Artificial Citrate Operon Confers Mineral Phosphate Solubilization Ability to Diverse Fluorescent Pseudomonads

Citric acid is a strong acid with good cation chelating ability and can be very efficient in solubilizing mineral phosphates. Only a few phosphate solubilizing bacteria and fungi are known to secrete citric acids. In this work, we incorporated artificial citrate operon containing NADH insensitive citrate synthase (gltA1) and citrate transporter (citC) genes into the genome of six-plant growth promoting P. fluorescens strains viz., PfO-1, Pf5, CHAO1, P109, ATCC13525 and Fp315 using MiniTn7 transposon gene delivery system. Comprehensive biochemical characterization of the genomic integrants and their comparison with plasmid transformants of the same operon in M9 minimal medium reveals the highest amount of ∼7.6±0.41 mM citric and 29.95±2.8 mM gluconic acid secretion along with ∼43.2±3.24 mM intracellular citrate without affecting the growth of these P. fluorescens strains. All genomic integrants showed enhanced citric and gluconic acid secretion on Tris-Cl rock phosphate (TRP) buffered medium, which was sufficient to release 200–1000 µM Pi in TRP medium. This study demonstrates that MPS ability could be achieved in natural fluorescent pseudomonads by incorporation of artificial citrate operon not only as plasmid but also by genomic integration.


Introduction
Phosphorous (P) is one of the major nutrients limiting plant growth. Soluble P-ion concentration in most soils varies from 0.1 to 10 mM while P required for optimal growth ranges from 1 to 60 mM [1]. 70% of the phosphate in soils is present in complexes with cations, which is unavailable to plants. Rock phosphates are converted into soluble forms by using strong inorganic acids and provided for increased plant growth and crop yields. However, there is gradual and rapid depletion of global high-grade P resources, which could pose a major challenge for sustaining the agricultural productivity [2][3][4]. Increasing the efficiency of P application and use of low grade P reserves could delay the depletion of phosphate reserves. Phosphate solubilization using microorganisms has been proposed to be effective for this purpose. However, variable efficacies in field conditions have become a major hurdle in harvesting their potentials [5]. Soil properties, nature of rock phosphate, nature of the root exudates, carbon catabolite repression and presence of plasmid are known to decrease the effectiveness of phosphate solubilizing microorganisms (PSMs). However, secretion of strong organic acids by native rhizobacteria results in higher efficacy of PSMs in field conditions [6].
The secretion of gluconic acid by direct oxidation of glucose is the major mechanism of P-solubilization by most Gram-negative bacteria [7]. Gluconic acid biosynthesis is carried out by periplasmic glucose dehydrogenase (GDH) enzyme involving pyrroloquinoline quinone (PQQ) as the co-factor. Incorporation of genes involved in PQQ biosynthesis and transport in non-PSMs resulted in increased gluconic acid secretion and mineral phosphate solubilizing (MPS) ability [8,9]. However, effectiveness of their MPS ability using gluconic acid is limited by the amount of aldose sugars naturally present in the root exudates. One potential solution for this problem is using stronger organic acids for P solubilization. For instance, in alkaline vertisol soils, P release is dependent on the nature and amount of organic acids in the order Acetic = Succinic = Lactic ,,Gluconic ,,Oxalic ,Tartaric = Citric [10]. Thus, rhizobacteria secreting citric acid could be more effective in solubilizing P in alkaline vertisols. An ideal rhizobacteria that can be employed for this purpose is fluorescent pseudomonads which possess very good plant growth promotion abilities including biocontrol againstbacterial and fungal plant pathogens by secretion of a variety of compounds [11]. Hence, fluorescent pseudomonads producing citric acid could be very effective as P biofertilizers in alkaline soils.
Previously, our studies demonstrated accumulation of high levels of citric acid but low secretion upon heterologous overexpression of phosphoenolpyruvate carboxylase (ppc) and citrate synthase (cs) genes in Pseudomonas fluorescens ATCC13525 [12,13].
Overexpression of these genes independently and together showed that CS activity chiefly regulates citric acid biosynthesis. CS enzymes in E. coli and Pseudomonas sp. are strongly inhibited by NADH [14][15][16]. Secretion of low citric acid, in spite of high intracellular citrate, suggested that native H + dependent citrate transporter is not effective in citrate efflux, though it is very efficient in citrate uptake as reflected in terms of supporting the growth of the P. fluorescens ATCC 13525 on citrate as sole carbon source. Thus citric acid secretion could be dependent on the amount of intracellular citrate as well as the nature of citrate transporters. Citrate utilization in Salmonella typhimurium is mediated by Na + dependent transporters (citC) belonging to hydroxy carboxylate transporter family [17]. Citrate in complex with Mg 2+ is the major citrate-uptake system (CitM) in soil bacterium Bacillus subtilis during growth on citrate under aerobic conditions [18,19]. The CitC and CitM are best characterized transporters used for citrate utilization. These transporter proteins are involved in pseudosymmetrical transport, can alternate between atleast two preferred conformations [20,21]. Functional characterization and structural determination of the Na + dependent dicarboxylate transporter VcINDY from Vibrio cholera also revealed the possible reversibility [22]. Therefore, the decarboxylase systems could also function in reversible manner wherein the direction of operation depends on the cation gradient and free energy change under the conditions of the physiological steady state [23]. Additionally, regulation of citric acid biosynthesis in bacteria is governed by anabolic and catabolic pathways at the anaplerotic node comprising of phosphoenol pyruvate-oxaloacetate-pyruvate [24] which makes any targeted genetic modification unstable. Nature and number of the enzymes involved at the anaplerotic node in fluorescent pseudomonads vary in different strains reflecting their catabolic diversity. Thus, it is necessary to understand the effect of alterations carried out at the anaplerotic node in different strains for universal application of the system. Present study demonstrates the selection of suitable NADH insensitive cs and citrate transporter genes for citric acid secretion and developing plasmid transformants and genomic integrants containing artificial citrate operon of the selected genes under lac promoter in several fluorescent pseudomonads. The effect of plasmid transformants and genome integrants on citric acid secretion and MPS ability demonstrates that genetic modifications can be effective strategy even in strains with variable metabolic traits.

Bacterial Strains, Plasmids and microbiological methods
The bacterial strains and plasmids used in this work are summarized in Table 1. All E. coli strains were grown in Luria-Bertani medium [25] at 37 uC unless otherwise stated. Media were supplemented with ampicillin (100 mg ml 21 ), chloramphenicol (30 mg ml 21 ); kanamycin (50 mg ml 21 ), gentamycin (50 mg ml 21 ) and tetracycline (20 mg ml 21 ) as required. For routine biochemical characterization, Pseudomonas fluorescens and its recombinants were grown on M9 minimal medium supplemented with a micronutrient solution and iron with glucose [0.5% (w/v)], as the carbon source as well as TRP buffered medium (with 75 mM Tris HCl, pH 8.0) containing 75 mM glucose. For the minimal media the concentrations of the antibiotics were reduced to one fourth of the concentrations added to the rich media mentioned above. Cultures were incubated at 30uC and shaken on an orbital platform (Orbitek, Scigenics Biotech, Chennai, India) operating at 200 strokes per minute. All strains were stored at 220uC in MicroBank vials and subcultured not more than twice prior to experimentation. Agar (1.5% w/v) was added to solidify the bacteriological media when necessary before sterilization. Aerobic growth was carried out under shaking in Erlenmeyer flasks containing 0.1 volumes of medium. Growth was measured spectrophotometrically at 600 nm.

Molecular Biology techniques
All molecular biology procedures such as plasmid extraction, transformation and in vitro manipulations of DNA, PCR amplification were carried out according to standard protocols [25]. Taq DNA polymerase and its buffer, dNTPs, PCR primers and other enzymes for DNA in vitro manipulations were obtained from Genei Merck Pvt. Ltd., Bangalore, India and Sigma Chemicals Pvt. Ltd., Bangalore, India, respectively, and were used according to manufacturer's instructions.

Construction of plasmids containing CS and Citrate transporter and artificial citrate operon
The E. coli NADH insensitive citrate synthase (CS) encoding gene gltA, with three different modifications viz gltA1, gltA2 and gltA3 were obtained as 1281 bp fragments flanked by SacI and BamHI sites by PCR amplification of plasmid DNA carrying the respective genes using pairs of primers: 59CGAGCTCGGGCC-CTTTTTCACGGAGGAAACCACAATG GCT GAT ACA A-AA GC 39 and 59CGGGATTCCGGATCCGTTA ACG CTT GAT ATC GC 39. Ribosomal binding site corresponding to Pseudomonas sp. has been added (underlined) while restriction enzymes SacI and KpnI in forward and ApaI and BamHI site in reverse of cs primers (bold and italicized) were added. The appropriately digested fragments were then inserted in to SacI/ BamHI site within the pAB8 multiple cloning site (MCS). pY145F, pR163L and pK167A plasmids containing the NADH insensitive gltA genes gltA1, gltA2 and gltA3, respectively, in the correct orientation with respect to the lac promoter of pAB8 were selected by restriction enzyme mapping and PCR amplification using gene specific primers. A 1.3 kb HindIII fragment of pGM160 plasmid containing gentamycin resistance gene was inserted at HindIII site of pUCPM18 plasmid to obtain pUCPM18Gm and used as a cloning vector for the citrate transporter genes. A 1281 bp DNA fragment containing citC gene was amplified from S. typhimurium genomic DNA with primer pair: GGGGGCCCCGGGATCCC-GCACGGAGGAATCAACTT ATG ACC AAC ATG ACC CA-G GCT TC {59 primer, ApaI/BamHI bold italicized, ribosomal binding site (RBS)underlined} and 59 GGGGTACCCCGCTC-TAGAGC TTA CAC CAT CAT GCT GAA CAC GAT GC (39 primer, KpnI/XbaI bold italicized). Similarly, a 1341 bp DNA fragment containing citM gene was obtained from B. subtilis genomic DNA using primer pair: 59 GGGGGCCCCGGGATC-CCGCACGGAGGAATCAACTT ATG TTA GCA ATC TTA GGC TTT CTC ATG ATG(59 primer, ApaI/BamHI bold italicized, RBS underlined) and 59 GGGGTACCCCGCTCTA-GAGC TTA TAC GGA AAT AGA GAT CGC ACC G (39 primer, KpnI/XbaI bold italicized). Taq DNA polymerase and its buffer, dNTPs and primers were obtained from Bangalore Genei Pvt. Ltd., India and Sigma Chemicals Pvt. Ltd., Bangalore, India, respectively, and were used according to manufacturer's instructions. The citC gene was blunted and inserted into EcoRV site of pBluescript KS and further subcloned in XbaI site in the MCS of pUCPM18Gm to generate pcitC. The citM gene was cloned into pTZ5/7R using InsT/Aclone PCR Product Cloning Kit, MBI Fermentas, (pJE4) and subcloned into EcoRI-SalI site in the MCS of pUCPM18Gm to generate pcitM plasmid. The presence of the appropriate plasmid was checked by PCR using gene specific primer and M13 primer and by restriction enzyme mapping. To construct the artificial citrate (yc) operon 1281 bp gltA1 gene was excised from pY145F with SacI-BamHI and ligated to the upstream of citC gene in pcitC yielding pYC.

Functional complementation study of gene constructs in E. coli
Functional expression of gltA genes was checked by complementation of glutamate auxotrophy of E. coli W620 CS mutant grown in minimal medium supplemented with IPTG (0.1 mM) and thiamine (100 mg/ml) with and without glutamate (340 mg/ ml). Functionality of citrate transporters was monitored by growth of E. coli DH5a transformants in Koser citrate broth supplemented with IPTG (0.1 mM). Development of P. fluorescens transformants harbouring the recombinant plasmids P. fluorescens PfO-1 was independently transformed with NADH insensitive plasmids by using the NaCl/CaCl 2 method [26] to obtain the recombinant strains expressing E. coli NADH insensitive cs gene. pAB7 and pAB8 transformants were used as a control in all experiments. P. fluorescens PfO-1 strains coexpressing the citrate transporter and gltA1 genes (pYFCitC or pYFCitM) were developed by transforming both the plasmids independently or as an operon in same plasmid to obtain pYC (with CitC transporter). The single and double transformants were selected on Pseudomonas agar containing accordingly kanamycin, gentamycin or both. pUCPM18Gm and pAB8 transformants were used as controls for all the experiments.

Genomic integration and PCR analysis of chromosomal insertion of citrate operon
For genomic integration, the citrate operon along with constitutive lac promoter was amplified using constitutive plac forward and citC reverse primers. The 2.7 kb fragment containing plac-gltA1-citC operon was blunted and inserted into the SmaI site of integration delivery plasmid AKN69 to generate pYC-Int plasmid. Prior to cloning the chloramphenicol resistance gene of the plasmid was disrupted using PvuII. The clone was confirmed by double digestion with EcoRI-XbaI restriction site and PCR. The construct along with the helper pUXBF13 plasmid containing the transposase genes were delivered into P. fluorescens strains by the electroporation method [27]. Electroporation was carried out using BioRad electroporator (Bio-Rad Laboratories Chennai (India) Pvt. Ltd.) with the following settings: 25 mF, 200 V, 2.5 kV (the time constant should be ,5 ms). Preliminary confirmation of the genomic integrants were done by monitoring the growth on gentamycin plate, fluorescence and reconfirmed by PCR using the primer pair: 59ATATCGACCCAAGTACCGCC (Tn7Gm, nt 509 from the start site of aaC1 gene) and 59CGGGATTCCGGATCCGTTAACGCTTGATATCGC (39 primer of gltA gene).

Physiological experiments
Fresh cultures grown with overnight shaking at 30uC in 3 ml Luria-Bertani broth were aseptically harvested, washed thrice with sterile 0.85% (w/v) saline and resuspended in 1 ml of the same under sterile conditions. The resultant bacterial cell suspension was used to inoculate M9 minimal medium supplemented with 100 mM glucose and micronutrient cocktail to give an initial cell density of OD 600 0.01-0.03 (,2.66106 c.f.u./ml). 75 mM Tris-Cl (pH 8.0) buffered medium with rock phosphate (TRP) was used to represent P deficient medium supplemented with 75 mM glucose. Inoculum for TRP medium was prepared in M9 medium as described earlier [28] and an initial cell density of 1.4610 8 cfu/ml was maintained. Batchculture studies were performed by shaking 150 ml conical flasks with 30 ml of the inoculated medium on an Orbitek rotary shaker (Orbitek, Scigenics Biotech, Chennai, India) at 30uC and 2006g.

Analytical techniques
Change in cell density (OD 600 nm) using Helios c spectrophotometer (Thermo Spectronics, Cambridge, UK) was considered as the measure of growth while drop in pH of the media was taken as the measure of acid production. The observations were continued till the medium pH reduced to less than 4 or O.D at 600 nm was greater than 2.1 ml Samples drawn at regular intervals were centrifuged at 9,2006g for 1 min at 4uC, filtered through 0.2 mm nylon membrane (MDI Advanced Microdevices, India) and used to measure residual glucose enzymatically (using the GOD-POD kit, Enzopak, Reckon Diagnostics Pvt. Ltd, India) and organic acid analysis using HPLC (Varian Microsorb RP-18 column) operated at room temperature using mobile phase of 0.01 M Na 2 HPO 4 + 5% acetonitrile at a flow rate of 0.2 ml min 21 , column effluents were monitored using a UV detector at 210 nm. The physiological parameters like growth rate, specific glucose depletion rate and biomass yield were calculated as described earlier [28]. Organic acid yields were expressed as grams of organic acid formed per gram of glucose depleted per gram of dry cell weight. Graph Pad Prism (version 3.0) for Windows, Graph Pad Software, San Diego, CA, USA (http://www.graphpad.com) and Microsoft Excel (Microsoft, 2003 Microsoft Excel computer software, Richmond, WA) were used for statistical analysis of the parameters. Each parameter has been represented as mean 6 SD or mean 6 SEM as specified in the legends to figures and tables. One-way ANOVA and Bonferroni's multiple comparison tests were performed for all treatments using Graph Pad prism software.

Preparation of cells/cell free extracts and enzyme assays
Citrate Synthase (CS, 4.1.3.7), Isocitrate dehydrogenase (ICDH, 1.1.1.42), glucose-6-phosphate dehydrogenase (G-6-PDH, EC 1.1.1.49) were assayed using both mid log and late log-stationary phase cultures whereas Isocitrate lyase (ICL, 4.1.3.1), glucose dehydrogenase (GDH) and pyruvate carboxylase (PYC) were assayed in late log to stationary phase. Preparation of cell free extracts and enzyme activities for G-6-PDH, PYC, ICDH and ICL assays as well as whole cell suspension for GDH assay were carried out as described earlier [12,28]. All enzymes were assayed at room temperature (3062uC) against appropriate controls without the substrate and enzyme source in the reaction mixture. Total protein concentration of the crude extract and whole cell suspensions was measured by a modified Lowry method [29] using bovine serum albumin as standard, along with appropriate corrections to rule out the interference of Tris [30]. One unit of specific enzyme activity was defined as the amount of protein required to convert 1 nmole of substrate per minute.

Mineral phosphate solubilization (MPS) ability
The dicalcium phosphate solubilizing ability of wild type and gltA transformants of fluorescent pseudomonads were tested on Pikovskaya's agar [31]. Three microliters of the bacterial inoculums prepared as described above, was aseptically spotted on the agar plates and was allowed to dry completely followed by incubation at 30uC. Preparation of bacterial inoculum for phosphate-solubilization studies was similar to that for the abovementioned physiological experiments. Rock phosphate solubilizing ability was monitored using 75 mM Tris buffered (TRP) minimal medium with 75 mM glucose. P-solubilization on Pikovskaya's agar was determined by monitoring the phosphate solubilization index (PSI) whereas drop in the medium pH from 8.0 to ,5 on TRP broth indicated by formation of a red zone on buffered Senegal rock phosphate. Methyl Red was used in agar plates to observe rock phosphate solubilization. The P liberated from rock phosphate in the TRP medium was measured by the Ames method [32].

Results
Overexpression of citrate synthase, transporter and artificial citrate operon in P. fluorescens PfO-1 Three gene variants of the gltA gene, encoding NADH insensitive CS were cloned independently in broad host range vector pUCPM18Gm under lac promoter and the functionality was confirmed by complementation of glutamate auxotrophy in E. coli W620 ( Figure S1). Amongst three cs mutants, P. fluorescens PfO-1 (pY145F) showed highest CS activity of 424.6616.1 U and 333.468.5 U in the mid log and stationary phase, respectively, on M9 minimal medium which is about 4.7 and 5.6 fold higher than that the control strain Pf (pAB8) (Figure 1). Hence, for further constructs of artificial citrate operon this mutant gene was used. The transformant accumulated 50 mM citrate intracellularly and secreted 3.2 mM ( Figure 2B). However, the specific growth rate, specific total glucose utilization rate and total amount of glucose utilized after 30 h remained unaffected. The amount of glucose consumed intracellularly was reduced by 1.49 fold in Pf (pY145F) as compared to Pf (pAB8). The increase in CS activity in Pf(pY145F) strain improved the biomass yield by 3.2, 2.38 and 2 fold compared to WT, Pf (pAB8) and Pf (pAB7), respectively ( Table 2).
E. coli DH5a transformants of citC and citM transporter cloned in Pseudomonas stable vector pUCPM18Gm could grow on Koser citrate broth medium containing citrate as a sole carbon source upon induction with 0.1 mM IPTG, demonstrating the functionality of the respective protein (data not shown). pYC plasmid was constructed by incorporating the csY145F and citC genes both under the lac promoter in pUCPM18Gm plasmid as an artificial citrate operon. After confirming the functionality in E. coli cs mutant by growth on citrate ( Figure S2), the plasmid was incorporated in P. fluorescens PfO-1. Extracellular citric acid levels in Pf(pYC) increased by 2.3 and 84.6 folds as compared to Pf (pY145F) and Pf(pAB8) strains, respectively ( Figure 2B). Corresponding extracellular citrate yield increased by 2.07 fold and 76 fold compared to respective controls. Thus the efficient expression of the genes encoding citrate producing enzyme under the lac promoter provided P. fluorescens with the ability to secrete citric acid. Stationary phase culture supernatents of Pf (pYC) showed 3.91 and 4.81 fold elevated gluconic and acetic acid levels, respectively, with concomitant 1.73 fold decreased pyruvic acid level as compared to Pf(pGm) ( Figure 2C).
In Pf (pYC), the periplasmic GDH activity increased by 1.46 fold as compared to the Pf (pGm) in late log to stationary phase of growth. Similarly, a significant increase in G6PDH, PYC and CS activities by 1.46, 4.74 and 4.5 fold, respectively, were observed as compared to the controls. However, ICL and ICDH activities in the stationary phase cultures remained unaltered (Figure 2A). Significant pH drop was found in case of Pf (pYC) compared to wild type and vector control strain. Specific growth rate, specific total glucose utilization rate after 30 h remained unaffected. Total amount of glucose utilized increased by 1.44 fold and glucose consumption is reduced by 1.27 fold in Pf (pYC) as compared to Pf (pGm). Pf (pYC) strain had improved the biomass yield by 2.58 fold compared to Pf (pGm) ( Table 2).
Growth and biochemical characterization of diverse P. fluorescens genomic integrants containing artificial citrate operon Artificial citrate operon was integrated into six P. fluorescens strains using Mini-Tn7 transposon system. The integrants were selected by fluorescence and gentamycin resistance ( Figure 3A). Presence of yc operon in the genome was confirmed by PCR amplification of 2.8 kb using specific primers ( Figure 3B).
The growth characteristics of P. fluorescens yc operon genomic integrants were determined upon growth under buffered-RP (TRP, 75 mM Tris-Cl pH 8.0) broth minimal medium containing 75 mM glucose. All genomic integrants had growth equivalent to wild type and plasmid transformant. On the other hand, P. fluorescens Fp315 genomic integrants showed an enhanced growth rate by 1.6 fold as compared to the plasmid transformants while growth rate of P. fluorescens Pf5 plasmid transformants was drastically reduced (1.7 fold) as compared to the wild type culture and genomic integrants ( Table S1). Genomic integrants of P. fluorescens PfO-1 and Fp315 strains acidified the medium pH below 5 within 48 h while other strains required 72-96 h.
P. fluorescens CHAO1genomic integrants showed an increase in total amount of glucose utilized by 1.36 fold and 1.25 fold as compared to the wild type and plasmid transformant, respectively, whereas the P. fluorescens P109 genomic integrant showed a 1.3 fold decrease in the total glucose utilization.Total glucose utilization remains unaltered in P. fluorescensPfO-1, Pf5 and ATCC 13525 integrants. The amount of glucose consumed decreased by 1.57 fold and 1.3 fold as compared to wild type and Pf(pYC) in P. fluorescensPfO-1 genomic integrant, respectively. Similar decrease was observed in P. fluorescens Pf5, ATCC13525 and Fp315 strains by 2.5, 1.26 and 2.26, respectively, as compared to wild type strain. On the other hand, an increased glucose consumption by 1.26 fold and 3 fold were observed in ATCC13525 and P109 genomic integrant, respectively, as compared to wild type strain. Glucose consumption remain unaltered in P. fluorescens CHAO1 genomic integrant. The biomass yield and specific glucose utilization rate were remained unaltered in P. fluorescens CHAO1 and ATCC13525 genomic  rate were monitored in Pf5 as compared to the wild type control. All these variations in results are shown in Table S1.
On TRP mediumcontaining 75 mM glucose, P. fluorescens CHAO1, ATCC13525 and Fp315 plasmid transformants and integrants, intracellular citrate level was increased upto 4.6 fold as compared to the control strain ( Figure 4A). However, the intracellular citrate levels were found to be similar in both genomic integrant and plasmid bearing strains. The CS activity correlated with the intracellular citrate levels. P. fluorescens Pf5(Int) and P109(Int) showed a similar increase in intracellular citrate level by 3.5 and 3.27 fold, respectively, as compared to the control. Exception of this effect was PfO-1(Int), in which the intracellular citrate level is 1.3 fold less as compared to PfO-1(pYC). This data when compared with the respective plasmid  bearing strain showed a significant reduction in citrate level by 1.27 and 1.5 fold, respectively. An enhancement of extracellular citrate levels was achieved in both plasmid bearing and genomic integrants of P. fluorescens strains when compared to their respective control Pf (pGm) strain ( Figure 4B). All the strains harboring citrate operon showed enhanced gluconic acid secretion up to 2 fold as compared to the vector control except Pf(P109) where decreased secretion was monitored ( Figure 4C). Genomic integration of citrate operon improved the gluconic secretion in Pf (CHAO1) and Pf (Fp315) as compared to their respective plasmid transformants.
The effect of genomic integration of plac yc operon on the key enzymes involved in periplasmic direct oxidation, intracellular phosphorylative pathway and anaplerotic reactions were monitored in P. fluorescens ( Table 3). GDH activity increased upto 2.7 fold in PfO-1(Int), Pf5(Int), CHAO1(Int) and Fp315(Int) as compared to the respective vector control strain whereas significicant decrease in GDH activity was observed in ATC-C13525(Int) (1.4 fold) and P109(Int) (2.76 fold). Improvement of GDH activity monitored in PfO-1(Int), Pf5(Int), CHAO1(Int) as compared to their respective plasmid control hence correlating with gluconic acid secretion. Reduction in G6PDH activity upto 50%was observed in both PfO-1 (Int) and PfO-1(pYC) as compared to control with exceptions in CHAO1 strain which showed a significant enhancement of G6PDH activity by 3 fold as compared to the vector control. The percentage change in G6PDH activity was correlated with change in GDH activity. Thus, the citrate operon genomic integrants showed the diversion of glucose flux more towards the direct oxidation pathway by increase in GDH activity with concommitant decrease in G6PDH activity as well as glucose consumption. However strain wise variation was also monitored.
All genomic integrants showed no alteration in ICDH and ICL activities with an exception of Pf5 (pYC) and Pf5 (Int) which showed an increase in ICDH activity by 1.7 fold and ICL activity by 2.7 fold, respectively, as compared to wild type strain. Interestingly, majority of genomic integrants of yc operon showed CS activity at par with the plasmid transformants. The CS activity is increased by 3.17 and 2.64 fold in P. fluorescensPfO-1yc operon genomic integrant and plasmid bearing strains, respectively. Similarly, the increase was 4.73 and 4.6 fold in Pf5, 4.1 and 2.7 fold in CHAO1, 6 fold each in ATCC13525, 1.9 and 2.2 fold in P109 and 3.4 and 2.65 fold in Fp315 as compared to the respective wild type controls. Remarkably, the CHAO1 genomic integrant showed 1.5 fold higher CS activity than that of CHAO1 (pYC) strain. PYC activity remain unaltered in P. fluorescensPf5, CHAO1, P109 and Fp315 genomic integrants whereas an increase by 5 fold and 2.8 fold in P. fluorescens PfO-1(Int) and ATCC13525(Int), respectively, were observed compared to control ( Table 3).

MPS ability of P. fluorescens yc transformants and genomic integrants
All yc bearing strains, monitored on both Pikovskaya's agar and TRP agar media, showed significant improvement on solubilization index as compared to their respective wild type control. Genomic integrants P. fluorescens PfO-1 and Fp315 yc operon showed maximum dicalcium phosphate solubization ability ( Figure 5A and 5B). Solubilization index of both the strains varied in the order of WT,pGm ,pYF ,pYC ,Intyc. On TRP medium, all cultures showed an improvement in the zone of coloration as compared to the wild type ( Figure 5C). Moreover, the genomic integrants of PfO-1, Pf5 and Fp315 showed faster rock phosphate solubilization as evident from the appearance of zone of coloration within 48-72 h of growth as compared to 96-120 h of plasmid transformants and wild type (Data not shown). Except Fp315 wild type, none of the wild type cultures could show zone of coloration on TRP medium containing 75 mM Tris HCl (pH 8.2) and 75 mM glucose. All P. fluorescens yc operon plasmid transformants and genomic integrants showed significant enhancement of Pi release as compared to the wild type strains with an exception of P109 wherein compared to the wild type no improvement of Pi release was observed. P. fluorescens Fp315 genomic integrant showed a maximum Pi release of 983.4658.78 mM which was even 1.37 fold higher as compared to plasmid transformants. In contrast, ATCC13525 genomic integrant showed a decrease in Pi level by 1.5 fold as compared to the plasmid transformants ( Figure 5D). The Pi release among the P fluorescens genomic integrants was in the increasing order of ATCC13525,P109 ,Pf5 ,CHAO1 , PfO-1,Fp315.

Discussion
In present study, metabolic engineering was employed to enhance citrate productionin Pseudomonas fluorescens to increase the mineral phosphate solubilization efficiency. In an earlier investigation, P. fluorescens ATCC13525 overexpressing E. coli wild type cs gene (allosterically inhibited by NADH) reported to increase CS activity by 2 fold with accumulation of 18.3 mM intracellular citrate [12]. Since native CS enzyme is subject to allosteric inhibition by NADH, it was hypothesized that overexpression of NADH insensitive variants of the CS would lead to further improvement of the citrate production in the genetically modified strains. Three site-directed allosteric cs mutants were reported with decreased affinity of NADH binding in the order R163L,,K167A ,,Y145F with a maximum Ki value of 7906210 mM for Y145F as against wild type CS (2.860.4 mM) [14]. In the present study, overexpression of these NADH insensitive gltA mutants in P. fluorescensPfO-1 and demonstrated increased CS activity in the order of pAB8,,pAB7#R163L# K167A ,Y145F ( Figure 1) and citrate accumulation in the order of pAB8,,pAB7,R163L,K167A ,Y145F ( Figure 2B) suggesting that enhanced citric acid biosynthesis in P. fluorescens PfO-1 is directly correlated with loss of NADH inhibition of CS enzyme. Earlier studies of overexpression of phosphoenolpyruvate carboxylase (ppc) and gltA genes alone and together showed that CS regulates citrate biosynthesis [13]. In addition, present study demonstrates that NADH mediated inhibition of CS activity regulates the flux of TCA cycle in P. fluorescens. Remarkably, the higher citrate accumulation in Pf (pY145F) did not hinder growth. Our study is supported by the fact that the mechanisms for pH sensing and cytoplasmic pH homeostasis enable most bacteria to tolerate or grow at pH values that are outside the cytoplasmic pH range they must maintain for growth [33].
Although extracellular citrate levels were enhanced in Pf (pY145F) as compared to Pf (pAB7), the amount secreted corresponds to only 6% of the intracellular level ( Figure 2B), Table 3. Activities of key glucose catabolic enzymes in P. fluorescens strains wild type (WT), plasmid transformants harbouring yc operon, Int (genomic integrants of yc operon) grown in TRP broth (pH 8.0) minimal media. suggesting limitation of the bacterial ability to secrete the acid into the extracellular medium. P. fluorescens possesses H + -dependent citrate transporter [34,35], which normally supports their growth on citrate as a sole carbon source. Our results suggest that this transporter seems to be inefficient in the efflux of citrate as also observed in our previous studies with the wild type citrate synthase gene [12]. The active pH homeostasis favors this in mesophilic bacteria [36]. The active transport system for citrate excretion appears to be the main rate-determining factor in citrate overproduction by yeasts [37]. Present study demonstrates that S. typhimurium CitC transporter is able to enhance the citrate excretion and is more efficient than B. subtilis CitM transporter suggesting that Na + ion gradient is much more favorable than Mg +2 ions for citrate efflux in fluorescent pseudomonads. Plasmid dependent metabolic load reported to cause alterations in the central carbon metabolism in P. fluorescens 13525 and lose in MPS ability in Enterobacter asburiae PSI3 [38,39]. Our study has also included genome integration instead of plasmid mediated expression to study the effect of the additional copy of cs and citrate transporter gene on metabolism avoiding the plasmid load effect. Genomic integration of citrate operon in six different P. fluorescens strains showed enhanced CS activity and citric acid secretion. This effect turned out to be similar or even better for genomic integrants compared to mid log to stationary phase cultures from plasmid transformants containing the same operon. This result suggests that the metabolic load of maintenance of the plasmid counteracts the beneficial effect of increased copy number. Overexpression of polyphosphate kinase (ppk) gene using low copy mini F plasmid was reported to cause a 20-fold increase in polyphosphate (poly P) content in the early stage of growth whereas the enhancement was 80% in the stationary phase [40]. On the other hand, profile using multicopy plasmid depicted a different behavior wherein PolyP increased dramatically within the first 3 h, but the polyP levels at 24 h were low and similar to that of low-copy plasmid.
The influence of metabolic diversity of catabolically versatile fluorescent pseudomonads overexpressing citrate operon was reflected in the qualitative and quantitative changes in their metabolism. Glucose catabolism in fluorescent pseudomonads consists of direct oxidative and phosphorylative pathways that determine the glucose utilization whereas TCA cycle flux is controlled at the anaplerotic node towards anabolic and oxidative phosphorylation pathways depending on the availability of electron acceptors. Variations in different strains arise from the nature and number of enzymes participating in these pathways especially at the anaplerotic node [24]. In our study, all genomic integrants showed similar effects but differed in a few parameters. Increased gluconic acid levels with concomitant reduction in pyruvic acid levels and increased PYC activity indicate probable diversion of pyruvate flux towards increased OAA biosynthesis in response to increased CS activity. Even in A. niger the enhancement of anaplerotic reactions replenishing TCA cycle intermediates predisposes the cells to form high amounts of citric acid [41]. Enhancement of biosynthetic reactions due to shortage of TCA cycle intermediates was also observed in citric acid accumulating E. coli K and B strains in the form of increased glyoxylate pathway [42,43]. However in our study, similar increase in flux through glyoxylate shunt was not apparent as evident from very low and unaltered ICL activity ( Table 3). Low ICL activity was consistent with earlier reports in P. fluorescens ATCC13525 and P. indigofera in which ICL contributed negligibly to glucose metabolism [12,44]. Enhanced CS activity also increased the periplasmic glucose oxidation, which is reflected, by increase in GDH activity and gluconic acid production.
Overexpression of citrate operon enhanced direct oxidative pathway, phosphorylative pathway but decreased TCA cycle flux and glyoxylate pathway in all fluorescent pseudomonads. Enhanced citric acid accumulation had no effect on anabolic process as evident from similar generation time of the bacteria ( Table 2). No change in the generation time of the transformants also suggests that the increased ATP generation via direct oxidative pathway appear to be coupled with decrease in ATP generation mediated by TCA flux.
Incorporation of citrate operon enhanced secretion of total citric and gluconic acids in all fluorescent pseudomonads that was correlated with enhanced MPS ability in both Pikovskaya's and TRP medium. Similar artificial citrate operon also enhanced P solubilization efficacy of Herbaspirillum seropedicae Z67 [45] and in Enterobacter hormaechei DHRSS [46]. The amount of citric acid secreted was similar in P. fluorescens PfO1 and E hormaechei DHRSS, but significantly lower in H. seropedicae Z67. However, genetically modified H. seropedicae Z67 enhanced rice plant growth. It will be interesting to know the effectiveness of Pseudomonas genomic integrants in supplementing P to plants. Recently, P. fluorescens 13525 containing an artificial oxalate operon was demonstrated to secrete oxalate and demonstrated good MPS ability in TRP and alfisols along with enhancement of mung bean plants grown in alfisols [47]. Thus, metabolic engineering approaches could be useful in developing effective PSMs. Figure S1 Complementation of E. coli W620 mutant phenotype by wild type and NADH insensitive cs plasmids. WT E. coli W620 represents E. coli W620 deletion mutant of cs gene pR163L, pK167A and pY145F represent plasmids containing NADH insensitive cs genes. All plasmid bearing strains were induced with 0.1 mM IPTG. Growth was monitored on M9 minimal medium with 0.2% glucose. +/2 at the top of each image indicates presence and absence of respective supplements in the media. (TIF) Figure S2 Growth of E. coli DH5a on Koser's citrate broth. E. coli DH5a containing pGm and pYC plasmids grown on media containing citrate as sole carbon source. All plasmid bearing strains are supplemented with 100 mg/ml thiamine, antibiotic 1/4th of the recommended dose and without supple-mentation of glutamate. +/2 at the top of each image indicates presence and absence of respective supplements in the media. (TIF) Table S1 Physiological variables from P. fluorescenstransformant and integrant strains grown on75 mM glucose in TRP minimal medium. Growth rate (m), Biomass yield (BMY) and specific glucose utilization rate (Q Glc) were estimated from mid log phase cultures and total glucose depleted (TGD) and glucose consumption (