Genomic Insights into Denitrifying Methane-Oxidizing Bacteria Gemmobacter fulva sp. Nov., Isolated from an Anabaena Culture

The genus Gemmobacter grows phototrophically, aerobically, or anaerobically, and utilizes methylated amine. Here, we present two high-quality complete genomes of the strains con4 and con5T isolated from a culture of Anabaena. The strains possess sMMO (soluble methane monooxygenase)-oxidizing alkanes to carbon dioxide. Functional genes for methane-oxidation (prmAC, mimBD, adh, gfa, fdh) were identified. The genome of strain con5T contains nirB, nirK, nirQ, norB, norC, and norG genes involved in dissimilatory nitrate reduction. The presence of nitrite reductase gene (nirK) and the nitric-oxide reductase gene (norB) indicates that it could potentially use nitrite as an electron acceptor in anoxic environments. Taxonomic investigations were also performed on two strains through polyphasic methods, proposing two isolates as a novel species of the genus Gemmobacter. The findings obtained through the whole genome analyses provide genome-based evidence of complete oxidation of methane to carbon dioxide. This study provides a genetic blueprint of Gemmobacter fulva con5T and its biochemical characteristics, which help us to understand the evolutionary biology of the genus Gemmobacter.

Comparative genomics analyses revealed that members of the genus Gemmobacter, including Gemmobacter aquatilis, Gemmobacter lutimaris, Gemmobacter sp. HYN0069, Gemmobacter caeni, and Gemmobacter sp. LW-1, utilize methylated amine. These species have related genes encoding the enzymes trimethylamine (TMA) dehydrogenase, TMA monooxygenase, and TMA demethylase in their genome, indicating metabolic potential of using the TMA oxidation pathway to convert trimethylamine to dimethylamine [3,8]. In a recent study, we isolated two alphaproteobacterial strains, con4 and con5 T , from an Anabaena culture that contained all genes related to the oxidization of methane to carbon dioxide.
Here, we report a comparative analysis of the genomes of two strains, con4 and con5 T , together with a taxonomic proposal based on their phylogenetic, genomic, physiological, and chemotaxonomic characteristics.

Isolation and Culture Conditions of the Strains
Strains con5 T and con4 were recovered from a culture of Anabaena using a serial dilution method [16]. For the Anabaena culture, Anabaena variabilis FBCC010004 strain was obtained from the Freshwater Bioresources Culture Collection (FBCC, https://fbcc.nibr.re. kr/fbcc, accessed on 25 August 2021) of the Nakdonggang National Institute of Biological Resources (NNIBR, South Korea). Anabaena cells were cultured in 500 mL standard cell culture flasks with Blue-Green (BG11) broth (Merck, St. Louis, MO, USA) under the conditions: 20 • C, 35% humidity, 12 h:12 h light-dark photoperiod, 20 µmol m −2 s −1 irradiance, and 200 rpm agitation [17]. From the Anabaena culture, a 100 µL sub-sample of the suspended material was aseptically spread onto modified R2A agar (Difco, NJ, USA) [18] and incubated at 25 • C under heterotrophic conditions. Two yellow-pigmented strains, con5 T and con4, were isolated after three days and routinely sub-cultivated on R2A agar at 30 • C for 48 h and kept in a glycerol solution (20%, v/v) at −70 • C for long-term preservation. For most experiments, all strains were cultivated on R2A agar at 30 • C.
The whole-cell fatty acid profile was analyzed by gas chromatography (Hewlett Packard 6890, Kyoto, Japan) with the TSBA 6 database using Sherlock software v6.1 at KCTC (Korean Collection for Type Cultures) service center. The cells of the strains con5 T and con4 and four Gemmobacter-type strains were collected after being grown at 30 • C for 48 h on the same sectors of R2A agar plates. Cell-harvesting standardization was done following the method described by Jin et al. [18]. The extraction of respiratory quinones and polar lipids was completed following the method described previously [21,22]. The respiratory quinones were identified using HPLC (Shimadzu, Kyoto, Japan) with an YMC-Pack ODS-A column, and the polar lipids were identified with two-dimensional TLC plates (60 F 254 , Merck).

Phylogenetic and Genomic Analyses
For a 16S rRNA gene analysis, genomic DNA of strains con5 T and con4 was extracted using a FastDNA TM SPIN kit. Amplification of the 16S rRNA genes was done by PCR with 27F/1492R primer sets [23]. The 16S rRNA gene sequences of strains con5 T and con4 were compared with the available sequences in the EzBioCloud server [24]. Evolutionary analyses were performed with MEGA7 software [25]. Sequence edition and multiple alignment were done using the programs BIOEDIT [26] and CLUSTAL X [27], respectively. Phylogenetic trees were reconstructed using three algorithms: Neighbor-Joining (NJ), Maximum-Parsimony (MP), and Maximum-Likelihood (ML) [28][29][30]. The ranch robustness of the phylogenetic trees was estimated by bootstrap analyses based on 1000 re-samplings of the sequences [31]. A phylogenomic tree was re-constructed from the available type strains of species of the genus Gemmobacter with whole genome sequences on the TYGS (Type Strain Genome Server) [32]. The phylogenomic tree was inferred using FastME 2.1.4 [33] from GBDP (Genome Blast Distance Phylogeny) distances calculated from genome sequences. Branch lengths were calculated using the GBDP distance formula d5 [34]. The whole genomes of strains con5 T and con4 were sequenced using both the PacBio RSII platform and Illumina next-generation sequencing technology at Macrogen Inc. (Seoul, South Korea). The chromosomes and plasmids were assembled using the software package of SMRT portal (v.3.2.0), Pacific Biosciences, (CA, USA) [35]. The de novo genome annotation of strains con5 T and con4 was performed with the Prokka (v.1.4.6) pipeline [36], and a sequence-based comparison was made using the SEED Viewer [37]. The predicted coding sequences (CDSs) were submitted to the COG database to create the functional categories [38,39]. The values of average nucleotide identity (ANI) were calculated by using the OrthoANI tool in EZBioCloud server [40], and the digital DNA-DNA hybridization (dDDH) values were calculated using the genome-to-genome distance calculator (GGDC 2.1) based on draft genome sequences [34].

Phylogenetic and Genomic Analysis: The Taxonomic Status
The 16S rRNA sequences of strains con5 T and con4 share 100% identity between them and 94.9-97.9% identity with those of the closest species within the genus Gemmobacter ( Table 2). The 16S rRNA gene sequences of strains con5 T and con4 were compared with the 16S rRNA gene sequences of representative species within the genus Gemmobacter and related genera in the EzTaxon-e server. The two strains share over 97.0% similarity with G. aquaticus A1-9 T (97.9%), G. caeruleus N8 T (97.7%), G. lutimaris YJ-T1-11 T (97.4%), and G. tilapiae KCTC 23310 T (97.3%) and less than 97% with the remaining species within the genus Gemmobacter. Strains con5 T and con4 also share high similarities with other species than members of Gemmobacter: 96.7% with Cypionkella collinsensis 4-T-34 T , 96.7% with Cypionkella psychrotolerans PAMC 27389 T , and 96.4% with Cypionkella aquatica DC2N1-10 T . However, it was clear from the topology of the phylogenetic tree ( Figure 1) that strains con5 T and con4 clustered clearly with the species of Gemmobacter. In addition, the phylogenomic tree reconstructed on the TYGS provided clearer evidence for the taxonomic position of the two strains within the genus Gemmobacter (Supplementary Figure S3). The genomic DNA G + C content of the two strains was 64.1 mol%, which is in the range reported previously for the genus Gemmobacter (61.4-69.4 mol%) [3,4] Figure S4), respectively, which were much lower than the species boundaries of ANI or dDDH of 95-96% and 70%, respectively, and fall in the intergeneric range [41][42][43].

Genome Properties
The genome of strain con5 T was 4.7 Mb and contained a circular chromosome of 3.4 Mb and six plasmids sized 34.0-425.5 kb (CP076361-CP076367) (Supplementary Tables S1 and S2). Of 4534 genes, 4472 were protein-coding genes and 62 were RNA genes (nine rRNA genes, 52 tRNA genes, and one tmRNA gene). For strain con4 (JAH-HWR000000000), of 4417 genes, 4317 were protein-coding genes and 53 were RNA genes (four rRNA genes, 46 tRNA genes, one tmRNA gene, and two ncRNA genes) (Supplementary Figure S5). The DNA G + C content of both strains was 64.1 mol% (Supplementary Tables S1 and S2).

Genome Analyses for Denitrification and Methane Oxidation
It is generally understood that methanotrophic bacteria are mostly active at the oxicanoxic transition zone in stratified lakes, using oxygen to oxidize methane. The methanotrophs produce methane monooxygenase to utilize methane as a carbon source [44,45]. The methanotrophs express two kinds of methane monooxygenase, soluble methane monooxygenase (sMMO) and particulate methane monooxygenase (pMMO), and these enzymes can also oxidize various alkanes [46][47][48].
Strain con5 T possesses sMMO (prmAC, mimBD) but does not have a pMMO gene in its genome sequence. The presence of one PQQ-dependent alcohol dehydrogenase (ADH) (PQQ-MDH: adh, adh1, and adhB), two NAD + -dependent ADH (adh, adh1), and another alcohol dehydrogenase (adhB) in strain con5 T makes it a good candidate for the conversion of methane to aldehyde (Figure 2). Genes involved in the glutathione (GSH)-dependent pathway to metabolize formaldehyde are identified. Formaldehyde (HCHO)-activating enzyme (gfa), a GSH-dependent formaldehyde dehydrogenase (fdh), and S-formyl-GSH hydrolase (fgh) convert formaldehyde to formate, and then the formate is oxidized by formate dehydrogenase (fdh) to the final product, carbon dioxide (CO 2 ) (Figures 2 and 3). Those genes were also found in the strain con4 ( Figure 4). of strain con5 T contains nirB, nirK, nirQ, norB, norC, and norG genes involved in dissimilatory nitrate reduction, but no dissimilatory nitrate reductase (narG) or nitrous oxide reductase (nosZ) genes are detected, which is incomplete for the denitrification pathway (Figures 2 and 3). This suggests that strain con5 T cannot perform nitrate reduction, itself, unless it utilizes unknown or incorrectly classified reduction pathways up to date. The genome of strain con5 T contains the nitrite reductase gene, nirK, and the nitric-oxide reductase gene norB, and it could potentially use nitrite as an electron acceptor in anoxic environments.  Some methanotrophs comprise genes encoding enzymes for the nitrate reduction pathway, which was confirmed to be related to methane oxidation under anoxic conditions [49], and these methanotrophic species encode a complete nitrate reduction pathway to use nitrate as a terminal electron acceptor when oxygen is depleted [50]. The genome of strain con5 T contains nirB, nirK, nirQ, norB, norC, and norG genes involved in dissimilatory nitrate reduction, but no dissimilatory nitrate reductase (narG) or nitrous oxide reductase (nosZ) genes are detected, which is incomplete for the denitrification pathway (Figures 2 and 3). This suggests that strain con5 T cannot perform nitrate reduction, itself, unless it utilizes unknown or incorrectly classified reduction pathways up to date. The genome of strain con5 T contains the nitrite reductase gene, nirK, and the nitric-oxide reductase gene norB, and it could potentially use nitrite as an electron acceptor in anoxic environments.

Conclusions
In this study, the strains con5 T and con4, representing methane oxidizing species from an Anabaena culture belonging to the genus Gemmobacter, were investigated using genomic and polyphasic methods. The findings obtained through the whole genome analyses provide genome-based evidence of complete oxidation of methane to carbon dioxide. This study provides a genetic blueprint of Gemmobacter fulva con5 T and its biochemical characteristics, which help us to understand the evolutionary biology of the genus Gemmobacter. Based on the phylogenetic position and the genotypic, chemotaxonomic, and physiological differences, we propose that strains con5 T and con4, Gemmobacter fulva sp. nov., should be assigned as a novel species of the genus Gemmobacter in the family Rhodobacteraceae (Table 3).

Conclusions
In this study, the strains con5 T and con4, representing methane oxidizing species from an Anabaena culture belonging to the genus Gemmobacter, were investigated using genomic and polyphasic methods. The findings obtained through the whole genome analyses provide genome-based evidence of complete oxidation of methane to carbon dioxide. This study provides a genetic blueprint of Gemmobacter fulva con5 T and its biochemical characteristics, which help us to understand the evolutionary biology of the genus Gemmobacter. Based on the phylogenetic position and the genotypic, chemotaxonomic, and physiological differences, we propose that strains con5 T and con4, Gemmobacter fulva sp. nov., should be assigned as a novel species of the genus Gemmobacter in the family Rhodobacteraceae (Table 3).

Nucleotide Sequence Accession Numbers
The GenBank/EMBL/DDBJ accession numbers for the 16S rRNA gene sequences of strains con5 T and con4 are MZ317888 and MZ317889, respectively. Accession numbers for the genome sequences of strains con5 T and con4 are CP076361-CP076367 and JAHHWR000000000, respectively.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/microorganisms9122423/s1, Figure S1. Transmission electron micrograph of strains con5 T (A) and con4 (B) grown on R2A for 48 h at 30 • C. Bar,1 µm. Figure S2. Polar lipid profile for strains con5 T (A) and con4 (B). All the polar lipids were stained with 5% molybdatophosphoric acid (for total lipids), molybdenum blue (for phospholipids), ninhydrin (for amino lipids), and anisaldehyde/sulfuric acid (glycolipids). AL, unidentified aminolipid; APL, unidentified aminophospholipid; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PC, phosphatidylcholine; GL, unidentified glycolipid; PL, unidentified phospholipid. 1, first dimension of TLC; 2, second dimension of TLC. Figure S3. Phylogenomic tree based on genome sequences of the strains con5 T and con4 in the TYGS (https://tygs.dsmz.de/, accessed on 25 August 2021). The branch lengths are calculated in terms of GBDP distance formula d5. The numbers above branches are GBDP pseudo-bootstrap support values > 60% from 100 replications, with an average branch support of 95.1%. The tree was rooted at the midpoint [34]. Figure S4. Cluster analysis of the profiles obtained from ANI (A) and DDH values (B). Dendrograms were calculated on the basis of the average nucleotide identity scores based on the whole genomes using the unweighted pair group method with the arithmetic averages clustering algorithm (UPGMA). Figure S5. Graphic representation of circular genome plot of strain con5 T . The circles from inside to outside represent GC skew ([G − C]/[G + C]) plot of the genome, GC content, CDS-reverse, CDS-forward, contigs, and position label. Table S1. General genome features of strains con5 T and con4. Table S2. Genomic feature of type strain con5 T .