Hydrogen concentrations in methane-forming cells probed by the ratios of reduced and oxidized coenzyme F420

Most methanogenic archaea derive their energy for growth from the hydrogen-dependent reduction of CO₂ into methane (reaction 1). The amount of energy that can be gained in the process depends on the in situ hydrogen concentration, which may vary by orders of magnitude in natural habitats and during growth under laboratory conditions.

A central electron carrier in methane metabolism is the 8-OH-5-deazaflavin derivative coenzyme F₄₂₀. The compound is present in high concentrations. Oxidized F₄₂₀ shows an intense blue fluorescence when excited at 420 nm (DiMarco et al., 1990; Eirich et al., 1978, 1979). UVvisible light and fluorescence spectral properties are pH-dependent, making F₄₂₀ a useful probe to measure the pH inside the cell (intracellular pH or pH_i) (de Poorter & Keltjens, 2001; von Felten & Bachofen, 2000). F₄₂₀ is reduced to the non-fluorescent species (F₄₂₀H₂) by the action of F₄₂₀-reducing hydrogenase (reaction 2) (Fox et al., 1987; Thauer, 1998). F₄₂₀H₂ is the substrate in two consecutive reactions in the methanogenic pathway, viz. the reduction of N⁵,N¹⁰-methenyl-tetrahydromethanopterin (H₄MPT) and N⁵,N¹⁰-methylene-H₄MPT (reactions 3 and 4). The reactions are catalysed by F₄₂₀-dependent methylene-H₄MPT dehydrogenase and methylene-H₄MPT reductase, respectively. Reactions (24) are reversible (Thauer, 1998). The enzymes involved display high turnover numbers (k_cat) and each represents as much as 0·51 % of the total cellular protein (Enßle et al., 1991; Ma & Thauer, 1990; Schwörer & Thauer, 1991; te Brömmelstroet et al., 1990, 1991a, b). Thus, the catalytic capacities of the hydrogenase, dehydrogenase and reductase substantially exceed the specific rate of methane formation. Under these conditions, the concentration ratios of reduced and oxidized coenzyme F₄₂₀ are predicted to be in thermodynamic equilibrium with the hydrogen partial pressure ().

Taking advantage of the fluorescent properties of F₄₂₀, ratios of reduced and oxidized species were measured in H₂CO₂-metabolizing cells of Methanothermobacter thermautotrophicus and in methanol- and acetate-utilizing Methanosarcina barkeri. It was found that the ratios were, indeed, in close thermodynamic equilibrium with the hydrogen concentrations applied (02 %). For reasons discussed, this did not hold for acetate-converting Methanosarcina barkeri. The results of the study indicate that coenzyme F₄₂₀ is not only a useful probe to measure pH_i, but also to determine the in situ hydrogen concentration in H₂-metabolizing methanogens.

Materials.
Coenzyme F₄₂₀ was purified from whole cells of Methanothermobacter thermautotrophicus and cell extracts of the organism were prepared by using established procedures (te Brömmelstroet et al., 1991b). Gases were supplied by Hoek-Loos. To remove traces of oxygen, hydrogen-containing gases were passed over a BASF RO-20 catalyst at room temperature and nitrogen-containing gases over a pre-reduced BASF R3-11 catalyst at 150 °C. The catalysts were a gift from BASF Aktiengesellschaft. All other chemicals used were of the highest grade available.

Culturing methods.
Methanothermobacter thermautotrophicus (formerly Methanobacterium thermautotrophicum) strain ΔH^T=DSM 1053^T was grown at 65 °C and pH 7·0 in a 3·5 l fermenter (MBR) containing 2·5 l mineral medium and gassed with H₂/CO₂ (80 : 20 %, v/v) at 1500 r.p.m. Mineral medium contained the following constituents (g l¹): KH₂PO₄ (6·8), Na₂CO₃ (3·3), NH₄Cl (2·1), trace elements as described by Schönheit et al. (1979) and sodium resazurin (0·1 mg l¹), and cysteine hydrochloride (0·6 g l¹) and Na₂S₂O₃ (0·5 g l¹) as reducing agents. At regular time intervals, samples were collected anoxically for the determination of OD₆₀₀, F₄₂₀ measurement, pH_i determination and for cell-suspension incubations. The dissolved and medium pH were monitored online with an amperometric (Ag/Ag₂O) H₂ probe (de Poorter et al., 2003; Schill et al., 1996) and a pH electrode (Ingold, Elscolab Nederlands BV), respectively.

Alternatively, Methanothermobacter thermautotrophicus was cultured in 115 ml serum bottles containing 50 ml mineral medium supplemented with 0·6 g Na₂S.2H₂O l¹. Growth was performed at various temperatures (5065 °C) and pH values (6·07·5) to an OD₆₀₀ of 0·20·3. Incubation took place in a rotary-shaking water bath operating at 150 r.p.m. After inoculation, cultures were pressured daily with H₂/CO₂ (80 : 20 %, v/v; 200 kPa).

Methanosarcina barkeri strain Fusaro (=DSM 804) was cultured in 50 ml amounts in 115 ml serum bottles. Media were prepared as described previously (Hutten et al., 1981) and contained 10 g sodium acetate l¹ (122 mM) or 10 ml methanol l¹ (200 mM) as a carbon and energy source. Cells were grown without shaking at 35 °C under an N₂/CO₂ (80 : 20 %, v/v; 120 kPa) atmosphere to an OD₆₀₀ of 0·10·2.

Reduction of coenzyme F₄₂₀.
Purified coenzyme F₄₂₀ was reduced enzymically by using cell extract of Methanothermobacter thermautotrophicus as described previously (Vermeij et al., 1997). Reaction mixtures (3 ml) were incubated in 25 ml serum bottles under 080 % H₂, 20 % CO₂, complemented with N₂ (800 %). After reactions had come to equilibrium, anoxic acetone was added and fluorescence spectra were recorded immediately as described below.

Cell-suspension incubations.
Cells were collected from 3·5 l fed-batch cultures or were obtained from serum-bottle cultures. Inside an anaerobic glove box, 2 ml portions of cells were divided over a series of 115 ml serum bottles. Cell suspensions with an OD₆₀₀ of >1 were diluted with anoxic mineral medium. After filling, bottles were closed with butyl rubber stoppers and aluminium-crimped seals, evacuated and pressured with mixtures of H₂/CO₂ (80 : 20 %, v/v) and N₂/CO₂ (80 : 20 %, v/v) to obtain the values specified in the text. Hereafter, titanium citrate (1 mM) was added to remove traces of oxygen (Zehnder & Wuhrmann, 1976). Ethane (1 ml) was added as an internal standard for methane measurements (Gijzen et al., 1991). Serum bottles were subsequently placed in a water bath without shaking at the specified temperatures. At regular times, headspace samples were withdrawn to follow methane formation. As soon as methanogenesis had started, incubations were continued for 30 min at 150 r.p.m. (Methanothermobacter thermautotrophicus) or 100 r.p.m. (Methanosarcina barkeri) rotation. Reactions were then stopped by cooling the serum bottles rapidly in ice-cold water and samples were immediately withdrawn with a gas-tight syringe for F₄₂₀ fluorescence analysis.

Coenzyme F₄₂₀ fluorescence measurements.
A known volume of cells from the fermenter (15 ml) or from cell-suspension incubations (1 ml) was injected under anoxic conditions into a serum bottle closed with a bromobutyl rubber stopper and containing ice-cold anoxic acetone kept under N₂/CO₂ (80 : 20 %, v/v). Before use, acetone was stored overnight in an anaerobic glove box to remove traces of oxygen. Immediately afterwards, cellacetone mixtures were pipetted into cuvettes placed inside the glove box. Cuvettes were closed with bromobutyl stoppers and the contents were analysed by anaerobic fluorescence spectroscopy. This gave the fluorescence intensities of oxidized F₄₂₀ present in the samples (F_ox). To determine the fluorescence of total coenzyme F₄₂₀ (F_tot), cell samples were mixed, after brief exposure to air, with oxic acetone and spectra were measured under aerobic conditions. To correct for background fluorescence (F_b), cell samples were incubated under (H₂/CO₂) 80 : 20 % at 65 °C, added to cold anoxic acetone and measured anaerobically.

Fluorescence emission was recorded at room temperature on an Aminco SPF-500 fluorimeter with excitation wavelength at 427 nm (band pass, 4 nm) and emission wavelength at 471 nm (band pass, 2 nm). Alternatively, excitation spectra (340470 nm) were recorded at an emission wavelength of 471 nm. The concentration ratios of F₄₂₀H₂ and F₄₂₀ were calculated as (F_totF_ox)/F_ox. The experimental values (F_tot, F_ox) were corrected for background fluorescence (F_b) measured for the fully (80 % H₂) reduced cell samples. Acetone extracts were alkaline (pH 910). Under these conditions, oxidized F₄₂₀ is measured exclusively as the phenolatequinoid anionic species (see Appendix).

Other analytical methods.
Methane-production rates during the fermenter culturing were calculated from the flow rate and methane content of the outflow gas, which were measured by use of a soap-film meter and by GC, respectively. GC was performed on an HP 5890 gas chromatograph equipped with a Poropak Q column and a flame-ionization detector. Cellular dry weights (DW) to determine specific methane-forming activities were derived from the OD₆₀₀ value of the culture. Previous research established the linear relation between both parameters, at which 1 l culture showing an OD₆₀₀ of 1 equalled 425 mg dry cells (unpublished results). pH_i values were measured by a previously described method, using the pH-dependent fluorescence properties of oxidized coenzyme F₄₂₀ (de Poorter & Keltjens, 2001).

Hydrogen-dependent reduction of coenzyme F₄₂₀
F₄₂₀ was incubated in the presence of cell extract in a series of serum bottles under different (080 %; 00·8 bar), and fluorescence-excitation spectra were recorded after reactions had come to equilibrium (Fig. 1). F₄₂₀ incubated under an N₂/CO₂ atmosphere (80 : 20 %, v/v) showed maximal fluorescence emission at 427 nm excitation. The same fluorescence intensities of H₂-incubated reaction mixtures were found after exposure to air or after mixing with aerobic acetone. Incubations at increased hydrogen concentrations resulted in the concomitant decrease of the excitation spectra, characteristic of F₄₂₀ reduction. Under 80 % H₂, the spectrum was bleached almost completely. Concentration ratios of F₄₂₀H₂ and F₄₂₀, determined as described in Methods, were related linearly to the applied (Fig. 1, inset). From the slope of the plot and by using equation (A.8) from the Appendix, a ΔG⁰'of 15 kJ mol¹ was calculated at the experimental conditions (pH 7·0, 60 °C). Remarkably, the same value of ΔG⁰' was found under standard conditions (pH 7·0, 25 °C).

(21K): Fig. 1. Hydrogen-dependent reduction of coenzyme F420. F420 (5 µM) was reduced at the indicated values (%, v/v), using cell-free extract (15 µg protein) as described in Methods. Reactions were performed at 60 °C and pH 7·0. Excitation spectra were recorded at 471 nm emission. In the inset, concentration ratios of reduced (F420H2) and oxidized F420 are plotted against the applied hydrogen partial pressures (). a.u., Arbitrary units.

Whole cells incubated under hydrogen revealed excitation and emission spectra that were indistinguishable from those obtained for purified F₄₂₀ (data not shown). This demonstrated that other cellular components did not interfere with F₄₂₀ fluorescence measurements. The fluorescence characteristics were subsequently used to determine the concentration ratios of reduced and oxidized F₄₂₀ in metabolizing cells.

Coenzyme F₄₂₀ reduction in methane-forming cell suspensions of Methanothermobacter thermautotrophicus
To investigate the effect of the applied on coenzyme F₄₂₀ reduction in methane-producing cells, cell suspensions of Methanothermobacter thermautotrophicus were incubated under a variety of conditions and at 02 % (v/v) hydrogen in the gas phase (, 00·02 bar). Cell suspensions were obtained from different growth stages in the fed-batch fermenter (see below) or from serum-bottle cultures. At low , the specific rates of methanogenesis in the suspension incubations were linearly dependent on the applied. Specific activities at a of 0·02 bar were 550 % of the maximal values measured at 80 % H₂ [13 µmol CH₄ min¹ (mg DW)¹]. The former percentages depended on the hydrogen concentration at which growth had occurred and reflect changes in the affinities (K_m) of the cells for hydrogen. It is known that Methanothermobacter thermautotrophicus cells derived from cultures grown under low-hydrogen conditions display a higher hydrogen affinity (K_m approx. 2 % H₂) than cells grown at a high hydrogen concentration (K_m approx. 20 % H₂) (Pennings et al., 2000). In addition, maximal specific activities of the cultured cells varied in a growth phase- and growth condition-related way (Pennings et al., 2000; L. M. I. de Poorter & J. T. Keltjens, unpublished observations). This explains the differences in values measured at 80 % H₂ during the suspension incubations.

When cell suspensions collected from different growth stages in the fed-batch fermenter were incubated at 60 °C and pH 7, a linear relationship was found between the [F₄₂₀H₂]/[F₄₂₀] ratios and the values applied (Fig. 2). Slopes of the graphs measured with cells from different growth stages were identical. The massaction ratio was associated with RT ln q_r at +15 kJ mol¹. Above data established a ΔG_r⁰' of 15 kJ mol¹ at 60 °C and pH 7. From the resulting ΔG_r' of 0 kJ mol¹ (equation A.1), it is inferred that the concentrations of reduced and oxidized coenzyme F₄₂₀ within the cells are in thermodynamic equilibrium with the in the gas phase.

(11K): Fig. 2. Effect of on coenzyme F420 reduction in Methanothermobacter thermautotrophicus. Cell suspensions were incubated under 20 % CO2 and the indicated gas-phase values. Incubations took place at 60 °C and pH 7·0 as described in Methods. Suspensions were collected from fed-batch fermenter cultures at the different growth phases. Symbols: , early-exponential phase (OD600=0·2); , exponential phase (OD600=0·6); , linear phase (OD600=2); , stationary phase (OD600=6).

To investigate the effect of temperature and pH on the hydrogen-dependent reduction of coenzyme F₄₂₀, Methanothermobacter thermautotrophicus was cultured in serum bottles at a range of temperatures (5065 °C) and pH values (6·07·5). Cells were subsequently incubated under various , using medium pH values and temperatures at which culturing had occurred. Separate incubations were performed to measure the pH_i after incubation. At the experimental conditions, pH_i was found to be equal to the medium pH. As before, [F₄₂₀H₂]/[F₄₂₀] ratios were related linearly to the values applied (Figs 3a and 4a). Slopes were pH-dependent and an approximately tenfold decrease in the massaction ratio was observed when medium pH increased by 1 unit (Fig. 3a). This indicates that coenzyme F₄₂₀ reduction is described by equation (5), in which F₄₂₀ refers to (deprotonated) phenolate anion (Fig. 7):

(14K): Fig. 3. Effect of pH on the thermodynamics of F420 reduction in Methanothermobacter thermautotrophicus. (a) Cell suspensions were incubated at 60 °C under 20 % CO2 and the indicated gas-phase values as described in Methods. Reactions took place at the following pH values of the medium: 6·3 (), 6·5 (), 6·8 (), 7·1 () or 7·2 (•). Data represent the means and errors of triplicate experiments. (b) Plots of massaction ratio terms (RT ln q) () derived from the slopes presented in (a), calculated pH-dependent ΔG0 values (•) and net Gibbs free-energy changes (ΔG) ().

(11K): Fig. 4. Effect of temperature on the thermodynamics of F420 reduction in Methanothermobacter thermautotrophicus. (a) Cell suspensions were incubated at pH 7·0 under 20 % CO2 and the indicated headspace values as described in Methods. Reactions took place at the following temperatures: 50 °C (), 55 °C () and 65 °C (). Data represent the means and errors of triplicate experiments. (b) Plots of massaction ratio terms (RT ln q) () derived from the slopes presented in (a), calculated ΔG0 values (•) and net Gibbs free-energy changes (ΔG) (). The massaction ratio at 60 °C was obtained from data presented in Fig. 2.

(19K): Fig. 7. Structure of coenzyme F420 and its (de)protonation and redox reactions.

The plot of RT ln q_r versus pH gave a straight line (Fig. 3b). The slope (6·4 kJ mol¹ pH¹) at the incubation temperature (60 °C) was in full agreement with the net uptake of one proton. By use of equation (A.6) and the experimental ΔG⁰' of 15 kJ mol¹at pH 7, ΔG₆₀⁰ could be calculated for the different pH values (Fig. 3b). Again considering that the Gibbs free-energy change at 60 °C (ΔG₆₀) sums as ΔG₆₀⁰+RT ln q_r (equation A.1), a ΔG₆₀ of 0 kJ mol¹ was derived for all pH values tested, indicative of thermodynamic equilibrium (Fig. 3b). When incubated at pH 7, massaction ratios varied with the incubation temperatures (5065 °C), but the RT ln q_r term was constant (+15 kJ mol¹) and exactly opposite to the (temperature-independent) ΔG⁰' of 15 kJ mol¹, again demonstrating thermodynamic equilibrium (ΔG'=0) (Figs 4a and b).

By routine, cell-suspension incubations were performed at relatively low values (00·02 bar). When incubated at higher headspace-hydrogen concentrations, large variations in [F₄₂₀H₂]/[F₄₂₀] ratios were found among repeated experiments and the ratios were generally lower than expected. At the higher values, methane production and, in direct connection, hydrogen uptake took place at correspondingly enhanced rates. The consumption of dissolved hydrogen during the brief but variable period between rotary incubation and cooling of the samples (515 s) probably caused the variation in and underestimation of the [F₄₂₀H₂]/[F₄₂₀] ratios.

Coenzyme F₄₂₀ reduction in methanol- and acetate-metabolizing cell suspensions of Methanosarcina barkeri
Methanosarcina barkeri was grown in serum bottles on methanol (200 mM) or acetate (122 mM) as substrates to an OD₆₀₀ of 0·10·2. At this time, cultures still contained approximately 150 mM methanol and 80 mM acetate, whilst methane was formed with specific activities of 0·4 and 0·1 µmol min¹ (mg DW)¹, respectively. Portions (2 ml) of the cultures were subsequently incubated under 080 % hydrogen at 35 °C. Determination of the [F₄₂₀H₂]/[F₄₂₀] ratios revealed a linear relationship between the ratios and the values applied in the case of methanol-grown cells (Fig. 5). From the slope of the curve, an RT ln q_r of +15 kJ mol¹ could be calculated, which equals the above-determined values. From this, we conclude that methanol-metabolizing Methanosarcina barkeri cells maintain their [F₄₂₀H₂]/[F₄₂₀] ratios in thermodynamic equilibrium with the in the environment. In acetate-grown cells, the situation was different. Although clearly detectable by the sensitive fluorescence method used, the F₄₂₀ content was lower by more than a factor of ten than that in methanol-grown cells. Moreover, coenzyme F₄₂₀ was only present in the oxidized state [(F₄₂₀H₂)/(F₄₂₀)=0] (Fig. 5), even if incubations were performed under high hydrogen concentrations (up to 80 %).

(10K): Fig. 5. Effect of on coenzyme F420 reduction in Methanosarcina barkeri. Methanol () and acetate ()-metabolizing cell suspensions were incubated under 20 % CO2 and the indicated headspace values. Reactions took place at 37 °C and pH 7·0 as described in Methods.

Changes in the ratios of reduced and oxidized coenzyme F₄₂₀ during growth of Methanothermobacter thermautotrophicus in a fed-batch fermenter
Methanothermobacter thermautotrophicus was cultured in a fed-batch fermenter at constant gassing with 80 : 20 % H₂/CO₂ (Fig. 6). Growth was characterized by an exponential increase of cell density up to an OD₆₀₀ of 1·7 [specific growth rate, 0·24 h¹; doubling time (t_d), 2·9 h]. Hereafter, cell density increased linearly with time. During exponential growth, methane was formed with a specific activity of 1·52·5 mol min¹ (mg DW)¹. Considering that 4 mol hydrogen is used (mol methane formed)¹ (equation 1), the specific hydrogen-consumption rate amounted to 610 mol min¹ mg¹. Together with the increase in biomass, the overall hydrogen-consumption rate increased tenfold (0·66 mmol min¹). The increase in hydrogen consumption was accompanied by the decline in the dissolved from 70 to 3 % (0·7 to 0·03 bar). Remarkably, the intracellular pH of the cells decreased as well, in particular during the mid-exponential phase (Fig. 6). During the linear-growth phase (1012 h), the hydrogen-consumption rate and became constant at 6 mmol min¹ and 0·03 bar, respectively. Now, pH_i was about equal to the medium pH of 7·0.

(14K): Fig. 6. Changes in the concentration ratios of reduced and oxidized coenzyme F420 during growth of Methanothermobacter thermautotrophicus in a fed-batch fermenter. The organism was grown under 80 : 20 % H2/CO2 at a constant gassing rate of 218 ml min1. Culturing took place at 65 °C and pH 7·0 as described in Methods. Measurements started (t=0) 12 h after inoculation. Symbols: , OD600; •, in growth medium; , intracellular pH (pHi); , experimental concentration ratios of reduced and oxidized F420 (means and errors of triplicate fluorescence measurements); , [F420H2]/[F420] ratios (assuming thermodynamic equilibrium).

At regular time intervals, cells were collected anoxically from the fermenter and analysed for the [F₄₂₀H₂]/[F₄₂₀] ratios (Fig. 6). The apparent ratios tended to decrease, but became somewhat higher during the linear-growth phase. From the recorded and pH_i values, [F₄₂₀H₂]/[F₄₂₀] ratios were calculated theoretically, assuming thermodynamic equilibrium. It can be seen that experimental and theoretic ratios were about equal during the early-exponential (03 h) and linear (1012 h) phases, where was as high as 0·70 bar and as low as 0·03 bar, respectively. During the intermediary-exponential phase, however, experimental [F₄₂₀H₂]/[F₄₂₀] ratios were five- to 15-fold lower than the theoretical values. It is conceivable that, during this stage, the hydrogen concentration inside the cells was lower than that in the medium. However, at least part of the difference could be due to an underestimation of the [F₄₂₀H₂]/[F₄₂₀] ratios as a result of the sampling procedure. Sampling included the passage of the culture liquid through the device interconnecting the fermenter and the acetone-containing sample bottle, which took about 5 s. During the passage, a substantial part of the dissolved hydrogen could have been utilized, especially at high cellular hydrogen-uptake rates and at high medium , conditions that typically apply to the exponential phase. Indeed, when acetone mixtures were analysed by GC for dissolved hydrogen, levels in samples collected during the intermediary-exponential phase were lower by a factor of 515 than measured with the hydrogen probe. In contrast, GC determinations on liquids from early-exponential and linear-phase cells agreed well with those recorded in the fermenter (data not shown). Hydrogen-metabolizing cells of Methanothermobacter thermautotrophicus consistently maintained the concentration ratios of reduced and oxidized coenzyme F₄₂₀ in thermodynamic equilibrium with the , if below approximately 0·02 bar. However, equilibrium was also observed at values as high as 0·7 bar and at high specific hydrogen-consumption rates (see Fig. 6, early-exponential phase). Therefore, the relationship could be valid for all conditions, but this could not be substantiated by the method applied, due to the time delay in our sampling procedure. Online in situ fluorescence measurements might clarify this issue.

In the temperature range tested (2565 °C), the standard free-energy change at pH 7 related to the hydrogen-dependent reduction of coenzyme F₄₂₀ was constant (ΔG⁰', 15 kJ mol¹). As the midpoint potential of the H⁺/H₂ couple varies with temperature, E_m,F for the F₄₂₀/F₄₂₀H₂ couple has to show the same temperature dependency. On the basis of the experimental ΔG⁰' values, the H⁺/H₂ midpoint potentials and by using equation (A.7), E_m values of 340 and 385 mV are then calculated for the F₄₂₀/F₄₂₀H₂ couple at 25 and 60 °C, respectively, by the biochemical assay described here. The former value equals reported data (340 to 350 mV) determined at ambient temperature by electrochemical methods (Jacobson & Walsh, 1984; Pol et al., 1980).

Thermodynamic equilibrium was also found in methanol-utilizing Methanosarcina barkeri cells. This is remarkable, as the conversion of methanol into methane and CO₂ does not involve hydrogen (equation 6).

However, methanol-grown cells contain high levels of F₄₂₀-reducing hydrogenase (Michel et al., 1995), whilst F₄₂₀ serves as the electron carrier in two reactions of the methyl group-oxidation pathway, notably N⁵-methyl-H₄MPT and N⁵,N¹⁰-methylene-H₄MPT oxidation (reversed reactions 3 and 4) (Enßle et al., 1991; Schwörer & Thauer, 1991; te Brömmelstroet et al., 1991a; Thauer, 1998). During growth on methanol, the compound serves as both the energy and carbon source. As cell carbon is formally more oxidized than that in methanol, anabolism is associated with a net electron production. It is conceivable that the generation (or consumption) of hydrogen gas is required to balance electron flows in catabolic and anabolic reactions at which F₄₂₀-hydrogenase could act as a redox valve. Indeed, it is known that Methanosarcina growing on methanol accumulates small concentrations of hydrogen gas in the gas atmosphere (Lovley & Ferry, 1985). In contrast, acetate catabolism does not involve F₄₂₀-dependent reactions. Under these conditions, F₄₂₀-reducing hydrogenase, as well as F₄₂₀-dependent N⁵,N¹⁰-methylene-H₄MPT dehydrogenase and reductase, are repressed (Schwörer & Thauer, 1991; Vaupel & Thauer, 1998). As expected for a limited role in cellular metabolism, F₄₂₀ is present at only low levels (Heine-Dobbernack et al., 1988; this study). Furthermore, it was found here that hydrogen had no effect on the F₄₂₀ reduction state during acetate metabolism. Apparently, hydrogen does not equilibrate with the intermediary F₄₂₀ metabolism, serving now only some specific anabolic steps.

In nature, methanogenic archaea form part of densely packed, complex microbial consortia that degrade organic matter into methane and CO₂ (Zinder, 1993). Hydrogen is a central intermediate in the degradation and the gas is presumably present as steep spatial-concentration gradients. Detailed understanding of the processes will require methods to measure in situ hydrogen concentrations within the microsystems. By taking advantage of its fluorescent properties, coenzyme F₄₂₀ could serve as a probe to assess hydrogen concentrations by using, for example, non-invasive laser techniques.

Theory
Equation (2) in the Introduction formally describes the reduction of coenzyme F₄₂₀ into 1,5-dihydro-F₄₂₀ (F₄₂₀H₂) with hydrogen. The (Gibbs) free-energy change, ΔG_r (kJ mol¹), at specified reaction conditions (suffix r; temperature, pH) of the reaction is:
in which R is the gas constant (8·314·10³ kJ mol¹ K¹), T is the absolute temperature (K) and q_r is the massaction ratio:

q_r equals the slope in the experimental [F₄₂₀H₂]/[F₄₂₀] versus plots. It should be noted that [F₄₂₀] and [F₄₂₀H₂] represent total concentrations of the oxidized and reduced species, respectively. In the physiological pH range, the 5-deazaflavin chromophore of oxidized coenzyme F₄₂₀ contains one ionizable group, viz. 8-OH (pK_a1 6·186·47, depending on the temperature) (Jacobson & Walsh, 1984; Purwantini et al., 1992). Deprotonation of 8-OH results in the phenolate anion, which tautomerizes into the conjugated paraquinoid anion (Fig. 7). In (non-fluorescent) reduced F₄₂₀, NH(1) (pK_a2 6·9) and the 8-hydroxyl group (pK_a1' 9·7) are of relevance. Thus, oxidized and reduced F₄₂₀ are composed of a mixture of species that will affect the redox potential of the F₄₂₀/F₄₂₀H₂ couple in a pH-dependent fashion.

In addition, the free-energy changes of coenzyme F₄₂₀ reduction with hydrogen will vary with the pH:
Defining ΔG_r⁰' (kJ mol¹) as the free-energy change at pH 7 and at the temperature at which the reaction is followed and m as the net number of protons that are consumed or produced per reaction, the following relations hold:

In equation (A.6), the sign of the term is minus in a proton-consuming reaction. In equation (A.7), n is the number (2) of electrons involved, F is the Faraday constant (96·49 kJ V¹ mol¹) and ΔE_m,7 is the difference between the midpoint potentials (V) of the H⁺/H₂ (E_m,H) and F₄₂₀/F₄₂₀H₂ (E_m,F) redox couples, respectively, at pH 7 and the specified temperature. E_m,H is derived for each given temperature from the Nernst equation: E_m,H=2·303(7RT/F). E_m,F should be measured, or it can be calculated if ΔG_r⁰' (at pH 7) is known. The latter can be determined from the reaction at equilibrium. Considering that, under these conditions, ΔG_r=0 and that the massaction ratio (q_r) equals the equilibrium constant K_r, it follows from equation (A.1):
The work of L. M. I. de P. was supported by the Life Sciences Foundation (ALW), which is subsidized by the Netherlands Organization for Scientific Research (NWO). Dr A. P. R. Theuvenet of the Department of Cell Biology of the University of Nijmegen is greatly acknowledged for helpful advice and for usage of the fluorimetric equipment.

Footnotes

†Present address: Department of Biotechnology, Delft University of Technology, Delft, The Netherlands.