AceP from Ms. acetivorans was shown to transport acetate by a proton symport mechanism (Ribas et al., 2018). A homolog of AceP was shown to be required for acetate transport of acetate in Ms. mazei, and an AceP homolog is encoded in the genome of Methanothrix thermophila (Smith and Ingram-Smith, 2007; Welte et al., 2014). The transported acetate is converted to acetyl-CoA by acetate kinase (Ack) and phosphotransacetylase (Pta) in Methanosarcina, and by the AMP-forming acetyl-CoA synthetase (Acs) in Methanothrix (Berger et al., 2012). It was proposed that Ack and Pta were acquired by horizontal gene transfer from the genus Clostridium within the last 475 million years coinciding with evolution of aceticlastic pathways. This event resulted in a significant net increase of CH₄ leading to climate change in agreement with that proposed for the end-Permian mass extinction (Fournier and Gogarten, 2008; Rothman et al., 2014).

The catalytic mechanism for Ack from Methanosarcina thermophila proceeds by nucleophilic attack of the carboxyl group of acetate on the γ-phosphate of ATP with direct in-line transfer to acetate producing acetyl phosphate (Buss et al., 2001; Miles et al., 2002; Ferry, 2011). The mechanism for Pta, also from Ms. thermophila, involves base-catalyzed abstraction of the thiol proton of HS-CoA followed by nucleophilic attack of the thiolate anion (^–S-CoA) on the carbonyl carbon of acetyl phosphate forming acetyl-CoA (Iyer et al., 2004; Lawrence et al., 2006; Ferry, 2011). The crystal structure and biochemical characterization of Acs from Ms. acetivorans revealed the preference for medium chain substrates that excludes acetate, a result which indicates Acs functions other than activating acetate to acetyl-CoA (Ingram-Smith and Smith, 2007; Shah et al., 2009; Meng et al., 2010). The Acs of Methanothrix has a greater affinity for acetate than Ack of Ms. acetivorans which explains the dominance of Methanothrix in environments where acetate is in concentrations <0.1 mM (Berger et al., 2012). The acetyl-CoA is decarbonylated by the acetyl-CoA decarbonylase/synthase (ACDS) yielding a methyl group and CO. The methyl group is transferred to tetrahydrosarcinapterin (H₄SPT) yielding CH₃-H₄SPT and CO is oxidized to CO₂ with transfer of electrons to either ferredoxin (Fdx) or a novel flavodoxin (FldA) characterized from Ms. acetivorans (Prakash et al., 2019b).

The ACDS is predicted to be a component of the last universal common ancestor (LUCA) (Adam et al., 2018). Although of ancient origin and of central importance in the aceticlastic pathway, an atomic resolution structure of the intact ACDS complex from any methanogen is not reported. The enzymes from Methanosarcina and Methanothrix are known to have five subunits (αβγδε) based on the purified complexes and genomic analyses (Terlesky et al., 1986; Grahame and Demoll, 1996; Smith and Ingram-Smith, 2007). The β subunit catalyzes decarbonylation of acetyl-CoA while the αε subunits catalyze CO oxidation and the γδ subunits transfer the methyl group to H₄SPT producing CH₃-H₄SPT (Murakami and Ragsdale, 2000). The crystal structure of the αε component of Ms. barkeri identified the active site in the α subunit comprised of a pseudocubane Ni-Fe₃S₄ cluster bridged to an exogenous iron atom (Gong et al., 2008). A mechanism was proposed wherein the CO bound to Ni, and the OH^– bound to exogenous iron, H are coupled to form CO₂. A role for the ε subunit was proposed in which bound FAD directs electrons from the α subunit to Fdx. This proposal fits with the possibility that FldA accepts electrons from the ε subunit of the ACDS from Ms. acetivorans at the proposed FAD site. Spectroscopic studies of the β subunit from Ms. thermophila indicate an active site Fe₄S₄ cluster bridged to a binuclear Ni–Ni site in analogy to the homolog from an acetogen of the domain Bacteria that synthesizes acetyl-CoA (Gu et al., 2003; Funk et al., 2004; Ragsdale, 2007). Kinetic and EPR spectroscopy results indicate that alterations in the Ni coordination environment of the active site cluster promote C–C bond cleavage dependent on conformational changes (Gencic and Grahame, 2008). The γδ component transfers the methyl group of acetyl-CoA to H₄SPT involving a corrinoid coenzyme, although it is unknown which subunit interacts with H₄SPT and a crystal structure is not available (Grahame, 1993).

Acetate-grown Ms. acetivorans up regulates a γ class carbonic anhydrase (Cam) for which the crystal structure and biochemical characterization of the homolog from Ms. thermophila revealed the catalytic mechanism involving an active-site iron (Kisker et al., 1996; Iverson et al., 2000; Macauley et al., 2009; Zimmerman et al., 2013). Although homologs are present in acetate grown Methanosarcina and Methanothrix, the physiological function is not established. A plausible function involves diffusion of cytoplasmic CO₂ to the outer aspect of the membrane where AceP is located in a complex with Cam that hydrates CO₂ to HCO3-/H⁺ which supplies a local concentration of protons for symport of acetate by AceP (Figure 2). In this way, the proton gradient that drives ATP synthesis is not collapsed. The putative function for Cam is analogous to that reported for the α class carbonic anhydrase that supplies a proton for symport of lactate in mammalian cells (Peetz et al., 2014).

The methyl group of CH₃-H₄SPT is transferred to coenzyme M (HS-CoM) coupled to sodium extrusion by a membrane bound methyltransferase (MtrABCDEFGH). The CH₃-SCoM is reductively demethylated to CH₄ by the methyl coenzyme M reductase (McrABG) requiring coenzyme B (HSCoB) as the reductant. Post-translational modified residues N¹-methylhistidine (3-methylhistidine), 5-(S)-methylarginine, thioglycine, and S-methylcysteine are present in the active-sites of the catalytic McrA subunits from phylogenetically and metabolically diverse methanogenic and methanotrophic archaea (Grabarse et al., 2000; Kahnt et al., 2007). Mcr from Ms. acetivorans has emerged as a model for investigations of the modified residues. A unique radical SAM methyltransferase was shown required for methylation of the active-site arginine and concluded important for stability under imposed oxidative and heat stress (Deobald et al., 2018; Radle et al., 2019). Deletion of a homolog essential for arginine methylation in the obligate CO₂-reducing methanogen Methanococcus maripaludis resulted in a 40–60% loss in the rate of methanogenesis consistent with partial loss of Mcr activity (Lyu et al., 2020). Deletion of two genes essential for thioglycine synthesis in McrA of Ms. acetivorans produced mutants severely impaired in the rate of growth with acetate and when exposed to thermal and oxidative stress, results supporting a role for thioglycine in stabilizing the McrA active-site although not essential. Combinatorial deletion of genes responsible for incorporation of 5-(S)-methylarginine, thioglycine and S-methylcysteine generated Ms. acetivorans mutants with phenotypes consistent with altered thermal stability of McrA (Nayak et al., 2020). The studies suggest that residue modifications of Mcr function in important ways although not essential for catalysis. The CoMS-SCoB product of Mcr is reduced by a membrane bound electron transport chain ending with heterodisulfide reductase (HdrE₁D₁) that regenerates sulfhydryl forms of the coenzymes.

The electron transport pathways of all acetotrophic methanogens begin with the oxidation of Fdx and end with reduction of CoMS-SCoB by HdrE₁D₁ (Figure 2). As heterodisulfide is the terminal electron acceptor and generated internally, the process fits the definition of fermentative electron transport and energy conservation as opposed to respiration that requires an externally supplied electron acceptor. The aceticlastic pathways diverge in the mechanisms of membrane-bound electron transport that generates ion gradients driving ATP synthesis for growth (Figure 2). The H₂ dependent pathway (Figure 2A) has been investigated in Ms. barkeri and Ms. mazei for which the understanding is well developed (Welte and Deppenmeier, 2014). Reduced Fdx donates electrons to Ech hydrogenase that pumps protons and also reduces protons to H₂ that diffuses across the membrane where it is reoxidized at the outer aspect by the Vho hydrogenase, further contributing to the proton gradient (Welte and Deppenmeier, 2014; Kulkarni et al., 2018). Electrons from the oxidation of H₂ by Vho are transferred to HdrE₁D₁ by the quinone-like electron carrier methanophenazine (MP) accompanied by the vectoral translocation of protons that supplements the proton gradient. The proton gradient, together with the Mtr imposed Na⁺ gradient, drives ATP synthesis.

Several acetotrophic Methanosarcina lack Ech and Vho hydrogenases and are H₂ independent (Zhilina, 1978; Ollivier et al., 1984; Sowers et al., 1984a; Zinder et al., 1985; Elberson and Sowers, 1997; Von Klein et al., 2002; Shimizu et al., 2011; Ganzert et al., 2014). Ms. acetivorans is typical of H₂ independent Methanosarcina that instead utilize the membrane bound RnfCDGEAB complex to oxidize Fdx or FldA (Figure 2B; Li et al., 2006; Wang et al., 2011; Schlegel et al., 2012b; Prakash et al., 2019b). FldA accepts electrons from ACDS and is proposed to replace Fdx when growing in iron-limited environments (Prakash et al., 2019b). Fdx is an electron donor to the RnfB subunit of the Rnf complex (Suharti et al., 2014). It was further shown that the heterologously produced flavin-containing RnfG subunit is located on the outer aspect of the Escherichia coli membrane leading to the proposed model shown in Figure 3. Although MmcA is abundant in acetate-grown cells, its role in acetotrophic growth is questioned with the finding that a ΔmmcA mutant grows with acetate (Holmes et al., 2019). In contrast, the mutant is incapable of methanol-dependent respiratory growth with anthraquinone-2,6-disulfonate (AQDS), which suggests a role for MmcA in mediating electron transfer to external electron acceptors which fits the definition of respiratory electron transport and energy conservation. Rnf transfers electrons to MP for reduction of CoMS-SCoB by HdrE₁D₁ and pumps Na⁺ that thermodynamic considerations predict 3-4Na⁺/2 electrons (Schlegel et al., 2012b; Welte and Deppenmeier, 2014). Thus, electron transport generates H⁺ and Na⁺ gradients that, together with the Mtr-imposed Na⁺ gradient, drives ATP synthesis by the ATP synthase dependent on both H⁺ and Na⁺ (Schlegel et al., 2012a). It is proposed that the multi subunit Na⁺/H⁺ antiporter MrpABCDEFG adjusts the Na⁺/H⁺ ratio optimal for ATP synthesis (Jasso-Chavez et al., 2013, 2017). Although electron transport is remarkably different in Ms. barkeri and Ms. acetivorans, they have similar growth rates and yields in the absence of an exogenous electron acceptor which indicates that each conserve the same amount of energy (Sowers et al., 1984b). This result is consistent with equivalent H⁺ and Na⁺ gradients generated by electron transport and Mtr.

Methanosarcina acetivorans, Ms. barkeri and Ms. mazei each encode HdrE₁D₁, HdrA₁B₁C₁, HdrD₂, HdrA₂, and HdrC₂B₂. HdrE₁D₁ was shown to function in acetotrophic growth of Ms. acetivorans whereas HdrA₁B₁C₁ is apparently specific for methylotrophic growth (Buan and Metcalf, 2010; Catlett et al., 2015). It is proposed that reduced Fdx, generated in the oxidative branch, donates electrons to HdrA₁B₁C₁ that then reduces F₄₂₀ at the expense of CoMS-SCoB reduction in an electron bifurcation reaction (Buan and Metcalf, 2010). With this mechanism, electrons from Fdx are directed to the Fpo complex which results in additional energy conservation. A mechanism is proposed for the catalytic subunit HdrD that is distinct from the catalytic HdrB of the electron bifurcating HdrABC of obligate CO₂-reducing methanogens. Based on the crystal structure alone, a mechanism is proposed for HdrB involving two novel non-cubane 4Fe4S clusters (Wagner et al., 2017). This mechanism contrasts with that proposed for HdrD involving one conventional 4Fe4S cluster although based primarily on spectroscopic analyses (Walters and Johnson, 2004). However, both mechanisms propose that on reduction of CoMS-SCoB the sulfur atoms of the HSCoM and HSCoB are bound to iron in a five-coordinate manner. The electron pair for reduction of CoMS-SCoB derives from a membrane-bound electron transport chain that accepts electrons from either reduced Fdx or a flavodoxin (FldA) generated by ACDS (Figure 2B). The HdrE₁ subunit contains a b-type cytochrome that accepts electrons from MP for transfer to HdrD₁ (Welte and Deppenmeier, 2014).

Subunits of the recently characterized electron bifurcating HdrA₂B₂C₂ are up regulated in acetate-grown Ms. acetivorans consistent with a role in acetotrophic growth (Li et al., 2007; Buan and Metcalf, 2010; Rohlin and Gunsalus, 2010; Yan et al., 2017). Indeed, acetotrophic growth is impaired in a strain of Ms. acetivorans unable to synthesize HdrA₂B₂C₂ (Buan and Metcalf, 2010). Expression of the individual HdrA₂, HdrB₂, and HdrB₂C₂ subunits in E. coli, and biochemical characterization of the reconstituted active HdrA₂B₂C₂ complex, revealed a role for HdrA₂ in the oxidation of reduced coenzyme F₄₂₀ (F₄₂₀H₂) and FAD-dependent bifurcation of electrons that are transferred to Fdx and HdrC₂ (Figure 4; Yan et al., 2017). The HdrC₂ mediates electron transfer to HdrB₂ for reduction of CoMS-SCoB. The thermodynamically unfavorable reduction of Fdx is driven by the more favorable reduction of CoMS-SCoB. Although up regulated in acetate grown cells, the role for HdrA₂B₂C₂ in acetotrophic growth has not been established experimentally. It is postulated that the Rnf complex reduces coenzyme F₄₂₀ that is oxidized by HdrA₂B₂C₂ thereby recycling electrons to Fdx for oxidation by Rnf and an additional Na⁺ translocated, improving the thermodynamic efficiency (Buckel and Thauer, 2018). An unusual flavodoxin (FldA) can replace Fdx as electron donor to Rnf and acceptor for HdrA₂B₂C₂ (Prakash et al., 2019b). FldA is a potential advantage in periods of oxidative stress that damage the iron-sulfur clusters of Fdx, or when iron is limiting in the environment (Prakash et al., 2019b).

Considerably less is known of electron transport and energy conservation in Methanothrix. The genomes are void of genes encoding Ech hydrogenase or Rnf and, instead, encode F₄₂₀H₂ dehydrogenase (FpoABCDHIJKLMNO) although lacking the gene encoding FpoF that in Methanosarcina is the input module oxidizing F₄₂₀H₂ (Zhu et al., 2012). Thus, it is postulated that Fpo accepts electrons directly from Fdx with MP-mediated reduction of HdrED that is encoded in Methanothrix genomes (Zhu et al., 2012). Thermodynamic considerations predict 3H⁺ translocated by Fpo for a total of seven ions contributing to the gradient driving ATP synthesis (Welte and Deppenmeier, 2014). Although equivalent to gradients generated by H₂ dependent and H₂ independent Methanosarcina (Figure 2), Methanothrix requires two ATP for activation of acetate compared to one for Methanosarcina which predicts lower growth yields. However, this thermodynamic disadvantage is at least partially compensated by the ability of Methanothrix to metabolize acetate at lower concentrations compared to Methanosarcina (Jetten et al., 1992).

Methanosarcina acetivorans is capable of Fe(III)-dependent respiratory growth with acetate, a finding previously undocumented for acetotrophic methanogens (Prakash et al., 2019a). Growth and acetate consumption nearly doubles in the presence of ferrihydrite [Fe(OH)₃], the metal oxide form of Fe(III) that is common in the environment. Ferric iron is stoichiometrically reduced to ferrous iron. The ATP/ADP ratio also doubles indicating a higher energetic state consistent with increased growth. However, CH₄ is also produced indicating both fermentative and respiratory electron transport and energy conservation. The revised, ecologically relevant, pathway is shown in Figure 5. All one-carbon transformations leading to CH₄ are the same as in Figure 2. Two Na⁺ are translocated for each Fe(III) reduced to Fe(II) in respiratory electron transport (Yan et al., 2018). Although further research is necessary, the present results indicate that productive Na⁺ translocation by the Rnf complex is dependent on electron transfer to MmcA that reduces an exogenous electron acceptor which fits the definition of respiratory electron transport. Respiratory electron transport is dependent on oxidation of the methyl group from CH₃-H₄SPT by reversal of reactions in the CO-dependent pathway of CO₂ reduction to CH₄ and acetate in Ms. acetivorans which generates reduced coenzyme F₄₂₀ (F₄₂₀H₂) and additional reduced Fdx to enter the pool for both respiratory and fermentative electron transport (Lessner et al., 2006). The F₄₂₀H₂ dehydrogenase, essential for methylotrophic growth, is down regulated in acetate-grown cells leading to the proposal that oxidation of F₄₂₀H₂ is dependent on the electron bifurcating HdrA₂B₂C₂ (Yan et al., 2017). As FldA can replace Fdx as electron acceptor for HdrA₂B₂C₂, and donor to Rnf, either are available for initiating fermentative and respiratory electron transport (Prakash et al., 2019b). The combination of fermentative and respiratory electron transport generates both H⁺ and Na⁺ gradients that drive ATP synthesis by the ATP synthase dependent on both gradients (Schlegel et al., 2012a). It is proposed that the multi subunit Na⁺/H⁺ antiporter Mrp adjusts the Na⁺/H⁺ ratio optimal for ATP synthesis (Jasso-Chavez et al., 2013, 2017).

A respiratory pathway is also proposed for Ms. acetivorans grown with methanol when methanogenesis is inhibited by 2-bromoethanesulfonate (Figure 6; Holmes et al., 2019). The methyl group of methanol is oxidized to CO₂ with reduction of Fdx and F₄₂₀ for which the latter is reoxidized by the F₄₂₀H₂ dehydrogenase complex (Fpo and FpoF) that is up regulated in methanol grown cells. Fpo transfers the electrons to MP accompanied by the translocation of H⁺ which contributes to the ion gradient that drives ATP synthesis. Reduced MP transfers electrons to MmcA that reduces AQDS as the final electron acceptor. The reduced Fdx donates electrons to Rnf that also transfers electrons to MmcA with translocation of Na⁺ analogous to that proposed in the revised aceticlastic pathway (Figure 4). The imposed inhibition of methanogenesis precludes extrapolation to the environment although reinforces the discovery that Ms. acetivorans is capable of respiratory growth.

The revised aceticlastic pathway of Ms. acetivorans has important ecological and evolutionary implications. Without respiration, the amount of energy

available by methanogenesis alone, with equimolar reactants and products (Eq. 1), is barely enough to synthesize one ATP (Eq. 2). It is possible that growth by methanogenesis alone is only achievable in the laboratory with an abundant supply of acetate at optimal temperature, pH, and supply of nutrients whereas growth in the competitive and dynamic environment is dependent on additional energy gained by respiration. In environments where Fe(III) is limiting, energy conservation by methanogenic fermentation could afford an advantage over acetotrophic competitors that conserve energy only by respiration. Ms. acetivorans, and other Methanosarcina which are H₂ independent, may have an advantage over H₂ dependent Methanosarcina that are without multi-heme c-type cytochromes and incapable of respiratory growth.

The realization of Fe(III)-dependent AOM has implications for understanding the CH₄ and iron cycles, both past and present. It is postulated that symbiotic associations of ANME and sulfate-reducing species evolved from methanogenic species that first acquired the capacity to conserve energy by oxidizing CH₄ and reducing metals (Scheller et al., 2016). Moreover, it is postulated that Fe(III)-dependent AOM was largely responsible for oxidizing all the CH₄ produced on early Earth prior to the appearance of oxygen (Beal et al., 2009). It is further hypothesized that if only a small fraction of current global Mn(IV) and Fe(III) influx is used for AOM, it has the potential to consume a large amount of CH₄ (Beal et al., 2009). Ms. acetivorans was isolated from off shore marine sediments near locations with CH₄ seeps where single cells and aggregates of ANME are present and could play a role in non-symbiotic Fe(III)-dependent AOM (Sowers et al., 1984a; Orphan et al., 2002).