Siderophore biosynthesis genes of Rhizobium sp. isolated from Cicer arietinum L.

Rhizobium BICC 651, a fast-growing strain isolated from root nodule of chickpea (Cicer arietinum L.), produced a catechol siderophore to acquire iron under iron poor condition. A Tn5-induced mutant (B153) of the strain, BICC 651 impaired in siderophore biosynthesis was isolated and characterized. The mutant failed to grow on medium supplemented with iron chelator and grew less efficiently in deferrated broth indicating its higher iron requirement. The mutant produced less number of nodules than its parent strain. The Tn5 insertion in the mutant strain, B153, was located on a 2.8 kb SalI fragment of the chromosomal DNA. DNA sequence analysis revealed that the Tn5-adjoining genomic DNA region contained a coding sequence homologous to agbB gene of Agrobacterium tumefaciens MAFF301001. About 5 kb genomic DNA region of the strain BICC 651 was amplified using the primers designed from DNA sequence of agrobactin biosynthesis genes of A. tumefaciens MAFF 301001 found in the database. From the PCR product of the strain BICC 651, a 4,921 bp DNA fragment was identified which contained four open reading frames. These genes were designated as sid, after siderophore. The genes were identified to be located in the order of sidC, sidE, sidB, and sidA. Narrow intergenic spaces between the genes indicated that they constitute an operon. Phylogenetic analyses of deduced sid gene products suggested their sequence similarity with the sequences of the enzymes involved in biosynthesis of catechol siderophore in other bacteria.


Introduction
In iron deficient environment, bacteria meet their iron requirement by producing low-molecular-mass iron-chelating compounds called siderophores. The compounds bind available Fe 3? with high affinity to form complexes which are internalized by the cells with the help of cognate membrane proteins. The structures of these compounds are quite variable. Many of them are catecholate in nature. Catecholate siderophores were isolated from a variety of bacterial species including Escherichia coli which produces enterobactin, the prototype of catecholate siderophore. Enterobactin is biosynthesized from 2,3-dihydroxybenzoic acid (DHBA) which, in turn is synthesized from chorismate via the consecutive actions of three enzymes, viz., isochorismate synthase (EntC), isochorismatase (EntB) and 2,3-dihydro-2,3-dihydroxybenzoate dehydrogenase (EntA) (Earhart 1996). Although, the enzymes involved in the pathway of DHBA biosynthesis are common to all catechol siderophore-producing organisms, the organization of the biosynthetic genes within the operon and the sequences of the genes are different.
The members of the genus Rhizobium are soil bacteria that enter into symbiotic relationship with plant hosts. This relationship is iron dependent, since iron is required for nodule formation, and synthesis of leghemoglobin, nitrogenase complex, ferredoxin and other electron transport proteins to energise the nitrogenase system during symbiosis (Guerinot 1991). Because a part of life cycle of Rhizobium requires invasion, growth and differentiation within its host plant tissue, the role of siderophore for iron acquisition is important for the plant-microbe interaction. Improved iron scavenging properties of Bradyrhizobium sp. and Sinorhizobium meliloti 1021 positively correlate with rhizospheric growth and nodulation effectiveness in peanut and in alfalfa (O'Hara et al. 1988;Gill et al. 1991). Moreover, siderophore-producing ability helps in the sustenance of rhizobia in iron-deficient soils (Lesueur et al. 1995). Benson et al. (2005) reported that the gene fegA required for utilization of ferrichrome of B. japonicum 61A152 was also required for symbiosis. When the gene with its native, promoter was cloned and transferred to Mesorhizobium sp. GN25, nodulating peanut (Joshi et al. 2008) and Rhizobium sp. ST1, nodulating pigeonpea (Joshi et al. 2009), the transconjugants became symbiotically more efficient indicating importance of iron uptake protein in rhizobium-legume symbiosis.
Siderophore biosynthesis genes were studied in Rhizobium leguminosarum bv. viciae (Carter et al. 2002) and in Sinorhizobium meliloti 1021 (Lynch et al. 2001). In both the systems, siderophore biosynthesis genes were located on plasmids and were clustered close to the genes encoding their cognate membrane proteins. Siderophores produced by these strains were not of catechol type. Catecholate siderophores were isolated from R. leguminosarum bv. trifolii (Skorupska et al. 1989), R. ciceri (Roy et al. 1994;Berraho et al. 1997), Bradyrhizobium (cowpea) (Modi et al. 1985), Bradyrhizobium (peanut) (Nambiar and Sivaramakrishnan 1987), but the biosynthetic genes have not been investigated yet.
A fast-growing Rhizobium strain BICC 651 was isolated from a nodule produced on root of a chickpea (Cicer arietinum L.) plant grown in the Experimental Farm of Bose Institute at Madhyamgram, West Bengal, India. The strain produced a catechol siderophore and its cognate membrane receptor in response to iron deficiency (Roy et al. 1994). Structurally, the siderophore produced by Rhizobium BICC 651 contains 2,3-dihydroxybenzoic acid as core compound with two moles of threonine as the ligand. This suggested that the strain BICC 651 is likely to possess enzymes for siderophore biosynthesis similar to those of other catechol siderophore-producing organisms and the biosynthetic mechanism of the siderophore could be similar to that in other catechol siderophore-producing organisms. In the present study, transposon Tn5-mob insertional mutagenesis of the Rhizobium strain BICC 651 was performed to generate mutants impaired in siderophore production. One of the mutants was exploited to find out siderophore biosynthetic genes of the Rhizobium strain BICC 651.

Materials and methods
Bacterial strains, plasmid and culture media Bacterial strains and plasmid used in the study are listed in Table 1. Escherichia coli and Rhizobium strains were cultured in Luria-Bertani (LB) medium and yeast extract mannitol (YEM) medium, respectively. The complete medium described by Modi et al. (1985) [composition (g/l): K 2 HPO 4 , 0.5; MgSO 4 . 7H 2 O, 0.4; NaCl, 0.1; mannitol, 10; glutamine, 1; and NH 4 NO 3 , 1; pH 6.8] was used to study the growth and siderophore production of the strain BICC 651 and its mutants. For the purpose, the medium was deferrated with hydroxyquinoline (Meyer and Abdallah 1978). CAS-agar plate for detection of siderophore production by the organisms was prepared according to Schwyn and Neilands (1987) using the complete medium.

Tn5 mutagenesis
Transconjugates were generated by plate mating method (Mukhopadhyaya et al. 2000). Escherichia coli S17.1, harbouring the suicide plasmid pSUP5011::Tn5-mob (Simon 1984) was used as the donor and a rifampicin (100 lg/ml) resistant spontaneous mutant of the Rhizobium strain BICC 651, designated as Rhizobium BICC 651R was used as the recipient. Fresh cultures of both the donor and the recipient were mixed, centrifuged to collect the cells and spread on plates containing tryptone-yeast extract agar medium [composition (g/l): bactotryptone, 8; bactoyeast, 5; NaCl, 5; and agar 20; pH 6.8] (Beringer 1974). The mating mixture was incubated overnight at 30°C. The cells were then collected, suspended in YEM broth and plated on Rhizobium medium (Himedia) containing neomycin  Simon (1984) (50 lg/ml) and rifampicin (100 lg/ml). The plates were incubated at 28°C for the development of isolated colonies of transconjugants. The colonies were picked to prepare master plates for post mating selection. Each master plate was replica plated on CAS plate. The mutants which did not produce orange halo were selected as siderophore negative (Sid -) mutants.
Bacterial growth and siderophore production The parent strain BICC 651R, and its Sidmutant, B153, were examined for their growth on Rhizobium medium in presence of increasing concentrations of bipyridine, a synthetic iron chelator. Subsequently, the parent strain and the mutant were studied for their growth and siderophore production in deferrated complete medium. The cultures were incubated on a shaker at 28°C and growth was measured by monitoring OD 590 of the culture. Siderophore in the culture supernatant was assayed using the CAS reagent (Schwyn and Neilands 1987).

Plant inoculation and nodulation assay
For nodulation study, surface-sterilized seeds of chickpea were inoculated with Rhizobium strains and planted in sea sand in pots. Sea sand was treated with HCl, washed with distilled water until the pH of the washing was neutral. It was then air dried and fumigated with chloroform overnight under airtight condition, autoclaved and finally kept at 200°C in a hot air oven for 6 h. Five hundreds gram of sterile sand in an earthen pot was moistened with 100 ml of nitrogen-free plant nutrient medium for sowing the seeds. with sand-charcoal (1:3) containing 2 % aqueous sodium carboxymethyl cellulose. The mixture was used to coat the surface-sterilized seeds. The coated seeds (*10 8 bacteria/ seed) were kept in dark for overnight and the next day, five seeds were transferred to each pot containing the sterile sand soaked in nitrogen-free plant nutrient medium. A control set of seeds which did not receive any inoculum was also included in this study. The pots were kept under well-illuminated condition and watered when necessary to moisten the sand. At 1 week interval, 50 ml of 1/10th dilution of nitrogen-free plant nutrient medium was added to each pot. At 35 days of inoculation, plants were uprooted and observed for development of nodules.

DNA preparations and Southern hybridization
Total DNA of the parent, BICC 651R and the mutant, B153, was isolated, digested with SalI, electrophoresed on 1 % agarose gel according to standard protocol (Sambrook et al. 1989). DNA fragments from agarose gel were blotted on nylon membrane and hybridized with probe prepared from Tn5 fragment containing neomycin resistance (neo r ) gene. The probe was labelled by biotinylation following the manufacturer's instruction (NE Blot Phototope Kit, Biolabs).

DNA ligation and inverse PCR
The SalI digested DNA fragments showing positive signal following hybridization with labelled probe were eluted from the gel and allowed to self-ligate by incubating the DNA at a concentration of 0.3-0.5 lg/ml in presence of follows: initial denaturation for 10 min at 94°C; 35 cycles of denaturation for 30 s at 94°C, annealing for 30 s at 55°C, and extension at 72°C for 1 min, followed by a final extension at 72°C for 10 min. The identity of the PCR product was confirmed by nested PCR and also by ascertaining the profile of bands resulting from its specific endonuclease digestion. The PCR product was purified and sequenced by ABI PRISM 377 automated DNA sequencer (Perkin-Elmer, Applied Biosystem, Inc.).
Amplification of agb homologues from genomic DNA of BICC 651 and site of Tn5 insertion in Sidmutant To amplify the homologues of agrobactin biosynthetic genes from the strain BICC 651, PCR amplifications were carried out using the genomic DNA of BICC 651 as template, and agbCF and agbAR as primers (Table 2). A long PCR amplification kit was used for the purpose following the manufacturer's instruction (Genei, India). To identify the position of Tn5 insertion in the genomic DNA of the mutant, B153, two sets of PCR were carried out, one with the primer pair, agbCF and Tn5Int and another with the primer pair agbAR and Tn5Int. The reaction mixture for amplification contained 100 ng of genomic DNA as template and the thermal programme remained the same as described earlier for IPCR. To amplify larger ([2 kb) DNA fragment, extension time was increased by 1 min for each 1 kb of desired amplicon during final extension. The PCR products were analysed by agarose gel electrophoresis, purified and sequenced.
Analysis of the Tn5 adjacent DNA sequence Genomic DNA sequence was analyzed using BLAST version 2.2.1 of National Center for Biotechnology Information. Nucleotide BLAST and BLASTX were used to search for nucleotide sequences and derivative amino acid sequences, respectively.

Genomic context analysis
Gene context analysis of sidB gene of Rhizobium strain BICC 651 was carried out using the GeConT programme. The program allowed comparison of adjacent genes of sidB and visualization of genomic context of sidB homologue in the genomes of other organisms. The program is available at: http://www.ibt.unam.mx/biocomputo/gecont.html (Ciria et al. 2004).

Phylogenetic analysis
Evolutionary analyses were conducted using the software MEGA5 (Tamura et al. 2011). Multiple sequence alignment was carried out using a CLUSTALW and evolutionary history of the sequence was inferred using the neighbor-joining method (Saitou and Nei 1987). The optimal tree with the sum of branch length is equal to 4.31786458 is shown. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances are computed using the Poisson correction method (Zuckerkandl and Pauling 1965) and are in units of the number of amino acid substitutions per site.

Results and discussion
Transposon mutagenesis, and isolation of Sidmutants Transposon insertion mutagenesis in a siderophore producing Rhizobium sp. of chickpea was carried out to select the siderophore negative mutants to identify the structural genes required for biosynthesis of siderophore in the organism. Random mutagenesis of the Rhizobium strain BICC 651R by mobilization of Tn5-mob from the suicide vector pSUP5011 to the recipient cells occurred at a frequency of 10 -5 per donor as revealed by the number of neomycin-resistant transconjugants. This is about 1,000 times greater than the rate of spontaneous resistance of the recipients to neomycin (10 -8 ). Similar transposition frequency had also been reported for other bacteria (Mukhopadhyaya et al. 2000;Manjanatha et al. 1992). Upon screening of the 1,000 of transconjugants, five Sidmutants were isolated. Among these, one mutant (B153) was characterized in the present study.

Study of iron requirement
The amount of iron required for optimal growth of many Gram-negative bacteria ranged from 0.36 to 1.8 lM (Lankford 1973). The strain BICC 651R was able to grow on Rhizobium medium containing bipyridine up to a concentration of 400 lM but the Sidmutant, B153, could not grow even in the presence of 50 lM of bipyridine (Table 3). When the organisms were grown in the complete medium deferrated with hydroxyquinoline, the highest level of growth of the Sidmutant B153 (OD 590 = 1) was almost half as that of the parent strain (OD 590 [ 2). These observations suggested that because of impaired biosynthesis of siderophore, the mutant required more iron for its growth than the parent strain. Siderophore was not detected in the culture filtrate of the mutant at all stages of its growth, whereas the parent strain was observed to produce as much as 70 nmole of siderophore per ml of culture filtrate during late log phase of its growth (at 48 h) in the deferrated complete medium (Fig. 1). Table 4 presents the data of symbiotic performance of the siderophore non producing mutant B153 and its parent strain BICC 651R. The control plants receiving no inoculum produced no nodules. The plants inoculated with B153 produced less number of nodules than those inoculated with the parent strain BICC 651R. The wet weight of nodules produced by the mutant B153 was almost half as compared to those of its parents. Gill et al. (1991) reported that mutants of Sinorhizobium meliloti which were unable to produce siderophore were able to nodulate the plants but the efficiency of the nodules in nitrogen fixation was less as compared to wild type incited nodules indicating the importance of iron in symbiotic N 2 fixation.

Identification of the DNA fragments containing Tn5
SalI cuts the 7.7 kb Tn5-mob DNA into two parts: a 2.7 kb part with neo r gene and a 5 kb part containing the 1.9 kb mob region. When the Tn5 probe prepared from Tn5 fragment containing neo r gene was allowed to hybridize with SalI restricted genomic DNAs of BICC 651R and its Sidmutant, B153, it showed positive signal only as a single band of 2.8 kb genomic DNA of the mutant indicating insertion of a single copy of Tn5 in the genome of the mutant and no hybridization of the probe was detected with the genomic DNA of the parent, BICC 651R. In the mutant, the 2.8 kb SalI restricted genomic DNA fragment contained the 2.7 kb Tn5 fragment adjoining nearly a 0.1 kb genomic DNA fragment. The 2.8 kb SalI fragment was used to self-ligate more easily than using a larger ([7.7 kb) EcoRI restricted fragment. The Tn5 adjoining genomic DNA fragments were amplified by inverse PCR using the Tn5Int and NeoF primer pair (Table 2). A 1.3 kb amplicon containing an almost 0.1 kb genomic DNA was obtained (Fig. 2a). The genomic DNA fragment was sequenced and the sequence showed 94 % identity with the agbB gene sequence of Agrobacterium tumefaciens MAFF301001 (GenBank accession no. AB083344). The agbB gene product, isochorismatase, catalyzes the synthesis of 2,3-dihydro 2,3-dihydroxybenzoic acid from isochorismate (Sonoda et al. 2002). Thus, in the mutant, B153, Tn5 interrupted one of the DHBA biosynthetic genes and resulted in impaired synthesis of siderophore.
Nucleotide sequence analysis of the siderophore biosynthesis region Using the agbCF and agbAR primer pair, a 5 kb DNA fragment was amplified from the genomic DNA of BICC 651 (Fig. 2b). From the fragment, 4,921 bp was sequenced using appropriate primers and the DNA sequence was submitted to GenBank under accession no. GU251056. Upon nucleotide BLAST search, the 4,921 bp DNA region of BICC 651 was found to match (90 % identity) with the agrobactin biosynthetic (agb) genes of A. tumefaciens MAFF301001 (accession no. AB083344). Nucleotide ? -

--
The organisms were streaked on bipyridine containing Rhizobium medium and growth was scored after three days of incubation; '?' indicates growth, '-' absence of growth   (Fig. 3a). ORF1 (sidC) was 1,029 bp long (positions 213-1,241), potentially coding for a 342-amino-acidresidue-long protein. ORF2 (sidE), identified between positions 1,406-3,034, was 1,629 bp long and coded for a putative protein of 542 amino acid residues. At 97 bp, downstream from the TGA stop codon of ORF2 was found the ORF3 (sidB), 870 bp long, potentially coding for a 289-amino-acid-residue protein (positions 3,131-4,000). Finally, at 14 bp from the TGA stop codon of ORF3, the ATG start codon of ORF4 (sidA) (positions 4,014-4,772) was located. It was only 759 bp long, with a predicted translation product of 252-amino acid residues. The four ORFs identified in the 5 kb region were closely connected with narrow intergenic spaces indicating their polycistronic organization within an operon. In the sequence of 4,921 bp, a probable ribosome-binding site (AGGAGG) was identified six bp upstream of the ATG start codon of sidC of BICC 651 using Promoter prediction search tool (www. softberry.com). A presumable iron box (ACAAAACAT GATTAGC) was also identified 56 bp upstream of the sidC start codon similar to the one found in A. tumefaciens MAFF301001 (accession no. AB083344) and in Pseudomonas fluorescens (accession no. Y09356). Presence of the potential Fur box sequence indicated iron regulation of the functions of sid genes in the strain BICC 651.

Site of Tn5 insertion
From the genomic DNA of the mutant, B153, a 3.5 kb product was obtained using agbCF and Tn5Int primers (Fig. 2c) and a 1.5 kb product was amplified using agbAR and Tn5Int primers (Fig. 2d). Matching the sequences, it was found that Tn5 was inserted next to the 210th nucleotide position from the start codon of the sidB gene in the mutant B153 (Fig. 3a). catechol-producing organisms. Arrows represent genes with their relative orientation in the genomes. Genes are coloured according to their functional categories and labelled according to their original gene annotation in the database. Orange/black (isochorismatase); red (2,3-dihydroxybenzoate-AMP-ligase); green (isochorismate synthase); sky blue/grey (2,3-dihydro-2,3-dihydroxybenzoate) and yellow (entF) Comparison of genetic organization of sid genes of BICC 651 with other species The genomic context of sid DNA region in Rhizobium BICC 651, when compared to those in other catechol siderophore-producing members revealed that the organization of the region was not conserved; relative orientation and arrangement of the genes in the region of the genomes of different members varied considerably (Fig. 3b). Genetic organization of sidC, sidE, sidB, and sidA of Rhizobium BICC 651 was similar to that of entCEBA in E. coli for enterobactin biosynthesis (Crosa and Walsh 2002), in A. tumefaciens MAFF301001 for agrobactin (Sonoda et al. 2002), and siderophore biosynthetic genes in Brucella melitensis, Chromobacterium violaceum, and Klebsiella pneumoniae. In Aeromonas hydrophila, entA was missing immediately after entB, whereas in Erwinia carotovora, entF was present in between entB and entA.
The four biosynthesis genes were arranged as dhbACEB cluster in Bacillus subtilis (Rowland et al. 1996) as well as in Acinetobactor baumannii (Dorsey et al. 2003). In Vibrio cholerae for vibriobactin biosynthesis, these were organised as vibACEB cluster in which vibC and vibE were cotranscribed, while vibA and vibB belonged to two independent transcriptional units (Wyckoff et al. 1997). Since organization of catechol siderophore genes of any Rhizobium spp. was not found in the database, gene context analysis was carried out from available data of catechol siderophore-producing organisms.

Phylogenetic analysis
On the basis of the amino acid sequence identities data given in Table 5, phylogenetic tree of each of the four sid translational products was constructed using neighborjoining method (Fig. 5). In all the trees, the strain BICC 651 was placed in the same evolutionary branch as of Agrobacterium tumefaciens. This result along with 16S Rdna sequence analysis of the strain BICC 651 (accession no. DQ839132) demonstrates that the strain BICC 651 has a chromosomal background similar to that of A. tumefaciens. Since the strain BICC 651 was isolated from a nodule on the root of a legume and genes responsible for symbiosis including nodulation are located on Sym plasmids of rhizobia it might be possible that the strain BICC 651 had its ancestry in Agrobacterium but through horizontal gene transfer acquired the Sym plasmid from some hitherto unknown source. A somewhat similar observation was made with a strain T1K, reported to be Rhizobium trifolii (R. leguminosarum biovar trifolii), which nodulated red and white clover and shared many physiological properties with Agrobacterium tumefaciens. The strain was more closely related to Agrobacterium than to R. trifolii (Skotnicki and Rolfe 1978). Based on the gene sequence similarity, it is concluded that the stain BICC 651 might acquire the siderophore gene cluster from other bacteria via horizontal gene transfer during evolutionary process as suggested by Sonoda et al. (2002) who proposed that Agrobacterium tumefaciens MAFF301001 had obtained the agb gene cluster from Brucella melitensis, Pseudomonas fluorescens, Escherichia coli or Salmonella typhimurium in the same process. Nucleotide sequence analysis revealed that the genes, sidC, sidE, sidB, sidA constitute an operon. Disruption of sidB by Tn5 insertion results in poor growth under iron-limiting condition, absence of siderophore production and production of less number of nodules by the Sidmutant, B153 as compared to the wild type strain. The data indicate that the siderophore biosynthetic genes are essential for growth and production of catechol siderophore of the Rhizobium BICC 651, which in turn influences its symbiotic efficiency.