Phylogeny and Metabolic Potential of the Methanotrophic Lineage MO3 in Beijerinckiaceae from the Paddy Soil through Metagenome-Assembled Genome Reconstruction

Although the study of aerobic methane-oxidizing bacteria (MOB, methanotrophs) has been carried out for more than a hundred years, there are many uncultivated methanotrophic lineages whose metabolism is largely unknown. Here, we reconstructed a nearly complete genome of a Beijerinckiaceae methanotroph from the enrichment of paddy soil by using nitrogen-free M2 medium. The methanotroph labeled as MO3_YZ.1 had a size of 3.83 Mb, GC content of 65.6%, and 3442 gene-coding regions. Based on phylogeny of pmoA gene and genome and the genomic average nucleotide identity, we confirmed its affiliation to the MO3 lineage and a close relationship to Methylocapsa. MO3_YZ.1 contained mxaF- and xoxF-type methanol dehydrogenase. MO3_YZ.1 used the serine cycle to assimilate carbon and regenerated glyoxylate through the glyoxylate shunt as it contained isocitrate lyase and complete tricarboxylic acid cycle-coding genes. The ethylmalonyl-CoA pathway and Calvin–Benson–Bassham cycle were incomplete in MO3_YZ.1. Three acetate utilization enzyme-coding genes were identified, suggesting its potential ability to utilize acetate. The presence of genes for N2 fixation, sulfur transformation, and poly-β-hydroxybutyrate synthesis enable its survival in heterogeneous habitats with fluctuating supplies of carbon, nitrogen, and sulfur.


Introduction
Aerobic methane-oxidizing bacteria or methanotrophs are a distinct group of bacteria that use methane as their main carbon and energy source [1,2]. The currently described aerobic methanotrophs are affiliated to Alphaproteobacteria (also known as type II), Gammaproteobacteria (type I), and Verrucomicrobia. The two methanotrophic families within Alphaproteobacteria are Methylocystaceae and Beijerinckiaceae [3][4][5]. These methanotrophs convert methane to methanol by using methane monooxygenase (MMO), which exists in particulate (pMMO) or soluble (sMMO) forms [2]. The pmoA gene encoding the beta-subunit of pMMO is present in all aerobic methanotrophs except Methylocella, Methyloferula, and a species of Methyloceanibacter [6][7][8]. The phylogenetic analysis of pmoA gene sequences in the GenBank database shows that about 20 pmoA lineages contain cultured representatives, and there are also more than 20 pmoA lineages have no cultured representative, such as upland soil cluster alpha (USCα), upland soil cluster gamma (USCγ), Rice Paddy Clusters, and the Lake Washington Clusters [9,10].
Currently, the analysis of metagenome-assembled genomes (MAGs) is an important approach to investigate the metabolism of these uncultivated lineages and some novel methanotrophs [11]. The reconstruction and analysis of a MAG of USCγ (type I) confirmed the presence of a nearly complete serine pathway of type II methanotrophs rather than the

Phylogeny Analysis
The full lengths of pmoA, nifH, and 16S rRNA genes extracted from methanotrophic MAGs were used to construct phylogenetic trees by using MEGA (version 6.06) to infer their phylogeny among known methanotrophs. A maximum-likelihood phylogenomic tree was also constructed with the FastTree v.2.1.10 [47] and visualized with ITOL after identifying and aligning a concatenated set of 120 marker proteins by using the GTDB-Tk v1.7.0 [48]. The genomic average nucleotide identity (gANI) and genomic average aminoacid identity (gAAI) values among methanotrophic MAGs and their related genomes were calculated by JSpeciesWS Online Service [49] and CompareM (https://github.com/dparks1 134/CompareM accessed on 4 April 2022), respectively. Tools of Kostas lab were also used to calculate gANI and gAAI [50].

Succession of MOB in N-Free Medium
According to the amplicon-sequencing results of the 16S rRNA gene ( Figure 1A), MOB accounts for about 1.6% of the total microorganisms in the original paddy soil. Their proportion in the soil reached 55.9% after the headspace CH 4 was consumed, and after four additional rounds of enrichment in nitrogen-free liquid M2 medium, their proportion stabilized at about 36%. The dominant methanotroph in soil after enrichment is Methylosarcina (type I), which accounts for 84.3% of total methanotrophs. However, after four rounds of enrichment in N-free M2 medium, the dominant methanotrophs gradually changed into unclassified type II (Methylocystaceae), suggesting they are some novel taxa that have not been well-characterized. On the basis of the amplicon sequencing of the pmoA gene, we obtained similar results. After four rounds of enrichment, the dominant MOB is rapidly transformed from Methylosarcina to Methylosinus and MO3, of which the latter accounts for 33.4% of total MOB ( Figure 1B). The actual proportion of MO3 may be much higher, because Beijerinckiaceae methanotrophs (type IIb) to which MO3 belongs generally have a single pmoCAB operon [22,51], whereas Methylocystaceae (type IIa) and other type I methanotrophs commonly have two pmoCAB operons in their genomes [52][53][54].
dominant MOB is rapidly transformed from Methylosarcina to Methylosinus and MO3, of which the latter accounts for 33.4% of total MOB ( Figure 1B). The actual proportion of MO3 may be much higher, because Beijerinckiaceae methanotrophs (type IIb) to which MO3 belongs generally have a single pmoCAB operon [22,51], whereas Methylocystaceae (type IIa) and other type I methanotrophs commonly have two pmoCAB operons in their genomes [52][53][54]. In most methanotroph-enrichment experiments using paddy soil, MO3 is rarely enriched [25,55,56]. We are not able to enrich it with NMS (nitrate mineral salts), nitrate-free NMS, and M2 media. The M2 medium is a fivefold dilution of M1 medium and is first designed for methanotrophs from freshwater wetlands and mildly acidic soils [27], and nitrate-free M2 medium is subsequently successfully used for enrichment and/or maintenance of multiple strains of Beijerinckiaceae methanotrophs, such as Methylocella palustris [57], Methylocapsa acidiphila [19], Methylocella tundra [58], and Methylocapsa palsarum [59]. Therefore, MO3 should have physiological characteristics similar to other Beijerinckiaceae methanotrophs, such as the ability to fix N2 and low-concentration inorganic salt requirements.

Reconstruction of MO3 MAGs
DNA from the fourth round of enrichment is used for metagenomic sequencing. After reads assembly using three methods and contig binning using two methods, we obtained seven high-quality MOB MAGs (Table S1) with completeness > 92.5% and contamination < 2.63%. According to the classification results of GTDBkit, three MAGs belong to Methylomagnum, one MAG belongs to Methylosinus, and three MAGs belong to unknown Beijerinckiaceae. The gANI similarity of these three Beijerinckiaceae MAGs is over 99.6%, indicating that they belong to the same species, of which Bin.033 contains only eight contigs with completeness of 98.59% and contamination of 0.75% (Table 1). In In most methanotroph-enrichment experiments using paddy soil, MO3 is rarely enriched [25,55,56]. We are not able to enrich it with NMS (nitrate mineral salts), nitrate-free NMS, and M2 media. The M2 medium is a fivefold dilution of M1 medium and is first designed for methanotrophs from freshwater wetlands and mildly acidic soils [27], and nitrate-free M2 medium is subsequently successfully used for enrichment and/or maintenance of multiple strains of Beijerinckiaceae methanotrophs, such as Methylocella palustris [57], Methylocapsa acidiphila [19], Methylocella tundra [58], and Methylocapsa palsarum [59]. Therefore, MO3 should have physiological characteristics similar to other Beijerinckiaceae methanotrophs, such as the ability to fix N 2 and low-concentration inorganic salt requirements.

Reconstruction of MO3 MAGs
DNA from the fourth round of enrichment is used for metagenomic sequencing. After reads assembly using three methods and contig binning using two methods, we obtained seven high-quality MOB MAGs (Table S1) with completeness > 92.5% and contamination < 2.63%. According to the classification results of GTDBkit, three MAGs belong to Methylomagnum, one MAG belongs to Methylosinus, and three MAGs belong to unknown Beijerinckiaceae. The gANI similarity of these three Beijerinckiaceae MAGs is over 99.6%, indicating that they belong to the same species, of which Bin.033 contains only eight contigs with completeness of 98.59% and contamination of 0.75% (Table 1). In addition, we detected a complete operon of ribosomal rRNA genes and complete operon of pmoCAB and nifHDKENX genes in Bin.033 ( Figure 2, Tables S2-S4). Therefore, MAG Bin.033 was selected for subsequent analysis and labeled as MO3_YZ.1 (YZ indicates that this MAG originates from the soil sample collected from Yangzhou City). addition, we detected a complete operon of ribosomal rRNA genes and complete operon of pmoCAB and nifHDKENX genes in Bin.033 ( Figure 2, Tables S2-S4). Therefore, MAG Bin.033 was selected for subsequent analysis and labeled as MO3_YZ.1 (YZ indicates that this MAG originates from the soil sample collected from Yangzhou City).   Table S4 shows the full names of enzymes encoded by these genes. The other three contigs less than 30k in length were not shown.

Phylogeny of MO3
MO3_YZ.1 has a pmoA gene length of 873 bp, is within the pmoA length range of type IIb (Beijerinckiaceae) methanotrophs, and is much longer than that of other methanotrophs ( Figure 3, Table S5). The length of the pmoA gene can also serve as a taxonomic feature of methanotrophs. The pmoA genes of most type I methanotrophs are 744 bp in length, and only a few genera of type Ia such as Methylomarinum, Methylomonas, and Methyloprofundus have pmoA genes of 750 bp in length. Type IIa methanotrophs, including all species of Methylocystis and Methylosinus, have pmoA genes of 759 bp in length except Methylocystis bryophila S285 (762 bp). When the length of the pmoA-like sequence is 771 or 753 bp, it must be pmoA2 or pxmA ( Figure 3, Table S5). Therefore, in the future, when analyzing a methanotrophic MAG, the length of its pmoA gene sequence can help us make a preliminary judgment on the taxa to which it belongs.  Table S4 shows the full names of enzymes encoded by these genes. The other three contigs less than 30k in length were not shown.

Phylogeny of MO3
MO3_YZ.1 has a pmoA gene length of 873 bp, is within the pmoA length range of type IIb (Beijerinckiaceae) methanotrophs, and is much longer than that of other methanotrophs ( Figure 3, Table S5). The length of the pmoA gene can also serve as a taxonomic feature of methanotrophs. The pmoA genes of most type I methanotrophs are 744 bp in length, and only a few genera of type Ia such as Methylomarinum, Methylomonas, and Methyloprofundus have pmoA genes of 750 bp in length. Type IIa methanotrophs, including all species of Methylocystis and Methylosinus, have pmoA genes of 759 bp in length except Methylocystis bryophila S285 (762 bp). When the length of the pmoA-like sequence is 771 or 753 bp, it must be pmoA2 or pxmA (Figure 3, Table S5). Therefore, in the future, when analyzing a methanotrophic MAG, the length of its pmoA gene sequence can help us make a preliminary judgment on the taxa to which it belongs.
The phylogenetic analysis of the pmoA gene from MO3_YZ.1 confirms its affiliation to the MO3 lineage, which is closely related but distinct from Methylocapsa, the sole pmoAcontaining genus of Beijerinckiaceae ( Figure 4A). The phylogeny of nifH genes also shows a close relationship of MO3_YZ.1 to Beijerinckiaceae methanotrophs ( Figure S1). However, when its 16S rRNA gene is used for phylogenetic tree construction, MO3_YZ.1 undoubtedly falls into the group of Methylosysits/Methylosinus, i.e., Methylocystaceae methanotrophs (type IIa, Figure 4B), and shows 98.5% of 16S rRNA sequence identity with Methylosinus sp. C49. The phylogenomic tree based on a concatenated set of 120 marker proteins confirms the placement of the MO3_YZ.1 within Beijerinckiaceae ( Figure 5A). The maximum values of gANI and gAAI between MO3_YZ.1 and other known Beijerinckiaceae MOB genomes are 74% (by JSpeciesWS Online Service) and 71% (by CompareM), respectively ( Figure 5B). When tools of Kostas lab are used for calculation, the maximum values of gANI and gAAI are 79% and 69%, respectively ( Figure S2). Based on these similarity values, whether MO3 should be a new genus of Beijerinckiaceae or a new species of Methylocapsa cannot be concluded yet.  Table S5 shows more details.
The phylogenetic analysis of the pmoA gene from MO3_YZ.1 confirms its affiliation to the MO3 lineage, which is closely related but distinct from Methylocapsa, the sole pmoA-containing genus of Beijerinckiaceae ( Figure 4A). The phylogeny of nifH genes also shows a close relationship of MO3_YZ.1 to Beijerinckiaceae methanotrophs ( Figure S1). However, when its 16S rRNA gene is used for phylogenetic tree construction, MO3_YZ.1 undoubtedly falls into the group of Methylosysits/Methylosinus, i.e., Methylocystaceae methanotrophs (type IIa, Figure 4B), and shows 98.5% of 16S rRNA sequence identity with Methylosinus sp. C49. The phylogenomic tree based on a concatenated set of 120 marker proteins confirms the placement of the MO3_YZ.1 within Beijerinckiaceae ( Figure  5A). The maximum values of gANI and gAAI between MO3_YZ.1 and other known Beijerinckiaceae MOB genomes are 74% (by JSpeciesWS Online Service) and 71% (by CompareM), respectively ( Figure 5B). When tools of Kostas lab are used for calculation, the maximum values of gANI and gAAI are 79% and 69%, respectively ( Figure S2). Based on these similarity values, whether MO3 should be a new genus of Beijerinckiaceae or a new species of Methylocapsa cannot be concluded yet.    Table S5 shows more details.
The phylogenetic analysis of the pmoA gene from MO3_YZ.1 confirms its affiliation to the MO3 lineage, which is closely related but distinct from Methylocapsa, the sole pmoA-containing genus of Beijerinckiaceae ( Figure 4A). The phylogeny of nifH genes also shows a close relationship of MO3_YZ.1 to Beijerinckiaceae methanotrophs ( Figure S1). However, when its 16S rRNA gene is used for phylogenetic tree construction, MO3_YZ.1 undoubtedly falls into the group of Methylosysits/Methylosinus, i.e., Methylocystaceae methanotrophs (type IIa, Figure 4B), and shows 98.5% of 16S rRNA sequence identity with Methylosinus sp. C49. The phylogenomic tree based on a concatenated set of 120 marker proteins confirms the placement of the MO3_YZ.1 within Beijerinckiaceae ( Figure  5A). The maximum values of gANI and gAAI between MO3_YZ.1 and other known Beijerinckiaceae MOB genomes are 74% (by JSpeciesWS Online Service) and 71% (by CompareM), respectively ( Figure 5B). When tools of Kostas lab are used for calculation, the maximum values of gANI and gAAI are 79% and 69%, respectively ( Figure S2). Based on these similarity values, whether MO3 should be a new genus of Beijerinckiaceae or a new species of Methylocapsa cannot be concluded yet.   The phylogenies of 16S rRNA and pmoA genes from MO3_YZ.1 are not congruent as they affiliate to different families. Such case has not been reported within the known type II methanotrophs. The 16S rRNA genes often fail to assemble and bin due to their conserved and repetitive nature [60]. It should be treated with caution when the 16S rRNA gene of one MAG appears incongruent taxonomic classification with the taxonomic identity of this MAG [61]. Due to the conservation of the 16S rRNA gene, it is expected that the 16S rRNA gene of MO3_YZ.1 should be most related to Methylocapsa. Therefore, in this study, the assembled Methylosinus-like 16S rRNA gene in MO3_YZ.1 very likely does not belong to this MAG. It may be a fragment of contaminating sequence from a Methylosinus species due to the large proportion of Methylosinus in the enriched culture.

Methane-Oxidation Pathway of MO3
We reconstructed the central metabolic pathways of MO3 on the basis of the gene-function annotation of MO3_YZ.1 ( Figure 6). MO3_YZ.1 possesses a complete operon of pmoCAB genes coding the particulate methane monooxygenase and has two orphan pmoC genes ( Figure 2). According to alignment of the deduced amino-acid sequences of pmoA genes, the amino acid of His38, Met42, Asp47, Asp49, and Glu100 for the tricopper cluster site is highly conserved in MO3_YZ.1 and other methanotrophs ( Figure S3) as previously reported [62]. Like Methylocapsa species, other pmoA-like genes (pxmA and pmoA2) and the soluble methane monooxygenase coding genes are absent in MO3_YZ.1 [3]. We further identified coding genes of mxaF-and xoxF-type methanol dehydrogenase (MDH), which require calcium and lanthanide in their active center, respectively [63,64]. The xoxF-type MDH is a homodimer of the canonical mxaF-type MDH, and appears to be more widespread than the later. The xoxF-type MDH uses rare-earth elements as part of its catalytic center, and therefore the expression and activity of these two MDHs depends on the availability of rare-earth elements [63]. The xoxF gene of MO3 shows an amino-acid identity of 86.6% to that of Methylocapsa aurea (WP_036262132), and The phylogenies of 16S rRNA and pmoA genes from MO3_YZ.1 are not congruent as they affiliate to different families. Such case has not been reported within the known type II methanotrophs. The 16S rRNA genes often fail to assemble and bin due to their conserved and repetitive nature [60]. It should be treated with caution when the 16S rRNA gene of one MAG appears incongruent taxonomic classification with the taxonomic identity of this MAG [61]. Due to the conservation of the 16S rRNA gene, it is expected that the 16S rRNA gene of MO3_YZ.1 should be most related to Methylocapsa. Therefore, in this study, the assembled Methylosinus-like 16S rRNA gene in MO3_YZ.1 very likely does not belong to this MAG. It may be a fragment of contaminating sequence from a Methylosinus species due to the large proportion of Methylosinus in the enriched culture.

Methane-Oxidation Pathway of MO3
We reconstructed the central metabolic pathways of MO3 on the basis of the genefunction annotation of MO3_YZ.1 (Figure 6). MO3_YZ.1 possesses a complete operon of pmoCAB genes coding the particulate methane monooxygenase and has two orphan pmoC genes (Figure 2). According to alignment of the deduced amino-acid sequences of pmoA genes, the amino acid of His38, Met42, Asp47, Asp49, and Glu100 for the tricopper cluster site is highly conserved in MO3_YZ.1 and other methanotrophs ( Figure S3) as previously reported [62]. Like Methylocapsa species, other pmoA-like genes (pxmA and pmoA2) and the soluble methane monooxygenase coding genes are absent in MO3_YZ.1 [3]. We further identified coding genes of mxaFand xoxF-type methanol dehydrogenase (MDH), which require calcium and lanthanide in their active center, respectively [63,64]. The xoxFtype MDH is a homodimer of the canonical mxaF-type MDH, and appears to be more widespread than the later. The xoxF-type MDH uses rare-earth elements as part of its catalytic center, and therefore the expression and activity of these two MDHs depends on the availability of rare-earth elements [63]. The xoxF gene of MO3 shows an aminoacid identity of 86.6% to that of Methylocapsa aurea (WP_036262132), and more than 79% to that of other Beijerinckiaceae methanotrophs, such as Methylocapsa palsarum NE2 [51], Ca. Methyloaffinis lahnbergensis [13], and Methylocella silvestris [65]. We also recovered a complete gene set of the tetrahydromethanopterin-dependent pathway (H 4 MPT pathway) for C1-carbon transfer during the oxidation of formaldehyde to formate, and fdh gene for the nonreversible formate dehydrogenase. MO3_YZ.1 catalyzes the final oxidation step of formate to CO 2 and produces NADH, which can further drive the production of ATP through the respiratory chain. However, neither the coding genes of the carbon-monoxide dehydrogenase nor those of [NiFe] hydrogenase are identified in MO3_YZ.1, suggesting that MO3 cannot use CO and H 2 as alternative energy sources as Methylocapsa gorgona MG08 [22].
Microorganisms 2022, 10, x FOR PEER REVIEW 8 of 13 more than 79% to that of other Beijerinckiaceae methanotrophs, such as Methylocapsa palsarum NE2 [51], Ca. Methyloaffinis lahnbergensis [13], and Methylocella silvestris [65]. We also recovered a complete gene set of the tetrahydromethanopterin-dependent pathway (H4MPT pathway) for C1-carbon transfer during the oxidation of formaldehyde to formate, and fdh gene for the nonreversible formate dehydrogenase. MO3_YZ.1 catalyzes the final oxidation step of formate to CO2 and produces NADH, which can further drive the production of ATP through the respiratory chain. However, neither the coding genes of the carbon-monoxide dehydrogenase nor those of [NiFe] hydrogenase are identified in MO3_YZ.1, suggesting that MO3 cannot use CO and H2 as alternative energy sources as Methylocapsa gorgona MG08 [22].

Carbon Assimilation of MO3
We detected a complete gene set of the serine cycle for the assimilation of C1 from formate. Formate was condensed with tetrahydrofolate (H4F) to form formyl-H4F, which was transformed to methylene-H4F via the H4F pathway, and then methylene-H4F reacted with glycine to form serine ( Figure 6). The regeneration of glyoxylate is a key pathway for the carbon assimilation of type II methanotrophs possessing serine cycle [66]. The coding gene (aceA) of the key enzyme (isocitrate lyase) of glyoxylate shunt and a complete gene set of the tricarboxylic acid (TCA) cycle in MO3_YZ.1 are observed, implying that the acetyl-CoA produced in the serine cycle can be subsequently oxidized to glyoxylate in assistance of some TCA cycle enzymes. This regeneration pathway of glyoxylate is common in type IIb but absent in type IIa methanotrophs, which use the ethylmalonyl-CoA (EMC) pathway to accomplish the same task [67]. Although many encoding genes of the EMC pathway-related enzymes are also detected in MO3_YZ.1, the encoding genes of four enzymes are absent (croR for 3-hydroxybutyryl-CoA dehydratase, ccr for crotonyl-CoA carboxylase/reductase, msd for 2-methylfumaryl-CoA hydratase and mcd for methenyltetrahydromethanopterin cyclohydrolase), indicating that MO3, like other Beijerinckiaceae methanotrophs, cannot regenerate glyoxylate through the EMC

Carbon Assimilation of MO3
We detected a complete gene set of the serine cycle for the assimilation of C1 from formate. Formate was condensed with tetrahydrofolate (H 4 F) to form formyl-H 4 F, which was transformed to methylene-H 4 F via the H 4 F pathway, and then methylene-H 4 F reacted with glycine to form serine ( Figure 6). The regeneration of glyoxylate is a key pathway for the carbon assimilation of type II methanotrophs possessing serine cycle [66]. The coding gene (aceA) of the key enzyme (isocitrate lyase) of glyoxylate shunt and a complete gene set of the tricarboxylic acid (TCA) cycle in MO3_YZ.1 are observed, implying that the acetyl-CoA produced in the serine cycle can be subsequently oxidized to glyoxylate in assistance of some TCA cycle enzymes. This regeneration pathway of glyoxylate is common in type IIb but absent in type IIa methanotrophs, which use the ethylmalonyl-CoA (EMC) pathway to accomplish the same task [67]. Although many encoding genes of the EMC pathway-related enzymes are also detected in MO3_YZ.1, the encoding genes of four enzymes are absent (croR for 3-hydroxybutyryl-CoA dehydratase, ccr for crotonyl-CoA carboxylase/reductase, msd for 2-methylfumaryl-CoA hydratase and mcd for methenyltetrahydromethanopterin cyclohydrolase), indicating that MO3, like other Beijerinckiaceae methanotrophs, cannot regenerate glyoxylate through the EMC pathway. For MO3, the acetyl-CoA produced in the serine cycle can also be converted to poly-β-hydroxybutyrate (PHB, Figure 6). This carbon-storage polymer is also an endogenous source of reducing power [68], and may help MO3 adapt to environments with fluctuating substrate supplies [55,69].
As expected, the major carbon-assimilation pathway in type I methanotrophs, the RuMP pathway, is not retrieved in MO3_YZ.1 because the coding genes of the two key enzymes (hps for 3-hexulose-6-phosphate synthase and phi for 6-phospho-3-hexuloisomerase) of the RuMP pathway are absent in MO3_YZ.1. The coding gene of ribulose-bisphosphate carboxylase, the key enzyme of the Calvin-Benson-Bassham (CBB) cycle for CO 2 fixation, is also absent in MO3_YZ.1. Thus, in this respect, MO3 is similar to Methylocapsa gorgona MG08 [22] and different to several other type IIb strains including Methylocapsa acidiphila [19], Methylocapsa palsarum NE2 [51], Methylocella silvestris BL2 [70], and Methyloferula stellata AR4 [71] which have a complete CBB cycle. MO3_YZ.1 encodes the Embden-Meyerhof-Parnas and pentose phosphate pathways for carbohydrate metabolism. In addition, MO3_YZ.1 carries all the necessary genes for enzymes involved in acetate metabolism, such as acs for acetate-CoA synthetase, ackA for acetate kinase, and pta for phosphotransacetylase ( Figure 6). However, whether MO3 can grow using acetate as sole substrate like Methylocapsa aurea [72] is unknown because Methylocapsa gorgona MG08, which also carries these genes, cannot grow on acetate as the sole carbon source as expected [22]. An efficient membrane transporter for acetate (acetate permease ActP) may be necessary, but we currently know very little about this [52,73].

Nitrogen and Sulfur Metabolism of MO3
For nitrogen metabolism, MO3_YZ.1 possesses a complete nifHDKENX operon for molybdenum-containing nitrogenase like other type II methanotrophs [22,74] and genes for assimilatory nitrate reduction (nasAB, and nirA), dissimilatory nitrite reduction to ammonium (nirBD), ammonium transporter (amt), nitrate/nitrite transport protein (nrt), and putrescine transport-system protein (potFGHI) (Figure 6). The presence of these genes suggests that MO3 can utilize multiple types of nitrogen sources. As expected, genes encoding the denitrification pathway are missing in MO3_YZ.1 as many other aerobic methanotrophs [75]. For sulfur metabolism, MO3_YZ.1 possesses a series of genes in the sulfur-assimilation pathway ( Figure 6). These genes include cysUWA (encodes sulfate/thiosulfate transport system permease/ATP-binding proteins); cysNC, cysH, and cysJ (encodes enzymes catalyze the subsequent sulfate-reduction steps to sulfide); and genes for sulfur-containing amino-acid production from sulfide (such as cysE and cysK) (Table S2). In addition, some genes encoding sulfur-oxidation enzymes, such as sulfite dehydrogenase (sor), thiosulfate/3-mercaptopyruvate sulfurtransferase (sseA) and S-sulfosulfanyl-Lcysteine sulfohydrolase (sox), are present in MO3_YZ.1. However, studies and discussions on the sulfur metabolism of aerobic methanotrophs are relatively few [14,76]. Whether sulfur metabolism is related to the carbon metabolism, energy acquisition, and environmental adaptability of methanotrophs remains to be investigated.

Conclusions
We enriched the uncultured Beijerinckiaceae methanotroph MO3 from paddy soil by using the nitrogen-free M2 medium and reconstructed a nearly complete genome of this lineage. Based on phylogenomic analysis, the closest relative of MO3 was Methylocapsa. In terms of the carbon-assimilation pathway, MO3 also exhibited similar characteristics to Methylocapsa. Its 16S rRNA gene was most related to Methylosinus rather than Methylocapsa, probably due to the typical misassembly of 16S rRNA gene from metagenomic data. MO3 encoded diverse metabolisms related to nitrogen, sulfur, and PHB, implying its ability to survive in a variety of stress environments such as low nitrogen availability.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms10050955/s1, Figure S1: Neighbor-joining phylogenetic tree of the nifH gene from MO3_YZ.1. The tree based on 290 amino-acid positions is constructed using MEGA software (version 6.06) and evaluated with 1000 bootstraps. Bootstrap values higher than 50% are given at the branch nodes. Scale bar indicates 2% amino-acid sequence divergence; Figure S2: Matrix of pairwise genomic average nucleotide identity (gANI) and genomic average amino-acid identity (gAAI) values of MO3_YZ.1 and its relatives. The genomes' sequences were ordered as in Figure 4. The gANI was presented in the lower-left triangle and the gAAI was presented in the upper-right triangle. Bothe of gANI and gAAI in this figure were calculated by tools of Kostas lab. Figure S3. Alignments of amino-acid sequences of PmoA subunit from methanotrophs. The amino acids that form the tricopper cluster site are shown in blue. Table S1: Genome statistics of the obtained metagenome-assembled genomes (MAGs) affiliated to methanotrophs; Table S2: Gene features of Bin.033 predicted by prokka v1.14.5; Table S3: Gene features of Bin.033 annotated by RAST tool kit; Table S4: Gene functions of Bin.033 annotated by BlastKOALA through against the Kyoto Encyclopedia of Genes and Genomes database; Table S5: Length of pmoA-like genes in genomes of currently known aerobic methanotrophs.
Author Contributions: Conceptualization, Y.C. and Z.J.; methodology, Y.C. and J.Y.; data curation, Y.C. and J.Y.; writing-original draft preparation, Y.C.; writing-review and editing, Y.C., J.Y. and Z.J.; project administration, Y.C. and Z.J.; funding acquisition, Y.C. and Z.J. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.