The Proposed Molecular Mechanisms Used by Archaea for Fe(III) Reduction and Fe(II) Oxidation

Iron (Fe) is the fourth most abundant element in the Earth’s crust where ferrous Fe [Fe(II)] and ferric Fe [Fe(III)] can be used by archaea for energy conservation. In these archaea-Fe interactions, Fe(III) serves as terminal electron acceptor for anaerobic respiration by a variety of archaea, while Fe(II) serves as electron donor and/or energy sources for archaeal growth. As no Fe is incorporated into the archaeal cells, these redox reactions are referred to as dissimilatory Fe(III) reduction and Fe(II) oxidation, respectively. Dissimilatory Fe(III)-reducing archaea (FeRA) and Fe(II)-oxidizing archaea (FeOA) are widespread on Earth where they play crucial roles in biogeochemical cycling of not only Fe, but also carbon and sulfur. To reduce extracellular Fe(III) (oxyhydr)oxides, some FeRA transfer electrons directly to the Fe(III) (oxyhydr)oxides most likely via multiheme c-type cytochromes (c-Cyts). These multiheme c-Cyts may form the pathways similar to those found in bacteria for transferring electrons from the quinone/quinol pool in the cytoplasmic membrane to the Fe(III) (oxyhydr)oxides external to the archaeal cells. Use of multiheme c-Cyts for extracellular Fe(III) reduction by both Domains of Archaea and Bacteria emphasizes an ancient mechanism of extracellular electron transfer, which is well conserved. Other FeRA, however, reduce Fe(III) (oxyhydr)oxides indirectly via electron shuttles. Similarly, it is proposed that FeOA use pathways to oxidize Fe(II) on the surface of the cytoplasmic membrane and then to transfer the released electrons across the cytoplasmic membrane inward to the O2 and NAD+ in the cytoplasm. In this review, we focus on the latest understandings of the molecular mechanisms used by FeRA and FeOA for Fe(III) reduction and Fe(II) oxidation, respectively.


INTRODUCTION
By mass, iron (Fe) is the most abundant element on Earth and the fourth most abundant element in the Earth's crust (Morgan and Anders, 1980). In the Earth's biosphere, Fe exists mainly as metal ion of two different oxidation states: divalent ferrous Fe [Fe(II)] and trivalent ferric Fe [Fe (III)]. At the circumneutral pH and under oxic condition, Fe(II) is quickly oxidized to Fe(III) that is readily precipitated to solid Fe(III) (oxyhydr)oxides in the absence of chelator (Emerson et al., 2010). Thus, Fe(III) (oxyhydr)oxides are abundant in a variety of environments (e.g., soils, sediments and subsurface) where they can be used by microorganisms, including archaea, as the terminal electron acceptors for anaerobic respiration White et al., 2016;Jiang et al., 2019). For example, the hyperthermophilic archaeon Pyrobaculum islandicum couples H 2 oxidation to reduction of Fe(III) (oxyhydr)oxides for growth (Vargas et al., 1998). As Fe is not assimilated into microbial cells, microbial Fe(III) reduction for anaerobic respiration is referred to as dissimilatory Fe(III) reduction. Notably, microbial dissimilatory Fe(III) reduction is believed to be an ancient form of respiration (Vargas et al., 1998). In addition to H 2 , dissimilatory Fe(III)-reducing archaea (FeRA) also couple oxidation of methane (CH 4 ), acetate and other organic compounds to Fe(III) reduction (Tor and Lovley, 2001;Kashefi et al., 2002;Beal et al., 2009). Moreover, FeRA are found in different environmental settings, ranging from the sediment of freshwater lakes to deep sea hyperthermal vents, where they are involved in biogeochemical cycling of not only Fe, but also carbon (C) (Slobodkina et al., 2009;Yamada et al., 2014;Weber et al., 2017).
In addition to oxygen (O 2 ), the stability of Fe(II) in solution is also pH-dependent: the more acidic, the more stable (Emerson et al., 2010). Thus, Fe(II) can be oxidized by bacteria as well as archaea under anoxic condition or at acidic pH (Hafenbradl et al., 1996;Edwards et al., 2000;Valdes et al., 2008). For example, the hyperthermophilic archaeon Ferroglobus placidus oxidizes Fe(II) under anoxic condition and at circumneutral pH (Hafenbradl et al., 1996). The extremely acidophilic archaeon Ferroplasma acidarmanus oxidizes Fe(II) at pH 0-2.5 and the Fe concentration that is as high as 111 g/L (Edwards et al., 2000). Furthermore, the acidophilic dissimilatory Fe(II)-oxidizing archaea (FeOA) are prevalent in the environments of low pH, such as acid mine drainages (AMD) and acidic hot springs, where they play crucial roles in not only AMD formation, but also biogeochemical cycling of Fe, C as well as sulfur (S) (Edwards et al., 2000;Baker and Banfield, 2003;Johnson et al., 2012;Chen et al., 2016;Huang et al., 2016). Notably, the FeOA Ferroplasma spp. are the key members of microbial consortia used in biomining of copper (Cu) and gold (Au) that associate the minerals rich in Fe and S (Dopson et al., 2004;Hawkes et al., 2006).
The electron exchanges between microorganisms and the metal ions or others that are external to microbial cells are termed as microbial extracellular electron transfer White et al., 2016;Jiang et al., 2019). FeRA can reduce extracellular Fe(III) (oxyhydr)oxides either directly or indirectly. For example, the hyperthermophilic archaeon Geoglobus ahangari reduces Fe(III) (oxyhydr)oxides most likely via its cell surface-exposed multiheme c-type cytochromes (c-Cyts) (Manzella et al., 2013. All the currently available data suggest that FeOA oxidize Fe(II) on the exterior side of the cytoplasmic membrane probably via their cell surface-exposed redox proteins, such as b-type cytochromes and the Cu-containing proteins Castelle et al., 2015). This review focuses on our current understandings of the molecular mechanisms underlying the ability of FeRA and FeOA to exchange electrons with Fe(III) and Fe(II), respectively.

IRON (III) REDUCTION
Dissimilatory Fe(III)-Reducing Archaea Thermophilic and Hyperthermophilic Archaea Thermophilic and hyperthermophilic archaea were among the first groups of archaea demonstrated conclusively to be capable of respiring on Fe(III) (Vargas et al., 1998;Slobodkin et al., 1999;Kashefi and Lovley, 2000;Kashefi et al., 2002Kashefi et al., , 2008. Among them, Geoglobus acetivorans and Geoglobus ahangari use only Fe(III) as the terminal electron acceptors, of which Fe(III) (oxyhydr)oxides are preferable (Kashefi et al., 2002;Slobodkina et al., 2009). Given that these thermophiles and hyperthermophiles are believed to be the most closely related to the last common ancestors of modern life, discovery of dissimilatory Fe(III) reduction by the thermophilic and hyperthermophilic archaea and bacteria provides microbiological evidences that dissimilatory Fe(III) reduction is an ancient form of microbial respiration (Liu et al., 1997;Pace, 1997;Vargas et al., 1998;Kashefi and Lovley, 2000). Consistent with these microbiological evidences are the geological evidences of abundance of H 2 and Fe(III) (oxyhydr)oxides on pre-biotic Earth where microbial dissimilatory Fe(III) reduction occurred earlier than microbial reduction of sulfate (SO4 2− ), nitrate (NO 3 − ) and O 2 for respiration (Walker, 1984(Walker, , 1987Lovley, 1991).

Methanogens
Methanogens are groups of strictly anaerobic archaea that are directly involved in CH 4 production (Thauer et al., 2008). Fe(III) can also be reduced for respiration by the methanogens that are phylogenetically and physiologically diverse (Vargas et al., 1998;Bond and Lovley, 2002;Bodegom et al., 2004;Liu et al., 2011a;Zhang et al., 2012Zhang et al., , 2013Yamada et al., 2014;Sivan et al., 2016;Prakash et al., 2019). For example, Methanosarcina barkeri couples H 2 oxidation to reduction of Fe(III) (oxyhydr)oxides and Fe(III)-containing clay minerals, which lower the efficiency of CH 4 formation by re-directing the electron originally for CH 4 generation to Fe(III) reduction (Bond and Lovley, 2002;Bodegom et al., 2004;Liu et al., 2011a). Furthermore, the acetotrophic methanogen Methanosarcina acetivorans produces CH 4 from the methyl group of acetate via fermentation whose G o = −36 kJ/mol (Deppenmeier and Muller, 2008;Prakash et al., 2019). Addition of ferrihydrite [Fe(III) (OH) 3 ], however, increases CH 4 production by M. acetivorans . This increase of CH 4 production is directly linked to the ferrihydrite reduction coupled to acetate oxidation whose G o = −707 kJ/mol (Roden and Lovley, 1993;Prakash et al., 2019). It is believed that a novel electron bifurcation mechanism facilitates the thermodynamically less favorable acetate fermentation to CH 4 with the thermodynamically more favorable respiratory reduction of ferrihydrite . This may result in increase of ATP level in the cytoplasm of M. acetivorans, which enhances production of methyl and carbonyl groups from acetate for methanogenesis Prakash et al., 2019).

The Proposed Molecular Mechanisms
Microorganisms may transfer electrons to extracellular Fe(III) (oxyhydr)oxides either directly or indirectly. Direct reduction requires physical contact between the redox proteins on the microbial surfaces, such as multiheme c-Cyts, and the surfaces of Fe(III) (oxyhydr)oxides ( Figure 1A; Xiong et al., 2006;Meitl et al., 2009). Indirect reduction may involve electron shuttles and Fe(III) chelators that are secreted by the microorganisms. The electron shuttles are first reduced by FeRA and the reduced shuttles then ferry the electrons to the Fe(III) (oxyhydr)oxides ( Figure 1B; Marsili et al., 2008;von Canstein et al., 2008). Different from the electron shuttles, the Fe(III) chelators solubilize the Fe(III) from Fe(III) (oxyhydr)oxides and then bring the chelator-Fe(III) complexes back to the FeRA for reduction ( Figure 1C; Gralnick and Newman, 2007;Marsili et al., 2008;Shi et al., 2009).

Direct Electron Transfer With Multiheme c-Cyts Thermophilic FeRA
The thermophilic FeRA G. ahangari reduces Fe(III) (oxyhydr)oxides directly (Manzella et al., 2013). Although addition of AQDS stimulates the reduction, G. ahangari itself cannot reduce the Fe(III) (oxyhydr)oxides entrapped in the alginate beads of 12-kDa pore size, which is large enough for AQDS to pass through, but not for the archaeal cells (Manzella et al., 2013). Furthermore, compared to the fresh medium, addition of the supernatants of the stationary culture grown with Fe(III) (oxyhydr)oxides does not increase the reduction rate by G. ahangari (Manzella et al., 2013). Finally, addition of artificial chelators has no impact on the reduction of entrapped Fe(III) (oxyhydr)oxides by G. ahangari (Manzella et al., 2013). Thus, G. ahangari produces no electron shuttle or chelator for reducing Fe(III) (oxyhydr)oxides (Manzella et al., 2013). Similarly, neither G. acetivorans nor F. placidus is able to reduce the Fe(III) (oxyhydr)oxides inside the alginate beads Smith et al., 2015). Thus, all these results show direct contact between Fe(III) (oxyhydr)oxides and G. acetivorans, G. ahangari and F. placidus is required for Fe(III) reduction (Manzella et al., 2013;Mardanov et al., 2015;Smith et al., 2015).
Results from investigating bacterial extracellular electron transfer mechanisms demonstrate the importance of multiheme c-Cyts in direct reduction of Fe(III) (oxyhydr)oxides by bacteria (Shi et al., 2007White et al., 2016;Jiang et al., 2019). Similarly, genomic sequencing results reveal that G. ahangari has 19 putative genes that encode multiheme c-Cyts and some of them are predicted to be on the archaeal cell surface . Indeed, some hemecontaining proteins can be mechanically sheared off from the cell surface of G. ahangari and this mechanic treatment renders the cells of G. ahangari unable to reduce Fe(III) (oxyhydr)oxides (Manzella et al., 2013). Moreover, transmission electron microscopy results show intimate association of the cells of G. ahangari with Fe(III) (oxyhydr)oxides (Manzella et al., 2013. Notably, G. ahangari possesses a putative decaheme c-Cyt of NapC/NirT family quinol dehydrogenase (GAH_01256) . Investigation with the dissimilatory Fe(III)-reducing bacterium Shewanella oneidensis MR-1 reveals that tetraheme c-Cyt CymA, which is also a member of NapC/NirT family quinol dehydrogenase, is the key member of an electron transfer pathway for extracellular reduction of Fe(III) (oxyhydr)oxides. In this pathway, CymA oxidizes quinol in the cytoplasmic membrane and then transfers the released electrons to other c-Cyts that relay electrons eventually to the Fe(III) (oxyhydr)oxides on the bacterial cell surface (Shi et al., 2007Marritt et al., 2012a,b;Mcmillan et al., 2012Mcmillan et al., , 2013. Further analyses also suggest that the c-Cyts GAH_01306, GAH_00286 of GAH_01534, and GAH_01253 maybe the terminal reductases of G. ahangari for extracellular reduction of Fe(III) . Although their proposed functions have not been experimentally verified, the c-Cyts GAH_01306, GAH_00286 of GAH_01534, GAH_01253 and GAH_01256 may function  like those identified in S. oneidensis MR-1 to form an pathway for transferring electrons from the quinone/quinol pool in the cytoplasmic membrane, across the cell wall to the Fe(III) (oxyhydr)oxides on the cell surface of G. ahangari (Figure 2A).
Similar to G. ahangari, G. acetivorans and F. placidus also possess numerous genes for multiheme c-Cyts Smith et al., 2015). Some of these c-Cyts of G. acetivorans contain the binding motifs for Fe(III) (oxyhydr)oxides, which were originally identified in the decaheme c-Cyts MtrC and OmcA. MtrC and OmcA are the terminal reductases used by S. oneidensis MR-1 for extracellular reduction of Fe(III) (oxyhydr)oxides (Xiong et al., 2006;Lower et al., 2008;Johs et al., 2010;Edwards et al., 2014Edwards et al., , 2015Edwards et al., , 2020Mardanov et al., 2015). Furthermore, expressions of some of the c-Cyts with binding motifs for Fe(III) (oxyhydr)oxides are up-regulated when G. acetivorans is cultured under Fe(III)-reducing conditions and some of them are on the cell surface. Thus, G. acetivorans and F. placidus may reduce Fe(III) (oxyhydr)oxides via their cell surface-exposed multiheme c-Cyts Smith et al., 2015). Given numerous c-Cyts found in these archaea, some of them may also form the electron transfer pathway similar to the proposed one of G. ahangari for extracellular reduction of Fe(III) (oxyhydr)oxides (Figure 2A; Mardanov et al., 2015).

Methanogen
The methyl-coenzyme M reductase (Mcr) is believed to catalyzes the key reaction of methane activation during anaerobic methane oxidation. The mcr genes are widespread in ANME. Addition of a mcr gene from an uncultivated ANME-1 enables the methanogen M. acetivorans to couple methane oxidation to Fe(III) reduction (Soo et al., 2016). Further characterizations show that the cytoplasmic membrane multiheme c-Cyt A (MmcA), heterodisulfide reductase (HdrDE), and methanophenazine (MP) in the cytoplasmic membrane are all capable of reducing the soluble Fe(III)-citrate either directly or indirectly ( Figure 2B; . Recent results suggested that MmcA was probably the terminal reductase for ferrihydrite .

ANME
Multiheme c-Cyts are also believed to be crucial in direct electron transfer from ANME to SRB as well as in electron transfer from ANME to Mn(IV) and Fe(III) (oxyhydr)oxides Wegener et al., 2015;Skennerton et al., 2017;Cai et al., 2018;Leu et al., 2020a,b). All these ANME genomes possess numerous multiheme c-Cyt-encoding genes. Some of these putative c-Cyts also contain S-layer domain, indicating that they are cell surface-exposed. Further chemical imaging analyses indeed detect cytochromes on these archaeal cell surface . Moreover, some of these multiheme c-Cyt-encoding genes are acquired by the ANME from the Fe(III)-reducing bacteria via lateral gene transfer (Skennerton et al., 2017;Leu et al., 2020b). Although their direct involvements in extracellular reduction of Fe(III) still remain to be demonstrated, the multiheme c-Cyts of these ANME may also form the pathways similar to those found in the bacteria and that proposed in other FeRA (Figure 2A) to transfer electrons from the quinone/quinol pool in the cytoplasmic membrane, across the cell wall to the Fe(III) (oxyhydr)oxides or the bacterial cells contacted on the archaeal cell surface (Figure 2C; McGlynn et al., 2015;Shi et al., 2016;Jiang et al., 2019;Leu et al., 2020b).
The crucial roles of multiheme c-Cyts in extracellular reduction of Fe(III) (oxyhydr)oxides and other extracellular substrates by both Domains of Archaea and Bacteria emphasizes that the multiheme c-Cyts-mediated pathway is an ancient and well conserved extracellular electron transfer mechanism.

Direct Electron Transfer With Other Redox Proteins
The hyperthermophilic FeRA P. islandicum reduces Fe(III) (oxyhydr)oxides and elemental sulfur [S(0)] directly as it is unable to reduce the Fe(III) (oxyhydr)oxides and S(0) inside the dialysis bags of 1.2-1.4 kDa pore size and produces no electron shuttle (Feinberg and Holden, 2006). It should be noted that P. islandicum lacks c-Cyt. Thus, P. islandicum must employ c-Cyt-independent mechanisms for direct reduction of Fe(III) (oxyhydr)oxides (Feinberg et al., 2008). Indeed, molybdopterin oxidoreductases of Pyrodictium delaneyi are suggested to transfer electrons directly to Fe(III) (oxyhydr)oxides (Kashyap and Holden, 2021).

Indirect Electron Transfer
Pyrobaculum aerophilum and Pyrobaculum arsenaticum reduce Fe(III) (oxyhydr)oxides indirectly as they can still reduce the Fe(III) (oxyhydr)oxides inside the dialysis bags of 1.2-1.4 kDa pore size that separate the archaeal cells and the (oxyhydr)oxides (Feinberg and Holden, 2006;Feinberg et al., 2008). Given that no chelated Fe(II) or Fe(III) is detected outside the bags, P. aerophilum is suggested to use electron shuttles to reduce the Fe(III) (oxyhydr)oxides inside the dialysis bags (Feinberg and Holden, 2006).

IRON (II) OXIDATION Dissimilatory Fe(II)-Oxidizing Archaea
The Neutrophilic FeOA The FeRA F. placidus was originally isolated from a shallow submarine hydrothermal system at Vulcano, Italy as an Fe(II) oxidizer that could couple Fe(II) oxidation to NO 3 − reduction with a pH optimum of 7 (Hafenbradl et al., 1996). Notably, F. placidus can fix CO 2 and reduce NO 3 − to N 2 O for chemolithoautotrophic growth when H 2 is used as an electron donor (Vorholt et al., 1997). However, whether F. placidus can chemolithoautotrophically grow with Fe(II) and NO 3 − still needs to be experimentally tested.

Acidophiles
Nearly all FeOA isolated so far are the acidophiles. They are abundant in the acidic and extremely acidic environments rich in Fe and S, such as acidic hot springs, AMD sites and bioleaching plants. These acidophilic FeOA are the key players in regulating biogeochemical cycling of C, Fe, and S, in AMD formation as well as in biomining of Cu, Au and other metals associated with the minerals that contain Fe and S (Baker and Banfield, 2003;Chen et al., 2016;Golyshina et al., 2016;Huang et al., 2016;Quatrini and Johnson, 2018). For example, Metallosphaera yellowstonensis was isolated from the Fe(III) oxide microbial mats in the acidic geothermal springs in Yellowstone National Park. It grows autotrophically with pyrite (FeS 2 ) and Fe(II) sorbed on ferrihydrite at the optimal pH 2-3 and temperature 65-75 • C (Kozubal et al., 2008). Notably, the Fe(III) oxide microbial mats also contain other microorganisms, such as Marsarchaeota and Geoarchaeota that reside under microaerobic condition and beneath the more oxic zone where M. yellowstonensis is abundant Bernstein et al., 2013;Jay et al., 2018). Furthermore, the Fe(III)-reducing archaea are also found in these microbial mats . Thus, an active archaea-mediated Fe(II)/Fe(III) cycling is believed to occur in the mats . Moreover, Metallosphaera sedula was isolated from an acidic drain of a hot spring near Naples, Italy. It can grow chemolithoautotrophically on FeS 2 , sphalerite [(Zn,Fe)S], chalcopyrite (CuFeS 2 ) and S(0) at pH 1-4.5 and temperature 50-80 • C (Huber et al., 1989). Other Metallosphaera spp. that are isolated from the acidic hot springs of different locations can also grow by oxidizing Fe(II) and other metal ions (Kozubal et al., 2008Liu et al., 2011b,c;Wheaton et al., 2019). Notably, Metallosphaera spp. belong to the order of Sulfolobales, among which Sulfolobus metallicus, Sulfolobus tokodaii, and Acidianus brierleyi (formerly known as Sulfolobus brierleyi) are all isolated from acidic hot springs and are all able to oxidize Fe(II) for growth (Brierley and Brierley, 1973;Segerer et al., 1986;Huber and Stetter, 1991;Suzuki et al., 2002;Bathe and Norris, 2007).
In addition to AMD sites, Ferroplasma spp. is also found in the bioleaching pilot plants globally. For example, F. acidiphilum was isolated from the bioreactor of a pilot plant at Tula, Russia for bioleaching the arsenopyrite (FeAsS) and FeS 2 ores that contain Au. F. acidiphilum is an aerobe that oxidizes Fe(II), including FeS 2 , with an optimum pH of 1.7 for growth (Golyshina et al., 2000). Originally identified as F. cupricumulans, Acidiplasma cupricumulans was isolated from an industrial-scale CuFeS 2 bioleach heap in Myanmar and it grew on Fe(II) at optimum pH 1-1.2 (Hawkes et al., 2006;Golyshina et al., 2009). It should be noted that all the tested Ferroplasma and Acidiplasma spp. are facultative anaerobes that can couple oxidation of yeast extract to Fe(III) reduction, which demonstrate that these Ferroplasma and Acidiplasma spp. are both FeOA and FeRA, depending on the ambient environmental conditions (Dopson et al., 2004;Zhou et al., 2008;Golyshina et al., 2009).

The Proposed Molecular Mechanisms
Fe(II) oxidation is a bioenergetic challenge for microorganisms, especially for autotrophic microorganisms. At pH 2, the redox potentials for Fe(II)/Fe(III) and O 2 /H 2 O couples are +0.77 V and +1.2 V, respectively. Thus, little energy can be derived from Fe(II) oxidation coupled to O 2 reduction . Furthermore, the redox potential for NAD + /NADH couple is −0.32 V at pH 6.5, which is the pH found in the microbial cytoplasm. Thus, without any energy input, Fe(II) oxidation coupled to NAD + reduction is thermodynamically unfeasible (Bonnefoy and Holmes, 2012). To overcome this thermodynamic barrier, the acidophilic FeOA use the energy from proton motive force to drive the electron transfer from Fe(II) oxidation to NAD + reduction (i.e., uphill electron transfer). Although energy generation is limited, the electron transfer from Fe(II) oxidation to O 2 reduction is thermodynamically feasible (i.e., downhill electron transfer) and thus can be catalyzed by microorganisms (Figure 3A; Bird et al., 2011;Bonnefoy and Holmes, 2012;Ilbert and Bonnefoy, 2013;Nitschke and Bonnefoy, 2016;Jiang et al., 2019).

Metallosphaera yellowstonensis
Sequencing results revealed that the genome of M. yellowstonensis contained a gene cluster that was also found in other FeOA, such as M. sedula and S. metallicus, in which this gene cluster was implicated in Fe(II) oxidation (Bathe and Norris, 2007;Kozubal et al., 2008). In this gene cluster, two of its protein products FoxCD are proposed to oxidize Fe(II) on the surface of the cytoplasmic membrane of M. yellowstonensis and transfer the released electrons in the cytoplasmic membrane to either O 2 via the proposed downhill pathway or NAD + via the proposed uphill pathway in the cytoplasmic membrane . The downhill electron transfer pathway may include multicopper oxidase (MCO) and terminal hemecopper oxidase FoxAB. The uphill electron transfer pathway may include MCO, Cytochrome bc complex, quinone/quinol pool and NADH dehydrogenase ( Figure 3B; Kozubal et al., 2011). Given that they also have the fox gene cluster, the acidophilic FeOA Acidianus copahuensis, M. sedula, and S. metallicus are also suggested to oxidize Fe(II) via the mechanism similar to that of M. yellowstonensis (Bathe and Norris, 2007;Bonnefoy and Holmes, 2012;Ilbert and Bonnefoy, 2013;Nitschke and Bonnefoy, 2016;Urbieta et al., 2017).

Ferroplasma acidiphilum
Sequencing results reveal no fox homolog in the genomes of Ferroplasma spp. (Tyson et al., 2004;Allen et al., 2007;Golyshina et al., 2017). Protein purification and characterization demonstrate that the blue copper protein sulfocyanin of F. acidiphilum, which is suggested to be on the exterior side of cytoplasmic membrane, oxidizes Fe(II) directly . Sulfocyanin is part of an 850 kDa protein complex that also contains the aa 3 -type cytochrome oxidase. Thus, this cytoplasmic membrane protein complex is proposed to couple Fe(II) oxidation to O 2 reduction ( Figure 3C; Castelle et al., 2015). A 150 kDa protein complex is also isolated from the membrane fraction. This complex is believed to be the Rieske-cytochrome b-type complex that is suggested to mediate electron transfer from sulfocyanin to quinone in the cytoplasmic membrane ( Figure 3C). The similar Fe(II) oxidation mechanism may also exist in other Ferroplasma spp. . Thus, F. acidiphilum and most likely other Ferroplasma spp. oxidize Fe(II) via a mechanism different from that of M. yellowstonensis. Given that they are also different from those of bacteria Bonnefoy and Holmes, 2012;Shi et al., 2012Shi et al., , 2016Ilbert and Bonnefoy, 2013;Jiang et al., 2019), the molecular mechanisms for Fe(II) oxidation by FeOA must have evolved independently.

CONCLUSION REMARKS
The phylogenetically and physiologically diverse groups of FeRA and FeOA have been isolated from a variety of ecosystems worldwide. They are directly involved in biogeochemical cycling of Fe, Mn, C, and S. Furthermore, FeRA reduce Fe(III) (oxyhydr)oxides either directly or indirectly and multiheme c-Cyts are believed to be involved in the direct reduction of extracellular Fe(III) (oxyhydr)oxides most likely by forming the extracellular electron transfer pathways. Finally, FeOA are proposed to oxidize Fe(II) on the surface of the cytoplasmic membrane via the b-type cytochromes and Cu-containing proteins that maybe the part of the electron transfer pathways. In bacteria, the extracellular electron transfer pathways connect extracellular redox transformation of Fe with intracellular metabolic activities. These pathways consist of redox and structural proteins as well as redox molecules (Table 1) 2019)]. Some of these pathways are rigorously characterized White et al., 2016;Jiang et al., 2019). In addition to the extracellular electron transfer pathways, extracellular extensions, such as Geobacter conductive pili (Gpili) and multiheme c-Cyt-based nanowires, have also been identified and characterized for transferring electrons to the Fe(III) (oxyhydr)oxides distant from bacterial surfaces ( Table 1; Filman et al., 2019;Lovley and Walker, 2019;Reguera and Kashefi, 2019;Wang et al., 2019;Yalcin et al., 2020).
Similar to those found in bacteria, multiheme c-Cyts and quinone/quinol pool are suggested to be the key components of the electron transfer pathways involved in extracellular reduction of Fe(III) by archaea. Likewise, Cu-containing proteins, NADH dehydrogenases, aa 3 oxidase, bc 1 complex, and quinone/quinol pool are implicated as the key components of the electron transfer pathways for Fe(II) oxidation by both bacteria and archaea (Table 1). Thus, archaea and bacteria may share these redox proteins and molecules for Fe(III) reduction and/or Fe(II) oxidation. However, identification of the redox proteins and molecules unique for archaeal Fe(III) reduction, such as molybdopterin oxidoreductases, HdrDE and MP, and those unique for bacterial Fe(III) reduction (e.g., flavins, flavin-binding proteins, G-pili and c-Cyt-based nanowires) demonstrate the diverse mechanisms that have evolved independently among different groups of archaea and bacteria for exchanging electrons with Fe(III) and Fe(II) ( Table 1).
Compared to those in bacteria, the molecular mechanisms for reduction of Fe(III) (oxyhydr)oxides and oxidation of Fe(II) by archaea have been much less characterized and remain speculative, which is mainly attributed to the lack of genetically tractable model FeRA and FeOA. Thus, development of the representative model systems is crucial for further understanding these electron transfer mechanisms at molecular level. Currently, M. acetivorans and M. barkeri are the two genetically tractable methanogens for investigating molecular mechanisms for Fe(III) reduction. Although it is demonstrated to reduce soluble Fe(III), whether MmcA of M. acetivorans can reduce Fe(III) (oxyhydr)oxides remains uncharacterized. Furthermore, a novel c-Cyt-independent extracellular electron transfer mechanism has recently been identified and characterized in Grampositive bacteria . Whether the similar mechanism also exists in archaea remains uninvestigated. Thus, future research should focus on bridging these knowledge gaps.
Identification and characterization of archaeal electron transfer mechanisms at molecular level will not only substantially advance our understanding of how microorganisms exchange electrons with Fe(III) and Fe(II), but also lay solid foundations for further developments of FeRA-and FeOA-based biotechnology. For example, molecular understanding of electron transfer mechanisms from archaeal cell surface to their cytoplasmic membranes will certainly help develop electromethanogenesis via the synthetic biology approach (Cheng et al., 2009).