Genomic analysis of halophilic bacterium, Lentibacillus sp. CBA3610, derived from human feces

Strain CBA3610 was isolated from a stool sample from a 28-year-old healthy male in Gwangju, Republic of Korea. The fecal sample was enriched in Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ) medium 372 broth under aerobic conditions at 37 °C for 7 days, after which 100 mL of the enriched broth was spread on DSMZ medium 372 agar plates to isolate bacterial strains under aerobic conditions at 37 °C for 24 h. Strain CBA3610 was isolated from several colonies, and subculturing was performed under the same conditions at least three times.

The genomic DNA of the isolated strain was extracted and purified using the MG genomic DNA purification kit (MGmed, Seoul, Korea). The whole genome sequencing was performed using Pacific Biosciences RS II (Pacific Biosciences, Menlo Park, CA) and Illumina HiSeq X Ten (Illumina, San Diego, CA). Each library used for sequencing was constructed using a 20-kb SMRTbell template preparation kit and a TruSeq Nano DNA High Throughput Library kit. The genome was assembled using the protocol of Unicycler ver. 0.4.6 with PacBio SMRT analysis ver. 2.3 [6] and Pilon ver. 1.21 with Illumina HiSeq for error correction [7]. The subread filtering of the PacBio sequences was performed based on the following criteria: minimum subread length 50, minimum polymerase read quality 75, and minimum polymerase read length 50. Adapter/primer contamination of HiSeq raw sequences was confirmed using FastQC (v0.11.9). The genome was annotated using the Pathosystems Resource Integration Center (PATRIC; https://​www.​patricbrc.​org/​) ver. 3.6.7, the bacterial bioinformatics database and analysis resource [8]. We constructed a phylogenetic tree based on 16S rRNA gene sequences. To construct the phylogenetic tree, the sequences of 16S rRNA gene of strain CBA3610 and related species were aligned using Clustal W [9]. Phylogenetic trees were constructed using MEGA 7, based on the neighbor-joining (NJ) [10], maximum parsimony (MP) [11], and maximum likelihood (ML) [12] algorithms using 1000 bootstrap value [13]. Functional genes were predicted and annotated using Rapid Annotation using Subsystem Technology (RAST; https://​rast.​nmpdr.​org/​) [14]. PathogenFinder (https://​cge.​cbs.​dtu.​dk/​services/​PathogenFinder/​) was used for predicting pathogenicity towards humans [15]. The presence of Clustered Regularly Interspaced Short Palindromic Repeats (CRISPRs) was detected using the CRISPRfinder server (https://​crispr.​i2bc.​paris-saclay.​fr/​Server/​) [16]. Prophages were confirmed using the PHASTER database (https://​phaster.​ca/​), a phage search tool [17].

Comparative genome analysis was performed using 12 reference strains belonging to the genus Lentibacillus along with strain CBA3610. The genome and amino acid sequences of 12 reference strains are available in the GenBank of National Center for Biotechnology Information (NCBI, Accessed 22 September 2020). The list of strains used in the analysis is summarized in Additional file 1: Table S1. Pan-genome analysis was performed using Bacterial Pan Genome Analysis tool (BPGA). The 50% sequence identity cut-off was applied to obtain the core genomes of a total of 13 strains using USEARCH (ver. 9.0) [18]. The core genome tree was constructed with the aligned amino acid sequences of common genes of 13 strains using MAFFT (ver. 7.471) [19] and the MEGA 7 with NJ algorithm [10, 13]. The OrthoANI value was calculated using the Orthologous Average Nucleotide Identity Tool (OAT) provided by EzBioCloud database [20].

Before genomic DNA extraction, the single colony of strain CBA3610 was transferred three times in DSMZ medium 372 to obtain pure single colony. After obtaining the whole genome sequence of strain CBA3610, the sequence of the 16S rRNA gene, extracted using RNAmmer 1.21 server, was confirmed through the EzBioCloud database.

After the PacBio subreads filtering process, the total number of bases was 1,186,149,844 and the number of reads was 111,990. After the HiSeq raw data filtering process, the total number of bases in the filtered dataset was 796,687,476 and number of reads was 5,276,076. In the de novo assembly process, long-reads of PacBio were assembled using the default option. After de novo assembly with PacBio subreads and error correction using HiSeq reads, the complete genome of Lentibacillus sp. CBA3610 consists of one chromosome (Total length: 4,035,571 bp). No plasmid was identified. Chromosome was circular with 42% G + C content. According to the PATRIC annotation results, the genome has 4714 predicted genes, 166 repeat regions, 64 tRNA genes, and 17 rRNA genes. The genome of Lentibacillus sp. CBA3610 was annotated as having one virulence factor, four transporters, four drug targets, and 37 antibiotic resistance genes. The circular map of the genome is shown in Fig. 1, and detailed genomic characteristics are listed in Table 1. A phylogenetic tree was constructed, based on the 16S rRNA gene sequences of the strains with close similarity to the Lentibacillus sp. CBA3610 (Fig. 2A). The similarity of the 16S rRNA gene sequence of strain CBA3610 with Lentibacillus salicampi SF-20^T, Lentibacillus salarius BH139^T and Lentibacillus halodurans 8-1^T was 95.81%, 95.63% and 95.61%, respectively. On the phylogenetic tree based on the 16S rRNA gene sequences, strain CBA3610 clustered with Lentibacillus halodurans 8-1^T and Lentibacillus salarius BH139^T. Strain CBA3610 can be considered as a novel candidate species of the genus Lentibacillus [21]. Based on the results of the RAST annotation, the following categories were classified in the SEED subsystem: amino acids and derivatives (346), carbohydrates (265), protein metabolism (194), cofactors, vitamins, prosthetic groups, and pigments (106) (Additional file 1: Figure S1). Among the 27 categories based on RAST annotation, 55 coding sequences (CDSs) existed in the ‘Virulence, Disease and Defense’. Among these, five CDSs were found to belong to the ‘Resistance to fluoroquinolones’ category related to antibiotic resistance. Based on the results of the PathogenFinder, this strain was not classified as a human pathogen because only one sequence classified as pathogenic, and 14 other sequences classified as non-pathogenic, were identified (Additional file 1: Table S2). The sequence classified as that belonging to the pathogenic family showed 84.78% similarity to those annotated with the function of 30S ribosomal protein S19 in the genome of Listeria monocytogenes 08-5578. CRISPRFinder detected five sequences presumed to be CRISPR candidates (Additional file 1: Table S3), and two incomplete prophage regions were found using PHASTER (Additional file 1: Table S4). Among the incomplete prophage regions, region 1 was confirmed to match PHAGE_Bacill_G_NC_023719 and region 2 matched to PHAGE_Brevib_Jimmer1_NC_029104.

Taking into consideration of the characteristics of Lentibacillus of being a halophilic bacterium that survives in a high salinity environment, genes related to osmotic stress of strain CBA3610 were analyzed. In the SEED subsystem of the RAST server, total 34 genes classified as related to ‘osmotic stress’, were identified. Among these, one gene was classified as related to osmoregulation and the remaining 33 genes were annotated to be related to choline and betaine uptake and betaine biosynthesis. The gene involved in osmoregulation encoded the aquaporin family protein, a transporter of glycerol across the cytoplasmic membrane that has limited permeability to small uncharged compounds, such as water. The remaining 33 genes are described as follows. As genes involved in the biosynthesis of osmoprotectant glycine betaine, there is one betA gene that encodes oxygen-dependent choline dehydrogenase, which converts choline to betaine aldehyde, and three betB genes that encode NAD/NADP-dependent betaine aldehyde dehydrogenase, which converts betaine aldehyde to glycine betaine. In addition, there are seven opuD genes that encode glycine betaine transporter, which are involved in glycine betaine uptake, and 12 genes belonging to the opuA gene family (including the opuAA, opuAB, and opuAC gene) that encode glycine betaine/carnitine/choline ABC transporter. Lastly, there are two ProV genes encoding glycine betaine/proline betaine transport system ATP-binding protein involved in glycine betaine and proline betaine uptake, four genes belonging to the opuB gene cluster, including opuBA, opuBB, opuBC, and opuBD genes, encoding glycine betaine/carnitine/choline ABC transporter, one opuCB gene encoding carnitine transport permease protein, and three soxA genes that encode sarcosine oxidase alpha subunit that converts sarcosine to glycine [22‐24].

Results of the pan-genome analysis using BPGA showed that between strain CBA3610 and 12 reference strains, 11,961 genes of pan-genome and 849 genes of core genome were found (Additional file 1: Figure S2). Strain CBA3610 had 2212 accessory genes (present in genome of 2–12 strains) of strain CBA3610 and 449 unique genes present only in genome (Additional file 1: Table S5). In the phylogenetic tree based on the core genome, strain CBA3610 was located close to Lentibacillus persicus, Lentibacillus amyloliquefaciens, and Lentibacillus halodurans; it was confirmed that strain CBA3610 belongs to the genus Lentibacillus (Fig. 2B). The calculated OrthoANI values between strain CBA3610 and the remaining 12 reference strains are summarized in Additional file 1: Table S6. The range of OrthoANI values between strain CBA3610 and the remaining 12 reference strains was 68.69–79.68%, showing the minimum value with strain Lentibacillus sp. JNUCC-1 and the maximum value with strain Lentibacillus halodurans CGMCC 1.3702. This result also supplements that of the phylogenetic analysis described above.