The pha2 gene cluster involved in Na+ resistance and adaption to alkaline pH in Sinorhizobium fredii RT19 encodes a monovalent cation/proton antiporter

Authors

  • Lifu Yang,

    1. Department of Microbiology and Immunology, College of Biological Sciences, China Agricultural University and Key Laboratory of Agro-Microbial Resource and Application, Ministry of Agriculture, Beijing, China
    2. Rubber Research Institute, Chinese Academy of Tropical Agricultural Sciences, and Key Laboratory of Physiology for Tropical Crops of Ministry of Agriculture, Hainan, China
    Search for more papers by this author
  • Juquan Jiang,

    1. Department of Microbiology and Immunology, College of Biological Sciences, China Agricultural University and Key Laboratory of Agro-Microbial Resource and Application, Ministry of Agriculture, Beijing, China
    Search for more papers by this author
  • Wei Wei,

    1. Department of Microbiology and Immunology, College of Biological Sciences, China Agricultural University and Key Laboratory of Agro-Microbial Resource and Application, Ministry of Agriculture, Beijing, China
    Search for more papers by this author
  • Bo Zhang,

    1. Department of Microbiology and Immunology, College of Biological Sciences, China Agricultural University and Key Laboratory of Agro-Microbial Resource and Application, Ministry of Agriculture, Beijing, China
    Search for more papers by this author
  • Lei Wang,

    1. Department of Microbiology and Immunology, College of Biological Sciences, China Agricultural University and Key Laboratory of Agro-Microbial Resource and Application, Ministry of Agriculture, Beijing, China
    Search for more papers by this author
  • Susheng Yang

    1. Department of Microbiology and Immunology, College of Biological Sciences, China Agricultural University and Key Laboratory of Agro-Microbial Resource and Application, Ministry of Agriculture, Beijing, China
    Search for more papers by this author

  • Editor: Aharon Oren

Correspondence: Susheng Yang, Department of Microbiology, College of Biological Sciences, China Agricultural University, No. 2 Yuanmingyuan West Road, Beijing, 100094, China. Tel.: +86 10 62732674; fax: +86 10 62731332; e-mail: yangssh@cau.edu.cn

Abstract

Sinorhizobium fredii RT19 can tolerate up to 0.6 M NaCl, whereas all its pha2-disrupted mutants, constructed by Tn5 mutagenesis, failed to grow in even the presence of 0.1 M NaCl. No growth difference was detected in pha2 mutants at a pH <7.5 in the presence or absence of K+, but growth reduction was observed in the presence of K+ when pH >7.5. The pha2 gene cluster was able to completely restore the growth of the pha2 mutants of S. fredii RT19 in 0.6 M NaCl. Measurement of monovalent cation intracellular content suggested that pha2 was involved in both Na+ (Li+) and K+ efflux. The pha2 mutants exhibited K+/H+, but no apparent Na+(Li+)/H+ antiporter activity in everted membrane vesicles. Taken together, these results indicated that the pha2 cluster of S. fredii RT19 encodes a monovalent cation/proton antiporter involved in resistance to Na+ and adaption to pH, which was very different from the pha1 cluster of Sinorhizobium meliloti, which encodes a K+/H+ antiporter.

Introduction

Monovalent cation/proton antiporters are secondary active transporters that catalyze the efflux of intracellular monovalent cations (such as Na+, K+ and Li+) in exchange for external protons, which play an essential role in reducing the cytoplasmic concentration of toxic cations and in supporting Na+/K+-dependent cytoplasmic pH homeostasis under alkaline conditions (Ito et al., 1999; Padan et al., 2005). Most of the monovalent cation/proton antiporters reported to date have been single gene products (Padan et al., 2001). Antiporters composed of multiple subunits (usually six to seven), that are proposed to function as an hetero-oligomeric complex, were first identified in Bacillus halodurans C-125 (Hamamoto et al., 1994). Activities for the homologues of these unusual antiporters, such as Mrp/Sha in Bacillus subtilis (Ito et al., 1999; Kosono et al., 2000), Pha in Sinorhizobium meliloti (Putnoky et al., 1998) and Mnh of Staphylococcus aureus (Hiramatsu et al., 1998) have also been reported. Widely distributed in bacterial genomes and classified as family CPA3 (cation:proton antiorter-3) in the Transporter Classification system (see http://www.tcdb.orgwebsite), all these tested Mrp-type multigene antiporters are involved in resistance to Na+ or K+ and adaptation to alkaline pH, as well as in specialized physiological functions.

Rhizobia can fix nitrogen in a symbiotic association with leguminous plants. However, environmental factors such as high osmotic stress often limit the effectiveness of these symbiotic systems and can lead to negative effects on rhizobia growth and nitrogen fixation (Swaraj & Bishnoi, 1999). Sinorhizobium fredii RT19 can tolerate NaCl at concentrations of up to 0.6 M, suggesting that it has special osmo-adaptation mechanisms for survival and growth, in order to maintain cytoplasmic Na+ homeostasis and pH levels. To date, most research on the salt tolerance of rhizobia has focused on their osmo-regulatory mechanisms (Osteras et al., 1998; Bosacri et al., 2002), and very little is known on how monovalent cation efflux systems contribute to salt resistance. The complete genome of S. meliloti 1021 contains two pha operons, namely phaA1C1D1E1F1G1 and phaA2B2C2D2E2F2G2. The pha1 gene cluster of S. meliloti is required for the invasion of nodule tissue to establish nitrogen-fixing symbiosis, and the pha1 mutants are K+-sensitive, but not Na+-sensitive, suggesting that the pha1 encodes a K+/H+ antiporter which is involved in pH adaption during the infection process (Putnoky et al., 1998). In our previous study, 21 salt-sensitive mutants were constructed by screening a transposon Tn5-1063 library from S. fredii RT19, nineteen of which showed disruptions in the pha2 gene region (eight in phaA2, seven in phaD2, two in phaF2, and two in phaG2). Since all the nineteen pha2 mutants of S. fredii RT19 showed similar growth phenotypes with respect to Na+ and Li+ sensitivity (Jiang et al., 2004), one of the pha2 mutants, RTa-1 (mutated in the phaA2 gene), was selected to study the roles of the pha2 cluster in resistance to Na+ and adaptation to pH.

Materials and methods

Strains, plasmids and growth conditions

Sinorhizobium fredii RT19 and its salt-sensitive Tn5 mutant RTa-1 were grown in TY or FY medium (Jiang et al., 2004) for 2 days at 28°C. The cloning and shuttle vector pBBR1-MCS5 (Gmr) was used for complementation experiments in rhizobium strains. The recombination plasmid pZBE-ZFC-MCS5 was constructed by ligating the complete pha2 cluster to pBBR1-MCS5. To determine the sensitivity of S. fredii RT19 and RTa-1 to KCl, 2-day-old cultures in TY medium were diluted 100-fold into YET medium containing 0.1% (w/v) yeast extract, 5 mM Tris (pH 7.5), 1 mM MgSO4, and different concentrations of KCl from 0 to 100 mM, as described previously (Putnoky et al., 1998). To examine the tolerance of S. fredii RT19 and RTa-1 to different pHs, YEM medium was used in which the Tris buffer was replaced by 5 mM MES-Tris buffer and adjusted to different pH values with or without 80 mM KCl. Sinorhizobium fredii RT19 and RTa-1 were grown in YET medium for 30 h, with shaking of 180 r.p.m. at 28°C, and then inoculated into the YEM medium at a 1% concentration under the same culture conditions. Gentamicin (Gm) and Kanamycin (Km) were used at the final concentrations of 50 μg mL−1. Growth was determined by monitoring cell concentrations at OD600.

Measurements of monovalent cation intracellular contents

To measure the intracellular concentrations of Na+, Li+ or K+, cells were harvested after 2 days' growth in liquid TY media at 28°C, and incubated in 0.6 M NaCl, 0.4 M LiCl or 0.6 M KCl/10 mM Tris-HCl (pH 7.2) for 0–60 min at 28°C. As described by Nakamura et al. (1982), the incubation reaction was terminated by the addition of 0.6 M or 0.4 M choline chloride/10 mM Tris-HCl (pH 7.2), after which the cells were washed three times with the same isotonic buffered solution. Monovalent cation intracellular contents were measured using a Hitachi Z-5000 Polarized Zeeman atomic absorption spectrophotometer. Monovalent cation concentrations in the cells were expressed as nmoles mg−1 protein.

Preparation of inverted membrane vesicles and measurement of monovalent cation/H+ antiporter activity

Cells of S. fredii RT19 and RTa-1 (also RTd-1, RTf-1 and RTg-1) were grown in TY medium up to the stationary phase at 28°C, and harvested by centrifugation at 5000 g, 4°C for 10 min. The cells were then washed with buffer containing 10 mM Tris-HCl (pH 7.4), 0.14 M choline chloride, 0.5 mM dithiothreitol and 0.25 M sucrose (buffer A). Everted membrane vesicles were prepared by the French press method at 2000 p.s.i. and collected by ultra-centrifugation at 100 000 g for 1 h as described by Rosen (1986). The vesicles were re-suspended in buffer A and stored at −70°C before use.

The antiporter activity of the everted membrane vesicles was measured based on their ability to collapse a trans-membrane pH gradient as monitored by acridine orange fluorescence (Rosen, 1986). The reaction mixture contained 10 mM Tris-HCl (pH 8.0), 0.14 M choline chloride for the K+/H+ antiporter assay or 0.3 M KCl for Na+/H+ and Li+/H+ antiporter assay, 5 mM MgCl2, 2 μM acridine orange, and 20 to 40 μg mL−1 protein of membrane vesicles. Quenching was initiated by the addition of 5 mM dl-lactate. Antiporter activity was estimated from the rate of fluorescence enhancement after the addition of 5 mM NaCl, LiCl or KCl, using a Hitachi F-4500 fluorescence spectrophotometer at excitation and emission wavelengths of 495 nm and 530 nm, respectively.

Complementation experiments

Plasmid pZBE-ZFC-MCS5 containing the full length pha2 cluster was introduced into the salt-sensitive mutant S. fredii RTa-1 by three-parental mating, as described by de Bruijn & Rossbach (1994). A growth test was carried out as follows: the rhizobial strains were grown in 5 ml TY media plus Km for 2 days, and then 1% culture was added into 5-mL aliquots of the FY media supplemented with different NaCl concentrations (0 to 0.6 M) and incubated for 4 days. Growth was determined by measuring optical densities at 600 nm.

DNA manipulation and sequence analysis

Plasmid DNA preparation, restriction enzyme digestions and ligation reactions were carried out as described previously (Sambrook et al., 1989). Nucleotide-nucleotide blastss were done using the NCBI blast program. Hydropathy analysis was done at the website http://www.sbc.su.se/~erikw/toppred2/.

Protein content determination

Protein content in both cells and everted membrane vesicles was determined by the method of Lowry et al. (1951) with bovine serum albumin as a standard.

Results and discussion

Primary structure characterization of the pha2 gene cluster

A series of highly salt-sensitive Tn5 mutants from a ∼6 kb DNA region were obtained from previous studies (Jiang et al., 2004). A sequence analysis showed that this region consisted of one putative common promoter and seven ORFs, most of which overlap with or terminate very closely to the following one, suggesting that they form a single transcription unit (Fig. 1). Each of them has an upstream ribosomal binding site sequence (RBS) and starts with ATG. The seven ORFs share the highest identity (81–84%) to the phaA2, B2, C2, D2, E2, F2 and G2 of S. meliloti 1021, respectively. These ORFs also showed a high similarity to mnhA, B, C, D, E, F and G of Agrobacterium tumefaciens, and nhaA, B, C, D, E, F and G of Coxiella burnetii, and both clusters encode subunits of a putative multi-subunit Na+/H+ antiporter.

Figure 1.

 Organization of the pha2 gene cluster in Sinorhizobium fredii. Seven genes of pha2 were designated in order from A2 to G2, respectively. P indicates a putative promoter, and the circles with crosses under A2, D2, F2 and G2 represented mutated genes.

Based on the deduced amino acid sequences, the putative protein (PhaA2 to PhaG2) consists of 791, 139, 125, 519, 158, 128 and 128 amino acid residues, respectively. An hydropathy profile analysis suggested that the corresponding products were highly hydrophobic trans-membrane proteins, having 21, 4, 3, 14, 3, 3 and 2 trans-membrane fragments, respectively.

K+ and pH sensitivity of the wild-type strain and pha2 mutant

Physiological studies showed that S. fredii RT19 can grow in the presence of 0.6 M NaCl while its pha2 mutants cannot do so, even at 0.1 M NaCl in FY medium, indicating that pha2 plays a key role in Na+ resistance. To test the effects of K+ concentration on growth, S. fredii RT19 and RTa-1 were grown in a YET medium containing different concentrations of KCl. As shown in Fig. 2, S. fredii RT19 grew well in the presence of 100 mM KCl, while RTa-1 growth was significantly inhibited when the concentration of KCl was more than 80 mM, suggesting that the RTa-1 mutant partially lost the capacity to transport K+ out of cells.

Figure 2.

 Effect of K+ concentration on the growth of Sinorhizobium fredii RT19 and its mutant RTa-1. RT19 (open circles) and RTa-1 (closed circles) were grown in YET medium containing different concentrations of KCl, from 0 to 100 mM. Growth was detected by monitoring at OD600. The average of three independent experiments is shown with the error bars indicating the standard deviation.

To determine pH sensitivity, S. fredii RT19 and RTa-1 were grown in YEM medium at different pHs, from 6 to 9, in the presence or absence of 80 mM KCl. As shown in Fig. 3, S. fredii RT19 had almost the same growth rate at each corresponding pH in both conditions; moreover, no growth difference was detected in the RTa-1 mutant at a pH below 7.5, with or without 80 mM KCl. Comparing this to the growth of RTa-1 without 80 mM KCl, a growth reduction was observed in the presence of 80 mM KCl when the pH was more than 7.5, indicating that phaA2 is important for the growth of S. fredii RT19 at alkaline pH. With retarded cell growth, RTa-1 could still grow to some extent at pHs from 8 to 8.5 in the presence of 80 mM KCl.

Figure 3.

 Sensitivity to pH of Sinorhizobium fredii RT19 and its mutant RTa-1. RT19 (open symbols) and RTa-1 (closed symbols) was grown in YEM medium adjusted to different pH values with (triangle symbols) or without (circle symbols) 80 mM KCl. Each value shows the average of three independent determinations.

Monovalent cation/proton antiporter activity assay

Monovalent cation/proton antiporter activity with everted membrane vesicles prepared from cells of S. fredii RT19 and RTa-1 was determined by measuring the dequenching of acridine orange fluorescence following the addition of 5 mM NaCl, KCl or LiCl. Sinorhizobium fredii RT19 showed high Na+/H+ and Li+/H+ antiporter activities, while no apparent Na+/H+ and Li+/H+ antiporter activity was detected in everted membrane vesicles prepared from RTa-1 (Fig. 4a and b). K+/H+ antiporter activities were also detected in membrane vesicles of both S. fredii RT19 and RTa-1, although the latter was relatively lower (Fig. 4c). Taken together, these results demonstrated that phaA2 exhibited both Na+ (Li+)/H+ and K+/H+ antiporter properties.

Figure 4.

 Comparison of monovalent cation/H+ antiporter activities of Sinorhizobium fredii RT19 with those of RTa-1. The activity measurements for Na+/H+ antiporter (left), Li+/H+ antiporter (middle) and K+/H+ antiporter (right) were performed at pH 8.0 in everted membrane vesicles prepared from cells of RT19 or RTa-1 by the French press method. At the time points indicated by downward arrows, lactate (final concentration of 5 mM) was added to the assay mixture to initiate respiration. At the time points indicated by upward arrows, NaCl, LiCl or KCl (each to a final concentration of 5 mM) was added to the assay mixture to initiate fluorescence quenching. Fluorescence quenching is shown in arbitrary units. In each of a, b and c, RT19 is shown on the left-hand side and RTa-1 on the right-hand side.

Besides RTa-1, K+/H+ antiporter activities were also detected in membrane vesicles from RTd-1, RTf-1 and RTg-1 with mutations in phaD2, phaF2, and phaG2, respectively, although no apparent Na+/H+ and Li+/H+ antiporter activities were obtained in membrane vesicles from all of them.

It has been reported that the mnh cluster (mnhA to mnhG) encoding a Na+/H+ antiporter in Staphylococcus aureus is required to confer significant resistance to Na+ and alkaline pH (Hiramatsu et al., 1998), and that the mrp of B. subtilis encoding both Na+/H+ and K+/H+ antiporters is responsible for both sodium and cholate resistance and pH homeostasis (Ito et al., 1999, 2000). Studies have also shown that shaA plays a significant role in both Na+ extrusion and sporulation initiation in B. subtilis (Kosono et al., 2000). Disruption of yufT, the first gene of the mrp cluster in B. subtilis, resulted in a decrease of Na+/H+ antiporter activity and growth impairment when the external sodium concentration increased, indicating that yufT encodes a Na+/H+ antiporter which has a dominant role in the expulsion of cytotoxic sodium ions (Kosono et al., 1999). Transposon mutagenesis of Anabaena sp. PCC7120 showed that the mrpA mutant exhibited a pronounced Na+ sensitivity and an inhibition of photosynthesis (Blanco-Rivero et al., 2005). In our present study, all 19 pha2 mutants of S. fredii RT19 performed similarly in physiological tests, and were unable to grow in the presence of 0.1 M NaCl (Jiang et al., 2004). It is possible that there is a certain functional relationship or interaction among the subunits of Pha2, an idea which is consistent with previous suggestions (Swartz et al., 2005).

Most of the assayed CPA3 family are primarily Na+(Li+)/H+ antiporters (Swartz et al., 2005). Our studies on the pha2 mutants of S. fredii RT19 constructed by Tn5 mutagenesis indicated that the pha2 cluster encodes a monovalent cation/H+ antiporter, which was very different from the K+/H+ antiporter pha1 of the closely related species S. meliloti (Putnoky et al., 1998) and many other Mrp-type antiporters. In our previous study, we were unable to detect mutation in any other antiporter locii among up to 30 000 mutants, and most of the salt-sensitive mutants (19 out of 21) obtained were mutated in the pha2 locus (Jiang et al., 2004). Moreover, no significant Na+(Li+)/H+ antiporter activity was obtained from pha2 mutants. All the above information indicates that pha2 perhaps plays a dominant role in the salt tolerance of S. fredii RT19.

Comparison of monovalent cation intracellular contents between the wild-type and pha2 mutant

To further identify whether the pha2 gene cluster was related to cation efflux, we measured the intracellular cation contents of S. fredii RT19 and RTa-1. Over time, the Na+ intracellular contents of the wild-type remained relatively stable, whereas RTa-1 accumulated more and more Na+ and Li+, until the intracellular content of RTa-1 was also higher than that of RT19 (Fig. 5). Both RT19 and RTa-1 maintained relatively consistent intracellular K+ concentrations, although the latter was a little higher (Fig. 5). A comparison of Na+ (Li+) contents between the wild-type strain and the pha2 mutant showed that a mutation in the pha2 genes resulted in an increased Na+/Li+ accumulation in the cells, suggesting a role for pha2 in Na+ (Li+) efflux. No large fluctuation was observed in K+ concentration between the wild-type and pha2 mutant, indicating that pha2 genes perhaps do not play a dominant role in the K+ efflux of S. fredii RT19.

Figure 5.

 Measurements of monovalent cation intracellular contents of Sinorhizobium fredii RT19 and RTa-1. Monovalent cation intracellular contents of RT19 (open symbols) and RTa-1 (closed symbols) were measured using a Hitachi Z-5000 Polarized Zeeman atomic absorption spectrophotometer after the cells were incubated at the indicated times in the presence of different cations. The circle symbols, square symbols and triangle symbols represent intracellular concentrations of Na+, K+ and Li+, respectively.

Complementation of the pha2 mutant

In order to determine that the salt sensitivity of the pha2 mutants resulted from pha2 itself, other than from polar and nonpolar mutations, the full length of pha2 was cloned and ligated into the shuttle vector pBBR1-MCS5 and then introduced into S. fredii RTa-1. Growth experiments showed that the pha2 cluster completely restored the growth of the pha2 mutant at 0.6 M NaCl (Fig. 6). Moreover, the pha2 cluster complemented the phaD2, phaF2 and phaG2 mutants (data not shown).

Figure 6.

 Growth of phaA2 mutant complemented with the pha2 gene cluster at different concentrations of NaCl. Sinorhizobium fredii RT19 (open circles), RTa-1 with the pha2 gene cluster (closed squares), and negative control of RTa-1 with the empty vector pBBR1-MCS5 (closed circles) were grown in FY medium supplemented with different NaCl concentrations (0 to 0.6 M) and growth was determined at OD600 nm. Each value shows the average of three independent determinations.

The mnh operon of Staphylococcus aureus can complement the Escherichia coli mutant KNabc lacking three main Na+/H+ antiporter genes (nhaA, nhaB and chaA) (Hiramatsu et al., 1998). In this study, pBBR1-ZFC-MCS5 was also introduced into E. coli KNabc by electroporation. However, the salt-sensitive phenotype of the mutant was not complemented. It is possible that pha2 was not expressed, or that it was expressed at insufficient levels in E. coli KNabc under the given conditions, and the double mutant E. coli EP432 nhaAnhB or single mutant NM81 nhaA could be used for further studies on the characterization of Pha2 of S. fredii.

Acknowledgements

This work was supported by the China Key Base Research Developing Project Program (973 program, 2001CB108905) and the Chinese National Program for High Technology Research and Development (863 program, 2003AA241150).

Ancillary