Methylobrevis albus sp. nov., isolated from freshwater lake sediment

An aerobic, Gram-stain-negative, rod-shaped bacterium with flagellum, designated L22T, was isolated from sediment of Hulun Lake, Inner Mongolia, China. The organism was found to grow optimally at 30° C in a medium containing 0–0.75% (w/v) NaCl at pH 7.5. The major fatty acid identified was summed feature 8 (C18:1ω7c). The dominant polar lipids were phosphomonoester, phosphatidylethanolamine, phosphatidylglycerol and phosphatidylcholine. The main respiratory quinone was Q-10. The draft genome sequence of strain L22T consisted of 4354,788 bp. The G + C content of genomic DNA was 69.8 mol %. The 16S rRNA gene sequences indicated that strain L22T was affiliated with the genus Methylobrevis within the family Pleomorphomonadaceae, being most closely related to Methylobrevis pamukkalensis JCM 30229T with 95.9% 16S rRNA gene sequences similarity. The AAI, ANI and dDDH values between strain L22T and M. pamukkalensis JCM 30229T were 72.5%, 80.7% and 22.7%. Based on taxonomic results in this study, we proposed that strain L22T a novel species in the genus Methylobrevis of the family Pleomorphomonadaceae, for which the name Methylobrevis albus sp. nov. is proposed. The type strain is L22T (=KCTC 72858T=MCCC 1H00432T).


Introduction
The genus Methylobrevis was first described by Poroshina et al. (2015) as a member of the order Rhizobiales within the class Alphaproteobacteria. Up to now only one described species, M. pamukkalensis (Poroshina et al. 2015), has been confirmed to belong to the genus Methylobrevis within the family Pleomorphomonadaceae. The family Pleomorphomonadaceae was recently revised by transferring the genus Chthonobacter (Kim et al. 2017), Hartmannibacter (Suarez et al. 2014), Methylobrevis (Poroshina et al. 2015), Mongoliimonas (Xi et al. 2017), Oharaeibacter (Lv et al. 2017) and Pleomorphomonas (Xie and Yokota 2005) into the family (Hördt et al. 2020). In October 2019, we isolated the strain L22 T from sediment of Hulun Lake, Inner Mongolia, China. Based on the phenotypic, genotypic and chemotaxonomic characteristics, we aimed to propose a novel species of the genus Methylobrevis within the newly identified family Pleomorphomonadaceae, represented by strain L22 T .
The GenBank accession number for the 16S rRNA gene sequence of Methylobrevis albus L22T is MW195049 and the draft genome has been deposited in GenBank under the accession number JADZLT000000000.

Inoculum source, isolation and preservation
Sample of fresh water and sediment was collected from the middle part of Hulun Lake (117°01′ 10" E, 48°30′ 40" N), Inner Mongolia, the fourth biggest freshwater lake in China in July 2019. The sediment sample was diluted serially to 10 -3 with sterile water. 0.1 ml aliquots of each dilution were spread onto 1/2-strength R 2 A agar. The R 2 A agar was made with 0.05% tryptone, 0.05% yeast, 0.05% casein, 0.05% starch, 0.05% glucose, 0.03% sodium pyruvate, 0.03% K 2 HPO 4 and 0.005% MgSO 4 •7H 2 O, all w/v. The pH was adjusted to pH 7.2. After incubating in an aerobic environment at 28° C for 15 days, strain L22 T was picked among several tiny white colonies and purified on R 2 A agar by routinely repetitious streaking. The strain was stored in sterile 15% (v/v) glycerol supplemented with 1% (w/v) NaCl suspensions at −80° C. Strain L22 T was cultivated on R 2 A agar at 30° C unless otherwise mentioned. For comparison and analysis, we used Methylobrevis pamukkalensis JCM 30229 T as a related strain, which was cultured under the same conditions as strain L22 T .

16S rRNA gene sequence analysis
Universal bacterial primers 27F and 1492R as previously described (Liu et al. 2014) were used for 16S rRNA gene sequence PCR amplification. All PCR production was purified and ligated into the PMD-18 vector (Takara). Cloned Escherichia coli DH5α cells (Trans-Gen Biotech) were then transformed with the obtained recombinant plasmids. The positive clones were selected by LB agar contained 0.01% (w/v) ampicillin and sequenced at BGI Co. Ltd (Qingdao, China) using the ABI 3730XL system. The almost complete 16S rRNA gene sequence of strain L22 T was compared with those available from the EzTaxon server (http:// www. ezbio cloud. net/ ident ify) (Kim et al. 2012) and NCBI database (hyyp://www. blast. ncbi.nlm.nih.gov/Blast.cgi) for further phylogenetic analysis. Sequences of the related type strains and the almost completed 16S rRNA gene sequence of strain L22 T were aligned using MUSCLE (Robert and Edgar 2004). A phylogenetic tree was reconstructed utilizing the maximum-likelihood (ML) algorithm (Felsenstein 1981) implemented in the software package MEGA 7.0 (www. megas oftwa re. net) (Sudhir et al. 2016). The overall mean distance of aligned 16S rRNA sequences was calculated and showed suitable compatibility, thus a neighbour-joining (NJ) (Saitou and Nei 1987) tree was reconstructed to verify the confidence of the ML tree. For all tree-making calculations, variance estimates were taken on for 1000 replication by the method of bootstrap analysis to ensure accurate calculation results.

Genome sequence analysis
The genomic DNA of strain L22 T was extracted and purified using a bacterial genomic DNA kit (Takara). The draft genomic sequence of strain L22 T was sent to Shanghai OE Biotech Co., Ltd (Shanghai, China) and sequenced using Illumina technology. An Illumina shotgun library using the Illumina TruSeq Nano DNA Sample Prep Kit was reconstructed and sequenced using the pair-end 350-bp protocol on the Illumina HiSeq Xten platform (Illumina Inc., San Diego, USA). Raw sequencing data were generated by Illumina base-calling software CASAVA v1.8.2 (http:// suppo rt. illum ina. com/) according to its corresponding manuscript. The sequenced reads were assembled using SOAP de novo software (Li et al. 2009). Gene content was annotated using the NCBI Prokaryotic Genome Annotation Pipeline and the genes involved in metabolic pathways were analysed in detail by using information from the KEGG database (Kanehisa et al. 2016). Protein-encoding regions were identified with the Rapid Annotations using Subsystem Technology (RAST) server (Aziz et al. 2008). A further phylogenetic and taxonomic multilocus sequence analysis was made to identify the accurate taxonomic status of strain L22 T . Genomes of the order Hyphomicrobiales which were available from the NCBI were aligned with MUS-CLE v.3.8.31. The phylogenomic relationship was analysed by Genome Taxonomy Database (GTDB; https:// gtdb. ecoge nomic. org), and the phylogenetic trees were constructed by using FastTree (Luke 2002) using GTR+CAT parameters and IQTree (Trifinopoulos et al. 2016) using GTR+F+I+G4 model and 1000 bootstrap replicates. The genome of Ochrobactrum lupini ATCC 49188 T served as the outgroup to compute the phylogeny for the DNA and protein sequences. The tree was collapsed and formatted using iTOL v4 (Ivica et al. 2019).

Morphological, physiological and biochemical analysis
Morphological features of strain L22 T were tested using the biomass incubated on R 2 A agar at 30 °C for 3 days. Light microscopy (E600; Nikon) and transmission electron microscopy (JEM1200, Japan) were used to observe the microstructures of cells of strain L22 T . The motility was tested using the hanging-drop method and gliding motility was confirmed as described by Bowman (Bowman 2000). Gram-staining was performed using a Gram stain kit (bio-Mérieux) according to the manufacturer's instructions. Different gradients of 0, 4, 10, 15, 20, 25, 28, 30, 33, 37 and 40 °C were set up to find the temperature range for growth on R 2 A agar. The tolerances to various NaCl concentrations were examined by incubating strain L22 T on R 2 A agar where 0.00-10.00% (w/v) NaCl at intervals of 0.25% (w/v) was added. The pH range for growth was determined by a series of adjusted R 2 A broth adding a series of 20 mM buffer: MES (pH 5.5 and 6.0), PIPES (pH 6.5 and 7.0), HEPES (pH 7.5 and 8.0), Tricine (pH 8.5), CAPSO (pH 9.0, 9.5 and 10.0) and CAPS (pH 10.5 and 11.0). Growth at an anaerobic (10% H 2 , 10% CO 2 and 80% N 2 ) environment was confirmed after 14 days of incubation on R 2 A agar with or without 0.1% (w/v) KNO 3 in an anaerobic bag. All experiments were performed with three replicates. The reduction of nitrate was tested following a previous statement of Cowan and Steel (Cowan and Steel 1974). Oxidase activity was tested using an oxidase reagent kit (bioMérieux) according to the manufacturer's instructions. Catalase activity was determined by the application of 3% (v/v) hydrogen peroxide solution and observing bubble production. Tests for the hydrolysis of alginate, carboxymethylcellulose, starch, casein, Tweens 20, 40, 60 and 80 were performed as previously mentioned (Cowan et al. 1966). DNase activity was investigated by DNase agar (HopeBio) according to the manufacturer's instruction. Sensibility to antibiotics was investigated on R 2 A agar using the disc diffusion method following the procedures outlined by the Clinical and Laboratory Standards Institute (CLSI 2018). Methanol utilization was tested in 500 ml flasks containing 100 ml of sterile marine ammonium mineral salt (MAMS) medium supplemented with 2% (v/v) methanol and 0.4 ml of a vitamin mixture (thiamin, biotin, folic acid and B 12 , 50 μg l -1 of each). The MAMS medium (Schäfer, 2007) contained (w/v): NaCl, 2%; K 2 HPO 4 , 0.234%; (NH 4 ) 2 SO 4 , 0.1%; MgSO 4 ⋅7H 2 O, 0.1%; KH 2 PO 4 , 0.036%; CaCl 2 , 0.015%; Na 2 MoO 4 ·2H 2 O, 0.002%; Na 2 WO 4 ·2H 2 O, 0.0003%; FeSO 4 ·7H 2 O, 0.0002%. The pH was adjusted to pH 7.0. Same flasks with no methanol were set as a blank control group. All flasks were incubated for one week at 30 °C with shaking (100 rpm). Strain L22 T and strain M. pamukkalensis JCM 30229 T were treated in the same way. Activities of other enzymes were evaluated in virtue of API ZYM (bioMérieux) kits according to the manufacturer's instruction. Oxidation of various carbons as the sole carbon source was assessed in Biolog GEN III microplates (http:// www. biolog. com/) according to the manufacturer's instruction. Acid production of different compounds as sole sources of carbon and energy was determined using API 50CHB (bioMerieux) kits according to the manufacturer's instruction. Other biochemical tests were performed with the API 20E (bioMerieux) kits according to the manufacturer's instruction. Each one of those API and Biolog tests was carried out using the biomass of strain L22 T grown on R 2 A agar at 30° C for 3 days and repeated for two times simultaneously with the related type strain.

Chemotaxonomic characterization
Late-exponential-growth-phase biomass of strain L22 T and its relative grew on R 2 A agar at 30° C were collected in centrifuge tubes and impermanently preserved at −20° C ready for imminent investigations. The cellular fatty acid methyl esters were prepared and extracted according to the standard protocol of MIDI (Sherlock Microbial Identification System, version 6.1) (Athalye et al. 2010). The subsequent identification was performed by an Agilent 6890 N gas chromatography basing on the TSBA40 database of the microbial identification system. Polar lipids were obtained by a series of centrifugation and extraction and separated via TLC (thin layer chromatography) as described previously (Minnikin, 1984). The respiratory quinone was extracted from 300 mg of freeze-drying biomass and analysed using reverse-phase HPLC (high-performance liquid chromatography), according to the method used previously (Minnikin et al. 1984).

16S rRNA gene sequence and phylogenetic analysis
The almost complete 16S rRNA (1457 bp) gene sequence of strain L22 T was obtained through conventional Sanger sequencing. The result of pairwise comparison of 16S rRNA gene sequences in the GenBank database of NCBI server (https:// www. ncbi. nlm. nih. gov) showed that the most related strain was Methylobrevis pamukkalensis JCM 30229 T with 95.9% 16S rRNA gene sequence similarity, followed by Chthonobacter albigriseus JCM 30603 T (95.0%), Oharaeibacter diazotrophicus SM30 T (94.5%), Nitratireductor lucknowense IITR-21 T (93.8%) and Nitratireductor indicus C115 T (93.8%). The 16S rRNA gene sequence of strain L22 T showed a maximum identity of 95.9% with validly published species, which is below the 98.7% species threshold identity but higher than the 94.5% genus threshold identity. Following the 16S rRNA-based taxonomic identity, the species rank thus seems appropriate for this strain. Besides, the phylogenetic maximum-likelihood tree based on 16S rRNA gene sequences (Fig. 1) showed that strain L22 T belonged to the genus Methylobrevis. The topology of the phylogenetic tree constructed by neighbour-joining arithmetic (Fig. S1) proved the conclusion that strain L22 T and M. pamukkalensis JCM 30229 T formed a stable cluster that had a distinct phylogenetic lineage with other genera within the family Pleomorphomonadaceae. These overall data indicated that strain L22 T was regarded as a novel species within the genus Methylobrevis.

Genome sequence analysis
The draft genome sequence of strain L22 T consisted of 4,354,788 bp. The combined assembly yielded 60 contigs with the largest being 618,762 bp and the contig N50 value was 249,401 bp. An average 473×coverage depth was accomplished. The DNA G+C content of strain L22 T was 69.8 mol%, which was following the reported value for strain M. pamukkalensis JCM 30229 T (67.9 mol%) (Poroshina et al. 2015). Gene prediction and annotation identified 3927 genes, 3866 protein-encoding genes, 51 tRNA genes, 6 rRNA genes. Two 16S rRNA gene sequences (1167 bp and 1473 bp) was extracted from the draft genome and were compared with the almost complete 16S rRNA (1457 bp) gene sequence of strain L22 T obtained through conventional Sanger sequencing to ensure the authenticity. Moreover, according to annotation from KEGG database, genes assigned to functional categories are involved in the environmental information processing (318), carbohydrate metabolism (250), protein families: genetic information processing (227), protein families: signalling and cellular processing (222), genetic information processing (164) and amino acid metabolism (147). Gene katE (encoding catalase) and katE-intracellular protease were found which meant strain L22 T got the ability to decompose hydrogen peroxide. A complete phosphatidate cytidylyltransferase [EC 2.7.7.41] pathway and gene pmtA were proved to exist and thus phosphatidylethanolamine (PE) can be produced and transformed into phosphatidylcholine (PC), in accord with the result of polar lipids experiment. Results on KEGG told that strain L22 T can reduce sulfate to APS, PAPS, sulfite and finally sulfide step by step. However, the experiment showed that sodium thiosulfate could not be reduced to H 2 S. Though the existence of gene xoxF, mdh1 and mxaF indicated that methanol could be oxidized into formaldehyde, strain L22 T was not capable of methylotrophic growth in the absence of formaldehyde dehydrogenase and formate dehydrogenase. Phylogenetic analysis of the genomic amino acid sequences performed with the IQTree and FastTree approaches demonstrates that strain L22 T formed a stable lineage within the genus Methylobrevis (Figs. 2, S2).
The AAI value between strain L22 T and M. pamukkalensis JCM 30229 T was 72.5%, significantly below the proposed cut-off for a species boundary of 85-90% and exceed the threshold value for a genus boundary 55-60% (Rodriguez-R and Konstantinidis 2014). The results of ANI and dDDH between strain L22 T and M. pamukkalensis JCM 30229 T were 80.7% and 22.7%, both below the thresholds (ANI: 95-96%; dDDH: 70%) for new species identification (Goris et al. 2007) (Meier-Kolthoff et al. 2013). Details of AAI, ANI and dDDH values between  Table 1. These results supported the finding that strain L22 T was a novel member of the genus Methylobrevis.

Chemotaxonomic characterization
The major fatty acid (> 10.0%) of strain L22 T was summed feature 8 (C 18:1 ω7c) (69.1%) same as the major fatty acid extracted from its related strain M. pamukkalensis JCM 30229 T . The polar lipids of strain L22 T comprised phosphomonoester (PME), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylcholine (PC), one unidentified glycolipid (GL) and seven unidentified lipids (L). What deserved to be mentioned was that strain L22 T and strain M. pamukkalensis JCM 30229 T both possessed phosphomonoester (PME), phosphatidylethanolamine (PE), phosphatidylglycerol (PG) and phosphatidylcholine (PC) as the major polar lipids (Fig. S3). Strain L22 T contained unidentified lipids 1 to 9, while the strain M. pamukkalensis JCM 30229 T did not have unidentified lipid 8 and unidentified lipid 9 but had unidentified lipid 10 to 13. Details of cellular fatty acids and polar lipids compositions are listed in Table 3. The major respiratory quinone of strain L22 T was Q-10, which is in accordance to the description of the family Pleomorphomonadaceae (Hördt et al. 2020).
According to the phylogenetic analyses, differential chemotaxonomic data and other phenotypic properties, strain L22 T was classified as a new member of the genus Methylobrevis.
Funding This work was supported by the National Natural Science Foundation of China (Grant No. 32070002 and Grant No. 31770002) and the National Science and Technology Fundamental Resources Investigation Program of China (Grant No. 2019FY100700). Table 3 Fatty acids compositions of strain L22 T and related type strain Strains: 1, L22 T ; 2, M. pamukkalensis JCM 30229 T Data are from this study. Tr, traces (< 1.0%). Fatty acids amounting (< 1.0%) are not shown, dominant fatty acids (≥ 10.0%) are highlighted in bold Summed feature 2 contained C 12:0 and/or aldehyde Summed feature 3 contained C 16:1 ω7c and/or C 16:1 ω6c; Summed feature 8 contained C 18:1 ω7c * Summed features are groups of two or three fatty acids that are treated together for evaluation in the MIDI system and include both peaks with discrete equivalent chain-lengths (ECLs) as well as those where the ECLs are not reported separately