Pedobacter schmidteae sp. nov., a new bacterium isolated from the microbiota of the planarian Schmidtea mediterranea

Pedobacter schmidteae sp. nov. strain EGT (Collection de Souches de l’Unité des Rickettsie CSUR P6417 = Colección Española de Cultivos Tipo CECT 9771) is a new Pedobacter species isolated from the planarian Schmidtea mediterranea. Schmidtea mediterranea are flatworms living in freshwater and exhibiting an unusual ability to regenerate amputated parts. To date, the gut microbiota of Schmidtea mediterranea remains poorly studied. Here, via the culturomics strategy that consists in using diversified culture conditions, we isolated a new bacterium, strain EG, that we characterized using the taxono-genomics approach that combines phenotypic assays and genome sequencing and analysis. Strain EG exhibits a 16S rRNA sequence similarity of 98.29% with Pedobacter nyackensis strain NWG-II14T, its closest neighbour with standing in nomenclature. It is an aerobic bacterium belonging to the family Sphingobacteriaceae. Colonies are small, round, smooth and transparent. Bacterial cells are Gram-negative, rod-shaped, motile and non-spore-forming bacilli with positive catalase and oxidase activities. The genome sequence is 6,198,518 bp–long with a G + C content of 41.13%, and the Ortho-ANI and dDDH values when compared to P. nyackensis are 77.34% and 21.50%, respectively. Strain EGT exhibits unique characteristics that classify it as the type strain of new bacterial species for which we propose the name Pedobacter schmidteae sp. nov.

Isolation and identification of bacteria from Schmidtea mediterranea. Following two weeks of starvation, S. mediterranea were washed in filter-sterilized water and then one ground worm was inoculated in Buffered Charcoal Yeast Extract (BCYE) (Oxoid Deutschland GmbH, Wesel, Germany), Luria Bertani (LB) and 5% sheep blood-enriched Columbia agar (bioMérieux, Marcy l' étoile, France). All inoculated media were incubated at 19, 28 and 37 °C. Each individual bacterial colony was harvested and identified by MALDI-TOF-MS (Microflex spectrometer; Bruker Daltonics, Bremen, Germany) 15 . The obtained spectra were imported into the MALDI Biotyper 3.0 software (Bruker Daltonics) and analysed against the reference spectra of bacteria included in the database (Bruker database constantly updated with the Mediterranee-Infection database (http://www. mediterranee-infection.com/article.php?larub=280&titre=urms-database). The MALDI Biotyper RTC software was used to interpret the results according to the obtained score values: a colony was likely identified at the species level for a score > 2.0, probably identified for a score between 1.99 and 1.7, but not identified for a score <1.7. phylogenetic analysis. Bacterial colonies that were not identified at the species level using MALDI-TOF MS were further tested using 16S rRNA sequencing. Genomic DNA was extracted using an EZ1 automate and the DNA tissue kit (Qiagen, Hilden, Germany). The complete 16S rRNA gene amplification and sequencing was performed using eight primers on an ABI Prism 3130xl Genetic Analyzer capillary sequencer (Applied Bio systems, Bedford, MA, USA). The primers used were Fd1 (5′-AGAGTTTGATCCTGGCTCAG-3′), Rp2 (5′ -AC GGCTAC CT TGT TAC GACT T-3′ ), F536 (5′ -CAGCAGC C GC GGTAATAC-3′ ) , R536 (5′ -GTAT TAC C GC GGCTGCTG-3′ ), F800 (5′ -AT TAGATAC C CTGGTAG-3′ ), R800 ( 5 ′ -C TAC C AG G G TAT C TA AT-3 ′ ) , F 1 0 5 0 ( 5 ′ -T G T C G T C AG C T C G T G -3 ′ ) a n d R 1 0 5 0 (5′-CACGAGCTGACGACA-3′) (Eurogentec, Angers, France). The CodonCode Aligner software was used for sequence alignment, assembly and correction (https://www.codoncode.com/). For taxonomic assignation, a BLASTn search was performed against the nr database 16 . A sequence similarity threshold of 98.65% by comparison with the phylogenetically closest species with standing in nomenclature was used to delineate a putative new species 17 . Phylogenetic relationships were inferred from the comparison of 16S rRNA sequences using the MEGA7 software 18 . phenotypic, biochemical and chemical characteristics of strain eG. Culture of strain strain EG and P. nyackensis strain NWG-II14 T (DSM19625) was attempted at various growth temperatures (6, 19, 30, 37 and 45 °C) in 5% sheep blood-enriched Columbia agar (bioMérieux) under anaerobic, aerobic and microaerophilic atmospheres using GasPak ™ EZ generators (Becton-Dickinson, Maryland, USA). Sporulation was tested by thermal shock, which consists in exposing bacteria to a temperature of 80 °C for 30 minutes and then monitoring their growth for 4 days. Various salinity (0, 8.5, 25, 50, 100 and 200 g/l) and pH (4, 5.5, 6, 7.5, 8, 9 and 10) conditions were tested. Gram staining and motility from fresh colonies were observed using a DM1000 photonic microscope (Leica Microsystems, Nanterre, France) with a 40 × objective lens. Catalase and oxidase activities were tested by using a BBL DrySlide according to the manufacturer's instructions (Becton Dickinson, Le Pont de Claix, France). The size of bacterial cells was measured using transmission electron microscopy. API strips (API ZYM [19][20][21] , API 20NE 22,23 , API 20E 24,25 and API 50CH [26][27][28][29] , bioMérieux) were used to study the biochemical characteristics of the strains. Bacterial susceptibility to benzylpenicillin, amoxicillin, ampicillin, ceftriaxone, imipenem, ciprofloxacin, amikacin, gentamicin, streptomycin, daptomycin, doxycycline, metronidazole, rifampicin, and vancomycin was assessed using E-tests and a 0.5 McFarland concentration of strains EG and NWG-II14 T . Cellular fatty acid methyl ester (FAME) analysis was performed by GC/MS. Two samples were prepared with approximately 120 mg of bacterial biomass per tube harvested from several culture plates. Fatty acid methyl esters were prepared as described by Sasser 30 and GC/MS analysis was carried out as previously described 31 . Briefly, FAMES were separated using an Elite 5-MS column and monitored by mass spectrometry (Clarus 500 -SQ 8S, Perkin Elmer, Courtaboeuf, France). Spectral database search was performed using MS Search 2.0 operated with the following Standard Reference Database 1 A (NIST, Gaithersburg, USA) and FAMEs mass spectral database (Wiley, Chichester, UK).
Sequencing, assembly, annotation and genomic comparison. The bacterial genomic DNA (gDNA) of strain EG was extracted using an EZ1 automate and the DNA tissue kit (Qiagen, Hilden, Germany) and then was quantified using a Qubit assay (Life Technologies, Carlsbad, CA, USA) at 82.6 ng/μL. The bacterial gDNA was prepared and sequenced as previously described 32 . Briefly, sequencing was performed using the Mate-Pair strategy and a Miseq sequencer (Illumina, San Diego, CA, USA). A concentration of 1.5 μg of gDNA prepared following the Nextera Mate-Pair Illumina guide was used to prepare the Mate-Pair library. The gDNA was simultaneously fragmented and tagged using a Mate-Pair junction adapter. Then the fragmentation pattern was validated using a DNA 7500 labchip on a BioAnalyzer (Agilent 2100, Agilent Technologies, Santa Clara, CA, USA). The size of the DNA fragments ranged from 1.5 kb to 11 kb. No size selection was performed and 662 ng of labelled fragments were circularized. Next, circularized DNA was mechanically sheared using a Covaris device S2 (Covaris, Woburn, MA, USA) into small fragments with an optimal size at 1200 bp. The library profile was analysed on a High Sensitivity Bioanalyzer LabChip (Agilent Technologies) and the concentration library was measured at 61.4 nmol/l. Then, the library was loaded on the sequencer after a denaturation step and a dilution at 15 pM. Automated cluster generation and sequencing were performed in a single 39-h run in a 2 × 251-bp and sequencing reads were assembled using the A5 pipeline. Genomic annotation was obtained using the Prokka software. A search for virulence factors was performed by comparaison with the VFDB database (http://www.mgc.ac.cn/ VFs/) using BLASTn 33,34 .
The analysis of 13 S. mediterranea revealed the presence of strain EG in 11 S. mediterranea sp. nov. Strain EG is a member of the family Sphingobacteriaceae within the phylum Bacteroidetes (Table 1).

Phenotypic, enzymatic and biochemical characteristics. After 4 days of culture on blood-enriched
Columbia agar, colonies from strain EG were small (0.4 mm of diameter), transparent, round with a convex shape and smooth. Bacterial cells were Gram-negative (Fig. 2), rod-shaped, motile, and non-spore-forming bacilli. Using the Image J software, their mean length and width were 1.98 µm and 0.69 µm, respectively (Fig. 3), without any flagellum. For the two strains EG and P. nyackensis NWG-II14 T , no growth was obtained in anaerobic or microaerophilic conditions. Both strains grew at temperatures ranging from 6 to 30 °C in aerobic atmosphere at pH values ranging from 5.5 to 9; the strains also grew at salinity concentrations lower than 25 g/L. Catalase and oxidase activities were positive for both strains.
Genomic characteristics. The genome sequence from strain EG was assembled into one contig of 6,198,518 bp with a G + C content of 41.13%. We identified a total of 5,012 predicted protein-coding genes, in addition to 3 complete rRNA operons, 52 tRNAs and 1 tmRNA. Comparison of these genomic data with those from of closely related species is presented in Table 5. The distribution of genes in functional categories (COGs) is shown within Table 6. In addition, using the Prokka software, we identified a two-component sensor histidine kinase system. This system exhibited an identity of 92% (E-value of 10 −3 ) and a score of 52 with the GacS     41,42 sensor histidine kinase/response regulator in Pseudomonas putida strain KT2440 according to the VFDB database. It has been reported that the two-component system signaling pathways are the major signaling mechanisms in bacteria and as well in Archaea 43 to monitor external and internal stimuli (including concentration of ions and gas, redox states, levels of nutrients and cell density) 44 . This pathway is also found in simple eukaryota and higher plants 45 . In opportunistic bacterial pathogens, the use of two-components systems is required to regulate the expression of genes necessary for the transition from the environmental reservoir to the host 46 . Digital DNA-DNA hybridization values (dDDH) obtained using the GGDC software are reported in Table 7. For strain EG, these values ranged from 18.90 with P. nutrimenti to 21.50% with P. nyackensis. Such values were lower than the 70% threshold recognized to delineate distinct species. Similarly, Ortho-ANI values (Fig. 4) ranged from 70.98% with P. nutrimenti to 74.65% with P. nyackensis, which is lower than the 95% threshold used to discriminate bacterial species. Thus, we could confirm that strain EG belonged to a separate Pedobacter species for which we propose the name Pedobacter schmidteae sp. nov.

conclusion
Using the taxono-genomic approach, we concluded that strain EG is the representative strain of the new species P. schmidteae sp. nov. Interestingly, P. schmidteae sp. nov. is present in 84.6% of S. mediterranea worms tested. To best of our knowledge, P. schmidteae sp. nov has never been identified anywhere else than in S. mediterranea. Indeed, no nucleotide sequence linked to this new strain has been found in the nr database (BLASTn https:// blast.ncbi.nlm.nih.gov). To date, we assume that P. schmidteae sp. nov is unique to S. mediterranea, but we cannot exclude formally that it might be present in other living organisms of the environment.  Table 3. Differential biochemical characteristics of Pedobacter schmidteae and phylogenetically-related species of the genus Pedobacter. Strains: 1, P. schmidteae strain EG T ; 2, P. nyackensis strain NWG-II14 T ; 3, P. heparinus strain DSM 2366 T ; 4, P. africanus DSM 12126 T . The data were completed using previously described characteristics 38,40 of taxa 3 and 4, and those obtained in the present study. Data presented for the taxa 3 and 4 were collected in previously published work, only results obtained with the same methodologies than used in the present study for the taxa 1 and 2 (see material and methods) were considered. +, positive; −, negative; V, variable; NA, data not available.