High quality draft genomic sequence of Flavobacterium enshiense DK69^T and comparison among Flavobacterium genomes

Abstract

Flavobacterium enshiense DK69^T is a Gram-negative, aerobic, rod-shaped, non-motile and non-flagellated bacterium that belongs to the family Flavobacteriaceae in the phylum Bacteroidetes. The high quality draft genome of strain DK69^T was obtained and has a 3,375,260 bp genome size with a G + C content of 37.7 mol % and 2848 protein coding genes. In addition, we sequenced five more genomes of Flavobacterium type strains and performed a comparative genomic analysis among 12 Flavobacterium genomes. The results show some specific genes within the fish pathogenic Flavobacterium strains which provide information for further analysis the pathogenicity.

Introduction

Flavobacterium enshiense DK69^T (= CCTCC AB2011144^T = KCTC 23775^T ) is a type strain that belongs to the genus Flavobacterium of the family Flavobacteriaceae [1]. In recent years, members of Flavobacterium were identified and widely distributed in soil, fresh water, marine water, sediment, microbial mat, and glaciers [2–5]. Some Flavobacterium strains are fish pathogens including Flavobacterium columnare ATCC 49512^T causing columnaris disease [6], Flavobacterium psychrophilum JIP02/86^T causing cold-water disease [7] and Flavobacterium branchiophilum FL-15^T causing bacterial gill disease [8].

The common characters of Flavobacterium strains are Gram-negative, non-spore-forming, yellow-pigmented, rod-shaped, aerobic and with a low DNA G + C content (30–41 mol %) [2–12]. The Flavobacterium strains contained iso-C_15:0 as the major fatty acid, phosphatidylethanolamine as the major polar lipid and menaquinone-6 as the major respiratory quinone [9–12].

In order to provide genome information of Flavobacterium species, we sequenced six Flavobacterium strains including F. enshiense DK69^T [1], Flavobacterium beibuense F44-8^T [13], Flavobacterium cauense R2A-7^T [14], Flavobacterium rivuli WB 3.3-2^T [15], Flavobacterium subsaxonicum WB 4.1-42^T [15] and Flavobacterium suncheonense GH29-5^T [2]. In this study, we compared 12 genomes including the six strains that we sequenced and other six available Flavobacterium genomes in the NCBI, Flavobacterium indicum GPTSA100-9^T [16], Flavobacterium frigoris PS1^T [17], Flavobacterium sp. F52 [18], Flavobacterium columnare ATCC 49512^T , Flavobacterium psychrophilum JIP02/86^T and Flavobacterium branchiophilum FL-15^T. Here, we present the description of the non-contiguous finished genomic sequencing of F. enshiense DK69^T and the comparative genome analysis of the 12 Flavobacterium genomes.

Organism information

Classification and features

F. enshiense DK69^T is a Gram-negative, strictly aerobic, yellow-pigmented rod shaped bacterium isolated from soil collected at a pharmaceutical company in Enshi, Hubei province, China. The total soil C, N, P, S and Fe concentrations were 39.83, 3.34, 0.68, 0.36, 33.80 g kg⁻¹, respectively, and the pH was 6.97 [1]. A neighbor-joining phylogenetic tree based on the 16S rRNA gene sequences was built using MEGA 6 [19] and showed that strain DK69^T was clustered within a branch containing other species in the genus Flavobacterium (Fig. 1). In addition, the sequence of F. enshiense DK69^T was compared with other sequenced strains of the family Flavobacteriaceae use BioLinux [20], and a total of 24 core protein sequences were obtained with 50 % identity and E-value exponent of e⁻¹⁰. A phylogenetic tree based on the 24 core protein sequences of the core genome (Fig. 2) is similar to the 16S rRNA gene based tree.

Fig. 1

A NJ phylogenetic tree of the strains within family Flavobacteriaceae based on 16S rRNA gene sequence comparisons. GenBank accession numbers are shown in parentheses. The sequences were aligned using CLUSTALX, and the phylogenetic tree was obtained using MEGA 6 [19] software of neighbor-joining method [39], with the bootstrap values of 500 replicates. *represents the strains sequenced by us

Fig. 2

A NJ phylogenetic tree of the strains within family Flavobacteriaceae based on core-protein sequence comparisons. GenBank accession numbers are shown in parentheses. *represents the strains sequenced by us

The colonies of F. enshiense DK69^T are smooth with regular edges, circular, yellowish and about 1 mm in diameter after grown on R2A agar at 28 °C for 48 h. Growth occurs at 4–32 °C, pH 6.0–8.0 on R2A and TSA, but not on NA or LB media, and NaCl is not required [1]. Cells are non-flagellated, non-spore-forming, non-motile, rod-shaped (Fig. 3). Oxidase- and catalase- positive. The DNA G + C content is 34.4 mol% [1]. The general description of this strain is shown in Table 1.

Fig. 3

A transmission electron micrograph of F. enshiense DK69^T cells

Table 1 Classification and general features of F. enshiense DK69^T according to the MIGS recommendations [21]

Chemotaxonomic data

The major cellular fatty acids of F. enshiense DK69^T were iso-C_15:0, iso-C_17:1 ω9c, C_15:0, iso-C_17:0 3-OH and iso-C_15:0 3-OH. The major polar lipids were phosphatidylethanolamine, one unidentified aminolipid and one unidentified lipid. F. enshiense DK69^T contained menaquinone 6 as the major quinone [1].

Genome sequencing information

Genome project history

Genome of F. enshiense DK69^T was sequenced by Majorbio Bio-pharm Technology Co., Ltd, Shanghai, China. The high-quality draft genome sequence was deposited in the National Center for Biotechnology Information. Contigs less than 200 bp were not included. The GenBank accession number is JRLZ00000000. The summary of the genome sequencing project information is shown in Table 2.

Table 2 Project information of F. enshiense DK69^T

Growth conditions and genomic DNA preparation

F. enshiense DK69^T was grown on R2A medium at 28 °C for 2 d with 160 rpm shaking. Cells in late-log-phase growth were harvested and lysed by EDTA, lysozyme, and detergent treatment, followed by proteinase K and RNase digestion. The DNA was extracted and purified using the QiAamp kit according to the manufacturer’s instruction (Qiagen, Germany). The quantity of DNA was measured by the NanoDrop Spectrophotometer to ensure that the DNA concentration is greater than 20 ng/μl, then 5 μg of DNA was sent to Majorbio (Shanghai, China) for sequencing.

Genome sequencing and assembly

The Illumina Hiseq2000 with the Paired-End library strategy was used to determine the whole-genome sequence of F. enshiense DK69^T . TruSeq DNA Sample Preparation Kits are used to prepare DNA libraries with insert sizes of 300–500 bp for single, paired-end, and multiplexed sequencing. The protocol used 1 μg of DNA sheared by either sonication or nebulization [28]. The genome raw data of F. enshiense DK69^T generated 8,329,997 x 2 reads totaling 1,682,659,394 bp data with an average coverage of 498.4 x. Then SOAPdenovo v1.05 [29] was used to perform the following steps to assemble the sequencing data: (1) removing the adapter sequences in the reads; (2) cutting the 5’ end bases without clear A, T, C and G; (3) trimming the quality read scores lower than 20; (4) removing the reads containing more than 10 % Ns; (5) removing the reads which the length were less than 25 bp. A total of 8,217,761 x 2 high quality reads totaling 1,645,393,073 bp data with an average coverage 487.4 × was generated. The assembled sequence contained 67 scaffolds with a genome size of 3.38 Mbp.

Genome annotation

The annotation of the genomic sequences was completed using the NCBI Prokaryotic Genome Annotation Pipeline which was combined using Best-placed reference protein set and the gene caller GeneMarkS⁺. SignalP [30] and SOSUI [31] were used to predict signal peptides and transmembrane helices. The predicted CDSs were also used to search against the Pfam protein family database [32]. The GenBank database [33] and the COG databases [34] BLASTP search were used to predict protein sequences.

Genome properties

The genome statistics are provided in Table 3 and Fig. 4. After genome annotation, the genome of F. enshiense DK69^T was found to have a total length of 3,375,260 bp, a G + C content of 1,273,385 bp (37.7 mol %) and 74 contigs. From a total of 3,054 genes predicted, 2,848 genes are protein-coding genes, 50 are RNA genes, 57.9 % are assigned with putative functions and the remaining are annotated as hypothetical proteins or proteins of unknown functions. The distribution of genes into COGs functional categories is shown in Table 4.

Table 3 Genome statistics of F. enshiense DK69^T

Fig. 4

A graphical circular map of F. enshiense DK69^T. From outside to inside, 1, 4 circles show forward strand or reverse strand protein-coding genes according to COG categories; 2, 3 circles show forward strand or reverse strand genes; ring 5 shows G + C% content, ring 6 shows GC skew

Table 4 Number of genes in F. enshiense DK69^T associated with general COG functional categories

Insights from the genome sequences

Profiles of metabolic network and pathway

The metabolic network and pathways of F. enshiense DK69^T (Fig. 5) were predicted using the Kyoto Encyclopedia of Genes and Genomes [35]. The metabolic network showed that F. enshiense DK69^T possesses glycolysis, TCA cycle and pentose phosphate pathways and could utilize casein, tyrosine, sucrose and D-mannitol. The genome analysis results are in agreement with the phenotypes [1].

Fig. 5

Metabolic network and pathways of Flavobacterium enshiense DK69^T as predicted using KEGG [35]. Green lines indicate pathways that are possessed by this strain

Comparison of the 12 Flavobacterium genomes

The genomic information of the 12 Flavobacterium genomes are summarized in Table 5. OrthoMCL [36] analysis was performed to identify the set of orthologs among the 12 Flavobacterium genomes. F. enshiense DK69^T shared 1,190 genes with the other 11 Flavobacterium strains, and had 437 strain-specific genes which may contribute to the species-specific features (Fig. 6).

Table 5 General features of the twelve Flavobacterium genomes

Fig. 6

A venn diagram indicates the twelve genomes of Flavobacterium analyzed by OrthoMCL [36] illustrate the number of the unique proteins and the common proteins among them

Three of the 12 Flavobacterium strains are fish pathogenic bacteria [6–8]. Using OrthoMCL [36] analysis, a total of ten proteins we found to be unique in the three fish-pathogenic species. Three of the putative proteins were reported to be related to the pathogenicity of pathogenic bacteria including polysaccharide deacetylase [37], ABC transporter ATPase and ABC transporter permease [38] (Table 6).

Table 6 Specific proteins of three pathogenic bacteria, F. branchiophilum FL-15^T, F. columnare ATCC 49512^T and F. psychrophilum JIP02/86^T

Conclusions

The genomic results of F. enshiense DK69^T and related strains reveled useful information. (1) The genome based phylogenetic analysis results is in agreement with the 16S rRNA gene based one; (2) The genomic data are correlated with some phenotypes of strain DK69^T; (3) Compared to the three fish pathogenic Flavobacterium strains, no pathogenic related genes was detected in the environmental strain DK69^T which indicated its non-pathogenicity; and (4) Some specific genes were found within the three fish pathogenic Flavobacterium strains which provides information for further analysis the pathogenicity.