Rhizospheric soil samples of rice (Oryza sativa L.) and banana (Musa spp. AAA) were collected from agricultural fields located at Puducherry, India. Samples were stored at 4?C before being processed within 24 h of collection. The soil was sand clay loam and its characteristics were as follows; pH 7.3, 74 ?g g^-1 nitrogen, 2.9 ?g g^-1 phosphorus, 5.29 ?g g^-1 potassium, 0.2 mOhms cm^-1 of electrical conductivity (EC). Isolation of fluorescent pseudomonad bacteria from rhizospheric soil was performed as described earlier [²⁵]. Briefly, soil suspension was obtained by shaking 10 g of soil sample having roots with tightly adhering soil in 90 ml of 0.1 M MgSO₄.7H₂O buffer for 10 min at 180 rpm on a rotary shaker. Resulting suspensions were serially diluted and 0.1 ml aliquots of each dilution were spread onto King's medium B (KB) agar in triplicates. After incubation at 28?C for 2 days, fluorescent pseudomonad colonies from replicate plates were identified under UV light (366 nm). Purified single colonies were further streaked onto KB agar plates to obtain pure cultures. Stock cultures were made in Luria Bertani (LB) broth containing 50% (w/v) glycerol and stored at -86?C.

To detect the phosphate solubilizing bacteria, strains were streaked onto Pikovskaya's agar medium, which contains (per liter): 0.5 g yeast extract, 10 g dextrose, 5 g Ca₃(PO₄)₂, 0.5 g (NH₄)₂SO₄, 0.2 g KCl, 0.1 g MgSO₄.7H₂O, 0.0001 g MnSO₄.H₂O, 0.0001 g FeSO₄.7H₂O and 15 g agar. After 3 days of incubation at 28?C, strains that induced clear zone around the colonies were considered as positive [²⁷]. Determination of phosphate solubilizing activity by the strains was carried out following standard method [²⁸]. Briefly, cells were grown in liquid medium (pH 7.0) at 28?C up to 10 days and on 1, 3, 5, 7 and 10 day an aliquot of 5 ml was collected and cells were removed by centrifugation at 9,000 g for 20 min. Soluble free phosphate in culture supernatant was estimated from the absorbance values obtained using the calibration curve with KH₂PO₄ at 600 nm for each strain. Also, pH variation in Pikovskaya's medium during the growth of each strain was also observed.

In order to determine the phenotypic diversity of 80 phosphate solubilizing bacteria characterization was done on the basis of fluorescence on King's B (KB) medium, levan formation, gelatin liquefaction, nitrate reduction and growth at 4?C and 42?C [²⁹,³⁰]. In order to identify fluorescence of the strains, overnight grown cells on KB agar medium were visualized under UV light (366 nm). Gram's reaction was determined by the KOH technique [³¹]. Briefly, visible amount of overnight grown cells from agar plate was smeared (1?2 cm² area) onto glass slide containing loopful (3 mm) of 3% aqueous KOH solution. Gram negative strains were identified as viscous gel that string out along with the loop. To test the presence of cytochrome oxidase, bacterial culture grown for 24 h on nutrient agar (NA) supplemented with 1% glucose was used. A loopful cells were rubbed onto a filter paper impregnated with 1% (wt/vol) aqueous N,N,N',N'-tetra-methyl-p-phenylenediamine dihydrochloride solution. A change in the color of the cultures to deep purple within 10 s was registered as a positive result. Presence of arginine dihydrolase and levan sucrase was tested on Thornley's medium and NA supplemented with arginine and sucrose (20 g l^-1), respectively [²⁹].

Carbon utilization profiles were tested using Hicarbohydrate? kit as descried by the manufacturer (Himedia Laboratories, Mumbai, India). Cells were grown in nutrient broth to reach density of 0.5 O.D. at 600 nm. An aliquot of 50 ?l of this suspension was inoculated to each well of Hicarbohydrate? kit, incubated at 30?C for 24 h and the results were registered according to the instructions of the manufacturer. The experiment was done with three replicates. On the basis of data derived from the carbon source utilization profiles, a matrix with binary code composing positive (1) and negative (0) values was made. SIMQUAL program was used to compute the symmetric matrix in the form of average taxonomic distances. Sequential, agglomerative, hierarchical and nested (SAHN) clustering was used for the cluster analyses [³²]. Dendrogram was constructed from the similarity matrix by the unweighted pair group with mathematical averages (UPGMA) using NTSYS-pc2.02a (Exeter software, New York, USA) numerical taxonomy and multivariate analysis system.

Bacteria were cultured in LB broth at 28?C for 18 h and the total genomic DNA was extracted as described [²⁵]. The BOX-A1R (5'-CTACGGCAAGGCGACGCTGACG-3') primer for genotypic analysis [³³] was synthesized by Integrated DNA Technologies Inc. (Coralville, IA, USA). PCR reaction (50 ?l) contained 50 pM of primer, 50 ng of genomic DNA, 1? Taq DNA polymerase buffer, 1 U of Taq DNA polymerase (Promega, Madison, WI), 0.2 mM of each deoxynucleotide triphosphate (dNTP), and 1.5 mM MgCl₂. Amplification was performed in a DNA thermal cycler (2400 cycler, Perkin Elmer International, Rotkreuz, Switzerland) programmed with an initial denaturation at 95?C for 7 min, 30 cycles at 94?C for 1 min, 53?C for 1 min, and 65?C for 8 min, with an extension at 65?C for 15 min. A 10 ?l of PCR product was separated using 1.5% agarose gel stained with ethidium bromide in 1? tris acetate ethylenediaminetetraacetic acid (TAE). The image of the gel was digitized by using BIO-CAPT system (Vilber Lourmat, France) and stored as TIFF files. BOX-PCR band profiles were detected by QuantityOne program (Bio-Rad Laboratories, CA, USA). Computer assisted analysis of genomic fingerprints was performed by using the BIO-GENE software program v11.02 (Vilber Lourmat, France). Similarity matrices of whole densitometric curves of the gels tracks were calculated by using the Dice coefficient. Cluster analysis of similarity matrices was performed by the un-weighted pair group with mathematical average (UPGMA) algorithm.

Amplification of 16S rRNA gene was performed from the genomic DNA of strains using universal primers fD1 (5'-GAGTTTGATCCTGGCTCA-3') and rP2 (5'-ACGGCTACCTTGTTACGACTT-3') [³⁴]. PCR cocktails (50 ?l) contained 50 pM of primer, 50 ng of genomic DNA, 1? Taq DNA polymerase buffer, 1 U of Taq DNA polymerase (Promega, Madison, WI, USA), 0.2 mM of each dNTP, and 1.5 mM MgCl₂. Amplification was performed in a DNA thermo cycler (2400 cycler, Perkin Elmer International, Rotkreuz, Switzerland) at 94?C for 3 min, followed by 30 cycles of 10 s at 94?C, 1 min at 56?C and 30 s at 72?C with an extension of 72?C for 5 min. A 5 ?l aliquot of each amplification product was electrophoresed on a 0.7% agarose gel in 1? TAE buffer at 50 V for 45 min, stained with ethidium bromide and the PCR products were visualised with a UV transilluminator. PCR products were purified using Quick PCR purification column (Promega, Madison, USA). Purified PCR products were sequenced with automated DNA sequencer with specific primers using the facility at Macrogen Inc. (Seoul, Korea). To perform molecular phylogenetic analyses, reference sequences required for comparison were downloaded from the EMBL database using the site . All the sequences of 16S rRNA were aligned using the multiple sequence alignment program CLUSTAL W [³⁵]. The aligned sequences were then checked for gaps manually, arranged in a block of 600 bp in each row [³⁶] and saved as molecular evolutionary genetics analysis (MEGA) format in software MEGA v3.0. The pair wise evolutionary distances were computed using the Kimura 2-parameter model [³⁷]. To obtain the confidence values, the original data set was resampled 1000 times using the bootstrap analysis method. The bootstrapped data set was used directly for constructing the phylogenetic tree using the MEGA v3.0 program for calculating the multiple distance matrixes [³⁸]. The multiple distance matrix obtained was then used to construct phylogenetic trees using neighbor-joining (NJ) method [³⁹].

The nucleotide sequences of 16S rRNA were deposited in GenBank. The accession numbers of the 16S rRNA nucleotide sequences of the strains are presented in Additional file 1.

Detection of the genes that encode for the production of antibiotics such as, 2,4-diacetylphloroglucinol (DAPG), phenazine-1-carboxylic acid (PCA), phenazine-1-carboxamide (PCN), pyrrolnitrin (PRN) and pyoluteorin (PLT) was done by PCR using gene-specific primers. Reference strains, Pseudomonas fluorescens Pf5, P. fluorescens 2?79, P. aureofaciens 30?84 (now considered as P. chlororaphis) and P. aeruginosa PAO1 were kindly supplied by Linda S. Thomashow, USDA, Washington State University, Pullman and P. fluorescens CHAO was kindly supplied by Genevi?ve D?fago, Swiss Federal Institute of Technology, Zurich. Oligonucleotide primers were synthesized by Integrated DNA Technologies Inc. (Coralville, IA, USA). The primer sets and the amplification conditions for the screening of genes encoding antibiotics are listed in Table 1. PCR reaction (50 ?l) contained 50 pM of each primer, 50 ng of genomic DNA, 1? Taq DNA polymerase buffer, 0.5 U of Taq DNA polymerase (Promega, Madison, WI, USA), 0.2 mM of each deoxynucleotide triphosphate, and 1.5 mM MgCl₂. Amplification was performed in a DNA thermal cycler. A 5 ?l aliquot of each amplified product was electrophoresed on a 0.7% agarose gel in 1? TAE buffer at 50 V for 45 min, stained with ethidium bromide and the PCR products were visualized with a UV transilluminator.

Antifungal metabolites were extracted as described [⁶,³⁶,⁴⁷]. Reference strains and antibiotic standards were kindly provided by Linda S. Thomashow, USDA, Washington State University, Pullman; and Genevi?ve D?fago, Swiss Federal Institute of Technology, Zurich. Thin layer chromatography (TLC) was carried out on silica gel G60 plate (20 ? 20 cm; 0.25 mm thick, Selecto Scientific, GA, USA). The plates were activated at 110?C for 30 min, cooled and spotted with ethanol solution containing standard antibiotic (0.5 ?g) and 20 ?l of the extract. Separation was performed using the solvent system: chloroform-methanol (9:1 vol/vol) for PCA and DAPG or chloroform-acetone (9:1 vol/vol) for PLT and PRN. The corresponding spots by PCA, and DAPG were detected by UV irradiation at 254 nm [⁴⁸]. PLT spots were detected by spraying with an aqueous 0.5% (wt/vol) Fast Blue RR salt solution 0.5% (wt/vol) and the PRN spots were detected by spraying the TLC plates with 2% p-dimethylaminobenzaldehyde dissolved in the ethanol-sulfuric acid (1:1 vol/vol). Detection and quantitative determination of antibiotics was done by analytical HPLC (Shimadzu, Kyoto, Japan) as described [³⁶]. Production of hydrogen cyanide (HCN) was determined by growing bacteria on KB agar medium supplemented with 4.4 g l^-1 glycine. A piece of filter paper impregnated with picric acid and sodium carbonate (0.5% and 2%, respectively) was placed in the lid of each Petri dish. Petri dish was sealed with parafilm and incubated at 28?C for 96 h. The production of cyanide was identified by the change in the color of filter paper from yellow to orange-brown [⁴⁷].

Of the 443 strains, 80 strains (18%) produced phosphate solubilization on Pikovskaya's agar medium by inducing clear zones (Fig. 1). The solubilization of tri-calcium phosphate was estimated for all strains. Soluble phosphate was estimated to be 29 to 105 ?g ml^-1 on 10 days of inoculation. A significant reduction in pH of Pikovskaya's liquid medium from pH 7.4 to pH 4.8 was observed on 10 days of incubation [See Additional file 2].

Out of 443 fluorescent pseudomonad strains tested, 80 strains showed phosphate solubilization potential. These phosphate solubilizing strains were Gram-negative, motile, rod shaped and tested positive for cytochrome oxidase, arginine dihydrolase, strains showed variability for traits such as gelatin hydrolysis, levan production and growth at 4?C and 42?C. All phosphate solubilizing strains utilized dextrose, galactose, mannose and citrate but exhibited varying degree of utilization towards other carbon sources such as lactose, xylose, fructose, melibiose, L-arabinose, glycerol, ribose, ?-methyl-D-mannoside, xylitol, esculin, D-arabinose, malonate, sorbose, trehalose, sorbitol, mannitol, adonitol and glucosamine. These strains did not utilize maltose, sucrose, inulin, salicin, dulcitol, inositol, ?-methyl-D-gluconate, rhamnose, cellobiose, melazitose, xylitol and ONPG. Numerical analysis of phenotypic characteristics revealed a high degree of polymorphism. All phosphate solubilizing strains were grouped into 3 major phenons at 0.76 similarity coefficient level (Fig. 2). The similarity coefficient range among phosphate solubilizing strains was 0.57 to 1.00. Phenons I, II and III consist a total of 61, 3 and 16 strains, respectively.

The cluster analysis based on the pair-wise coefficient similarity with UPGMA of BOX-PCR resulted into 3 distinct genomic clusters and at a 80% similarity coefficient generated 26 distinct BOX profiles (Fig. 3). A total of 20 strains were grouped into cluster I which shared 40% similarity with other strains. Cluster III consisted of 2 strains and a large number of strains grouped in cluster II comprising 58 strains. All the strains showed wide variations in fingerprinting pattern due to their high degree of genetic variability and distributed into different clusters. On the basis of these results, present study identified a high degree of genetic variability among different species of phosphate solubilizing fluorescent pseudomonads.

On the basis of phylogenetic analysis of 16S rRNA gene (600 bp), species of pseudomonads such as Pseudomonas monteilii, P. putida, P. plecoglossicida, P. fluorescens, P. fulva, P. mosselii and P. aeruginosa were identified. Phylogenetic analyses of the 80 fluorescent pseudomonad strains based on the NJ method with 1000 bootstrap sampling were resulted into 3 major clusters (Fig. 4). Of the 80 strains, cluster I formed with 58 strains, cluster II formed with 8 strains and cluster III formed with 14 strains. A total of 36 strains belong to P. monteilii, 10 strains belong to P. plecoglossicida, 12 strains belong to P. putida, 7 strains belong to P. fluorescens, 1 strain belongs to P. fulva, 12 strains belong to P. aeruginosa and 2 strains belong to P. mosselli (See Additional file 1; Fig. 4).

All the strains tested positive for the production of siderophore on the basis of change in colour of the CAS medium from blue to orange. Of the 80 phosphate solubilizing strains, 39 strains (49%) showed positive for the production of plant growth-promoting hormone, IAA. The ACC deaminase was observed in 13 strains (16%) by their growth on Dworking and Foster (DF) minimal medium containing ACC. Of the 80 strains, 22 strains (28%) showed proteolytic activity by inducing clear zones around the cells on skim milk agar medium, 7 strains (9%) showed chitinase activity by inducing clear zones around the cells on chitin agar medium, 33 strains (41%) produced pectinase and 25 strains (31%) produced cellulase [See Additional file 1].

Of the 80 phosphate solubilizing strains, 33 strains (41%) showed antifungal activity towards phytopathogenic fungi used in the study [See Additional file 3]. Strains induced growth-free inhibition zones (diameter) ranging from 10 to 36 mm towards phytopathogenic fungi.

Genomic DNA of strains when tested as templates using gene-specific primers, 12 strains (FPB16, FPB17, FPB51, FP2, FP3, FP5, FP7 FP10, FP14, Pw70, Pw71 and Pw72) amplified the DNA fragment of 745-bp of DAPG, 10 strains (FPB15, FPB16, FPB74, FPB75, FP2, FP3, FP5, FP7, FP14 and FP15) amplified the DNA fragment of 719-bp of PRN, 6 strains (FP2, FP3, FP5, FP7, FP9 and FP14) amplified DNA fragment of 779-bp of PLT and 6 strains (FPB18, FPB52, FPB75, Pw60, Pw61 and FP15) amplified DNA fragment of 1,150-bp of PCA (C, D).

The putative antibiotic strains identified by PCR were subjected to fermentation and antifungal metabolites such as DAPG (yellowish white), PCA (greenish yellow), PRN (light yellow) and PLT (yellowish white) were detected in the production cultures. TLC and HPLC analyses confirmed the production of PCA, DAPG, PLT and PRN by the strains. The retardation factor (Rf) values were found to be 0.77 for DAPG, 0.53 for PCA, 0.80 for PRN, 0.50 for PLT as determined by co-migration with pure standards. Strains grown in fermentation cultures yielded PCA (20 to 1124 ?g ml^-1), DAPG (13 to 93 ?g ml^-1), PLT (3 to 9 ?g ml^-1) and PRN (3 to 41 ?g ml^-1). HPLC analyses of active fractions of antibiotics yielded similar retention time to that of the standard antibiotics. PCA was detected at 257 nm with a retention time 4.94 min, DAPG was detected at 270 nm with a retention time 10.77 min, PLT was detected at 255 nm with a retention time of 17.6 min and PRN was detected at 220 nm with a retention time of 27.5 min. Of the 80 strains, 20 strains (25%) produced HCN [See Additional file 1].

The financial support from Department of Biotechnology (DBT), New Delhi, India, is gratefully acknowledged.