Devosia litorisediminis sp. nov., isolated from a sand dune

A Gram-negative, aerobic, non-motile, and rod-shaped bacterial strain, designated BSSL-BM10T, was isolated from sand of a dune that was collected from the Yellow Sea, Republic of Korea. It was subjected to a polyphasic taxonomic study. 16S rRNA gene sequence analysis showed that strain BSSL-BM10T fell phylogenetically within the radiation comprising type strains of Devosia species. The 16S rRNA gene sequence of strain BSSL-BM10T shared sequence similarities of 98.2% with the type strain of D. naphthalenivorans and 93.5–97.7% with type strains of other Devosia species. ANI and dDDH values between strain BSSL-BM10T and type strains of 18 Devosia species were 71.0–78.4% and 18.8–21.5%, respectively. The DNA G + C content of strain BSSL-BM10T was 60.9% based on its genomic sequence data. Strain BSSL-BM10T contained Q-10 as the predominant ubiquinone and 11-methyl C18:1ω7c, C18:1ω7c, summed feature 3 (C16:1ω7c and/or C16:1ω6c), and C16:0 as its major fatty acids. Major polar lipids of strain BSSL-BM10T were phosphatidylglycerol and two unidentified glycolipids. Strain BSSL-BM10T showed distinguishable phenotypic properties with its phylogenetic and genetic distinctiveness separated from recognized Devosia species. Based on data presented in this study, strain BSSL-BM10T should be placed in the genus Devosia. The name Devosia litorisediminis sp. nov. is proposed for strain BSSL-BM10T (= KACC 21633T = NBRC 115152T).


Introduction
The genus Devosia, a member of the family Devosiaceae (Hördt et al. 2020) of the class Alphaproteobacteria, was proposed by Nakagawa et al. (1996) with the transfer of "Pseudomonas riboflavina" to Devosia riboflavina (type species). The genus Devosia currently comprises 28 species with validly published names (https:// lpsn. dsmz. de/ genus/ devos ia; Parte 2018). Members of the genus Devosia have been isolated from various habitats (Bautista et al. 2010;Galatis et al. 2013;Jia et al. 2014;Kumar et al. 2008;Lin et al. 2020;Park et al. 2016;Quan et al. 2020;Romanenko et al. 2013;Yoon et al. 2007;Zhang et al. 2012). Recently, in the course of screening novel bacteria from a sand dune close to the Yellow Sea of Republic of Korea, many bacterial isolates have been obtained followed by taxonomic characterization. One of these bacterial isolates, designated as BSSL-BM10 T , showed the closest affiliation to members of the genus Devosia from the result of 16S rRNA gene sequence comparison. In this study, strain BSSL-BM10 T is characterized further using a polyphasic approach. diluted with 0.85% (w/v) saline solution and spread onto marine agar 2216 (MA; BD Difco). After incubation at 25 °C for 7 days, strain BSSL-BM10 T was isolated from the MA plate and speak onto trypticase soy agar (TSA; BD Bacto) at 30 °C. Cells of strain BSSL-BM10 T were suspended in a sterile solution containing 20% (w/v) glycerol and stored at -80 °C for long-term preservation. Devosia naphthalenivorans JCM 32509 T and Devosia riboflavina DSM 7230 T , the type strain of the type species, were obtained from the Japan Collection for Microorganisms (JCM; Japan) and Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ; Germany), respectively. Cells of strain BSSL-BM10 T and D. riboflavina DSM 7230 T obtained from culture grown for 3 days in trypticase soy broth (BD Bacto) at 30 °C were used to extract DNA and to analyze isoprenoid quinones and polar lipids. Cell masses for cellular fatty acid analysis were obtained under the following conditions: strain BSSL-BM10 T were harvested from TSA plates after cultivation for 3, 5, and 7 days at 30 °C, and D. naphthalenivorans JCM 32509 T and D. riboflavina DSM 7230 T were harvested from MA and TSA plates, respectively, after cultivation for 5 days at 30 °C.

Sequencing and phylogenetic analysis of 16S rRNA gene
Chromosomal DNA extraction was performed using a Wizard Genomic DNA isolation kit (Promega) according to the manufacturer's instruction. The 16S rRNA gene amplification was performed as described previously (Yoon et al. 1997) using PCR in which 9F (5′-GAG TTT GAT CCT GGC TCA G-3′) and 1512R (5′-ACG GTT ACC TTG TTA CGA CTT-3′) were used. Sequencing of the 16S rRNA gene followed by phylogenetic analysis were carried out as described by Yoon et al. (2003). Similarity between 16S rRNA gene sequences was calculated from alignment obtained using Clustal W program.

Genomic analysis
A TruSeq DNA LT Sample Prep kit (Illumina) was used to prepare a library for genomic sequencing. The library was sequenced using Illumina MiSeq platform. Sequencing data were assembled with SPAdes (Bankevich et al. 2012). Contamination of genome sequence was assessed using ContEst16S (Lee et al. 2017). Library construction and sequencing were performed by Chunlab Inc. (Republic of Korea). The ANI value based on BLAST + was calculated using JSpecies WS (http:// jspec ies. riboh ost. com/ jspec iesws/; Richter et al. 2015) or OrthoANI  in EZBioCloud. The dDDH value was estimated using TYGS (https:// tygs. dsmz. de/ user_ reque sts/ new) with BLAST + in which the recommended formula 2 (Meier-Kolthoff et al. 2013) was used. Phylogenetic tree was constructed based on genomic sequences using previous methods (Lefort et al. 2015;Meier-Kolthoff et al. 2013) described in the TYGS. Intergenomic distances inferred under the algorithm 'trimming' and distance formula d 5 (Meier-Kolthoff et al. 2013) and 100 distance replicates were calculated each. The resulting distances were used to infer a balanced minimum evolution tree with branch support via FASTME 2.1.6.1 including SPR postprocessing (Lefort et al. 2015).

Chemotaxonomic characterization
Extraction and analysis of isoprenoid quinones were performed as described by Komagata and Suzuki (1987) and Park et al. (2014), respectively. Fatty acid analysis was performed as described by Park et al. (2014) using the standard MIDI protocol (Sherlock Microbial Identification System, version 6.2B), GC (Hewlett Packard 6890), and TSBA6 database of the Microbial Identification System (Sasser 1990). Extraction of polar lipids were carried out according to procedures described by Minnikin et al. (1984). They were separated by two-dimensional TLC using the solvent systems as described by Embley and Wait (1994). The TLC plates were sprayed with various reagents as described by Park et al. (2014) and individual polar lipids were visualized followed by identified with heating at 150 °C for 3 min.

Morphological, cultural, physiological and biochemical characterization
Cell morphology, Gram reaction, anaerobic growth, pH range for growth, growth at various concentrations of NaCl, hydrolysis of gelatin and urea and susceptibility to antibiotics were investigated as described by Park et al. (2014). For transmission electron microscopy (JEM1010; JEOL), cells were negatively stained with 1% (w/v) phosphotungstic acid and air-dried. Grids were then examined. Growths at 4,10,20,25,28,30,35,37, and 40 °C on MA were measured to estimate the optimal temperature and temperature range for its growth. Nitrate reduction and hydrolysis of aesculin or Tween 80 were investigated as described previously (Lányí 1987) using artificial seawater (Bruns et al. 2001) for the preparation of the media. Hydrolysis of other substrates was tested as described by Barrow and Feltham (1993) with the modification that MA was used. Activities of catalase and oxidase were determined as described by Lányí (1987). Utilization of various substrates (each 0.2%) for growth was investigated as described by Kämpfer et al. (1991). Other biochemical and physiological properties were determined using API ZYM and API 20NE systems (bioMérieux; France). Enzyme activities by the API ZYM system were determined after incubation at 30 °C for 8 h. Other physiological and biochemical properties by the API 20NE system were determined after incubation at 30 °C for 2 days.

Phylogenetic analysis based on 16S rRNA gene sequence
The almost complete 16S rRNA gene sequence of strain BSSL-BM10 T had a continuous stretch of 1421 nucleotides, corresponding to positions 28-1491 (95%) of the Escherichia coli 16S rRNA sequence. Phylogenetic trees using three different algorithms (neighbor joining, maximum-likelihood and maximum-parsimony) showed that strain BSSL-BM10 T formed an independent lineage within the clade comprising type strains of Devosia species ( Fig. 1; Figs. S1 and S2). Strain BSSL-BM10 T shared the highest 16S rRNA gene sequence similarities (98.2%) with D. naphthalenivorans CM5-1 T . It shared 93.5-97.7% 16S rRNA gene sequence similarities with type strains of other Devosia species. These sequence similarities indicated that strain BSSL-BM10 T might be a species different from recognized Devosia species according to the threshold value (98.7%) recommended for delineation of a bacterial species by Kim et al. (2014).

Genomic features
The genome size of strain BSSL-BM10 T obtained from the assembly of sequencing reads was 3,743,297 bp with a sequencing depth of coverage of 414.33X. The genomic sequence of strain BSSL-BM10 T contained five contigs with N50 length of 2,671,820 bp. The complete 16S rRNA gene sequence from the genomic data of strain BSSL-BM10 T was extracted using ContEst16S (Lee et al. 2017). It was found to be identical to respective 16S rRNA gene information previously obtained by Sanger sequencing. This indicated that strain BSSL-BM10 T and its genomic data were not mislabeled. They did not originate from any source of contamination (Chun et al. 2018). Based on its genomic sequence data, the DNA G + C content of strain BSSL-BM10 T was 60.9%, a value in the range reported for Devosia species (Jia et al. 2014;Quan et al. 2020). The phylogenetic tree based on genomic sequences showed that strain BSSL-BM10 T formed a lineage within the clade comprising type strains of Devosia species (Fig. S3). The genomic sequence of strain BSSL-BM10 T had an ANI value of 77.2% with that of D. naphthalenivorans CM5-1 T and 71.0-78.4% with those of type strains of the other Devosia species as indicated in Table S1. Strain BSSL-BM10 T had dDDH values of 21.0% with D. naphthalenivorans CM5-1 T and 18.8-21.5% with type strains of the other 17 Devosia species (Table S1). These ANI values (71.0-78.4%) and dDDH values (18.8-21.5%) of genomic sequences between strain BSSL-BM10 T and type strains of Devosia species were lower than ANI and dDDH values (95-96% and 70%, respectively) recommended for delineation of a bacterial species (Goris et al. 2007; Konstantinidis and Tiedje 2005; Richter and Rosselló-Móra 2009).

Chemotaxonomic characteristics
The predominant isoprenoid quinone detected in strain BSSL-BM10 T was ubiquinone-10 (Q-10), consistent with results for members of Devosia species (Nakagawa et al. 1996;Quan et al. 2020). The major fatty acids (> 10% of total fatty acids in all growth phases) found in strain BSSL-BM10 T were 11-methyl C 18:1 ω7c, C 18:1 ω7c, summed feature 3 (C 16:1 ω7c and/or C 16:1 ω6c), and C 16:0 (Table S2). Fatty acid profiles of strain BSSL-BM10 T were similar to those of type strains of D. naphthalenivorans and D. riboflavina, with 11-methyl C 18:1 ω7c, C 18:1 ω7c, and C 16:0 being the major fatty acids. Nevertheless, there were differences in proportions of some fatty acids, including summed feature 3 (C 16:1 ω7c and/or C 16:1 ω6c) and cyclo C 19:0 ω8c, between strain BSSL-BM10 T and two reference strains (Table S2). Major polar lipids detected in stain BSSL-BM10 T were phosphatidylglycerol and two unidentified glycolipids. Minor amounts of diphosphatidylglycerol, two unidentified Only bootstrap values greater than 50% are shown at branching points. Filled circles indicate that the corresponding nodes were also recovered in the trees generated with the maximum-likelihood and maximum-parsimony algorithms. Stappia stellulata IAM 12621 T was used as an outgroup. Bar 0 0.01 substitutions per nucleotide position lipids, another unidentified glycolipid, and one unidentified aminolipid were also present (Fig. S4). The polar lipid profile of strain BSSL-BM10 T was similar to those of type strains of D. naphthalenivorans and D. riboflavina in that phosphatidylglycerol and two unidentified glycolipids were major components ( Fig. S4; Park et al. 2016).

Morphological, cultural, physiological and biochemical characteristics
Strain BSSL-BM10 T showed a Gram-negative, non-sporeforming, and non-flagellated properties and its cellular morphology was rod-shaped. Phenotypic characteristics of strain BSSL-BM10 T are given in the species description, Table 1, Table S3 and Fig. S5. Strain BSSL-BM10 T grew well on TSA and MA. Although the type strain of D. naphthalenivorans also grew well on MA, it grew poorly on TSA. Strain BSSL-BM10 T was resistant to gentamicin, whereas type strains of D. naphthalenivorans and D. riboflavina were susceptible to gentamicin (Table 1). Strain BSSL-BM10 T produced trypsin, α-glucosidase, and α-fucosidase, but the type strains of D. naphthalenivorans and D. riboflavin did not produce these three enzymes (Table 1).

Conclusion
Combined results obtained from phylogenetic, genomic, and chemotaxonomic analyses made it reasonable to assign strain BSSL-BM10 T as a member of the genus Devosia ( Fig. 1; Figs. S1, S2 and S3; Table S2). Strain BSSL-BM10 T was distinguished from type strains of D. naphthalenivorans and D. riboflavina by differences in several phenotypic characteristics, including nitrate reduction, acid production from D-glucose, utilization of some substrates, susceptibility to some antibiotics, and activities of some enzymes (Table 1). Based on polyphasic taxonomic data presented, strain BSSL-BM10 T is considered to represent a novel Devosia species, for which we propose the name Devosia litorisediminis sp. nov.
The GenBank/EMBL/DDBJ accession number for the 16S rRNA gene sequence and GenBank accession number for the whole genome shotgun sequence of strain BSSL-BM10 T are MN872411 and JAGXTP000000000, respectively.