![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article 
Applied and Environmental Microbiology, November 2006, p. 7111-7122, Vol. 72, No. 11
0099-2240/06/$08.00+0 doi:10.1128/AEM.01215-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Department of Plant Pathology, Washington State University, Pullman, Washington,1 USDA-ARS, Root Disease and Biological Control Research Unit, Pullman, Washington2
Received 26 May 2006/ Accepted 13 August 2006
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our research has focused on colonization of the wheat rhizosphere by Pseudomonas fluorescens Q8r1-96, which produces the polyketide antibiotic 2,4-diacetylphloroglucinol (2,4-DAPG). 2,4-DAPG-producing strains of P. fluorescens suppress root and seedling diseases of a variety of crops and play a key role in the natural biological control of take-all disease of wheat known as take-all decline (4, 19, 28, 49, 50, 65, 75). Strain Q8r1-96 is representative of D-genotype 2,4-DAPG producers as defined by restriction fragment length polymorphism analysis of the phlD gene (phlD+) (39) and by repetitive-sequence-based PCR using the BOXA1R primer (42). D-genotype strains accounted for the majority of phlD+ isolates obtained from wheat or pea plants grown in Washington state soils that naturally suppress take-all, caused by Gaeumannomyces graminis var. tritici, and Fusarium wilt of pea, caused by Fusarium oxysporum f. sp. pisi. Strain Q8r1-96 is exceptional because of its ability to establish and maintain large populations (up to 107 CFU/g of root) on the roots of wheat, pea (34, 51), and sugar beet (6) even when low doses are used. This special colonizing ability is typical of all D-genotype strains tested to date (34, 35, 51) and distinguishes these strains from typical rhizosphere pseudomonads.
Root colonization by introduced PGPR is a complex process that includes interactions among the introduced strain, the pathogen, and the indigenous rhizosphere microflora. All of these microorganisms interact with and influence each other in the context of the rhizosphere environment. Bacterial cell surface structures, such as flagella (18), fimbriae (11), and the O antigen of lipopolysaccharide (14), have been shown to influence the attachment of Pseudomonas cells to plant roots. Genes responsible for the biosynthesis of amino acids, vitamin B1 (67), a putrescine transport system (32), the NADH dehydrogenase NDH-1 (11), and ColR/ColS, a two-component regulatory system (16), also can influence the efficiency of root colonization. A site-specific sss recombinase gene, originally identified in P. aeruginosa 7NSK2 as an orthologue of the Escherichia coli site-specific recombinase gene xerC (26), has a role in root colonization in several strains (1, 15, 61) and even has been proposed as a potential target for improved colonization through genetic engineering (17). The protein encoded by sss, also known as xerC, belongs to the 
integrase family and plays a role in DNA rearrangement and phase variation (15, 61).
As one approach to identifying determinants of the unique rhizosphere competence of Q8r1-96 and related D-genotype strains, we hypothesized that these bacteria may interact with their plant hosts more intimately than other commensal pseudomonads interact with their plant hosts and that some of the broadly conserved bacterial genes critical to pathogenicity in such varied hosts as plants and animals (36, 54) may contribute to this unusual ability. Indeed, recent evidence of the presence of type III secretion genes in the PGPR strain P. fluorescens SBW25 (55, 56) and the presence of related genes in many other PGPR (41, 48, 58) suggests that such commonalities may occur more frequently than was previously anticipated. In this study, we focused on ptsP and orfT, two of the genes that contribute to the pathogenicity of P. aeruginosa in both plant and animal systems (53, 54) and are highly conserved (levels of identity, more than 70%) in the genomes of saprophytic rhizosphere pseudomonads, and on the sss recombinase gene because of its known contribution to the ability of bacteria to adapt to new environments. We identified and characterized the sss, orfT, and ptsP orthologues in P. fluorescens Q8r1-96, generated mutants with mutations in each gene, and determined the contributions of the genes to rhizosphere competence and strain competitiveness in the wheat rhizosphere.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
particles with a Gigapack Gold kit (Stratagene, La Jolla, Calif.), transduced into E. coli XL1-Blue MR, and selected on LB medium supplemented with tetracycline. Amplified and ordered copies of the genomic library were stored at –80°C in a freezing medium [2.5% (wt/vol) LB broth, 13 mM KH2PO4, 36 mM K2HPO4, 1.7 mM sodium citrate, 6.8 mM (NH4)2SO4, 0.4 mM MgSO4, 10% glycerol].
Screening of the genomic library by hybridization and PCR.
The library was arrayed on BrightStar-Plus nylon membranes (7.4 by 11.4 cm; Ambion, Inc., Austin, Tex.) by replicating clones from the glycerol stocks with a 96-pin multiblot replicator (V&P Scientific, Inc., San Diego, Calif.) and a library copier that permits arraying in a 386-sample format. After arraying, the clones were grown overnight at 37°C and lysed in 0.4 M NaOH with subsequent UV cross-linking as described elsewhere (7).
The sss and orfT hybridization probes were amplified by PCR performed with Taq DNA polymerase (Promega, Madison, Wis.) and primers SSS_UP and SSS_LOW and primers ORFT_UP and ORFT_LOW (Table 2), respectively. The primers were developed by using the Oligo 6.65 software (Molecular Biology Insights, West Cascade, Colo.), based on the sss recombinase sequence from P. fluorescens F113 (GenBank accession number AF416734) and the orfT sequence from P. fluorescens SBW25 (http://www.sanger.ac.uk/Projects/P-fluorescens). The cycling program with the sss primers included a 1.5-min initial denaturation at 94°C, followed by 30 cycles of 94°C for 30 s, 61°C for 20 s, and 72°C for 1.2 min and a final extension at 68°C for 5 min. PCR amplification with the orfT primers was performed similarly except that the annealing step was at 62°C for 30 s and extension was at 72°C for 1 min.
|
-32P] dATP using the rediprime II random primer labeling system (Amersham Pharmacia Biotech. Inc., Piscataway, N.J.) and were purified with a QIAGEN nucleotide removal kit (QIAGEN). The membranes on which the library was arrayed were prehybridized for 2 h at 60°C in a solution containing 3x SSC (60), 4x Denhardt's solution (60), 0.1% sodium dodecyl sulfate (SDS), and 300 µg per ml of denatured salmon sperm DNA (Sigma). Prehybridized membranes were incubated overnight with 
1 x 106 cpm of 32P-labeled probe under the same conditions and were washed with 2x SSC-0.1% SDS at room temperature (twice), 0.2x SSC-0.1% SDS at room temperature (twice), 0.2x SSC-0.1% SDS at 60°C (twice), and 0.1x SSC-0.1% SDS at 60°C (once). The conditions used for Southern hybridization with the orfT probe were the conditions described above, except that 4x SSC was used for prehybridization and high-stringency washes were performed at 58°C. Clones containing ptsP were identified by PCR performed with primers PTSP3 and PTSP4 (Table 2), which were developed from the genome sequence of P. fluorescens SBW25. The library was divided into "primary" pools (i.e., pooled clones from each 96-well plate) that were further subdivided into "secondary" pools (containing clones from each 96-well plate pooled by rows or by columns). The purified cosmid DNA from "primary" and then "secondary" pools was screened by PCR using a program that included 1 min of initial denaturation at 94°C, followed by 30 cycles of 94°C for 25 s, 60°C for 15 s, and 72°C for 20 s.
Shotgun sequencing and sequence analysis.
Selected cosmid clones carrying sss, ptsP, or orfT were sequenced by using the EZ::TN<Kan-2> transposition system (Epicenter Technologies, Madison, Wis.). Transposition reactions were performed in vitro according to the manufacturer's recommendations, and cosmids bearing insertions in the gene of interest were shotgun sequenced with transposon-specific primers by using an ABI PRISM BigDye Terminator v.3.0 Ready Reaction cycle sequencing kit (Applied Biosystems, Foster City, Calif.). The sequence data were compiled with the Vector NTI software (Invitrogen Corp., Carlsbad, Calif.) and were analyzed with the OMIGA 2.0 software (Accelrys, San Diego, Calif.). Database searches for similar protein sequences and protein motifs and domains were performed by using NCBI's BLAST network service (http://www.ncbi.nlm.nih.gov/BLAST) and the MyHits Internet engine (46).
Allelic replacement in Q8r1-96.
A spontaneous rifampin-resistant derivative of Q8r1-96 (34) was used for gene replacement mutagenesis. To construct an sss mutant, the sss gene interrupted by EZ::TN<Kan-2> was amplified with primers SSS_UP and SSS_LOW by using KOD Hot Start DNA polymerase. The cycling program included 2 min of initial denaturation at 94°C, followed by 30 cycles of 94°C for 15 s, 61°C for 30 s, and 68°C for 1.2 min and a final extension at 68°C for 5 min. The amplification product was cloned into the SmaI site of the gene replacement vector pNOT19 (62). The resultant Kanr plasmids were digested with NotI and ligated into a 5-kb fragment carrying a pMOB3 cassette (62) linearized with NotI and containing the sacB and cat genes. The resultant plasmid was electroporated into E. coli S17-1(-pir), selected on LB medium supplemented with chloramphenicol and kanamycin, and mobilized from E. coli S17-(-pir) into P. fluorescens Q8r1-96 by using a biparental mating technique. Mutant clones were first selected on LB medium supplemented with rifampin, kanamycin, and 5% sucrose. Sucrose- and kanamycin-resistant clones were screened for the absence of plasmid-borne sacB, bla, and catgenes by PCR performed with primers SAC1 and SAC2 (38), primers BLA1 and BLA2 (38), and primers Cm_UP and Cm_LOW (Table 2), respectively. Primers KAN_UP and KAN_LOW (Table 2) were used to detect the presence of a kanamycin resistance gene in sss recombinase mutants, which also were screened by PCR performed with primers SSS_UP and SSS_LOW to confirm the absence of the wild-type sss allele.
A similar strategy was used to construct a ptsP mutant. Briefly, the full-length ptsP gene was amplified from strain Q8r1-96 with primers PTSP5 and PTSP13 (Table 2) and Expand Long PCR polymerase (Roche Applied Science, Indianapolis, Ind.). The cycling program consisted of 2 min of initial denaturation at 94°C, followed by 30 cycles of 94°C for 10 s, 60°C for 30 s, and 68°C for 3.5 min and a final extension at 68°C for 5 min. The amplification product was treated with T4 DNA polymerase, cloned into the SmaI site of pNOT19, mutagenized in vitro with EZ::TN<Kan-2>, digested with NotI, and ligated with a 5-kb fragment carrying a pMOB3 cassette (62) as described above. The resultant plasmids were electroporated into E. coli S17-1(-pir) and used for gene replacement as described above, except that the mutants were screened by PCR performed with primers PTSP5 and PTSP13 to confirm the absence of the wild-type allele.
To construct the orfTmutant, the 1,386-bp 5' part of orfT was amplified by using KOD Hot Start DNA polymerase and oligonucleotides ORFT_KpnI and ORFT_Bam1 (Table 2). The cycling program included 2 min of initial denaturation at 94°C, followed by 35 cycles of 94°C for 15 s, 65°C for 30 s, and 68°C for 1 min and a final extension at 68°C for 5 min. The PCR product was digested with KpnI and BamHI, gel purified, and cloned into pNOT19. Next, a 1,246-bp fragment containing the 3' part of orfT was amplified with primers ORFT_Pst and ORFT_Bam2 (Table 2) by using the same cycling program. The PCR product was digested with BamHI and PstI, gel purified, and cloned into pNOT19 containing the 5' end of the orfT gene. These manipulations resulted in introduction of a unique BamHI site in orfT. This site was then used to insert the kanamycin resistance gene from EZ::TN<Kan-2>, yielding pNOT19-orfT-Kan. The resultant plasmids were digested with NotI, ligated with a 5-kb fragment of pMOB3 containing the sacB and cat genes, and electroporated into E. coli S17-1(-pir). Gene replacement mutagenesis was carried out essentially as described above, except that the mutants were screened by PCR performed with primers ORFT_seq1 and ORFT_seq2 to confirm the absence of the wild-type orfT allele. All mutant clones of each of the three genes were isogenic, and only one mutant clone was used for further experiments.
Construction of complemented mutants.
Full-length copies of sss, ptsP, and orfT were cloned into the stable broad-host-range plasmid vector pME6010 (23) by using Gateway Technology (Invitrogen). First, the genes were amplified by using a nested PCR protocol described by House et al. (27). Briefly, sss, ptsP, and orfT were amplified with gene-specific primers sssF and sssR, primers ptsPF and ptsPR, and primers orfTF and orfTR, respectively, and then attB sequences were introduced by reamplification with the secondary primers 2F and 2R (27) (or primers sss2F and 2R in the case of sss) (Table 2). All amplifications were carried out with KOD Hot Start DNA polymerase. The PCR products were then cloned into the entry plasmid pMK2010 (27) by using BP Clonase II (Invitrogen), sequenced to confirm gene integrity, and transferred with LR Clonase II (Invitrogen) to the destination vector pME6010 containing ccdB and attR cassettes (27). The resultant plasmids were electroporated (20) into Q8r1-96sss, Q8r1-96ptsP, or Q8r1-96orfT with a Gene Pulser II (Bio-Rad Laboratories, Hercules, Calif.).
Phenotypic characterization in vitro.
Polysaccharide production was scored visually after 3 days of growth on PsP by using a scale from 0 to 5, where 0 indicated a nonmucoid isolate and 5 indicated a moderately mucoid culture. Siderophore production was determined by measuring orange halos after 2 days of growth at 28°C on CAS agar plates (63). Exoprotease activity was detected by spotting 5 µl of an exponentially growing culture whose optical density at 600 nm (OD600) was adjusted to 0.1 on skim milk agar (59); a clearing zone surrounding the bacterial growth was measured after incubation for 48 and 72 h at 28°C. Production of hydrogen cyanide was monitored daily by using cyanide detection paper placed on petri dish lids for cultures grown on KMB agar amended with 0.44% glycine for 4 days at 28°C (3). All experiments described above were repeated twice with six replicates. Motility assays were performed on LB medium solidified with 0.3%, 0.5%, 1.0%, or 1.5% agar. Plates were inoculated with 5 µl of logarithmically growing bacterial cultures whose OD600 was adjusted to 0.1 and were incubated right side up at 28°C, and the diameter of outward expansion was measured after 24, 48, and 72 h. Experiments were repeated twice with four replicates per strain. Inhibition of G. graminis var. tritici by P. fluorescens Q8r1-96Gm and mutants of this strain was assayed on KMB agar as described previously (44). Carbon and nitrogen utilization profiles were generated by using Biolog SF-N2 and PM3 MicroPlates (Biolog, Inc., Hayward, Calif.), respectively, with 96 substrates each. Four independent repetitions were performed with each strain (40). Biolog assays were validated by using M9 minimal media supplemented with 0.4% D-galactose as a carbon source or with nitrogen sources at a concentration of 10 mM. Phloroglucinol compounds were extracted with ethyl acetate from bacterial cultures grown for 48 h at 27°C in KMB broth. The extracts were fractionated on a Waters NOVA-PAK C18 Radial-PAK cartridge (4 µm; 8 by 100 mm; Waters Corp., Milford, Mass.) as described previously (9). Two independent experiments with five replications were performed.
Rhizosphere colonization assays.
Rhizosphere colonization assays were performed with the sss recombinase, ptsP, and orfT mutants and Q8r1-96Gm, a gentamicin-resistant derivative of the parental strain tagged with mini-Tn7-gfp2 (71) to distinguish it from mutant strains in mixed-inoculation studies. Bacterial inocula were prepared and added to Quincy virgin soil as previously described (33) to obtain 1 x 104 CFU g–1 of soil and 0.5 x 104 CFU of each strain g–1 of soil (1:1 ratio) for single and mixed inoculations, respectively. The actual density of each strain was determined by assaying 0.5 g of inoculated soil as described by Landa et al. (33), and the control treatments consisted of soil amended with a 1% methylcellulose suspension. Experiments were repeated twice with six replicates per treatment. Spring wheat (Triticum aestivum L. cv. Penawawa) seeds were pregerminated on moistened sterile filter paper in petri dishes for 24 h in the dark and were sown in square pots (height, 6.5 cm; width, 7 cm) containing 200 g of Quincy virgin soil (33) inoculated with one or two bacterial strains. Wheat was grown for six successive cycles in a controlled-environment chamber at 15°C with a 12-h photoperiod. After 2 weeks of growth (one cycle), population densities of bacteria were determined as described by Mavrodi et al. (40). In order to determine the persistence of the mutants in the soil, the soil from pots that received the same treatment was decanted into plastic bags after the sixth cycle and stored at 20°C for 10 weeks before spring wheat was planted again. Plants were grown in a growth chamber for two successive 2-week cycles in the same controlled environment, processed, and analyzed as described below.
Population densities of the introduced strains were monitored by the modified dilution-endpoint method (43, 71). Briefly, individual wheat root systems were placed in 10 ml of sterile distilled water, vortexed, and sonicated. The soil suspensions were serially diluted in 96-well microtiter plates containing 1/3x KMB broth supplemented with rifampin, cycloheximide, ampicillin, and chloramphenicol and incubated for 72 h at room temperature (an OD600 of 
0.07 was scored as positive) (43). After 3 days, the cultures were replicated into fresh 96-well plates containing KMB broth amended with kanamycin or gentamicin to distinguish between strains in mixed inoculations. Population densities of total culturable heterotrophic bacteria were determined by performing the same assay in 0.1x tryptic soy broth supplemented with cycloheximide (43).
Data analysis.
All treatments in competitive colonization experiments were arranged in a complete randomized design. Statistical analyses were performed by using appropriate parametric and nonparametric procedures with the STATISTIX 8.0 software (Analytical Software, St. Paul, Minn.). All population data were converted to log CFU g–1 of soil or log CFU g–1 (fresh weight) of root. Differences in population densities among treatments were determined by standard analysis of variance, and mean comparisons among treatments were performed by using Fisher's protected least-significant-difference test (P = 0.05) or the Kruskal-Wallis all-pairwise comparison test (P = 0.05). The area under the colonization progress curve (AUCPC), which represented the total rhizosphere colonization for all six cycles (34), was calculated with SigmaPlot V. 8.0 (SYSTAT Software Inc., Richmond, Calif.). Data from phenotypic assays were compared by using a two-sample t test or the Wilcoxon rank sum test (P = 0.05).
Nucleotide sequence accession numbers.
The sss, ptsP, and orfT sequences of P. fluorescens Q8r1-96 have been deposited in the GenBank nucleotide sequence database under accession numbers AY172655, AY816321, and AY816322, respectively.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Included in sss-positive cosmid clone 7D10 was a 1,851-bp contig containing a predicted open reading frame with similarity to genes encoding numerous bacterial and phage site-specific recombinases (Fig. 1A). The sss gene encodes a predicted 299-amino-acid protein with a molecular mass of 33,772 Da and is flanked by two open reading frames coding for a conserved hypothetical protein and a putative hydrolase, YigB.
|
Shotgun sequencing of cosmid clone 7D2, which hybridized to the orfT probe, revealed a 5,176-bp contig (Fig. 1B) that included a predicted open reading frame with similarity to Orf338 from P. aeruginosa PA14 (69). The gene, referred to here as orfT, is preceded by a putative ribosome-binding site, GGAGA, and encodes a predicted 341-amino-acid protein with a molecular mass of 38,655 Da. The contig contains a number of other open reading frames, two of which are located immediately downstream of orfT and probably are cotranscribed with it. These genes encode a conserved hypothetical protein and a putative DnaJ-like protein.
Database searches revealed that orfT is highly conserved in sequenced bacterial genomes, and a blastp search against the nonredundant GenBank data set returned more than 40 hits with E values less than 1e–50. The deduced OrfT protein is most similar to its counterparts from P. fluorescens Pf-5 (NCBI accession number AAY94838; 84% identity), Pseudomonas syringae pv. syringae DC3000 (NCBI accession number AAO54097; 82% identity), P. putida KT2440 (NCBI accession number NP_742571; 74% identity), and P. aeruginosa PA14 (NCBI accession number AAD22455; 67% identity). OrfT is a member of COG3178, which contains predicted phosphotransferases (E value, 1e–90), and further analyses revealed the presence of a Pfam domain associated with the aminoglycoside phosphotransferase enzyme family (residues 152 to 234). The OrfT structure predicted by 3D-PSSM (29) is related to the structures of aminoglycoside 3'-phosphotransferases from Enterococcus faecalis (10) and Klebsiella pneumoniae (45).
ptsP-positive cosmid clone 5F9 contained a 2,586-bp contig that included a predicted open reading frame with similarity to numerous bacterial pstP genes (Fig. 1C). The ptsP gene encodes a predicted 759-amino-acid protein with a molecular mass of 83,208 Da and is flanked in Q8r1-96 by a well-conserved ribosome-binding site, GGAG, and a putative transcriptional terminator comprised of a 93-bp region with imperfect dyad symmetry.
The deduced PtsP protein is highly similar to its orthologues from P. fluorescens Pf-5 (NCBI accession number AAY95089; 96% identity), P. syringae pv. syringae DC3000 (NCBI accession number AAO58710; 92% identity), Azotobacter vinelandii (NCBI accession number CAA74995; 87% identity), and P. aeruginosa PAO1 (NCBI accession number NP_249028; 86% identity). The results of domain searches revealed a conserved PROSITE phosphoenolpyruvate (PEP)-utilizing enzyme signature (residues 620 to 638) and Pfam GAF (residues 17 to 154), N-terminal (residues 178 to 302), mobile (residues 318 to 399), and TIM barrel (residues 424 to 715) domains associated with PEP-utilizing enzymes. Finally, searches against the COG database indicated that the protein belongs to COG3605 containing bacterial PtsP proteins (E value, 0.0), and the protein structure predicted by 3D-PSSM (29) was related to the structures of pyruvate phosphate dikinases from Clostridium symbiosum (24) and Trypanosoma brucei (13).
Phenotypic characteristics of the sss, ptsP, and orfT mutations.
The motilities of wild-type Q8r1-96 and sss, ptsP, and orfT mutants of this strain were compared on LB medium solidified with different agar concentrations. The motility of the wild-type strain was significantly (P = 0.05) greater than the motilities of the ptsP and orfT mutants on 0.3% agar (Table 3). In contrast, the motility of the sss mutant did not differ from that of Q8r1-96. When higher concentrations of agar were used, neither the mutants nor the wild type showed the ability to swarm.
|
The growth kinetics of the sss, ptsP, and orfT mutants in 1/3x KMB and M9 media supplemented with glycerol were indistinguishable from those of wild-type strain Q8r1-96. Likewise, the carbon and nitrogen substrate utilization profiles of the sss and orfT mutants did not differ from those of Q8r1-96 on the 96 substrates contained in Biolog SF-N2 and PM3 microplates. In contrast, the ptsP mutant grew more slowly on D-galactose as a source of carbon and on L-cysteine as a source of nitrogen, both in the Biolog assays and in appropriately supplemented cultures grown in M9 medium.
Phloroglucinol production and fungal inhibition in vitro.
P. fluorescens Q8r1-96 and the sss, orfT, and ptsP mutants all produced detectable quantities of 2,4-DAPG, as well as monoacetylphloroglucinol (MAPG) and three other uncharacterized phloroglucinol-related compounds (Table 3). The sss and orfT mutants produced larger amounts of 2,4-DAPG and total phloroglucinol-related compounds than wild-type strain Q8r1-96 produced (for the orfT mutant only production of total phloroglucinol-related compounds was significantly greater [P = 0.05]), whereas the ptsP mutant produced significantly less MAPG and 2,4-DAPG. The reduced phloroglucinol production by the ptsP mutant was correlated with the diminished ability of this mutant to inhibit growth of G. graminis var. tritici in vitro; the hyphal inhibition indices (ratios of the distance between the edge of the bacterial colony and the fungal mat to the distance between the edge of the colony and the center of the mat) for this mutant and the parental strain were 0.0 and 0.16, respectively, after 6 days. In contrast, overproduction of 2,4-DAPG by the sss and orfT mutants was not correlated with increased fungal inhibition, probably due to differences in the relative amounts of phloroglucinol compounds accumulated during the different growth regimens. The hyphal inhibition indices for the sss and orfT mutants at 6 days postinoculation were 0.12 and 0.08, respectively.
Impact of the sss, ptsP, and orfT mutations on rhizosphere colonization by Q8r1-96.
The impact of the sss, ptsP, and orfT mutations on the rhizosphere competence of Q8r1-96 was assessed by performing competitive wheat root colonization assays under greenhouse conditions. The sss, ptsP, and orfT mutant strains were introduced into raw Quincy virgin soil either individually or, to test their competitiveness with the parental strain, in pairwise combinations (1:1 ratio) with strain Q8r1-96Gm. Figures 2A to C show the population dynamics of the strains over six 2-week growth cycles. In each case, the population densities of the introduced strains at cycle 0, the beginning of the experiments, were equivalent. In both single and mixed inoculations, the population sizes of the wild type and the sss recombinase and orfT mutant strains increased by 4 orders of magnitude by the end of cycle 1 and then slowly declined over the following six cycles.
|
|
Unlike the sss and orfT mutants, the ptsP mutant colonized the wheat rhizosphere significantly less (P = 0.05) than Q8r1-96Gm when the strains were introduced individually (Fig. 2C, cycles 1 and 4). The population densities of the two strains were similar in the other cycles, but the AUCPC and mean colonization values for Q8r1-96Gm were significantly higher than those for the ptsP mutant (Table 4). In mixed treatments, the ptsP mutant consistently colonized the wheat rhizosphere less than Q8r1-96Gm in cycles 1 through 6 (Fig. 2C), and the AUCPC and mean colonization values for the mutant were significantly lower (P = 0.05) than those for the wild type (Table 4).
The population densities for total culturable aerobic bacteria in the wheat rhizosphere for all experiments were more than log 8.4 CFU/g root. Mean colonization values of indigenous bacteria are shown in Table 4.
In order to evaluate survival of the wild type and the sss, orfT, and ptsP mutants in the absence of wheat roots, after the sixth cycle of the colonization experiments soils were stored at 20°C for 10 weeks, and then wheat seeds were sown again. The mean population densities recovered from the roots after two consecutive cycles are shown in Table 5. In soil into which the sss mutant and the wild-type strain had been introduced separately, comparable populations of the two strains were recovered from the rhizosphere, suggesting that the strains had survived equally well in the bulk soil during storage. In the soil into which the two strains had been coinoculated, a significantly larger population of the wild type than of the mutant was recovered, consistent with both the larger wild-type population at the time of storage and the tendency of the wild type to outcompete the mutant in mixed inoculations. In contrast, in the soils into which the ptsP or orfT mutant and the wild type had been introduced, whether separately or together, the recovered populations of the mutants were significantly smaller than those of the wild type (Table 5), suggesting that the mutants did not survive well in the absence of roots.
|
|
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The results of the sequence analysis suggest that the sss-like gene that we cloned from P. fluorescens Q8r1-96 indeed encodes a true xerC homologue. We included this gene in our studies of root colonization because it was shown previously to contribute to rhizosphere competence and phenotypic variation in fluorescent Pseudomonas spp. in a variety of crops (1, 15, 61) and because we wondered if it played a significant role in the unusual colonization properties of strain Q8r1-96. Under gnotobiotic conditions, the sss mutant of P. aeruginosa 7NSK2 colonized potato root tips 10- to 1,000-fold less than the wild type colonized potato root tips in mixed inoculations, and the corresponding mutant of P. fluorescens WCS365 was impaired in colonization of tomato root tips in potting soil and in colonization of radish and wheat root tips under gnotobiotic conditions (15). These gnotobiotic systems are important in identifying genes that may function in colonization, but the results obtained under such conditions must be validated in natural soil. In our studies, the bacteria were added to a natural field soil that was cropped to wheat for multiple cycles, allowing rhizosphere colonization to proceed for months under controlled conditions in the presence of indigenous microflora and percolating water. Under these conditions, the sss mutant of Q8r1-96 colonized the wheat rhizosphere to the same extent as its wild-type parent colonized the wheat rhizosphere, and a deficiency in rhizosphere competence of the mutant became apparent only when plants were inoculated with a mixture containing both strains (Table 4).
Inactivation of sss in P. fluorescens Q8r1-96 resulted in a phenotype similar to that of the sss mutant of P. fluorescens WCS365 (15). Q8r1-96sss did not differ from the wild-type parent in its carbon and nitrogen utilization profiles, growth rate, motility, or swarming behavior (Table 3). Among the minor differences observed were decreased siderophore production, also reported in P. aeruginosa 7NSK2 (26), and production of elevated amounts of exoprotease and phloroglucinol-related compounds (Table 3). The fact that the extracellular protease activity or HCN production of the sss mutant was not impaired suggests that the GacA/GacS regulatory circuitry, which coordinately regulates the production of secondary metabolites and exoprotease (77), was not disturbed in Q8r1-96sss. We have no immediate explanation for the phenotypic changes observed in the sss mutant, and we attribute them to the pleiotropic nature of the sss mutation.
The second gene studied in this work, orfT, was selected for analysis based on its role in interactions between P. aeruginosa PA14 and Arabidopsis thaliana (53). Like the sss mutant of Q8r1-96, the orfT mutant did not differ from the wild type in the ability to colonize the rhizosphere of wheat when it was introduced into the soil alone, but it colonized significantly less than the parental strain colonized when the two strains were introduced together (Table 4). Inactivation of orfT resulted in reduced motility and production of elevated total amounts of phloroglucinol-related compounds (Table 3). On the other hand, the orfT mutant did not differ from the wild type in carbon and nitrogen utilization profiles or the production of exoprotease, hydrogen cyanide, and siderophores. Furthermore, the mutant was indistinguishable from P. fluorescens Q8r1-96 on KMB and LB media supplemented with 2% glucose, but it was slightly more mucoid on PsP.
The results of computer analyses of OrfT indicate that its orthologues are found in all Pseudomonas genomes sequenced to date. The predicted proteins share a HRDxxxN motif with eukaryotic and some prokaryotic Ser/Thr and Tyr protein kinases (66) and some aminoglycoside phosphotransferases that are responsible for bacterial resistance to aminoglycoside antibiotics, such as streptomycin and kanamycin (45). The motif represents part of an active center where the conserved aspartate residue acts as a catalytic base (45). Despite these findings, this eukaryote-like protein kinase motif is too generic to provide any clues concerning the specificity of OrfT, and the similarity to aminoglycoside phosphotransferases is partial and limited to about 40% of the polypeptide chain. Thus, at this point the exact function of orfT in the rhizosphere competence of P. fluorescens Q8r1-96 remains unclear.
Like orfT, the third gene studied in this work, ptsP, was chosen for study based on its role in interactions between P. aeruginosa PA14 and A. thaliana (53). In contrast to the sss and orfT mutants, the rhizosphere competence of Q8r1-96ptsP was strongly impaired, and its rhizosphere population densities were significantly lower than those of the wild type in both single and mixed inoculations (Table 4 and Fig. 2C). The mutant lost the ability to maintain a density greater than 106 CFU/g root during extended cycling, which is a key characteristic of Q8r1-96 and all D-genotype isolates (6, 34, 35, 51). In cycles 5 and 6, the density of the mutant was 
103 CFU/g root, at least 3 orders of magnitude less than that of the wild type. The ptsP mutant exhibited altered morphology, reduced motility and exoprotease activity, and an increased level of fluorescence. It produced significantly smaller amounts of MAPG and 2,4-DAPG (ca. 54% and 60% of the wild-type levels). Carbon and nitrogen substrate utilization profiling revealed that Q8r1-96ptsP grew more slowly on D-galactose and L-cysteine, respectively.
The pleiotropic phenotype exhibited by Q8r1-96ptsP appears to reflect the proposed global regulatory function of the ptsP gene, which is highly conserved in the genomes of P. fluorescens Pf01, SBW25, and Pf-5. The product of ptsP forms part of an alternative PEP:carbohydrate phosphotransferase system (PTS). PTSs are found in a wide range of gram-positive and gram-negative microorganisms and have been best studied in enteric bacteria, where they are involved in sensing, transport, and metabolism of carbohydrates, as well as in catabolite repression and inducer exclusion. The first component of a typical PTS is a carbohydrate-specific multisubunit (or multidomain) enzyme II (EII) that forms a membrane translocation channel and phosphorylates the incoming sugar (47). The second part of a PTS is involved in the phosphorylation of all PTS carbohydrates. It lacks specificity and is comprised of cytoplasmic enzyme I (EI) and histidine protein (HPr).
In E. coli, ptsP encodes a nitrogen-specific EI paralogue called EINtr that, together with EIINtr and NPr (paralogues of EIIFru and HPr, respectively), forms a regulatory PTS phosphoryl transfer chain (52). Functional analyses have shown that this chain is involved in sugar-dependent utilization of certain amino acids and somehow links the metabolism of carbon and nitrogen. In P. putida, EIINtr either positively or negatively (and largely in a glucose-independent fashion) controls the expression of over 100 genes, some of which are members of the 
54 regulon (12). Based on the results of this study, it was proposed that EIINtr, which together with EINtr and NPr presumably forms an alternative PTS in P. putida, functions as a global regulatory factor rather than as a promoter-specific regulatory factor. Genes coding for EINtr also were described for A. vinelandii (64), Bradyrhizobium japonicum (31), P. aeruginosa (69), and Legionella pneumophila (25). In these microorganisms, as in E. coli, ptsP primarily plays a regulatory role and does not directly participate in phosphorylation or utilization of carbohydrates. Similarly, the ptsP mutant of Q8r1-96 did not differ considerably from the wild type in utilization of the sugars, amino acids, and organic acids included among the substrates present in Biolog MicroPlates. Recently, it was reported that ptsP plays an important role in the regulation of pyocyanin production but did not influence the quorum-sensing system in P. aeruginosa (78).
In conclusion, the data presented in this paper suggest novel functions for two genes, ptsP and orfT, that previously were linked with pathogenesis in P. aeruginosa. The ptsP and orfT mutants of Q8r1-96 did not have nonspecific growth defects in vitro, and the effects of the mutations became apparent only when the mutants were tested in the rhizosphere, either individually or in competition with the parental strain. In this respect, both genes fulfill the criteria for "true" rhizosphere colonization determinants, as described by Lugtenberg et al. (36). To our knowledge, this is the first report to provide evidence that ptsP is involved in rhizosphere colonization by fluorescent pseudomonads. The results of sequence analyses suggest that in P. fluorescens Q8r1-96, as in P. aeruginosa (57) and P. putida (12), PtsP (EINtr) forms part of the regulatory phosphoryl transfer chain that may coordinate the metabolism of carbon and nitrogen. Unfortunately, despite the fact that EINtr, EIINtr, and NPr have been purified and characterized biochemically, their exact functions remain poorly understood even in E. coli (70). However, if PtsP has a role in integrating carbon and nitrogen utilization, as it is thought to have in some bacteria, then the mutant phenotype might be manifested differently in the synthetic media used in our growth studies than in the complex mixture of sugars, organic acids, and amino acids present in root exudates. More comprehensive growth studies are necessary to determine whether such differences may account for the limited effects of the mutation on growth in vitro compared to the profound impact observed in the wheat rhizosphere.
|
ACKNOWLEDGMENTS |
|---|
We thank K. Hansen, A. Park, M. Young Son, and J. Mitchell for technical assistance and Nathalie Walter for comments on the manuscript.
|
FOOTNOTES |
|---|
Published ahead of print on 25 August 2006.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
