Geobacter species are well known for their metal reducing capabilities and are responsible for accelerating the bioremediation of radioactive and toxic metals in subsurface environments [1–3]. They are capable of anaerobic oxidation of organic contaminants with concomitant reduction of metals such as Fe(III) and other contaminants such as Uranium, and consequently are important for groundwater bioremediation [4–7]. Furthermore, Geobacter species are capable of direct electron transfer to an electrode motivating their use in microbial fuel cells [8]. The pure culture model of Geobacter sulfurreducens from Geobacteraceae family has been extensively studied in order to characterize its physiology and to determine the mechanisms associated with the extra-cellular electron transfer. Recently, several genome-wide studies have characterized the unique metabolic features of Geobacteraceae including the development of a genome-scale metabolic model of Geobacter sulfurreducens, the chemotaxis towards iron, and the synthesis of conductive pili [9–13]. The genome-based modeling revealed that, global proton balance was significantly different for Geobacter species, which rely on extra-cellular electron transfer to insoluble substrates, in comparison to aerobic organisms which reduce oxygen to water. Furthermore, improved understanding of the global proton balance in Geobacter species can provide insights on the physiology of other species capable of extra-cellular electron transfer such as Rhodoferax, Shewanella regardless of the mechanism of the electron transfer since protons are not consumed at the terminal electron accepting step in all cases.

Changes in the environmental conditions such as pH, temperature, osmolarity, and electron donor availability can alter the trans-membrane pH gradient, total proton motive force and affect the internal pH of the cells as well as the energetics [14]. Hence, in order to maintain energy homeostasis, cells have to actively regulate the internal pH by secreting or consuming protons. The medium can act as a source/sink for protons depending on the substrates that are present. The characterization of this proton exchange can provide additional insights on the metabolism and physiology of less studied organisms such as Geobacteraceae that use extra-cellular electron transfer for energy generation.

Several in silico models have been constructed to describe and predict the intracellular metabolism at the genome-scale for many organisms including E. coli, S. cerevisiae, and B. subtilis [15–20]. Specifically, the constraint-based modeling approach has been used to study E. coli metabolism for over ten years [21, 22]. These constraint-based models provide a framework to predict cellular physiology including growth and by-product formation across a range of growth environments [23]. Predictions of proton secretion/consumptions rates associated with E. coli grown under aerobic conditions in the presence of varying electron donors has been reported [24]. Consequently, the measurements of the proton secretion/consumption associated with the cellular growth and metabolism can be used to derive additional information on physiology that can be used to further validate the models.

The in silico analysis of Geobacter sulfurreducens metabolism revealed that differences in global intracellular proton balance can lead to lowered biomass yields during growth with extracellular electron acceptors such as Fe(III), relative to the growth with electron acceptors reduced in the cytosol, such as fumarate [25], highlighting the need for detailed analysis of the proton exchange associated with metabolism. One of the major concerns during high-density growth of E. coli on excess glucose under aerobic conditions is the formation of acidic by-products [26–29]. Several studies have been carried out to determine the effect of fermentation conditions on accumulation of acetate and other by-products [26, 30–33].

The bacterial cells grown in complex environment use the available substrate either preferentially or simultaneously depending on the growth condition. The metabolism of glucose and acetate in E. coli has been extensively investigated during the past 50 years [34] and are well characterized. Although dynamic models of growth on mixed substrates have been developed based on experimental data from substrate uptake rate and biomass production measurements [35–37], no previous experimental study, to our knowledge, measured the net production or consumption of protons in different environments, even though in the case of E. coli, modeling studies suggested that the direction of proton exchange was found to depend on the electron donor [24]. The scope of this work was to first confirm the previous model predictions of proton exchange for E. coli and subsequently, to measure the proton production or consumption during growth of Geobacter species on acetate with fumarate or ferric citrate and compare it with the in silico model predictions (Figure 1). Models that can predict the rate of proton secretion or consumption under a wide range of environmental conditions will be valuable in developing model-based approaches to the optimization of power production in microbial fuel cells and the in situ bioremediation of contaminated subsurface environments, where the effect of pH changes is critical.

It has been shown previously that for Geobacter species, global proton balance has been shown to be critical in determining biomass yields during the respiration of extracellular electron acceptors. Here, we have investigated the relation between the growth environment and proton generation/consumption for E. coli, G. sulfurreducens and G. metallireducens. Specifically, we studied the variation in proton exchange as a function of the electron donor (glucose, acetate) for E. coli and the electron acceptor (fumarate/Fe(III) citrate) for G. sulfurreducens and G. metallireducens. The results from these experiments are detailed in the following sections.

2.1 Analysis of Proton Production in E. coli

Recent genome-scale metabolic models incorporate detailed charge and elemental balance, and consequently, the number of protons consumed or generated for every metabolic reaction is represented in the model. This feature allows the calculation of proton flux generated or consumed during growth in varied environments. Reed et al. [24] described the variation of proton secretion flux with the carbon source. Depending on the limiting substrate, the exchange of protons across the membrane can produce either basic or acid environment. For E. coli, during growth with acetate as electron donor, the external medium environment became basic and acid was added to maintain neutral pH as shown in Figure 2 (panels c & d). The maximum observed OD_{550 nm} value was 0.45 and substrate was completely consumed after 25 hrs. Similar experiment was carried out for growth of E. coli on glucose. However, in the case of growth with glucose, base was added in order to maintain circumneutral pH (Figures 2a &2b). The variations in proton generation with the electron donor are also consistent with the reduced nature of glucose as compared to acetate. Thus, glucose oxidation resulted in generation of excess protons that are not consumed completely by the formation of water during the oxygen reduction. These excess protons are secreted into the medium along with other acid by-products resulting in an acidification of the medium forcing the addition of base to maintain a constant pH. Table S1 (Additional file 1) shows the initial and final biomass concentrations and the amount of acid/base added for all of the studies in this report.

2.2 Proton Production in GeobacterSpecies

G. sulfurreducens and G. metallireducens containing the fumarate transporter [38], were grown in a batch bioreactor, in the absence of a buffer, with acetate (10 mM) as electron donor and fumarate (40 mM) as the sole electron acceptor as previously described [39]. In control experiment(s), the inoculum(s) was omitted to investigate changes in medium pH. The medium pH remained neutral and there was no addition of either acid or base in all control experiments (data not shown).

Both the Geobacter strains showed a similar trend in the amount of acid added as compared to the aerobic growth of E. coli with acetate (Figure 3). The pH of the culture medium gradually increased and acid was added to maintain neutral pH as shown in Figures 3b &3c. G. sulfurreducens and G. metallireducens obtained a maximum OD value of 0.37 and 0.42 respectively. Reed et al[21, 24] showed that by choosing different electron donors, the medium can be made either acidic or basic. In addition to the variation of the electron donors, we have also investigated the influence of electron acceptors on the medium pH. Figure 4 shows the acetate consumption by G. sulfurreducens during growth in the presence of acetate (10 mM) and Fe(III) citrate (40 mM).

During the growth of G. sulfurreducens with Fe(III) citrate as sole electron acceptor, protons were produced by the cells. The cell growth was inferred indirectly by acetate oxidation and Fe(III) reduction since Optical Density (OD) measurements are not possible and cell counts lead to significant variances in the measurements. In control experiments, the acid addition started at the beginning and continued for 25 hrs until volume of 10 ml was dispensed into the bioreactor to maintain neutral pH (7.0), however reduction of Fe(III) was not observed. This observed increase in pH of the uninoculated medium upon N₂ sparging can be explained by an inorganic reaction equilibria model (data not shown). In contrast, during the growth of Geobacter sulfurreducens, the medium was maintained at neutral pH without external addition of either acid or base. This can be attributed to the secretion of protons during the growth phase G. sulfurreducens as a result of cellular metabolism.

However, as Fe(III) was depleted at 25 hr, the cells reached stationary phase, consequently proton production associated with growth stopped and the acid had to be gradually added externally by the pump until it reached volume of 21 ml. The change in pH of the medium containing Fe(III) citrate in control experiments could be attributed to the potential release of hydroxyl ions from inorganic equilibrium reactions. The hydroxyl ions are reactive species, and will react immediately to form Fe(OH)₃ which can precipitate in the solution. Similar results were reported by Francis and Dodge [40].

This study shows that proton exchange also depends on the choice of the electron acceptor. Specifically, for the case of the extracellular acceptors such as Fe(III), there appears to be net production of protons. This result is consistent with the model predictions of proton secretion during Fe(III) reduction [25] and provides additional evidence to support the hypothesis, that the biomass yield of G. sulfurreducens in the presence of Fe(III) citrate as the electron acceptor is lower than fumarate due to the need to maintain global proton balance by exporting protons from the cell at the expense of ATP generation. The comparison of predicted and measured proton fluxes is presented in Figure 5. Here, the positive proton flux represents secretion and negative flux represents uptake. The model predictions of proton production are consistent with the data and are within the experimental error for E. coli cultures, where as for the Geobacter species the model predictions qualitatively predict proton consumption although the extent of the predicted proton consumption is higher than what is observed experimentally.

Bacterial cells have to regulate the cytoplasmic pH to survive in the constantly changing environment as bacterial growth is dependent on substrate availability, as well as the redox potential and the pH of the medium. The regulation of internal pH involves proton export that requires energy in the form of ATP. The biochemical reactions in the cytoplasm associated with metabolism can lead to a net proton production or consumption. The variations in proton concentration in the cytoplasm during growth, can affect the kinetics and the thermodynamic feasibility of biochemical reactions necessitating active regulation of pH. The energetic cost of pH regulation via exporting protons associated with metabolism can impact the biomass yield of the organism and the extent of this effect can vary with the environment.

For example, in E. coli grown with glucose as the electron donor, organic acids released during growth greatly contribute to changes in pH of the medium. Previous studies [41–43] have shown that changes in pH of the culture medium affects the growth of microorganism. However, measurement of such pH changes can be used to analyze the metabolic state of the cells. Previously, changes in pH have been used to identify the onset of product synthesis using on-line parameter estimation in fermentation process San et al [44]. More recently, the in silico model by Reed et al [24] predicted the variation of proton secretion flux for E. coli with different electron donors. In this study, we have shown that the in silico model predictions were consistent with experimentally observed proton production or consumption in 1) cultures of E. coli grown aerobically on different electron donors, and 2) Geobacter species grown anaerobically on acetate with different electron acceptors.

During aerobic growth, intracellular protons along with electrons transferred from the electron transport chain combine with oxygen resulting in the formation of water as well as removal of intracellular protons. In contrast, during anaerobic growth with metals, the protons generated during substrate oxidation cannot be removed at the terminal electron accepting step. Consequently, proton generation profile during anaerobic growth will be markedly different from aerobic growth in similar conditions. In Geobacter species, acetate oxidation is coupled with reduction of fumarate which is the terminal electron acceptor. Biomass yield of G. sulfurreducens during acetate oxidation with fumarate as the electron acceptor is higher than the corresponding yield during Fe(III) reduction. This has been attributed to the differences in proton generation associated with metabolism, since during Fe(III) reduction, protons are generated and there is a net proton efflux that consumes energy. During the reduction of soluble acceptors such as fumarate, protons are consumed at the terminal electron accepting step resulting in a net proton uptake. In this study, during Fe(III) reduction, we observed a net production of protons, whereas, we found that during fumarate reduction acid was added to maintain pH balance indicating proton consumption.

The predicted proton uptake agreed with the measured acid addition for E. coli grown on glucose. However, for all other cases model predicted higher proton consumption by the cells. One reason for this discrepancy could be the incorrect representation of protons that are consumed during biomass production. Another factor could be either the bicarbonate produced during growth of E. coli that can buffer some of the pH changes experimentally or the production of organic acids that are not represented in the model. Yet another factor could be the uptake of NH₄⁺ followed by proton extrusion observed for bacteria and fungi [45–47]. While these mechanisms might explain the discrepancies for E. coli, it is important to note that CO₂ is also consumed by G. sulfurreducens and acids cannot be secreted during acetate oxidation. Hence, the factors that contribute to proton production during growth of Geobacter species needs to be further elucidated. These effects are not incorporated in the current stoichiometric model which does not include the acid/base equilibrium reactions in the reactor. In addition, since the proton exchange flux was calculated based on the acid/base added during batch growth experiments, errors in the measurements could also lead to additional discrepancy between the model and the data.

The results in this study clearly highlight the ability of the genome-scale models that incorporate detailed proton balanced biochemical reactions to predict, albeit qualitatively, the proton exchange rates and other aspects of physiology. Future research in this area would extend the scope of genome-scale models by including detailed inorganic equilibria reactions to accurately represent the abiotic reactions that can consume or produce protons in the medium. In order to effectively incorporate the abiotic reactions in the model, additional experimental measurements such as the CO₂ production/consumption rate along with NH₄ consumption rate will be required. Such an integrated description will be valuable in practical applications such as the in situ bioremediation or microbial fuel cells, where changes in environmental pH would greatly affect the respiration and metabolism. Therefore, models that can predict the rate of proton secretion or consumption under a wide range of environmental conditions will be valuable in predicting the growth physiology as well as the extent of respiration and for prioritizing strategies for bioremediation of contaminated subsurface environments and optimization of power production in microbial fuel cells.