Description of three new Peptoniphilus species cultured in the vaginal fluid of a woman diagnosed with bacterial vaginosis: Peptoniphilus pacaensis sp. nov., Peptoniphilus raoultii sp. nov., and Peptoniphilus vaginalis sp. nov.

Abstract Three previously unidentified Gram‐positive anaerobic coccoid bacteria, strains KhD‐2T, KHD4T, and Kh‐D5T, isolated from a vaginal swab, were characterized using the taxonogenomics concept. The phylogenic analysis, phenotypic characteristics, and genotypic data presented in this report attest that these three bacteria are distinct from previously known bacterial species with standing in nomenclature and represent three new Peptoniphilus species. Strain KhD‐2T is most closely related to Peptoniphilus sp. DNF00840 and Peptoniphilus harei (99.7% and 98.2% identity, respectively); strain KHD4T to Peptoniphilus lacrimalis (96%) and strain Kh‐D5T to Peptoniphilus coxii (97.2%). Strains KhD‐2T, KHD4T, and Kh‐D5T DNA G+C contents are, respectively, 34.23%, 31.87%, and 49.38%; their major fatty acid was C16:0 (41.6%, 32.0%, and 36.4%, respectively). We propose that strains KhD‐2T (=CSUR P0125 = DSM 101742), KHD4T (=CSUR P0110 = CECT 9308), and Kh‐D5T (=CSUR P2271 = DSM 101839) be the type strains of the new species for which the names Peptoniphilus vaginalis sp. nov., Peptoniphilus raoultii sp. nov., and Peptoniphilu pacaensis sp. nov., are proposed, respectively.


| Samples and ethics
The vaginal specimen from a French 33-year-old woman with bacterial vaginosis was sampled at Hospital Nord in Marseille (France) in October 2015 using a Sigma Transwab (Medical Wire, Corsham, United Kingdom). Bacterial vaginosis was diagnosed as previously described (Menard, Fenollar, Henry, Bretelle, & Raoult, 2008). The patient had not received any antibiotic for several months. The local IFR48 ethics committee in Marseille (France) authorized the study (agreement number: 09-022). In addition, the patient gave her signed informed consent.

| Bacterial strain isolation and identification
After sampling, the specimen was preincubated in a blood culture bottle (Becton-Dickinson Diagnostics, Le Pont-de-Claix, France).
The blood culture bottle was enriched with 3 ml of sheep blood (bioMérieux, Marcy l'Etoile, France) and 4 ml of rumen fluid, filtersterilized through a 0.2 μm pore filter (Thermo Fisher Scientific, Villebon-sur-Yvette, France). Various preincubation periods (1, 3, 7, 10, 15, 20, and 30 days) were tested. Then, 50 μl of the supernatant were inoculated on both Colistin-nalidixic acid (CNA) used for selective enrichment of Gram-positive bacteria and trypticase soy agar plates used for cultivation of nonfastidious and fastidious microorganisms (both BD Diagnostics), and then incubated for 4 days under anaerobic conditions at 37°C. Isolated colonies were purified and subsequently identified by matrix-assisted laser-desorption/ ionization time-of-flight (MALDI-TOF) mass spectrometry with a Microflex spectrometer (Bruker, Leipzig, Germany) that compared the new spectra with those present in the library (Bruker database and URMITE database, constantly updated), as previously reported (Seng et al., 2009). If the score was >1.99, the bacterium was considered as identified at the genus level (score between 2.0 and 2.299) or species level (score from 2.3 to 3.0). When the score was <1.7, no identification was considered reliable. The 16S rRNA sequence of unidentified isolates was obtained using an ABI Prism 3130xl Genetic Analyzer capillary sequencer (Applied Biosystems, Bedford, MA, USA), as previously described (Morel et al., 2015;Seng et al., 2009). Finally, the sequences were compared to the NCBI nr database using the BLAST algorithm (https://blast.ncbi.nlm.nih.gov/ Blast.cgi). If the 16S rRNA sequence similarity value was lower than 98.7%, the isolate was considered as a putative new species (Kim, Oh, Park, & Chun, 2014;Stackebrandt & Ebers, 2006;Yarza et al., 2014).

| Phylogenetic analysis
The 16S rRNA sequences of isolates not identified using mass spectrometry and those of members of the family Peptoniphilaceae with standing in nomenclature (downloaded from the nr database) were aligned using CLUSTALW (Thompson, Higgins, & Gibson, 1994) with default setting. The phylogenetic inferences were performed using both the neighbor-joining and maximum-likelihood methods with the software MEGA version 6 (Tamura, Stecher, Peterson, Filipski, & Kumar, 2013).

| Phenotypic characteristics
For each new isolate, cell morphology was visualized using optical and electron microscopy. Oxidase, catalase, motility, sporulation tests, as well as Gram stain were performed as already reported (Murray, Baron, Jorgensen, Landry, & Pfaller, 2007). Cells were fixed for electron microscopy for at least 1 hour at 4°C with 2.5% glutaraldehyde in a 0.1 mol L −1 cacodylate buffer. One drop of cell suspension was deposited for about 5 min on a glow-discharged formvar carbon film on 400-mesh nickel grids (FCF400-Ni, EMS). The grids were dried on a blotting paper. Then, the cells were negatively stained at room temperature for 10 s with a 1% ammonium molybdate solution in filtered water. Micrographs were obtained using a Tecnai G20 Cryo (FEI) transmission electron microscope operated at 200 keV.
For the analysis of cellular fatty acid methyl ester (FAME), gas chromatography/mass spectrometry (GC/MS) was achieved.

| Genome sequencing and analyses
After a pretreatment of 2 hr at 37°C using lysozyme, the genomic DNAs (gDNAs) of strains KhD-2 T , KHD4 T , and Kh-D5 T were extracted using the EZ1 biorobot and EZ1 DNA Tissue kit (Qiagen  The genome assembly was performed with a pipeline that enabled to create an assembly with various software such as Velvet (Zerbino & Birney, 2008), Spades (Bankevich et al., 2012), and Soap Denovo (Luo et al., 2012), on trimmed data with MiSeq and Trimmomatic (Bolger, Lohse, & Usadel, 2014) software or untrimmed data with only MiSeq software. In order to reduce gaps, GapCloser was used (Luo et al., 2012). Phage contamination was searched (blastn against Phage Phix174 DNA sequence) and eliminated.
Finally, scaffolds with sizes under 800 bp and scaffolds with a depth value lower than 25% of the mean depth were identified as possible contaminants and removed. The best assembly was considered by using several criteria including number of scaffolds, N50, and number of N. Spades gave the best assembly for the three studied strains with depth coverage of 518x.
Prodigal was used to predict open reading frames (ORFs) (Hyatt et al., 2010) using default parameters. However, the predicted ORFs were excluded if they spanned a sequencing gap region (containing Ns). The predicted bacterial protein sequences were analyzed as previously reported (Alou et al., 2017). tRNA genes were found using the tRNAScan-SE tool (Lowe & Eddy, 1997), while RNAmmer was used to find ribosomal RNAs (Lagesen et al., 2007).
Phobius was used to predict lipoprotein signal peptides and the number of transmembrane helices (Käll, Krogh, & Sonnhammer, 2004). ORFans were identified when the BLASTP search failed to provide positive results (E-value smaller than 1e −03 for ORFs with a sequence size larger than 80 aa or an E-value smaller than 1e −05 for ORFs with a sequence length smaller than 80 aa), as previously reported (Alou et al., 2017). For genomic comparison, the closest species with validly published names in the 16S RNA phylogenetic tree were identified with the Phylopattern software (Gouret, Thompson, & Pontarotti, 2009). The complete genome, proteome, and ORFeome sequences were retrieved for each selected species in NCBI. An annotation of the entire proteome in order to define the distribution of functional classes of predicted genes according to the COG classification of their predicted protein products was performed as already reported (Alou et al., 2017). Annotation and comparison processes were done using the DAGOBAH software as previously described (Alou et al., 2017;Gouret et al., 2005Gouret et al., , 2011.
Finally, in order to evaluate the genomic similarity between the genomes, we determined two previously described parameters: average amino acid identity (AAI) based on the overall similarity between two genomic datasets of proteins available at (http:// enve-omics.ce.gatech.edu/aai/index) and digital DNA-DNA hy-

| Strain identification and phylogenetic analysis
The MS identification of the three bacteria, secluded, respectively, after 24 hr (strains KhD-2 T and KHD4 T ) and 15 days (Kh-D5 T ) of preincubation, failed. This suggested that these isolates were not in the database and may be unknown species. Pairwise analysis of 16S rRNA sequences attested that strain KhD-2 T exhibited 92.8% and 87.4% sequence similarities with strains KHD4 T and Kh-D5 T , respectively, and strains KHD4 T and Kh-D5 T had an 88.7% identity.
BLASTN sequence searches demonstrated that the three strains were related to the genus Peptoniphilus, suggesting that each strain represented a new species within this genus. Strain KhD-2 T exhibited a 16S rRNA similarity of 99.7% with Peptoniphilus sp. strain The names P. vaginalis sp. nov., P. raoultii sp. nov., and P. pacaensis sp.

| Phenotypic features
Cells from all three novel strains (KhD-2 T , KHD4 T , and Kh-D5 T ) were Gram--positive cocci (mean diameter of 0.6-0.7 μm for each). After 4 days of incubation, colonies on blood agar were grey and circular, and all had a diameter ranging from 1 to 2 mm. For all the three strains, growth occurred only in anaerobic atmosphere.
The genome characteristics were detailed in Table 4. The repartition of genes into the 25 general COG categories was represented in Table 5 and Figure 4. When compared to other Peptoniphilus species, the three strains had genome sizes, G+C contents and total gene counts in the same range (Table 6, Figure 5). Although, base composition varies widely among bacterial species, the genes within a given genome are relatively similar in G+C content with the exception of recently acquired genes. As a matter of fact, DNA sequences acquired by horizontal transfer often bear unusual sequence characteristics and can be distinguished from ancestral DNA notably by a distinct G+C content (Lawrence & Ochman, 1997). The region between 100,000 and 600,000 bp of the chromosome from strain KhD-5 T showed a high variation in G+C content (Figure 3). Thus, 43 genes putatively acquired by horizontal gene transfer were identified in this region, including 25 genes specific for strain KhD-5 T and 18 genes shared with strain Peptoniphilus urinimassiliensis.
Consequently, the presence of these genes may play a role in the F I G U R E 1 Phylogenetic analysis based on the 16S RNA gene sequence highlighting the position of Peptoniphilus vaginalis strain KhD-2 T , Peptoniphilus raoultii strain KHD4 T , and Peptoniphilus pacaensis strain Kh-D5 T relative to other closely related strains. GenBank accession numbers are indicated in parentheses. Sequences were aligned using Muscle v3.8.31 with default parameters and, phylogenetic inferences were performed using the neighbor-joining (a) and maximum-likelihood (b) methods with the software MEGA version 6. The scale bar represents a 2% nucleotide sequence divergence significant difference in genomic G+C content observed between strain KhD-5 T and other compared Peptoniphilus species as well as the similar genomic G+C content observed between strain KhD-5 T and P. urinimassiliensis.

| D ISCUSS I ON
The aim of this study was to investigate, using culturomics, the vaginal flora of a woman with bacterial vaginosis. Indeed, bacterial vaginosis is a gynecologic disorder marked by a perturbation of the vaginal microbiota equilibrium with a loss of commensal Lactobacillus spp. and their replacement with anaerobic bacteria including Atopobium vaginae, Bacteroides spp., Mobiluncus spp., Prevotella spp., and numerous Gram-positive anaerobic cocci (Bradshaw et al., 2006;Onderdonk, Delaney, & Fichorova, 2016;Shipitsyna et al., 2013). Gram-positive anaerobic cocci were associated to various infections (Murdoch, 1998). They represent about 24%-31% of anaerobic bacteria cultivated in clinical specimens (Murdoch, Mitchelmore, & Tabaqchali, 1994). In this present study, three novel Gram-positive-staining, anaerobic cocci (KhD-2 T , KHD4 T , and Kh-D5 T ) were cultured in the vaginal discharge of a patient suffering from bacterial vaginosis. These bacteria exhibited sufficient MALDI-TOF MS profiles, 16S rRNA sequence, F I G U R E 2 Gel view comparing strains KhD-2 T , KHD4 T , and Kh-D5 T to other species within the genus Peptoniphilus. 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. The peak intensity is expressed by a gray scale scheme code. The right y-axis indicates the relation between the color of a peak and its intensity, in arbitrary units.    KhD-2 T , KHD4 T , and Kh-D5 T ferment ribose and tagatose. The study of their genomes revealed that strain Kh-D2 T had 75 genes associated to carbohydrate metabolism, including 4 genes (1 rbsA gene, 2 rbsR genes, and 1 rpiB gene) encoding proteins involved in fermentation of ribose; the genome from strain KHD4 T contained 61 genes associated to carbohydrate metabolism of which one rpiB gene is involved in fermentation of ribose; and strain KhD-5 T had 58 genes associated to carbohydrate metabolism with 3 genes implicated in ribose fermentation (2 rpiB genes and 1 rbsK) and 1 gene encoding a tagatose biphosphate aldolase enzyme involved in tagatose fermentation. In addition, the genomes of strains Kh-D2 T , KHD4 T , and KhD-5 T also had 25 genes (5 genes encoding proteins responsible for the degradation of histidine, 1 of lysine, 2 of threonine, 12 of methionine, and 5 of arginine), 20 genes (5 of histidine, 1 of lysine, 1 of threonine, 7 of methionine, and 6 of arginine), and 21 genes (14 which degraded methionine, 6 for arginine and 1 for lysine), associated to amino acid degradation, respectively.
Gram-stain-positive. Coccus-shaped bacterium with a mean diameter of 0.66 μm. Peptoniphilus vaginalis sp. nov. is a mesophilic bacterium; its optimal growth occurs at temperature 37°C, a pH ranking from 6.5 to 8.5, and a NaCl concentration lower than 5%.
In EMBL-EBI, the 16S rRNA gene sequence is deposited under accession number LN998068 and the draft genome sequence under accession number FMWM00000000. Strain KHD4 T (=CSUR P0110 = CECT 9308) is the type strain of P. raoultii sp. nov., which was cultured from the vaginal discharge of a woman suffering from bacterial vaginosis.
Gram-stain-positive. Coccus-shaped bacterium with a mean diameter of 0.7 μm. Peptoniphilus pacaensis sp. nov. is a mesophilic bacterium; its optimal growth occurs at temperature 37°C, a pH ranking from 6.5 to 8.5, and a NaCl concentration lower than 5%.
The type strain of P. pacaensis sp. nov. is strain Kh-D5 T (=CSUR P2270 = DSM 101839), which was cultured from the vaginal discharge of a woman suffering from bacterial vaginosis.