Genomic and phenotypic description of Streptococcus resistens sp. nov., Streptococcus buccae sp. nov. and Streptococcus mediterraneus sp. nov., three new members of the Streptococcus genus isolated from the human oral cavity

Phenotypic, phylogenetic and genomic studies were carried out on three unidentied Gram-stain positive, facultative anaerobic, and cocci-shaped bacteria isolated from the human oral cavity. The 16S rRNA gene of strains Marseille-P5794 T , Marseille-P6264 T and Marseille-P7376 T exhibited a sequence identity of 99,41%, 99.67% and 97.88%, respectively with Streptococcus cristatus, their closest phylogenetic relative with standing in nomenclature. Moreover, the rpoB gene sequence of strains Marseille-P5794 T and Marseille-P6264 T shared a similarity level with 96.1%, and 95.9% with Streptococcus cristatus whereas strain Marseille-P7376 T shared a 93.98% identity with Streptococcus sanguinis. Whole genome comparison of strains Marseille-P5794 T , Marseille-P6264 T and Marseille-P7376 T with their phylogenetic neighbours were under the threshold values set to dene new species using digital DNA-DNA hybridization and Orthologous Average Nucleotide Identity. The taxonogenomics analysis thus allowed the classication of these strains as new species within the Streptococcus genus named Streptoccocus resistens sp. nov. Strain Marseille-P5794 T (=CSUR P5794 = CECT9902), Streptococcus buccae sp. nov. Strain Marseille-P6264 T (=CSUR P6264 = CECT9910) and Streptococcus mediterraneus sp. nov. Strain Marseille-P7376 (=CSUR P7376 = CECT30035).


Introduction
The genus Streptococcus currently consists of 108 species and 27 subspecies according the list of prokaryotic names with standing in nomenclature (Parte et al. 2020). Members of this genus are Grampositive cocci, forming relatively long chains, facultative anaerobes that never possess a catalase (Facklam 2002). Streptococcus species are mostly found in the oral cavities of humans and animals (Facklam 2002;Shinozaki-Kuwahara et al. 2011;Bakour et al. 2016). The taxonomic classi cation of bacterial species belonging to the Streptococcus genus remains a challenge, as a result of a high genetic and phenotypic similarity between several Streptococcus species particularly when solely considering the 16S rRNA gene sequence which is the gold standard to discriminate bacterial species (Arbique et al. 2004;Kilian et al. 2008).
Although Streptococci are commensals of the human oral cavity, members of this genus have been associated to different diseases. For example, Streptococcus agalactiae have been associated with the development of pneumonia, septicemia or, more rarely, meningitis in neonates (Krzyściak et al. 2013). In addition, studies showed that Streptococcus canis, Streptococcus equinus or Streptococcus bovis can be responsible for epizootic opportunistic infections (Cole et al. 2008). Despite the wide diversity within the Streptococcus genus, using the culturomics method, our laboratory previously isolated new members of this genus (Ricaboni et al. 2016). This high throughput culture technique consists in using various culture media and physico-chemical conditions to mimic the original environment of microorganisms and explore as exhaustively as possible a given ecosystem (Lagier et al. 2012). It relies on the Matrix Assisted Laser Desorption Ionization Mass Spectrometry (MALDI-TOF MS) technology to identify bacterial isolates expeditiously as well as on the taxonogenomics strategy, which combines phylogenetic analysis, phenotypic criteria and genomic characterization to describe new species (Ramasamy et al. 2014;Ngom et al. 2020). Here, we present a detailed description according to the taxonogenomics concept (Ramasamy et al. 2014;Fournier et al. 2015) of three putative new species for which we suggest the names Streptoccocus resistens sp. nov. strain Marseille-P5794 T , Streptococcus buccae sp. nov. strain Marseille-P6264 T and Streptococcus mediterraneus sp. nov. strain Marseille-P7376 T .

Isolation and identi cation
Three oral samples were collected from a 39-year-old French female who spontaneously recovered from HIV (Colson et al. 2020) as well as from two healthy French males aged 27-and 28-year-old. Each subject gave an informed and written consent for this study which was approved by the ethics committee of the Institut Federatif de Recherche IFR48 under agreement number 2016-011. Samples were cultured in aerobic and anaerobic blood culture bottles (bioMérieux, Marcy l'Etoile, France) supplemented with 5% of sterile rumen uid and 5% of sterile de brinated sheep blood (bioMérieux). At different incubation timepoints, ten-fold serial dilutions of the culture were inoculated on 5% de brinated sheep blood enriched Colombia agar (COS, bioMérieux, Marcy l'Etoile, France) incubated at 37°C in aerobic or anaerobic atmosphere. After an incubation for 24-48h, bacterial colonies were puri ed by subculture and identi ed using MALDI-TOF MS (Bruker Daltonics, Bremen, Germany) as previously described (Lo et al. 2015a). The unidenti ed spectra were imported into MALDI Biotyper 3.0 software (Bruker Daltonics) and analyzed against two databases, that of Bruker and the constantly updated in lab URMS database (https://www.mediterranee-infection.com/acces-ressources/base-de-donnees/urms-data-base/).

16S rRNA gene sequencing
In order to classify the isolated bacteria, the 16S rRNA gene sequencing was performed. The primer pair fD1 and rP2 (Eurogentec, Angers, France) was used to amplify the 16S rRNA gene prior to sequencing using the Big Dye® Terminator v1.1 Cycle Sequencing Kit and a 3500xL sequencer (Thermo sher, Saint-Aubin, France) as previously described (Drancourt et al. 2000). The 16S rRNA nucleotide sequences were assembled using CodonCode Aligner software (http://www.codoncode.com). The obtained sequence was matched against the GenBank database using BLASTn (2015). A sequence similarity under 98.7% and 95% with the phylogenetically closest species was used to de ne a new species and a new genus respectively (Kim et al. 2014). A phylogenetic tree based on 16S rRNA gene was built with the MEGA 7 software (Edgar 2004) with sequences aligned using the Muscle software (Kumar et al. 2016).

Phenotypic and biochemical characterization
Optimal growth condition. The optimal growth of the studied strains was determined using COS agar. Several atmospheres namely aerobic, anaerobic and microaerophilic, were tested and generated using anaerobic and microaerophilic generators respectively (Thermo Fisher Scienti c, Dardilly, France). Different growth temperatures (25, 28, 37, 45 and 56°C) were tested in each atmosphere. Salt tolerance was tested for these strains using Columbia agar (Thermo Fisher Scienti c Dardilly France) supplemented with 5, 10 and 15% of NaCl as well. The pH range tolerated by the described isolates was determined by testing different pH levels (6, 6.5, 7, 7.5, 8 and 8.5) in the optimal growth condition.
Biochemical properties. Oxidase (Becton Dickinson) and catalase assays (bioMérieux) were carried out on each strain according to the manufacturer's instructions. Biochemical characteristics of these strains were further explored using API 50CH, API ZYM and API 20 Strep strips (bioMérieux) according to the manufacturer's instructions. Spore-forming capacity was explored by subjecting the described strains to a thermal shock through heating at 80°C for 20 minutes before inoculation on COS agar.
Morphological properties. Scanning microscopy electron (Hitachi TM4000) allowed the morphologic observation of the bacterial cells, as previously described (Belkacemi et al. 2019). The motility of each bacteria was studied by observing a suspension of fresh colonies in a slide and coverslip under a DM1000 photonic microscope (Leica Microsystems, Wetzlar, Germany) at a 100x magni cation.
Moreover, the cellular fatty acid methyl ester (FAME) pro le was established using gas chromatography/ mass spectrometry (GC/MS). Brie y, two samples were prepared with approximately 70 mg of bacterial biomass per tube harvested from several culture plates. FAMEs were prepared as described by Sasser (Sasser). GC/MS analyses were performed as previously described (Dione et al. 2016). FAMEs were separated using an Elite 5-MS column and monitored by mass spectrometry (Clarus 500 -SQ 8 S, Perkin Elmer, Courtaboeuf, France). A spectral database search was then performed using MS Search 2.0 operated with the Standard Reference Database 1A (NIST, Gaithersburg, USA) and the FAMEs mass spectral database (Wiley, Chichester, UK).
Antibiotic susceptibility pro le. The antibiotic susceptibility pro le of the strains was assessed using the disc diffusion method, speci cally E-test strips (Citron et al. 1991;Balouiri et al. 2016) to determine the minimal inhibitory concentration (MIC) of the following antibiotics: amoxicillin, erythromycin, clindamycin, rifampicin, oxacillin, penicillin, vancomycin, ceftriaxone, linezolid, gentamicin, fosfomycin and doxycycline.
Genome sequencing and analysis DNA extraction and sequencing. To study the genome properties of the studied strains, the genomic DNA (gDNA) was extracted using the EZ1 biorobot (Qiagen, Courtaboeuf, France) with the EZ1 DNA tissue kit prior to sequencing on the MiSeq technology (Illumina, San Diego, CA, USA) with the Nextera XT Paired end (Illumina), as previously described (Lo et al. 2015b). Then, the quality control of raw data from DNA sequencing was performed using the FastQC software (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/).
Open Reading Frames (ORF) were predicted using Prodigal (Hyatt et al. 2010) whereas protein-coding sequences were predicted by matching ORFs to the NR database using BLASTP with the following parameters: E-value of 1e -03 (1e -05 for a sequence shorter than 80 amino acids, coverage 0.7 and identity percentage of 30%). The resulting protein sequences were matched against the Clutster of Orthologous Groups (COG) database (Tatusov et al. 2000(Tatusov et al. , 2015. RNA sequences, speci cally ribosomal RNA (rRNA) and tranfer RNA (tRNA) were predicted using the RNAmmer (Lagesen et al. 2007) and tRNAScanSE (Lowe and Eddy 1997) tools respectively.
Genomic comparison. RpoB gene sequences extracted from the genomes of the studied strains were used to further classify the aforementioned isolates within the Streptococcus genus though the construction of a phylogenetic tree based built with the Muscle (Kumar et al. 2016) and MEGA 7 softwares (Edgar 2004).To assess the level of genomic similarity of the described strains with closely related species, genomic data from the following species were used: Enterococcus hirae strain R17 (CP015516.1), Streptococcus constellatus subsp. pharyngis C1050 (NC_022238.1), Streptococcus cristatus ATCC 51100 (NZ_CP050133.1) and Streptococcus gordonii strain FDAARGOS_257 (CP020450.2). The Type Strain Genome Server web server available online (https://tygs.dsmz.de/) (Meier-Kolthoff and Göker 2019) was used for genomic comparison through in silico digital DNA-DNA hybridization (dDDH) (Meier-Kolthoff et al. 2013) which allowed the estimation of the percentage of nucleotide identity between the compared genomes. In addition, the Orthologous Average Nucleotide Identity (OrthoANI) was assessed using the OAT software (Lee et al. 2016).

Strain identi cation
Strain isolation and growth conditions. Strain Marseille-P5794 T was isolated from the oral sample of a 39-year-old French female who spontaneously recovered from HIV while strain Marseille-P6264 T was isolated from the oral sample of a 28-year-old healthy French male. Similarly, strain Marseille-P7376 T was isolated from the oral sample of a 27-year-old healthy French male. Strains Marseille-P6264 T and Marseille-P7376 T were isolated after a 7-day pre-incubation in an aerobic blood culture bottle (bioMérieux) while strain Marseille-P5794 T was isolated after two days of pre-incubation in an anaerobic blood culture bottle. An identi cation attempt using MALDI-TOF MS allowed to the creation of the reference spectrum of each strain (Figure 1) which was subsequently incremented in our database that did not contain any matching spectrum.
Strain identi cation. The 16S rRNA gene of strains Marseille-P5794 T , Marseille-P6264 T and Marseille-P7376 T exhibited a 99.47%, 99.77% and 97.9 % sequence identity respectively with Streptococcus cristatus strain ATCC 51100 (Genbank accession number EU156757), the phylogenetically closest species with standing in nomenclature ( Figure 2). As the use of the 16S rRNA gene was not conclusive for the classi cation of species within the Streptococcus genus, the complete genome sequencing was performed and the rpoB gene sequences were extracted in silico as this gene was previously described as a useful marker to delineate species within this genus (Drancourt et al. 2004). Using the rpoB gene sequence, strains Marseille-P5794 T and Marseille-P6264 T shared a 96.1% and 95.9% nucleotide sequence similarity respectively with Streptococcus cristatus ATCC 51100 (NZ_AFUE00000000.1), the phylogenetically closest species with a validly published name. As for strain Marseille-P7376 T , it showed a 93.98% identity with Streptococcus sanguinis strain CGMH010 (CP040556.1, Figure 3).

Phenotypic characteristics
Optimal growth condition. The optimal growth of the three strains was recorded in aerobic condition at 37°C after a 24-hour incubation although growth was recorded in aerobic, microaerophilic and anaerobic conditions at temperatures ranging from 25 to 45°C classifying as facultative anaerobes. Growth was observed at pH 6.0, 6.5, 7.0 and 7.5 for strains Marseille-P5794 T and Marseille-P6264 T whereas strain Marseille-P7376 T grew at pH levels ranging from 7 to 9. However, no salt tolerance was exhibited by any of the strains.
Morphological characteristics. Cells from the three strains were Gram-stain positive, non-spore-forming and non-motile cocci which formed cream-colored colonies on COS agar. Scanning electron microscopy revealed that strains Marseille-P5794 T , Marseille-P6264 T and Marseille-P7376 T had a mean diameter of 550nm, 736nm and 500nm respectively (Figure 4). Fatty acids with chain saturated structures were the most commonly found in the three strains of interest. In fact, the main cellular fatty acids found were hexadecanoic acid, tetradecanoic acid and dodecanoic acid. Several unsaturated structures were detected with minor amounts (Table 3).
Biochemical characteristics. Strains Marseille-P5794 T , Marseille-P6264 T and Marseille-P7376 T possessed neither catalase nor oxidase activities. Using an API 50CH strip, all strains exhibited a positive reaction for D-sorbitol, N-acetylglucosamine, D-saccharose and D-maltose (Table S1). Using an API ZYM strip, the three strains presented high enzymatic activity for leucine arylamidase (Table S2). Finally, using an API Strep, all strains were also able to ferment for esculin iron citrate, L-arginine and D-ribose (Table S3). The main phenotypic characteristics of these strains are summarized in Table 1.

Discussion
The taxonogenomics analysis performed on these putative new species revealed a close taxonomic distance from the Mitis group of the Streptococcus genus (Kawamura et al. 1995;Kilian et al. 2014).
Consequently, the high similarity observed between the 16S rRNA gene sequences warranted the use of the rpoB housekeeping gene to discriminate these strains (Drancourt et al. 2004). Based on their phylogenetic position as well as their respective percentage of similarity with the rpoB gene of the closest validly published species, these three strains were classi ed as new members within the Streptococcus genus. In addition, the highest dDDH value obtained by comparing the genomes of these strains was 59.2% shared between S. cristatus and strain Marseille-P5794 T (Table 4). This value was below the recommended threshold of 70% to describe a new bacterial species [28] demonstrating that all the genomic sequences compared here were taxonomically distinct. Similar dDDH values were observed in the description of other novel Streptococcus spp. (Ricaboni et al. 2016;Martínez-Lamas et al. 2020). Consequently, based on their phylogenetic, phenotypic, biochemical, and genomic characteristics, we classify strains Marseille-P6264 T , Marseille-P5794 T and Marseille-P7376 T within the Streptococcus genus, as novel species distinct from their closest phylogenetic neighbours. Therefore, we formally Streptococcus resistens (re.sis'tens L. part. adj. resistens, referring to the HIV resistance exhibited by the patient from whom the strain was isolated). Cells are facultative anaerobes and Gram-stain positive cocci with a mean diameter of 550 nm. They are non-motile and non-spore-forming, catalase and oxidase negative. Strain Marseille-P5794 T is mesophilic with an optimal growth at 37°C. Colonies are circular and cream-colored. S. resistens is neither halotolerant nor halophilic but grows on agar between pH 6 and pH 7.5. Using an API ZYM strip, a positive reaction is observed for esterase (C4), leucine arylamidase, valine arylamidase, cystine arylamidase, acid phosphatase, naphtalo-AS-BI-phosphohydrolase and βgalactosodase. Using an API 50 CH strip, reaction is positive for salicine, D-maltose, D-lactose (bovine origin), D-saccharose (sucrose), D-trehalose, D-sorbitol, N-acetylglucosamine and arbutine. Finallys, using API 20 Strep, S.resistens can ferment esculin iron citrate, L-arginine, D-ribose, D-lactose (bovin origin), Dtréhalose, Inuline and Glycogen. C 16:0 (39.8%), C 14:0 (29.5%) and C 12:0 (16.9%) were the major fatty acids.
The genome of S. resistens was 2.07 Mbp long with 42.3 % of G+C content. Strain Marseille-P5794 was isolated from an oral sample provided by a 39-year-old French female who spontaneously recovered from HIV. The 16S rRNA gene and genome sequences are available in GenBank under accession numbers LR597666 and CABPTQ010000000 respectively. This strain is deposited in the "Collection de Souches de l'Unité des Rickettsies" (CSUR) and "Spanish Type Culture Collection" (STCC) under number CSURP5794 and CECT 9902 respectively.
Description of Streptococcus buccae sp. nov. Streptococcus buccae sp. nov. (buc.ca'e L. gen. fem. n. buccae, of the mouth, referring to the habitat where the strain was isolated). Streptococcus buccae consists of Gram-stain positive cocci which are facultative anaerobes. Cells are non-motile and non-spore-forming with a mean diameter of 736 nm. Catalase and oxidase activities are negative. S. buccae was mesophilic with an optimal growth temperature at 37°C and was able to tolerate pH levels ranging from 6 to 7.5. Using an API ZYM strip, a positive reaction was observed for leucine arylamidase, valine arylamidase, cystine arylamidase, naphthol-AS-BI-phosphohydrolase, α-galactosidase and β-galactosodase. Using an API 50 CH strip, positive reactions were obtained for D-maltose, D-saccharose (sucrose), D-galactose, D-ra nose, Dglucose, D-fructose, D-mannose, D-sorbitol and N-acetylglucosamine. Finally, using API 20 Strep, S.buccae can also ferment Sodium pyruvate, esculin iron citrate, 6-bromo-2-naphthyl-α Dgalactopyranoside, L-arginine, D-ribose, D-lactose (bovin origin), starch and glycogen. The major fatty acids were C 16:0 (45.1%), C 14:0 (25.7%) and C 12:0 (14.7%). Its genome was 2.04 Mbp long with 42.5 % of G+C content. The accession numbers of 16S rRNA and genome sequences are LR699782 and CABPTO010000000 respectively in the Genbank database. The type strain Marseille-P6264 was isolated from the oral sample of a 28-year-old healthy French male and is available in the CSUR and STCC collections under number CSURP6264 and CECT 9910 respectively.
Description of Streptococcus mediterraneus sp. nov. Streptococcus mediterraneus sp. nov. (me.di.ter.ra'ne.us L. masc. adj. mediterraneus, from the Mediterranean Sea, which borders Marseille, where the strain was isolated). Cells are Gram-stain positive cocci which were facultatively anaerobic, non motile and non-spore-forming with a mean average diameter of 500 nm. Catalase and oxidase activities are negative. S. mediterraneus grows optimally on COS agar after 24 hours in aerobic condition at 37°C at pH 7.5. Using an API ZYM strip, a positive reaction is observed only for leucine arylamidase. Using an API 50 CH strip, positive reactions were obtained with N-acetylglucosamine, D-galactose, Dglucose, D-fructose, D-mannose, D-mannitol, D-sorbitol, arbutin, esculin, salicine, D-cellobiose, D-maltose, D-lactose, D-saccharose and D-trehalose. In addition, using API 20 Strep, S.mediterraneus can also ferment esculin iron citrate, 6-bromo-2-naphthyl-α D-galactopyranoside, L-arginine and D-ribose. The major fatty acids were C 16:0 (56 %), C 14:0 (19 %) and C 12:0 (7.2%). Its genome was 2.29 Mbp long with 41.9 % of G+C content. The 16S rRNA gene and genome sequences are available in the Genbank database under number LR699798 and CABPTS010000000, respectively. The type strain Marseille-P7376 was isolated from the oral sample of a 27-year-old healthy French male and it is deposited in the CSUR and STCC collections under number CSURP7376 and CECT 30035 respectively.

Declarations
Funding This study was supported by the Méditerranée Infection Foundation and the National Research Agency under the program « Investissements d'avenir », reference ANR-10-IAHU-03. This work was supported by the Région Provence-Alpes-Côte d'Azur and European funding FEDER PRIMI Con icts of interest/Competing interests None to declare

Availability of data and material
The datasets presented in this study can be found in online repositories. Associated accession numbers are cited in the manuscript.

Code availability
The accession numbers can be found in this article or in the supplementary Material.

Consent to participate
Each subject gave an informed and written consent for this study.

Consent for publication
All authors had access to this publication and approved its submission.