The Wolfe cycle comes full circle

Rudolf K. Thauer1 Max Planck Institute for Terrestrial Microbiology, D-35043 Marburg, Germany

n 1988, Rouvière and Wolfe (1) H - ΔμNa+ 2 CO2 suggested that methane formation + MFR from H and CO by methanogenic + 2H+ *Fd + H O I 2 2 ox 2 archaea could be a cyclical process. j O = Indirect evidence indicated that the CoB-SH + CoM-SH fi *Fd 2- a rst step, the reduction of CO2 to for- red R mylmethanofuran, was somehow coupled + * H MPT 2 H2 Fdox 4 to the last step, the reduction of the het- h erodisulfide (CoM-S-S-CoB) to coenzyme CoM-S-S-CoB b MFR M (CoM-SH) and coenzyme B (CoB-SH). H Over 2 decades passed until the coupling C 4 10 mechanism was unraveled in 2011: Via g flavin-based electron bifurcation, the re- CoB-SH duction of CoM-S-S-CoB with H provides 2 H+ the reduced ferredoxin (Fig. 1h) required c + Purines for CO2 reduction to formylmethanofuran ΔμNa + H MPT 4 f H O (2) (Fig. 1a). However, one question still 2 remained unanswered: How are the in- termediates replenished that are removed CoM-SH for the biosynthesis of cell components H Methionine d from CO2 (orange arrows in Fig. 1)? This Acetyl-CoA e anaplerotic (replenishing) reaction has F420 F420H2 recently been identified by Lie et al. (3) as F420 F420H2 the sodium motive force-driven reduction H i of ferredoxin with H2 catalyzed by the i energy-converting hydrogenase EhaA-T H2 (green arrow in Fig. 1). H2 Thymidylate Coenzyme A Biological methane formation from H2 and CO2 (ΔG°′ = −131 kJ/mol) is not only a quantitatively important process but Fig. 1. Wolfe cycle of CO2 reduction to methane with 4 H2 in hydrogenotrophic methanogenic ar- possibly one of the oldest (4). It involves chaea. Orange arrows indicate biosynthetic reactions that remove intermediates(thicknessofthear- rows reflects the quantitative importance), the green arrow illustrates the anaplerotic reaction such coenzymes as methanofuran, tetrahy- catalyzed by EhaA-T, and yellow highlighting represents the electron-bifurcating reaction. Reactions f dromethanopterin (H4MPT), and CoM-SH and j are catalyzed by membrane-associated enzyme complexes. All other reactions are catalyzed by

as C1-unit carriers. These coenzymes were cytoplasmic enzymes. F420,coenzymeF420;*Fd,specific ferredoxin; H4MPT, tetrahydromethanopterin; first thought to be unique to anaerobic MFR, methanofuran; ΔμNa+, electrochemical sodium ion potential. Enzymes: a, formylmethanofuran methanogenic archaea (5) but were later dehydrogenase; b, formylmethanofuran/H4MPT formyltransferase; c, methenyl-H4MPT cyclohydrolase; found also in methanotrophic bacteria (6) d, methylene-H4MPT dehydrogenase; e, methylene-H4MPT reductase; f, methyl-H4MPT/coenzyme M and in some other microorganisms (7, 8). methyltransferase; g, methyl-coenzyme M reductase; h, electron-bifurcating hydrogenase–heterodisulfide CO reduction to methane begins with its reductase complex; i, F420-reducing hydrogenase; j, energy-converting hydrogenase catalyzing the so- 2 dium motive force-driven reduction of ferredoxin with H .WithMethanothermobacter marburgensis, reduction to formylmethanofuran with 2 − it has been shown that one of the two C1 carbons in purines is derived from formate (C-2) and the other reduced ferredoxin (Fd 2 )(9–11), which red is derived from formyl-H4MPT/methenyl-H4MPT (C-8) (25). is regenerated by reduction of oxidized ferredoxin (Fdox)withH2 (12). The redox − − potential of the Fd /Fd 2 couple of which, in turn, drives both ATP synthesis 55 kJ/mol) catalyzed by a cytoplasmic ox red – fi −500 mV is almost 200 mV more negative via the membrane-associated AoA1-ATP hydrogenase heterodisul de reductase + synthase (not shown) and ferredoxin re- complex (Fig. 1h), which couples this re- than that of the 2H /H2 couple at a pH of 7 and the H partial pressure of 10 Pa duction with H2 via the membrane-asso- action with the endergonic reduction of 2 ΔG ′ prevailing in the anaerobic habitats of ciated and energy-converting hydrogenase ferredoxin with H2 ( ° = +16 kJ/mol) hydrogenotrophic methanogens (11). The complexes EhaA-T (Fig. 1j) and EhbA-Q and renders the methanogenic pathway endergonic reduction of ferredoxin (not shown) (11). cyclical (2). Coupling is via the newly dis- covered mechanism of flavin-based elec- with H2 must therefore somehow be cou- The second exergonic reaction is the pled to one of the three downstream reduction of methyl-coenzyme M with tron bifurcation (16). exergonic reactions. CoB-SH to methane and the hetero- Using elegant genetic experiments, Lie The first exergonic reaction is the disulfide CoM-S-S-CoB (ΔG°′ = −30 et al. (3) show that when the genes for six transfer of the methyl group from methyl- kJ/mol) catalyzed by cytoplasmic methyl- of the seven chromosomally encoded H4MPT to CoM-SH (ΔG°′= −30 kJ/mol), coenzyme M reductase (Fig. 1g). As far as which is catalyzed by the membrane-asso- known, this reaction is not coupled with ciated and energy-conserving enzyme ferredoxin reduction or energy conserva- Author contributions: R.K.T. wrote the paper. complex MtrA-H (13) (Fig. 1f). The re- tion (14, 15). The author declares no conflict of interest. action is associated with the build-up of an The third exergonic reaction is the re- See companion article on page 15473. 1 electrochemical sodium ion potential, duction of CoM-S-S-CoB with H2 (ΔG°′ = E-mail: [email protected].

15084–15085 | PNAS | September 18, 2012 | vol. 109 | no. 38 www.pnas.org/cgi/doi/10.1073/pnas.1213193109 Downloaded by guest on September 26, 2021 COMMENTARY

hydrogenases of Methanococcus mar- the reduction of CO2 with H2 to methane which catalyzes the sodium motive force- ipaludis are deleted, this model organism proceeds only in the presence of cata- driven reduction of ferredoxin with H2, can still grow on formate, but only when lytical amounts of methyl-coenzyme M. has a purely anabolic function and cannot H2 is also present. Only the energy-con- Later, it was found that methyl-coenzyme substitute for the EhaA-T complex cata- verting hydrogenase EhaA-T appears to M could be substituted by catalytical lyzing the same reaction. This can be be essential. During growth on formate, amounts of CoM-S-S-CoB (19). This explained by assuming that the two hy- the hydrogenase–heterodisulfide re- finding was reminiscent of the observa- drogenases prefer to use different fer- ductase complex (Fig. 1h) is substituted tion of Hans Adolf Krebs (20) that in redoxins. Indeed, in the genome of all by an electron-bifurcating formate de- liver slices, the formation of urea from methanogens analyzed, several genes for hydrogenase–heterodisulfide reductase ammonia and CO2 is stimulated by cata- ferredoxins with quite different predicted complex and the F420-reducing hydroge- lytical amounts of ornithine, which led to structures are encoded (23). An alterna- nase (Fig. 1i) is substituted by F420-re- the discovery of the first metabolic cycle, tive explanation is that ferredoxin pools ducing formate dehydrogenase. The H2 is the urea cycle. This was followed by are compartmentalized. The finding that used in substoichiometric amounts (< 0.2 Krebs’ elucidation of the tricarboxylic the hydrogenase–heterodisulfide re- mol of H2 per mol of CH4) that approxi- acid cycle for acetate oxidation. Later, ductase complex (Fig. 1h) and the for- mately comprise the amount of H2 re- Hans L. Kornberg (21) showed that ana- mylmethanofuran dehydrogenase complex quired for autotrophic CO2 fixation, which plerotic reactions were necessary to re- (Fig. 1a) form super complexes points in suggests an anaplerotic role of EhaA-T. plenish the intermediates of the Krebs this direction (24). Why is an anaplerotic reaction so im- cycle used for biosynthesis. With the How widespread is the Wolfe cycle? portant? It is because intermediates are anaplerotic reaction found by Lie et al. Biochemical and genomic information continuously withdrawn from the cycle for (3), the cycle of CO2 reduction to meth- indicates that with modifications, the the biosynthesis of purines (DNA and ane is finally complete. This methano- cycle operates in all members of the RNA); thymidylate (DNA); CoA; methio- genic cycle is now referred to as the Methanobacteriales, Methanopyrales, nine (protein and S-adenosyl-methionine); Wolfe cycle, in honor of Ralph S. Wolfe Methanococcales, Methanomicobiales, and, quantitatively most important, acetyl- (22), who has been the motor driving and Methanocellales that can grow on CoA (Fig. 1), which is synthesized from its elucidation. H2/CO2 and/or formate. Only the rela- methyl-H4MPT, CO, and CoA in autotro- One can argue that the Wolfe cycle is tively few members of the Meth- phic methanogens (17). If reduced ferre- not a cycle comparable to the Krebs cycle anosarcinales that grow on H2 and CO2 doxin is not replenished, the cycle would because the first and last steps are coupled are exceptions, although these members come to a halt. The same would happen if via the electron carrier ferredoxin rather appear to contain all the genes required electron bifurcation is imperfectly coupled than by a carbon compound. Also, the for the operation of the cycle. In Meth- (11). However, imperfect coupling appears Embden–Meyerhof pathway would be anosarcina barkeri, for example, the re- not to be of quantitative importance, as a cycle if formulated with ATP as an in- duction of CoM-S-S-CoB is catalyzed by evidenced by the relatively low H2 re- termediate connecting the first step and two membrane-associated enzyme com- quirement during growth on formate of the last step, namely, the hexokinase re- plexes and is associated with the build-up the M. maripaludis mutant lacking the six action and the pyruvate kinase reaction. of an electrochemical proton potential, hydrogenases. This finding is another im- However, the Wolfe cycle differs from the which, in turn, drives the reduction of portant result of the work of Lie et al. (3). Embden–Meyerhof pathway in that the the ferredoxin required for CO2 reduc- The elucidation of the cycle all began ferredoxin involved appears to be used tion to formylmethanofuran via an in 1977 with an experiment by Gunsalus mainly in the cycle, whereas the ATP of energy-converting hydrogenase (11). In and Wolfe (18), which showed that in cell the Embden–Meyerhof pathway is also M. barkeri during growth on H2 and CO2, extracts of Methanobacterium thermoau- used elsewhere. Evidence for a specific the function of the energy-converting totrophicum strain ΔH (now Meth- ferredoxin comes from the finding that the hydrogenase is therefore catabolic rather anothermobacter thermoautotrophicus), energy-converting hydrogenase EhbA-Q, than anaplerotic.

1. Rouvière PE, Wolfe RS (1988) Novel biochemistry of Methanobacterium thermoautotrophicum. Eur J Bio- 17. Berg IA, et al. (2010) Autotrophic carbon fixation in methanogenesis. J Biol Chem 263:7913–7916. chem 226:811–818. archaea. Nat Rev Microbiol 8:447–460. 2. Kaster AK, Moll J, Parey K, Thauer RK (2011) Coupling 10. Vorholt JA, Thauer RK (1997) The active species of 18. Gunsalus RP, Wolfe RS (1977) Stimulation of CO2 re- of ferredoxin and heterodisulfide reduction via elec- ‘CO2’ utilized by formylmethanofuran dehydrogenase duction to methane by methylcoenzyme M in extracts tron bifurcation in hydrogenotrophic methanogenic from methanogenic Archaea. Eur J Biochem 248: Methanobacterium. Biochem Biophys Res Commun 76: – – archaea. Proc Natl Acad Sci USA 108:2981 2986. 919–924. 790 795. 3. Lie TJ, et al. (2012) Essential anaplerotic role for the 11. Thauer RK, Kaster AK, Seedorf H, Buckel W, 19. Bobik TA, Wolfe RS (1989) Activation of formylmetha- energy-converting hydrogenase Eha in hydrogenotro- Hedderich R (2008) Methanogenic archaea: Ecologi- nofuran synthesis in cell extracts of Methanobacterium – phic methanogenesis. Proc Natl Acad Sci USA 109: cally relevant differences in energy conservation. Nat thermoautotrophicum. J Bacteriol 171:1423 1427. – fl 15473 15478. Rev Microbiol 6:579–591. 20. Krebs HA (1981) Reminiscences and Re ections (Clar- – 4. Balch WE, Fox GE, Magrum LJ, Woese CR, Wolfe RS 12. Thauer RK, et al. (2010) Hydrogenases from methano- endon Press, Oxford), pp 51 60. (1979) Methanogens: Reevaluation of a unique biolog- 21. Kornberg HL (1965) Anaplerotic sequences in microbial genic archaea, nickel, a novel cofactor, and H2 storage. ical group. Microbiol Rev 43:260–296. metabolism. Angew Chem Int Ed 4:558–565. Annu Rev Biochem 79:507–536. 5. DiMarco AA, Bobik TA, Wolfe RS (1990) Unusual coen- 22. Wolfe RS (1991) My kind of biology. Annu Rev Micro- 13. Gottschalk G, Thauer RK (2001) The Na(+)-translocat- zymes of methanogenesis. Annu Rev Biochem 59: biol 45:1–35. ing methyltransferase complex from methanogenic ar- 355–394. 23. Kaster AK, et al. (2011) More than 200 genes required chaea. Biochim Biophys Acta 1505(1):28–36. 6. Chistoserdova L, Vorholt JA, Thauer RK, Lidstrom ME for methane formation from H2 and CO2 and energy 14. Thauer RK (1998) Biochemistry of methanogenesis: A (1998) C1 transfer enzymes and coenzymes linking conservation are present in Methanothermobacter methylotrophic bacteria and methanogenic Archaea. tribute to Marjory Stephenson. Microbiology 144: marburgensis and Methanothermobacter thermauto- – Science 281:99–102. 2377 2406. trophicus. Archaea, 10.1155/2011/973848. 7. Chistoserdova L, et al. (2004) The enigmatic planctomy- 15. Scheller S, Goenrich M, Boecher R, Thauer RK, Jaun B 24. Costa KC, et al. (2010) Protein complexing in a metha- cetes may hold a key to the origins of methanogenesis (2010) The key nickel enzyme of methanogenesis ca- nogen suggests electron bifurcation and electron de- and methylotrophy. Mol Biol Evol 21:1234–1241. talyses the anaerobic oxidation of methane. Nature livery from formate to heterodisulfide reductase. Proc 8. Krishnakumar AM, et al. (2008) Getting a handle on 465:606–608. Natl Acad Sci USA 107:11050–11055. the role of coenzyme M in alkene metabolism. Micro- 16. Buckel W, Thauer RK (2012) Energy conservation via 25. Schick M, et al. (2012) Biosynthesis of the iron-guany- biol Mol Biol Rev 72:445–456. electron bifurcating ferredoxin reduction and proto- lylpyridinol cofactor of [Fe]-hydrogenase in methano- 9. Bertram PA, Thauer RK (1994) Thermodynamics of the n/Na(+) translocating ferredoxin oxidation. Biochim Bi- genic archaea as elucidated by stable-isotope labeling. formylmethanofuran dehydrogenase reaction in ophys Acta, 10.1016/j.bbabio.2012.07.002. J Am Chem Soc 134:3271–3280.

Thauer PNAS | September 18, 2012 | vol. 109 | no. 38 | 15085 Downloaded by guest on September 26, 2021