Báo cáo Y học: Intracellular pH homeostasis in the filamentous fungus Aspergillus niger

doi:10.1046/j.1432-1033.2002.03042.x

Eur. J. Biochem. 269, 3485–3494 (2002) (cid:1) FEBS 2002

Intracellular pH homeostasis in the filamentous fungus Aspergillusniger

Stephan J. A. Hesse1, George J. G. Ruijter2, Cor Dijkema1 and Jaap Visser2,† 1Department of Biophysics, Wageningen University, the Netherlands; 2Department of Microbiology, section Fungal Genomics, Wageningen University, the Netherlands

cyanide m-chlorophenylhydrazone, CCCP) to the perfusion buffer led to decreased ATP levels, increased respiration and a partial (1 lM CCCP), transient (2 lM CCCP) or perma- nent (10 lM CCCP) collapse of the vacuolar membrane –, DpH. Nonlethal levels of the metabolic inhibitor azide (N3 0.1 mM) caused a transient decrease in pHcyt that was closely paralleled by a transient vacuolar acidification. Vacuolar H+ influx in response to cytoplasmic acidification, also observed during extreme medium acidification, indicates a role in pH homeostasis for this organelle. Finally, 31P NMR spectra of citric acid producing A. niger mycelium showed that despite a combination of low pHex (1.8) and a high acid- secreting capacity, pHcyt and pHvac values were still well maintained (pH 7.5 and 6.4, respectively).

Keywords: Aspergillus niger; intracellular pH; pH homeo- stasis; 31P NMR; perfusion.

Intracellular pH homeostasis in the filamentous fungus Aspergillus niger was measured in real time by 31P NMR during perfusion in the NMR tube of fungal biomass immobilized in Ca2+-alginate beads. The fungus maintained constant cytoplasmic pH (pHcyt) and vacuolar pH (pHvac) values of 7.6 and 6.2, respectively, when the extracellular pH (pHex) was varied between 1.5 and 7.0 in the presence of citrate. Intracellular metabolism did not collapse until a DpH over the cytoplasmic membrane of 6.6–6.7 was reached (pHex 0.7–0.8). Maintenance of these large pH differences was possible without increased respiration compared to pHex 5.8. Perfusion in the presence of various hexoses and pentoses (pHex 5.8) revealed that the magnitude of DpH values over the cytoplasmic and vacuolar membrane could be linked to the carbon catabolite repressing properties of the carbon source. Also, larger DpH values coincided with a higher degree of respiration and increased accumulation of (carbonyl polyphosphate. Addition of protonophore

During operation of cellular metabolism under aerobic conditions net intracellular production of protons takes place mainly by formation of tricarboxylic acid cycle acids, CO2/H2CO3 and protein synthesis [1]. Consequently, tight control of proton fluxes, in combination with the ability to maintain pH gradients across cellular membranes, is a crucial aspect of cellular energetics. Large deviations of cytoplasmic pH (pHcyt) need to be avoided to keep control of fundamental intracellular processes which are sensitive to pH, such as DNA transcription, protein synthesis and enzyme activities [2]. In order to ensure optimal activity of major metabolic pathways, constant removal of free protons from the cytoplasm is required [3]. In lower eukaryotes and plants this process is mediated through the action of the plasma membrane P-ATPase at the expense of ATP hydrolysis, which results in pH and electrical potential differences across the plasma membrane. The P-ATPase is involved in intracellular pH (pHin) regulation, maintenance of a proper ion balance and generation of the electrochemi- cal proton gradient (proton motive force, Dp) across the cytoplasmic membrane which drives an array of secondary

transport systems [4]. In Saccharomyces cerevisiae, intracel- lular pH is thought to be additionally regulated through the action of alkali-cation/H+ antiporters, such as the Nha1 antiporter with a H+/K+ (Na+) exchange mechanism [5]. Intracellular pH homeostasis in the filamentous fungus Neurospora crassa has been suggested to be achieved by parallel operation of the H+-extruding P-ATPase and a high-affinity proton symport uptake system for K+, yielding a net 1 : 1 exchange of K+ for cytoplasmic H+ [6]. Other major ATPases in fungal cells are the vacuolar membrane V-ATPase and the mitochondrial membrane F1F0-ATPase. The action of the former, ATP-dependent transport of protons into the vacuole, is thought to contribute to cytoplasmic pH homeostasis as well [7]. The resulting electrochemical proton potential is able to drive amino acid and ion transport across the vacuolar mem- brane, probably through proton antiport systems. The F1F0-ATPase uses the proton motive force generated by the electron transport chain across the inner mitochondrial membrane to drive phosphorylation of ADP to ATP. Consequently, intracellular pH and pH gradients are directly linked to cellular energy levels and metabolism [8]. The response of intracellular pH values to different conditions, especially variation in extracellular pH (pHex), reveals clues to mechanisms of pH regulation [9]. In a previous study we reported on a system based on long-term acquisition of 31P NMR spectra of constantly perfused and well-oxygenated immobilized mycelium for the determin- ation of compartmental pH values in the filamentous fungus Aspergillus niger [10]. A. niger is industrially important for

Correspondence to S. J. A. Hesse, Department of Biophysics, Wageningen University, Dreijenlaan 3, 6703 HA Wageningen, the Netherlands. Fax: +31 317 484 011, Tel.: +31 317 484 692, E-mail: stephan.hesse@algemeen.mgim.wau.nl Abbreviations: CCCP, chlorophenylhydrazone. (cid:1)Present address: Postbus 396, 6700 AJ Wageningen, the Netherlands. (Received 12 April 2002, accepted 30 May 2002)

3486 S. J. A. Hesse et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

0.2 gÆL)1 KH2PO4,

was filled with 125 mL immobilized conidia and 375 mL medium optimized for citric acid production (140 gÆL)1 1.25 gÆL)1 decationized glucose, (NH4)2SO4, 0.25 gÆL)1 MgSO4Æ7H2O, 1.3 mgÆL)1 ZnSO4Æ 7H2O, 6.5 mgÆL)1 FeSO4Æ7H2O). Under these conditions no growth of biomass outside of the beads occurred. The culture pH was not regulated and was initially set at 3.5. Cultures were sparged with 1.5 LÆmin)1 air, and after 24 h 15 lL of a solution containing 30% (v/v) polypropylene- glycol in alcohol was added to the reactor to prevent excessive foaming. The reactors were run for 2 or 7 days at 30 (cid:2)C. The fermentation volume was periodically adjusted to 0.5 L with double-distilled water.

Perfusion conditions

large-scale production of organic acids (e.g. citric acid) due to a high intrinsic excretion capacity for this compound [11]. Despite this capacity, quantitative analysis of metabolism using kinetic models and metabolic engineering, comple- mentary to traditional strain improvement, is still a promising approach to increase citric acid production or to shorten fermentation times [12]. However, a more predictive accuracy for kinetic (mechanistic) models requires a more detailed description of the conditions under which the enzymes involved operate in vivo. With the perfusion system developed, problems related to fungal morphology during long-term in vivo NMR measurements have been overcome. With this method it is now possible to obtain more information on intracellular pH, one of the key parameters affecting enzyme activity, and its homeostasis. Also, the ability of A. niger to acidify its medium to pH values below 2.0 during production of large quantities of organic acids implies that a very efficient pH-homeostatic system exists in these cells. As protein synthesis and intracellular enzyme activities are sensitive to pH, mainten- ance of intracellular pH under extreme conditions (especi- ally low pHex) is crucial to ensure optimal cellular activity during (industrial) fermentations. So far, however, data on intracellular pH in filamentous fungi have been scarce. A few reports have dealt with cytoplasmic pH of citric acid- producing A. niger mycelium [13,14]. More detailed know- ledge about intracellular pH homeostasis in filamentous fungi is to date only available for N. crassa [1,15,16]. To investigate the ability of A. niger to maintain cellular energy levels, cells were subjected to several stresses like low pHex, citric acid producing conditions, increased proton permeab- ility by an uncoupler and inhibition of ATP synthesis by sodium azide. Also tested was the effect of different carbon sources on steady-state DpH values. Together our results provide reliable data on intracellular pH under various extracellular conditions, and demonstrate the ability of the fungus to maintain cellular energetics under extreme conditions with a formidable tolerance towards extracellular acidity.

Immobilized biomass from shake-flask cultures was har- vested, washed with a buffer (30 (cid:2)C) containing 25 mM sodium citrate pH 5.8, 0.25 gÆL)1 NH4NO3, 0.2 gÆL)1 KCl, 0.2 gÆL)1 MgSO4Æ7H2O, 0.2 mM KH2PO4, 0.3 mM CaCl2, and perfused within the NMR tube with 1 L of the same buffer saturated with oxygen. The extracellular pH was varied by using perfusion buffer with pH values ranging from 1.0 to 7.0 or by direct titration of the buffer reservoir with 2 M HCl. The same buffer supplemented with 50 mM Tris was used under alkaline extracellular conditions (pHex, 7.0–9.0). The effect of the presence of various sugars on intracellular pH values was tested after a 2-h transfer of immobilized biomass to 150 mL of perfusion buffer pH 5.8 supplemented with 10 mM of sugar (D-glucose, D-fructose, D-xylose or L-arabinose). Subsequently, the beads were perfused with the same buffer saturated with oxygen for 3 h. Carbonyl cyanide m-chlorophenylhydrazone (CCCP) and sodium azide (NaN3) were directly added to the buffer reservoir (pH 5.8) from a 10-mM stock solution in ethanol and a 10% stock solution in water, respectively. Oxygen consumption during perfusion (DO2) was expressed as the percentage of oxygen removed from an oxygen saturated buffer after passage through the immobilized cell plug.

E X P E R I M E N T A L P R O C E D U R E S

Strain, immobilization of conidia and culture conditions

Immobilized citric acid producing mycelium from 2- and 7-day-old fermentations was directly transferred from the bubble column reactor to the NMR tube and perfused with filtered culture medium (500 mL) from 2-day- and 7-day- old fermentations, respectively. In all cases, a 4-cm high plug of immobilized biomass (± 12.5 mL beads) was perfused at a rate of 15 mLÆmin)1.

31P NMR spectroscopy 31P NMR spectra were recorded at 121.5 MHz at 30 (cid:2)C on a AMX300 wide-bore spectrometer (Bruker, Germany), equipped with a 20-mm switchable 31P/13C probe tuned at the 31P nucleus, and collected in 15, 20 or 60-min blocks (4500, 5700 or 18000 FIDs, respectively) using acquisition parameters described previously [10]. Methylene diphos- phonic acid (0.2 M, pH 8.9), contained in an in situ capillary, was used as an internal reference, resonating at 16.92 p.p.m. relative to 85% H3PO4 (0 p.p.m).

Analyses

Condiospores of A. niger NW131 (cspA1 goxC17) lacking glucose oxidase activity [17] were propagated at 30 (cid:2)C on complete medium [18] solidified by 1.5% agar and contain- ing 1% D-glucose. Conidiospores were harvested with a solution containing 0.9% NaCl and 0.05% (v/v) Tween-80. Immobilization of conidia in Ca2+-alginate beads (diameter 1 mm) was performed as described previously [10]. A 2.5% solution of Manugel DJX (ISP Alginates, Tadworth, Surrey, UK) was used in all immobilization experiments. After harvesting and washing with demineralized water, immobilized conidia were cultured in 500-mL Erlenmeyer flasks. Immobilized mycelium was obtained by incubating the beads (10 g) for 40–44 h in a rotary shaker at 250 r.p.m and 30 (cid:2)C in 100 mL minimal medium (2 gÆL)1 NH4NO3, 1.5 gÆL)1 KH2PO4, 0.5 gÆL)1 KCl, 0.5 gÆL)1 MgSO4Æ7H2O pH 6.0), supplemented with 1.5% D-glucose, 0.02% (v/v) of a trace element solution [19] and 0.05% yeast extract.

To obtain immobilized mycelium under citric acid- producing conditions a 500-mL bubble column reactor

Cytoplasmic and vacuolar pH values were determined by comparing the pH-sensitive chemical shifts of cytoplasmic

Intracellular pH homeostasis in A. niger (Eur. J. Biochem. 269) 3487

(cid:1) FEBS 2002

R E S U L T S

The dependence of intracellular pH values on ambient pH

(Pcyt) and vacuolar inorganic phosphate (Pvac) with a calibration curve for inorganic phosphate (Pi). This curve has been referred to in many other NMR studies on both yeasts and fungi [16,20,21], and shows the pH dependence of the chemical shift of Pi measured in a medium made up to approximate concentrations of the major cationic compo- nents in yeast [22]. The perfusion buffer pH (pHex) was monitored by pH electrode measurements. Intracellular pH values in the presence of various carbon sources and in mycelium producing high levels of citric acid were deter- mined from 12–16 and 5–7 different spectra, respectively; final values represent mean values that were statistically analysed by a two-tailed Student’s t-test (a ¼ 5%). Relative increases in polyphosphate levels were determined from 31P NMR spectra by relating the integral of the polyphosphate peak to the integral of the internal reference peak. Citrate, ammonium and phosphate levels in culture filtrate samples of citric acid fermentations were determined as described before [17]. Perfusion buffer and transfer medium samples were analysed for sugars and organic acids by HPLC using an HPX-87H column (Bio-Rad) eluted with 25 mM HCl at 50 (cid:2)C with UV (210 nm) and refractive index detection, and for polyols using a Carbopac MA-1 column (Dionex) eluted with 480 mM NaOH at 20 (cid:2)C with pulsed amperometric detection. Dry weight determinations on immobilized biomass were carried out by dissolving the Ca2+-alginate beads in a 100-mM solution of the Ca2+-scavenger sodium hexametaphosphate (Fluka AG, Buchs, Switzerland). Mycelium was then collected by filtration, washed with demineralized water, frozen in liquid nitrogen, lyophilized and weighed.

A. niger has the ability to acidify its environment to values as low as pH 1.5 [17]. To investigate to what extent extracellular pH (pHex) affects intracellular pH (pHin) values, we determined pHcyt and pHvac as a function of ambient pH using 31P NMR as described previously [10]. Surprisingly, the cells were able to maintain pHcyt and pHvac at 7.6 and 6.2, respectively, when pHex was varied between 1.5 and 7.0, implying that a very steep DpH over the cytoplasmic membrane of 6.1 can be sustained by the cells (Fig. 1A). The DpH over the vacuolar membrane was maintained at a constant value of 1.4. At pHex 1.5, this gradient was about 0.1 pH unit larger due to a slightly more acidic vacuole. Further acidification to pHex 1.0 caused pHcyt to drop to pH 7.4, in parallel to an even larger vacuolar acidification than at pHex 1.5. A comparable observation was made at pHex 8.0: pHcyt was still reason- ably well regulated, whereas the vacuoles became more alkaline (pHvac 6.5). Interestingly, in the pHex range of 1.0– 7.0 in the presence of 25 mM citrate, cells consumed the same amount of oxygen once a steady-state was reached (DO2 ¼ 8–9%). Cellular metabolism collapsed and O2 consumption rapidly dropped to 0% when pHex reached a value of about 0.7–0.8. Cells only moderately increased their oxygen consumption (from 9 to 11%) just before the collapse. A representative 31P NMR spectrum of immobi- lized A. niger mycelium perfused in the presence of 25 mM citrate pH 1.5 and 0.2 mM Pi is shown in Fig. 1B.

Fig. 1. The dependence of A. niger NW131 cytoplasmic pH (pHcyt) and vacuolar pH (pHvac) on extracellular pH (pHex) in the presence of oxygen- saturated buffer (A) and typical 31P NMR spectrum of immobilized A. niger NW131 mycelium (B). (A) Buffer contained 25 mM citrate (pHex 1.0–7.0) or 25 mM citrate and 50 mM Tris (pHex 7.0–8.5). Values are the mean of two experiments. (B) Mycelium cultured for 42 h and perfused with oxygen-saturated buffer containing 25 mM citrate at pHex 1.5. Abbreviations and chemical shifts of the assignments: SME: sugar phospho- monoesters, 4.9 p.p.m.; SDE: sugar phosphodiesters, 4.5 p.p.m.; Pcyt: cytoplasmic inorganic phosphate, 2.9 p.p.m.; Pvac: vacuolar inorganic phosphate, 1.2 p.p.m.; Pex: extracellular inorganic phosphate, 0.6 p.p.m.; c-ATP: ) 4.9 p.p.m.; P2: pyrophosphate and terminal phosphate of polyphosphate, )5.8 p.p.m.; a-ATP: )9.9 p.p.m.; NAD(H): )10.6 p.p.m.; UDPG: uridine diphosphoglucose, )10.6 and )12.3 p.p.m.; P3 and P4: penultimate phosphates of polyphosphate, )17.7 and )19.7 p.p.m., respectively; a-ATP: )18.6 p.p.m.; Pn: polyphosphate, )22.5 p.p.m. Data were collected in a 60-min block. The internal reference is not shown.

3488 S. J. A. Hesse et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

membrane (Table 1). Younger mycelium (18–24 h old) contained hardly any polyphosphate (spectra not shown).

The effect of CCCP on intracellular pH

Different bioenergetic states with different sugars For the yeast Candida tropicalis it was demonstrated by 31P NMR that cells aerobically metabolizing glucose were more energized than xylose-fed cells [23]. Besides higher UDPG and polyphosphate levels and higher rates of Pi assimilation, cells metabolizing glucose had a slightly higher pHcyt and a slightly lower pHvac. To investigate to what extent different nutritional conditions result in different steady-state intracellular pH values in A. niger, mycelium was perfused for 3 h with perfusion buffer pH 5.8 containing one of the following sugars: L-arabi- nose, D-xylose, D-fructose or D-glucose. In this sequence these sugars represent poor to good carbon sources for A. niger. The initial oxygen consumption (DO2) of perfused biomass was 10–15% (using buffers saturated with O2). No steady oxygen consumption was reached for any of the sugars tested during 3 h of data acquisition. Instead, a gradual increase in oxygen consumption was observed. In the absence of sugar, citrate was the only available carbon source, and under these circumstances the initial oxygen consumption was lower and remained constant throughout the experiment (8–9%). Catabolism of glucose, fructose and xylose resulted in a more pronounced cytoplasmic alkalinization compared to arabinose or citrate only, whereas a clearly stronger vacuolar acidification was observed only in the presence of glucose and fructose (Table 1). Although the differences between the deter- mined pHin values were only small, an increased oxygen consumption coincided with higher pHcyt and lower pHvac values. As inferred from the 31P NMR spectra, however, no significant differences in ATP levels could be observed for the conditions tested (spectra not shown). Replacement of citrate by a 25 mM Mes buffer (pH 5.8) resulted in similar pH values in the case of glucose (results not shown), indicating that the effect of citrate on pHin is only minor. HPLC analysis of perfusion buffer samples showed that no polyols or organic acids were excreted during 3 h of perfusion, ensuring constant extracellular conditions during data acquisition. Besides generating ATP and establishing pH gradients, an alternative way for the cells to store energy generated by catabolism may be polyphos- phate synthesis. The relative increase in polyphosphate levels during 3 h of perfusion appeared to coincide with increased pH gradients over the cytoplasmic and vacuolar

A strong argument in favour of Mitchell’s chemiosmotic theory [24] was the fact that it could explain the mode of action of lipid-soluble weak acids that are able to carry out electrogenic proton transport across biological and artificial membranes, thereby dissipating both the electrical mem- brane potential (DY) and the proton gradient (DpH). These uncouplers abolish the tight coupling of electron transport to oxidative phosphorylation and allow respiration to proceed without control by phosphorylation. The decreased ability of the cytoplasmic and vacuolar membrane to maintain transmembrane proton gradients in vivo in the presence of increasing extracellular concentrations of the uncoupler CCCP is shown in Fig. 2. It should be noted that after transfer of the cells to the perfusion system, pHin values (especially pHvac) reached their steady-state values only after 1.5–2 h of perfusion. To shorten experimental times, CCCP (and azide, see below) were added before this time point, resulting in slightly different initial pHcyt and pHvac values in Figs 2 and 3. Addition of 1 lM CCCP to the perfusion buffer caused pHcyt to drop from 7.6 to 7.1 and pHvac to increase from 6.3 to 6.6 during the first 45 min after addition (Fig. 2A). At the same time, ATP and vacuolar phosphate levels decreased and cytoplasmic phosphate levels increased (spectra not shown). During this period oxygen consumption increased from approximately 9 to 21%. A complete collapse of the vacuolar membrane pH gradient did not occur. The cells started to recover 60– 75 min after the addition. At this point initial ATP levels were nearly restored. The absolute pHin values determined after recovery were slightly higher than initial values, in particular pHcyt. Consequently, the re-established pH gradient over the vacuolar membrane (1.5) was somewhat higher than before addition (1.3). In the presence of 2 lM CCCP essentially the same changes were observed as described for 1 lM CCCP. In this case, however, the vacuolar membrane pH gradient was completely dissipated (Fig. 2B). After 30 min no distinct Pcyt or Pvac could be recorded. Instead, both peaks had merged into one large intracellular phosphate resonance, corresponding to an intracellular pH value of 6.9. The cells reacted to the

Table 1. Steady-state pHin values, increase in oxygen consumption and relative increase in polyphosphate levels during 3 h of perfusion in the presence of oxygen-saturated buffer containing 25 mM citrate pH 5.8, supplemented with 10 mM sugar.

Increase in

Sugar [polyphosphate]b pHcyt pHvac O2 consumptiona

a Expressed as the percentage of oxygen removed from an oxygen-saturated buffer after passage through the immobilized cell plug. Initial oxygen consumption in the presence of sugar: 10–15%; in the presence of citrate only: 8–9%. b Relative increases in polyphosphate levels were determined from 31P NMR spectra by relating the integral of the polyphosphate peak to the integral of the internal reference peak at t ¼ 0 and t ¼ 3 h.

Glucose Fructose Xylose Arabinose — 7.76 ± 0.02 7.73 ± 0.02 7.72 ± 0.03 7.64 ± 0.02 7.58 ± 0.01 6.05 ± 0.04 6.08 ± 0.02 6.15 ± 0.04 6.19 ± 0.03 6.21 ± 0.02 24% 21% 17% 10% 0% 1.82 1.48 1.39 1.31 1.02

Intracellular pH homeostasis in A. niger (Eur. J. Biochem. 269) 3489

(cid:1) FEBS 2002

membrane than lethal doses of uncoupler. The residual pH gradient observed 2 h after azide addition (0.5 mM) was 0.3– 0.4 pH unit, which is considerably lower than when CCCP was used (0.9 pH unit). Eventually, pHin became equal to

imposed stress condition once more by increasing their oxygen consumption from 9 to 30% within 60 min of addition. Surprisingly, 90 min after CCCP was added the large intracellular phosphate resonance split up again into two separate resonances: Pcyt, shifting to the left (indicating alkalinization within that compartment) and Pvac, shifting to the right (indicating acidification within that compart- ment). Again, the onset of pHin recovery coincided with the (partial) restoration of original ATP levels (spectra not shown). With 10 lM CCCP, an irreversible collapse of the vacuolar membrane pH gradient was observed within 15 min (Fig. 2C). The intracellular pH, deduced from the chemical shift of a large intracellular phosphate resonance, became 6.7 and remained around that value during another 105 min of perfusion. The oxygen consumption increased during the first 5 min following the CCCP addition, but then rapidly dropped to 0%, indicating cell death. Con- comitantly, a rapid loss of ATP was observed, whereas UDPG, cofactor and polyphosphate levels gradually decreased during the 120 min that spectra were acquired. A complete collapse of the residual DpH across the cytoplasmic membrane (approximately 0.9 pH unit) did not occur, even when CCCP levels were doubled to 20 lM.

The effect of azide on intracellular pH

The inhibitory mode of action of azide on cellular ATP synthesis is twofold by inhibiting both the mitochondrial F1F0-ATPase and cytochrome-c-oxidase in the terminal part of the electron transport chain. The effect of transient and permanent depletion of ATP due to azide addition on pHin is –, Fig. 3A) a shown in Fig. 3. In the former case (0.1 mM N3 transient decrease in pHcyt was observed with a minimal value of 7.0 after 30 min. In contrast with CCCP, the drop in pHcyt was paralleled by a decrease in pHvac from 6.3 to 5.9. Spectra showed that ATP levels dropped sharply, whereas both cytoplasmic and vacuolar phosphate levels were slightly higher compared to the original situation (Fig. 4A and B). At the same time, cells increased their oxygen consumption from 9 to 23%. Interestingly, the observed rise in both pHcyt and pHvac after 45 min was accompanied by a small increase in ATP level (Fig. 4C). In the new steady-state, pHin values were identical to those observed before the addition of the inhibitor, although ATP and UDPG levels were somewhat lower (Fig. 4D). Polyphosphate levels remained unchanged during 2.5 h of perfusion. Cells lost all of their ATP permanently when 0.5 mM azide was used. Moreover, an effective and permanent dissipation of the vacuolar mem- brane pH gradient was observed (Fig. 3B), visualized in the spectra as one large intracellular phosphate resonance (data not shown). Lethal azide concentrations were much more effective in dissipating the pH gradient across the cytoplasmic

Fig. 2. The effect of CCCP addition on cytoplasmic pH (pHcyt) and vacuolar pH (pHvac) in immobilized A. niger NW131 mycelium, grown for 42 h and perfused with oxygen-saturated buffer containing 25 mM citrate pH 5.8. CCCP was added at t ¼ 0, and the applied concen- trations were 1 lM (A), 2 lM (B) and 10 lM (C). Addition occurred before pHcyt and pHvac reached steady-state values, resulting in slightly different initial pH values (see text). Data points represent the mean of two experiments.

3490 S. J. A. Hesse et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

Fig. 4. 31P NMR spectra of immobilized A. niger NW131 mycelium in the presence of 25 mM citrate pH 5.8, showing the effect of addition of 0.1 mM sodium azide. (A) The situation before addition. Pcyt, cyto- plasmic inorganic phosphate resonance; Pvac, vacuolar inorganic phosphate resonance. The internal reference is not shown. (B) After 30 min both Pcyt and Pvac had moved to a lower chemical shift (i.e. compartmental acidification), accompanied by a sharp decrease in intensities of the c-ATP and the a-ATP peaks as indicated by the arrows. (C) After 45 min ATP levels started to restore again, and Pcyt and Pvac moved to a higher chemical shift again (i.e. compartmental alkalinization). (D) In the new steady-state (t ¼ 120 min) compart- mental pH values were identical to those observed before addition. See Fig. 3A for corresponding pH values.

pHex after 8 h of perfusion. During this whole period only a very small breakdown of polyphosphate could be observed.

The bioenergetic state of citric acid producing mycelium

+ and PO4

A. niger has the capacity to produce high levels of citric acid from hexoses and disaccharides in traditional citric acid producing processes when two important criteria are met [25]: a low pH (< 2) and absence of manganese ions

(Mn2+). Low pH is necessary to avoid production of gluconic acid and oxalic acid. In our experiments interfer- ence by gluconic acid production was prevented by using an A. niger N400 derivative, strain NW131, lacking glucose oxidase activity. The amount of citric acid accumulated from glucose in 7 days by immobilized biomass in a bubble column reactor was approximately 50 gÆL)1. In a typical fermentation the pH dropped from 3.5 to 1.8 in (cid:2) 3 days, and remained around that value. The cells consumed all 3– during the first 24 h (data not shown), NH4 and no citrate was produced in this period. Dry weight determinations of 2- and 7-day-old cultures indicated a small decrease in biomass content during the fermentation (17.4 and 16.1 gÆL)1, respectively). It was decided to

Fig. 3. The effect of sodium azide addition on cytoplasmic pH (pHcyt) and vacuolar pH (pHvac) in immobilized A. niger NW131 mycelium, grown for 42 h and perfused with oxygen-saturated buffer containing 25 mM citrate pH 5.8. Azide was added at t ¼ 0, and the applied concentrations were 0.1 mM (A) and 0.5 mM (B). Addition occurred before pHcyt and pHvac reached steady-state values, resulting in slightly different initial values. Data points represent the mean of two experiments.

Intracellular pH homeostasis in A. niger (Eur. J. Biochem. 269) 3491

(cid:1) FEBS 2002

function as sequestrator or donor of protons), its exact physiological role is not yet fully understood. Our results showed that with increased cellular energy levels polyphos- phate accumulated to higher levels, suggesting a role in cellular energy storage for the polymer. Recent studies on E. coli cells revealed a more regulatory role for polyphos- phate [26]. Cells deficient in polyphosphate were unable to express many genes that are needed for adaptation to deficiencies and environmental stresses during the stationary phase, and lost their viability relatively quickly. Increased polyphosphate synthesis may therefore greatly enhance the chances of survival in stationary phase cells. If so, it is crucial for cells to accumulate as much polyphosphate as possible in times of energy excess. Our results are in agreement with this hypothesis as polyphosphate accumu- lation was maximal in the presence of glucose at pHex 5.8. In the presence of a poor carbon source at the same pHex (citrate), no accumulation of polyphosphate was observed. In the presence of glucose at pHex 1.0, no increase in polyphosphate levels occurred either (results not shown). A key role for polyphosphate in stationary phase cells is further corroborated by the fact that polyphosphate was practically absent in younger (18–24-h-old) mycelium. In N. crassa, the ratio of polyphosphate to orthophosphate in vacuoles increased from 2.4 in early log phase cells to 13.5 in stationary phase cells [15]. When early log phase cells were exposed to a hypo-osmotic shock, both pHcyt and pHvac increased and cells lost 95% of their total polyphos- phate content. In contrast, hypo-osmotic shock of station- ary phase cells did not cause any changes in intracellular pH or polyphosphate levels, showing that these cells were much more effective in handling osmotic stress.

investigate a younger (2-day old) and an older (7-day old) culture. To mimic citric acid production conditions as closely as possible, immobilized cell mass was perfused with filtered medium of 2-day- and 7-day-old cultures, respect- ively, saturated with oxygen. The specific citric acid production rates of mycelium after transfer to the perfusion set up were 0.20 gÆLh)1 (2-day-old mycelium) and 0.32 gÆLh)1 (7-day-old mycelium), whereas in the bioreactor values of 0.37 gÆLh)1 (2-day-old mycelium) and 0.19 gÆLh)1 (7-day-old mycelium) were obtained. The values obtained for perfused mycelium confirmed that spectra were acquired under true citric acid producing conditions, although transfer to the perfusion system appeared to have an effect upon specific production rates. 31P NMR spectra of 2- and 7-day-old-citric acid-producing mycelium showed a remark- able constancy with respect to metabolite levels (including polyphosphate) and pHin values during the first 4–6 h of perfusion (spectra not shown). For correct assignment of Pcyt and Pvac, CCCP (2 lM) was used to partially collapse the DpH between these two compartments. No difference could be detected between steady-state pHcyt values of 2- and 7-day-old mycelium (7.53 ± 0.05 and 7.54 ± 0.04, respectively) although cytoplasmic phosphate, sugar phos- phate and ATP levels were clearly higher in 2-day-old mycelium. Compared to pHcyt values obtained in the presence of various carbon sources (Table 1) or under extreme extracellular acidity (Fig. 1A, pHex 1.5–2.0), citric acid-producing mycelium appeared to have only a slightly more acidic cytoplasm. The vacuoles of 2-day-old mycelium were relatively alkaline (pH 6.41 ± 0.03), whereas an even higher pHvac was found in vacuoles of 7-day-old mycelium (pH 6.50 ± 0.04).

D I S C U S S I O N

The proper functioning of cells relies on maintenance of their intracellular pH within relatively narrow limits, as large deviations of pH from normal values would be severely inhibitory to metabolism based on pH optima of cytoplasmic enzymes [1]. Our results show that A. niger is indeed capable of tightly maintaining its intracellular pH values within a narrow range. In the presence of various carbon sources at pHex 5.8, pHcyt varied only from 7.58 (citrate) to 7.76 (glucose), whereas pHvac ranged from 6.05 (glucose) to 6.19 (citrate) (Table 1). As no significant differences in ATP levels could be observed in the 31P spectra, an energetically more favourable carbon source may lead to a larger availability of ATP to the P-ATPase and V-ATPase due to increased ATP turnover, reflected by a higher oxygen consumption and larger pH gradients. Of the carbon sources tested, glucose has the largest carbon catabolite repressing effect in A. niger, and citrate the (glucose > fructose > xylose > arabinose > cit- lowest rate; G. J. G. Ruijter, Wageningen, the Netherlands, personal communication). Thus, the carbon catabolite repressing capacity of the carbon source could be linked to steady-state DpH values.

Interestingly, larger pH gradients and a higher oxygen consumption coincided with increased polyphosphate syn- thesis in A. niger (Table 1). Polyphosphate has been suggested to function as a cellular phosphate or high-energy reserve. Although polyphosphate may have additional functions (e.g. chelation of cations, or a pH-homeostatic

A striking observation was the ability of A. niger to maintain constant intracellular pH values during extracel- lular acidification to pH values as low as 1.0 without having to change its steady-state oxygen consumption. In yeast, filamentous fungi and higher plant cells, the proton pumping activity of the plasma membrane P-ATPase has been recognized as the major factor responsible for pHcyt homeostasis [27–29], possibly in combination with high affinity potassium uptake in symport with protons. Oper- ating in parallel in N. crassa, these two systems yield a net 1 : 1 exchange of K+ for cytoplasmic H+ [6]. In N. crassa, changes of pHex between 3.9 and 9.3 affect pHcyt linearly with a slope of approximately 0.1 unit pHcyt per unit pHex [1]. In S. cerevisiae, both pHcyt and pHvac became more acidic at pHex 3.5 compared with pHex 6.5 whether glucose was present or not [30]. Intracellular pH homeostasis in respiring Escherichia coli cells was good (pHcyt 7.6 ± 0.2) over a pHex range of about 5.5–9.0 [31]. Finally, in sycamore (Acer pseudoplatanus L) cells pHcyt and pHvac values were maintained when pHex was varied from 4.5 to 7.5 [29]. Oxygen consumption measurements of these cells in a perfusion setup revealed a progressive acceleration of the rate of O2 consumption towards the uncoupled O2 uptake rate (+ 1 lM FCCP) as pHex decreased from 6.5 to 4.5. Our results clearly show that after addition of CCCP (2 lM, pHex 5.8) a much higher oxygen consumption could be achieved (DO2 ¼ 25–30%) than the steady-state oxygen consumption (DO2 ¼ 8–9%) observed in the presence of citrate only (25 mM, pHex 1.0–7.0). Apparently A. niger can maintain its intracellular pH while keeping its oxygen from the uncoupled value. An obvious uptake far

3492 S. J. A. Hesse et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

two acidophilic eukaryotes:

possibility to generate and sustain a positive-inside DY. Indications for such a mechanism have been reported the alga Dunaliella for acidophila and the yeast Metschnikowia reukaufii [38]. Whether A. niger relies on generation of a positive-inside DY, a low plasma membrane proton permeability at low pH, or a combination of both, still remains to be investigated.

1.0 would be

explanation for this high tolerance towards extreme extracellular acidity would be to contribute this behaviour to plasma membranes with an unusual lipid composition, rendering them highly impermeable to protons. For acido- philic prokaryotes (both bacteria and archaea) it has been shown that a link exists between the lipid composition of their plasma membranes and an acidophilic mode of existence [32]. The proton permeability (P, cmÆs)1) in biological membranes has been found to be extremely pH dependent, with values ranging from 10)3 to 10)6 cmÆs)1 [33]. Based on results obtained by Sanders and Slayman [1], Burgstaller argued that the proton permeability of N. crassa plasma membranes is probably much lower than 10)3 to explain their results [33]. Using lipid bilayer membranes composed of bacterial phosphatidylethanolamine, Gutkn- echt found a 106 times lower P at pH 2 compared to pH 7, indicating much lower values for P at low pH [34]. Using a value for P of 10)6 cmÆs)1 for A. niger at pHex 1.5, and assuming hydrolysis of 1 ATP/H+ expelled by the P- ATPase, the ATP requirement to maintain a pHcyt value of + 7.6 can be calculated from the passive proton flux JH (molÆg dw)1Æs)1) ¼ PÆAÆD[H+]. For A (m2Æg dw)1), a value of 2.9 has been found for A. niger NW131 (B. R. Poulsen, Department of Microbiology, Fungal Genomics Section, Wageningen University, the Netherlands, personal commu- nication). Using these values the ATP turnover for pH homeostasis is 10)6 molÆg dw)1Æs)1. This value is within the same range of ATP turnover necessary for cellular main- tenance (2 · 10)6 molÆg dw)1Æs)1), assuming a maintenance coefficient m of 0.034 g glucoseÆg dw)1Æh)1 and 38 mol ATP formed per mole glucose (B. R. Poulsen, personal commu- nication). Although the exact value for P in fungal membranes at low pH is not known, these values suggest that with a low intrinsic proton permeability the energy costs to maintain such large DpHs are relatively low. This means that physical protection by the cytoplasmic mem- brane alone may be sufficient to keep pHcyt close to neutrality in an extremely acidic environment. If we assume DY to be 0 mV, then the proton motive force (Dp) generated around )400 mV (Dp ¼ at pHex DY ) ZDpH). This is near the theoretical limit if we assume the maximal amount of energy available to the P-ATPase to result from ATP hydrolysis and if a stoichiometry of 1 H+ expelled per ATP hydrolysed is assumed [35].

The capacity of the various energy-transducing mem- branes to maintain proton gradients appeared to be quite different. CCCP was much more efficient in dissipating DpH across the vacuolar membrane than across the cytoplasmic membrane, a finding that had already been reported for S. cerevisiae [39]. This could mean that the ability of the V-ATPase to be stimulated by increased proton leak is rather poor. ATP levels (and pH gradients) could be maintained or restored as long as the respiratory rate could still be stimulated after uncoupler addition, even when a temporary collapse of the vacuolar membrane DpH occurred. A complete collapse of DpH across the cytoplas- mic membrane in the presence of lethal CCCP levels (10–20 lM) could not be observed. A similar result has been reported for S. cerevisiae [39], and the authors suggested that besides ATP-dependent ion pumps an additional role in pH homeostasis may be reserved for the cytoplasmic buffering capacity. However, Sanders and Slayman [1] have clearly shown that a large involvement of the cytoplasmic buffering capacity in intracellular pH homeostasis is not very likely. A more obvious explanation would be that, although true uncouplers are able to reduce the H+ electrochemical potential difference across a membrane to zero when the concentration applied is high enough, the pH difference can only be dissipated if the charge imbalance as a result of the H+ transport is compensated by the movement of other ions [40]. Hence, if the permeability of the A. niger plasma membrane for other ions is low, a lack of compensating charge fluxes could account for the observed residual DpH. Lethal levels of azide resulted in a smaller residual cytoplasmic membrane DpH than lethal levels of CCCP (0.3 and 0.9 pH units, respectively). Besides cytoplasmic acidification due to specific inhibition of ATP synthesis, azide has been suggested to have additional uncoupling abilities [41]. This combined effect may account for the observed difference. However, as lethal doses of CCCP also lead to a complete loss of ATP, it is more tempting to speculate that azide, upon uptake into the cell, is able to alter ion conductivity in the cytoplasmic membrane. As a consequence, larger compensating ion fluxes may occur that allow a larger dissipation of cytoplasmic membrane DpH. Indeed, azide was found to have a specific effect on ion transport (probably K+/H+ exchange) in plasma membranes of S. cerevisiae [42].

In the presence of nonlethal azide concentrations, chan- ges in pHvac followed a course that was similar to pHcyt. Vacuolar H+ influx in response to increased cytoplasmic H+ levels indicates a role in pHcyt homeostasis for this organelle. These results are in accordance with observations made in S. cerevisiae [30] and in higher plant cells (Acer pseudoplatanus L.) [29].

The observed recovery of A. niger from nonlethal CCCP levels, even under conditions of complete vacuolar mem- brane DpH dissipation, may be attributed to induced expression of membrane-bound ATP-dependent transpor-

Compared to the pH-homeostatic properties of the organisms mentioned above, A. niger behaves like a typical acidophile. So far, bacteria and archaea have been the main focus of studies on intracellular pH homeostasis in acido- philes. Comparable to A. niger, the acidophile Thiobacillus ferrooxidans is capable of maintaining its intracellular pH constant at a value of 6.5 over a range of pHex from 1.0 to 8.0 [36]. Acidophiles are able to sustain such large cytoplasmic DpH values by counteracting the large inwardly directed H+ gradient with a positive-inside DY that results from Donnan or H+ diffusion potentials [37]. Thus, the proton motive force across the cytoplasmic membrane is reduced in order to decrease the proton leak into the cells and to decrease the back-pressure for H+ extrusion by the P-ATPase. In this way H+ influx eventually becomes self- limiting with an increasing positive-inside DY. A K+/H+ symport uptake system operating in parallel with the P-ATPase, combined with a low intrinsic cation permeab- ility of the plasma membrane, offers the acidophile a

Intracellular pH homeostasis in A. niger (Eur. J. Biochem. 269) 3493

(cid:1) FEBS 2002

9. Kugel, H., Mayer, A., Kirst, G.O. & Leibfritz, D. (1987) In vivo P-31 NMR measurements of phosphate metabolism in Platymonas subcordiformis as related to external pH. Eur. Biophys. J. 14, 461– 470.

ters of the ATP-binding cassette superfamily involved in multidrug resistance. Transcription of two ATP-binding cassette transporter encoding genes (atrA and atrB) in A. nidulans was shown to be enhanced within a few minutes of treatment with several drugs, including antibio- tics, azole fungicides and plant defence toxins [43]. A subsequent energy-dependent drug efflux activity may also confer such resistance to A. niger after CCCP addition, provided that a certain minimal ATP level can be maintained within the cell.

10. Hesse, S.J.A., Ruijter, G.J.G., Dijkema, C. & Visser, J. (2000) Measurement of intracellular (compartmental) pH by 31P NMR in Aspergillus niger. J. Biotechnol. 77, 5–15. 11. Bennet, J.W. (1998) Mycotechnology: the role of fungi in bio- technology. J. Biotechnol. 66, 101–107.

12. Ruijter, G.J.G. & Visser, J. (1999) Strain improvement. In Citric Acid Biotechnology (Kristiansen, B., Mattey, M. & Linden, J., eds), pp. 55–68. Taylor & Francis, London.

13. Legisˇ a, M. & Kidric, J. (1989) Initiation of citric acid accumula- tion in the early stages of Aspergillus niger growth. Appl. Micro- biol. Biotechnol. 31, 453–457.

14. Legisˇ a, M. & Kidric, J. (1991) The influence of intracellular pH of spores on citric acid production by Aspergillus niger. J. Biotechnol. 20, 313–318.

15. Yang, Y., Bastos, M. & Chen, K.Y. (1993) Effects of osmotic stress and growth stage on cellular pH and polyphosphate meta- bolism in Neurospora crassa as studied by 31P nuclear magnetic resonance spectroscopy. Biochim. Biophys. Acta 1179, 141–147. 16. Pilatus, U. & Techel, D. (1991) 31P-NMR-studies on intracellular pH and metabolite concentrations in relation to the circadian rhythm, temperature and nutrition in Neurospora crassa. Biochim. Biophys. Acta 1091, 349–355.

17. Ruijter, G.J.G., Panneman, H. & Visser, J. (1997) Overexpression of phosphofructokinase and pyruvate kinase in citric acid- producing Aspergillus niger. Biochim. Biophys. Acta 1334, 317– 326.

Industrial-scale production of citric acid is performed mainly by fermentation processes using A. niger [12]. Previously reported values for pHcyt in citric acid producing A. niger mycelium vary from 6.5 [13] to 6.8 [14]. We report here a considerable higher pHcyt of 7.5 (in both 2- and 7-day-old-mycelium) under fully oxygenated conditions. The low pHcyt values reported earlier probably resulted from difficulties to keep the mycelium well-energized during spectra acquisition. With metabolic engineering and mod- elling studies as a promising approach to improve produc- tivity [44], a detailed description of pHin during citric acid production provides valuable additional information about the conditions under which the enzymes involved operate in vivo. To reach a higher degree of accuracy for Aspergillus kinetic models, more information on the internal metabolic changes that take place during the transition to citric acid producing mycelium is needed in the future. Our results have shown that, at least with respect to intracellular pH, these changes are minor: the combination of low pHex and a high acid-secreting capacity only led to a slightly lower cytoplasmic pH.

18. Pontecorvo, G., Roper, J.A., Hemmons, L.J., MacDonald, K.J. & Bufton, A.W.J. (1953) The genetics of Aspergillus nidulans. Adv. Genet. 5, 141–238. 19. Vishniac, W. & Santer, M. (1957) The thiobacilli. Bacteriol. Rev. 21, 195–213.

A C K N O W L E D G E M E N T S

20. Nicolay, K., Scheffers, W.A., Bruinenberg, P.M. & Kaptein, R. (1982) Phosphorus-31 nuclear magnetic resonance studies of intracellular pH, phosphate compartmentation and phosphate transport in yeasts. Arch. Microbiol. 133, 83–89. The authors acknowledge financial support from the EC in the framework of the Eurofung Cell Factory project (QTRK3-1999– 00729).

R E F E R E N C E S

21. Shipanova, I., Bartoshevich, Y., Sibeldina, L., Zaslavskaya, P. & Michtchenko, A. (1995) Relationship between intracellular pH and antibiotic biosynthesis in Fusidium coccineum. Appl. Micro- biol. Biotechnol. 43, 514–517.

1. Sanders, D. & Slayman, C.L. (1982) Control of intracellular pH: predominant role of oxidative metabolism, not proton transport, in the eukaryotic microorganism. Neurospora. J. Gen. Physiol. 80, 377–402. 22. den Hollander, J.A., Ugurbil, K., Brown, T.R. & Shulman, R.G. (1981) Phosphorus-31 nuclear magnetic resonance studies of the effect of oxygen upon glycolysis in yeast. Biochemistry 20, 5871– 5880. 2. Madshus, I.H. (1988) Regulation of intracellular pH in eukaryotic cells. Biochem. J. 250, 1–8.

23. Lohmeier-Vogel, E.M., Hahn-Ha¨ gerdal, B. & Vogel, H. (1995) Phosphorus-31 and carbon-13 nuclear magnetic resonance studies of glucose and xylose metabolism in Candida tropicalis cell sus- pensions. Appl. Environ. Microbiol. 61, 1414–1419.

24. Mitchell, P. (1961) Coupling of phosphorylation to electron and hydrogen transfer by a chemiosmotic type of mechanism. Nature 191, 144–148.

3. Bracey, D., Holyoak, C.D., Nebe-von Caron, G. & Coote, P.J. (1998) Determination of the intracellular pH (pHi) of growing cells of Saccharomyces cerevisiae: the effect of reduced-expression of the membrane H+-ATPase. J. Microbiol. Methods 31, 113–125. 4. Serrano, R. (1988) Structure and function of proton translocating ATPase in plasma membranes of plants and fungi. Biochim. Bio- phys. Acta 947, 1–28.

25. Ruijter, G.J.G., van de Vondervoort, P.J.I. & Visser, J. (1999) Oxalic acid production by Aspergillus niger: an oxalate-non-pro- ducing mutant produces citric acid at pH 5 and in the presence of manganese. Microbiology 145, 2569–2576. 5. Sychrova´ , H., Ramı´ rez, J. & Pen˜ a, A. (1999) Involvement of Nha1 antiporter in regulation of intracellular pH in Saccharomyces cerevisiae. FEMS Microbiol. Lett. 171, 167–172.

26. Kornberg, A., Rao, N.N. & Ault-Riche´ , D. (1999) Inorganic polyphosphate: a molecule of many functions. Annu. Rev. Bio- chem. 68, 89–125. 6. Blatt, M.R. & Slayman, C.L. (1987) Role of (cid:3)active(cid:4) potassium transport in the regulation of cytoplasmic pH by nonanimal cells. Proc. Natl Acad. Sci. USA 84, 2737–2741.

27. Venema, K. & Palmgren, M.G. (1995) Metabolic modulation of transport coupling ratio in yeast plasma membrane H+-ATPase. J. Biol. Chem. 270, 19659–19667. 7. Klionsky, D.J., Herman, P.K. & Emr, S.D. (1990) The fungal vacuole: composition, function, and biogenesis. Microbiol. Rev. 54, 266–292.

28. Sanders, D., Hanssen, U.-P. & Slayman, C.L. (1981) Role of the plasma membrane proton pump in pH regulation in non-animal cells. Proc. Natl Acad. Sci. USA 78, 5903–5907. 8. Sureshkumar, G.K. & Mutharasan, R. (1993) Intracellular pH based controlled cultivation of yeast cells: I. measurement methodology. Biotechnol. Bioeng. 41, 118–128.

3494 S. J. A. Hesse et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

29. Gout, E., Bligny, R. & Douce, R.

(1992) Regulation of intracellular pH values in higher plant cells. J. Biol. Chem. 267, 13903–13909. 38. Gla¨ ser, H.-U., Sekler, I. & Pick, U. (1990) Indications for a K+/H+ cotransport system in plasma membranes from two acidophilic microorganisms. Biochim. Biophys. Acta 1019, 293– 299.

30. Carmelo, V., Santos, H. & Sa´ -Correia, I. (1997) Effect of extra- cellular acidification on the activity of plasma membrane ATPase and on the cytosolic and vacuolar pH of Saccharomyces cerevisiae. Biochim. Biophys. Acta 1325, 63–70.

39. Beauvoit, B., Rigoulet, M., Raffard, G., Canioni, P. & Gue´ rin, B. (1991) Differential sensitivity of the cellular compartments of Saccharomyces cerevisiae to protonophoric uncoupler under fer- mentative and respiratory energy supply. Biochemistry 30, 11212– 11220.

31. Slonczewski, J.L., Rosen, B.P., Alger, J.R. & Macnab, R.M. (1981) pH homeostasis in Escherichia coli: measurement by 31P nuclear magnetic resonance of methylphosphonate and phos- phate. Proc. Natl Acad. Sci. USA 78, 6271–6275.

40. Ginzburg, M., Ratcliffe, R.G. & Southon, T.E. (1988) Phosphorus metabolism and intracellular pH in the halotolerant alga Duna- liella parva studied by 31P-NMR. Biochim. Biophys. Acta 969, 225– 235. 32. Albers, S., van de Vossenberg, J.L.C.M., Driessen, A.J.M. & Konings, W.N. (2001) Bioenergetics and solute uptake under extreme conditions. Extremophiles 5, 285–294.

33. Burgstaller, W. (1997) Transport of small ions and molecules through the plasma membrane of filamentous fungi. Crit. Rev. Micriobiol. 23, 1–46. 41. Lohmeier-Vogel, E., Skoog, K., Vogel, H. & Hahn-Ha¨ gerdal, B. (1989) 31P nuclear magnetic resonance study of the effect of azide on xylose fermentation by Candida tropicalis. Appl. Environ. Microbiol. 55, 1974–1980. 34. Gutknecht, J. (1984) Proton/hydroxide conductance through lipid bilayer membranes. J. Membrane Biol. 82, 105–112. 42. Riemersma, J.C. (1968) Effects of sodium azide and 2,4-dinitro- phenol on phosphorylation reactions and ion fluxes in Saccharo- myces cerevisiae. Biochim. Biophys. Acta 153, 80–87.

35. Warncke, J. & Slayman, C.L. (1980) Metabolic modulation of stoichiometry in a proton pump. Biochim. Biophys. Acta 591, 224– 233.

43. Del Sorbo, G., Andrade, A.C., van Nistelrooy, J.G.M., van Kan, J.A.L., Balzi, E. & de Waard, M.A. (1997) Multidrug resistance in Aspergillus nidulans involves novel ATP-binding cassette trans- porters. Mol. Gen. Genet. 254, 417–426. 36. Cox, J.C., Nicholls, D.G. & Ingledew, W.J. (1979) Transmem- brane electrical potential and transmembrane pH gradient in the acidophile Thiobacillus ferro-oxidans. Biochem. J. 178, 195–200. 37. Matin, A. (1990) Bioenergetics parameters and transport in obli- gate acidophiles. Biochim. Biophys. Acta 1018, 267–270. 44. Ruijter, G.J.G., Panneman, H. & Visser, J. (1998) Metabolic engineering of the glycolytic pathway in Aspergillus niger. Food Technol. Biotechnol. 36, 185–188.