiol groups. This drop in surface thiol concentration was not due to their stoichiometric oxidation by Fe3", since no counterpart in ferrous iron generation was observed (as measured with the ferrous iron-trapping reagent ferrozine; data not shown). Iron may have catalyzed the autoxidation of the surface thiol groups, as do other metal ions (22). The amount of intracellular reduced thiols was higher in cells grown under iron limitation than in iron-rich cells. There is, thus, a difference in the redox status of the two types of cells. Addition of Fe(III)-EDTA to the resting cells resulted in a progressive decrease in the low mol wt reduced thiol pool in irondeprived cells only (Fig. 6). However, it seems unlikely that glutathione acts as a direct intracellular electron donor for extracellular ferric reduction, because reduction of Fe(III)EDTA by isolated plasma membranes was not stimulated by 1 mM glutathione, even in the presence of cytosolic extracts (data not shown).

Plasma Membrane Potential The ferri-reductase activity (Fe[III]-citrate) of whole cells was strongly inhibited (80-95%) by the plasma membrane H+-ATPase inhibitors vanadate (1 mM), diethylstilbestrol (50 AM), and SW26 (2,2,2-trichloroethyl 3,4-dichlorocarbanilate [0.2 mM]). It was also strongly inhibited (>95%) by permeabilizing agents such as 2.5% toluene-ethanol (1:4, v/v) and digitonin (0.1% w/v), and by the uncoupler carbonyl cyanide m-chlorophenyl hydrazone (70-80% inhibition at 0.5 ml). Moreover, the thermosensitive ATPase-deficient mutant strain RS373 (4, 9) showed similar rates of decrease in proton excretion and ferri-reductase activity (measured with either ferricyanide or Fe[III]-citrate; data not shown). These observations suggest that the glucose-stimulated redox process requires the integrity of the plasmalemma, with the ATPasegenerated proton gradient probably playing a role. As previously observed by van de Mortel et al. (29), adding glucose to resting cells was followed within a few minutes by hyperpolarization of the cells (Fig. 7). That glucose-in-


-r

T'

''i'n

775

dous increase in the cell ferri-reductase activity that occurs a few hours after transferring the cells to iron-deficient growth medium. This response results in the cells having an increased capacity for reductive iron uptake. The present work also shows that copper could play a physiological role in regulating cell ferri-reductase activity, confirming the previous observation of Funk and Schneider (10) that there are interactions between iron and copper metabolisms in S. cerevisiae. Our data suggest that a reductive mechanism could be involved in copper assimilation. Another major change that occurs when cells are transferred to iron-deficient culture medium is an increase in intracellular reduced glutathione and in the increased capacity of the cells to expose sulfhydryl groups at the exofacial surface. Reglinski et al. (20) showed that in erythrocytes, there is a link between the intracellular glutathione pool and the exofacial sulfhydryl groups and that an extracellular oxidative stimulus can deplete intracellular glutathione and change the redox balance of the cells via oxidation of the surface thiols. Similar mechanisms of oxidative stress may

rv1

0O.90

0.6I

I

zn

03-

105

0.0Fel Fe2 0

u

0

Cu I

z

Cr

LUI u (A LU

0= 70-

z I5

10 TIME (min)

15

Figure 6. The cell surface-associated thiols (exofacial thiols: 0, 0) and the intracellular low mol wt thiols (O, *) in iron-deficient (open symbols) and iron-rich resting cells (closed symbols) of the strain T8A. Experimental conditions as in Figure 5 (results are means ± SE from three experiments).

duced polarization was increased by adding Zn2+ to the growth medium and was decreased by adding Cu2" or Co"; Fe3" had no effect (Fig. 7). These results suggest that some cations, but not iron, could act on cell ferri-reductase activity by modifying the cell plasma membrane polarizability. Addition of ferric iron to suspensions of iron-deprived or ironrich cells resulted in a rapid initial depolarization followed, in iron-rich cells only, by a phase of slow depolarization (Fig. 7). Some mechanism, perhaps increased efflux of proton, should, therefore, prevent extensive plasma membrane depolarization in iron-deprived cells during trans-plasma membrane electron transfer to extracellular Fe3". DISCUSSION Iron deprivation induces a series of adaptive responses in yeast. One of the most spectacular changes is in the tremen-

-i J

(

LU

L, LU

0 t//

Cu2

uL- 35-

0

Cu 2

6

LU

9

aI

~~~~~LU

!3 0 /-~IM 1 TmEn

40 80 Ii20 FERRI-REDUCTASE ACTIVITY (nmol h-' mg-,)

Figure 7. Relationship between the level of glucose-induced polarization of cells grown in the presence of different cations and cell ferri-reductase activity. Cells of the strain FL200 were grown for 6 h in iron-deficient medium without addition (0) or supplemented with 1 $M ZnSO4 (Zn), 200 AM CrCI3 (Cr), 1 Mm Fe(IlIl)-citrate (Fe 1), 100 AM Fe(IIl)-citrate (Fe 2), 0.1 MM CuSO4 (CU 1), 10 MM CuSO4 (CU 2), 20 Mm CuSO4 (Cu 3), or 200 AM CoCI2 (Co). Cell ferri-reductase activity was determined and the glucose-induced polarization of cells was recorded as the maximum increase in DMP fluorescence emission when glucose was added to the cell suspensions. Inset, Fluorescence emission of the reagent DMP added to resting suspensions of cells grown in iron-deficient (a) or in iron-rich medium (b). First arrow, Addition of glucose (5% w/v); second arrow, addition of Fe(IlIl)-citrate (360 AM).

776

LESUISSE AND LABBE

account for the changes in intracellular glutathione, surface thiols, and intracellular NADH that occur in yeast cells upon addition of ferric ions to the resting cells. However, the nature of the electron source and of the metabolic pathways involved in extracellular ferric reduction remains unknown. The increased capacity of iron-deficient cells to reduce extracellular electron acceptors should obviously not be considered simply as the result of an increase in the intracellular level of NAD(P)H, but rather in terms of a flux of reducing power. The results of Yamashoji and Kajimoto (31), who observed oxidation of NADH when ferricyanide was added to resting cells, do not necessarily imply that NADH acts as the direct electron source for the reduction process. As shown in the present work, such a drop in NADH concentration occurs in all types of cells, even those with a deficient ferri-reductase activity. Our observations on the ferri-reductase activity of isolated plasma membranes suggest that NADPH is the best candidate as the electron source for extracellular ferric iron reduction. It should be pointed out, however, that the in vitro measurements are difficult to interpret because several different reductase activities seem to be associated with the isolated plasma membranes, and it is not known which of these activities is (are) responsible for the reduction of iron in vivo. No relationship was found between the induction of one of the NADPH-generating pathways and the induction of the ferri-reductase activity. Our results also indicate that it is no longer possible to assume that the hexose monophosphate pathway is the only source of reducing equivalents for the iron reduction process. The importance of the hexose monophosphate pathway as the main source of NADPH in yeast has recently been questioned (26). It should also be stressed that the level of NADPH in T8A strain (deficient in glucose6-phosphate dehydrogenase activity) was comparable to that in a wild strain, and that the T8A strain exhibited no growth deficiency. The relative importance of the cytosolic NADP+linked isocitrate dehydrogenase(s) in generating NADPH remains essentially unknown (2) and could be generally underestimated. Last, the possible role of malic enzyme(s) or nucleotide kinase(s) in generating NADPH in yeast is poorly documented. The trans-plasma membrane redox system described here differs from that described in plants in several aspects (no apparent involvement of intracellular citrate [24] and malate [27], no substantial difference in the transmembrane electrical potential of iron-rich and iron-deprived cells [32]). Ferric iron reduction by plants is accompanied by an increased efflux of protons that depends on the activation of the ATP-driven plasmalemma proton pump [27]). This could also be the case in yeast, since inhibitors of the plasmalemma H+-ATPase strongly inhibit the reduction process. The direct involvement of the plasmalemma H+-ATPase in the reduction process by yeast cells could account for the lower depolarizing effect of ferric ions on iron-deprived cells than on iron-rich cells. The ATP-driven proton pump would be activated by both the initial depolarization and the decrease in intracellular pH that is expected to occur during the transfer of electrons across the plasmalemma (23). This electron transfer would be faster in iron-deprived cells, mainly because of the increased cata-


lytic capacity of the membrane-bound electron transfer system in those cells. Our results suggest that far from being simply related to the iron status of the cells, the plasma membrane ferrireductase activity in S. cerevisiae can be controlled by a number of factors, several of which could act by modifying the redox status of the cells. Copper is suspected to specifically induce several enzymes (glutathione peroxidase, catalase, Zn,Cu-superoxide dismutase) involved in the defense against oxidative stress in yeast (11, 12). Our data show that the effects of Fe3" addition (as Fe[III]-EDTA) on a suspension of iron-deprived cells are essentially those expected from an oxidative stress. Conversely, cells not exposed to iron and copper during growth and protected from oxidative conditions (excess oxygen, menadione) have a higher capacity to transfer reducing equivalents from the intracellular to the extracellular compartment, either via direct electron transfer to an extracellular acceptor or via exposure of exofacial sulfhydryl groups. We therefore suggest that the plasma membrane reductase not only is involved in reductive iron uptake (6, 16, 17), but also plays a more general role in adapting the redox potential of the cells to the growth conditions, since its activity results in a modification of the redox status of the cell, whereas its expression itself is subject to regulation via intra- or extracellular redox stimuli. ACKNOWLEDGMENTS We thank Professor P.K. Maitra and Dr. F. Portillo for the gift of strains T8A and RS373, respectively, and Dr. C.O. Parkes for help in preparing the manuscript.

LITERATURE CITED 1. Bienfait F (1987) Biochemical basis of iron efficiency reactions in plants. In G Winkelmann, D van der Helm, JB Neilands, eds, Iron Transport in Microbes, Plants and Animals. VCH, Weinheim, pp 339-352 2. Bruinenberg PM (1986) The NADP(H) redox couple in yeast metabolism. Antonie Leeuwenhoek 52: 411-429 3. Buckhout TJ, Bell PF, Luster DG, Chaney RL (1989) Ironstress induced redox activity in tomato (Lycopersicon esculentum Mill) is localized on the plasma membrane. Plant Physiol 90: 151-156 4. Cid A, Serrano R (1988) Mutations of the yeast plasma membrane H+-ATPase which cause thermosensitivity and altered regulation of the enzyme. J Biol Chem 263: 14134-14139 5. Crane FL, Roberts H, Linnane AW, Low H (1982) Transmembrane ferricyanide reduction by cells of the yeast Saccharomyces cerevisiae. J Bioenerg Biomembr 14: 191-205 6. Dancis A, Klausner RD, Hinnebusch AG, Barriocanal JG (1990) Genetic evidence that ferric reductase is required for iron uptake in Saccharomyces cerevisiae. Mol Cell Biol 10: 2294-2301 7. Dufour JP, Amory A, Goffeau A (1989) Plasma membrane ATPase from the yeast Schizosaccharomyces pombe. Methods Enzymol 157: 513-538 8. Emery T (1987) Reductive mechanisms in iron assimilation. In G Winkelmann, D van der Helm, JB Neilands, eds, Iron Transport in Microbes, Plants and Animals. VCH, Weinheim, pp 235-250 9. Famaey JP, Whitehouse MW (1975) Interaction between nonsteroidal anti inflammatory drugs and biological membranes Part 4: effects of nonsteroidal anti inflammatory drugs and of various ions on the availability of sulfhydryl groups on lymphoid cells and mitochondrial membranes. Biochem Pharmacol 24: 1609-1615

TRANS-PLASMA MEMBRANE ELECTRON TRANSFER IN SACCHAROMYCES CEREVISIAE 10. Funk F, Schneider W (1989) The dichotomy of mechanisms of copper accumulation in yeast induced by enhanced levels of copper as well menadione in growth media. Chem Speciation Bioavailability 1: 131-138 11. Galiazzo F, Schiesser A, Rotilio G (1987) Glutathione peroxidase in yeast. Presence of the enzyme and induction by oxidative conditions. Biochem Biophys Res Commun 147: 1200-1205 12. Galiazzo F, Schiesser A, Rotilio G (1988) Oxygen-independent induction of enzyme activities related to oxygen metabolism in yeast by copper. Biochim Biophys Acta 965: 46-51 13. Gilbert H F (1990) Molecular and cellular acpects of thiol disulfide exchange. Adv Enzymol 63: 69-172 14. Klingenberg M (1974) Nicotinamide adenine dinucleotides (NAD, NADP, NADH, NADPH). Spectrophotometric and fluorimetric methods. In Bergmeyer HU, ed, Methods of Enzymatic Analysis. Academic Press, New York, pp 2045-2059 15. Lesuisse E, Crichton RR, Labbe P (1990) Iron-reductases in the yeast Saccharomyces cerevisiae. Biochim Biophys Acta 1038: 253-259 16. Lesuisse E, Labbe P (1989) Reductive and non-reductive mechanisms of iron assimilation by the yeast Saccharomyces cerevisiae. J Gen Microbiol 135: 257-263 17. Lesuisse E, Raguzzi F, Crichton RR (1987) Iron uptake by the yeast Saccharomyces cerevisiae: involvement of a reduction step. J Gen Microbiol 133: 3229-3236 18. Lobo Z, Maitra PK (1982) Pentose phosphate pathway mutants of yeast. Mol Gen Genet 185: 367-368 19. Portillo F, Mazon MJ (1986) The Saccharomyces cerevisiae start mutant carrying the cdc 25 mutation is defective in activation of plasma membrane ATPase by glucose. J Bacteriol 168: 1254-1257 20. Reglinski J, Hoey S, Smith WE, Sturrock RD (1988) Cellular response to oxidative stress at sulfhydryl group receptor sites on the erythrocyte membrane. J Biol Chem 263: 12360-12366 21. Romheld V (1987) Existence of two different strategies for the acquisition of iron in higher plants. In G Winkelmann, D van

22.

23. 24. 25. 26.

27.

28.

29.

30. 31.

32.

777

der Helm, JB Neilands, eds, Iron Transport in Microbes, Plants and Animals. VCH, Weinheim, pp 353-374 Russo A (1988) Determination and quantitation of biological sulfhydryls. Methods Biochem Anal 33: 165-241 Serrano R (1988) Structure and function of proton translocating ATPase in plasma membranes of plants and fungi. Biochim Biophys Acta 947: 1-28 Sijmons PC, Bienfait HF (1986) Development of Fe3" reduction activity and H' extrusion during growth of iron-deficient bean plants in a rhizostat. Biochem Physiol Pflanzen 181: 283-299 Sijmons PC, van den Briel W, Bienfait HF (1984) Cytosolic NADPH is the electron donor for extracellular Fe"' reduction in iron-deficient bean plants. Plant Physiol 75: 219-221 Thomas D, Cherest H, Surdin-Kerjan Y (1991) Identification of the structural gene for glucose-6-phosphate dehydrogenase in yeast. Inactivation leads to a nutritional requirement for organic sulfur. EMBO J 10: 547-553 Trockner V, Marre E (1988) Plasmalemma redox chain and H+ extrusion.II. Respiratory and metabolic changes associated with fusicoccn-induced and with ferricyanide-induced H+ extrusion. Plant Physiol 87: 30-35 Urban-Grimal D, Labbe-Bois R (1981) Genetic and biochemical characterization of mutants of Saccharomyces cerevisiae blocked in six different steps of heme biosynthesis. Mol Gen Genet 183: 85-92 van de Mortel JBJ, Mulders D, Korthout H, Theuvenet APR, Borst-Pauwels GWFH (1988) Transient hyperpolarization of yeast by glucose and ethanol. Biochim Biophys Acta 936: 421-428 Yagi K (1971) Simultaneous microdetermination of riboflavin, FMN and FAD in animal tissues. Methods Enzymol 18: 290-296 Yamashoji S, Kajimoto G (1986) Decrease of NADH in yeast cells by external ferricyanide. Biochim Biophys Acta 852: 25-29 Zocchi G, Cocucci S (1990) Fe-uptake mechanism in Fe-efficient cucumber roots. Plant Physiol 92: 908-911