Mineralogical and Genomic Constraints on the Origin of Microbial Mn Oxide Formation in Complexed Microbial Community at the Terrestrial Hot Spring

Manganese (Mn) oxides are widespread on the surface environments of the modern Earth. The role of microbial activities in the formation of Mn oxides has been discussed for several decades. However, the mechanisms of microbial Mn oxidation, and its role in complex microbial communities in natural environments, remain uncertain. Here, we report the geochemical, mineralogical, and metagenomic evidence for biogenic Mn oxides, found in Japanese hot spring sinters. The low crystallinity of Mn oxides, and their spatial associations with organic matter, support the biogenic origin of Mn oxides. Specific multicopper oxidases (MCOs), which are considered Mn-oxidizing enzymes, were identified using metagenomic analyses. Nanoscale nuggets of copper sulfides were, also, discovered in the organic matter in Mn-rich sinters. A part of these copper sulfides most likely represents traces of MCOs, and this is the first report of traces of Mn-oxidizing enzyme in geological samples. Metagenomic analyses, surprisingly, indicated a close association of Mn oxides, not only in aerobic but also in anaerobic microbial communities. These new findings offer the unique and unified positions of Mn oxides, with roles that have not been ignored, to sustain anaerobic microbial communities in hot spring environments.


General Background
Manganese (Mn) is ubiquitous in the Earth's lithosphere and hydrosphere. Mn(II) is stable in solution under relatively acidic or anoxic conditions, whereas Mn(III) and Mn(IV) are favored under oxic or high pH conditions and, mainly, exist as Mn hydroxides, oxyhydroxides, or oxides [1]. The Mn cycling on the modern Earth is operated by shuttling between soluble Mn(II) and insoluble Mn(III) and Mn(IV). Mn(IV) oxides are found in diverse environments, including metal-contaminated streams [2,3], submarine hydrothermal fields [4,5], the ocean floor, where they occur as ferromanganese nodules and crusts [6][7][8], and terrestrial hot springs [9][10][11] (Table S1). Microbial Mn(II) oxidation is, generally, faster than abiotic Mn(II) oxidation processes [1]. This kinetic advantage implies that biological Mn(II) oxidation is thought to be widespread and significant, in natural environments over time [12][13][14]. However, little is known about the mechanisms of microbial Mn oxidation.
Direct mineralogical evidence of the biogenesity of Mn(IV) oxides is, still, obscure, due to the difficulties in distinguishing biogenic Mn(IV) oxides from abiotic Mn(IV) oxides. In contrast, several genetic pathways have been proposed for biogenic Mn(IV) oxides, including step-by-step enzymatic oxidation [15][16][17] and disproportionation of early biotic oxide [18]. Hence, more case studies of coupled examination of mineralogy, physiology, and enzymatic genomics, using natural samples, are required to, further, understand microbial Mn oxidation.

Site of Study
The Hachikuro (HK) hot spring is located in Akita, Japan ( Figure 1A,B). Hot spring water is anoxic and rich in bicarbonate, at a source-venting area with moderate concentrations of Fe(II) and Mn(II) ( Table S7). Ca-carbonate sinters of a mixture of aragonite and Fe-(hydro)oxides were deposited near the vent (Site 1, Figure 1C). The sinters were 2-5 cm thick and have laminated or columnar structures with a dark red color (sample HKs-Fe, Figure 1D). Some sinters were exposed to the air and covered by cyanobacterial mats. Black layers, confirmed as Mn oxides, commonly appeared beneath the cyanobacteria mats (e.g., sample HKs-Mn, Figure 1D). Fe-(hydro)oxides were less dominant in the downstream zones. Instead, the black-and brown-colored Mn-oxide layers became more significant. The hot spring water became slightly oxic, and the temperature decreased along the drainage path. The chemical composition, including Mn(II) concentration, did not change substantially (Table S7). Sinters at all sites were composed of Ca carbonates, but the chemistry changed from Fe-rich to Mn-rich, with increasing distance from the vent. Unsolidified carbonates were deposited at the midstream, approximately 10 m from the source (Site 2, sample HKm, Figure 1E,F). The thickness of the sediments was less than 5 cm, and they were covered by cyanobacterial mats. Black layers were, often, found beneath the cyanobacterial mat at this site. Sample HKd was a part of a Mn-rich soil that developed downstream, approximately 25 m from the source (Site 3). Sample HKd was brecciated and cemented with Mn oxides. Samples HKs-Mn, HKs-Fe, HKm, and HKd were used for further metagenomic, mineralogical, and geochemical analyses. After samples HKs-Mn, HKs-Fe, HKm, and HKd were retrieved, they were taken in sterile bags and, immediately, preserved in a freezer (−30 °C) in a laboratory. Weathered surfaces were removed to avoid contamination. Only the black part in the remnants were picked up and powdered, using a sterile mortar and pestle in a desktop clean bench for metagenomic and geochemical analyses.

Site of Study
The Hachikuro (HK) hot spring is located in Akita, Japan ( Figure 1A,B). Hot spring water is anoxic and rich in bicarbonate, at a source-venting area with moderate concentrations of Fe(II) and Mn(II) (Table S7). Ca-carbonate sinters of a mixture of aragonite and Fe-(hydro)oxides were deposited near the vent (Site 1, Figure 1C). The sinters were 2-5 cm thick and have laminated or columnar structures with a dark red color (sample HKs-Fe, Figure 1D). Some sinters were exposed to the air and covered by cyanobacterial mats. Black layers, confirmed as Mn oxides, commonly appeared beneath the cyanobacteria mats (e.g., sample HKs-Mn, Figure 1D). Fe-(hydro)oxides were less dominant in the downstream zones. Instead, the black-and brown-colored Mn-oxide layers became more significant. The hot spring water became slightly oxic, and the temperature decreased along the drainage path. The chemical composition, including Mn(II) concentration, did not change substantially (Table S7). Sinters at all sites were composed of Ca carbonates, but the chemistry changed from Fe-rich to Mn-rich, with increasing distance from the vent. Unsolidified carbonates were deposited at the midstream, approximately 10 m from the source (Site 2, sample HKm, Figure 1E,F). The thickness of the sediments was less than 5 cm, and they were covered by cyanobacterial mats. Black layers were, often, found beneath the cyanobacterial mat at this site. Sample HKd was a part of a Mn-rich soil that developed downstream, approximately 25 m from the source (Site 3). Sample HKd was brecciated and cemented with Mn oxides. Samples HKs-Mn, HKs-Fe, HKm, and HKd were used for further metagenomic, mineralogical, and geochemical analyses. After samples HKs-Mn, HKs-Fe, HKm, and HKd were retrieved, they were taken in sterile bags and, immediately, preserved in a freezer (−30 • C) in a laboratory. Weathered surfaces were removed to avoid contamination. Only the black part in the remnants were picked up and Life 2022, 12, 816 4 of 16 powdered, using a sterile mortar and pestle in a desktop clean bench for metagenomic and geochemical analyses.

Electron Microscope Observation
Micron-scale observations for the examined samples were performed, using fieldemission scanning electron microscopy (FE-SEM, JSM-7001F, JEOL, Tokyo, Japan). Crosssections of the alternating layers of Mn oxides and organic matter ( Figure S1)

Elemental and Stable Isotope Analyses
The elemental compositions of hot spring water at Site 1 and Site 3 as well as the trace elemental compositions of the Mn-rich soil (sample HKd) were measured, using inductively coupled plasma mass spectrometry (ICP-MS, Agilent 8800, Agilent, Santa Clara, CA, USA). The major elemental compositions of HKd were measured, using energy dispersive X-ray fluorescence (XRF, Epsilon 5, Panalytical, Almelo, The Netherlands). Concentrations of total carbon, total organic carbon, and total nitrogen were measured, using an elemental analyzer (Flash 2000, Thermo Fisher Scientific, Waltham, MA, USA). The total organic carbon content was determined, from the HCl-treated samples. Stable carbon and nitrogen isotope compositions were measured, using a Delta V Advantage Isotope Ratio Mass Spectrometer (EA-IRMS, Thermo Fisher Scientific, Waltham, MA, USA), through the Con-FlowIV interface. The methods of the stable isotope analyses were followed by [42]. Briefly, approximately 15-30 µg of C and 50 µg of N were used for analyses. The isotopic ratios are reported, as δ 13 C values against the international standard Pee Dee Belemnite (PDB), δ 13 C = {( 13 C/ 12 C) sample /( 13 C/ 12 C)V PDB − 1} × 1000, or as δ 15 N values against the international standard of atmospheric air, δ 15 N = {( 15 N/ 14 N) sample /( 15 N/ 14 N) Air − 1} × 1000. The precision of the isotope analyses was confirmed, by repeated analyses of in-house reference material (histidine calibrated against IAEA reference materials), as ±0.2‰ (1 standard deviation). The microbial compositions were clarified by 16S rRNA amplicon paired-end sequencing, using the Illumina MiSeq platform (Illumina, San Diego, CA, USA). Sequencing was conducted at Seibustu-Giken Co. Ltd. (Kanagawa, Japan), and the V4 region (515F-806R) of the 16S rRNA gene was amplified by the bacterial primers 515F (5 -GTGCCAGCMGCCGCGGTAA-3 ) and 806R (5 -GGACTACHVGGGTWTCTAAT-3 ). Quantification of the 1st PCR products was confirmed, using the QuantiFluor dsDNA System. A 2nd PCR was conducted, using sequencing primers. Library concentrations and qualities were measured, using a Synergy H1 microplate reader (BioTek, Agilent, Santa Clara, CA, USA) with a QuantiFluor dsDNA System and on a Fragment Analyzer (Agilent, Santa Clara, CA, USA) with a dsDNA 915 Reagent Kit (Agilent, Santa Clara, CA, USA), respectively, in accordance with the instructions of the manufacturer. Paired-end sequencing (2 × 150 bp) was performed on the Illumina MiSeq platform (Illumina, San Diego, CA, USA), with an MiSeq Reagent Kit v3 (Illumina, San Diego, CA, USA). All library preparation, pooling, quality controls, and sequencing were conducted at the Seibustu-Giken Co. Ltd.
Putative Mn-oxidizing genes were extracted, based on four criteria: (1) >30% homology with known Mn-oxidizing genes; (2) compatibility with domain; (3) preservation of amino acid sequences of the four conserved copper-binding sites of MCOs [28,81], T1, T2, T3a, and T3b sites ( Figure S2); and (4) location in the same clade with the assigned known Mn-oxidizing gene in the phylogenetic tree. The reference database of known Mn-oxidizing genes compiled in this study was screened against the concatenated protein FASTA sequences from MAGs, using BLASTP and HMMER ver. 3.3.2 (e-value cut-off of 10 −5 against Pfam-A database), to extract putative Mn-oxidizing genes and confirm their domain. Afterwards, possessing four conserved-copper-binding sites of MCOs and the phylogenetic relationship tree between the queried sequences and assigned, the known Mn-oxidizing genes were confirmed. Alignment was performed, using MAFFT (ver. 7) [82] with default parameters. After removing the suspicious or poorly aligned regions, using trimAI (ver. 1.4) [83], phylogenic trees were constructed, using RaxML (ver. 8.2.12) [84] with the command "raxmlHPC-PTHREADS -m PROTGAMMAAUTO -p 12345 -x 12345 -T 8 -N 100". MAGs possessing putative Mn-oxidizing genes, which met all the criteria, are listed in Table S5. References for the Mn-oxidizing genes used in this study are listed in Table S6.
Potential metabolic pathways were evaluated, using module completion ratios (MCR) and Q-value on Genomaple TM (ver. 2.3.2) [85]. Following the official manual, functional modules with a Q value < 0.5 were adopted, to consider the metabolic pathways of queried MAGs. Gene annotation was performed, using DFAST [86,87].

Biogenic Mn Oxides in Fe-and CO 2 -Rich Hot Spring
Assembled spherical Mn oxides (<5 µm in diameter, Figure 2A) in each sample were observed by FE-SEM. The spheres consisted of alternating layers of Mn oxides and organic matter ( Figure 2B-F). Such a spatial relationship between organic matter and Mn oxides is unique and has, rarely, been reported in modern marine Mn nodules or crusts. Organic matter is encrusted by linear, folded, and fibrous forms of Mn oxides, at the nanoscale ( Figures 3A,B,D,E and S1, HKs-Mn, HKm, and HKd). Most Mn oxides in the examined samples were amorphous in phases, but some of the Mn oxides showed randomly stacked lattices in the TEM images ( Figure 3B,E). High-angle annular dark field scanning (HAADF) TEM analyses revealed that the examined samples were composed of poorly crystalline phyllomanganates. δ-MnO 2 (vernadite, Figure 3C Figure 3A,B,D,E and Figure S1) are, apparently, different from synthetic triclinic MnO 2 [88]. The biogenesity and genetic sequences from vernadite, hausmannite, and birnessite have been discussed, by previous investigators [12,89,90]. Incubation experiments of Mn-oxidizing bacteria, also, produced poorly crystalline birnessite [39]. Our observations are consistent with the previously proposed biogenic origin models of these phyllomanganates [12,[89][90][91].

Biogenic Mn Oxides in Fe-and CO2-Rich Hot Spring
Assembled spherical Mn oxides (<5 µm in diameter, Figure 2A) in each sample were observed by FE-SEM. The spheres consisted of alternating layers of Mn oxides and organic matter ( Figure 2B-F). Such a spatial relationship between organic matter and Mn oxides is unique and has, rarely, been reported in modern marine Mn nodules or crusts. Organic matter is encrusted by linear, folded, and fibrous forms of Mn oxides, at the nanoscale ( Figures 3A,B,D,E and S1, HKs-Mn, HKm, and HKd). Most Mn oxides in the examined samples were amorphous in phases, but some of the Mn oxides showed randomly stacked lattices in the TEM images ( Figure 3B,E). High-angle annular dark field scanning (HAADF) TEM analyses revealed that the examined samples were composed of poorly crystalline phyllomanganates. δ-MnO2 (vernadite, Figure 3C), hausmannite (Mn 2+ Mn 3+ 2O4, Figure 3F), and birnessite ((Na,Ca,K) × (Mn 4+ , Mn 3+ 2)2O4·1.5H2O) were identified in the examined samples. TEM images (Figures 3A,B,D,E and S1) are, apparently, different from synthetic triclinic MnO2 [88]. The biogenesity and genetic sequences from vernadite, hausmannite, and birnessite have been discussed, by previous investigators [12,89,90]. Incubation experiments of Mn-oxidizing bacteria, also, produced poorly crystalline birnessite [39]. Our observations are consistent with the previously proposed biogenic origin models of these phyllomanganates [12,[89][90][91].
Previous studies have reported the enzymatic or biogenic formation of Mn-oxide nanoparticles as a nascent phase [33,92]. However, nanoparticles of Mn oxides were not found in the examined samples, suggesting early rapid merging of nanocrystals into larger polycrystals, in the hot spring environments. Tunnel-structured manganates, such as todorokite, are common in Mn nodules on the modern ocean floor [93][94][95], but these were not present in the examined samples in the present study.  Previous studies have reported the enzymatic or biogenic formation of Mn-oxide nanoparticles as a nascent phase [33,92]. However, nanoparticles of Mn oxides were not found in the examined samples, suggesting early rapid merging of nanocrystals into larger polycrystals, in the hot spring environments. Tunnel-structured manganates, such as todorokite, are common in Mn nodules on the modern ocean floor [93][94][95], but these were not present in the examined samples in the present study.

Complex Microbial Community at HK
Metagenomic analyses for biogenic Mn oxides indicated different phylogenies at each sampling point. The phylum-and class-level community compositions are illustrated in Figure 4 and Table S3. One-third of the operational taxonomic units (OTUs) at each site were, generally, composed of members of Gammaproteobacteria and Alphaproteobacteria. The abundance of Proteobacteria was the same in HKs-Mn and HKs-Fe.
Actinobacteriota and Bacteroidota are abundant in HKm and HKd, compared to HKs-Mn and HKs-Fe. Acidobacteriota and Chloroflexi were more abundant in HKm than in HKd. OTUs of Pastescibacteria, which was referred to as candidate phyla radiation, were abundant in HKd, but were not found in HKm. OTUs of Gallionellaceae were not found in HKm and HKd, corresponding to a lower abundance of Fe-(hydro)oxides. Cyanobacteria accounted for 6.5% and 2.3% of the microbial community in HKs-Fe and HKd, respectively. HKs-Mn and HKm, which were not exposed to the surface, did not show OTUs of cyanobacteria. The genera Pedobacter sp., Candidatus Kaiserbacteria sp., and Massilia sp. were also found in the Mn-rich samples (HKd).

Complex Microbial Community at HK
Metagenomic analyses for biogenic Mn oxides indicated different phylogenies at each sampling point. The phylum-and class-level community compositions are illustrated in Figure 4 and Table S3. One-third of the operational taxonomic units (OTUs) at each site were, generally, composed of members of Gammaproteobacteria and Alphaproteobacteria. The abundance of Proteobacteria was the same in HKs-Mn and HKs-Fe.
Actinobacteriota and Bacteroidota are abundant in HKm and HKd, compared to HKs-Mn and HKs-Fe. Acidobacteriota and Chloroflexi were more abundant in HKm than in HKd. OTUs of Pastescibacteria, which was referred to as candidate phyla radiation, were abundant in HKd, but were not found in HKm. OTUs of Gallionellaceae were not found in HKm and HKd, corresponding to a lower abundance of Fe-(hydro)oxides. Cyanobacteria accounted for 6.5% and 2.3% of the microbial community in HKs-Fe and HKd, respectively. HKs-Mn and HKm, which were not exposed to the surface, did not show OTUs of cyanobacteria. The genera Pedobacter sp., Candidatus Kaiserbacteria sp., and Massilia sp. were also found in the Mn-rich samples (HKd).

Bacteria Associated with Mn Oxidation
The taxonomy of MAGs was determined, based on 120 concatenated single-copy bacterial genes (Table S4). Nine MAGs possessed Mn-oxidizing genes, indicating the presence of putative Mn-oxidizing bacteria (Table S5). In the HKs-Mn, MAGs of putative Mnoxidizing bacteria, genus Rhizobiaceae_RCIO01 (HKs107), previously reported as Mn-oxidizing bacteria [96], was successfully constructed.
In the HKm, MAGs of the putative Mn-oxidizing bacteria Ramlibacter sp. (HKm46) was constructed (Table S5). The genus Ramlibacter sp. Is, generally, an aerobic heterotroph and has not been recognized as a Mn-oxidizing bacteria. Other MAGs of putative Mnoxidizing bacteria were assigned to thermophile, the class of Blastocatellia (HKm2), which, also, has not been reported to have Mn-oxidation capacity.
Sample HKd showed different characteristics. MAGs of the putative Mn-oxidizing bacteria, Herminiimonas sp. (HKm161) and Hydrogenophaga sp. (HKd102) were constructed (Table S5). Herminiimonas sp. has not been recognized as Mn-oxidizing bacteria, but Hydrogenophaga sp. was, previously, reported as Mn-oxidizing bacteria [46]. These are candidates for major Mn-oxidizing bacteria at each site. Beside those common Mn-oxidizing bacteria, some anaerobic bacteria were found to have Mn-oxidizing genes (see Section 3.5).

MCOs Utilization for Biological Mn Oxidation in Nature
Our metagenomic data indicated the prevalence of putative Mn-oxidizing genes encoding MCOs in the examined samples (Table S5). Among the nine MAGs in the present study, moxA (locus ID; CAJ19378) was the top hit, with an identity of approximately 70% of the HKs-Mn, HKm, and HKd (HKs107, HKm46, HKd161), respectively. Other putative Mn-oxidizing genes in HKs-Mn (HKs85, 166, 176, 177) were moxA, mcoA (locus ID; ABY98562), and mnxG (locus ID; PputGB1_2447), while those in HKm and HKd were

Bacteria Associated with Mn Oxidation
The taxonomy of MAGs was determined, based on 120 concatenated single-copy bacterial genes (Table S4). Nine MAGs possessed Mn-oxidizing genes, indicating the presence of putative Mn-oxidizing bacteria (Table S5). In the HKs-Mn, MAGs of putative Mn-oxidizing bacteria, genus Rhizobiaceae_RCIO01 (HKs107), previously reported as Mnoxidizing bacteria [96], was successfully constructed.
In the HKm, MAGs of the putative Mn-oxidizing bacteria Ramlibacter sp. (HKm46) was constructed (Table S5). The genus Ramlibacter sp. Is, generally, an aerobic heterotroph and has not been recognized as a Mn-oxidizing bacteria. Other MAGs of putative Mn-oxidizing bacteria were assigned to thermophile, the class of Blastocatellia (HKm2), which, also, has not been reported to have Mn-oxidation capacity.
Sample HKd showed different characteristics. MAGs of the putative Mn-oxidizing bacteria, Herminiimonas sp. (HKm161) and Hydrogenophaga sp. (HKd102) were constructed (Table S5). Herminiimonas sp. has not been recognized as Mn-oxidizing bacteria, but Hydrogenophaga sp. was, previously, reported as Mn-oxidizing bacteria [46]. These are candidates for major Mn-oxidizing bacteria at each site. Beside those common Mn-oxidizing bacteria, some anaerobic bacteria were found to have Mn-oxidizing genes (see Section 3.5).

MCOs Utilization for Biological Mn Oxidation in Nature
Our metagenomic data indicated the prevalence of putative Mn-oxidizing genes encoding MCOs in the examined samples (Table S5). Among the nine MAGs in the present study, moxA (locus ID; CAJ19378) was the top hit, with an identity of approximately 70% of the HKs-Mn, HKm, and HKd (HKs107, HKm46, HKd161), respectively. Other putative Mnoxidizing genes in HKs-Mn (HKs85, 166, 176, 177) were moxA, mcoA (locus ID; ABY98562), and mnxG (locus ID; PputGB1_2447), while those in HKm and HKd were moxA and mcoA Life 2022, 12, 816 9 of 16 (HKm2 and HKd102). These data confirm that MCOs were the dominant Mn-oxidizing genes around venting and downstream sites.
Nanoscale textures and chemistry of organic matter in Mn-oxide spherules were analyzed, using high-resolution transmission electron microscopy (HR-TEM) and STEM. Cu-bearing nuggets (<300 nm, mostly 100-200 nm Figure 5A) have been, newly, found in the spherules. Such nuggets only occurred in organic matter in the spherules, and Mn oxides or carbonates in the same spherules never contained the Cu-bearing nuggets. Quantitative analyses by HR-TEM indicated that the nuggets were mostly made of Cu x S y ( Figure 5B-F), although the determination of specific stoichiometry was difficult. Cu-bearing nuggets are relevant for natural covellite (CuS) or chalcocite (Cu 2 S). Such Cu-bearing nuggets in organic matter have not been reported, previously, in terrestrial hot spring environments. The stability field of Cu x S y was estimated, using the chemical data of HK hot spring water ( Figure S3). The stability field was incompatible with the conditions of the samples, in which aragonite and goethite precipitate. These facts suggest that abiotic precipitation of Cu x S y from hot spring water is not thermodynamically favored. In addition, FeS 2 or FeS were not found in the examined samples, although hot spring water contains significant amounts of Fe 2+ . This suggests that Cu x S y was not a simple product of microbial sulfate reduction (e.g., [97]). In nature, Mn oxides act as sponges to adsorb trace elements (e.g., [98]). Mn nodules or crusts on the modern ocean floor are known to abiotically accumulate Cu and other heavy metals, and they are comparable to several hundreds to thousands parts per million (ppm) (e.g., [99]). On the other hand, the Mn oxides in the examined samples did not show the enrichment of Cu (1.5 ppm) and other heavy metals (Table S8). Such observations suggest a unique mechanism to form Cu x S y in the examined samples, rather than simple adsorption and enrichment on the surfaces of Mn oxides. We interpret this to mean that the novel Cu nuggets are traces of MCOs, after significant diagenetic modification from their original forms.

Role of Mn Oxidation in the Sinter Ecosystem
Phylogenetic analyses indicated that the microbial communities in Mn oxides differed at the sampling locality. Mn oxides in the venting site harbored a remarkable proportion of anaerobic microorganisms, such as sulfate-reducing bacteria (SRB) and Mnreducing bacteria. In contrast, Mn oxides at the downstream harbored the aerobic hetero- Nano-scale aggregations of biogenic metal sulfides within organic matter were reported, previously, from natural samples [100]. Metals in metal-binding proteins are bound with sulfur in amino acids, proteins, and polypeptides belonging to the sulfhydryl group. Nano-particles of metal sulfides were formed, after degradation of the original proteinmetal compounds [97]. Similar nano-particles of various metal sulfides (e.g., Zn, Hg, Fe, Cd) have been found, in natural organic-rich samples. It is interpreted that Cu-binding proteins (MCOs) were degraded after cell death, and Cu and sulfur from organic molecules were trapped in non-permeable organic layers cemented in carbonates. H 2 S from deep sulfate reduction might join, as a part of sulfur, in this closed system. Aggregations were promoted by binding protein-rich organic matter with metal sulfides. In particular, cysteine stimulates large aggregations, up to~100 nm diameter [100]. MCOs, generally, contain cysteines bound with Cu. Such high concentrations of Cu and organic molecules, including cysteine in closed systems, were responsible for Cu x S y formation in the examined samples. Other organic sulfurs, also, contribute to form Cu sulfides. The finding of Cu x S y is consistent with the detection of genes encoding MCOs in the same samples, supporting that MCOs were the major Mn-oxidizing genes in the venting area and downstream sites.

Role of Mn Oxidation in the Sinter Ecosystem
Phylogenetic analyses indicated that the microbial communities in Mn oxides differed at the sampling locality. Mn oxides in the venting site harbored a remarkable proportion of anaerobic microorganisms, such as sulfate-reducing bacteria (SRB) and Mn-reducing bacteria. In contrast, Mn oxides at the downstream harbored the aerobic heterotrophs. Putative Mn-oxidizing bacteria at the venting site were different from those at downstream sites.
The temperature of hot spring water was lower and more oxic at downstream sites, compared to the venting site. These factors are considerable reasons for the differences in the microbial community structures and Mn-oxidizing bacteria at each site ( Figure 4, and Table S3). At both sites, biological Mn oxidation benefits the entire microbial community, and Mn oxides are utilized as electron acceptors. Alternatively, Mn oxides are utilized for the degradation and storage of organic matter in the microbial community [101,102].
Our analyses indicate that the following anaerobic bacteria have putative Mn-oxidizing genes: Thermodesulfovibrionales (HKs177), Desulfobacterota (HKs166), Thermoanaerobaculia (HKs85), and Chloracidobacteriales (HKs176) ( Table S5). Desulfobacterota and Thermodesulfovibrionales were SRB. Finding putative Mn-oxidizing genes in those anaerobic bacteria is enigmatic, and it is still uncertain whether those bacteria are actively oxidizing Mn(II) at the examined site.
Recent incubations of anaerobic phototrophs [40] and aerobic chemolithoautotrophs [39] showed microbial Mn(II) oxidation, with a help from other aerobic and anaerobic microbial communities. These studies indicate the importance of biogenic Mn oxides, for developing microbial communities at the interface of oxic and anoxic environments.
Yu and Leadbetter (2020) suggested that the class of putative Mn-oxidizing bacteria are phylogenetically closed to the phylum of Nitropsirae, which contains SRB classes. Our findings and previous results, further, imply that SRB, potentially, acquired the anaerobic Mn-oxidizing ability during its evolution, although conclusive evidence is, still, unavailable. Yu and Leadbetter (2020) proposed metabolic pathways for chemoautotrophic Mn oxidation and emphasized the postulation of Fe-S clusters with Cu-bearing protein, to transfer electrons in vitro. This model, further, implies the necessity of sulfur for Mn oxidation. The presence of SRB might be beneficial for Mn-oxidizing bacteria in the same microbial community, so that Mn-oxidizing bacteria could uptake essential sulfur species easily from SRB. This could be alternative explanation for detection of Mn oxides in SRB-bearing complexed microbial community. The presence of Mn oxides was also suggested to be beneficial to SRB for energy conservation (i.e., buttery) through metabolic electron transfers. This, further, implies that the inorganic Mn oxides are unified with microbial mats and have essential roles to sustain anaerobic microbial communities.

Conclusions
Geochemical, mineralogical, and metagenomic analyses were performed on Mn-oxiderich sinters in Japan. Sub-micron scale spherical aggregates of Mn oxides were observed. HAADF analyses revealed that the Mn oxides were composed of poorly crystalline phyllomanganates, including δ-MnO 2 (vernadite), hausmannite (Mn 2+ Mn 3+ 2 O 4 ), and birnessite ((Na,Ca,K) × (Mn 4+ , Mn 3+ 2 ) 2 O 4 ·1.5H 2 O). Nanoscale layers of Mn oxides in each sphere were, often, intercalated with layers of organic matter, which are, rarely, found in marine Mn crust or nodules. The low crystallinities of the spherical Mn oxides and their close associations with organic matter support the biogenic origin of Mn oxides.
Several putative Mn-oxidizing genes encoding MCOs were identified, using metagenomic analyses. The predominant putative Mn-oxidizing genes were moxA and mcoA. Nanoscale nuggets of copper sulfides were, also, discovered in the layers of organic matter. Thermodynamic calculations indicated that conditions in the examined hot spring environment were not favorable for the abiotic precipitation of copper sulfides. In addition, other mineralogical and geochemical data excluded the possibility of the product, by microbial sulfate reduction or simple adsorption and enrichment on the surfaces of Mn oxides. Therefore, the novel copper sulfides are, most likely, degradation products of MCO-bearing proteins.
Enzymatically produced Mn oxides, most likely, acted as electron acceptors or helped in the degradation and storage of organic matter. These actions would help sustain and develop the overall aerobic and anaerobic microbial communities. Nine MAGs of putative Mn-oxidizing bacteria were detected. In particular, four of them appeared to be close associations of Mn-oxidizing genes with anaerobic bacteria, including SRB, although there was high uncertainty regarding whether anaerobic bacteria anaerobically oxidized Mn(II). The findings of the present study suggest that Mn oxides became a part of meso to thermophilic microbial mats and offer essential roles to sustain anaerobic microbial communities.  Figure S2: Alignment of four conserved copper binding sites among putative Mn-oxidizing genes in nine MAGs in this study and known Mn-oxidizing genes, Figure S3: Eh-pH diagram of the Cu-S-Fe-O system with coexisting stability area for both aragonite and goethite, described by Geochemist Workbench Standard 12.0. Yellow represents the stability field of Cu x S y (in this case covellite and chalcocite). Red represents the coexisting stability area for both aragonite and goethite. Blue indicates the pH conditions of the hot spring in HK. Calculations are conducted on the basis of the water chemistry of hot spring as follows: aCu = 10 −7.06 , aCa = 10 −1.796 , aHCO 3 = 10 −1.678 , aNa + = 10 −1.678 , aCl − = 10 −1.602 , aMg 2+ = 10 −2.301 , aSO4 2-= 10 −2.097 , aFe 2+ = 10 −4.162 , aMn 2+ = 10 −4.189 . (References [103][104][105][106][107][108][109][110][111][112][113][114][115][116][117][118] are cited in the supplementary materials).
Author Contributions: Y.T. and T.K. designed and wrote the overall manuscript; Y.T. conducted all experiments, analyses, and bioinformatics. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author, Y.T.