Staphylococcus ursi sp. nov., a new member of the ‘Staphylococcus intermedius group’ isolated from healthy black bears

Six Staphylococcus strains were isolated from healthy black bears (Ursus americanus) in the Great Smoky Mountains National Park, Tennessee, USA. Phylogenetic analysis based on complete genome, 16S rRNA, dnaJ, hsp60, rpoB and sodA genes, and MALDI-TOF-MS main spectral profiles revealed that the strains belonged to one species and showed the closest relatedness to members of the ‘ Staphylococcus intermedius group’ (SIG), which include Staphylococcus intermedius , Staphylococcus pseudintermedius, Staphylococcus delphini and Staphyloccoccus cornubiensis. The strains were positive in SIG-specific and negative in individual species-specific PCR assays for the nuc gene. The strains can be differentiated from the other SIG species by the absence of sucrose fermentation, from S. intermedius DSM 20373T, S. pseudintermedius CCUG 49543T and S. cornubiensis DSM 105366T by the absence of methyl β-d-glucopyranoside fermentation and from S. delphini DSM 20771T by fermentation of trehalose. DNA relatedness of the type strain MI 10-1553T with the type strains of S. delphini , S. pseudintermedius , S. intermedius and S. cornubiensis was ≤48.2 % by digital DNA–DNA hybridization and ≤92.3 % by average nucleotide identity calculations. Iso-C15:0, anteiso-C15 : 0 and anteiso-C17 : 0 were the most common fatty acids. Polar lipids consisted of phosphadidylglycerols, phospholipids, glycolipid, diphosphatidylglycerol and aminophospholipid. Cell-wall peptidoglycan was of type A3α l-Lys-Gly3 (Ser; similar to A11.2 and A11.3). The respiratory quinone belonged to menaquinone 7 (MK-7). The G+C content of MI 10-1553T was 39.3 mol%. The isolated strains represent a novel species of the genus Staphylococcus , for which we propose the name Staphylococcus ursi sp. nov. The type strain is MI 10-1553T (=ATCC TSD-55T=CCOS 1900T).


INTRODUCTION
Over a number of years, bacteria that share colony morphotypes and overlapping phenotypic characteristics with Staphylococcus aureus were taxonomically segregated and assigned to new species, e.g. Staphylococcus intermedius [1], Staphylococcus delphini [2], Staphylococcus pseudintermedius [3], and Staphylococcus cornubiensis [4]. The above species are commonly referred to as the 'Staphylococcus intermedius group' (SIG). Phenotypic traits and 16S rRNA gene sequencing do not readily differentiate between these closely related species [5]. SIG members are usually associated with various animal hosts; however, their presence in human infections may be underestimated due to the potential for misidentification as S. aureus [6][7][8]. Such misidentification can negatively affect interpretation of susceptibility test results and lead to inappropriate antimicrobial therapy [8,9]. As with S. aureus in people, S. pseudintermedius has been well characterized as an opportunistic pathogen in dogs and the frequency of infections caused by methicillin-and multidrug-resistant strains has been increasing in recent years [10]. Overall, little is known about the host range and population distribution of species within the SIG. It has been suggested, for example, that strains identified as S. pseudintermedius by traditional means from hosts other than dogs should be more thoroughly investigated by phylogenetic analysis before reporting [5]. The objectives of this work were to seek SIG isolates from healthy wild bears and to confirm their species identity by polyphasic taxonomic characterization. Such strains will be useful for future comparisons to existing collections for the purpose of examining their diversity and potential to exchange virulence and resistance traits.
During a surveillance study for the presence of staphylococci, cutaneous samples for culture were obtained from 15 healthy black bears (Ursus americanus) that had been captured and tranquilized for relocation, or, in one case, hit by a car and euthanized, in the Great Smoky Mountains National Park, Tennessee, USA, during 2010. All animal handling and sample collections were performed by trained and certified National Park Service wildlife biologists. Culture swabs (Becton Dickinson) were used to collect samples from the nasal cavity and a minimum of two additional pre-defined sites (nasal, oral lip folds, inguinal skin or external ear) from each animal. Swabs in Amies medium (Becton Dickinson) were kept cool while in the field and transported for processing within 24-48 h. Swabs were inoculated onto colistin-nalidixic acid agar (Becton Dickinson) containing 5 % sheep blood. The plates were incubated at 35 °C for 24-48 h. A representative of all colony types resembling those of staphylococci and that were catalase-positive, Gram-stain-positive cocci were inoculated onto trypticase soy agar containing 5 % sheep blood (TSA-SB; Becton Dickinson) and incubated at 35 °C for 24 h; subsequently, cell lysates were used as template DNA for universal bacterial 16S rRNA gene PCR followed by direct Sanger sequencing to obtain partial nucleotide sequences as described previously [11]. The 16S rRNA gene sequences of the strains were compared to sequences in the GenBank database using the blastn algorithm [12]. Strains resembling members of the SIG were further screened by partial thermonuclease gene (nuc) PCR using previously described group-specific [13] and species-specific [14] methods. Of the 108 strains screened, 61 were provisionally identified as belonging to the genus Staphylococcus and were distributed among 10 different species or groups of closely related species. Six strains, found in cultures from sample sites of five animals, had large (>5 mm), smooth, off-white colony morphologies resembling those of SIG members, displayed α hemolysis on sheep blood agar and were catalase-positive, Gram-stain-positive cocci. The strain designations and their corresponding animal, gender and sample site from which they were isolated were as follows: MI 10-1549, bear 1, female, nasal; MI 10-1553, bear 2, female, oral lip fold; MI 10-1558, bear 3, female, oral lip fold; MI 10-1605, bear 4, male, inguinal skin; MI 10-1708, bear 5, female, oral lip fold; MI 10-1710, bear 5, female, external ear canal. The six strains contained 16S rRNA gene sequences that differed from those of known Staphylococcus species clustering with strains of the S. hyicus-intermedius species group [15] and appeared to be most closely related to members of the SIG (Fig. 1). The strains could not be assigned to any of the known species of the SIG using a species-specific nuc PCR assay which is the currently reliable and recommended method to rapidly identify members of this group [14]. The strains were, however, positive in the SIG-specific [13] PCR assay for the nuc gene. These six strains were not clonally related as demonstrated by pulsed-field gel electrophoresis (PFGE; Fig. S1, available in the online version of this article) and were selected for further identification and characterization. PFGE was carried out using DNA digested with Cfr9I as described previously [16]. PFGE was run in 1 % agarose gel in 0.5×TBE on a Bio-Rad CHEF-DR III system for 18 h with pulse from 0.5 to 15 s and for 5 h with pulse from 20 to 25 s and at 14 °C and 5.6 V cm −1 .
Whole-genome sequencing was performed for strain MI 10-1553 T using the MinION R.9.4.1 flow cell (Oxford Nanopore Technology), Novaseq6000 (Illumina), Ion Torrent sequencing technology (Thermo Fisher Scientific) and assembled with Unicycler (version 0.4.4). The resulting assembled sequence consisted of a 2 896 994 bp chromosome which was annotated using the NCBI Prokaryotic Genome Annotation Pipeline (GenBank acc. no. CP048279). The genome sequence was used to perform digital DNA-DNA hybridization (dDDH) and average nucleotide identity (ANI) calculations, whole genome comparative analysis, and to obtain DNA sequences for the 16S rRNA, nuc, dnaJ, hsp60, rpoB and sodA genes as well as the operon coding for teichoic acid biosynthesis. The 16S rRNA gene sequence of strain MI 10-1553 T (GenBank acc. no. KY019172) clustered into the phylogenetic branch of the SIG (Fig. 1) and shared ≤99.7 % identity to the sequence of the other SIG type strains using the lalign program [17] ( Table 1). The 16S rRNA of the novel species shared 99.2 % identity with Staphylococcus schleiferi subsp. coagulans ATCC 49545 T , which is the next closest Staphylococcus species to those of the SIG group, was provisionally also isolated from some of the bear samples in this study and that is commonly found in body sites occupied by S. pseudintermedius in dogs [10]. Specific regions of the dnaJ [18], hsp60 [19], rpoB [20] and sodA [21] genes, which have been frequently used for molecular identification of staphylococci, were also amplified by PCR from the S. ursi strains using primers and conditions presented in Table S1 and sequenced. Alignment of these regions with those of the S. hyicus-intermedius group showed that the six strains were closely related but differed from those of SIG members (  (Table 1). Since PCR assay targeting the nuc region is the currently recommended method to rapidly identify members of the SIG [14], specific primers nuc-S. ursi-F (5′-GCAG ACAC CTCG AAAT CAATGTG) and nuc-S. ursi-R (5′-GGTA TCCC CATC TATC ACGCGT) were designed in this study to specifically amplify a 593 bp fragment of the nuc region of the S. ursi sp. nov. strains to distinguish S. ursi sp. nov. from the four type strains of other members of the SIG. PCR were performed using Taq polymerase and an initial denaturation step at 94 °C  The ANI were calculated using OrthoANIu [24], and the dDDH was determined using the Genome-to-Genome Distance Calculator (GGDC) [23]. Identity between the genetic markers was determined using the lalign program [17]; nd, No cut off defined; CI, confidential interval. The genome sequences were uploaded to the Type (Strain) Genome Server (TYGS) hosted at the Leibniz Institute DSMZ, Germany (https:// tygs. dsmz. de) for whole-genome comparative analysis and dDDH calculations [22]. All pairwise comparisons among the set of genomes were conducted using genome blast distance phylogeny (GBDP) and intergenomic distances inferred under the algorithm 'trimming' and distance formula d5 with a 100 distance replicates calculation [23]. The dDDH values and confidence intervals were calculated using the recommended settings and formula d4 of the Genome-to-Genome Distance Calculator (GGDC version 2.1) [23]. The ANI value of the genome of S. ursi MI 10-1553 T as compared to the other SIG type strains has been calculated using orthoANIu [24].
A balanced minimum evolution tree of strain MI 10-1553 T and its next-related Staphylococcus species was reconstructed using the intergenomic distances obtained from the GBDP analysis and with branch support (100 pseudo-bootstrap replicates) via fastme 2.1.4 and including subtree pruning and regrafting postprocessing [25]. The tree was visualized using iTOL (https:// itol. embl. de/) [26] (Fig. 2). Phylogenetic relationships of the novel strains and other staphylococcal species were also determined for 16S rRNA gene, sodA, hsp60, dnaJ and rpoB, and analysing sequences using BioNumerics 7.5 (Applied Maths). The trees were generated by the neighbour-joining method [multiple alignment (OG 100 %; UG 0 %), discard unknown bases, use active zones only] using the Jukes-Cantor correction. Bootstrap values were determined from 1500 replications. The phylogenetic trees showed that the novel strains, as represented by strain MI 10-1553 T , clustered within the S. hyicus-S. intermedius group [15] and are most closely related to the SIG (Figs 1, 2 and S2).
Strain MI 10-1553 T also differed from the type strains of species within the SIG by dDDH and ANI analyses. A threshold value of 70 % DNA-DNA relatedness and of 95-96 % ANI for the definition of bacterial species was considered as recommended [27,28] [25]. The genome of M. canis KM45013 T was used as outgroup to root the tree. The tree was visualized and rooted using iTOL (https://itol.embl.de/) [26] (=CCUG 49543 T ), S. intermedius NCTC 11048 T (=CCUG 20373 T ), and S. cornubiensis NW1 T (=DSM 105366 T ), which were the closest relatives by 16S rRNA, dnaJ, hsp60, rpoB, sodA and whole genome sequence comparison ( Table 1). Fatty acids iso-C 15 : 0 , anteiso-C 15 : 0 and anteiso-C 17 : 0 were the most abundant fatty acids in strain MI 10-1553 T . With the exception of anteiso-C 17 : 0 , these fatty acids were also the most common in the other species of the SIG ( Table 2).
The peptidoglycan structure of strain MI 10-1553 T was determined as previously described using cells grown in brain heart infusion (BHI) broth (BBL, Becton Dickinson) at 37 °C for 18 h [32]. Total hydrolysates (4 N HCl, 16 h, 100 o C) of the peptidoglycan in strain MI 10-1553 T contained the amino acids Lys, Ala, Ser, Gly and Glu in molar ratios of 0.9, 1.4, 0.3, 2.8, and 1.0, respectively. The identity of amino acids was confirmed by GC/MS (GC/MS 320 Singlequad, Varian). Partial hydrolysates of the peptidoglycan (4 N HCl, 0.75 h at 100 °C), determined by 2D-TLC, revealed the presence of peptides l-Ala-d-Glu, Ala-Gly, l-Lys-Gly, d-Ala-l-Lys-Gly, oligo-Gly. The peptidoglycan type of MI 10-1553 T was A3α l-Lys-Gly 3 (Ser) and was similar to types A11.2 and A11.3 according to the system proposed by Schleifer and Kandler [33] for the characterization and representation of Respiratory lipoquinones and polar lipids were extracted from 100 mg freeze-dried cell material from a 24 h culture in a trypticase soy yeast extract medium at 37 °C using a the two-stage method described by Tindall [34,35]. Respiratory lipoquinones were extracted first followed by the polar lipids [34,35]. Respiratory lipoquinones were separated into their different classes (menaquinones and ubiquinones) by TLC on silica gel (Macherey-Nagel art. no. 805 023), using hexane/ tert-butyl methyl ether (9; 1, v/v) as solvent. UV-absorbing bands corresponding to menaquinones or ubiquinones were removed from the plate and analysed further by HPLC. This step was carried out on an LDC Analytical HPLC (Thermo Separation Products) fitted with a reversed-phase column (Machery-Nagel; 2×125 mm, 3 µm, RP18) using methanol as the eluent. Respiratory lipoquinones were detected at 269 nm. Strain MI 10-1553 T contained menaquinone 7 (MK 7).
The presence of teichoic acid in the novel species was determined by the detection within the genome of strain MI 10-1553 T of the Tar/Tag operon (TagAHGBD) which is involved in the cell-wall teichoic acid biosynthesis [37].
Matrix-assisted laser desorption/ionization time-of-flight mass spectroscopy (MALDI-TOF-MS) was performed from colonies grown on TSA-SB agar plates for 18 h at 37 °C using the ethanol-formic acid extraction method for better resolution and following the manufacturer's instructions (Application Note, MT-80, Microflex LT, Bruker Daltonics). Each of the strains generated similar ribosomal protein spectra, but no reliable identification was obtained using the manufacturer's database. Following manual inclusion of spectra of strain MI 10-1553 T to the database, all the novel bear strains were re-analysed and matched to S. ursi with score above 2.2, whereas the next closest species was S. delphini with scores underneath 1.8 (Fig. S4). It was determined that the bear strains had specific ribosomal protein spectra which were not related to any of those of the Staphylococcus species already represented in the manufacturer's database at the time of testing.
Phenotypic characterization of the novel bear strains and reference strains S. intermedius (DSM 20373 T ), Staphylococcus cornubiensis DSM 105366 T , S. delphini (DSM 20771 T ) and S. pseudintermedius (CCUG 49543 T ) was initially tested using both Vitek 2 GP card (bioMérieux) and API ID 32 Staph V3.0 (bioMérieux) system following the manufacturer's recommendations. Results obtained with Vitek 2 system (bioMérieux) were confirmed for overlapping tests with the API ID 32 STAPH system (bioMérieux) ( In addition to their similarities to S. aureus and members of the SIG, colony phenotypes of the bear strains on TSA-SB (large, smooth, off-white with alpha hemolysis) also resembled those described for S. schleiferi [41]. In this study the novel bear strains were phylogenetically most closely related to members of the SIG and S. schleiferi occupied the next closest clade (Figs. 1 and 2). The novel bear strains can be differentiated from S. schleiferi by the absence of acetoin production.
Based on the combined phenotypic and genotypic characteristics of the bear strains, these strains, represented by strain MI 10-1553 T , meet criteria for being assigned to a new Staphylococcus species, for which we propose the name Staphylococcus ursi sp. nov.