Comparative genomic analysis and characterization of Staphylococcus sp. AOAB, isolated from a notoriously invasive Mnemiopsis leidyi gut revealed multiple antibiotic resistance determinants

Here, we describe the isolation and characterization of a coagulase-negative, vancomycin and oxacillin-susceptible novel bacterium of the genus Staphylococcus. Staphylococcus sp. strain AOAB was isolated from the stomodeum (gut) of the Mnemiopsis leidyi from Mobile Bay, Alabama USA. A polyphasic taxonomic approach comprised of phenotypic, chemotaxonomic and genotypic characteristics was used for analysis. The dominant respiratory quinone detected was MK-7 (100%). Major cellular fatty acids were anteiso-C15:0 (40.52%), anteiso-C17:0 (13.04 %), C-18:0 (11.53%) and C-20:0 (10.45%). The polar lipid profile consisted of glycolipid, phospholipid, phosphatidylglycerol and diphosphatidylglycerol. Although strain AOAB had a 16SrRNA gene sequence similarity of 99% with S. warneri SG1, S. pasteuri, S. devriesei KS-SP_60, S. lugdunensis HKU09-01, S. epidermidis RP62A, S. haemolyticus JCSC1435 and S. hominis DM 122, it was be distinguished from those species based on Multi-Locus Sequence Analysis (MLSA) using 6 marker genes (16S rRNA, hsp60, rpoB, dnaJ, sodA and tuf). MLSA revealed strain AOAB to be closely related to S. warneri SG1 and S. pasteuri SP1 but distinct from two hitherto known species. These results were confirmed by Average Nucleotide Identity (closest ANI of 84.93% and 84.58% identity against S. warneri SG1 and S. pasteuri SP1 respectively). In-silico DNA-DNA hybridization was <70% (33.1 % and 32.8% against S. warneri SG1 and S. pasteuri SP1 respectively), which further confirmed that the strain was a potential novel Staphylococcus species.


Materials and Methods
Ctenophore gut (stomodeum) samples were collected from animals from Dauphin Island Marina, Mobile Bay, Alabama. Sterile toothpicks were inserted into the gut from the oral end and gently rotated, in order to collect a mucus-rich sample without damage to the stomodeal lining. Tips containing the mucoid gut samples were transferred into 1.5 mL tubes, stored on ice and transported to the laboratory for enrichment.
Phenotypic tests were carried out following minimum recommended standards to aid in discriminating species (Freney et al., 1999) using using carbohydrate fermentation with the API CH50 systems (bioMe´rieux). Optimal growth was obtained at 37C after 24-48 hrs.
Antimicrobial susceptibility was performed using a standard disk diffusion method, following cutoff ranges as outlined by the National Committee for Clinical Laboratory Standards (NCCLS) (Miller et al., 2003). Isolates were inoculated using spread plating with 100 µL of inoculum at log phase from LB broth onto LB agar before impregnating with antibiotic discs. Zones of inhibition (ZOI) were used to determine the ability of the antibiotics to inhibit the growth. The antibiotics used are shown in Table 1. Results were interpreted using commonly accepted zone breakpoints for Staphylococcus (Howe & Andrews, 2012). Control discs were sterile 6mm discs without antibiotics.
Genomic DNA was isolated using the CTAB protocol (Andreou, 2013) with minor modifications that included the use of 0.5 mm silica beads (Biospec Products, Inc. Cat. No. 110791052z), shaken using a specialized MO BIO Vortex-Genie R 2 (MO BIO Laboratories). DNA purity was checked using a NanoDrop reader (ND-2000, NanoDrop Technologies, Wilmington, DE, USA) and precision-quantified using Qubit HS reagents (Life Technologies). The DNA template concentration was adjusted to 5ng prior for use in touchdown PCR reactions. Amplification for most of the 16S rRNA gene was achieved using the universal primers 63f (5′-CAG GCC TAA CAC ATG CAA GTC-3′) and 1387r (5′-GGG CGG WGT GTA CAA GGC-3′) as described by Suriyachadkun et al. (Suriyachadkun et al., 2009).
Contents of a 25 µL PCR mixture included 12.5 µL of EconoTaq Plus Green 2X Master Mix (Lucigen), 0.5 µL of 20 µM of forward and reverse primers, 10.5 µL water and 1 µL of 5 ng µL -1 genomic template DNA. PCR was carried out using 'touchdown' conditions: initial denaturation was at 95 °C for 5 min, followed by 20 cycles at 95 °C for 1min 61°C for 45 sec, and extension at 72 °C for 90 sec. The touchdown method was followed by another 30 cycles at 95 °C (1min) 51 °C (45 sec) 72 °C (90 sec). The final extension was done for 7 min at 72 °C.
The study also targeted a 370 bp tuf gene which is well established for Staphylococcus taxonomy (Martineau et al., 2001). Primer selection and melting temperature determination was done using The PCR reagent ratios were similar to those used in the 16S rRNA gene amplification.
Touchdown PCR conditions were as previously described (Martineau et al., 2001) with modifications that included 5 min at 95°C, 30 cycles of 30 s at 95°C, 30 s at 55°C and 45 s at 72°C, and final extension for 7 min at 72°C. Amplified PCR products were purified using QIAquick (Qiagen, Maryland, USA) purification kit following manufacturer's instructions prior to sequencing using ABI 3100 DNA Genetic Analyzer at Auburn University's Genomics & Sequencing Laboratory (GSL).
Genomic DNA of the isolate was used for library construction. We used the Agilent 2100 BioAnalyzer (Agilent Technologies, USA) to perform size fractionation and quantification of DNA. DNA was then fragmented and libraries prepared using Nextera XT (Illumina) according to manufacturer's protocol before being run on an Illumina Miseq ( for 2 x250 bp paired-end reads) sequencing platform at the Auburn University's Biological Sciences Department.
FASTA-formatted genomic sequences of closely related Staphylococcus genomes were obtained from the RefSeq database on GenBank and uploaded to PATRIC for annotation.
Protein coding genes in genomes were identified using GeneMarkS (Borodovsky & Lomsadze, 2014). Non-coding RNA prediction was achieved using RNAMMER 1.2 online server (Lagesen et al., 2007). Predictions for tRNA and tmRNA genes were done with the ARAG ORN tRNA and tmRNA prediction program (Laslett & Canback, 2004). Functional annotation of the protein gene models was achieved using multiple bioinformatic softwares including the RAST server Annotation for antibiotic resistance genes was achieved using PATRIC via BLASTP sequence homology search from the Antibiotic Resistance Genes Database (ARDB) (Liu & Pop, 2009) and the Comprehensive Antibiotic Resistance Database (CARD) (McArthur et al., 2013) databases. To remove database-based redundancy, replicates were removed.
Analysis for occurrence of metal resistance genes (MRG) was performed using BLASTX against the Antibacterial Biocide and Metal Resistance Genes Database (MRDB) database with an evalue cutoff of 0.01 (Altschul et al., 1997). To achieve this, the BaCMet experimentally confirmed database of MRDB was used (Pal, Bengtsson-Palme, Rensing, Kristiansson, & Larsson, 2014).
Using the 16S rRNA gene sequences of strain AOAB and closely related Staphylococcus representatives were structurally aligned using SSU-ALIGN v0.1.1 (Nawrocki, 2009) and used for reconstruction of neighbor-joining tree. Separate sequence alignments was done using ClustalW algorithm in MEGA7 (Kumar, Stecher, & Tamura, 2016) for unrooted neighborjoining tree (Supplementary Figure S3) . FASTA sequences of housekeeping genes from the closest species were obtained from GenBank. Concatenated DNA sequences from six marker genes (16S rRNA, tuf, sodA, dnaJ, hsp60 and rpoB) were used for MLSA. Sequences were aligned using MUSCLE (Edgar, 2004). The evolutionary history of strain AOAB was inferred by using the Maximum Likelihood method based on the Jukes-Cantor model (Jukes & Cantor, 1969). Evolutionary analyses were conducted in MEGA7 (Kumar et al., 2016).
To infer genomic distance between strain AOAB and the closest species, pairwise average nucleotide identity (ANI) was computed using IMG/M system (Varghese et al., 2015). With a species cut off set at 96%, this method has been found to be robust in delineating bacteria based on genome sequence data (M. Kim, Oh, Park, & Chun, 2014).
The presence of fatty acids, ai-C15 : 0, i-C15 : 0, i-C17 : 0, ai-C17 : 0, confirmed Genus of isolates as Figure S4). Cluster analysis of the fatty composition of our isolate based on Euclidean distance revealed that our isolate does not belong to any known species (Figure 4) as MIDI dendrogram software places same species link at about 10 Euclidian Distance (http://www.midilabs.com/fatty-acid-analysis). The presence of menaquinone  in the cytoplasmic membrane helped confirm the isolated as coagulase negative (Heß & Gallert, 2015).
Phylogenetic analysis using the 16S rRNA gene using neighbor-joining method indicated that the isolate was closely related to S. pasteuri and S warneri (Figure 2). Further phylogenetic analysis using MLSA (Figure 3) clustered the isolate as novel bacteria. Both ANI and DDH using genome sequence data confirmed delineation of the isolate as a novel species. None of the closest species met the ANI cut off of 96% or the DDH cut off of 70% (Table 3).
Whole genome sequencing yielded a total of 905,410 paired reads for strain AOAB. The PATRIC annotated genome size was 2,617,061 bp. A number of genomic features differentiated it from the two closest relatives. Unlike S. mnemiopsis, S. pasteuri was experimentally confirmed to be susceptible to kanamycin (Chesneau et al., 1993), which helped in discrimination between the two species (Table 4).
Genomic characterization of strain AOAB revealed that the isolate harbors antibiotic resistance determinants (virulence factors, antibiotic resistance and drug targets and heavy metal resistance genes) (Table 5). In addition to antibiotic resistance genes, MRG-like sequences were detected in the genome of strain AOAB. The most dominant ones were copper (22), arsenic (11) and zinc (4) metal resistance genes. S.warneri had copper (12), zinc (7) cadmium (7) and arsenic (5) as the dominant metal resistance genes and for S. pasteuri SP1, it was copper (12), zinc (7), cadmium (6) and arsenic (5), thus aiding in further discrimination. This study forms the first report of Staphylococcus isolation from the stomodeum of M. leidyi. The results also suggest that M.
leidyi, a notoriously invasive zooplankton harbors culturable but previously uncharacterized Staphylococcus species, some of which harbor antibiotic resistance genes. Future efforts will involve investigation of host-microbe interaction.

Data Availability
The Whole Genome Shotgun project has been deposited at DDBJ/ENA/GenBank under the accession LZFL00000000. The version described in this paper is version           Table Legends   Table 1: Antibiotic susceptibility: ZOI (Zone of inhibition) as measured (in mm) from the edge of disc to edge of inhibition zone (Kaushik, Kessel, Ratnayeke, & Gordon, 2015). Resistant (R) and susceptible (S) assigned based on previously defined breakpoints (Howe & Andrews, 2012). Below the breakpoints, a given isolate is resistant. Antibiotics with unrevised break points are indicated as dash (-). Experiments done in triplicate.