Genome sequence of Shinella sp. strain DD12, isolated from homogenized guts of starved Daphnia magna

Shinella sp. strain DD12, a novel phosphite assimilating bacterium, has been isolated from homogenized guts of 4 days starved zooplankton Daphnia magna. Here we report the draft genome of this bacterium, which comprises 7,677,812 bp and 7505 predicted protein-coding genes. Electronic supplementary material The online version of this article (doi:10.1186/s40793-015-0129-3) contains supplementary material, which is available to authorized users.


Introduction
Shinella sp. strain DD12 was isolated from homogenized guts of 4 days starved zooplankton Daphnia magna in the frame of a study, describing the importance, diversity and stability of bacterial communities inside the Daphnia guts. Structural diversity of the bacterial communities were investigated over time, while D. magna were fed with different food sources or were let starve for 4 days, or starved to death [1,2].
Daphnia spp. are small filter-feeding cladoceran zooplankton organisms which play the role of key members in the freshwater food webs. Heterotrophic bacteria can contribute significantly to the nutrition of Daphnia species [3,4]. Furthermore, bacteria compared with many algae, are superior competitors for phosphorus and are often characterized by high P:C values [5]. This suggests that bacteria are a rich source of phosphorus for zooplankton [6].
Female D. magna were grown in water from the oligotrophic and low-phosphorus content (below 10 mg.m -3 concentration of total phosphorus in the water column) Lake Constance. Phosphorus in form of phosphate has been identified as the major limiting agent of phytoplankton growth in this lake [7,8]. Studies performed in the 1990s, after a long period of active care aiming to lower the phosphorus content in the lake water, showed that the primary production of phytoplankton was not influenced substantially after the decrease of the phosphorus content [9]. This phenomenon together with the fact that some bacteria can assimilate reduced inorganic and organic phosphorus compounds (phosphite [+III] and organophosphonates) under phosphate starvation [10][11][12][13][14][15][16][17][18][19], led us to investigate the newly isolated Shinella sp. strain DD12 in this aspect.
Shinella sp. strain DD12 was chosen for sequencing as it is able to assimilate phosphite under phosphate starvation and use it as single P-source to support its growth. We also focus on the following specific features of this genome -the assimilation of inorganic and organic phosphonates, providing that the organophosphonates are known to serve not only as P-, but as C-and N-sources for different bacteria. This is the first report on a genome sequence of a member of genus Shinella.
The cells of strain DD12 are short rounded rods with blunt ends and size of 0.6-1 μm in length, and 0.3-0.5 μm in width. Cells are motile via monotrichous flagellum ( Fig. 1, Left).
Shinella sp. DD12 forms colonies within 3 to 5 days, when grown on nutrient agar at 18°C (Fig. 1, Center). Colonies are circular, raised to convex, smooth milkywhite in color, slightly opaque with pronounced translucent halo-like edges. In liquid media cells form white fluffy aggregates with finger-like or tree-like morphology (Fig. 1, Right). The strain grows at the temperature range of 10-30°C. No growth was observed at 37°C. At 18°C the strain grows poorly on nutrient broth. At 21°C it grows with a doubling time of 54-61 h on nutrient broth. By employing a newly developed chemically defined medium (MDS3) with phosphate as the phosphorus source the doubling time was reduced to 32-33 h at 21°C. The composition of MDS3 medium and the conditions of the tests for phosphite assimilation are available in Additional file 1.
Shinella sp. strain DD12 is positive for catalase, catalase-peroxidase, β-galactosidase and β-glucosidase activity as described for all members of the genus [20,23]. Strain DD12 can grow oxidatively with the production of acid on different sugars and sugar alcohols. Shinella sp. strain DD12, like other Shinella species except S. fusca, cannot grow on melibiose or starch [20,[23][24][25][26]. It does not either grow on inulin as is found for S. kummerowiae, whereas there is no data reported for the rest of Shinella strains. Strain DD12 however, shows some specificity in substrate Evidence codes -IDA: Inferred from Direct Assay; TAS: Traceable Author Statement (i.e., a direct report exists in the literature); NAS: Non-traceable Author Statement (i.e., not directly observed for the living, isolated sample, but based on a generally accepted property for the species, or anecdotal evidence). These evidence codes are from the Gene Ontology project [54] assimilation, as the lack of growth on D-arabinose, while all Shinella strains can grow on this substrate with exception of S. yambaruensis [20,26]. Analogously, a weak growth on salicin was observed for Shinella sp. strain DD12, where five of the six Shinella strains cannot grow on this substrate. S. granuli growth on salicin remains undetermined [20,25]. We compared 16S gene sequences of Shinella sp. DD12 with the non-redundant nucleotide collection of NCBI using NCBI MegaBLAST [27,28]. This comparison revealed that the strain shares 99 % (1445/1453 bp) and 99 % (1438/1446 bp) sequence identity to the 16S rRNA gene sequences of Rhizobium sp. R-24658, and S. zoogloeoides 81 g, respectively. Figure 2 shows the phylogenetic neighborhood of Shinella sp. DD12 in a 16S rRNA sequence based tree of all Shinella type species.
The phylogenetic tree was calculated with MEGA5 [29] using the Maximum Likelihood method based on the Jukes Cantor model [30]. Sequences were downloaded from the RDP [31], aligned by CLUSTALW [32] and tested by the bootstrap approach with 1000 resamplings. The length of the tree branches was scaled according the number of substitutions per site (see size bar). Shinella sp. DD12 clustered together with S.granuli Ch06T KCTC12237.
The minimum information about the genome sequence (MIGS) is provided in Table 1, according to MIGS recommendations [33].

Genome project history
This bacterium was selected for sequencing on the basis of its environmental relevance to issues in global P-and N-cycles, and still widely unrecognized reduced P-cycle in nature. Prior to sequencing, Shinella sp. strain DD12 was tested for growth in a newly developed chemically defined liquid medium MDS3 supplemented with 1 mM sodium phosphite as single P-source. The growth and the phosphite assimilation ability of this isolate were confirmed at physiological level (three successive passages in triplicate). The genome project has been deposited in GenBank database (AYLZ00000000) and as an improved high-quality-draft genome sequence in IMG. Genome Sequencing and annotation were done at Göttingen Genomics Laboratory; while cultivation and analysis were performed at the University of Konstanz. The project information and its association with MIGS version 2.0 compliance [33] are presented in Table 2.  Growth conditions and genomic DNA preparation Shinella sp. DD12 was grown either in nutrient broth or on nutrient agar. The medium was adjusted to pH 7.0 and autoclaved for 25 min at 125°C. MDS3 medium was used to assay carbohydrate, phosphite and phosphate assimilation by the strain. The chemical composition of the MDS3 medium is given in Additional file 1.
The genomic DNA of the strain was isolated as follows: the cells from 4 ml of a well grown culture in nutrient broth reaching an OD 600 of 0.348 ± 0.050 were harvested at 13,000 × g in a benchtop microfuge for 5 min. Cell pellet was suspended in the cell lysis solution of the Purgene Core Kit B (Qiagen, Hilden, Germany). Further, the genomic DNA extraction processed as recommended by the manufacturer. DNA quantity was determined with NanoDrop ND-1000 to ensure that the concentration is greater than 30 ng/μl. One nanogram of the genomic DNA was used for sequencing.

Genome sequencing and assembly
Extracted DNA was used to prepare shotgun libraries for the Genome Analyzer II (Illumina, San Diego, CA, USA). Libraries were prepared according to the manufacturer protocol. Sequencing resulted in 7,118,226 paired-ends Illumina reads of 112 bp and a 72.54-fold coverage. Reads were trimmed using Trimmomatic 0.32 [34] to remove sequences with quality scores lower than 20 (Illumina 1.9 encoding) and remaining adaptor sequences. The initial hybrid de novo assembly employing the SPAdes 2.5 [35] software resulted in 236 contigs larger than 0.5 kb of which 162 were larger than 1 kb including 139 contigs larger than 3 kb. The final assembly had an N50 value of 97,231 bp and an N90 value of 24,331 bp.

Genome annotation
YACOB and GLIMMER [36] software tools were used for automatic gene prediction. RNAmmer [37] and tRNAscan [38] were used for identification of rRNA and tRNA genes, respectively. Functional annotation of the predicted protein-coding genes was carried out with the IMG/ER system [39] and was manually curated by using the Swiss-Prot, TrEMBL, and InterPro databases [40].

Genome properties
The genome statistics are provided in Table 3. The pseudogenes may also be counted as protein coding or RNA genes, so they are not additive under total gene count.
The draft genome of Shinella sp. DD12 consists of 236 contigs comprising 7.678 Mb and an overall GC content of 63.40 mol%. The genome harbors 7555 putative genes, of which 7505 are protein-encoding and 50 RNAs (2 rRNA and 48 tRNA). The tRNAs included tRNA necessary for selenocystein incorporation (SHLA_2c001070). Protein encoding genes with a putative function prediction are 6241 (82.61 %) of all proteins in the genome and 1264 (16.73 %) were annotated as hypothetical proteins. The majority of the protein-encoding genes 5394 (71.40 %) were assigned to one of the known COG categories [41]. The distribution of these genes with respect to assigned functions is presented in Table 4.

Insights from the genome sequence
The genome of Shinella sp. DD12 consists of a circular chromosome and at least 7 plasmids as we could detect  7 different repABC gene clusters located on 7 different contigs. Further database analysis revealed that all complete sequenced Rhizobiaceae genomes harbor usually between 2 and 6 plasmids, but species with up to 9 plasmids have been found, as in Ensifer fredii HH103 [42]. The strain is aerobe and its aerobic respiratory chain contains all genes encoding Complex I to Complex V. In addition, strain DD12 possesses a complete denitrification pathway via periplasmic cytochrome c [43]. The pathway found in this genome includes the genes encoding a periplasmic nitrate reductase napABC (SHLA_29c000730 -SHLA_29c000770), NO-forming nitrtite reductase nirK (SHLA_5c000410), nitric oxide reductase norCBD (SHLA_5c000290 -SHLA_5c000340) and a nitrous oxide reductase nosZ (SHLA_36c000580). The genome analysis of the strain DD12 revealed the potential abilities of this isolate to reduce nitrogen via the dissimilatory nitrate reduction to ammonia (DNRA) pathway, and to assimilate nitrate to L-glutamine and L-glutamate. The genes encoding nitrogen fixation ability such as N-acetylglucosaminyl transferase (nodC) or nitrogenase reductase (nifH) are absent from the genome. This is consistant with the previously reported lack of nitrogen fixation ability in free-living Shinella species, except for the only known symbiont S. kummerowiae [20,23,25].
Shinella sp. strain DD12 is able to utilize reduced inorganic phosphonate (phosphite) and presumably organophosphonates as the single P-sources to support its growth. The phosphite oxidation most probably proceeds through a periplasmic alkaline phosphatase (phoA), analogously to E.coli [18], or through the carbon-phosphorus (C-P) lyase complex [10][11][12]15]. The latter complex is known to have broader substrate specificity, including the oxidation of phosphite and the assimilation of the most common organophosphonate -methylphosphonate. The C-P lyase complex, although the presence of the conserved structural phnGHIJKLM gene cluster, shows low conservation level of the gene sequences arrangement amongst representatives of Alpha-Beta-Gammaand Deltaproteobacteria (Fig. 3). However, this drastically changes within the Rhizobiaceae members harboring a C-P lyase complex. The C-P complex shows highest conservation level of the gene sequences and their arrangement amongst the Rhizobiaceae members that harbors it. The total is based on the total number of protein coding genes in the genome In addition, the Shinella sp. DD12 genome harbors a 2-aminoethylphosphonate (2-AEP) degradation pathway, which operates through the phosphonoacetaldehyde dehydrogenase -phosphonoacetate hydrolase (phnWAY) [16,44]. The 2-AEP (ciliatine) is a common phosphonate constituent of the phospholipids in a variety of marine invertebrates, including ciliated protozoa, sees anemones, some plants and animals. Recently, the synthesis of sphingophosphonolipids was found in some bacterial species including Bacteriovorax stolpii, a facultative predator which parasitizes larger Gram-negative bacteria [45]. A Tblastx comparison of the phnWAY encoding operon from Shinella sp. strain DD12 with another 3 species belonging to Alphaproteobacteria, two of which members of Rhizobiaceae is shown on Fig. 4. An analysis of all genomes available at IMG (as of April 1, 2015) against phosphonoacetaldehyde dehydrogenase encoding gene (phnY) revealed its presence in 431 gene clusters. However, the complete phnWAY operon was present in only 92 genomes of which 41 belong to Rhizobiaceae species. Furthermore, the phnWAY operon was placed in the majority of the Alphaproteobacteria genomes in close proximity to the fbpABC transporter involved in the utilization of xenosiderophores as iron sources in a TonB-independent manner. It is known that the fbpABC gene cluster is transcribed as separate operon in Neisseria meningitidis [46]. However, whether this cluster plays a role in phosphonate uptake in the cell is unclear.

Conclusions
The draft genome sequence of Shinella sp. strain DD12 described here is the first genome sequence of a member of the genus Shinella. The genome of the strain DD12 suggests the presence of 7 plasmids, which is often found amongst members of Rhizobiaceae. Fig. 3 Tblastx comparison of the C-P lyase complex. An E-value cutoff of 1e -10 was used and visualization was done with the program Easyfig [55]. Functional genes of the C-P lyase complex were marked in green, accessory genes in orange, the ABC-type transporter in blue and the regulatory subunit in red. Genes not directly associated with this pathway are marked in grey Fig. 4 Tblastx comparison of Alphaproteobacteria phnWAY operon. An E-value cutoff of 1e -10 was used and visualization was done with the program Easyfig [55]. The phnWAY cluster was marked in orange tones; sodium/phosphonate symporter (yjbB) was marked in purple; the genes coding ABC-type ferric uptake system (fbpABC) were marked in blue; the transcriptional regulator (lysR) is shown in green The genome analysis of Shinella sp. strain DD12 indicates that the bacterium is a denitrifier, as it harbours two complete sets of genes encoding: i) the dissimilatory nitrate reduction to ammonia pathway and ii) assimilative nitrate reduction to L-glutamine, and L-glutamate pathway. Shinella sp. strain DD12 cannot fix nitrogen, similarly to the other free-living known Shinella species, whereas the symbiotically growing S. kummerowiae is a nitrogen fixing bacterium.
Finally, the genome of Shinella sp. DD12 encodes three complete pathways for assimilation of phosphonates. The presence of these three pathways indicates relatively broad abilities to utilise reduced phosphonates as P-and/or C-and N-sources, compared to the remaining genomes of Rhizobiaceae members and even to Alphaproteobacteria as a whole. This could be a great advantage for the strain DD12 in environments where other bacteria can face growth limitations, providing that the inorganic-and organophosphonates are naturally occurring compounds. Furthermore, the presence of the genes encoding the complete pathway for 2-AEP containing biomolecules might provide a defence mechanism against predator and parasite bacteria.

Competing interests
The authors declare that they have no competing interests.
Authors' contributions DDS and AP designed research. DDS characterized the strain DD12. HF has contributed for the chemotaxonomy. AP carried out genome analyses with the help of DDS. DDS, AP and RD wrote the manuscript. All authors read and approved the final manuscript.