Thermodynamics of Hydrogen Metabolism in Methanogenic Cocultures Degrading Ethanol Or Lactate

119

Thermodynamics of hydrogen metabolism in methanogenic cocultures degrading ethanol or lactate

H.-J. Seitz 1, B. Schink 2 and R. Conrad 1

I Fakultiitfiir Biologie, Unioersiti~'t Konstanz, Konstanz, F.R.G. and 2 Lehrstuhl Mikrobiologie I, Universit&'t Tiibingen, Tiibingen, F.R.G.

Key words: Hydrogen partial pressure; Proton reduction; Methanogenesis; Syntrophic coculture; Thermodynamics

1. SUMMARY 2. INTRODUCTION Anaerobic mineralization of organic matter to CO 2 and methane requires a complex food web of Pure cultures of Desulfovibrio vulgaris or Pelo- microorganisms [1,2]. Molecular hydrogen is a key bacter acetylenicus do not grow with lactate or intermediate in these degradation processes, and ethanol, respectively, under obligately proton- re- influences H z-producing as well as hydrogen-con- ducing conditions. However, a small part of these suming bacteria. It has been shown that a mini- substrates was oxidized and molecular hydrogen mum H 2 partial pressure is necessary for net H z was produced up to 4.2 and 3.2 kPa, respectively. oxidation in pure cultures of hydrogen-utilizing During growth in syntrophic methanogenic cocul- bacteria [3,4]. On the other hand, it is well estab- tures with Methanospirillum hungatei as partner, lished by pure culture studies that increased H z maximum hydrogen partial pressures were signifi- partial pressures shift the electron flow in ferment- cantly lower (0.7 to 2.5 kPa) than in the corre- ing bacteria towards reduced organic compounds sponding pure cultures. Calculation of Gibbs free [5-7] or even completely inhibit oxidation of cer- energies for the prevailing culture conditions tain substrates which can only be oxidized via showed that H 2 partial pressures were kept in a proton reduction [8-10]. range at which both, H z-producing and H z-con- H z concentrations permissive for anaerobic de- suming reactions, were thermodynamically per- gradation of various substrates have been esti- missive in pure as well as in syntrophic mixed mated from theoretical calculations [11-14] but cultures. have not yet been determined by measurements in defined cultures. We therefore measured H 2 metabolism of physiologically well-characterized fermenting bacteria under obligately proton-re- ducing conditions in pure culture as well as in Correspondence to: H.-J. Seitz, Fakult~t fiir Biologie, Uni- coculture with a Hz-utilizing methanogenic versitat Konstanz, Postfach 5560, D-7750 Konstanz, F.R.G. bacterium. The data were used to calculate the 120

Gibbs free energies of the H 2 transformation at CO 2 and pH using published equations [19]. Al- various growth phases. Our results show that H 2 cohols and volatile fatty acids were assayed by partial pressures are being kept in a range at standard gas chromatography procedures with which both, H2-producing and H2-consuming re- injector and detector temperatures being 130 or actions were thermodynamically permissive. 170°C, oven temperature being 100 or 130°C, respectively. For analysis of volatile fatty acids, samples were acidified with formic acid to a final 3. MATERIALS AND METHODS concentration of 0.5 mol formic acid/1 [16]. Lactate was determined enzymatically using lactate 3.1. Organisms and cultivation dehydrogenase (Boehringer, Mannheim) according Desulfovibrio vulgar& strain Marburg, DSM to [20]. Sulfide was determined by the methylene 2119, was kindly provided by Prof. Thauer, Mar- blue method [21]. The concentration of HS- was burg, F.R.G. Pelobacter acetylenicus strain calculated from total sulfide and the actual pH, WoAcyl, DSM 2348, and Methanospirillum assuming a pKa of 7.0. Sulfate was measured hungatei strain Mlh, were taken from the culture photometrically after precipitation with barium collection of our laboratory. chloride [22, modified after 23]. t Cells were grown at an initial pH of 7.0-7.2 in The standard Gibbs free energies (AG ° ) of the mineral medium described [15], except that H2-producing and Hz-utilizing reactions were Na2S was increased to 1.5 mmol/1. Acetate (2 calculated from the tabulated Gibbs free energies mmol/1) was added as additional carbon source if of formation (AG O of the individual reactants hydrogen was the sole electron donor. Prior to and products [24]. H 2 and CH 4 were assumed as inoculation, media were reduced with a few crystals gaseous compounds; all other compounds as dis- of sodium-dithionite (less than 0.1 mmol/1). solved. The Gibbs free energy (AG) of a reaction Precultures of hydrogen-consuming bacteria under non-standard conditions was calculated were grown in shaken 120 ml serum bottles under from its standard Gibbs free energy (AG °' ) and an atmosphere of H2/CO 2 (80/20 v/v). P. the actual partial pressures or concentrations of acetylenicus was grown on acetoin (10 mmol/1) in the reactants and products involved, the actual 50 ml screw cap bottles without a gas phase. D. temperature and H+-concentration [25]. vulgar& was cultivated similarly on lactate (40 mmol/1) and limiting amounts of sulfate (5 mmol/1). Experiments with pure cultures and 4. RESULTS defined cocultures were carried out basically as recently described [4]. Purity controls were per- 4.1. H2-metabolism in pure cultures formed microscopically at the beginning and the Pure cultures of Desulfovibrio vulgaris or Pelo- end of each experiment. bacter acetylenicus were unable to grow by fermentation of lactate or ethanol, respectively, 3.2. Analytical procedures according to the following equations [24]: Methane was measured with a Perkin Elmer gas chromatograph with flame ionization detector (1) Ethanol + H20---) acetate- + H + + 2 H 2 [16]. H 2 partial pressures above 100 Pa were (AG °' = + 9.6 kJ/mol ethanol) analyzed in a gas chromatograph (Carlo Erba) (2) Lactate- + 2 H20---) with a thermal conductivity detector. H 2 partial pressures below 100 Pa were measured using a H 2 acetate- + H + + HCO 3 + 2 H 2 analyzer based on the HgO-Hg vapour conversion (AG °'= -4.2 kJ/mol lactate) technique [17,18] with a lower detection limit of 0.2 mPa H z. CO 2 was determined with an infrared However, both strains were able to produce analyzer (UNOR, Maihak, Hamburg, F.R.G.). Bi- hydrogen which accumulated in the culture carbonate concentrations were calculated from headspace to partial pressures of up to 3.2 kPa 121

Table 1 5OOO Gibbs free energies of H2-producing and H2-consuming reac-

fo tions in pure cultures of fermentative and methanogenic o bacteria / Organism Substrate Maximum or AG /o minimum H2 (kJ/mol) "0 30OO partial pressure fit. (Pa) o Desulfovibrio lactate 4 200 - 34.6/lactate -r / 2000 o vulgaris (-sulfate) Pelobacter / acetylenicus ethanol 3200 - 9.5/ethanol 10120 Methanospirillum hungatei hydrogen 2.5 - 26.0/methane

0 ~ i i .~0 IO0 with ethanol and up to 4.2 kPa with lactate as Time [ h] electron donor (Fig. 1). With both strains, hydro- Fig. 1. Production of hydrogen from ethanol (e) or lactate (o) gen evolution ceased before the H 2 partial pres- by pure cultures of Pelobacter acetylenicus and Desulfooibrio sures reached values at which fermentation of oulgaris. The data are mean values of duplicate experiments. lactate or of ethanol would become endergonic The maximum theoretical H 2 partial pressure would be 24000 (Table 1). Pa, if the total amount of organic electron donor added (10 mmol/l) would have been fermented according to the equa- In cell suspensions of Methanospirillum hungatei tions given in RESULTS. H 2 was oxidized according to the following equa-

0,10 - 100 ¢- t~ 4000 0 0 E ~" o -8o ~, E 3000 O -60 0 .12 D 0,~. 2ooo LLI C "6 (.) - ~0 E .t o E 0 X -6 0 a 1000 - 20 £-9 -1- 2 ~ o < "6"

0 0 2 /, 6 o 2 ~ 6

Time [ d ] Time [d] Fig. 2. Fermentation of ethanol by a homogeneously mixed coculture of Pelobacter acetylenicus and Methanospirillum hungatei: (a) time course of substrate degradation and product formation, (b) change of Gibbs free energies during syntrophic growth. 122

Table 2 5. DISCUSSION Gibbs free energies of Hz-producing and H2-consuming reactions in obligately syntrophic cocultures of fermentative and In the absence of external electron acceptors, methanogenic bacteria Desulfovibrio vulgaris and Pelobacter acetylenicus Coculture Sub- H 2 partial AG behaved like typical obligately syntrophic H+-re - strate pressure (k J/tool) ducing bacteria. Thus, growth coupled to oxida- (Pa) tion of organic substrates like lactate or ethanol maxi- mini- was only possible with concomitant hydrogen re- mum mum moval. This reaction can be carried out by hydro- P. acetylenicus - 13.7/ethanol gen-oxidizing bacteria in syntrophic coculture + ethanol 2 500 2.8 [10,26-30]. Recently, effective chemical mecha- M. hungatei - 31.3/methane nisms have also been demonstrated [31]. The D. vulgaris - 34.3/lactate + lactate 1800 2.2 necessity for H 2 removal can easily be explained M. hungatei - 25.7/methane by the fact that the H2-producing reactions are endergonic or only poorly exergonic at standard conditions and pH 7.0, i.e., at a hydrogen partial pressure of approximately 100 kPa. During growth of these strains in syntrophic coculture with M. tion [24]: hungatei, H z partial pressures were about 40 to HCO~- + H + + 4 H 2---~ CH 4 + 3 H20 150 times lower (0.7-2.5 kPa) and thus always kept in a range at which hydrogen producers as × (AG °'= -135.6 kJ/mol methane) well as hydrogen consumers could obtain suffi- cient energy for growth (Fig. 2). This corresponds H 2 oxidation stopped at H 2 partial pressures at well with data based on H 2 measurements in which Ha-dependent production of methane was various methanogenic environments [25]. still exergonic (Table 1). While in homogeneously mixed cocultures H 2 partial pressures were always lower than the maxi- 4.2. H2-metabolism in syntrophic cocultures mum partial pressures in pure cultures of the In syntrophic coculture with the hydrogen- fermentative bacteria, this was different when H 2 oxidizing M. hungatei, D. vulgaris and P. producers were spatially separated from H 2 con- acetylenicus were able to grow with lactate or sumers [32]. Experiments with membrane-sep- ethanol, respectively (Fig. 2). Hydrogen partial arated cultures recently demonstrated that H 2 pressures during syntrophic substrate degradation partial pressures were dramatically higher in the were generally lower than the maximum value of compartment of the ethanol-fermenting P. the corresponding hydrogen producing strain. As acetylenicus than in the compartment of the hy- a consequence, Gibbs free energies (AG) of lactate drogen-utilizing Acetobacterium woodii [32]. Un- or ethanol fermentation usually were more ex- der these conditions H2 partial pressures were ergonic in mixed than in pure cultures (Tables 1 reached which were almost identical to the maxi- and 2). mum H~ partial pressure in pure cultures of P. At the end of the exponential growth phase, acetylenicus, presented in this study (Fig. 1). hydrogen oxidation continued in the syntrophic Hydrogen partial pressures observed in syn- culture until a minimum partial pressure was re- trophic cocultures and consequently, Gibbs free ached. Generally, this H 2 threshold partial pres- energies (AG) available for H2-producing and sure was identical to that observed in pure culture H2-consuming reactions apparently were depen- of the H2-utilizing methanogen, and did neither dent on culture conditions and in addition, depend on the H2-producing strain nor on the changed during incubation of the batch cultures. electron donor (ethanol, lactate) used for H2 pro- Therefore, it was not possible to correlate the duction (Table 2). observed Gibbs free energies (AG) with growth 123 yields of the fermenting or the methanogenic nig for many helpful suggestions and for critically bacteria. We are presently establishing syntrophic reading the manuscript. This study was supported chemostat cultures to study these relationships. by a grant of the Deutsche Forschungsgemein- However, our data demonstrate the thermody- schaft (Schwerpunkt 'Methanogene Bakterien'). namic boundary conditions of H2-producing and H2-consuming reactions. Hydrogen-producing fermenting bacteria such REFERENCES as P. acetylenicus or D. vulgar& and H2-consuming methanogenic bacteria such as M. hungatei did [1] Zehnder, A.J.B. (1978) In: Water Pollution Microbiology (R. Mitchell, ed.), pp. 349-376, John Wiley, New York. not metabolize H 2 beyond a particular partial [2] Zeikus, J.G. (1983) In: Microbes in their natural environ- pressure although actual Gibbs free energies (AG) ments (J.H. Slater, and J.W. Whittenbury, eds.), pp. would still have been exergonic. Thus, a certain 423-462, Cambridge University Press, London. minimum energy appears to be necessary for pro- [3] Lovley, D.R. (1985) Appl. Environ. Microbiol. 49, duction or consumption of H 2 in these bacteria. 1530-1531. [4] Cord-Ruwisch, R., Seitz, H.J. and Conrad, R. (1988) In syntrophic cocultures, H2 was kept in a range Arch. Microbiol. 149, 350-357. where fermentation as well as methanogenesis were [5] Kim, B.H. and Zeikus, J.G. (1985) Dev. Ind. Microbiol. exergonic by at least -13.7 to -34.3 kJ per 26, 549-556. reaction (Table 2). These Gibbs free energies (AG) [6] Zeikus, J.G. (1980) Ann. Rev. Microbiol. 34, 423-464. represent about one to two third of the AG (-44 [7] Ianotti, E.L., Kafkewitz, D., Wolin, M.J. and Bryant, M.P. (1973) J. Bacteriol. 114, 1231-1240. kJ mol-1) necessary for the reversible synthesis of [8] Johns, A.T. and Barker, H.A. (1960) J. Bacteriol. 80, one ATP and thus, are consistent with the hy- 837-841. pothesis of Thauer and Morris [33] that one third [9] Ahring, B.K. and Westermann, P. (1987) Appl. Environ. of ATP is the minimum biological energy quan- Microbiol. 53, 429-433. tum. In pure culture, P. acetylenicus produced H 2 [10] Bryant, M.P., Wolin, E.A., Wolin, M.J. and Wolfe, R.S. (1967) Arch. Microbiol. 59, 20-31. from ethanol until the energy yield was as low as [11] Wolin, M.J. (1976) In: Symposium on microbial produc- -9.5 kJ per mol of ethanol. However, ethanol tion and utilization of gases (H.G. Schlegel, G. Got- fermentation had already stopped under these tschalk, and N. Pfennig, eds.), pp. 141-150, E. Goltze, conditions. GiSttingen. The minimum Gibbs free energy (AG) of [12] Mclnerney, M.J. and Bryant, M.P. (1980) In: Anaerobes and anaerobic infections (G. Gottschalk, ed.), pp. 117-126, ethanol fermentation by P. acetylenicus was sig- Gustav Fischer Verlag, Stuttgart and New York. nificantly lower than that of lactate fermentation [13] Speece, R.E. (1983) Environ. Sci. Technol. 17, 416A-427A. by D. vulgar&, both in pure and in syntrophic [14] Harper, S.R. and Pohland, F.G. (1986) Biotechnol. Bio- mixed cultures. It is presently unknown whether eng. 28, 585-602. the maximum possible H 2 partial pressure (or [15] Widdel, F. and Pfennig, N. (1981) Arch. Microbiol. 129, minimum AG) of fermenting bacteria is species- 395-400. [16] Schink, B. and Pfennig, N. (1982) Arch. Microbiol. 133, dependent or is determined by the biochemical 195-201. degradation pathway of the substrate. Just re- [17] Seiler, W. (1978) In: Environmental Biogeochemistry and cently it has been shown that energy is required to Geomicrobiology Vol. 3, (W.E. Krumbein, ed.), pp. produce H 2 during the oxidation of lactate to 773-810. Ann Arbor Science Publishers, Ann Arbor, pyruvate in D. vulgar& [34]. In case of H2-con- Michigan. [18] Seiler, W., Giehl, H. and Roggendorf, P. (1980) Atmos. suming anaerobes, on the other hand, the mini- Technol. 12, 40-45. mum possible H 2 partial pressure (H 2 threshold) [19] Stumm, W. and Morgan, J.J. (1981) Aquatic Chemistry, was shown to decrease with increasing redox 2rid ed. John Wiley, New York. potential of the electron acceptor used [4]. [20] Bergmeyer, H.U. (Hsg) (1977) Grundlagen der enzyma- tischen Analyse, 1 Auflage. Verlag Chemie. [21] Cline, E. (1969) Limnol. Oceanogr. 14, 454-458. ACKNOWLEDGEMENTS [22] Tabatabai, M.A. (1974) Sulphur Inst. J. 10, 11-13. [23] Cypionka, H. and Pfennig, N. (1986) Arch. Microbiol. The authors are indebted to Prof. Dr. N. Pfen- 143, 396-399. 124

[24] Thauer, R.K., Jungermann, K. and Decker, K. (1977) [29] Boone, D.R. and Bryant, M.P. (1980) Appl. Environ. Bact. Rev. 41, 100-180. Microbiol. 40, 626-632. [25] Conrad, R., Schink, B. and Phelps, T.J. (1987) FEMS [30] Schink, B. (1985) Arch. Microbiol. 142, 295-301. Microb. Ecol. 38, 353-360. [31] Kaspar, H.F., Holland, A.J. and Mountfort, D.O. (1987) [26] Bryant, M.P., Campbell, L.L., Reddy, C.A. and Crabill, Arch. Microbiol. 147, 334-339. M.R. (1977) Appl. Environ. Microbiol. 33, 1162-1169. [32] Stieb, M. and Schink, B. (1987) FEMS Microbiol. Ecol. [27] Mclnerney, M.J. and Bryant, M.P. (1981) Appl. Environ. 45, 71-76. Microbiol. 41, 346-354. [33] Thauer, R.K. and Morris, J.G. (1984) Symp. Soc. Gen. [28] Mclnerney, M.J., Bryant, M.P., Hespell, R.B. and Cos- Microbiol. 36, 123-168. terton, J.W. (1981) Appl. Environ. Microbiol. 41, [34] Pankhania, I.P., Spormann, A.M., Hamilton, W.A. and 1029-1039. Thauer, R.K. (1988) Arch. Microbiol. 150, 26-31.