Novel properties of photofermentative biohydrogen production by purple bacteria Rhodobacter sphaeroides: effects of protonophores and inhibitors of responsible enzymes

Background Biohydrogen (H2) production by purple bacteria during photofermentation is a very promising way among biological H2 production methods. The effects of protonophores, carbonyl cyanide m-chlorophenylhydrazone (CCCP), 2,4-dinitrophenol (DNP), and inhibitors of enzymes, involved in H2 metabolism, metronidazole (Met), diphenyleneiodonium (DPI), and dimethylsulphoxide (DMSO) on H2 production by Rhodobacter sphaeroides MDC6522 isolated from Jermuk mineral springs in Armenia have been investigated in both nitrogen-limited and nitrogen-excess conditions. Results With the increase of inhibitors concentrations H2 yield gradually decreased. The complete inhibition of H2 production was observed in the presence of DPI and CCCP. DPI’s solvent—DMSO in low concentration did not significantly affect H2 yield. N,N′-dicyclohexylcarbodiimide (DCCD)-inhibited the FOF1-ATPase activity of bacterial membrane vesicles was analyzed in the presence of inhibitors. Low concentrations of DPI and DMSO did not affect ATPase activity, whereas Met and CCCP stimulated enzyme activity. The effect of DNP was similar to CCCP. Conclusions and significance The results have shown the low concentration or concentration dependent effects of protonophores and nitrogenase and hydrogenase inhibitors on photofermentative H2 production by Rh. sphaeroides in nitrogen-limited and nitrogen-excess conditions. They would be significant to understand novel properties in relationship between nitrogenase, hydrogenase and the FOF1-ATPase in Rh. sphaeroides, and regulatory pathways of photofermentation. The inhibitors of nitrogenase and hydrogenase can be used in biotechnology for regulation of H2 production in different technology conditions and development of scale-up applications, for biomass and energy production using purple bacterial cells.

world production of H 2 is more than 50 million tons, and it is increased quickly over the world, it will decrease H 2 production costs, which can be competitive with other fuels, such as oil and natural gas [3]. Thus, shortly H 2 can become one of the main fuels in global energy economy.
During photofermentation purple bacteria can oxidize some organic carbon substrates to CO 2 , protons and electrons in tricarboxylic acid cycle (TCA) by generating NADH [7,8]. The protons are pumped through the bacterial membrane during the photosynthetic electron transport with generation of proton motive force (Δp). Under nitrogen-limited conditions or in the absence of N 2 , upon light H 2 production by purple bacteria is mainly mediated by nitrogenase, which catalyzes conversion of protons to H 2 by using energy from ATP, which is generated via the proton-translocating F O F 1 -ATPase; while hydrogenases in Rh. sphaeroides are usually involved in H 2 uptake. But these hydrogenases can be reversible depending on the conditions: the reversibility of hydrogenases might be similar to the situation with hydrogenases in Escherichia coli or cyanobacteria [1,3,8,13].
Carbonyl cyanide m-chlorophenylhydrazone (CCCP) and 2,4-dinitrophenol (DNP), protonophores functioning as uncouplers and dissipating Δp, and metronidazole (Met), a low-range electron acceptor (redox potential (E h ) is equal to −325 mV), have been shown to inhibit nitrogenase activity in cyanobacteria [14][15][16][17][18][19][20]. Diphenyleneiodonium (DPI) was established as an inhibitor of hydrogenase activity in Rh. capsulatus and Chlamydomonas reinhardtii [21,22]. Dimethylsulphoxide (DMSO) (solvent of DPI) affected the bacterial growth properties and membrane stability [23,24]. In our previous works, we demonstrated the inhibitory effects of high concentrations of DPI, DMSO and Met on H 2 production by Rh. sphaeroides strain MDC6521 isolated from Arzni mineral springs in Armenia [25,26]. The hydrogenase activity in Rh. sphaeroides and its relationship with nitrogenase and the F O F 1 -ATPase were suggested. Moreover, light and dark alternations affected H 2 production by Rh. sphaeroides [9]. However, there are no data on effects of protonophores and those inhibitors at low concentrations on H 2 production ability of Rh. sphaeroides. It is known, that many chemicals show biological effects at low and ultra-low concentrations [27,28]. This is very interesting phenomenon; however, the mechanisms of low and ultralow concentrations effects are not clear. It is interesting, how the effects of compounds used in low concentrations differ from those of relatively high concentrations and for other bacterial strains. Appropriate mechanisms of photofermentation and H 2 production are still not clear; further studies will be needed.
In the present study we have investigated the effects of protonophores such as CCCP, DNP, inhibitors of nitrogenase and hydrogenase such as Met, DPI, and DMSO at different concentrations on H 2 production ability depending on the nitrogen-limited and nitrogen-excess conditions in Rh. sphaeroides strain MDC6522, isolated from the other mineral springs in Armenian mountains-Jermuk. Novel and significant experimental data about the concentration-dependent effects of protonophores and inhibitors of nitrogenase and hydrogenase in Rh. sphaeroides have been obtained. The results would improve our understanding of mechanisms, regulatory pathways of bacterial H 2 metabolism and bioenergetics of photofermentation. They can be helpful for determining the role of various enzymes and the interaction between them in H 2 production depending on the growth conditions. Thus they might lead to optimization of the technology conditions for efficient H 2 production. Importantly, the effects might be applied in H 2 biotechnology, energy production using purple bacteria.

Effect of various inhibitors on Rh. sphaeroides growth properties
Photofermentative H 2 production by purple non-sulfur bacteria is known to be catalyzed by nitrogenase and hydrogenase. During the photosynthetic electron transport protons are pumped through the membrane with generation of Δp, which is used to generate ATP via the F O F 1 -ATPase and to transfer electrons to ferredoxin (Fd). It is known that then Fd and ATP are used to generate H 2 via nitrogenase [2,7,8]. As it was shown in our previous papers [25,26], H 2 production was strongly inhibited by high concentrations of DPI and Met. The effect of various compounds such as hydrogenase inhibitor DPI and its solvent DMSO, nitrogenase inhibitor Met, protonophore CCCP and their concentrations on growth peculiarities and photofermentative H 2 production by Rh. sphaeroides strain MDC6522, in comparison with the other strain MDC6521, isolated from Arzni mineral springs, was studied.
The growth properties were determined during anaerobic growth of Rh. sphaeroides MDC6522 upon illumination. The compounds used affected the specific growth rate of bacterial culture. Figure 2 shows a comparison of the growth specific rates of Rh. sphaeroides, grown in the presence of different inhibitors. With the increase of reagents concentrations, the specific growth rate gradually decreased. 1-2 µM DPI decreased the specific growth rate 6-to 12-folds (p < 0.001), whereas 1 mM Met suppressed growth rate ~5-fold (p < 0.001) (see Fig. 2). Rh. sphaeroides was unable to grow in the medium with CCCP and DNP, and both uncouplers were added after 24 h growth of culture, after then bacterial growth was strongly inhibited (not shown). The effect of DPI's solvent DMSO on the culture specific growth rate was also studied for revealing the inhibitory effect of DPI. In culture with 1 mM DMSO this rate was 1.3-fold (p < 0.01) lower than that of the control, whereas 5-10 mM DMSO suppressed the specific growth rate ~1.5-to 2-folds (p < 0.01) in comparison with the control (Fig. 2). These data were similar to the results on DPI and DMSO effects obtained for the other strain MDC6521 [26], but Rh. sphaeroides MDC6521 was more sensitive to the inhibitors used.

Effect of various inhibitors for enzymes on medium pH and E h during Rh. sphaeroides anaerobic growth
The growth medium pH is an important parameter for bacterial growth under different conditions [9,[29][30][31]. During the anaerobic growth of Rh. sphaeroides MDC6522 control cells up to 72 h in nitrogen-limited anaerobic conditions, the pH of medium has risen from 7.0 ± 0.2 (initial pH) up to 8.8 (Fig. 3a). This increase can be caused by the carbon source utilization and OH − ions efflux or by the polyhydroxybutyrate formation [29]. After then, during growth up to 96 h, pH decreased, which can be caused by the generation of photofermentation end-products, particularly acids, which could decay with H 2 evaluation. Also during H 2 generation the coevolved CO 2 can moderate pH change.
All inhibitors affected the pH of culture growth medium (Fig. 3). The pH value increased to ~9.0 by addition of 5-10 µM DNP (after 24 h growth) (Fig. 3a). The other kinetics of pH was observed by the addition of CCCP (after 24 h growth): pH increased to ~7.8-8.0 in the presence of 0.5-1 µM CCCP, whereas in the presence of 2 µM CCCP and 50 µM DNP pH of medium was not changed (see Fig. 3a). In the presence of 0.5-1 µM DPI the pH of medium was not changed much, and it decreased to ~6.7 in medium with 2 µM DPI (Fig. 3b). By addition of 1-5 mM DMSO pH changes during bacterial growth were similar to the control, and pH was not changed much in the presence of 10 mM DMSO (see Fig. 3b). In the presence of Met pH change during Rh. Redox potential (E h ) is another significant parameter of the bacterial growth medium, which can be determined as the ability of a biological system to oxidize or reduce different substrates [9,[31][32][33]. According to the Nernst equation E h depends on the reduced and oxidized products of fermentation, as well as on pH [29][30][31]. E h of Rh. sphaeroides MDC6522 control cells decreased to −650 ± 20 mV during growth up to 72 h in nitrogen-limited anaerobic conditions (Fig. 4a). In the medium with 5 µM DNP E h decreased to −580 ± 20 mV; whereas with 50 µM DNP E h dropped to −170 ± 10 mV only (see Fig. 4a). In contrast to DNP, the other protonophore-CCCP delayed drop in E h . The inhibition of bacterial growth may arise from the effect of CCCP (1-2 µM) on E h , which was decreased to −315 ± 20 and −200 ± 15 mV, respectively, during 72 h culture growth (see Fig. 4a). In the medium with 0.5 mM Met E h decreased to −410 ± 5 mV; whereas with 1 mM Met E h dropped to −210 ± 10 mV only (not shown). Therefore, the used compounds affect the E h in a concentrationdependent manner. Change of E h can be caused by indirect effect of these reagents on E h , or by redox processes on a surface of bacterial membrane. Thus, the negative values of E h and reduced medium are required for bacterial growth.
E h of Rh. sphaeroides str. MDC6522 control cells, grown in nitrogen-excess anaerobic conditions up to 72 h, decreased to −600 ± 15 mV (Fig. 4b). The addition of DPI into the growth medium also affected E h : in the medium with 0.5 µM DPI E h decreased to −520 ± 10 mV; whereas with 2 µM DPI E h did not change much (Fig. 4b).
1 mM DMSO increased E h up to −640 ± 20 mV. At the same time E h gradually decreased during the growth from 5 to 10 mM DMSO: E h drop was more intensive in  Fig. 4b).

Effect of various inhibitors for enzymes on H 2 photoproduction during Rh. sphaeroides anaerobic growth
The analysis of E h changes gives information not only on main redox processes but also on H 2 yield during bacterial anaerobic growth. There is a direct relationship between changes of E h and H 2 production by these bacteria; the reduction of protons to H 2 is observed under strong reducing conditions [9,31,32,34]. H 2 yield of Rh. sphaeroides MDC6522 control cells during growth up to 72 h in nitrogen-limited anaerobic conditions was 6.91 mmol H 2 (g dry weight (DW)) −1 , whereas in nitrogen-excess conditions H 2 yield was ~1.2fold lower (Fig. 5). As it can be seen from Fig. 5, four reagents used, except 1-5 mM DMSO, inhibited H 2 production by Rh. sphaeroides. With the increase of inhibitors concentrations, the H 2 yield gradually decreased. By addition of 0.1 and 0.5 mM Met H 2 yield has decreased ~1.6-fold (p < 0.01) and ~7.6-fold (p < 0.001), respectively (see Fig. 5). The effect of Met on H 2 yield might be coupled with change of photosynthetic electron transfer with Met as a preferred acceptor of electrons, instead of nitrogenase.
CCCP and DNP are two well-known protonophores, which are used to dissipate proton gradient responsible for ATP generation via the F o F 1 -ATPase and to uncouple photophosphorylation from photosynthetic electron transfer [35]. The Fig. 5 shows the H 2 yield level in the presence of various concentrations of CCCP and DNP. In medium with 1-2 µM CCCP H 2 production by bacterium has not been observed during 72 h growth, and has decreased ~5.5-fold (p < 0.001) in the presence of 0.5 µM CCCP (see Fig. 5). H 2 yield lowered ~1.6-fold (p < 0.01) and ~8.0-fold (p < 0.001) in the medium with 5 and 10 µM DNP, and was not observed in the presence of 50 µM DNP. It is suggested that protonophores can decrease H 2 generation by inhibiting of ATP synthesis by photophosphorylation, which is significant for nitrogenase-dependent photofermentative H 2 production. These data were similar to the results obtained by Skizim and co-workers for cyanobacteria Cyanothece [35].
In medium with 1-2 µM DPI, the production of H 2 by Rh. sphaeroides has not been observed during 72 h growth, and has decreased ~3.1-fold (p < 0.001) by addition of 0.5 µM DPI in comparison with the control (see Fig. 5). As shown in Fig. 5, DPI's solvent-DMSO in concentrations of 1-5 mM did not affect H 2 production by Rh. sphaeroides, whereas the high concentrations (10 mM) suppressed the H 2 yield (~3.3-fold). These results were similar to those obtained for MDC6521, and they indicated that DMSO could inhibit H 2 production depending on its concentration [26]. H 2 production by the other strain-Rh. sphaeroides MDC6521 is more sensitive to the inhibitors action. DPI, DNP and CCCP irreversibly repress the H 2 production; however DMSO and Met reversibly inhibit this process. When Met and DMSO was added into the growth medium, H 2 started to be produced after 144-168 h (not shown) growth, probably, according to the recovery of enzyme activity.

Effects of various inhibitors for enzymes on ATPase activity of Rh. sphaeroides membrane vesicles
The F O F 1 -ATPase activity of Rh. sphaeroides MDC6522 membrane vesicles was analyzed in the presence of inhibitors to reveal the role of ATPase in H 2 production. The F O F 1 -ATPase of purple bacteria belongs to F-type ATPase [36]. The membrane vesicles of bacteria, grown in the absence of inhibitors, demonstrated definite ATPase activity. By addition of 0.2 mM N,N′dicyclohexylcarbodiimide (DCCD) ATPase activity was suppressed ~2-fold (p < 0.01) (not shown). Low concentrations of DPI and DMSO did not affect the enzyme activity, whereas CCCP (2 µM) stimulated ATPase activity on 10 % (Fig. 6). Similar data were obtained by the addition of DNP (not shown) and Met (0.5 and 1 mM), enhancing this enzyme activity on ~40-45 % (see Fig. 6). This effect can be attributed to the F O F 1 -ATPase, because DCCD specifically inhibits the F O F 1 -ATPase in various bacteria [31,32]. It is possible, that these effects might be a result of inhibitors effect on ATPase via binding with enzyme and changing its activity, or on E h , which can regulate the F O F 1 -ATPase activity.

Discussion
Purple non-sulfur bacteria are the most studied photosynthetic bacteria due to their demonstrated high H 2 production yield. Two enzymes-nitrogenase and hydrogenase are involved in H 2 metabolism in these bacteria. In the anaerobic nitrogen-limited conditions, during bacterial non-oxygenic photosynthesis, organic carbon sources are oxidized to CO 2 , protons and electrons. H + can be recombined by a nitrogenase to produce H 2 using energy of ATP, which is generated by the F O F 1 -ATPase during the work of photosynthetic apparatus [2,7,8]. The participation of hydrogenase in H 2 production is also suggested under the other-nitrogen-excess conditions [1,25,26]. But appropriate mechanisms of photofermentation and H 2 production depending on conditions are not clear yet. Thereby, two aspects of photofermentative H 2 production by Rh. sphaeroides are interesting: 1st-type of enzyme (nitrogenase or hydrogenase), which is responsible for H 2 production depending on the nitrogen-limited or nitrogen-excess conditions, and the relationship between these enzymes; and 2nd-a role of the F O F 1 -ATPase in photofermentation and H 2 production by Rh. sphaeroides.
In this study the comparative analysis of protonophores and various inhibitors low concentrations effects on photofermentative H 2 production during Rh. sphaeroides MDC6522 anaerobic growth in nitrogen-limited and nitrogen-excess conditions are presented. Two possible pathways of H 2 generation in nitrogen-limited and nitrogen-excess conditions can be suggested (Fig. 7). H 2 production from various carbon sources by purple bacteria is observed in anaerobic conditions under illumination (see Fig. 1). H 2 yield of Rh. sphaeroides control cells during growth up to 72 h in nitrogen-excess anaerobic conditions was ~1.2-fold lower in comparison with nitrogen-limited conditions (see Fig. 5). In nitrogenexcess conditions nitrogenase catalyzes the reduction of N 2 to ammonia according to the reaction: N 2 + 8H + + 8e − + 16ATP → 2NH 3 +H 2 + 16ADP + 16P i , which leads to the generation of 1 mol H 2 per mole of N 2 fixed [4,7,8]. However, in nitrogen-limited conditions nitrogenase catalyzes reduction of protons to H 2 according to the reaction: 8H + + 8e − + 16ATP → 4H 2 + 16ADP + 16 Pi. Thereby, in these conditions 4 times more H 2 can be produced.
In the presence of protonophores and Met H 2 yield has decreased ~5-8-folds. It is known, that CCCP and DNP are protonophores, which dissipate the Δp and inhibit the synthesis of ATP via the F O F 1 -ATPase [14][15][16][17]35]. Indeed, these protonophores can inhibit nitrogenasedependent photofermentative H 2 production by inhibiting the synthesis of ATP (Fig. 7a). The mechanisms of the Met inhibition are not clear yet, but it is known, that Met as a low-range electron acceptor can interact with the low potential electron carriers (Fd, flavodoxin) in photosynthetic electron transfer chain. The inhibitory effects of Met on the H 2 yield in Rh. sphaeroides may be associated with dysfunction of the photosynthetic electron transport chain (Fig. 7a). Met penetrates into the bacterial cell through passive diffusion, where its nitro-group is reduced to reactive cytotoxic nitro-radicals by reduced Fd [37,38]. Fd works as electron acceptors of nitrogenase, hydrogenase and other enzymes in anaerobic bacteria. It is known, that the selective toxicity of Met for anaerobic microorganisms is due to the redox potential of their electron transport components, which are sufficiently negative to reduce the nitro-group of Met [38].
Reversibility of hydrogenases is suggested for various chemotropic and phototrophic bacteria [1,3,12,13,39], so hydrogenase in Rh. sphaeroides might be bidirectional involved in H 2 production (Fig. 7b). The complete inhibition of H 2 production by Rh. sphaeroides was observed in the presence of 1-2 µM DPI, whereas DPI's solvent-DMSO did not significantly affect H 2 yield. The results with the inhibitory effects of DPI on H 2 yield in Rh. sphaeroides provide a new evidence of involvement of hydrogenase in H 2 production by these bacteria.
Then, to understand the role of ATPase in H 2 production by Rh. sphaeroides the F O F 1 -ATPase activity of bacterial membrane vesicles was investigated. Low concentrations of DPI and DMSO did not affect the enzyme activity, whereas CCCP (2 µM) and Met (0.5-1 mM) enhanced ATPase activity on 10 % and ~40-45 %, respectively. These effects might be a result of inhibitors effect on ATPase via binding with the enzyme and changing its activity. Indeed, CCCP suppressed transfer of H + by whole cells of Rh. sphaeroides, as shown before  Fig. 6 ATPase activity of Rh. sphaeroides MDC6522 membrane vesicles, which was calculated by colorimetric determination of liberation of inorganic phosphate (P in ) per time and protein upon ATP adding (see "Methods") [40], which confirms the role of the Δp in the activity of enzymes responsible for H + transfer and H 2 production. Thus, two possible routes of H 2 production by Rh. sphaeroides can be suggested (Fig. 7). Oxidation of organic acids generates electrons, which are passed through various photosynthetic electron transfer carriers to Fd, and protons, which are pumped through the membrane generating a Δp. The latter derives the synthesis of ATP from ADP and inorganic phosphate (P i ) via the

Conclusions and significance
The data have shown low concentration or concentration dependent effects of protonophores and nitrogenise and hydrogenase inhibitors on photofermentative H 2 production by Rh. sphaeroides in nitrogen-limited and nitrogen-excess conditions. The results obtained are significant to understand the relationship between nitrogenase, hydrogenase and the F O F 1 -ATPase and their roles during photofermentation and H 2 production in Rh. sphaeroides. The relationship if any can be considered as a novel property of these enzymes. Importantly, the relationship depends on the nitrogen-excess and nitrogen-limited conditions. Thus, protonophores and nitrogenase and hydrogenase various inhibitors at different concentrations can be applied in the development of scale-up H 2 production biotechnology, for biomass and energy production using purple bacterial cells.

Bacterial strain and growth conditions
In the present work we used Rh. sphaeroides strain MDC6522 (Microbial Depository Center, Armenia, WDCM803), which was isolated from Jermuk mineral waters (pH 6.5-8.5, 57-64 °C) in Armenian mountains [9,30]. Bacteria were grown in batch culture anaerobically upon illumination (~36 W m −2 ) in Ormerod medium with succinate as a carbon source and yeast extract as a nitrogen source as described previously [9,30,31]. The growth of bacterial culture was recorded by changes in optical density (OD 660 ) using a Spectro UV-Vis Auto spectrophotometer (Labomed, USA), and by determining DW of bacterial biomass, which was correlated with OD 660 according the equation: DW (g L −1 ) = OD 660 × 0.48. The specific growth rate was calculated, as described previously [9,26,31].
In order to create conditions of nitrogen source limitation, the media was supplied with yeast extract (2 g L −1 ); whereas to create nitrogen-excess conditions the concentration of yeast extract was increased 2.5-fold (the media was supplied with 5 g L −1 yeast extract). Yeast extracts contain various amino acids, vitamins and other growth stimulating compounds and therefore it can be used as a component of growth media for the cultivation of various microorganisms [30,41].
The concentrations of DPI and CCCP added into the growth medium ranged from 0.5 to 2 µM; DNP-from 5 to 50 µM; Met-from 0.1 to 2 mM, DMSO-from 1 to 10 mM.

Determinations of pH, E h and H 2 yield
The initial pH of the culture medium was maintained to 7.0 ± 0.1 by 0.1 M NaOH or 0.1 M HCl and determined at certain time intervals (0-96 h) by a pH-meter (HANNA Instruments, Portugal) with selective pH electrode, as described [9,25,31].
The medium E h was determined during Rh. sphaeroides growth using a pair of redox (platinum (Pt) and titaniumsilicate (Ti-Si)) and reference (Ag/AgCl) electrodes, as described before [9,26,31]. Note Ti-Si electrode measures the overall E h , whereas Pt electrode (sensitive to O 2 and H 2 ) under anaerobic conditions detects only H 2 [30,32]. E h kinetics determined using redox electrodes during culture growth gives information about main redox processes and also H 2 generation [31,32]. The H 2 yield was evaluated by the drop of E h to low negative values using correlation between E h change and H 2 evolution and was expressed in mmol H 2 (g DW) −1 [9,31]: This determination of H 2 is close to the method with Clark-type electrode employed by other authors [42,43]. H 2 generation was confirmed by the chemical method, as described [26,44].

ATPase activity assay
ATPase activity of Rh. sphaeroides bacterial membrane vesicles was determined by the liberation of inorganic phosphate (P in ) in the reaction with ATP by the spectrophotometric method, as described [31,32], and it was expressed in nmol P in per L per µg protein in 1 min. Membrane vesicles were prepared by the method, as described previously [26,31]. For inhibitors effects studies, the membrane vesicles were incubated with inhibitors for 10 min.

Reagents and data processing
CCCP, Met, DPI, DNP, DMSO, DCCD, ATP were obtained from Sigma, Aldrich (USA); yeast extract was purchased from Carl Roth GmbH (Germany) and succinic acid was obtained from Unichem (China). The other reagents of analytical grade were used in this study.
Each experiment was repeated three times to determine deviations, which are displayed as error bars on figures. The standard errors were calculated and Student criteria (p) were employed to validate the difference in average data between different series of experiments, as described previously [26,31]. Dry weight (g L −1 ) .