Organic acid production in vitro and plant growth promotion in maize under controlled environment by phosphate-solubilizing fluorescent Pseudomonas

Background Phosphorus deficiency is a major constraint to crop production due to rapid binding of the applied phosphorus into fixed forms not available to the plants. Microbial solubilization of inorganic phosphates has been attributed mainly to the production of organic acids. Phosphate-solubilizing microorganisms enhance plant growth under conditions of poor phosphorus availability by solubilizing insoluble phosphates in the soil. This paper describes the production of organic acids during inorganic phosphate solubilization and influence on plant growth as a function of phosphate solubilization by fluorescent Pseudomonas. Results Nineteen phosphate-solubilizing fluorescent Pseudomonas strains of P. fluorescens, P. poae, P. trivialis, and Pseudomonas spp. produced gluconic acid, oxalic acid, 2-ketogluconic acid, lactic acid, succinic acid, formic acid, citric acid and malic acid in the culture filtrates during the solubilization of tricalcium phosphate, Mussoorie rock phosphate, Udaipur rock phosphate and North Carolina rock phosphate. The strains differed quantitatively and qualitatively in the production of organic acids during solubilization of phosphate substrates. Cluster analysis based on organic acid profiling revealed inter-species and intra-species variation in organic acids produced by Pseudomonas strains. The phosphate-solubilizing bacterial treatments P. trivialis BIHB 745, P. trivialis BIHB 747, Pseudomonas sp. BIHB 756 and P. poae BIHB 808 resulted in significantly higher or statistically at par growth and total N, P and K content over single super phosphate treatment in maize. These treatments also significantly affected pH, organic matter, and N, P, and K content of the soil. Conclusion The results implied that organic acid production by Pseudomonas strains is independent of their genetic relatedness and each strain has its own ability of producing organic acids during the solubilization of inorganic phosphates. Significant difference in plant growth promotion by efficient phosphate-solubilizing Pseudomonas strains point at the need for selecting potential strains in plant growth promotion experiments in conjunction with various phosphate substrates for their targeted application as bioinoculants.


Background
Phosphorus (P) is an essential macronutrient often limiting the plant growth due to its low solubility and fixation in the soil. Improving soil fertility by releasing bound phosphorus by microbial inoculants is an important aspect for increasing crop yield. Phosphorus release from insoluble phosphates reported for several soil microorganisms has been attributed mainly to the production of organic acids and their chelation capacity [1][2][3]. Direct periplasmic oxidation of glucose to gluconic acid is considered as the metabolic basis of inorganic phosphate solubilization by many Gram-negative bacteria as a competitive strategy to transform the readily available carbon sources into less readily utilizable products by other microorganisms [1,4].
Increased solubilization of fixed soil phosphates and applied phosphates ensuring higher crop yields has been reported on inoculation of phosphate-solubilizing bacteria including Pseudomonas, Bacillus, Rhizobium, Micrococcus, Flavobacterium, Burkholderia, Achromobacter, Erwinia, and Agrobacterium [5,6]. Several Pseudomonas species have been reported among the most efficient phosphate-solublizing bacteria and as important bio-inoculants due to their multiple biofertilizing activities of improving soil nutrient status, secretion of plant growth regulators, and suppression of soil-borne pathogens [5,[7][8][9].
Soils in the cold deserts of Lahaul and Spiti in the Indian trans-Himalayan region latitude 31° 44' 57" and 32° 59' 57" North and 76° 46' 29" and 78° 41' 34" East are often with low moisture status, high alkalinity, and low organic matter content. The nutritional problems in such soils are often specific in respect of the low phosphorus availability resulting from their high phosphorus-fixing capacity due to high calcium content [10]. The vast potential of microorganisms for improving productivity in the region remains unexploited [11]. Previously we have reported the isolation, selection, and characterization of stress-tolerant and efficient phosphate-solubilizing fluorescent Pseudomonas from the cold deserts of the Himalayas [8,9]. The aim of the present study was to explicate organic acid production during solubilization of inorganic phosphates and effect on plant growth as a function of phosphate solubilization by fluorescent Pseudomonas.

Bacterial strains
Nineteen phosphate-solubilizing fluorescent Pseudomonas included in the present studies were isolated from the rhizosphere of Hippophae rhamnoides growing in the cold deserts of Lahaul and Spiti in the trans-Himalayas and characterized based on their phenotypic characters and 16S rDNA gene sequencing [8,9]. The bacterial strains were maintained at -70°C in nutrient broth supplemented with 20% (v/v) glycerol.

Production of organic acids during phosphate solubilization
The bacterial strains grown in triplicate in 10 ml NBRIP broth supplemented with 0.5% tricalcium phosphate (TCP), Mussoorie rock phosphate (MRP), Udaipur rock phosphate (URP) and North Carolina rock phosphate (NCRP) at 28°C for 5 days at 180 rpm in a refrigerated incubator shaker (Innova Model 4230, New Brunswick Scientific, USA) were centrifuged at 10,000 rpm for 10 min. and passed through 0.22 μm nylon filter. Quantitative estimation of P-liberated from inorganic phosphates was done using vanado-molybdate method as described earlier [8]. Detection and quantification of organic acids was done on Waters 996 High Performance Liquid Chromatogram (HPLC) equipped with PDA detector, Waters 717 plus autosampler, Waters 600 controller, Waters™ pump, Waters inline degasser AF, and Lichrosphere RP-18 column 250 mm × 4.6 mm and 5 μm particle size (Merck, Germany). The mobile phase was 0.1% ortho-phosphoric acid (Merck, Germany) in the gradient of flow rate as given in Table 1. Eluates were detected at λ 210 nm and identified by retention time and co-chromatography by spiking the sample with the authentic organic acids. The organic acids were quantified by reference to the peak areas obtained for the authentic standards for gluconic acid (Sigma-Aldrich, USA), 2-ketogluconic acid (Sigma, USA), and lactic acid, oxalic acid, malic acid, succinic acid, formic acid, citric acid, malonic acid, propionic acid and tartaric acid (Supelco, USA). Each replicate was analyzed in a single run on HPLC for 76 samples for the four phosphate substrates. The values were presented as the mean of three replicates.

Inoculum preparation
The bacterial strains grown in 20 ml trypticase soya broth (TSB) for 48 h at 28°C were centrifuged at 10,000 rpm for 10 min. and the pellets suspended in 0.85% NaCl (OD 600 = 1.0). The bacterial suspensions were separately mixed with sterilized activated charcoal (4:6 v/w) to give a CFU of approximately 10 7 /g of charcoal-based bacterial inoculants.

Plant growth under controlled environment
Seeds of Zea mays var. Girija surface sterilized with 20% sodium hypochlorite for 3 min. and washed thrice with sterile distilled water were germinated at 25°C in moist sterile vermiculite. Uniformly germinated seeds were coated with the water slurry of charcoal-based microbial

Chemical analyses
The soil samples were air dried and sieved for determining pH, available N, P, K, Ca and organic matter content. The plant samples were oven-dried and powdered for estimation of total N, P and K. Organic matter was determined by the modified Walkley and Black method [12]. Estimation of total N was done by modified Kjeldahl's method, total P by vanado-molybdate yellow colour method, total and available K by flame photometric method, and available Ca in ammonium-acetate extracts [13]. Estimation of available P was estimated by sodium biocarbonate method [14] and available N by alkaline permanganate method [15].

Experimental design and data analyses
Randomized block design with two factor factorial arrangement was adopted for conducting the experiments. The data were subjected to one-way analysis of variance (ANOVA) and the mean of treatments compared by Duncan's Multiple Range Test at p ≤ 0.01 using SPSS Software version 7.5. Cluster analysis based on the organic acid profiles was performed using STATISTICA data analysis software system version 7 (StatSoft ® Inc. Tulsa, USA, 2004).

Production of organic acids
HPLC analysis of the culture filtrates was done to identify and quantity the organic acids produced during the solubilization of TCP, MRP, URP and NCRP by Pseudomonas fluorescens strain, three Pseudomonas poae strains, ten Pseudomonas trivialis strains, and five Pseudomonas spp. strains (Fig. 1). During TCP solubilization all strains showed the production of gluconic and 2-ketogluconic acids ( Table  2). Apart from one Pseudomonas sp. strain no other strain showed oxalic acid production. All strains exhibited the production of malic acid excepting one Pseudomonas sp. strain and succinic acid excluding one Pseudomonas sp. strain. The production of lactic acid was restricted to one strain of both P. trivialis and Pseudomonas sp., formic acid to six P. trivialis, P. fluorescens and two Pseudomonas spp. strains, and citric acid to three P. trivialis strains and one strain each of P. poae and Pseudomonas sp., and P. fluorescens strain.
During URP solubilization the production of oxalic and gluconic acid was detected for all the strains ( Table 3). The production of other organic acids was restricted to some strains: 2-ketogluconic acid to three Pseudomonas spp. strains and one strain each of P. trivialis, P. poae and P. fluorescens; lactic acid to five P. trivialis, P. fluorescens and two Pseudomonas spp. strains; succinic acid to one strain each of P. trivialis, P. fluorescens and Pseudomonas sp.; formic acid to two P. trivialis strains; and malic acid to four P. trivialis, two P. poae and four Pseudomonas spp. strains. None of the strains showed citric acid production during URP solubilization.
During MRP solubilization the production of oxalic and gluconic acid was also detected for all the strains ( Table  4). The production of 2-ketogluconic acid was shown by one Pseudomonas poae, P. fluorescens and four Pseudomonas spp. strains, lactic acid by five P. trivialis, one P. poae and three Pseudomonas spp. strains, succinic acid by three Pseudomonas spp. strains, formic acid by three P. trivialis and three Pseudomonas spp. strains, formic acid by P. fluorescens and three P. trivialis strains, malic acid by two P. trivialis, one P. poae, P. fluorescens and four Pseudomonas spp. strains, and citric acid by one Pseudomonas sp. strain.
In NCRP solubilization the production of oxalic acid and gluconic acid was detected for all the strains (   Values are the mean of three replicates ± standard error of the mean; ND = not detected; 2-KGA = 2-ketogluconic acid. Values are the mean of three replicates ± standard error of the mean; ND = not detected; 2-KGA = 2-ketogluconic acid. Values are the mean of three replicates ± standard error of the mean; ND = Not detected; 2-KGA = 2-ketogluconic acid.  Cluster analysis based on the organic acid profiles during TCP, URP, MRP and NCRP solubilization generated Pseudomonas groups with strains belonging to the same or different species (Fig. 2). For TCP solubilization a single cluster was obtained at 2000 linkage distance, while Pseudomonas sp. BIHB 751 and Pseudomonas sp. BIHB 811 stood outside the cluster (Fig. 2a). Pseudomonas sp. BIHB 751 differed from the other strains in producing oxalic acid, lack of succinic acid production, and producing the lowest quantity of gluconic acid and the highest quantity of 2-ketogluconic acid. Pseudomonas sp. BIHB 811 showed dissimilarity in not producing malic acid. In URP solubilization a single cluster of three sub-clusters and single branches of Pseudomonas sp. BIHB 811, P. trivialis BIHB 769 and P. fluorescens BIHB 740 were formed at 2000 linkage distance, while Pseudomonas sp. BIHB 751 and P. trivialis BIHB 763 stood independently outside the cluster (Fig. 2b). Pseudomonas sp. BIHB 751 differed in producing the lowest quantity of gluconic acid and the highest quantities of 2-ketogluconic and malic acids. Pseudomonas trivialis BIHB 763 was separate from other strains in producing the highest quantities of gluconic and formic acids (Fig. 2b). During MRP solubilization a single cluster including six sub-clusters and two single branches of P. trivialis BIHB 745 and P. poae BIHB 752 were observed at 2000 linkage distance. Pseudomonas sp. BIHB 751 stood separately outside the cluster in producing the lowest quantity of gluconic acid and the highest quantity of malic acid (Fig. 2c). In NCRP solubilization P. trivialis BIHB 747, Pseudomonas sp. BIHB 751 and Pseudomonas sp. BIHB 811 stood outside the cluster as independent branches at 600 linkage distance (Fig 2d). The cluster incorporated 5 sub-clusters and separate branches of Pseudomonas sp. BIHB 740 and P. trivialis BIHB 759. Pseudomonas trivialis BIHB 747 differed in the highest gluconic acid production, Pseudomonas sp. BIHB 751 in the highest malic acid production, and Pseudomonas sp. BIHB 811 in producing the lowest quantity of gluconic acid and the highest quantity of 2-ketogluconic, lactic, and succinic acids.

Influence on plant growth
Significant difference was observed for the growth parameters in maize among PSB treatments and uninoculated control treatments (

Plant NPK content
The treatments showed significant difference in the nutrient content of roots and shoots ( Table 6). The shoot N was statistically higher in seven PSB treatments over NP 0 K and two PSB treatments over NP 0 K, NP TCP K and NP SSP K. A non-significant difference in the shoot N was observed with NP 0 K, NP TCP K and NP SSP K. The shoot P was significantly higher in ten PSB treatments over NP 0 K, NP TCP K and NP SSP K. The highest P content obtained with NP TCP K+Pt BIHB 745. The treatments NP TCP K and NP SSP K were statistically at par with NP 0 K. The shoot K was significantly higher in NP TCP K+Psp BIHB 756, NP TCP K+Pt BIHB 759 and NP TCP K+Pt BIHB 745 over NP 0 K, NP TCP K and NP SSP K.
The root N was significantly higher in eight PSB treatments over NP 0 K, NP TCP K and NP SSP K. The N content was statistically at par in NP 0 K, NP TCP K and NP SSP K. The highest N was obtained with NP TCP K+Pt BIHB 736. The root P was significantly higher in three PSB treatments over NP SSP K. The maximum increase was obtained with NP TCP K+Pt BIHB 745, followed by NP TCP K+Pp BIHB 752 and NP TCP K+Psp BIHB 756. The P content was significantly higher in NP SSP K over NP 0 K and NP TCP K. The root K was significantly higher in NP TCP K+Pt BIHB 745 and NP TCP K+Pt BIHB 728 over NP 0 K, NP TCP K and NP SSP K. Other treatments were statistically at par with NP TCP K and NP SSP K.

Soil properties
The soil pH, organic matter and available N, P, K contents were significantly affected by PSB treatments ( Table 7). The final pH with non-significant difference among various treatments was less than the initial pH. The highest decrease recorded with NP TCP K+Pt BIHB 757 was statisti-cally at par with all other PSB treatments but significantly lower than NP 0 K, NP TCP K and NP SSP K. The organic matter content was significantly higher in four PSB treatments than NP 0 K, NP TCP K and NP SSP K.
The soil N content was significantly higher in five PSB treatments than NP 0 K, NP TCP K and NP SSP K and statistically at par among NP 0 K, NP TCP K and NP SSP K. The soil P content was significantly higher in three PSB treatments over NP 0 K, NP TCP K and NP SSP K. The highest available P content was obtained with NP TCP K+Pt BIHB745 among PSB treatments and with NP SSP K among uninoculated treatments. The soil K content was significantly higher in nine PSB treatments than other PSB treatments, NP 0 K, NP TCP K and NP SSP K. The highest available K was recorded for NP TCP K+Psp BIHB 756. The available Ca was significantly higher in three PSB treatments than other PSB treatments, NP 0 K, NP TCP K and NP SSP K.

Discussion
The organic acid production during solubilization of inorganic phosphates by the efficient phosphate-solubilizing strains of Pseudomonas trivialis, Pseudomonas poae, Pseudomonas fluorescens and Pseudomonas spp., corroborated their involvement in phosphate solubilization [1,3,6].
The results indicated that the quantity of organic acids produced differed with the nature of phosphate substrates and Pseudomonas strains (Tables 2, 3 , 4, 5). The higher solubilization of TCP than URP, MRP and NCRP could possibly be due to the higher gluconic acid production in presence of TCP. The lower production of gluconic acid and lower TCP solubilization by Pseudomonas sp. BIHB 751 than other Pseudomonas strains substantiated the involvement of gluconic acid in solubilization of calciumbound phosphates. Succinic acid also appeared contribut- ing to TCP solubilization as it was produced by high TCPsolubilizing strains and not by low TCP-solubilizing Pseudomonas sp. BIHB 751 strain. The lack of oxalic acid production by efficient phosphate-solubilizing Pseudomonas strains signified non involvement of oxalic acid in TCP solubilization though this acid has been implicated besides citric, gluconic, lactic and succinic acids in phosphate solubilization in alkaline vertisols [20]. Pseudomonas sp. strain BIHB 751 producing the highest quantity of 2-ketogluconic acid but showing the lowest TCP and URP solubilization also differed from Enterobacter intermedium reported for the enhanced phosphate solubilization with increasing 2-ketogluconic acid production [21]. Likewise, no relationship could be ascertained between the quantity of organic acids produced and the solubilization of rock phosphates by Pseudomonas strains as the highest solubilization observed for NCRP among the rock phosphates was coupled to the lowest production of total organic acids (Tables 3, 4, 5). Previously also the quantities of solubilized phosphorus could not be correlated with the quantities of organic acids in the culture medium [22]. UPR, MRP and NCRP have fluorapatite structure with the highest substitution of phosphate with carbonate in NCRP [23]. The higher solubilization and lowered quantities of organic acids detected in the presence of NCRP could be due to the higher reactivity and greater diversion of organic acids in the neutralization of free carbonates in the solubilization of NCRP as compared to MRP and URP [23,24]. Likewise, the higher solubilization and higher production of organic acids in the presence of TCP could be attributed to its amorphous nature with simple structure and absence of any free carbonates as compared to the crystalline lattice structure of the rock phosphates [25].
Cluster analysis of organic acid profiles generated different groups revealing inter and intra-specific variation in the production of organic acids by Pseudomonas strains (Fig. 2). The strains clustered together and those standing outside the clusters or sub-clusters belonged to different Pseudomonas species characterized previously by 16S rRNA gene sequencing [8,9]. The strains standing outside the clusters differed qualitatively and/or quantitatively from other strains in the production of organic acids (Tables 2,  3, 4, 5). The results implied that Pseudomonas strains are independent of their genetic relatedness in their phosphate-solubilizing ability and organic acid production even under similar set of culture conditions. Phosphate solubilization is a complex phenomenon which depends on the nutritional, physiological and growth conditions of the culture [26].
The enhanced growth and higher N, P and K contents in maize with PSB treatments underlined the advantage of phosphate-solubilizing activity of microorganisms for plant growth promotion (Table 6 and 7). The increased growth and P uptake have been reported on PSB inoculations with Pseudomonas sp. and Serratia marcescens in maize [17], Pseudomonas fluorescens in peanut [27], Bacillus circulans in mungbean [28] and Pseudomonas sp. in wheat [29]. The TCP solubilization in soil by fluorescent Pseudomonas strains as evidenced by in vitro TCP solubilization, increased soil P availability and higher plant P content would be useful particularly in the cold deserts of Lahaul and Spiti where soil P deficiency is attributed mainly to the reaction of P with calcium carbonate and calcium sulphate forming insoluble di-and tricalcium phosphates. The rock phosphates recommended for acid soils are reportedly not effective in alkaline soils as P source for the crops [30]. The significantly higher plant growth and N, P, and K content in plant tissues and soil with some PSB treatments over NP SSP K might be due to the immobilization of applied P by native soil microbiota and physico-chemical reactions in the soil. The increased and continuous P availability in the soil promotes biological nitrogen fixation [27]. No correlation among TCP solubilization, production of organic acids and plant growth promotion could be established as the highest solubilization and plant growth promoting activity was observed for P. trivialis BIHB 745 not showing the highest organic acid production. However, the lowest organic acid production and plant growth promotion by Pseudomonas sp. BIHB 751 showing the lowest TCP solubilization suggested that phosphate solubilization is an important mechanism of plant growth promotion. Pseudomonas strains exhibiting high TCP solubilization in vitro differed significantly in enhancing the plant growth in the soil indicating interplay of some other growth factors besides phosphate-solubilization (Tables 2, 6, and 7). Apart from making P available to the plants, phosphate-solubilizing microorganisms improve plant health directly by the production of phytohormones [31]. Pseudomonas strains have been reported to vary in their ability for phytohormone production [32][33][34]. The bacterial strains also differ in utilizing root exudates in producing biologically active substances and root colonizing ability known to influence the plant growth-promoting action of rhizobacteria [35]. Plant-microbe interaction is a complex phenomenon with the interplay of several mechanisms and environmental factors.
The decrease in soil pH in PSB treatments indicated the production of organic acids by Pseudomonas strains as also reported for phosphate-solubilizing Aspergillus niger and A. tubingensis [36]. However, less pH decline in soil during plant growth promotion experiments than phosphate solubilization in culture medium could be due to the buffering nature of soil [20]. The inorganic acids and H + ions of microbial origin and H + ions released from the plant roots during ammonium assimilation are also reported to influ-ence the soil pH [22,30,37]. The studies have shown potential for plant growth promotion by P. trivialis BIHB 745, P. trivialis BIHB 747, Pseudomonas sp. BIHB 756 and P. poae BIHB 808 in the presence of TCP as the phosphate source. The native phosphate-solubilizing and stress-tolerant Pseudomonas strains are expected to cohabitate as effective microbial inoculants with the crops grown in the cold deserts of Lahaul and Spiti.

Conclusion
The present study revealed that the innate ability of organic acid production by Pseudomonas strains is independent of their genetic relatedness. Significant difference in plant growth promotion among the efficient phosphate-solubilizing Pseudomonas strains point at the need for selecting the potential strains based on plant growth promotion in the soils supplemented with insoluble phosphates for their targeted application. The PSB strains with high potential for TCP solubilization appear promising for application in the Ca-rich and P-deficit soils in the cold deserts of Lahaul and Spiti for which field studies are required.

Authors' contributions
PV carried out the experiments on phosphate solubilization, organic acid profiling, plant growth promotion and chemical analyses, data analyses, and manuscript writing. AG contributed in experimental designing, interpretation of results, co-ordination and supervision of the experimental work, manuscript writing and editing.