Regulation of Phosphate Assimilation in Rhizobium (Sinorhizobium) meliloti

Corresponding author: Turlough M. Finan, Department of Biology, McMaster University, 1280 Main St. West, Hamilton, Ontario, L8S 4K1, Canada, finan{at}mcmaster.ca (E-mail).

Communicating editor: P. L. FOSTER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We report the isolation of phoB and phoU mutants of the bacterium Rhizobium (Sinorhizobium) meliloti. These mutants form N₂-fixing nodules on the roots of alfalfa plants. R. meliloti mutants defective in the phoCDET (ndvF ) encoded phosphate transport system grow slowly in media containing 2 mM Pi, and form nodules which fail to fix nitrogen (Fix^-). We show that the transfer of phoB or phoU insertion mutations into phoC mutant strains restores the ability of these mutants to: (i) form normal N₂-fixing root-nodules, and (ii) grow like the wild type in media containing 2 mM Pi. We also show that expression of the alternate orfA pit encoded Pi transport system is negatively regulated by the phoB gene product, whereas phoB is required for phoCDET expression. We suggest that in R. meliloti cells growing under Pi limiting conditions, PhoB protein activates phoCDET transcription and represses orfA pit transcription. Our results suggest that there are major differences between the Escherichia coli and R. meliloti phosphate regulatory systems.

THE presence of two phosphate transport systems is a common feature in bacteria, and such systems have been identified in Escherichia coli (WILLSKY and MALAMY 1980 ), Acinetobacter johnsonii (VAN VEEN et al. 1993A ), and Bacillus cereus (ROSENBERG et al. 1969 ). Differences in specificity and affinity for phosphate between the two Pi transporters, as observed in E. coli and A. johnsonii (VAN VEEN et al. 1993B , VAN VEEN et al. 1994 ) suggest that the organisms have evolved two Pi transport systems in order to adjust to the different phosphate concentrations found in their environment.

The high-affinity phosphate-specific transport system of E. coli, PstSCAB, is a periplasmic-binding protein, ABC-type transporter, and pstSCAB transcription is induced when cells are starved for phosphate (WANNER 1996 ). Induction of the pstSCAB genes is regulated by the PhoB and PhoR proteins, which constitute a two-component regulatory system (RONSON et al. 1987 ; PARKINSON 1993 ; TOMMASSEN et al. 1982 ). The environmental sensor histidine kinase protein, PhoR, autophosphorylates under low phosphate conditions and then phosphorylates PhoB; in high phosphate conditions, PhoR probably dephosphorylates phospho-PhoB (MAKINO et al. 1989 , MAKINO et al. 1994 ). Phosphorylated PhoB binds to DNA sequences, called Pho boxes, which are found in the -35 promoter region of PhoB-regulated genes (MAKIND et al. 1994, MAKIND et al. 1996). Under Pi limiting conditions, PhoB activates transcription of a number of genes, which together constitute the Pho regulon. These include the pstSCAB genes and the phoA gene encoding alkaline phosphatase.

In E. coli, the PstSCAB and PhoU proteins appear to be involved in the sensing of environmental Pi, as many pstSCAB phoU mutations result in constitutive expression of the Pho regulon (AMEMURA et al. 1982 ; COX et al. 1981 ). On the other hand, Pi transport per se does not appear to be a sensory signal, as one PstA and two PstC missense mutations which abolish Pi transport show wild-type regulation of the Pho regulon (COX et al. 1988 , COX et al. 1989 ). STEED and WANNER 1993 showed that the phoU gene of E.coli is not required for Pi transport. However, phoU mutants grew poorly, suggesting that PhoU has another function in addition to its role in regulation of the Pho regulon.

Wild-type Rhizobium meliloti form N₂-fixing nodules on alfalfa, whereas mutants defective in the phoCDET (originally designated ndvF ) encoded phosphate transport system form nodules which contain few bacteria and fail to fix N₂ (Fix^-) (BARDIN et al. 1996 ; CHARLES et al. 1991 ). The Fix^- phenotype of phoCDET mutants can be suppressed to Fix⁺ by spontaneous mutations which have been separated into two genetic classes, designated I and II (ORESNIK et al. 1994 ). The Class I suppressor allele, sfx1, was shown to map to the promoter region and increase transcription of an operon, orfA-pit, which encodes a second phosphate transport system, and we hypothesized that suppression of the phoCDET Fix^- phenotype resulted from increased phosphate uptake via the Pit system (S. D. BARDIN, R. VOEGELE, N. FALCIONI and T. M. FINAN, unpublished results; VOEGELE et al. 1997 ).

In this article, we report the isolation of phoU and phoB mutants of R. meliloti. The phoU/B mutations are shown by transduction to map to the Class II suppressor locus (sfx2) and also to suppress the Fix^- phenotype of phoCDET mutants to Fix⁺. We show that phoB is required for phoCDET expression, and that phoB regulates expression of the orfA-pit operon in a phosphate-dependent manner.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Bacterial strains, plasmids, media, and growth conditions:
The strains and plasmids employed in this work are listed in Table 1. The growth media used were Luria-Bertani (LB) or LB containing 2.5 mM MgSO₄ and 2.5 mM CaCl₂ (LBmc) with antibiotic concentrations as previously described (CHARLES and FINAN 1991 ). The phosphate-free media was morpholinopropane sulfonic acid (MOPS)-buffered minimal media described in BARDIN et al. 1996 , except this medium was supplemented with 14 µl/liter of a yeast extract fraction which stimulates growth of R. meliloti in defined medium (BOB WATSON, personal communication). For the growth experiments, 24-hr LBmc cultures were washed and resuspended in phosphate-free MOPS media (MOPS P0). Twenty microliters of these cells were used to inoculate 5 ml of MOPS P0 (OD₆₀₀ ~0.05). The cells were grown for 24 hr under agitation. The resulting culture densities were adjusted to an OD₆₀₀ of 0.2, and 5 µl were then used to inoculate 5 ml of MOPS P0 and MOPS supplemented with 2 mM ortho-phosphate (MOPS P2) or 2 mM aminoethylphosphonate (AEP) (Sigma Chemical Co., St. Louis) media. The phosphate starvation step reduced the cellular phosphate reserves (presumably polyphosphate), which otherwise allowed significant growth in MOPS P0 media.

For experiments employing lacZ gene fusions, LBmc-grown cells, supplemented with tetracycline (Tc) 2 µg/ml for strains carrying pMP220-derived plasmids (SPAINK et al. 1987 ), were washed once and resuspended in MOPS P0. 20 µl and 5 µl of these cells were then used to inoculate 5 ml of MOPS P0 and MOPS P2 media, respectively. These cultures were grown for 38 hr (OD₆₀₀ reaching 0.3–0.4 in MOPS P0, and 0.8–1 in MOPS P2 media) before performing the ß-galactosidase and alkaline phosphatase (AP) assays.

The low osmolarity media (GYM) was prepared as stated in ORESNIK et al. 1994 . The AP phenotype of colonies was determined on solid LB agar medium containing 60 µg/ml 5-bromo-4-chloro-3-indolyl phosphate (LBXPhos). AP⁺ colonies are blue, Ap^- colonies are white.

The chromosomal transcriptional lacZ fusions to pit were constructed as outlined in Figure 8. The structure of the recombinants was confirmed by Southern blot of the genomic DNA digested with HindIII, SacI, and HindIII/SacI, and probed with the 4.8-kb HindIII/SacI fragment of pTH90. All ß-galactosidase values reported in Figure 6 were corrected for the ß-galactosidase activity obtained from recombinants with the lacZ gene in the opposite orientation (ort II) to the direction of orf-pit transcription (e.g., RmH695). This background activity remained constant at 10 Miller Units, regardless of the background or whether the fusion was to the sfx1 or wild-type locus.

View larger version (20K): In this window In a new window Download PPT slide   Figure 1. —Alkaline phosphatase (AP) activity of R. meliloti strains Rm1021 (wild type), RmG490 (phoC ), RmH363 (phoC sfx2), RmG497 (sfx2), RmG762 (phoC sfx1), and RmG591 (sfx1) after 60 hr growth in MOPS-buffered minimal media containing no added phosphate (MOPS P0, solid box) or 2 mm inorganic phosphate (MOPS P2, open box). The AP activity of Rm1021 and RmG490 was also determined after 80 hr growth in MOPS medium containing 2 mm aminoethylphosphonate (MOPS AEP, hatched box). Each activity represents the average of triplicate values ±SE.

View larger version (31K): In this window In a new window Download PPT slide   Figure 2. —Genetic linkage map showing the location of two Tn5 insertions (5258 and 5259) and the 5263::Tn5-233 insertion relative to sfx2. The cotransduction frequency is represented in percent, the arrow tail and head refer to the selected and unselected markers, respectively. The table presents the result of a three-factor cross used to determine the order of 5259 relative to sfx2 and 5263. Neomycin resistance (Nmr) was transduced from RmG552 into RmG640. 50 transductant colonies selected on LB Nm were screened by patching onto LB Gm-Sp and LBXPhos media. Linkage values to 5258 are based on data not shown. Abbreviations used: AP+, produce alkaline phosphatase; AP-, deficient in alkaline phosphate production; wt, wild type; (s) and (r), meaning sensitive and resistant, respectively.

View larger version (10K): In this window In a new window Download PPT slide   Figure 3. —Map of the phoU-phoB locus of R. meliloti showing the location of the phoU10, phoB3, and phoB8 insertion mutations. The map was drawn from the complete DNA sequence of phoB and partial sequence of phoU from GenBank (accession number M96261; P. McLean, C. Liu, C. Sookdeo and F. Cannon, unpublished results). The hatched part of the phoU gene represents the remaining portion of the gene, as predicted from alignment to the phoU gene of E. coli. Also indicated are the TnV subclones of phoU10 (pTH292), phoB3 (pTH287), and phoB8 (pTH311), as well as subsequent subclones used for sequencing. The four sequenced fragments (nos. 1–4) are indicated.

View larger version (18K): In this window In a new window Download PPT slide   Figure 4. —ß-galactosidase activities from plasmid pTH21 phoD::lacZ and pTH21 phoE::lacZ gene fusions in the following strains: (1) wild type (RmG212), (2) RmH617 (phoU10), and (3) Rm615 (phoB3). The control dataset 4 shows negligible activity from RmG212 cells lacking the fusion plasmids. The assays were performed after 32 hr growth in MOPS minimal media with no phosphate added (MOPS P0, solid box) or in media supplemented with 2 mm Pi (MOPS P2, open box). Each value corresponds to the mean of triplicate values ±SE. The pTH21 lacZ plasmid fusions 7A(phoD) and 19(phoE) are described in BARDIN et al. 1996 . ß-galactosidase activities are expressed as Miller Units.

View larger version (15K): In this window In a new window Download PPT slide   Figure 5. —Growth of R. meliloti strains in MOPS medium containing 2 mm Pi. The strains presented are Rm1021 (wt, ), RmG490 (phoC, ), RmH625 (phoB3 phoC, ) and RmH838 (phoB3, ). The growth characteristics of RmH623 (phoU10 phoC ), RmH624 (phoB8 phoC ), and RmH363 (sfx2 phoC ) were similar to the growth of RmH625 and are not presented here to simplify the figure. Each data point represents the mean of triplicate values.

View larger version (42K): In this window In a new window Download PPT slide   Figure 6. —Alkaline phosphatase- (a) and ß-galactosidase- (b) specific activities of an orfA-pit ::lacZ chromosomal fusion in wild type (wt), phoU10 (#2 and 7), phoB8 (#3 and 8), phoB3 (#4 and 9) and sfx2 backgrounds. pit expression from both wild-type and sfx1 loci were tested in these backgrounds. The assay was performed after 32 hr growth in MOPS minimal medium with no phosphate added (MOPS P0, solid box) or in a medium supplemented with 2 mm Pi (MOPS P2, open box). Each value corresponds to the mean of triplicate assays ±SE. Datasets are as follows: (1) RmH662, (2) RmH754, (3) RmH755, (4) RmH756, (5) RmH765, (6) RmH771, (7) RM-H861, (8) RmH862, (9) RmH863, and (10) RmH860. ß-galactosidase activities are expressed as Miller Units and AP specific activities were determined as described in materials and methods.

View larger version (53K): In this window In a new window Download PPT slide   Figure 7. —Alkaline phosphatase (a) and ß-galactosidase (b) activities of a plasmid-born pit::lacZ fusion in wild type (wt), phoC 490 (phoC), phoU10 phoC, phoB8 phoC (datasets 4 and 9), phoB3 phoC (datasets 5 and 10), and sfx2 phoC backgrounds. The assays were performed after 38 hr growth in MOPS-buffered minimal medium with no phosphate added (MOPS P0, solid box) or in medium supplemented with 2 mm Pi (MOPS P2, open box). Each value corresponds to the mean of triplicate values ±SE. ß-galactosidase activities are expressed as Miller Units and AP specific activities were determined as described in materials and methods.

View larger version (19K): In this window In a new window Download PPT slide   Figure 8. —Construction of the chromosomal lacZ fusions to pit. (A) The 0.5-kb EcoRI fragment (R1-R2) from pTH90 (containing the 3'-end of sfx1-pit) was subcloned into pUC118 so that R1 was on the polylinker side. The subclone was digested with SacII (located 11 nucleotides upstream from the pit translational stop codon) and PstI (located in the polylinker), and the fragment was replaced by a linker (with SacII/PstI protruding ends) to create pTH343. The linker was constructed by hybridizing two complementary oligos. The two oligos (5'-GGACCTCGTCGCCTGACCCGGGCTGCA-3' and 5'-GCCCGGGTCAGGCGACGAGGTCCGC-3') were synthesized so that the end of the pit gene was reconstituted (in order to obtain a functional gene). The translational stop codon was followed by a SmaI site used to clone the lacZ cassette. pTH343 was then digested with PstI/EcoRI and the fragment was cloned in pBR322 digested with the same enzymes to create pTH344. The lacZ cassette of the pLMS clone (pUC18 with the Spr mob lacZ cassette, M. Hynes, unpublished results) was subcloned as a SmaI fragment into pTH344 to create pTH351 and pTH352 with the lacZ gene in the same and opposite orientation as pit, respectively. (B) Following conjugative transfer from E. coli into R. meliloti, the plasmids recombined into the chromosome of a wild type and a sfx1 strain by single crossover homologous recombination in the 0.3kb SacII/EcoRI(2) fragment. Symbols used: poly: polylinker; H: HindIII; Pt: PstI; R: EcoRI; ScII: SacII; Sm: SmaI.

DNA manipulation and genetic techniques:
Cloning procedures, including DNA isolation, restriction digests, ligation, and transformation were performed according to SAMBROOK et al. 1989 .

Conjugal mating using MT616 as a helper strain, M12 generalized transduction, TnV and Tn5-233 replacement, and determination of the linkage between two markers by transduction were performed as described previously (CHARLES and FINAN 1991 ; FINAN et al. 1988 , FINAN et al. 1986 , FINAN et al. 1984 ).

TnV possesses the NptII gene (Km^r) and the pSC101 origin of vegetative replication (able to replicate in E. coli but not in R. meliloti), flanked by the inverted repeats IS50_L and IS50_R (FURUICHI et al. 1985 ). Following the replacement of the pho10, pho8, and pho3 Tn5-132 insertions with TnV, total genomic DNA from these strains was digested with Sal I, religated, and transformed into DH5-competent cells. As TnV does not contain a Sal I restriction site, the religated plasmid carried flanking genomic DNA. Subclones of the resulting plasmids were sequenced using the IS50 primer (5'-TCACATGGAAGTCAGATCCT-3').

Plant growth, alkaline phosphatase, and ß-galactosidase assays:
Plant growth experiments in a nitrogen-free environment and acetylene reduction assays were performed as stated in CHARLES et al. 1991 , except that a Hewlett-Packard (Palo Alto, CA) 5890 gas chromatograph (air 34 psi, H₂ 12 psi, N₂ 65 psi) was used, and the ethylene peaks were integrated using the HP3365 (Hewlett-Packard) Series II Chemstation computer program. Plant dry weight was determined by weighing the plant shoots of one pot, dried for one week in an oven, and dividing by the number of plants.

The alkaline phosphatase activity was measured as described in CHARLES et al. 1991 , except that the cells were spun down before the OD₄₂₀ was measured. The alkaline phosphatase activity was calculated using the formula (1000 x OD₄₂₀)/(OD₆₀₀ x T), with T being the reaction time (min).

The ß-galactosidase assay was performed by mixing 0.5 ml of cell, for which OD₆₀₀ was determined, with 0.5 ml Buffer Z (pH 7; 60 mM, Na₂HPO₄; 40 mM, NaH₂PO₄; 10 mM, KCl; 1 mM, MgSO₄ and 2.7 ml/liter of 2-mercaptoethanol added just before use), 20 µl chloroform and 10 µl 0.1% SDS. The tubes were equilibrated at 30° for 5 min, and the reaction was started by adding 0.2 ml of 4 mg/ml o-nitrophenyl-ß-D-galactoside (ONPG). When the solution turned yellow, the reaction was stopped by adding 0.5 ml of 1 M Na₂CO₃. The optical density at 420 nm (OD₄₂₀) was determined after centrifuging the cells for 5 min at 12,000 rpm. The ß-galactosidase activity in Miller Units was calculated using the formula (1000 x OD₄₂₀)/(OD₆₀₀ x T x V), with T being the reaction time (min) and V representing the initial volume of culture used (ml).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains carrying sfx2 are deficient in alkaline phosphatase production:
Measurements of alkaline phosphatase activity in R. meliloti cells can be used to monitor the physiological status of cultures with respect to phosphate availability (BARDIN et al. 1996 ). Thus, the level of AP activity detected in Rm1021 wild-type cells cultured in MOPS-buffered minimal media with no added phosphate (Pi 0 mM) is 10–20-fold higher than the AP activity found in cells grown in the same media containing 2 mM inorganic phosphate (Pi 2 mM) (Figure 1). Similar measurements for cells of the R. meliloti phoC mutant, RmG490, revealed a high AP activity even in media containing 2 mM Pi (Figure 1). Thus, even in the presence of 2 mM Pi, RmG490 cells appeared to be starved for phosphate. This phenotype is consistent with the observation that RmG490 grows poorly in media containing 2 mM Pi (S. D. BARDIN, R. VOEGELE, N. FALCIONI and T. M. FINAN, unpublished results). As we had previously found that R. meliloti phoCDET mutants grew like the wild type in MOPS media containing 2 mM AEP as sole source of phosphorus (BARDIN et al. 1996 ), we measured AP activity in cultures of RmG490 and the wild-type Rm1021 following growth in this medium (Figure 1). Both RmG490 and wild-type cells contained background AP activities after growth with 2 mM AEP. Thus, unlike what is observed in E. coli, where mutants defective in the high-affinity transport system (pstSCAB) show constitutive AP expression (COX et al. 1989 ), in R. meliloti, AP expression in the phoC mutant was repressed when a readily assimilated phosphorus source like AEP was provided in the growth media.

To examine the effect of the previously isolated phoCDET (ndvF ) symbiotic suppressor mutations (ORESNIK et al. 1994 ) on alkaline phosphatase expression, we measured AP activity in the Class I and Class II suppressor strains RmG762 (phoC sfx1) and RmH363 (phoC sfx2). AP activity in both these strains was repressed in the presence of 2 mM Pi (Figure 1). Unexpectedly, however, no alkaline phosphatase activity was detected when RmH363 (phoC sfx2) was grown in MOPS medium with no Pi added. This result was in contrast with what we observed with the wild type (wt), RmG591 (sfx1), and RmG762 (phoC sfx1) strains. Subsequent analysis revealed that RmG497 cells that carry the sfx2 mutation in an otherwise wild-type background, as well as strain RmG425 carrying the other Class II allele, sfx3, in a phoCDET background, also lacked alkaline phosphatase activity (AP^- phenotype) (Figure 1 and data not shown). The AP^- phenotype of strains carrying a Class II mutation, together with the suppression of the mucoid colony and symbiotic phenotypes associated with the phoCDET (ndvF ) mutations (ORESNIK et al. 1994 ), suggested that sfx2 and sfx3 mutations affected a regulatory gene involved in phosphorus assimilation similar to the phoBR genes of E. coli (WANNER 1993 ).

Isolation of phoUB mutants:
To isolate mutants defective in the phosphate-signaling regulatory pathway, we screened for alkaline phosphatase negative mutants (AP^-). Tn5-132 (oxytetracycline resistance) and Tn5 insertion mutants of Rm1021 were plated on LB agar containing the AP chromogenic substrate 5-bromo-4-chloro-3-indolyl phosphate (LBXPhos). One Tn5 insertion mutant (RmH405, pho27 ) and 26 Tn5-132 mutants which formed white colonies (AP^-) were identified and purified. To check for linkage of the AP^- mutations to the sfx2 locus, we utilized the insertions 5258::Tn5, 5259::Tn5, and 5263::Tn5-233, which were previously identified to be linked to the sfx2 locus (ORESNIK et al. 1994 ; I. J. ORESNIK and T. M. FINAN, unpublished results). The order and linkage of these insertions with respect to sfx2 were deduced from the three-factor cross described in Figure 2.

Nm^r was transduced from strain RmG551 (5258::Tn5 ) into the 26 Tn5-132 AP^- mutants, and transductants were screened for the presence of blue colonies (AP⁺) on LBXPhos plates. Three of the Tn5-132 insertion mutations designated pho10, pho8, and pho3 showed 60, 64, and 66% linkage to 5258::Tn5, respectively. In further transductions, Nm^r from insertion 5259::Tn5 (strain RmG552) showed 48, 44, and 42% linkage to the pho10, pho8, and pho3 mutations, respectively. These linkage values were similar to those between the two insertions 5258 and 5259, and sfx2, suggesting that the pho10, pho8, and pho3 insertion mutations and sfx2 map to the same locus (see Figure 2). The remaining 23 AP^- (white) Tn5-132 insertion mutants showed no linkage to 5258::Tn5; however, these were 100% linked in transduction to the Tn5 insertion in pho27 (strain RmH405). These mutants were not examined further.

To identify the gene(s) in which the pho10, pho8, and pho3 Tn5-132 insertions were located, we first replaced the Tn5-132 insertions with TnV (FURUICHI et al. 1985 ), and then cloned the TnV together with the flanking DNAs as SalI fragments into E. coli. Following subcloning, the DNA sequence extending into the flanking genomic DNA was determined using a primer which anneals 50 bp inside the end of the IS50 (see MATERIALS AND METHODS). Blast searches of the DNA data bases revealed that these sequences were identical with sections of the DNA sequence of the phoB and part of the phoU genes from R. meliloti Rm1021 which have been determined previously (GenBank accession no. M96261, P. MCLEAN, C. LIU, C. SOOKDEO, and F. CANNON, unpublished results). A schematic map of the phoUB region, together with the locations of the three pho::Tn5-132/TnV insertions is shown in Figure 3. The phoU and phoB genes lie adjacent to one another in the order phoU-phoB. The deduced R. meliloti PhoB protein is 47.6% identical to the PhoB protein of E. coli, while the deduced C-terminal 117 amino acids from PhoU are 36% identical with the corresponding region of the E. coli PhoU protein. The pho10 insertion was located 84 nucleotides into the unsequenced region of phoU as determined from sequence alignment to the PhoU protein sequence of E. coli. Insertions pho3 and pho8 were located at position 157 aa and 188 aa of the R. meliloti PhoB sequence, respectively. In view of these results, the pho10, pho8, and pho3 alleles were designated phoU10, phoB8, and phoB3, respectively.

The location of the phoUB locus on the R. meliloti genome was determined by conjugation using seven Tn5-mob insertion strains as described in ORESNIK et al. 1994 . The Tn5-mob insertions were transduced into strain RmG640 (5263,Tn5-233, 50% linked to sfx2; Figure 2). Following purification, the resulting transductants were crossed in conjugation with the rifampicin resistant (Rif ^r) Rm5000 strain. The frequency of Rif ^r, Gm^r Sp^r colonies placed the phoUB locus between trp-33 and pyr49 on the R. meliloti chromosome (KLEIN et al. 1992 ). More refined mapping was not carried out.

The phoUB genes are required for phoCDET expression:
We have previously shown that transcription of the pho-CDET transport genes is induced when R. meliloti cells are starved for phosphate, and that the phoC promoter contained two elements which are similar in sequence to the consensus-binding site for the E. coli PhoB protein (BARDIN et al. 1996 ). To determine whether the phoUB genes are required for phoCDET transcription, plasmids carrying the phoD7::lacZ and phoE19::lacZ gene fusions (BARDIN et al. 1996 ) were transferred into the strain RmG212 (lac^-) and its phoU and phoB derivatives. ß-galactosidase and AP activities were determined in cells cultured both in media without added Pi and with 2 mM Pi added. The AP activity of wild-type and pho mutant cells (not shown) was similar to the data shown in Figure 6A. The ß-galactosidase activity results clearly showed that neither phoD nor phoE were expressed in the phoB background (Figure 4, dataset 3). A low level of phoD and phoE expression was detected in the phoU mutant background (Figure 4, dataset 2).

phoU and phoB mutations suppress the symbiotic and associated phenotypes of phoCDET mutants:
The phoU10, phoB8, and phoB3 mutations had similar phenotypic effects to the sfx2 suppressor mutation. As the sfx2 allele was originally identified as a suppressor of the symbiotic Fix^- phenotype of phoCDET (ndvF ) mutants, we tested the three phoB and phoU insertion mutations for their ability to suppress the Fix^- phenotype of a phoC mutant. We also examined the ability of phoB and phoU mutations to suppress the slow growth of phoCDET mutants in media containing 2 mM Pi, and the mucoid colony phenotype which phoCDET mutants exhibit when plated on low osmolarity media (see below). The phoUB-phoC strains were constructed by transducing the phoU10, phoB8, and phoB3::TnV (Nm^r) insertion mutant alleles from strains RmH399, RmH428, and RmH430, respectively, into RmG490 (phoC 490) to create strains RmH623, RmH624, and RmH625, respectively.

The symbiotic phenotype of the various mutants was determined from the analysis of alfalfa plants 28 days after inoculation with the various R. meliloti strains. Plants inoculated with the phoC mutant were small and chlorotic, and showed little evidence of N₂-fixation as measured by acetylene reduction and plant dry weight determinations (Table 2). Plants inoculated with the phoC, phoU/B double mutants had shoot dry weight and acetylene reduction values comparable to the wild-type strain Rm1021 (Table 2). Thus, the phoU10, phoB8 and phoB3 mutations suppressed the Fix^- phenotype of the phoC to Fix⁺. We also note that the individual phoB, and phoU insertion mutants showed no reduction in symbiotic effectiveness compared to the wild-type strain.

The ability of the various R. meliloti strains to grow in MOPS-buffered minimal media containing 2 mM Pi was determined (Figure 5). While the phoC mutant grew poorly, the single phoB, phoU or sfx2 mutants, and the double mutants carrying phoC together with phoU10, phoB8, phoB3 or the sfx2 allele grew as well as the wild-type strain Rm1021 (Figure 5, data not shown). Given that phoU/B mutants do not express the phoCDET genes (Figure 4), it appears that the transport system(s) which allows Pi uptake into phoU/B mutants may also be responsible for allowing the phoU/B, phoC double mutants to grow normally in media containing 2 mM phosphate.

ORESNIK et al. 1994 observed that phoCDET (ndvF ) mutants form mucoid colonies, as opposed to dry wild-type colonies when the strains are plated on low osmolarity GYM medium. Moreover, phoCDET sfx1 and pho-CDET sfx2 double mutants formed dry colonies like the wild type on this medium. The mucoid phenotype was dependent on genes required for synthesis of the exopolysaccharide II (EpsII) of R. meliloti; these genes were known to be expressed under phosphate starvation conditions (ZHAN et al. 1991 ). We investigated whether the phoUB mutations were able to suppress the mucoid phenotype of phoC mutants by plating RmH623 (phoC phoU10), RmH624 (phoC phoB8), and RmH625 (phoC phoB3) on GYM agar. In all cases, a dry type of colony morphology comparable to that of the wild type was obtained (Table 2). Moreover, when the phoC mutant RmG490 was plated on GYM agar supplemented with 2 mM aminoethylphosphonate (a P source the phoC mutant strain can utilize), colonies with a dry wild-type morphology were obtained. These data suggest that the mucoid phenotype of phoCDET mutants is a direct consequence of the phosphate starvation state of these cells when plated on GYM medium.

In summary, the results presented in this section show that phoB and phoU mutations suppress the symbiotic and phosphate-dependent phenotypes associated with mutations in the phoCDET locus.

Mechanism of phoUB mediated suppression of phoCDET mutants:
In view of the above results, we suspected that phoB and/or phoU mutations led to increased expression of the recently characterized orfA-pit locus, which appears to encode an alternate phosphate transport system (S. D. BARDIN, R. VOEGELE, N. FALCIONI and T.M. FINAN, unpublished results). To investigate this possibility, we transduced phoB, phoU, and sfx2 mutations into strain RmH662, which carries a chromosomally located transcriptional lacZ fusion to the wild-type orfA-pit locus [to give strains RmH754 (phoU10), RmH755 (phoB8), RmH756 (phoB3), and RmH765 (sfx2)], and into strain RmH771, which carries the lacZ fusion to sfx1 orfA-pit locus [to give strains RmH861 (phoU10), RmH862 (phoB8 ), RmH863 (phoB3), and RmH860 (sfx2)]. We note that the lacZ fusion construct employed in these experiments was prepared so that lacZ was inserted immediately downstream of the pit translational stop codon (see MATERIALS AND METHODS and Figure 8 for details of this construction). Thus, in the wild-type and sfx1-directed gene fusions, both the orfA- and pit-encoded proteins should be fully functional. (This was confirmed as the sfx1 constructs in a phoC mutant background grew like wild type in MOPS medium containing 2 mM Pi, data not shown.) The level of orfA-pit expression (ß-galactosidase activity) was measured after culturing the cells for 32 hr in MOPS-buffered medium with no added Pi (P0) and with 2 mM Pi added (Figure 6B). To monitor the physiological status of the cells with respect to phosphate, we also measured alkaline phosphatase activity (Figure 6A).

Strains with a wild-type background (Figure 6A, datasets 1 and 6) showed the expected high AP activity in P-starved cells and low AP activity in cells grown with excess Pi (2 mM). As expected, the various phoU/B and sfx2 mutant strains contained minimal AP activities regardless of the level of Pi in the growth medium.

The level of orfA-pit expression (ß-galactosidase activity) directed from the sfx1 promoter was higher than the level directed from the wild-type orfA-pit promoter (Figure 6B, datasets 1–5 vs. 6–10); and in the wild-type background, both wild-type and sfx1-directed orfA-pit expression was phosphate-regulated as the ß-galactosidase activity was two- to fourfold higher in cells grown with 2 mM phosphate compared to phosphate-starved cells (Figure 6B, datasets 1 and 6). In the phoU, phoB, and sfx2 backgrounds, orfA-pit expression increased 4.5–5.5-fold when the cells were cultured in the absence of added Pi, and 1.5–2-fold when the cells were grown with excess Pi (2 mM) relative to orfA-pit expression in a wild-type background (Figure 6B, datasets 1 vs. 2–5, and 6 vs. 7–10). These data suggest that phoB (phoU ) is negatively regulating orfA-pit expression. Moreover, in the phoB and sfx2 backgrounds the phosphate-dependent regulation of orfA-pit expression was dramatically reduced, suggesting that phosphate regulation of orfA-pit expression was probably mediated via PhoB. The results outlined above were obtained from the chromosomal orfA-pit ::lacZ gene fusions. We obtained similar data when pit expression from a plasmid-borne gene fusion was assayed in these strains (data not shown).

As phoB and phoU mutations suppressed the Fix- and phosphate-growth phenotype of phoCDET mutants, we wished to examine the influence of the phoU, phoB, and sfx2 mutations on the level of orfA-pit expression observed in a phoCDET mutant background. In these experiments, expression from plasmid-borne wild-type and sfx1-directed orfA-pit::lacZ fusions in phoC phoB and phoC phoU backgrounds was monitored, and again AP activities were measured as controls (Figure 7, a and b). We note that phoC mutant cells behave as if they are phosphate-starved, even in the presence of excess phosphate (Figure 7A, dataset 2; Figure 1). In the phoC background, under both culture conditions, orfA-pit expression directed from the wild-type promoter was as low as its expression in phosphate-starved wild-type cells (Figure 7B, datasets 1 and 2). The introduction of a phoB or phoU mutation resulted in a greater than fivefold increase in the level of orfA-pit expression, suggesting that the repression of orfA-pit expression in phoC cells was mediated by PhoB or PhoU (Figure 7B, dataset 2 vs. 3–5). In the case of sfx1-directed orfA-pit expression, the level of expression observed in the phoC phoB and phoC phoU double mutants was two- to fourfold higher than the basal level observed in phosphate-starved wild-type cells and in phoC cells (Figure 7B, datasets 8–10 vs. 6 and 7). Thus, under phosphate starvation conditions, PhoB and/or PhoU appear to repress sfx1-directed orfA-pit expression. In cells cultured in 2 mM Pi, the level of sfx1-directed orfA-pit transcription was slightly higher in the wild-type and phoC background than in the phoC phoB and phoC phoU double mutants (Figure 7, datasets 6 and 7 vs. 8–10) Thus, the combination of sfx1 and phoB or phoU does not have an additive effect on the level of pit transcription. Lastly, we note that in a phoC background, sfx1-directed orfA-pit expression, but not AP synthesis, was responsive to the media phosphate concentrations (Figure 7, a and b, dataset 7). This difference between wild-type vs. sfx1-directed orfA-pit expression was previously observed (S. D. BARDIN, R. VOEGELE, N. FALCIONI and T.M. FINAN, unpublished results).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Defined phoB and phoU mutants of R. meliloti were isolated and used to examine the role of phoB and phoU in the expression of the phoCDET and orfA-pit genes. Expression of both phoD::lacZ and phoE::lacZ gene fusions was completely abolished in phoB and phoU mutant strains (Figure 4). We showed earlier that phoCDET expression is strongly activated in response to Pi starvation, and had noted that two appropriately positioned PhoB-like binding sites are located in the phoC promoter (BARDIN et al. 1996 ). Collectively, these data represent strong evidence that PhoB is a positive regulator of phoCDET transcription.

Our data also suggest that phoB plays an important role in determining the level of orfA-pit transcription (Figure 6B and Figure 7). In Pi-starved cultures, transcription of pit and orfA was clearly higher in the phoB and phoU mutants than in the wild type (Figure 6B, dataset 1 vs. 2–5, data for orfA not shown). In addition, while the wild-type pit fusions showed increased expression in Pi-sufficient compared to Pi-deficient cells, no such Pi regulation was observed when phoB or phoU mutants were examined (Figure 6B, dataset 1 vs. 2–4). While we assume that PhoB is directly interacting with the orfA-pit promoter, our evidence is indirect.

We found that phoB, phoU, and sfx2 mutants, which are phenotypically phoCDET ^- (Figure 4 and data not shown), formed wild-type Fix⁺ nodules on alfalfa and grew like wild type in media containing 2 mM Pi (Table 2; Figure 5). This indicated that an alternative Pi transport system was induced in these mutants. Our data on the expression of pit ::lacZ fusions in the phoB and phoU mutants (Figure 6B) also strongly suggest that the alternate uptake system is encoded by the orfA-pit locus. We recently demonstrated that the kinetics and specificity of Pi uptake in phoB mutant cells and in phoC sfx1 mutant cells is very similar (VOEGELE et al. 1997 ). As phoC sfx1 double mutant cells appear to transport Pi via the OrfA-Pit system (S. D. BARDIN, R. VOEGELE, N. FALCIONI and T.M. FINAN, unpublished results), the transport data referred to above are consistent with the suggestion that the OrfA-Pit transport system is employed for Pi uptake in phoB or phoU mutant cells.

To account for the reciprocal pattern of orfA-pit and phoCDET expression observed in wild-type cells cultured in the presence and absence of Pi, we suggest that when the concentration of Pi is high (2 mM), PhoB is inactive and orfA-pit is derepressed (expressed), and the phoCDET system is not expressed. Conversely, under conditions of phosphate starvation, PhoB is activated, orfA-pit expression is repressed, whereas phoCDET expression is activated by PhoB. The pattern of phoCDET and orfA-pit expression reflects different physiological characteristics of the two transport systems. Thus, the OrfA-Pit system is a low-affinity Pi transporter and the PhoCDET system has a high-affinity for Pi (VOEGELE et al. 1997 ).

When R. meliloti phoC insertion and phoCDET deletion mutants are cultured in media containing 2 mM Pi, the cells contain high AP activity and thus physiologically behave as though they are starved for phosphate (Figure 7A, dataset 2, data not shown). Under these conditions, the orfA-pit genes are repressed and the cell is then phenotypically both orfA-pit and phoCDET negative. The poor growth of phoCDET mutants in media containing 2 mM Pi therefore indirectly results from the lack of orfA-pit expression; and wild-type growth is restored via mutations which increase orfA-pit expression, such as sfx1, phoB or phoU.

Why phoCDET mutations lead to reduced orfA-pit expression is not clear. In E. coli, mutations in the PstSCAB transport system generate a Pho constitutive phenotype, which is not unlike that observed for phoCDET mutants of R. meliloti. The PstSCAB system is thus believed to be part of the Pi sensory system (WANNER 1996 ). If the PhoCDET system has a sensory role in R. meliloti, we believe there must be at least another backup sensory system, as when phosphorus in the form of 2 mM AEP is supplied to R. meliloti phoC mutants, AP activity is completely repressed, and hence a phosphorous sensory system appears to function (Figure 1). Moreover, E. coli pstSCAB mutants, unlike R. meliloti phoCDET mutants, are reported to grow normally in media containing 1 mM Pi (WILLSKY and MALAMY 1980 ). Phosphate transport in these pstSCAB mutants is believed to occur via a constitutive Pit phosphate transport system (ROSENBERG 1987 ).

We draw attention to the fact that the phoB gene is located 111 bp downstream from phoU in R. meliloti (GenBank accession no. M96261). As both of these genes are transcribed in the same direction, it is probable that phoU and phoB are transcribed as a single mRNA, in which case phoU insertion mutants would also be phenotypically PhoB^-. If this is the case, elucidation of the role of phoU in R. meliloti will require the construction of defined phoU ^- phoB⁺ strains. The low Pi-regulated expression of phoD and phoE observed in a phoU10 mutant background (Figure 4, dataset 2), together with the apparent Pi-dependent regulation of sfx1-directed orfA-pit expression also observed in the phoU10 background, (Figure 6, dataset 7) suggest that the phoU10 ::TnV insertion may allow a low level transcription of phoB.

The AP^- phenotype of the R. meliloti phoB mutants is similar to that observed for phoB mutants in other bacteria (LEE et al. 1989 ; ANBA et al. 1990 ). On the other hand, the AP^- phenotype of the phoU mutant is unusual as phoU mutants of both E. coli and Pseudomonas aeruginosa are constitutive for AP activity (STEED and WANNER 1993 ; KATO et al. 1994 ). E. coli phoU mutants grow poorly in MOPS media containing 2 mM Pi (STEED and WANNER 1993 ). While we have not seen any analogous phenotype with the R. meliloti phoU insertion mutant, we again note that it is very probable that these mutants are also phoB ^-.

Our screening experiment did not result in the isolation of any phoR mutants. It is possible that in the absence of PhoR other kinases cross-activate PhoB as observed in E. coli (WANNER and WILMES-RIESENBERG 1992 ; WANNER 1992 ).

Most studies of the Pho regulon have focused on genes whose expression increases in response to Pi limitation. However, over 20 years ago, WILLSKY and MALAMY 1976 showed that three E. coli periplasmic proteins were only synthesized during growth in excess phosphate medium and not during growth in phosphate-limited medium. Moreover, synthesis of two of these proteins was clearly derepressed in a phoB mutant background. More recently, VANBOGELEN et al. 1996 estimated that the synthesis rate of 413 proteins in E. coli was modified under phosphate limitation, of which 208 were induced and 205 repressed. They noted that the promoter regions of three repressed genes, ompF, pfl, and ssb do contain putative PhoB boxes. In addition, SMITH and PAYNE 1992 have suggested that PhoB may repress expression of periplasmic peptide transport-binding proteins under low phosphate conditions. They identified a putative PhoB box that seems to overlap with the RNA polymerase-binding site.

sfx1 and sfx2 were originally identified as mutations which suppressed the symbiotic Fix^- phenotype of phoCDET(ndvF ) mutants (ORESNIK et al. 1994 ). As sfx1, sfx2, increase orfA-pit expression and presumably OrfA-Pit-mediated Pi transport (S. D. BARDIN, R. VOEGELE, N. FALCIONI and T. M. FINAN, unpublished results; and this work, data not shown), we conclude that the symbiotic phenotype of phoCDET (ndvF ) mutants is a direct consequence of their inability to transport sufficient Pi for cellular growth.

	ACKNOWLEDGMENTS

We are grateful to BELINDA SCHOEMAN for technical assistance, to MICHAEL HYNES for the use of plasmid pLMS prior to publication and to RALF VOEGELE, MAGNE OSTERAS and ALISON COWIE for comments and discussions. This work was supported by operating and strategic grants from the Natural Sciences and Engineering Research Council of Canada to T.M.F.

Manuscript received September 10, 1997; Accepted for publication December 22, 1997.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AMEMURA, M., H. SHINAGAWA, K. MAKINO, N. OTSUJI, and A. NAKATA, 1982  Cloning of and complementation tests with alkaline phosphatase regulatory genes (phoS and phoT ) of Escherichia coli.. J. Bacteriol. 152:692-701[Abstract/Free Full Text].

ANBA, J., M. BIDAUD, M. L. VASIL, and A. LAZDUNSKI, 1990  Nucleotide sequence of the Pseudomonas aeruginosa phoB gene, the regulatory gene for the phosophate regulon. J. Bacteriol. 172(8):4686-4689.

BARDIN, S. D., S. DAN, M. OSTERAS, and T. M. FINAN, 1996  A phosphate transport system is required for symbiotic nitrogen fixation by Rhizobium meliloti.. J. Bacteriol. 178:4540-4547[Abstract/Free Full Text].

CHARLES, T. C. and T. M. FINAN, 1991  Analysis of a 1600-kilobase Rhizobium meliloti megaplasmid using defined deletions generated in vivo.. Genetics 127:5-20[Abstract].

CHARLES, T. C., W. NEWCOMB, and T. M. FINAN, 1991  ndvF, a novel locus located on megaplasmid pRmeSU47b (pEXO) of Rhizobium meliloti, is required for normal nodule development. J. Bacteriol. 173:3981-3992[Abstract/Free Full Text].

COX, G. B., D. WEBB, and H. ROSENBERG, 1989  Specific amino acid residues in both the PstB and PstC proteins are required for phosphate transport by the Escherichia coli Pst system. J. Bacteriol. 171:1531-1534[Abstract/Free Full Text].

COX, G. B., D. WEBB, J. GODOVAC-ZIMMERMANN, and H. ROSENBERG, 1988  Arg-220 of the PstA protein is required for phosphate transport through the phosphate-specific transport system in Escherichia coli but not for alkaline phosphatase repression. J. Bacteriol. 170:2283-2286[Abstract/Free Full Text].

COX, G. B., H. ROSENBERG, J. A. DOWNIE, and S. SILVER, 1981  Genetic analysis of mutants affected in the Pst inorganic phosphate transport system. J. Bacteriol. 148:1-9[Abstract/Free Full Text].

FINAN, T. M., E. HARTWIEG, K. LEMIEUX, K. BERGMAN, and G. C. WALKER et al., 1984  General transduction in Rhizobium meliloti.. J. Bacteriol. 159:120-124[Abstract/Free Full Text].

FINAN, T. M., B. KUNKEL, G. F. DE VOS, and E. R. SIGNER, 1986  Second symbiotic megaplasmid in Rhizobium meliloti carrying exopolysaccharide and thiamine synthesis genes. J. Bacteriol. 167:66-72[Abstract/Free Full Text].

FINAN, T. M., I. ORESNIK, and A. BOTTACIN, 1988  Mutants of Rhizobium meliloti defective in succinate metabolism. J. Bacteriol. 170:3396-3403[Abstract/Free Full Text].

FURUICHI, T., M. INOUYE, and S. INOUYE, 1985  Novel one-step cloning vector with transposable element: application to the Myxococcus xanthus genome. J. Bacteriol. 164:270-275[Abstract/Free Full Text].

KATO, J., Y. SAKAI, T. NIKATA, and H. OHTAKE, 1994  Cloning and characterization of a Pseudomonas aeroginosa gene involved in the negative regulation of phosphate taxis. J. Bacteriol. 176:5874-5877[Abstract/Free Full Text].

KLEIN, S., K. LOHMAN, R. CLOVER, G. C. WALKER, and E. R. SIGNER, 1992  A directional high frequency chromosomal mobilization system for genetic mapping of Rhizobium meliloti.. J. Bacteriol. 174:324-326[Abstract/Free Full Text].

LEE, T.-Y., K. MAKINO, H. SHINAGAWA, M. AMEMURA, and A. NAKATA, 1989  Phosphate regulon members of the family Enterobacteriaceae : comparison of the phoB-phoR operons of Escherichia coli, Shigella dysenteriae, and Klebsiella pneumoniae.. J. Bacteriol. 171:6593-6599[Abstract/Free Full Text].

MAKINO, K., H. SHINAWAGA, M. AMEMURA, T. KAWAMOTO, and M. YAMASSA et al., 1989  Signal transduction in the phosphate regulon of Escherichia coli involves phosphotransfer between PhoR and PhoB proteins. J. Mol. Biol. 210:551-559[Medline].

MAKINO, K., M. AMEMURA, T. KAWAMOTO, S. KIMURA, and H. SHINAGAWA et al., 1996  DNA binding of PhoB and its interaction with RNA polymerase. J. Mol. Biol. 259:15-26[Medline].

MAKINO, K., M. AMEMURA, S.-K. KIM, A. NAKATA and H. SHINAWAGA, 1994 Mechanism of transcriptional activation of the phosphate regulon in Escherichia coli, pp. 5–12 in Phosphate in Microorganisms: Cellular and Molecular Biology, edited by A. TORRIANI-GORINI, E. YAGIL and S. SILVER. ASM Press, Washington, D.C.

ORESNIK, I. J., T. C. CHARLES, and T. M. FINAN, 1994  Second site mutation specifically suppresses the Fix^- phenotype of Rhizobium meliloti ndvF mutations on alfalfa: identification of a conditional ndvF-dependent mucoid colony phenotype. Genetics 136:1233-1343[Abstract].

PARKINSON, J. S., 1993  Signal transduction schemes in bacteria. Cell 73:857-871[Medline].

RONSON, C. W., B. T. NIXON, and F. M. AUSUBEL, 1987  Conserved domains in bacterial regulatory proteins that respond to environmental stimuli. Cell 49:579-581[Medline].

ROSENBERG, H., 1987 Phosphate transport in prokaryotes, pp. 205–248 in Ion Transport in Prokaryotes, edited by B. P. ROSEN and S. SILVER. Academic Press, San Diego.

ROSENBERG, H., N. MEDVECZKY, and J. M. LA NAUZE, 1969  Phosphate transport in Bacillus cereus.. Biochim. Biophys. Acta 193:159-167[Medline].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning. A Laboratory Manual, Second Edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SMITH, M. W. and J. W. PAYNE, 1992  Expression of periplasmic binding proteins for peptide transport is subject to negative regulation by phosphate limiting condition in Escherichia coli.. FEMS Microbiology Letters. 79:183-190[Medline].

SPAINK, H. P., R. J. H. OKKER, C. A. WIJFFELMAN, E. PEES, and B. J. J. LUGTENBERG, 1987  Promoters in the nodulation region of the Rhizobium leguminosarum Sym plasmid pRL1J1. Plant Mol. Biol. 9:27-39.

STEED, P. M. and B. L. WANNER, 1993  Use of the rep technique for allele replacement to construct mutants with deletions of the pstSCAB-phoU operon: evidence of a new role for the PhoU protein in the phosphate regulon. J. Bacteriol. 175:6797-6809[Abstract/Free Full Text].

TOMMASSEN, J., P. DE GEUS, B. LUGTENBERG, J. HACKETT, and P. REEVES, 1982  Regulation of the pho regulon of Escherichia coli K-12. Cloning of the regulatory genes phoB and phoR and identification of their gene products. J. Mol. Biol. 157:265-274[Medline].

VANBOGELEN, R. T., E. R. OLSON, B. L. WANNER, and F. C. NEIDHARDT, 1996  Global analysis of proteins synthesized during phosphorus restriction in Escherichia coli.. J. Bacteriol. 178:4344-4366[Abstract/Free Full Text].

VAN VEEN, H. W., T. ABEE, G. J. J. KORTSTEE, W. N. KONINGS, and A. J. B. ZEHNDER, 1994  Translocation of metal phosphate via the phosphate inorganic transport system of Escherichia coli.. Biochemistry 33:1766-1770[Medline].

VAN VEEN, H. W., T. ABEE, G. J. J. KORTSTEE, W. N. KONINGS, and A. J. B. ZEHNDER, 1993a  Characterization of two phosphate transport systems in Acinetobacter johnsonii 210A. J. Bacteriol. 175:200-206[Abstract/Free Full Text].

VAN VEEN, H. W., T. ABEE, G. J. J. KORTSTEE, W. N. KONINGS, and A. J. B. ZEHNDER, 1993b  Mechanism and energetics of the secondary phosphate transport system of Acinetobacter johnsonii 210A. J. Biol. Chem. 268:19377-19383[Abstract/Free Full Text].

VOEGELE, R. T., S. D. BARDIN, and T. M. FINAN, 1997  Characterization of the high and low affinity phosphate uptake systems in Rhizobium meliloti.. J. Bacteriol. 179:7226-7232[Abstract/Free Full Text].

WANNER, B. L., 1992  Is cross regulation by phosphorylation of two-component response regulator proteins important in bacteria. J. Bacteriol. 174:2053-2058[Free Full Text].

WANNER, B. L., 1993  Gene regulation by phosphate in enteric bacteria. J. Cell. Biochem. 51:47-54[Medline].

WANNER, B. L., 1996 Phosphorus assimilation and control of the phosphate regulon, pp. 1357–1381 in Escherichia coli and Salmonella: Cellular and Molecular Biology, Ed. 2, edited by F. C. NEIDHARDT. ASM Press, Washington, DC.

WANNER, B. L. and M. R. WILMES-RIESENBERG, 1992  Involvement of Phosphotransacetylase, acetate kinase, and acetyl phosphate synthesis in control of the phosphate regulon in Escherichia coli.. J. Bacteriol. 174:2124-2130[Abstract/Free Full Text].

WILLSKY, G. R. and M. H. MALAMY, 1980  Characterization of two genetically separable inorganic phosphate transport systems in Escherichia coli.. J. Bacteriol. 144:356-365[Abstract/Free Full Text].

WILLSKY, G. R. and M. H. MALAMY, 1976  Control of the synthesis of alkaline phosphatase and the phosphate-binding protein in Escherichia coli.. J. Bacteriol. 127:595-609[Abstract/Free Full Text].

ZHAN, H., C. C. LEE, and J. A. LEIGH, 1991  Induction of the second exopolysaccharide (EPSb) in Rhizobium meliloti SU-47 by low phosphate concentrations. J. Bacteriol. 173:7391-7394[Abstract/Free Full Text].

This article has been cited by other articles:

				L. S. Basconcillo, R. Zaheer, T. M. Finan, and B. E. McCarry A shotgun lipidomics approach in Sinorhizobium meliloti as a tool in functional genomics J. Lipid Res., June 1, 2009; 50(6): 1120 - 1132. [Abstract] [Full Text] [PDF]

				F. Santos-Beneit, A. Rodriguez-Garcia, E. Franco-Dominguez, and J. F. Martin Phosphate-dependent regulation of the low- and high-affinity transport systems in the model actinomycete Streptomyces coelicolor Microbiology, August 1, 2008; 154(8): 2356 - 2370. [Abstract] [Full Text] [PDF]

				S. Gebhard, S. L. Tran, and G. M. Cook The Phn system of Mycobacterium smegmatis: a second high-affinity ABC-transporter for phosphate. Microbiology, November 1, 2006; 152(Pt 11): 3453 - 3465. [Abstract] [Full Text] [PDF]

				Z.-C. Yuan, R. Zaheer, and T. M. Finan Regulation and Properties of PstSCAB, a High-Affinity, High-Velocity Phosphate Transport System of Sinorhizobium meliloti J. Bacteriol., February 1, 2006; 188(3): 1089 - 1102. [Abstract] [Full Text] [PDF]

				Z.-C. Yuan, R. Zaheer, R. Morton, and T. M. Finan Genome prediction of PhoB regulated promoters in Sinorhizobium meliloti and twelve proteobacteria. Nucleic Acids Res., January 1, 2006; 34(9): 2686 - 2697. [Abstract] [Full Text] [PDF]

				T. Danhorn, M. Hentzer, M. Givskov, M. R. Parsek, and C. Fuqua Phosphorus Limitation Enhances Biofilm Formation of the Plant Pathogen Agrobacterium tumefaciens through the PhoR-PhoB Regulatory System J. Bacteriol., July 15, 2004; 186(14): 4492 - 4501. [Abstract] [Full Text] [PDF]

				R. M. Harris, D. C. Webb, S. M. Howitt, and G. B. Cox Characterization of PitA and PitB from Escherichia coli J. Bacteriol., September 1, 2001; 183(17): 5008 - 5014. [Abstract] [Full Text] [PDF]

				K. E. Mendrygal and J. E. González Environmental Regulation of Exopolysaccharide Production in Sinorhizobium meliloti J. Bacteriol., February 1, 2000; 182(3): 599 - 606. [Abstract] [Full Text]

				M. Gonin, E. M. Quardokus, D. O'Donnol, J. Maddock, and Y. V. Brun Regulation of Stalk Elongation by Phosphate in Caulobacter crescentus J. Bacteriol., January 15, 2000; 182(2): 337 - 347. [Abstract] [Full Text]

				L. M. Botero, T. S. Al-Niemi, and T. R. McDermott Characterization of Two Inducible Phosphate Transport Systems in Rhizobium tropici Appl. Envir. Microbiol., January 1, 2000; 66(1): 15 - 22. [Abstract] [Full Text]

				M. L. Summers, M. C. Denton, and T. R. McDermott Genes Coding for Phosphotransacetylase and Acetate Kinase in Sinorhizobium meliloti Are in an Operon That Is Inducible by Phosphate Stress and Controlled by PhoB J. Bacteriol., April 1, 1999; 181(7): 2217 - 2224. [Abstract] [Full Text]

				G. F. Parker, T. P. Higgins, T. Hawkes, and R. L. Robson Rhizobium (Sinorhizobium) meliloti phn Genes: Characterization and Identification of Their Protein Products J. Bacteriol., January 15, 1999; 181(2): 389 - 395. [Abstract] [Full Text]

				S. D. Bardin, R. T. Voegele, and T. M. Finan Phosphate Assimilation in Rhizobium (Sinorhizobium) meliloti: Identification of a pit-Like Gene J. Bacteriol., August 15, 1998; 180(16): 4219 - 4226. [Abstract] [Full Text]