Microbial Culturomics to Map Halophilic Bacterium in Human Gut: Genome Sequence and Description of Oceanobacillus Jeddahense Sp

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/301533830

Microbial Culturomics to Map Halophilic Bacterium in Human Gut: Genome Sequence and Description of Oceanobacillus jeddahense sp. nov

Article in Omics: a journal of integrative biology · April 2016 Impact Factor: 2.36 · DOI: 10.1089/omi.2016.0004

12 authors, including:

Fehmeeda Imran Olivier Croce King Abdulaziz University French National Centre for Scientific Research

34 PUBLICATIONS 135 CITATIONS 62 PUBLICATIONS 190 CITATIONS

SEE PROFILE SEE PROFILE

Asif Jiman-Fatani Catherine Robert King Abdulaziz University Aix-Marseille Université

34 PUBLICATIONS 75 CITATIONS 273 PUBLICATIONS 21,597 CITATIONS

SEE PROFILE SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Jean-Christophe Lagier letting you access and read them immediately. Retrieved on: 02 May 2016 OMICS A Journal of Integrative Biology Volume 20, Number 4, 2016 ª Mary Ann Liebert, Inc. DOI: 10.1089/omi.2016.0004

Microbial Culturomics to Map Halophilic Bacterium in Human Gut: Genome Sequence and Description of Oceanobacillus jeddahense sp. nov.

Saber Khelaiﬁa,1* Jean-Christophe Lagier,1* Fehmida Bibi,2* Esam Ibraheem Azhar,2,3 Olivier Croce,1 Roshan Padmanabhan,1 Asif Ahmad Jiman-Fatani,4 Muhammad Yasir,2 Catherine Robert,1 Claudia Andrieu,1 Pierre-Edouard Fournier,1 and Didier Raoult1,2

Abstract

Culturomics is a new omics subspecialty to map the microbial diversity of human gut, coupled with a taxono- genomic strategy. We report here the description of a new bacterial species using microbial culturomics: strain S5T,(= CSUR P1091 = DSM 28586) isolated from a stool specimen of a 25-year-old obese patient from Saudi Arabia. The strain S5T was a Gram-positive, strictly aerobic rod, which was motile by a polar ﬂagellum, spore- forming, and exhibited catalase and oxidase activities. It grows optimally at 37C, with a pH of 7.5 and 10% of NaCl. 16S rRNA gene-based identiﬁcation revealed that strain S5T has 98.6% 16S rRNA sequence similarity with the reference O. oncorhynchi, phylogenetically the closest validated Oceanobacillus species. Here, we further describe the phenotypic characteristics of this organism and its complete genome sequence and annotation. The 5,388,285 bp long genome exhibits a G + C content of 37.24% and contains 5109 protein-coding genes and 198 RNA genes. Based on the characteristics reported here, we propose classifying this novel bacterium as representative of a new species belonging to the genus Oceanobacillus, Oceanobacillus jeddahense sp. nov. In a broader context, it is noteworthy that halophilic bacteria have long been overlooked in the human gut, and their role in human health and disease has not yet been investigated. This study thus further underscores the usefulness of the culturomics approach exploring the bacterial diversity of the gut.

Introduction genera, leading to some limitations (Kokcha et al., 2012). Genomic data for many bacterial species are now available culture-based emerging omics concept is the ﬁeld due to the emergence of high-throughput sequencing tech- Aof ‘‘culturomics’’ that has recently been introduced, niques (Mishra et al., 2012). revealing high bacterial diversity in human stool samples, We have recently outlined a new method (taxono- including halophilic organisms (Lagier et al., 2012a). Cul- genomics), including genomic data in a polyphasic approach turomics has helped to map the microbial diversity of human in order to describe new bacterial species (Ramasamy et al., gut, coupled with a taxono-genomic strategy. Notably, 16S 2014). This strategy combines phenotypic characteristics rRNA sequence identity and phylogeny (Rossello-Mora, with MALDI-TOF MS spectrum and genomic analysis (La- 2006; Wayne et al., 1987), genomic G + C content diversity gier et al., 2012b; Mishra et al., 2012). Using this method, we and DNA–DNA hybridization (DDH) (Ramasamy et al., isolated the strain S5T belonging to the genus Oceanoba- 2014; Welker and Moore, 2011) have been the traditional cillus. The ﬁrst description of genus Oceanobacillus was parameters used to delineate a bacterial species. However, made by Lu et al. (2001) and was then amended (Yumoto the cut-off values vary dramatically between species and et al., 2005). These bacteria belong to the phylum Firmicutes

1Unite´ de Recherche sur les Maladies Infectieuses et Tropicales Emergentes, CNRS (UMR 7278), IRD (198), INSERM (U1095), AMU (UM63), Institut Hospitalo-Universitaire Me´diterrane´e-Infection, Faculte´ de Me´decine, Aix-Marseille Universite´, Marseille, Franca. 2Special Infectious Agents Unit, King Fahd Medical Research Center, and Departments of 3Medical Laboratory Technology, Faculty of Applied Medical Sciences, and 4Medical Microbiology and Parasitology, Faculty of Medicine, King Abdulaziz University, Jeddah, Saudi Arabia. *These three authors contributed equally.

248 CULTUROMICS OF OCEANOBACILLUS JEDDAHENSE 249 within the Bacillaceae family. The Oceanobacillus genus Strain identification using MALDI-TOF currently includes 13 recognized species and two subspecies. and 16S rRNA sequencing These bacteria are Gram-positive, motile rods, obligate MALDI-TOF MS (matrix-assisted laser-desorption/ aerobic or facultative anaerobic, and obligate or facultative ionization time-of-flight) protein analysis was carried out as alkaliphilic. previously described (Seng et al., 2009). The resulting twelve Oceanobacillus T Bacteria from the genus were essentially spectra of strain S5 were imported into the MALDI Bio- ., isolated from deep-sea sediment core (Lu et al 2001; Kim Typer software (version 2.0, Bruker) and analyzed using et al., 2007; Yumoto et al., 2005), deteriorated mural paint- standard pattern matching (with default parameter settings) ings (Heyrman et al., 2003), salt fields (Lee et al., 2006), against the main spectra of over 6252 bacteria, including freshwater fish (Yumoto et al., 2005), algal mat (Romano the spectra from the most closely related species: Oceano- et al., 2006), freshwater insects (Raats and Halpern, 2007), a bacillus oncorhynchi CIP108867T, O. picturae CIP 108264T, Bacillus -dominated wastewater treatment system in Korea O. profundus CIP 109535T, O. chironomi CIP 109536T, O. (Nam et al., 2008), fermented shrimp paste samples (Nam- iheyensis CIP 107618T, and O. oncorhynchi subsp. in- wong et al., 2009), soy sauce production equipment (Tomi- caldanensis CIP 109235T. The 16S rRNA PCR and sequenc- naga et al., 2009), marine solar saltern (Lee et al., 2010), ing were performed as previously described (Mourembou activated sludge in a bioreactor (Whon et al., 2010), tradi- et al., 2015). tional Korean fermented food (Hirota et al., 2013), fermented Polygonum indigo liquor sample (Yang et al., 2010), and, finally, a human stool sample (Roux et al., 2013). Growth conditions Here, we used high salt culture conditions in order to The optimum growth temperature of the strain S5T was cultivate halophilic bacteria from human faeces and a novel tested on our home-made solid medium by inoculating 105 Oceanobacillus isolate, strain S5T ( = CSUR P1091 = DSM CFU/mL of an exponentially growing culture incubated 28586), was isolated from a stool sample of a 25-year- aerobically at 28,37,45, and 55C. The growth atmo- old obese Saudi patient, as part of culturomics study of in- sphere was tested under aerobic conditions, in the presence of testinal microflora, as a new representative of the genus 5% CO2, and also in microaerophilic and anaerobic condi- Oceanobacillus. tions created using GENbag microaer and GENbag anaer (BioMe´rieux, Marcy l’Etoile, France), respectively. The Materials and Methods optimum NaCl concentration required for growth was tested at 0%, 0.5%, 1.5%, 5%, 7.5%, 10%, 15%, and 20% of NaCl. Ethics and collection of the strain The optimum pH was determined by growth testing at pH 5,

The stool specimens were collected from a 25-year-old 6, 7, 7.5, 8. and 9. obese Saudi patient after defecation into sterile plastic con- tainers, which were sampled and stored at -80C until use. Biochemical, sporulation, and motility assays (Supplementary Table S1. Supplementary material is available The commercially available Api ZYM, Api 20 NE, Api 50 online at www.liebertpub.com/omi). Informed and signed CH strips (bioMe´rieux), supplemented by 10% NaCl (w/v), consent was obtained from the patient. The study and consent were used to characterize the biochemical properties of the procedure were approved by the Ethics Committees of the strain according to the manufacturer’s instructions. Strain King Abdulaziz University (King Fahd Medical Research sporulation was tested by thermic-shock at 60C for 20 Center, Saudi Arabia), under agreement number 014- T minutes. A fresh culture of the strain S5 was observed under CEGMR-2-ETH-P, and by the Institut Fe´de´ratif de Recherche DM1000 photonic microscope (Leica Microsystems, Nan- 48 (Faculty of Medicine, Marseille, France), under agreement terre, France) at 100X to assess the motility of the bacteria. number 09-022. The salt concentration of the stool specimen The colony’s surface was observed on our home-made cul- was determined using a digital refractometer (Fisher Scientiﬁc, ture medium after 24 h incubation under aerobic conditions at Illkirch, France) and the pH by using a pH-meter. 37C.

Isolation of the strain Antibiotic susceptibility testing T Strain S5 was isolated in December 2013 using aerobic Susceptibility to antibiotics was tested using antibiotic cultivation on a home-made culture medium consisting of a disks (B. Braun Medical SAS, Boulogne, France) containing Columbia broth culture medium (Sigma-Aldrich, Saint- the following antibiotics: doxycycline, rifampicin, vanco- Quentin Fallavier, France) modified by adding (per liter): mycin, nitrofurantoin, amoxicillin, erythromycin, ampicillin, MgCl2 6H2O, 5 g; MgSO4 7H2O, 5 g; KCl, 2 g; CaCl2 2H2O, ceftriaxone, ciprofloxacin, gentamicin, penicillin, trimethoprim/ 1g;NaBr,0.5g;NaHCO3, 0.5 g; glucose 2 g, and 10% of NaCl. sulfamethoxazole, imipenem, and metronidazole. The pH was adjusted to 7.5 with 10 M NaOH before auto- claving. Approximately 1 g of stool specimen was inoculated Microscopy and Gram test aerobically into 100 mL of our home-made liquid medium in flasks at 37C, stirring at 150 rpm. Subcultures were conducted Transmission electron microscopy of the strain S5T was after 3, 10, 15, 20, and 30 days incubation. Serial dilutions of performed after negative staining using a Morgani 268D 10-1 to 10-10 were then performed in the home-made liquid (Philips, Verdun, France) at an operating voltage of 60kV. culture medium and plated on agar medium. After 3 days’ in- The Gram stain was performed using the colour Gram 2 kit cubation at 37C, all colonies that appeared were picked and (Biomerieux) and observed using a DM1000 photonic mi- streaked replant several times to obtain pure cultures. croscope (Leica Microsystems). 250 KHELAIFIA ET AL.

DNA extraction Genome comparison DNA extraction of the strain S5T was conducted on ten To estimate the mean level of nucleotide sequence sim- Petri dishes of fresh culture that were harvested and re- ilarity at genome level between the strain S5T and the suspended in 4X 100 lL of TE buffer. 200 lL of this sus- four closest species with an available genome, we used the pension was then diluted in 1 mL TE buffer for lysis Average Genomic Identity of Orthologous gene Sequences treatment that included 30 min incubation with 2.5 lg/lLof (AGIOS), an in lab pipeline. Briefly, this pipeline combines lysozyme at 37C, followed by overnight incubation with the Proteinortho software (with the following parameters: e- 20 lg/lL of proteinase-K at 37C. Extracted DNA was then value 1e-5, 30% identity, 50% coverage, and algebraic con- purified using three successive phenol-chloroform extrac- nectivity of 50%) for detecting orthologous proteins between tions and ethanol precipitation at -20C overnight. After genomes compared pairwise, and retrieves the corresponding centrifugation, the DNA was resuspended in 160 lLTE genes and determines the mean percentage of nucleotide buffer. The yield and concentration was measured using a sequence identity between orthologous ORFs using the Quant-it Picogreen kit (Invitrogen, Villebon sur Yvette, Needleman-Wunsch global alignment algorithm (Ramasamy France) on the Genios-Tecan fluorometer. et al., 2014).

Genome sequencing and assembly Results T Genomic DNA (gDNA) of the strain S5 was sequenced Phenotypic and biochemical characterization using MiSeq Technology (Illumina Inc, San Diego, CA, T USA) with the mate pair strategy. The gDNA was barcoded The colonies of the strain S5 were circular, grayish, shiny, in order to be mixed with 11 other projects with the Nextera and smooth with a diameter of 2–5 mm. Cells stained Gram- Mate Pair sample prep kit (Illumina). The gDNA was quan- positive and were rod shaped (Fig. 1). They were motile by tified by a Qubit assay using the high sensitivity kit (Life polar flagella, spore-forming, had a length of 2–4 lm and a Technologies, Carlsbad, CA).The mate pair library was diameter of 0.5 lm, and generally occurred singly or in pairs T prepared with 1 lg of genomic DNA using the Nextera (Fig. 2). Strain S5 was mesophilic and grew at temperatures Amate pair Illumina guide. The gDNA sample was simulta- ranging from 15 to 45C, with optimum growth at 37C. The neously fragmented and tagged with a mate pair junction isolate required NaCl for growth and grew at salinity ranging adapter. The profile of the fragmentation was validated on an from 0.5 to 20% of NaCl (optimum at 10%). The optimum Agilent 2100 BioAnalyzer (Agilent Technologies Inc., Santa pH for growth was 7.5 (range between pH 5 to 9). The strain T Clara, CA) with a DNA 7500 labchip. The DNA fragments S5 was aerobic and also grew in microaerophilia and in were sorted by size from 1 kb up to 10 kb. No size selection the presence of 5% CO2 but did not grow in an anaerobic was performed and only 14 ng of tagmented fragments atmosphere. T were circularized. The circularized DNA was mechanically The strain S5 exhibited catalase and oxidase activity. sheared to small fragments with an optimal at 696 bp on the Using an Api ZYM strip and Api 20 NE, a negative reaction Covaris device S2 in microtubes (Covaris, Woburn, MA, was observed only for alkaline phosphatase, b-galactosidase, USA).The library profile was visualized on a High Sensitivity N-acetyl-b-glucosaminidase, and urease. Positive reactions Bioanalyzer LabChip (Agilent Technologies. Inc.). were observed for nitrate reduction. Using API 50CH strip The libraries were normalized at 2 nM and pooled. After a and Api 20E at 37C, negative reactions were obtained for D- denaturation step and dilution at 10 pM, the pool of libraries lactose, L-arabinose, D-galactose, and D-ribose. Positive re- was loaded onto the reagent cartridge and then onto the in- actions were obtained for D-glucose, D-fructose, D-mannose, strument along with the flow cell. The automated cluster D-mannitol, D-maltose, and D-sucrose. Phenotypic charac- generation and sequencing run were performed in a single 42- teristics were compared to those of the most closely related hour run in a 2 · 251-bp. Illumina reads were trimmed using species (Table 1). Antimicrobial susceptibility testing dem- T Trimmomatic (Lohse et al., 2012), then assembled through onstrate that the strain S5 was susceptible to doxycycline, Spades software (Bankevich et al., 2012; Nurk et al., 2013). rifampicin, vancomycin, nitrofurantoin, amoxicillin, eryth- Contigs obtained were combined together using SSpace romycin, ampicillin, ceftriaxone, ciprofloxacin, gentamicin, (Boetzer et al., 2011) and Opera software v1.2 (Gao et al., penicillin, and imipenem, but resistant to trimethoprim/ 2011), with assistance from GapFiller V1.10 (Boetzer et al., sulfamethoxazole and metronidazole. 2012) to reduce the set. Some manual refinements using CLC Genomics v7 software (CLC bio, Aarhus, Denmark) and MALDI-TOF and phylogenic analysis homemade tools improved the genome. MALDI-TOF results showed that the spectra we obtained for strain S5T did not match with one of our database (Bruker Genome annotation constantly increment with our data), suggesting that our Non-coding genes and miscellaneous features of the strain isolate was a new species. We added the spectrum from strain S5T were predicted using RNAmmer (Lagesen et al., 2007), S5T to our database (Fig. 3). In contrast, PCR-based identi- ARAGORN (Laslett et al., 2004), Rfam (Griffiths-Jones fication of the16S rRNA of the strain S5T (HG931338) et al., 2003), PFAM (Punta et al., 2012), and Infernal (Na- demonstrated 98.6% 16S rRNA gene sequence homology wrocki et al., 2009). Coding DNA sequences (CDSs) were with reference O. oncorhynchi (DQ089679), the phyloge- predicted using Prodigal (Hyatt et al., 2010) and functional netically closest validated Oceanobacillus species (Fig. 4). annotation was achieved using BLAST+ (Camacho et al., This value is under the threshold that allows for identification 2009) and HMMER3 (Eddy 2009) against the UniProtKB of new species, establish by Stackebrandt and Ebers (2006). database (The UniProt Consortium, 2011). Consequently, our strain is considered as a new species, CULTUROMICS OF OCEANOBACILLUS JEDDAHENSE 251

FIG. 1. Gram staining of Oceanobacillus jeddahense strain S5T.

FIG. 2. Transmission electron microscopy of Oceanobacillus jeddahense strain S5T. The scale bar represents 500 nm.

T Table 1. Differential Characteristics of Oceanobacillus jeddahense strain S5 with Other Strains* Properties O. jeddahense O. oncorhynchi O. iheyensis O. sojae O. massiliensis O. chironomi O. picturae O. indicireducens Cell diameter (lm) 0.4–0.6 0.4–0.6 0.6–0.8 0.4–0.7 0.8–1 0.5–0.6 0.4–0.8 Oxygen requirement Aerobic Aerobic Aerobic Aerobic Aerobic Aerobic Aerobic Aerobic Gram stain ++ ++++++ Salt requirement ++ ++++++ Motility ++ ++++++ Endospore formation ++ ++-++- Indole - na +- - na - Production of Alkaline phosphatase -- --na - na Catalase ++ ++++w + Oxidase ++ v ++ + w + Nitrate reductase ++ ---+w + 252 Urease - na --- na -- b-Galactosidase -+ +--na -+ N-Acetyl-glucosamine - na -+-+ - + Acid from L-Arabinose -- na -- na -+ Ribose -- -+-+-+ D-Mannose ++ ++++w + D-Mannitol + na na ++ na w + D-Sucrose ++ -+-na w w D-Glucose ++ ++++w + D-Fructose ++ ++++w + D-Maltose ++ ++++w + D-Lactose -- --na -- - Habitat Human gut Rainbowtrout skin Deep-sea sediment Soy sauce Human gut Chironomid egg Mural paintings Algal mat

*Oceanobacillus oncorhynchi (Yumoto et al., 2005), Oceanobacillus iheyensis (Lu et al., 2001), Oceanobacillus sojae (Tominaga et al., 2009), Oceanobacillus massiliensis (Roux et al., 2013), Oceanobacillus chironomi (Raats et al., 2007), Oceanobacillus picturae (Lee et al., 2006), Oceanobacillus indicireducens (Hirota et al., 2013). na = data not available; v = variable reaction; w = weak. CULTUROMICS OF OCEANOBACILLUS JEDDAHENSE 253

FIG. 3. Reference mass spectrum from Oceanobacillus jeddahense strain S5T. Spectra from 10 individual colonies were compared and a reference spectrum was generated. named Oceanobacillus jeddahense. Finally, the gel view et al., 2000) with BLASTP default parameters. Table 3 pres- demonstrates the spectral differences with other members of ents the distribution of genes in COGs functional categories. the genus Oceanobacillus (Fig. 5). Genome comparison Genome information and properties Brief comparisons were made with Oceanobacillus The O. jeddahense strain S5T genome was sequenced as iheyensis strain HTE831 (BA000028), which is currently the part of culturomics study aiming to isolate all bacterial spe- closest available annotated sequenced genome. The draft cies colonizing the human gut (Lu et al., 2001) and because of genome sequence of O. jeddahense strain S5T is bigger in size its phylogenetic affiliation to the Oceanobacillus genus. than the O. iheyensis (respectively, 5.39 Mb vs. 3.63 Mb). O. Strain S5T represents the fourth genome sequenced in the jeddahense strain S5T exhibit more genes (5217 genes com- Oceanobacillus genus and the first genome of O. jeddahense pared to 3593 genes) and also has a similar ratio of genes per sp. nov. The O. jeddahense strain S5T genome contains Mb (967.9 genes/Mb compared to 989.8 genes/Mb).The 5,388,285 bp with a G + C content of 37.24% (Fig. 6) and G + C content is close to Bacteroides intestinalis (37.24% and consists of 11 scaffolds with 12 contigs. 35.7%, respectively). The gene number difference (in per- The draft genome (Genbank accession number centage) related to each COG categories between O. jedda- CCDM010000001.1) was shown to encode at least 198 pre- hense and O. iheyensis is presented in Table 3. The maximum dicted RNA including 17 rRNA, 65 tRNA, 1 tmRNA, and difference relates to the COG ‘‘Transcription’’ which does 115 miscellaneous RNA (Table 2). 5109 protein-coding not exceed 2.32%. In terms of the other categories, the two genes were also identified, representing a coding capacity of species are highly similar. 4,382,541 bp (coding percentage: 81.3%). Of these genes, In addition, a whole genome comparison was performed 860 (16.83%) were found to be putative proteins and 1109 between O. jeddahense, O. massiliensis, O. kimchii (uid (21.71%) were assigned as hypothetical proteins. Moreover, 199920), O. iheyensis (uid 57867), and Halobacillus halo- 4303 genes matched against at least one sequence in the philus (uid 162033). At the time of analysis, only three whole Clusters of Orthologous Groups (COGs) database (Tatusov genome sequences of Oceanobacillus were available at 254 KHELAIFIA ET AL.

FIG. 4. Phylogenetic tree highlighting the position of Oceanobacillus jeddahense strain S5T relative to other type strains within the genus Oceanobacillus. GenBank accession numbers are indicated in paren- theses. Sequences were aligned using CLUSTALW, and phylogenetic inferences obtained using the maximum-likelihood method within the MEGA software. Halobacillus litoralis was used as outgroup. The scale bar represents 0.005% nucleotide sequence divergence.

FIG. 5. Gel view comparing Oceanobacillus jeddahense strain S5T to other species within the genus Oceanobacillus and Ornithinibacillus. The gel view displays the raw spectra of loaded spectrum ﬁles arranged in a pseudo-gel like look. The x-axis records the m/z value. The left y-axis displays the running spectrum number originating from subsequent spectra loading. Peak intensity is expressed by a gray scale scheme. The color bar and the right y-axis indicate the relation between the colour a peak is displayed with and the peak intensity in arbitrary units. Displayed species are indicated on the left. CULTUROMICS OF OCEANOBACILLUS JEDDAHENSE 255

FIG. 6. Circular representation of the Oceanobacillus jeddahense strain S5T chromosome. Circles from the center to the outside: GC skew (green/purple), GC content (black), RNA (dark red, purple, blue, beige)on forward strand, genes on forward strand colored by COGs categories, scaffolds in alternative grays, genes on reverse strand colored by COGs, RNA (dark red, purple, blue, beige) on reverse strand.

NCBI. Among the species of Oceanobacillus spp., the draft between the O. jeddahense strain S5T and other compared genome of O. jeddahense strain S5T (5.4 Mb) is the largest, species are roughly half compared to the total proteins in the followed by O. kimchii (3.8 Mb), O. iheyensis (3.6 Mb), and strain S5T, suggesting a high number of hypothetical proteins O. massiliensis (3.5 Mb), respectively. H.halophilus has a in this genome. genome size of 4.1 Mb. The AGIOs values in Oceanobacillus genus for the four Discussion studied genomes range from 67.70% to 89.62%. The AGIOs This work describes the isolation of a moderately halo- values for O. jeddahense strain S5T and other species of philic bacterium associated with the human microbiota using Oceanobacillus range from 67.70% (O. massiliensis) to culture-based detection methods. The culturomics approach 69.85% (O. iheyensis and O. kimchii), thus confirming its was applied to render a dynamic description of new bacteria new species status (Table 4). The orthologous genes shared species and genera from the human gut, including samples from various geographical origins. In this way, several new halophilic bacteria species were recently isolated in our Table 2. Nucleotide Content and Gene laboratory, including two new species recently described Count Levels of the Genome (Khelaifia et al., 2015; Seck et al., 2015). Here, we report the phenotypic characterization and genomic description of a Attribute Value % of Total* new halophilic bacterium isolated from a stool specimen collected from 25-year-old obese patient from Saudi Arabia. Genome size (bp) 5,388,285 100 This strain was deposited in two different collections of mi- DNA coding region (bp) 4,382,541 81.3 DNA G + C content (bp) 2,006,597 37.24 croorganisms, Deutsche Sammlung von Mikroorganismen Total genes 5,217 100 und Zellkulturen (DSMZ) and the Collection de Souches de rRNA 17 0.33 l’Unite´ des Rickettsies (CSUR). After this study, O. jedda- tRNA 65 1.24 hense was isolated from three different stool specimens and tmRNA 1 0.02 one duodenum simple (unpublished data), confirming it is not miscRNA 115 2.2 uncommon in the human microbiota. Protein-coding genes 5,109 97.93 This work supports the usefulness of the culturomics ap- Genes with function prediction 3,910 76.53 proach exploring the bacterial diversity of the gut and its new Genes assigned to COGs 4,303 97.46 approach proposed to describe the new species of microor- *The total is based on either the size of the genome in base pairs ganism based on a phenotypic characterization and genomic or the total number of protein coding genes in the annotated description. Indeed, as this work demonstrates, halophilic genome. bacteria have long been overlooked in the human gut, and 256 KHELAIFIA ET AL.

Table 3. Percentage of Genes Associated with 25 General COG Functional Categories for O. jeddahense Strain S5Tand O. iheyensis strain HTE83 O. jeddahense O. iheyensis Difference Code Description % of total % of total (in %) J Translation, ribosomal structure and biogenesis 4.43 5.15 -0.72 A RNA processing and modification 0.1 0.1 0 K Transcription 9.2 6.88 2.32 L Replication, recombination and repair 3.97 4.32 -0.35 B Chromatin structure and dynamics 0.21 0.13 0.08 D Cell cycle control, cell division, chromosome partitioning 0.96 1.45 -0.49 Y Nuclear structure 0 0.03 -0.03 V Defense mechanisms 2.17 1.66 0.51 T Signal transduction mechanisms 3.99 4.11 -0.12 M Cell wall/membrane biogenesis 4.25 4.99 -0.74 N Cell motility 1.66 1.99 -0.33 Z Cytoskeleton 0.02 0.16 -0.14 w Extracellular structures 0 0 0 U Intracellular trafficking and secretion, and vesicular transport 1.68 1.97 -0.29 O Post-translational modification, protein turnover, chaperones 2.59 3.26 -0.67 C Energy production and conversion 5.28 5.31 -0.03 G Carbohydrate transport and metabolism 8.54 7.07 1.47 E Amino acid transport and metabolism 8.57 8.54 0.03 F Nucleotide transport and metabolism 2.29 2.54 -0.25 H Coenzyme transport and metabolism 3.38 3.44 -0.06 I Lipid transport and metabolism 2.83 3.05 -0.22 P Inorganic ion transport and metabolism 5.88 5.95 -0.07 Q Secondary metabolites biosynthesis, transport and catabolism 2.34 1.84 0.5 R General function prediction only 12.83 12.6 0.23 S Function unknown 12.81 13.46 -0.65 their role in the human health and disease has not yet been intestinal microbiota remains partially unknown and its di- investigated. Culturomics is so far the only approach that has versity has yet to be fully explored. looked at halophilic organisms colonizing the human gut. It should be noted that this work does not intend to describe the Description of Oceanobacillus jeddahense sp. nov. medical or biotechnological importance of this new bacteri- Oceanobacillus jeddahense strain S5T (jed.dah..en’.se um; it aims to extend the repertoire of microorganisms that L.gen. neutr. n.; from Jeddah, Saudi Arabia, where the colonize colonise the human gut in order to better understand specimen was obtained from an obese Saudi patient). Growth the composition of its microflora. of O. jeddahense strain S5T occurred between 15 to 45C (optimal growth at 37C), in aerobic conditions with a pH Conclusion ranging from pH 5 to 9 (optimum pH 7.5). It required NaCl Based on phenotypic, phylogenetic, and genomic analyses, for growth, ranging from 0.5% to 20% (optimum at 10%). we formally propose the creation of Oceanobacillus jedda- Colonies of O. jeddahense strain S5T were circular, grayish, hense sp. nov., represented here by the type strain S5T. The shiny, and smooth, with a diameter of between 2 and 5 mm. strain has been isolated from a stool specimen of a Saudi male Cells are Gram-positive and rod-shaped, measuring 2 to 4 lm individual suffering from morbid obesity. Several other in length and 0.5 lm in diameter. They were motile because bacterial species as yet undescribed were also isolated from of polar flagella, and spore-forming. different stool specimens as part of the culturomics study The O. jeddahense strain S5T exhibited catalase and oxi- using different culture conditions, suggesting that the human dase activities. Using an Api ZYM strip and Api 20 NE,

Table 4. Numbers of Orthologous Proteins Shared Between Genomes (Below Diagonal), AGIOS Values (Above Diagonal), and Numbers of Proteins per Genome (bold numbers) Halobacillus O. jeddahense O. iheyensis O. kimchii O. massiliensis halophilus O. jeddahense 5,109 69.85 69.85 67.70 63.67 O. iheyensis 2,293 3,500 89.62 68.84 63.86 O. kimchii 2,321 2,553 3,798 68.83 63.86 O. massiliensis 1,934 2,076 2,103 3,444 64.39 Halobacillus halophilus 1,753 1,878 1,900 1,714 4,090 CULTUROMICS OF OCEANOBACILLUS JEDDAHENSE 257 negative reactions were observed for alkaline phosphatase, b- that reduces an indigo dye. Int J Syst Evol Microbiol 63, galactosidase, N-acetyl-b-glucosaminidase, and urease, and 1437–1442. positive reactions were observed for nitrate reduction. Using Hyatt D, Chen GL, Locascio PF, Land ML, Larimer FW, and API 50CH strip and Api 20E, negative reactions were ob- Hauser LJ (2010). Prodigal: Prokaryotic gene recognition and tained for D-lactose, L-arabinose, D-galactose and D-ribose translation initiation site identification. BMC Bioinformatics and positive reactions were obtained for D-glucose, D- 11, 119. fructose, D-mannose, D-mannitol, D-maltose and D-sucrose. Kim YG, Choi DH, Hyun S, and Cho BC. (2007). Oceanoba- O. jeddahense strain S5T was susceptible to doxycycline, cillus profundus sp. nov., isolated from a deep-sea sediment rifampicin, vancomycin, nitrofurantoin, amoxicillin, eryth- core. Int J Syst Evol Microbiol 57, 409– 413. romycin, ampicillin, ceftriaxone, ciprofloxacin, gentamicin, Khelaifia S, Croce O, Lagier JC, et al. (2015). Non-contiguous Virgibacillus penicillin, and imipenem, but resistant to trimethoprim/ finished genome sequence and description of massiliensis sp. nov., a moderately halophilic bacterium sulfamethoxazole and metronidazole. The G+C% content isolated from human gut. New Microbes New Infect 8, of the genome is 37.24%. The 16S rRNA and genome se- 78–88. quences are deposited in GenBank under accession numbers Kokcha S, Mishra AK, Lagier JC, et al. (2012). Non-contiguous HG931338 and CCDM010000001.1 respectively. The habi- finished genome sequence and description of Bacillus timo- tat of the microorganism is the human digestive tract. The T nensis sp. nov. Stand Genomic Sci 6, 346–355. type strain of O. jeddahense S5 ( = CSUR P1091 = DSM Lagesen K, Hallin P, Rodland EA, Staerfeldt HH, Rognes T, 28586) was isolated from a stool specimen of a Saudi male and Ussery DW. (2007). RNAmmer: consistent and rapid suffering from morbid obesity. annotation of ribosomal RNA genes. Nucleic Acids Res 35, 3100–3108. Acknowledgments Lagier JC, Armougom F, Million M, et al. (2012a). Microbial culturomics: Paradigm shift in the human gut microbiome This project was funded by the Deanship of Scientific study. Clin Microbiol Infect 18, 1185–1193. Research (DSR), King Abdulaziz University, Jeddah, under Lagier JC, El Karkouri K, Nguyen TT, Armougom F, Raoult D, grant no. (3-140-1434-HiCi). The authors, therefore, ac- and Fournier PE. (2012b). Non-contiguous finished ge- knowledge DSR technical and financial support with thanks. nome sequence and description of Anaerococcus senegalensis This study was also funded by the ‘‘Fondation Me´diterraneé sp. nov. Stand Genomic Sci 6, 116–125. Infection’’. Laslett D, and Canback B. (2004). ARAGORN, a program to detect tRNA genes and tmRNA genes in nucleotide se- Author Disclosure Statement quences. Nucleic Acids Res 32, 11–16. Lee JS, Lim JM, Lee KC, Lee JC, Park YH, and Kim CJ.

The authors declare that there is no financial conflict of Virgibacillus koreensis interest. (2006). sp. nov., a novel bacterium from salt field, and transfer of Virgibacillus picturae to the genus Oceanobacillus as Oceanobacillus picturae comb. nov. References with emended descriptions. Int J Syst Evol Microbiol 56, Bankevich A, Nurk S, Antipov D, et al. (2012). SPAdes: A new 251–257. genome assembly algorithm and its applications to single-cell Lee SY, Oh TK, Kim W, and Yoon JH. (2010). Oceanobacillus sequencing. J Comput Biol 19, 455–477. locisalsi sp. nov., isolated from a marine solar saltern. Int J Boerzer M, and Pirovano W. (2012). Toward almost closed Syst Evol Microbiol 60, 2758–2762. genomes with GapFiller. Genome Biol 13, R56. Lohse M, Bolger AM, Nagel A, et al. (2012). RobiNA: A user- Boetzer M, Henkel CV, Jansen HJ, Butler D, and Pirovano W. friendly, integrated software solution for RNA-Seq-based (2011). Scaffolding pre-assembled contigs using SSPACE. transcriptomics. Nucleic Acids Res 40, W622–627. Bioinformatics 27, 578–579. Lu J, Nogi Y, and Takami H. (2001). Oceanobacillus iheyensis Camacho C, Coulouris G, Avagyan V, et al. (2009). BLAST+: gen. nov., sp. nov., a deep-sea extremely halotolerant and Architecture and applications. BMC Bioinformatics 10, 421. alkaliphilic species isolated from a depth of 1050 m on the Eddy SR. (2011). Accelerated profile HMM searches. PLoS Iheya Ridge. FEMS Microbiol Lett 205, 291–297. Comp Biol 7(10), e1002195. Mishra AK, Lagier JC, Robert C, Raoult D, and Fournier PE. Gao S, Sung WK, and Nagarajan N. (2011). Opera: Re- (2012). Non-contiguous finished genome sequence and de- constructing optimal genomic scaffolds with high-throughput scription of Clostridium senegalense sp. nov. Stand Genomic paired-end sequences. J Comput Biol 18, 1681–1691. Sci 6, 386–395. Griffiths-Jones S, Bateman A, Marshall M, Khanna A, and Eddy Mourembou G, Yasir M, Azhar EI, et al. (2015). Rise of SR. (2003). Rfam: An RNA family database. Nucleic Acids Microbial Culturomics: Non-contiguous finished genome se- Res 31, 439–441. quence and description of Beduini massiliensis gen. nov., Heyrman J, Logan NA, Busse HJ, et al. (2003). Virgibacillus sp. nov. OMICS 19, 766–776. carmonensis sp. nov., Virgibacillus necropolis sp. nov. and Nam JH, Bae W, and Lee DH. (2008). Oceanobacillus caeni Virgibacillus picturae sp. nov., three novel species isolated sp. nov., isolated from a Bacillus-dominated wastewater from deteriorated mural paintings, transfer of the species of treatment system in Korea. Int J Syst Evol Microbiol 58, the genus Salibacillus to Virgibacillus,asVirgibacillus 1109–1113. marismortui comb. nov. and Virgibacillus salexigens comb. Namwong S, Tanasupawat S, Lee KC, and Lee JS. (2009). nov., and amended description of the genus Virgibacillus. Int Oceanobacillus kapialis sp. nov., from fermented shrimp J Syst Evol Microbiol 53, 501–511. paste in Thailand. Int J Syst Evol Microbiol 59, 2254–2259. Hirota K, Aino K, Nodasaka Y, and Yumoto I. (2013). Ocea- Nawrocki EP, Kolbe DL, and Eddy SR. (2009). Infernal 1.0: nobacillus indicireducens sp. nov., a facultative alkaliphile inference of RNA alignments. Bioinformatics 25, 1335–1337. 258 KHELAIFIA ET AL.

Nurk S, Bankevich A, Antipov D, et al. (2013). Assembling Wayne LG, Brenner DJ, Colwell RR, et al. (1987). Report of single-cell genomes and mini-metagenomes from chimeric the ad hoc committee on reconciliation of approaches to MDA products. J Comput Biol 20, 714–737. bacterial systematics. Int J Syst Microbiol 37, 463–464. Punta M, Coggill PC, Eberhardt RY, et al. (2012). The Pfam Welker M, and Moore ER. (2011). Applications of whole- protein families database. Nucleic Acids Res 40, D290–D301. cell matrix-assisted laser- desorption/ionization time-of-flight Raats D, and Halpern M. (2007). Oceanobacillus chironomi mass spectrometry in systematic microbiology. Syst Appl sp. nov., a halotolerant and facultatively alkaliphilic species Microbiol 34, 2–11. isolated from a chironomid egg mass. Int J Syst Evol Mi- Whon TW, Jung MJ, Roh SW, Nam YD, Park EJ, Shin KS, and crobiol 57, 255–259. Bae JW. (2010). Oceanobacillus kimchii sp. nov. isolated Ramasamy D, Mishra AK, Lagier JC, et al. (2014). A poly- from a traditional Korean fermented food. J Microbiol 48, phasic strategy incorporating genomic data for the taxonomic 862–866. description of new bacterial species. Int J Syst Evol Microbiol Yang JY, Huo YY, Xu XW, Meng FX, Wu M, and Wang CS. 64, 384-3-91. (2010). Oceanobacillus neutriphilus sp. nov., isolated from Romano I, Lama L, Nicolaus B, Poli A, Gambacorta A, and activated sludge in a bioreactor. Int J Syst Evol Microbiol, 60, Giordano A. (2006). Oceanobacillus oncorhynchi subsp. in- 2409–2414. caldanensis subsp. nov, an alkalitolerant halophile isolated Yumoto I, Hirota K, Nodasaka Y, and Nakajima K. (2005). from an algal mat collected from a sulfurous spring in Oceanobacillus oncorhynchi sp. nov., a halotolerant obligate Campania (Italy), and emended description of Oceanoba- alkaliphile isolated from the skin of a rainbow trout (On- cillus oncorhynchi. Int J Syst Evol Microbiol 56, 805–810. corhynchus mykiss), and emended description of the genus Rossello-Mora R. (2006). DNA-DNA reassociation methods Oceanobacillus. Int J Syst Evol Microbiol 55, 1521–1524. applied to microbial taxonomy and their critical evaluation. In: Molecular Identification, Systematics, and Population Structure of Prokaryotes. Stackebrandt E, ed. Springer, Address correspondence to: Berlin. Prof. Didier Raoult Roux V, Million M, Robert C, Magne A, and Raoult D. (2013). URMITE, CNRS (UMR 7278), IRD (198) Non-contiguous finished genome sequence and description of INSERM (U1095), AMU (UM63) Oceanobacillus massiliensis sp. nov. Stand Genomic Sci 9, Faculte´ de Me´decine 370–384. Aix-Marseille Universite´ Seck E, Rathored J, Khelaifia S, et al. (2015). Virgibacillus 27 Boulevard Jean Moulin senegalensis sp. nov., a new moderately halophilic bacte- 13385 Marseille Cedex 5 rium isolated from human gut. New Microbes New Infect 8, France 116–126 Seng P, Drancourt M, Gouriet F, La Scola B, Fournier PE, E-mail: [email protected] Rolain JM, and Raoult D. (2009). Ongoing revolution in bacteriology: Routine identification of bacteria by matrix- assisted laser desorption ionization time-of-flight mass spec- Abbreviations trometry. Clin Infect Dis 49, 543–551. CSUR ¼ Collection de Souches de l’Unite´ Stackebrandt E, and Ebers J. (2006). Taxonomic parameters des Rickettsies revisited: Tarnished gold standards. Microbiol Today 33, DSM ¼ Deutsche Sammlung von 152–155. Mikroorganismen Tatusov RL, Galperin MY, Natale DA, and Koonin EV. (2000). MALDI-TOF MS ¼ Matrix-assisted laser-desorption/ The COG database: A tool for genome-scale analysis of ionization time-of-flight mass protein functions and evolution. Nucleic Acids Res 28:33–36. spectrometry The UniProt Consortium. (2011). Ongoing and future devel- TE buffer ¼ Tris-EDTA buffer opments at the Universal Protein Resource. Nucleic Acids SDS ¼ sodium dodecyl sulfate Res 39, D214–D219. URMITE ¼ Unite´ de Recherche sur les Tominaga T, An SY, Oyaizu H, and Yokota A. (2009). Ocea- Maladies Infectieuses et nobacillus soja sp. nov. isolated from soy sauce production Tropicales Emergentes equipment in Japan. J Gen Appl Microbiol 55, 225–232.