In this work we used two different strains of P. ochrochloron, a fungus, which is known to tolerate high heavy metal concentrations (Tuthill et al., 2001). The first strain, P. ochrochloron CBS 123823, an isolate from the soil of the copper works Montanwerke Brixlegg (Tyrol, Austria) in 1986, was incorrectly identified in 1989 by the Centraalbureau voor Schimmelcultures (CBS, The Netherlands) as Penicillium simplicissimum (Franz et al., 1991). The second strain, P. ochrochloron CBS 123824 emerged from the wild type strain CBS 123823 over 20 years of passaging on rye. Both strains were recently re-identified by three independent institutions as the close related species P. ochrochloron (Vrabl et al., 2008).

All bioreactors were inoculated with filamentous mycelium pregrown in a HEPES–glucose medium for 60 h as previously described (Gallmetzer et al., 1998).

Chemostat experiments were performed in a Biostat M bioreactor (Sartorius/Braun, working volume of 1.7 L, bioreactor equipped with an overflow device) under ammonium limitation (medium composition see Gallmetzer and Burgstaller, 2002; cultivation conditions as described in Gallmetzer and Burgstaller, 2001). For pH-shift experiments the culture pH was raised stepwise (Gallmetzer, 2000): starting with an initial pH of 2, the culture pH was adjusted with sodium hydroxide to the desired next pH level (pH 4, 6, and 7). The chemostat cultures typically reached a new steady state (assumed with onset of constant oxygen consumption and biomass concentration) after passing four to six generation times following a shift in pH. Once a steady state was established, samples were withdrawn regularly from the bioreactor over a minimum time period of 30 h (at least four sampling points; for each sampling point three samples were taken and averaged) to monitor residual glucose concentration, organic acid production, and biomass concentration.

Prior to inoculation to the bioreactor, the preculture was harvested under sterile conditions by filtration (0.2 μm, cellulose acetate) and washed with approximately 200 mL sterile distilled water. The mycelial cake was transferred aseptically to the bioreactors (either Biostat M, Sartorius/Braun, Melsungen, Germany or KLF 2000 Bioengineering, Wald, Switzerland). Both bioreactors were operated with a working volume of 1.7 L. Equipment and culture conditions were according to Vrabl et al. (2008, 2009). The pH was kept constant with 1 M sodium hydroxide. External pH and dissolved oxygen tension were monitored with an online data acquisition system (Fa. Gatterbauer Messtechnik, Sipbachzell, Austria).

The medium for bioreactor batch cultivation was a modified preculture medium and concentrations are the same for ammonium and phosphate-limited conditions unless stated otherwise (mM): glucose × 1 H₂O (400), (NH₄)₂SO₄ (NH₄ limitation: 5, PO₄ limitation 6.25), NH₄Cl (NH₄ limitation: 10, PO₄ limitation: 12.5), KH₂PO₄ (NH₄ limitation: 5.8, PO₄ limitation: 0.5), MgSO₄ × 7 H₂O (1.6), 1 g L⁻¹ antifoam agent (Clerol FBA 5075, Cognis, Germany) and 10 mL L⁻¹ of trace element solution (composition of the trace element solution as in Vrabl et al., 2008). Glucose and salts were sterilized at 121°C separately, whereas the trace element solution was added after sterilization via filtration. Prior to inoculation, the pH was adjusted with 1 M sodium hydroxide to either pH 5 or 7. Each nutrient limitation was performed with both strains at pH 5 and 7. All conditions were performed at least in triplicates, except for strain CBS 123824 the phosphate limitation at pH 5 was performed in duplicate. The single experiments were limited by the small working volume (1.7 L) of the bioreactors. To disturb the cultivations as little as possible by the high sampling frequency, only one sample per time point was taken. Samples (10 mL) were withdrawn from the bioreactor at regular time intervals and immediately filtrated (cellulose acetate filter, pore size 0.45 μm). The mycelial cake was used for dry weight estimation. The filtrate was rapidly aliquotated into Eppendorf tubes and frozen in a cold bath (−40°C; Frykatherm) for later analysis of extracellular organic acids and residual nutrient concentrations. Cultivations were monitored at least for 40 h after inoculation.

A phosphate-limited culture with the strain CBS 123824 was started at a pH of 7. Without sodium hydroxide titration overnight, the fungus acidified the medium during the lag and exponential phase until the next morning. From then the pH was kept constant at pH 5 throughout the experiment. These experiments were performed twice.

Extracellular nutrient and organic acid concentrations were determined in culture filtrates, which were diluted prior to analysis if necessary. Glucose was measured with a Megazyme Test Kit for d-Glucose (GOPOD Format, K-GLUC, Megazyme International Ireland Ltd., Wicklow, Ireland), ammonium according to the macroscale method of Rhine et al. (1998) and phosphate as described by Illmer (1996). Organic acids were determined according to Womersley et al. (1985) on a Bio-Rad Aminex HPX-87H column with 4 mN sulfuric acid as mobile phase, a temperature of 41°C and a flow rate of 0.6 mL min⁻¹. Dry weight was estimated as previously described (Vrabl et al., 2009).

Rates for nutrient uptake, organic acid excretion, or sodium hydroxide titration were calculated by differentiation of the respective nutrient, organic acid, or sodium hydroxide versus time curves as previously described (Vrabl et al., 2009).

Although bioreactor batch experiments showed comparable tendencies, there were slight temporal shifts between parallel experiments, e.g., caused by small differences in culture history. These temporal shifts are particularly important when considering dynamic processes like the short spike in pH shown in Figure 4 or 6: Averaging such curves can conceal real dynamics and should be avoided. We therefore depicted one representative (e.g., Figures 4 and 5) dataset or all three datasets for each experimental condition.

Under all tested conditions, total organic acid excretion in P. ochrochloron increased with rising external pH. This observation was independent whether the fungus was grown under ammonium-limited conditions in a chemostat between pH 2 and 7 (Figure 2) or in highly dynamic bioreactor batch cultivations under ammonium or phosphate-limited conditions (Figures 3–5). In contrast, the spectrum of excreted organic acids varied strongly with pH, strain and nutrient limitation: While oxalate and citrate were the predominant organic acids of the strain CBS 123823 of both nutrient limitations at pH 7 (Figures 3A,C), solely citrate was excreted in considerable amounts at pH 5. In strain CBS 123824 malate was the main excreted organic acid during ammonium-limited batch cultures at pH 7, whereas at pH 5 succinate and malate were predominant (Figure 3B). The same strain excreted mainly citrate during phosphate limitation (Figure 3D).

In addition to cultivations at pH 5 and 7, phosphate-limited cultures of the strain CBS 123824 were exposed to a transient in pH from pH 7 to 5 during the lag and exponential phase, i.e., starting with an initial pH of 7, the fungus was allowed to acidify the medium until reaching pH 5 which was kept constant until the end of cultivation. In these experiments the amount of excreted organic acids was in the magnitude of cultivations at pH 5 (data not shown).

In contrast to chemostat studies, bioreactor batch experiments depict complex nutritional and physiological changes. Because both strains showed comparable dynamics (Figures 4 and 5), we demonstrate the main findings with the P. ochrochloron strain CBS 123823, i.e., the strain which excretes more organic acids (as demonstrated in Figure 3). At any pH P. ochrochloron excreted organic acids during the whole cultivation, i.e., low amounts of organic acids were excreted during exponential growth but the excretion was strongly enhanced on depletion of either ammonium or phosphate (exception see Unusual Findings).

The course of ammonium or phosphate uptake in ammonium-limited cultures (Figure 4A) did not significantly differ between pH 5 and 7 (note that phosphate uptake more or less halted after ammonium had been exhausted). In contrast to ammonium and phosphate, glucose was consumed faster at pH 7 than at pH 5 despite a lower biomass concentration at pH 7 (Figure 4A). This is also reflected in a decreasing biomass yield on glucose with increasing pH (Table 1).

In phosphate-limited cultures at pH 5 (Figure 5A), P. ochrochloron utilized ammonium considerably faster than at pH 7, leading surprisingly to an early double limitation. Thus, biomass at pH 5 and 7 remained approximately at the same level. Similar to ammonium-limited experiments, the course of phosphate uptake did not differ between pH 5 and 7. When measuring the glucose concentrations of phosphate-limited cultures, we often observed an extreme scattering of the values, especially at pH 7. This scattering was independent of the analytical method, i.e., photometrical assay or HPLC, or the bioreactor used. Interestingly we did not observe this problem with any ammonium-limited experiment, although sampling and sample preparation were the same. From experiments with less scattering such as depicted in Figure 5, the glucose consumed per g dry weight seemed to be almost in the same order of magnitude.

The pH was kept constant with sodium hydroxide throughout all cultivations, because we expected only an acidification of the medium in these early stages. Surprisingly the external pH shortly increased in ammonium-limited cultures of the strain CBS 123823 at pH 5 after exhaustion of ammonium (Figure 4C). This effect was even more distinct with the strain CBS 123824 (Figure 6A). In this strain the short spike in pH after exhaustion of ammonium was followed by a wavelike sequence of increases and decreases.

Although glucose was present in excess at this time, some organic acids seemed to be reutilized (Figure 6B): While citrate and fumarate seemed to be almost completely taken up, oxalate excretion increased steadily (data for oxalate not shown). On the other hand, pyruvate, malate, and succinate (data for succinate not shown) remained on the same level or followed a wavelike progress. Likewise, the strain CBS 123823 occasionally seemed to reutilize organic acids (e.g., pyruvate in Figure 4B) despite extracellular abundance of glucose.

The main finding of this study was that the total amount of excreted organic acids in P. ochrochloron increased with rising pH – irrespective of the mode of cultivation (chemostat, batch cultivation), nutrient limitation (ammonium, phosphate limitation), or strain (Figures 2–5). As far as we are aware, only two other studies address this topic for a broader range of organic acids, i.e., a pH-stat study of N. crassa (Slayman et al., 1990; C. L. Slayman and P. Kaminski, unpublished results, Figure 1 with kind permission) and a bioreactor batch study of A. niger (Andersen et al., 2009). Both studies observe increasing organic acid excretion with rising external pH.

Apparently this observation holds true despite huge differences in experimental setting (e.g., time frame, cultivation mode, or nutrient limitation) or physiological status of the mycelia (e.g., chemostat mycelium growing at a defined specific growth rate versus batch cultures at varying growth rates or non-growing hyphae in a pH-stat). This confirms our hypothesis that this observation might be very well a general response of filamentous fungi to increased ambient pH. It could thus provide a key point in understanding the function and mechanisms of organic acid excretion in these organisms.

The physiological reason behind this phenomenon still remains unclear and becomes particularly intriguing when one considers the comparably high metabolic costs involved in organic acid production. Why would total organic acid excretion be increased at higher pH levels at all?

One possible explanation is that at higher pH levels the glycolytic flux is elevated and leads as a consequence of energy spilling mechanisms to enhanced level of organic acid excretion (theoretical details on the “overflow metabolism hypothesis” see Box 1). In P. ochrochloron, the data suggest an elevated glycolytic flux at higher pH levels (present study; Gallmetzer, 2000). Similar results can be found for other filamentous fungi, for instance, glucose uptake rates increase with external pH in A. niger (Mischak et al., 1984; Torres et al., 1996) and in Trichoderma harzianum (Delgado-Jarana et al., 2003). In the latter study the authors, however, observed a slight decrease in glucose uptake at pH 6 compared to pH 5. Also decreased yields on glucose at higher pH levels (Table 1) hint in this direction and are frequently reported (e.g., current study, Pitt and Bull, 1982; Andersen et al., 2009).

A second possible explanation for increased organic acid production at high external pH is that organic acid anion excretion serves as charge balance mechanism (theoretical details on the “charge balance hypothesis” see Box 2). In literature there is some experimental evidence for an electrical compensation of proton efflux via organic anion excretion, e.g., in non-growing and multiple starved hyphae of N. crassa (personal communication C. L. Slayman), the proton efflux was almost matched by organic anion excretion (Slayman et al., 1990; Rivetta et al., 2011). Roos and Luckner (1984) reported for P. cyclopium that as long as NH4+ was present in the medium, NH4+ uptake provided the necessary charge balance for proton efflux via the plasma membrane H⁺-ATPase with a 1:1 stoichiometry. After NH4+ was exhausted, however, charge balance for proton efflux was ensured by the stoichiometric excretion of organic acid anions. A stoichiometric relationship between ammonium uptake and proton release (or sodium hydroxide titration) has also been reported for several other fungi (e.g., Mattey, 1992; Burgstaller et al., 1994).

A third possible explanation for increased organic acid excretion at higher pH levels originates from a hypothesis of Andersen et al. (2009) for A. niger. The authors suggest that fungi excrete at high external pH more protons (in form of organic acids) to acidify the environment and to outcompete rivals (theoretical details on the “aggressive acidification hypothesis” see Box 3). Simulations predicted oxalic acid as the major source of protons at pH levels above pH 3 for strains which are able to excrete all organic acids.

In this context three issues are of special interest. Firstly, some authors attribute the major mechanism behind medium acidification to proton release by the plasma membrane H⁺-ATPase (e.g., Sanders, 1988; Jernejc and Legisa, 2001) and not to organic acid excretion. The latter would require the transport of fully protonated organic acids across the plasma membrane which is questionable (in depth discussion in “Are Organic Acids Excreted as Anions or Fully Protonated?”). Secondly, in P. ochrochloron but also in other fungi, medium acidification seems to be at least partly a matter of the N-source. When ammonium is present as sole N-source, P. ochrochloron rapidly acidifies the medium to very low pH values (Burgstaller et al., 1994), but not when nitrate is the sole N-source (although organic acids were excreted at the same time; Vrabl, Schinagl, and Burgstaller unpublished results). Similar observations have been reported for other filamentous fungi including A. niger: External pH typically decreases in cultures with ammonium as N-source, but on other N-sources like nitrite or glutamate a slower acidification or even an alkalinization of the medium can be observed (Usami, 1978; Amrane et al., 1999; Jernejc and Legisa, 2001). Thirdly, whilst the model explained the experimental data for A. niger very well (Ruijter et al., 1999; Andersen et al., 2009), it does not seem to apply to the data available from P. ochrochloron (e.g., Figures 2 and 3). For example, oxalic acid was the predominant acid for the strain CBS 123823 only at pH 7 and never for the strain CBS 123824, although both strains are capable of excreting all organic acids. Citric acid was excreted at both pH levels. It was even the predominant excreted organic acid for strain CBS 123823 at pH 5 (Figures 3A,C) and for phosphate-limited cultures of strain CBS 123824 at pH 7 (Figure 3D). These results indicate that oxalic acid is not the major source of protons in P. ochrochloron. Thus in P. ochrochloron the shift in the spectrum of excreted organic acids has to have other reasons than to most efficiently acidify the medium and might be explained as a consequence of changes in cytoplasmatic pH. Although filamentous fungi keep their cytoplasmatic pH surprisingly stable over a broad range of external pH, the cytoplasmatic pH is not entirely unaffected and changes slightly (Hesse et al., 2002). Because the activity of several enzymes is very sensitive to changes between pH 6 and 7 (Mattey, 1992), a shift in external pH can cause a shift in enzyme activity and thus also a shift in the spectrum of excreted organic acids.

Evidently, in each of the three above mentioned explanations organic acid excretion fulfills a completely different metabolic function. Moreover, the assumption on how organic acids are transported across the plasma membrane is mutually exclusive concerning the charge balance hypothesis (transport of organic acid anions) and the aggressive acidification hypothesis (transported fully protonated). We will therefore consider this topic in more detail in the following section.

The charge balance hypothesis and the aggressive acidification hypothesis differ in their assumption on which organic acid species are transported across the plasma membrane. This is of particular interest when one considers the question, if organic acids are the major source of protons for ambient acidification or not. The charge balance hypothesis requires excretion of organic acid anions. Thus, depending on the degree of deprotonation and the ambient pH, organic acid anions would contribute only little. In contrast, the aggressive acidification hypothesis postulates organic acids to be excreted fully protonated (Andersen et al., 2009), and consequently, to contribute strongly to the medium acidification.

So far, experimental data on the transported species of organic acids are scarce (Rivetta et al., 2011), but there is a significant argument in favor of anions as the transported species (Mattey, 1977; Roos and Luckner, 1984; Burgstaller, 2006; Rivetta et al., 2011). The cytoplasmic pH is around neutral (Hesse et al., 2002) or, maybe, slightly more acidic near the interior surface of the plasma membrane (Roos and Slavik, 1987). Considering the pK_a values of the organic acids and the cytoplasmic pH it follows that organic acids are almost completely in their anionic forms in the cytoplasm. Inversely, the concentrations of fully protonated organic acids are extremely low (Burgstaller, 2006; Garcia and Torres, 2011). This implies that the excretion rates, if calculated from the concentrations of fully protonated species, are far too low to explain the measured excretion rates of organic acids (Burgstaller, 2006). Taken together, it is unlikely that organic acids are transported fully protonated across the fungal plasma membrane and therefore they cannot be considered as major source of protons for extracellular acidification.

What does this mean for medium acidification in general? The plasma membrane H⁺-ATPase is most probably the main source of protons for medium acidification in fungi (Sanders, 1988; Sigler and Höfer, 1991; Jernejc and Legisa, 2001). Net extracellular acidification, however, results from an interplay between the excreted species of organic acids, proton excretion by the plasma membrane H⁺-ATPase, the extracellular pH, and the extracellular buffering capacity. Net extracellular acidification may thus be similar whether or not organic acids are excreted fully protonated or as anions accompanied by protons expelled via the H⁺-ATPase. Because the latter is the more probable scenario, organic acid excretion must serve another purpose (e.g., charge balance, energy spilling, or chelation of trace elements) than merely acidification of the external medium.

Although filamentous fungi typically tend to acidify their medium especially when ammonium serves as N-source, we observed a short uprise in pH after exhaustion of ammonium at pH 5 for both P. ochrochloron strains (Figure 4 and 6) which might be explained the reuptake of organic acids (Figure 6). Several studies report protons disappearing from the medium during dicarboxylic acid uptake, for example in Schizosaccharomyces pombe (Sousa et al., 1992) or Hansenula anomala (Corte-Real and Leao, 1990), which could be the consequence of a dicarboxylate proton symport (Corte-Real and Leao, 1990; Sousa et al., 1992). In some fungi like Candida utilis the transport is repressed when glucose is present (Saayman et al., 2000), whereas in other fungi like S. pombe the transporter seems to be constitutively active, even at high external glucose concentrations (Sousa et al., 1992; Saayman et al., 2000). This might also be the case in P. ochrochloron. Strikingly, in the bacterium Salmonella typhimurium the formate transporter FocA changes its function depending on the external pH, i.e., performing as an exporter at high ambient pH and as an importer at acidic pH values (Lü et al., 2011).

In this respect the work of Roos and Luckner (1984) is of particular interest. After adding [1.5-¹⁴C] citric acid to cultures of P. cyclopium under organic acid excreting conditions (i.e., glucose excess), they found a continuous uptake and excretion of citric acid. Uptake dominated at later phases of cultivation which resulted in a net uptake of citric acid. This experiment points out two very important issues: First, external organic acid concentrations require cautious interpretation because they might be the consequence of a continuous influx and efflux and thus represent only a net flux. Second, organic acids may serve an additional, yet unknown metabolic function. Clearly this issue needs more experimental evidences.