INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Na⁺/H⁺ antiporters are integral membrane proteins present in virtually all cell types from bacteria to higher eukaryotes (9). In mammalian cells their function is primarily to maintain a neutral cytoplasmic pH (~7.2) by exchanging intracellular H⁺ for extracellular Na⁺. In bacteria, Na⁺/H⁺ antiporters normally function in the opposite direction, using the proton motive force to extrude Na⁺ or Li⁺ when present at toxic concentrations in the cytoplasm or to maintain a lower cytoplasmic pH in an alkaline environment (5, 19, 21). There are many isoforms of Na⁺/H⁺ antiporters in bacteria. The genome of Bacillus subtilis encodes two multifunctional antiporters, TetA(L) (6) and MrpA (16), and four additional Na⁺/H⁺ antiporters (22). TetA(L) is a tetracycline-metal/H⁺ antiporter that also exhibits a net K⁺ uptake and a monovalent cation/H⁺ antiporter mode, which has roles in pH homeostasis and Na⁺ resistance (5, 6, 10). MrpA (or ShaA) is a Na⁺(K⁺)/H⁺ antiporter that is required for the initiation of sporulation (20). MrpA is encoded by the first gene of the mrp operon that is required for the maintenance of pH homeostasis and resistance to cholate and Na⁺ (16). While TetA(L) and MrpA antiporters play key roles in Na⁺-dependent pH homeostasis and Na⁺ resistance, NhaC appeared to have only a modest role for maintaining pH homeostasis (17, 38). This was the case even in tetA(L) deletion strains in which the expression of NhaC was increased (38). B. subtilis carries a second NhaC-like antiporter, YqkI, that shows not only Na⁺/H⁺ exchange activity but couples malate-lactate exchange to proton uptake and Na⁺ efflux (39).

When the growth of B. subtilis becomes limited by the availability of phosphate, genes of the Pho regulon are either activated or repressed (24). These genes include phoA, phoB, and phoD encoding alkaline phosphatases (APases) and a phosphodiesterase-APase (13, 14); the pstSACB1B2 operon encoding a high-affinity phosphate transporter (34); the tuaABCDEFGH, tagAB, and tagDEF operons involved in teichuronic acid and teichoic acid synthesis (23, 25, 28); glpQ encoding a glycerophosphoryl diester phosphodiesterase involved in the hydrolysis of deacylated phospholipids (2); and two genes, ydhF and ykoL, of unknown function (2, 35). The members of the Pho regulon are controlled by the interaction of at least three two-component signal transduction systems (13). The center of this regulatory network is the PhoP-PhoR sensor-regulator system (15). During phosphate limitation, the PhoP response regulator is activated by its cognate sensor-kinase, PhoR. Phosphorylated PhoP (PhoP~P) is required for the induction or repression of genes in the Pho regulon and to enhance the transcription of the phoPR and resABCDE operons. The second signal-transduction system, ResD-ResE, is required for the full induction of the Pho regulon, while the third response regulator, Spo0A, terminates the phosphate response and initiates sporulation if phosphate starvation conditions persist.

We report here the influence on APase production, growth, and expression of the phoPR operon and phoA and phoB genes of inactivating and overexpressing nhaC. We also report the effects of Na⁺ and other monovalent cations on the growth, APase production, and expression of nhaC itself. Finally, we investigated the influence of NhaC on the production of secretory proteins which are not members of the Pho regulon.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, plasmids, primers, and media. Bacterial strains and plasmids are listed in Table 1, and primers are described in Table 2. Strains were grown in Luria-Bertani (LB) medium, low-phosphate medium (LPM), or high-phosphate medium (HPM) (31). The concentration of phosphate was 0.42 mM in LPM and 5.0 mM in HPM. Both media contained 6.8 mM sodium citrate. In LPMK and HPMK, sodium citrate was replaced by 6.8 mM potassium citrate, while LPMK20Na was supplemented with 20 mM NaCl, and LPMK200Na was supplemented with 200 mM NaCl. NB20Na contained 8 g of Nutrient Broth (Merck, Darmstadt, Germany) per liter, 1 mM MgSO₄, and 20 mM NaCl. In agar medium NB20Na was solidified with 1.5% agar. When required, the concentrations of antibiotics were 100 µg of ampicillin and 50 µg of kanamycin per ml for Escherichia coli and 0.3 µg of erythromycin, 25 µg of lincomycin, 12.5 µg of tetracycline, 5 µg of chloramphenicol, and 10 µg of kanamycin per ml for B. subtilis. 5-Bromo-4-chloro-3-indolylphosphate (BCIP) was used at 100 µg per ml, 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) was used at 100 µg per ml, and isopropyl--D-thiogalactopyranoside (IPTG) was used at 1 mM.

DNA manipulation and general methods. Plasmid DNA extraction, restriction endonuclease digestion, ligation, agarose gel electrophoresis, and transformation of electrocompetent E. coli cells were carried out as described previously (33). Enzymes, molecular size markers and deoxynucleotides were purchased from Roche Diagnostics, Ltd. (Lewes, United Kingdom), or Amersham Pharmacia Biotech, Ltd. (Little Chalfont, United Kingdom). Extraction of B. subtilis DNA and transformation of B. subtilis by the Groningen method was according to Bron (4). PCR was carried out with Pfu DNA polymerase (Stratagene Europe, Amsterdam, The Netherlands) or Taq DNA polymerase (Promega UK, Ltd., Southampton, United Kingdom) using the following cycling program: 1 cycle of 5 min at 94°C and then 35 cycles of 1 min at 94°C, 1 min at 60°C, and 1 min to 4 min, depending on the size of the PCR product, at 72°C.

Construction of plasmids. Primers NHACK-FOR and NHACK-REV (Table 2) were used for PCR amplification of a 488-bp internal fragment of nhaC, and primers NHACF-FOR and NHACF-REV were used to amplify a 332-bp fragment containing the putative ribosome-binding site (RBS) and 5' end of nhaC. Primers YHEKF-FOR and YHEKF-REV were used to amplify a 307-bp fragment containing the putative RBS and 5' end of yheK, and primers PSTSK-FOR and PSTSK-REV were used for PCR amplification of a 225-bp internal fragment of pstS. The PCR reactions were carried out with Pfu DNA polymerase using chromosomal DNA of B. subtilis 168 as template. After HindIII and BamHI digestion, the PCR fragments were ligated into HindIII- and BamHI-digested pMUTIN4 integrational vector (37) and transformed into electrocompetent cells of E. coli XL1-Blue (Stratagene Europe, Amsterdam, The Netherlands). Transformants were selected on LB agar medium supplemented with ampicillin. The resulting plasmids pNHACK, pNHACF, pYHEKF, and pSTSK were confirmed by restriction digestion and PCR using the insert-specific primers and the plasmid-specific primers MUT-FOR and MUT-REV (Tables 1 and 2).

Construction of mutants. The recombinant plasmids pNHACK, pNHACF, pYHEKF, and pSTSK were transformed into competent cells of B. subtilis, and transformants were selected on LB agar plates containing erythromycin and lincomycin. The mutants were analyzed by PCR to confirm the integration of a single copy of the plasmids into the target genes on the chromosome using a strategy similar to that described previously (32). In order to introduce a phoR deletion into B. subtilis 168, the integrational mutants were transformed with chromosomal DNA from MH5124 (15), selecting for tetracycline resistance (Tc^r). The insertion of the Tc^r marker in phoR of strains 168-PR, NHACF-PR, and YHEKF-PR was verified by PCR using PHOR1-FOR and PHOR2-REV primers. The sizes of the PCR product were 1,841 bp in the case of the wild-type and 3,538 bp in the case of phoR mutant strains.

Enzyme assays. Overnight cultures grown in HPM or HPMK were diluted 500-fold in fresh medium. The cultures were grown at 37°C with shaking at 220 rpm. Samples were collected at hourly intervals for the determination of the optical density at 600 nm, the APase activity (29), and the -galactosidase activity (27), as described previously (31). The concentration of P_i in the medium was assayed (11) after removal of the cells by filtration through a 0.45-µm-pore-size filter. P_i uptake assays (34) were carried out using ³²P_i (Amersham Pharmacia Biotech, Ltd.) in LPM containing 10 µM P_i rather than 0.42 mM (31).

Proteases and -amylase assays. Plate assays were used to monitor the secretion of -amylase and proteases (26). Overnight cultures of 168-Q and BFA1681-Q were diluted 50-fold in fresh NB20Na medium supplemented with chloramphenicol for 168-Q and with chloramphenicol and erythromycin for BFA1681-Q. After 3 h of growth at 37°C, strains were streaked onto NB20Na agar containing 1% starch (Sigma, St. Louis, Mo.) for the production of -amylase and 1% skim milk (Oxoid, Ltd., Basingstoke, United Kingdom) for the production of proteases. After 24 h of incubation at 37°C, the halos produced as a result of protein or starch hydrolysis were visualized directly in the case of proteases or after staining with 1% iodine solution in the case of -amylase.

Bioinformatical analyses. Database searches were carried out using the BLAST program (1). Multiple sequence alignments were carried out using the web site at the Institut National de la Recherche Agronomique (http://www.toulouse.inra.fr/multalin.html) (8). The symbol comparison table was Blosum62, the gap weight was 12, and the gap length weight was 2. For the prediction of transmembrane helices and topology of proteins, the HMMTOP automatic server (http: //www.enzim.hu/hmmtop) was used (36).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of BFA mutants influencing APase production. Within the framework of the B. subtilis functional analysis project (12), BFA mutants were constructed in unknown reading frames (URFs) without homologs of known function using the integration vector, pMUTIN (37). These mutants were used to identify URFs which influence the expression of the members of Pho regulon in a primary screen using the resident APase genes of the mutants as reporters. The mutants were tested on both low-phosphate agar (LPA) and high-phosphate-agar (HPA) plates supplemented with BCIP, which is a chromogenic substrate for APases. Under these conditions wild-type B. subtilis 168 gives blue colonies on LPA and white colonies on HPA. A mutation in a gene having an inducing effect on the synthesis of APases should give a white colony, while a mutation in a gene having a repressive effect should give a blue colony, irrespective of the phosphate concentration in the medium. Of the 1,146 BFA mutants tested, 1 showed a decrease in APase production, while 10 mutants showed dark blue colonies on LPA and light blue colonies on HPA. One of these mutants was BFA1681. The mutant phenotype was confirmed by transforming wild-type B. subtilis 168 with chromosomal DNA from BFA1681. All of the 100 erythromycin resistant (Em^r) transformants tested produced an increased amount of APases on LPA and HPA supplemented with BCIP as the parent, BFA1681. This confirmed that the APase overproduction was linked to the plasmid integration, and it suggested that nhaC (Fig. 1), encoding an Na⁺/H⁺ antiporter (38), influences the expression of members of the APase gene family and/or Pho regulon.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Construction of B. subtilis mutants in genes yheK and nhaC. Schematic representation of the integration of pMUTIN4 in mutants BFA1681, NHACF, and YHEKF. Filled thick arrows indicate structural genes, and putative Rho-independent terminators are shown with stem-loop structures. The promoter of yheK is marked with a fine broken arrow. Striped boxes show the tandem duplication of the internal part of nhaC in BFA1681 and the RBS and 5' end of nhaC and yheK in NHACF and YHEKF. nhaC' and yheK' are the 5' ends of nhaC and yheK, respectively. `nhaC is the 3' end of nhaC. Plasmid pMUTIN4 is shown as a thick line. The spoVG-lacZ reporter gene, lacI, Apr and Emr genes are marked with fine arrows; promoter Pspac is marked with a fine broken arrow above that of pMUTIN4. The region in pMUTIN4 used for replication in E. coli is labeled "ori," and three terminators (t1t2t0) upstream of Pspac are represented by a single stem-loop structure.

Production of APases in BFA1681. B. subtilis 168 and BFA1681 were grown in liquid LPM and HPM. BFA1681 exhibited a 1.3- to 1.5-fold increase in growth yield compared to that of strain 168, irrespective of the concentration of phosphate in the culture medium (Fig. 2A). In LPM, APase was induced at the same point in the growth cycle, T₁ (i.e., 1 h before the transition from exponential to stationary phase), in both 168 and BFA1681 (Fig. 2B). However, the kinetics of induction were different, and BFA1681 synthesized significantly more APase than the wild-type at all subsequent time points; 11-fold higher at time T₀ and 3- to 4-fold higher at T₃ and T₄. In contrast to the result of the primary screen, BFA1681 showed wild-type phenotype in HPM, producing no APases (Fig. 2B). A similar APase-negative phenotype was observed in BFA1681-PR, an nhaC-phoR double mutant in LPM (Fig. 2B). These data suggest that NhaC affects the synthesis of APases indirectly, probably via the expression or phosphorylation of one of the two-component systems that influence the regulation of the Pho regulon.

View larger version (22K): [in this window] [in a new window]   FIG. 2.   Effect of a knockout mutation in nhaC on growth (A) and APase production (B). Results for B. subtilis strains 168 (), 168-PR (168 carrying a phoR mutation) (*), BFA1681 (), and BFA1681-PR (BFA1681 carrying a phoR mutation) (+) in LPM and B. subtilis strains 168 () and BFA1681 () in HPM are as indicated. The gray symbols in panel B show the concentration of Pi in LPM containing B. subtilis 168 () and BFA1681 ().

View larger version (17K): [in this window] [in a new window]   FIG. 3.   Pi uptake by B. subtilis. Cells were grown in LPM and starved for phosphate, and Pi uptake was measured in medium containing 10 µM Pi as described in Materials and Methods. Results for B. subtilis strains 168 (), 168-PR (), BFA1681 () and PSTSK () are as indicated.

NhaC is responsible for the overproduction of APases. To determine whether the APase overproduction phenotype was due to the inactivation of nhaC rather than a polar effect on a downstream gene, BFA1681 was grown in LPM in the presence of 1 mM IPTG. Under these conditions, the P_spac promoter of the integrated pMUTIN4 plasmid (Fig. 1) should induce the expression of genes downstream of nhaC. BFA1681 exhibited the same increase in growth yield and APase production in the presence of IPTG (data not shown) as was observed in its absence (Fig. 2). Thus, APase overproduction and increased growth yield are due to the inactivation of nhaC rather than to a polar effect on downstream gene(s). This conclusion was supported by the presence of a putative Rho-independent transcription terminator (G = 28.4 kcal/mol) immediately downstream of nhaC (Fig. 1) (30).

Effect of sodium on growth and APase production. Since NhaC is an Na⁺/H⁺ antiporter, the influence of Na⁺ concentration was investigated with mutant NHACF, in which the expression of nhaC was under the control of the P_spac promoter (Fig. 1). LPM, which contains about 20.4 mM Na⁺, was replaced by LPMK. NHACF and B. subtilis 168 were grown in LPMK, LPMK20Na (LPMK supplemented with 20 mM NaCl; low Na⁺ medium), and LPMK200Na (LPMK supplemented with 200 mM NaCl; high Na⁺ medium). Whereas the growth rate and yield were unaffected in NHACF by increasing Na⁺ concentrations, both were reduced somewhat in the wild type and reduced markedly when NHACF was grown in the presence of IPTG (Fig. 4). This suggests that NhaC is involved in the uptake rather than the extrusion of Na⁺.

View larger version (23K): [in this window] [in a new window]   FIG. 4.   Influence of Na+ on growth and APase production. The growth (solid symbols) and APase production (open symbols) of B. subtilis 168 (diamonds), NHACF (triangles), and NHACF in the presence of 1 mM IPTG (circles) are as indicated. (A) LPMK medium. (B) LPMK20Na medium (LPMK with 20 mM NaCl). (C) LPMK200Na medium (LPMK with 200 mM NaCl).

Specificity to sodium. Since sodium affected the growth and APase synthesis of the wild-type and NHACF in the presence of IPTG (Fig. 4), the influence of other monovalent cations was investigated. NHACF and 168 were grown in LPMK supplemented with 200 mM LiCl, 200 mM KCl, or 200 mM NaCl. The growth and the production of APases of the wild type (Fig. 5B) and NHACF in the presence of IPTG (Fig. 5C) were affected by Na⁺ but not by Li⁺ or K⁺. When the expression of nhaC was not induced in NHACF, both parameters were unaffected, even in the presence of Na⁺ (Fig. 5A). These results indicate that the inhibition of the growth and production of APases is specific for the presence of NhaC and sodium and not lithium or potassium.

View larger version (20K): [in this window] [in a new window]   FIG. 5.   Influence of Na+, Li+, and K+ on growth and APase production. The growth (closed symbols) and APase production (open symbols) in LPMK supplemented with 200 mM NaCl (triangles), 200 mM LiCl (diamonds), and 200 mM KCl (circles) are as indicated for B. subtilis strains NHACF (A), 168 (B), and NHACF in the presence of 1 mM IPTG (C).

Influence of YheK on growth and APase production. To determine the effect of yheK, located upstream of nhaC, on growth and APase production, yheK was placed under the control of the P_spac promoter in mutant YHEKF (Fig. 1). If yheK and nhaC are in an operon, as was reported recently (38), the integration of pMUTIN4 into yheK should have a polar effect on nhaC, and therefore YHEKF should show the same phenotype as NHACF. In contrast to NHACF (Fig. 4B), YHEKF grown in LPMK20Na showed no increase in the growth yield and APase production (data not shown). The absence of a polar effect indicates that yheK is not in the same transcriptional unit as nhaC. In the presence of IPTG, the growth and APase activities of YHEKF were similar to that of the wild-type (data not shown), in contrast to the marked reduction in APase activity and growth yield observed in NHACF (Fig. 4B). These data suggest the presence of a potential transcription terminator between yheK and nhaC.

Transcriptional activities of yheK and nhaC. The transcription of yheK and nhaC were monitored using the spoVG-lacZ transcriptional reporter gene of the integrated pMUTIN4 in NHACF and YHEKF (Fig. 1). In the case of NHACF, a peak of activity was observed toward the end of the exponential phase, decreasing thereafter by about twofold (Fig. 6A) irrespective of the phosphate concentration. In the case of LPMK, the point at which nhaC expression decreased, T₁, coincided with the point at which APase production increased (Fig. 4A). In the phoR-null background (NHACF-PR), the expression of nhaC in LPMK continued to increase and reached a plateau at about T₁ (Fig. 6A). These data appeared to suggest that nhaC is a member of the Pho regulon, since its expression seemed to be repressed by phosphorylated PhoP (PhoP~P). However, this was not supported by data obtained in HPMK, where PhoP is not phosphorylated. In this case, the expression of nhaC increased during the exponential growth phase but started to decrease at T₄. Additionally, no putative PhoP~P binding sites (Pho boxes [25]) were found immediately upstream or within the 5' end of nhaC.

View larger version (21K): [in this window] [in a new window]   FIG. 6.   Transcriptional activities of yheK and nhaC. The specific -galactosidase activities of strains NHACF (A) and YHEKF (B) grown in HPMK (), LPMK (), LPMK20Na (), LPMK in the presence of 1 mM IPTG (), and LPMK20Na in the presence of 1 mM IPTG () are shown, as are the specific -galactosidase activities of strains NHACF-PR (A) and YHEKF-PR (B) in LPMK (*).

1

1

Effect of NhaC on transcription of phoPR. Our results (Fig. 2 and 3) indicated that NhaC influenced the synthesis of APases indirectly, probably via the regulatory network of the Pho regulon, which contains at least three two-component signal transduction systems, namely, PhoP-PhoR, ResD-ResE, and Spo0A (13). Among these regulators, PhoP and PhoR are the center of this regulatory network and are essential for the expression of members of Pho regulon. To investigate the effect of NhaC on the transcription of phoPR operon, a derivative of NHACF was constructed (NHACF-Km) in which the lacZ reporter downstream of the nhaC promoter was inactivated by insertion of a Km^r gene. A phoP-lacZ transcriptional fusion was then integrated into the amyE gene of strains NHACF-Km or 168, using chromosomal DNA of MH4050 (15) for the transformation. The resulting strains, NHACF-KmP and 168-P, were grown in LPMK20Na (Fig. 7). In strains 168-P and NHACF-KmP phoPR showed a constitutive, low-level expression during exponential growth phase but was induced at T₀ as the cells became phosphate starved. In the absence of NhaC (NHACF-KmP), the expression of phoPR and APase production were about threefold higher than in 168-P. When nhaC was induced by the addition of IPTG, the transcription of phoPR and the synthesis of APases at T₀ were abolished. These data suggest that the presence of NhaC influences the synthesis of APases by affecting the transcription of the phoPR operon.

View larger version (25K): [in this window] [in a new window]   FIG. 7.   Influence of NhaC on transcription of the phoPR operon. APase production (closed symbols) and specific -galactosidase activities (open symbols) for strains 168-P (amyE::phoP-lacZ) (diamonds), NHACF-KmP (amyE::phoP-lacZ nhaC::pNHACFSacI-ClaI::Kmr) (triangles), and NHACF-KmP in the presence of 1 mM IPTG (circles) grown in LPMK20Na medium (LPMK with 20 mM NaCl).

Effect of NhaC on transcription of phoA and phoB. APaseA and APaseB are responsible for 98% of the APase activity synthesized in response to phosphate starvation. APaseA, encoded by phoA, accounts for 60 to 80%, while APaseB, encoded by phoB, accounts for 25 to 35% of APase activity in phosphate-starved vegetative cells (15). To study the effect of NhaC on the transcription of these APase genes, phoA-lacZ or phoB-lacZ transcriptional fusions, derived from strains MH4040 (15) or MH6192, respectively, were integrated into the amyE gene of strains NHACF-Km and 168. The resulting strains, NHACF-KmA, NHACF-KmB, 168-A, and 168-B, were grown in LPMK20Na (Fig. 8). In the wild-type background and the nhaC mutant, the expression of phoA and phoB (Fig. 8B) and the synthesis of APases (Fig. 8A) were induced as the cells entered stationary phase. At all subsequent time points the transcriptional activity of phoA was ~2-fold higher than that of phoB. In the absence of NhaC the expression of phoA and phoB was enhanced two- to threefold (Fig. 8B), while the presence of IPTG-induced NhaC abolished the expression of both APase genes.

View larger version (21K): [in this window] [in a new window]   FIG. 8.   Influence of NhaC on transcription of phoA and phoB. APase production (A) and specific -galactosidase activities (B) for strains 168-A (amyE::phoA-lacZ) (), 168-B (amyE::phoB-lacZ) (), NHACF-KmA (amyE::phoA-lacZ nhaC::pNHACFSacI-ClaI::Kmr) (), NHACF-KmB (amyE::phoB-lacZ nhaC::pNHACFSacI-ClaI::Kmr) (), and NHACF-KmA (), and NHACF-KmB in the presence of 1 mM IPTG () grown in LPMK20Na medium (LPMK with 20 mM NaCl).

Production of proteases and AmyQ -amylase by BFA1681. Since NhaC has an effect on the production of APases, we determined whether the absence of NhaC enhances the production of secretory proteins which are not members of the Pho regulon. The production of the native proteases and the heterologous B. amyloliquefaciens -amylase, AmyQ, were monitored on NB20Na agar containing either skim milk or starch and supplemented with 20 mM NaCl. After incubation for 24 h at 37°C, the sizes of colonies and zones of hydrolysis produced by 168-Q and BFA1681-Q were similar (data not shown). These data indicate that the absence of NhaC influences the production of APases specifically (Fig. 2B) and not of secretory proteins which are not the members of the Pho regulon.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In B. subtilis, the presence of NhaC has a repressive effect on growth and the synthesis of APases in the presence of sodium. Conversely, in its absence the growth yield was higher and was not influenced by increasing amounts of sodium, while APase production was only slightly reduced at the higher (200 mM) sodium concentration (Fig. 4). In the case of wild-type (B. subtilis 168) and overproduced (NHACF grown in the presence of IPTG) levels of NhaC, growth and APase production were both reduced in the presence of Na⁺ (Fig. 4). Although we have not determined the level of NhaC overproduction with the P_spac promoter, the level of -galactosidase produced under the control of the P_spac promoter of pMUTIN4 is significantly higher (433 Miller units per mg protein [37]) than that produced under the control of the nhaC promoter (Fig. 6A). The data are consistent with this interpretation, since growth and APase production were reduced to a greater extent in the overexpressing mutant than in B. subtilis 168 (Fig. 4B and C) grown in the presence of sodium. These phenotypic changes were specific to Na⁺ and were not observed in the presence of other monovalent cations, such as Li⁺ and K⁺ (Fig. 5). The induced level of NhaC has no influence on growth and APase production in the absence of Na⁺ (Fig. 4A) or in the presence of other monovalent cations (Fig. 5C). If the overproduction of NhaC per se had a nonspecific toxic effect on the cell, as do certain other membrane proteins, it should be independent of the presence of specific cations.

In a previous report, when NhaC was expressed on a high-copy-number plasmid the resulting overproduction of this protein led to a marked increased sensitivity to cobalt (38). These authors suggested that the cells become sensitive to Co²⁺ because the barrier function of the membrane is compromised nonspecifically by the elevated levels of NhaC. In the case of the overproduction of NhaC in NHACF, it is unlikely that the increased sensitivity to Na⁺ is due to such a nonspecific effect on the barrier function of the membrane. Increasing Na⁺ concentrations leads to a decrease in growth and APase production both in NHACF in the presence of IPTG and in the wild type. Growth and APase production were decreased to similar levels in the wild type at high sodium concentrations (Fig. 4C), as they were in the induced NHACF mutant at low sodium concentrations (Fig. 4B).

In bacteria, Na⁺/H⁺ antiporters exchange intracellular Na⁺, Li⁺, or tetracycline for extracellular protons, either to maintain pH homeostasis upon pH upshift or to reduce their toxic effects on the cell (5, 9, 21). In B. subtilis, two multifunctional antiporters, TetA(L) and MrpA, play a key role in pH homeostasis and the extrusion of toxic Na⁺ or tetracycline from the cytoplasm (6, 16). The absence of TetA(L) or MrpA results in a major growth defect in the presence of Na⁺. In a nonpolar mrpA mutant, the expression of a distally located copy of mrpA from a P_spac promoter gave wild-type levels of Na⁺ resistance and Na⁺-dependent pH homeostasis and even faster protonophore-sensitive Na⁺ efflux (16). NhaC appears to have only a modest role in Na⁺-dependent pH homeostasis and no role in Na⁺ resistance (38). However, in Bacillus firmus the absence of NhaC led to a growth defect in medium containing low (25 mM) Na⁺ concentrations at pH 7.5 (17). In contrast, our data showed that growth rate and yield were unaffected by sodium in the absence of NhaC but were reduced progressively in both the wild type and the induced mutant as the concentration of Na⁺ was increased (Fig. 4). Similarly, B. subtilis JC112, in which the expression of nhaC was increased ninefold, showed a marked decrease in the MIC of Na⁺ compared with that for the wild type (38). Together, these results point to a role of NhaC in the uptake rather than the extrusion of Na⁺. In B. subtilis, the expression of certain ^H-dependent early sporulation genes is affected by Na⁺ in the absence of MrpA (20). In contrast, our data show that the production of APases was affected by Na⁺ in the presence of NhaC (Fig. 4) and indicates that these transporters function in opposite directions: MrpA in the extrusion of Na⁺ and NhaC in its uptake.

When Na⁺ is present in the medium, the absence of NhaC resulted in the hyperinduction of the phoPR operon, whereas IPTG-induced NhaC overproduction repressed this operon. Since NhaC has the characteristics of a membrane protein and is therefore unlikely to be a DNA-binding regulatory protein, these data suggest that Na⁺ may influence the activity of the PhoP-PhoR two-component signal transduction system. If, for example, Na⁺ prevented or reduced the formation of PhoP~P, this would account both for the reduced expression of phoPR (Fig. 7) and, consequently, of phoA and phoB (Fig. 8B).

In the NhaC mutant, the rate of removal of P_i from LPM was similar to that of the wild type (Fig. 2B), while the mutant showed a higher growth yield (Fig. 2A) and a reduced rate of P_i uptake (Fig. 3). These data indicate that the intracellular P_i concentration in the mutant was lower than that of the wild type. However, a mutation in the pst operon, encoding a high-affinity phosphate transporter, also affected the rate of P_i uptake (Fig. 3) but did not influence APase production (34). This suggests that factors other than intracellular P_i concentration, such as intracellular Na⁺, could also be important in modulating the expression of the Pho regulon.

Genes yheK and nhaC are adjacent to each other and in the same orientation on the chromosome. No transcription terminators were reported in the 127-bp intergenic region, although a putative terminator was identified immediately downstream of nhaC (30). Previous transcription studies have suggested that these genes comprise a bicistronic operon (38). In contrast, our data suggest that yheK and nhaC are in different transcriptional units, separated by a potential transcription terminator. This conclusion is based on the following findings: (i) the absence of a polar effect on the expression of nhaC in YHEKF; (ii) the overexpression of yheK did not lead to the overproduction of NhaC, as judged by an unaltered sensitivity to Na⁺ and APase production; and (iii) the transcriptional activities of yheK and nhaC were markedly different (Fig. 6). Analysis of the nucleotide sequence in the intergenic region of yheK and nhaC revealed the presence of two inverted repeats (IR). IR1, the upstream inverted repeat (G = 16.6 kcal/mol), had the characteristics of a Rho-independent transcription terminator. Its 5' end was located 3 bp upstream of the yheK stop codon and was followed by a T-rich region. IR2 (G = 20 kcal/mol) was located 51 bp downstream of IR1 and 22 bp upstream of the nhaC start codon. It was not followed by a T-rich region and, in view of its location, might be involved in controlling the expression of nhaC. These data suggest that nhaC is in a monocistronic operon, the promoter of which is likely to be downstream of IR1 and in the 86-bp region that includes IR2.

Our data show that when NhaC is overproduced, the expression of nhaC was repressed in the presence Na⁺ but not in the presence of Li⁺ or K⁺. In contrast, the expression of nhaC was reported to either not be affected by or slightly induced by 100 mM NaCl, using a translational nhaC-lacZ fusion in the amyE locus of B. subtilis (38).

An alignment of NhaC and 12 Na⁺/H⁺ antiporter homologs revealed a conserved motif, GDXXSXXSD (data not shown), containing two aspartyl and two serine residues. In the NhaA family of Na⁺/H⁺ antiporters, the aspartyl residues, located in two transmembrane segments, have been shown to be important in cation binding and transport (9). NhaC contains 12 putative membrane-spanning segments with the conserved motif located on the largest extracytoplasmic loop, between the fifth and sixth membrane-spanning segments. After the integration of pNHACF in mutant NHACF, the truncated NhaC protein (NhaC-97) has 97 amino acid residues of the N-terminal end of the native NhaC. In BFA1681 the truncated protein (NhaC-317) has 317 amino acid residues of the N-terminal of NhaC. Both NhaC-97 and NhaC-317 were unable to restore the wild-type phenotype. NhaC-317 has 70% of the 453 amino acid residues of NhaC containing 8 of the 12 membrane-spanning segments and 7 of the 11 loops, including the loop containing the GDXXSXXSD motif. These data indicate that the C-terminal end (317 to 453 amino acids) of NhaC is essential for function.