Pseudomonas canadensis sp. nov., a biological control agent isolated from a field plot under long-term mineral fertilization

The bacterial strain 2-92T, isolated from a field plot under long-term (>40 years) mineral fertilization, exhibited in vitro antagonistic properties against fungal pathogens. A polyphasic approach was undertaken to verify its taxonomic status. Strain 2-92T was Gram-reaction-negative, aerobic, non-spore-forming, motile by one or more flagella, and oxidase-, catalase- and urease-positive. The optimal growth temperature of strain 2-92T was 30 °C. 16S rRNA gene sequence analysis demonstrated that the strain is related to species of the genus Pseudomonas. Phylogenetic analysis of six housekeeping genes (dnaA, gyrB, recA, recF, rpoB and rpoD) revealed that strain 2-92T clustered as a distinct and well separated lineage with Pseudomonassimiae as the most closely related species. Polar lipid and fatty acid compositions corroborated the taxonomic position of strain 2-92T in the genus Pseudomonas. Phenotypic characteristics from carbon utilization tests could be used to differentiate strain 2-92T from closely related species of the genus Pseudomonas. DNA–DNA hybridization values (wet laboratory and genome-based) and average nucleotide identity data confirmed that this strain represents a novel species. On the basis of phenotypic and genotypic characteristics, it is concluded that this strain represents a separate novel species for which the name Pseudomonas canadensis sp. nov. is proposed, with type strain 2-92T (=LMG 28499T=DOAB 798T). The DNA G+C content is 60.30 mol%.

Species of the genus Pseudomonas are aerobic, Gramreaction-negative gammaproteobacteria, ubiquitous in agricultural soils, and are well adapted to grow in the rhizosphere. This genus includes species that are of significant environmental importance such as plant growth promoters, xenobiotic degraders and biocontrol agents [1][2][3]. The fluorescent pseudomonads are uniquely capable of synthesizing many metabolites that play a role in maintaining soil health leading to bioprotection of crops against pathogens [4,5].
In 2009, we started a prospective study focused on cultivable members of the genus Pseudomonas in a soil ecosystem under long-term (>40 years) applications of inorganic fertilizer (NPK). Soil samples were collected from a corn-alfalfa rotation plot located at Woodslee, Ontario, Canada (42.22 N 82.73 W). Triplicates of bulk soil (10 g each) were suspended in 90 ml 0.85 % NaCl and vortexed vigorously. The soil suspensions were serially diluted in 0.85 % NaCl, plated on King's B (KB) agar (Sigma-Aldrich) and incubated at 28 C for 48 h to isolate fluorescent pseudomonads. Colonies were screened for fluorescence under UV light. Single colonies were obtained after repeated streaking and plated on Pseudomonas-F agar medium (BD Difco). A total of 148 bacterial isolates were obtained of which 99 % fluoresced under UV light. All isolates were evaluated for in vitro antagonistic activities against the following fungal pathogens: Fusarium graminearum, Rhizoctonia solani and Gaeumannomyces graminis. Twelve of the isolates were able to inhibit the mycelial growth of at least two of the fungal pathogens. All 12 biological control isolates were identified as members of the genus Pseudomonas on the basis of phenotypic features and 16S rRNA sequence analyses. Of the three strains (2-92, 2-36 and 2-114) that potently inhibited the growth of R. solani and G. graminis, two were assigned to known species of the genus Pseudomonas: Pseudomonas simiae 2-36 and Pseudomonas extremorientalis 2-114. Strain 2-92 T did not match any known species of the genus Pseudomonas.
In the present study, the taxonomic status of strain 2-92 T was investigated using electron microscopy, phenotypic tests, chemotaxonomic traits, analyses of the 16S rRNA gene and six house-keeping genes (dnaA, gyrB, recA, recF, rpoB and rpoD), DNA-DNA hybridization and DNA G+C content determination as well as analysis of draft genome sequences. Based on this polyphasic characterization, a novel species, Pseudomonas canadensis sp. nov. is proposed.
Cells of strain 2-92 T were purified on Pseudomonas-F agar medium (BD Difco), and cell suspensions in Luria-Bertani (LB; BD Difco) broth supplemented with 30 % (v/v) glycerol were maintained at À80 C. Cells were routinely grown overnight in liquid LB with shaking or on LB agar medium and incubated at 28 C. Strain 2-92 T was Gram-reactionnegative based on the 3 % KOH assay [6] and oxidasepositive based on API 20 NE strips (BioM erieux). It was catalase-positive based on 3 % (v/v) hydrogen peroxide solution. Cell growth was tested at different temperatures (5-40 C, in steps of 1 C below 5 C then at intervals of 5 C), and salt tolerance (NaCl) was determined in the range 0-6 % (w/v) as described by Gonz alez et al. [7]. Strain 2-92 T grew at 4 C, showed optimal growth at 30 C and did not grow at 40 C. This strain was tolerant to different NaCl concentrations up to 4 % and was non-spore-forming based on the Schaeffer and Fulton method [8]. Cell morphology was investigated using scanning (SEM) and transmission (TEM) electron microscopy. Bacteria were cultured in LB broth overnight at 28 C, processed as described by Greco-Stewart et al. [9] and imaged using a Philips XL-30 ESEM scanning electron microscope (data not shown). TEM was performed as previously reported by Hayat and Miller using 1 % phosphotungstic acid (pH 7.0) [10], and images were captured with a Hitachi H7000 microscope (Fig. 1). Cells were rod-shaped with sizes in the range of 0.5-0.6 µm wide and 2.1-2.6 µm long, consistent with species of the genus Pseudomonas [3], and had one or more polar flagella. Motility was demonstrated using triphenyl tetrazolium [11] in semisolid medium (per litre: 3.0 g beef extract, 10.0 g pancreatic digest of casein, 5.0 g NaCl, 4.0 g agar). Fluorescent pigment was produced on KB medium [12]. After 48 h of incubation at 28 C on KB, the colonies of 2-92 T were white-yellowish, circular (mean 4 mm in diameter) and convex with regular margins.
Genomic DNAs were extracted using a Wizard SV Genomic DNA Purification kit (Promega), and the purity of each 0.5 µm sample was determined by agarose gel electrophoresis. The DNA concentrations were determined fluorimetrically using a FLUOstar OPTIMA micro-plate reader (BMG-LABTECH) with picogreen chemistry (Invitrogen). PCR amplifications of almost full-length (1466 bp) 16S rRNA genes were performed as described previously [20] using primer pairs 16F27 (5¢-AGAGTTTGATCMTGGCTCAG-3¢) and 16R1492 (5¢-TAC GGYTACCTTGTTACGACTT-3¢) [21]. Sequencing (ABi 3300xl analyzer; Applied Biosystems) of the DNA fragments was performed in-house as reported previously [22] using BigDye terminator chemistry. All the other 16S rRNA gene sequences (53) of reference species of the genus Pseudomonas were retrieved from GenBank. Sequences of strain 2-92 T were compared with other species by pairwise distance calculations; multiple sequence alignments and phylogenetic analysis of the 16S rRNA gene sequences were performed using MUSCLE [23] and MEGA5.0 [24], respectively. Phylogenetic analysis was performed on almost-complete 16S rRNA gene sequences using maximum-likelihood (ML) with a general time reversible best-fit substitution model as implemented in jModelTest version 2 [25]. The topological robustness of the trees was evaluated by bootstrap analysis based on 1000 replicates. The 16S rRNA gene tree confirmed that strain 2-92 T was unique and a member of the genus Pseudomonas (Fig. 2). Pairwise sequence similarity values were between 98.8 and 99.5 % with several type strains of the P. fluorescens group [26], for example P. simiae CCUG 50988 T , P. salomonii LMG 22120, P. extremorientalis KMM 3447 T and P. azotoformans LMG 21611 T (Fig. 2). Phylogenies reconstructed using neighbour-joining and minimum-evolution algorithms showed similar topologies (Fig. S2).
DNA-DNA hybridization (DDH) studies were employed as previously described by Ramisse et al. [32] between strain 2-92 T and seven closely related species of the genus Pseudomonas (Table 2). Hybridizations were performed in triplicates with reciprocal reactions. DDH values of 49.0, 35.6 and 46.7 % between strain 2-92 T and P. simiae CCUG 50988 T , P. extremorientalis LMG 19695 T and P. salomonii LMG 22120 T , respectively (Table 2), were clearly below the threshold level (<70 %) for species definition [33]. Also, genome-sequence-based digital DDH (dDDH; [34]) and MUMmer-based average nucleotide identity (ANIm; [35]), which have been proposed to replace wet-lab DDH [34,36], were employed to confirm the taxonomic position of strain 2-92 T . The dDDH values were calculated using the genome-togenome distance calculator (GGDC) version 2.1 (http://ggdc. dsmz.de; [34]). ANIm similarity values were computed as described by Kurtz et al. [35] and implemented in JSpecies [37]. The dDDH values between strain 2-92 T and the seven closely related species of the genus Pseudomonas were all below the threshold of 70 % for species delineation (  Table 2). The DNA G+C content of strain 2-92 T was 60.3 mol% using the HPLC method [38], which is well within the range reported for species of the genus Pseudomonas [3]. The DNA G+C content was confirmed by wholegenome sequencing and analysis of strain 2-92 T (AYTD00000000; 6.4 Mb genome size) [30].

DESCRIPTION OF
The type strain is 2-92 T (=DOAB 798 T =LMG 28499 T ), isolated from a soil sample from Woodslee, Ontario, Canada. The DNA G+C content of strain 2-92 T is 60.3 mol%.

Funding information
This work was supported by Agriculture and Agri-Food Canada (AAFC) projects 1800 and 1136; and the support for genome analysis was provided by projects CRTI09-462RD and J-00409).