Intracellular Na (cid:49) Regulates Transcription of the ntp Operon Encoding a Vacuolar-type Na (cid:49) -translocating ATPase in Enterococcus hirae *

The Gram-positive bacterium Enterococcus hirae has a vacuolar-type Na (cid:49) -translocating ATPase that is encoded by the ntp operon ( ntpFIKECGABDHJ ) (Takase, K., Kakinuma, S., Yamato, I., Konishi, K., Igarashi, K., Kakinuma, Y. (1994) J. Biol. Chem. 269, 11037–11044). Primer extension experiments identified the start site of transcription of this operon upstream of the ntpF gene. In parallel with the increases of both Na (cid:49) -pumping activity in whole cells and Na (cid:49) -stimulated ATPase activity in the membranes, the amounts of the two major sub- units (A and B) of this enzyme increased remarkably in cells grown on medium containing high concentrations of NaCl but not on medium containing KCl or sorbitol. Chloramphenicol completely abolished the increases of the enzyme activity and the amounts of A and B subunits, suggesting that the Na (cid:49) -ATPase level increased by de novo synthesis of the enzyme with the stimulation of high concentrations of the external sodium ions. Finally, Western blot and Northern blot experiments re- vealed that the increase in the Na (cid:49) -ATPase level with the external Na (cid:49) was further accelerated by addition of an ionophore, such as monensin, which rendered the cell membrane permeable to

All living cells show Na ϩ circulation across the cell membrane. This circulation is driven by active transport systems that extrude sodium ions and maintain the Na ϩ concentration gradient directed inward (1)(2)(3); Na ϩ reenters the cell via a Na ϩ gradient consumer, such as Na ϩ -coupled secondary cotransport systems, as the widespread route (4). In animal cells, the familiar Na ϩ ,K ϩ -ATPase expels sodium ions, to which K ϩ uptake is tightly coupled. In bacteria, secondary Na ϩ -H ϩ antiporters, which are driven by the proton electrochemical gradient generated by proton pumps, are widely distributed to perform this work. However, it is now generally accepted that bacteria have evolved multiple sodium extrusion systems (primary and secondary ones) for the purpose of coping with environmental fluctuations (5).
The Gram-positive bacterium Enterococcus hirae has two sodium extrusion systems: the secondary Na ϩ /H ϩ antiporter (6,7) and an ATP-driven primary sodium pump (8), a vacuolartype Na ϩ -translocating ATPase (9 -14). This bacterium lacks the respiratory chain; ATP is produced by substrate-level phosphorylation via the glycolytic pathway. The proton electrochemical gradient is generated by proton expulsion via the F 0 F 1 , H ϩ -translocating ATPase, whose activity is optimal at around pH 6.5 (15,16). The proton gradient (the electrochemical potential of proton) is generated well at low pH, but it is minimal at high pH (17). Therefore, the secondary Na ϩ /H ϩ antiporter works well at extruding Na ϩ at low pH, but does not functional at high pH. Therefore, it can be considered that the Na ϩ -ATPase is especially important for Na ϩ extrusion at high pH (18).
The Na ϩ -ATPase is encoded by the ntp operon consisting of 11 ntp genes (ntpFIKECGABDHJ) (10 -12); Northern blotting experiments have revealed that this operon is transcribed as a single mRNA of more than 10 kilobase pairs (19). This Na ϩ -ATPase in E. hirae is not constitutive. We have observed that both Na ϩ -stimulated ATP hydrolytic activity of the membranes and sodium-pumping activity in whole cells were relatively high in the cells grown in high Na ϩ medium (20,21). Since (i) the V max value of the Na ϩ -stimulated ATPase activity of the membranes increased in the cells grown in high Na ϩ medium (20,22), and (ii) Western blotting revealed that the amount of the catalytic V 1 moiety of the Na ϩ -ATPase increased in parallel with the Na ϩ -stimulated ATPase activity in the membranes (13,19), the change in the Na ϩ -ATPase activity is likely to be due to a change of its abundance but not its enzymatic activation. We also found that the Na ϩ -ATPase activity and sodiumpumping activity increased by increasing the pH of the medium (22). Furthermore, an additional increase in its activity was observed where the proton electrochemical gradient was dissipated by a mutation of the F 0 F 1 , H ϩ -ATPase or by the inclusion of a protonophore such as carbonyl cyanide m-chlorophenylhydrazone in the medium (20,21). Since the Na ϩ /H ϩ antiporter does not lower the cytoplasmic Na ϩ concentration ([Na ϩ ] in ) 1 under these conditions, these results suggest that an increase in the Na ϩ -ATPase activity is induced by an increase of [Na ϩ ] in as the signal (18).
In this paper we examine the change in the activity and the quantity of the Na ϩ -ATPase or the amount of the mRNA for the ATPase under various growth conditions. The results indicate that de novo synthesis of E. hirae Na ϩ -ATPase is regulated specifically by [Na ϩ ] in at transcriptional level. * This work was supported by a grant-in-aid (to Y. K. and I. Y.) for scientific research on priority areas of channel-transporter correlation from the Ministry of Education, Science, Sports and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

MATERIALS AND METHODS
Strain and Growth Conditions-The E. hirae strain used was ATCC 9790 (wild-type strain), obtained from the American Type Culture Collection. Cells were grown in the complex medium KNY (0.2% Bacto neopeptone, 0.2% Bacto yeast extract, 1% glucose, and 0.85% KH 2 PO 4 ). The pH of the medium was adjusted to 7.5 with K 2 CO 3 . This medium was contaminated with less than 1 mM Na ϩ . When required, NaCl, KCl, or sorbitol was added to KNY medium. In some experiments, the standard medium (21), KTY (1% Bacto tryptone, 0.5% Bacto yeast extract, 1% glucose, and 1% K 2 HPO 4 ), or NaTY (1% Bacto tryptone, 0.5% Bacto yeast extract, 1% glucose, and 0.85% Na 2 HPO 4 ), was used. The pH of the medium of NaTY was alkalinized with Na 2 CO 3 . The cell growth was monitored by a photometer at 600 nm.
Transport Assay--Sodium extrusion was monitored using 22 Na ϩ as described previously (6). Cells harvested at the late-log phase were used directly as K ϩ -loaded cells. Washed cells were suspended at 4 mg (dry weight) per ml in 50 mM K ϩ -HEPES buffer (pH 7.5) containing 100 mM maleate-KOH with 20 mM 22 NaCl (2.315 MBq/mmol) and incubated at 25°C for 60 min. The reaction was initiated by addition of 10 mM glucose. In some assays, DCCD (N,NЈ-dicyclohexylcarbodiimide), TCS (tetrachlorosalicylanilide), and valinomycin were added to the assay mixture 10 min before starting the reaction so as to block the generation of the proton electrochemical gradient. At intervals, the samples (0.2 ml) were filtered through membrane filters (pore size, 0.45 m; Toyo Roshi Co., Tokyo, Japan) and washed with the same buffer, and the radioactivity was measured with a liquid scintillation counter.
Western Blotting-Total cell lysates were prepared as follows: cells (0.1-0.2 mg of protein) were harvested and suspended in 1 ml of buffer (10 mM Tris-maleate, 2 mM MgSO 4 , 20 M FUT175 (serine protease inhibitor (23)) and 20 g of N-acetylmuramidase, pH 7.5) without washing. The cell suspension was incubated at 37°C for 10 min. During this period, more than 90% of the cells were lysed. After addition of 0.5 ml of 15% trichloroacetic acid, the mixture was heated at 70°C for 15 min and then was centrifuged at 12,000 ϫ g. The precipitate was washed twice with ethyl ether and was dissolved in the sample buffer for SDS-polyacrylamide gel electrophoresis. For the quantitative determination of the amount of V 1 in the lysates, the sample (3 g of protein) was electrophoresed, immunoblotted with antiserum against purified V 1 of Na ϩ -ATPase as described elsewhere (13), and reacted with 125 Ilabeled protein A (300 MBq/mg). The radioactivity was measured by BioImaging analyzer (BAS-2000II; Fuji Film Co., Tokyo, Japan).
Primer Extension-Determination of the transcription start site of the ntp operon was achieved by the primer extension as described (24). Total RNA was extracted as described (25) from E. hirae cells grown in various media. The primer oligonucleotide (5ЈCTTGTCAAAATTCG-CGC), which is located on the noncoding strand just downstream of the putative initiation codon (ATG) for the ntpF gene, was end-labeled with [␥-32 P]ATP using polynucleotide kinase. Total RNA extract (20 g) and the labeled primer (4 ϫ 10 4 cpm) were mixed and incubated at 60°C and at room temperature as described (24). The primer extension reaction was started in the reaction mixture (24) with reverse transcriptase. The product was precipitated with ethanol and analyzed on a sequencing gel with a size marker.
Northern Blotting-Total RNA was extracted as described elsewhere (25) from E. hirae cells grown in various media, fractionated through formaldehyde gels, and transferred to nitrocellulose membranes. A DNA fragment labeled by the random priming method (25) was used as a probe.
Miscellaneous Methods-The catalytic V 1 moiety of Na ϩ -ATPase was purified as described previously (13). The cell membranes were prepared by the standard procedure as described previously (26), and, if necessary, stored frozen at Ϫ80°C. The ATPase activity was assayed in the presence of 0.2 mM DCCD by a procedure described elsewhere (21); the Na ϩ -stimulated ATPase activity of the membranes was determined at pH 8.5 in the presence or absence of 25 mM NaCl. Denatured polyacrylamide gel electrophoresis was carried out using the system of Laemmli with 10% polyacrylamide (27) and stained with Coomassie Brilliant Blue R-250. Protein was determined by the method of Lowry et al. (28) with bovine serum albumin as standard.
First, the immunoblotting using antiserum raised against the purified catalytic moiety was performed with the cells grown on KNY medium (limited Na ϩ ) or KNY medium containing 0.3 M NaCl (high Na ϩ ) (Fig. 1). The Na ϩ -stimulated ATP hydrolytic activities of the membranes prepared from these cells were 0.01 (limited Na ϩ ) and 0.25 (high Na ϩ ) mol/min/mg of protein, respectively. In parallel with an increase in the Na ϩ -ATPase activity in high Na ϩ conditions, the amounts of the two major subunits (A and B) of Na ϩ -ATPase increased in the membranes (Fig. 1, lanes 3 and 4). Although the purified enzyme contained the 29-kDa (D) subunit (13), this subunit was not detected clearly in this immunoassay (Fig. 1, lane 2); the antiserum probably contained only a small amount of antibodies against this subunit. The amounts of the A and B subunits in total cell protein also increased in high Na ϩ conditions (Fig. 1, lanes 5 and 6), disproving that an increase in FIG. 1. Western blots with anti-V 1 serum. Cells were grown in KNY medium (limited Na ϩ ) or KNY medium containing 0.3 M NaCl (high Na ϩ ) and harvested at late log phase, and the cell lysates were prepared as described under "Materials and Methods." The cell membranes were prepared as described previously (26). Electrophoresis in a denatured system and Western blotting of these samples using antiserum against purified V 1 moiety of Na ϩ -ATPase was performed, and V 1 was detected with the radioactivity of 125 I-labeled protein A by a BioImaging analyzer (BAS2000) (lanes 2-6). The amounts of proteins applied to gels were 3 g for the membranes and 15 g for the cell lysates. Lane 1, purified V 1 stained with Coomassie Brilliant Blue; lane 2, purified V 1 (0.3 g); lane 3, membranes (limited Na ϩ ); lane 4, membranes (high Na ϩ ); lane 5, total cell lysate (limited Na ϩ ); lane 6, total cell lysate (high Na ϩ ).

TABLE I
Changes in the amount of Na ϩ -ATPase Cells were grown in Na ϩ -limited medium (KNY), and at A 600 ϭ 0.1, NaCl, KCl, or sorbitol was added to the medium at the concentration of 1 M. After two generations (A 600 ϭ 0.4), cells were harvested, and the membranes and total cell lysates were prepared. The Na ϩ -stimulated ATPase activity of the membranes and the amount of A plus B subunits of Na ϩ -ATPase in total cell protein were measured as described under "Materials and Methods." Na ϩ -ATPase in membranes is caused by an elevated rate of enzyme incorporation into the membranes. We also observed a small amount of V 1 moiety in the soluble fraction; we do not know whether these V 1Ј moieties are the fraction released from the V 0 moiety in the membrane during cell disruption or into the V 1 pool before incorporation into the membrane (29). By measuring the radioactivity from 125 I-labeled protein A bound to the A and B subunits, the amount of Na ϩ -ATPase was compared in the cells grown in various media (Table I). Cells were grown on KNY medium containing NaCl, KCl, or sorbitol at a concentration of 1 M. The Na ϩ -stimulated ATPase activity of membranes prepared from cells grown in medium containing NaCl was 24-fold higher than the value of those from cells in KNY medium. Likewise, the amount of the A plus B subunits in total cell protein was increased 17-fold by addition of NaCl. On the other hand, both Na ϩ -ATPase activity and amounts of the A plus B subunits were little affected by addition of KCl or sorbitol (Table I). Fig. 2 shows active 22 Na ϩ extrusion from the cells grown in these media. In these experiments, DCCD, TCS, and valinomycin were all added in the assay mixture so as to block the generation of the proton gradient; sodium extrusion via the Na ϩ /H ϩ antiporter was neglected. The proton gradientindependent 22 Na ϩ extrusion was observed in the cells grown at high Na ϩ condition, but negligible in the cells in high K ϩ or sorbitol media (Fig. 2). The effect of LiCl or CsCl on the amounts of the A and B subunits was also examined by immunoblotting. Experiments were performed at the concentration of 150 mM because these salts are toxic for this organism. Whereas the amounts of A and B subunits were increased 5.5-fold by the addition of 150 mM NaCl to the medium, the amounts increased 2.6-and 1.1-fold when cells were grown in high LiCl and high CsCl, respectively. Furthermore, the effect of other osmolytes, such as glutamate, proline, or betaine, at a concentration of 0.2 M, was negligible on the amounts of these subunits. These results indicate that an increase in the amount of Na ϩ -ATPase is the specific response to high concentration of sodium ions or lithium ions, not to the ionic strength or osmotic change.
Sodium Stimulation of de Novo Synthesis of Na ϩ -ATPase- Fig. 3 shows the effect of chloramphenicol on the increase in the amount of Na ϩ -ATPase. To follow the quantitative change during cell growth, the cell culture was started in KNY medium, and then 0.8 M NaCl was added. Cell growth was slowed down by addition of 0.8 M NaCl, but recovered after about a 1-h lag period (Fig. 3A, closed circles). Although the growth was slightly slower than that in medium without NaCl, the growth yields in both media were similar (data not shown). The amounts of the A plus B subunits of the Na ϩ -ATPase were limited but constant in the cells growing in this Na ϩ -depleted medium (Fig. 3B, open circles). Addition of NaCl remarkably increased the amounts of these subunits preceding the growth recovery (Figs. 3, A and B, closed circles), and at 1 h after the addition of NaCl, the amounts of these subunits increased by 7-fold of the value before addition of NaCl and 15-fold at 2 h. When chloramphenicol (100 g/ml) was added with NaCl, an increase in the amounts of these subunits was not observed (Fig. 3B, closed triangles). Furthermore, the proton gradientindependent Na ϩ -pumping activity was not observed in the cells grown on KNY medium (Fig. 4A, closed circles); in the absence of the ionophores, however, sodium ions were actively extruded via the Na ϩ /H ϩ antiporter (Fig. 4A, closed triangles) (6). By addition of 0.5 M NaCl into KTY medium, the proton gradient-independent Na ϩ extrusion was induced (Fig. 4B). However, the addition of 100 g/ml chloramphenicol with NaCl abolished the induction of proton gradient-independent Na ϩpumping activity (Fig. 4C). An increase in the Na ϩ -stimulated

FIG. 3. Time course of a change of the amount of the A and B
subunits of Na ؉ -ATPase during cell growth. Cell culture was started in KNY medium, and 0.8 M NaCl and chloramphenicol (100 g/ml) were added (arrow). At intervals, the cell growth (A) was monitored, and the amounts of the A plus B subunits (B) in total cell proteins were immunochemically measured with 125 I-labeled protein A as described under "Materials and Methods." Open circles, without additions of NaCl and chloramphenicol; closed circles, with addition of NaCl and without addition of chloramphenicol; closed triangles, with additions of NaCl and chloramphenicol.

FIG. 2. Sodium-pumping activity by E. hirae cells grown in various media.
Cells were grown in KNY medium, and at A 600 ϭ 0.1, NaCl (A), KCl (B), or sorbitol (C), at a concentration of 1 M, was added to the medium. After two generations (A 600 ϭ 0.4), cells were harvested and 22 Na ϩ extrusion was assayed as described under "Materials and Methods." DCCD (0.2 mM), TCS(5 M), and valinomycin (5 M) were added to the assay mixture at Ϫ5 min, and the reaction was initiated by 10 mM glucose (indicated by arrows) at 5 min. Open circles, without glucose; closed circles, with glucose.
ATPase activity of the membranes at high Na ϩ condition was also blocked by chloramphenicol (data not shown). All these results suggest that protein synthesis is required for the increase in the amount of Na ϩ -ATPase. When rifamycin SV, instead of chloramphenicol, was added with NaCl, the increase in Na ϩ -ATPase activity in membranes was prevented and the basal amount of Na ϩ -ATPase subunits was slightly decreased (data not shown). Although we cannot exclude a possibility that a change in the rate of enzyme degradation increases the Na ϩ -ATPase level at such a high Na ϩ condition, the results obtained suggest that a stimulation of de novo synthesis of the Na ϩ -ATPase by sodium ions elevates the enzyme level.
In contrast to rifamycin SV, chloramphenicol did not decrease the enzyme basal level, although it inhibited the increase (Fig. 3B, closed triangles). It appears that protein synthesis was not inhibited completely by chloramphenicol under our conditions. In fact, E. hirae can grow in the presence of chloramphenicol under such conditions, although the growth rate is very low. Fig. 5A shows the arrangement of the ntp genes in the Na ϩ -ATPase operon (12). Northern blot experiments with several DNA probes corresponding to these ntp genes have revealed that the ntp operon is transcribed as a single mRNA of more than 10 kilobase pairs (19). Primer extension analysis was performed so as to determine the transcription initiation site(s) and to estimate the amount of mRNA for this operon. Total RNA was prepared from cells grown in several media; low Na ϩ medium (KTY) (pH 7.5), high Na ϩ medium (NaTY medium containing 0.5 M NaCl; pH 7.5), and high Na ϩ and high pH medium (NaTY medium containing 0.5 M NaCl; pH 10.0). The Na ϩ -ATPase activities of the membranes of these cells were 0.02, 0.25, and 0.51 mol/min/mg of protein, respectively. When the primer, which corresponded to the complementary sequence of the ntpF gene (Fig. 5C, underlined) was used, we found three spots as the initiation sites of the ntp operon just upstream of the ntpF gene (Fig. 5B). The primer extension was attempted using another primer corresponding to the 3Ј part of the ntpF gene, FIG. 4. Effect of chloramphenicol on induction of the sodium-pumping activity by E. hirae cells. Cells were grown in Na ϩ -limited medium (KNY) (A), and at A 600 ϭ 0.1, 0.5 M NaCl was added to the medium (B). After two generations (A 600 ϭ 0.4), cells were harvested, and 22 Na ϩ extrusion was assayed as described under "Materials and Methods." When chloramphenicol (100 g/ml) was added with 0.5 M NaCl (C), cells were harvested at 2 h after the addition. DCCD, TCS, and valinomycin were added at Ϫ5 min, and the reaction was initiated by addition of 10 mM glucose at 5 min. Open circles, without glucose; closed circles, with glucose; closed triangles, with glucose in the absence of DCCD, TCS, and valinomycin.

FIG. 5. Primer extension analysis.
A, structure of the ntp operon. The ntp operon is composed of 11 genes: ntp-FIKECGABDHJ. The arrow indicates the direction of the operon. The DNA segment designated by the shaded box represents the probe used for Northern blotting as shown in Fig. 7. B, primer extension. Total RNA was extracted from the cells grown in NaTY medium containing 0.5 M NaCl at pH 10.0 (lane 1), NaTY medium containing 0.5 M NaCl (lane 2) or KTY medium (lane 3). Primer extension was carried out using the primer (the coding sequence corresponding to the primer is underlined in C) labeled with [␥-32 P]ATP as described under "Materials and Methods." The tentative initiation site of transcription was represented by an asterisk. C, DNA sequence of the promotor region of the ntp operon. The sequence corresponding to the primer used for primer extension was underlined. The putative Shine-Dalgarno (SD) sequence (boldface) and Ϫ10 and Ϫ35 boxes (overlined) were represented. and the same initiation spots were observed. 2 At this time, it is hard to determine the unique transcription initiation site among these three spots (G, T, or C), but we tentatively assigned G as the site, located 40 bp upstream of the ntpF gene.
The amount of primer extension products changed drastically in parallel with the activities of the Na ϩ -ATPase in the membranes under various growth conditions. Amount of the products was limited in the cells grown on the Na ϩ -limited medium (Fig. 5B, lane 3). However, the amount of the spots increased at high Na ϩ condition (Fig. 5B, lane 2), and it further increased in high Na ϩ and high pH medium (Fig. 5B, lane 1). In a previous paper (19), we performed the Northern blotting analysis with the probes corresponding to several parts from ntpA to ntpJ genes and total RNA extract prepared from the cells cultured in these media. The increases in the mRNA for the ntp operon were in parallel with those observed here by the primer extension. Western blotting with anti-V 1 serum also revealed that the increase in the amount of the A subunit of the Na ϩ -ATPase in cell lysates was in parallel with the increases of transcripts in these conditions (19). These results suggest that de novo synthesis of Na ϩ -ATPase is regulated at the transcriptional level by sodium ions.
Intracellular Na ϩ Regulates Transcription of the Na ϩ -ATPase Operon-We have speculated that an increase in the internal Na ϩ concentration ([Na ϩ ] in ) triggers stimulation of Na ϩ -ATPase biosynthesis (18). Fig. 6 shows the effect of monensin, a Na ϩ /H ϩ exchanging ionophore, on the amounts of the A plus B subunits of the Na ϩ -ATPase at the various concentrations of the external Na ϩ ([Na ϩ ] out ). In the absence of monensin, the amounts of these subunits increased in proportion with the increase in [Na ϩ ] out , and, at 1 M NaCl, the amounts of these subunits increased up to about 15-fold (Fig. 6, open circles). When the membranes were rendered permeable to sodium ions by monensin, the amounts of the A plus B subunits increased by 6-fold of that of the untreated cells even in Na ϩlimited KNY medium; the contaminated Na ϩ was less than 1 mM. The amounts of these subunits in the presence of monensin were also affected by [Na ϩ ] out , but they remarkably increased at the lower concentrations of [Na ϩ ] out . At 100 mM NaCl, the amounts of these subunits increased by 22-fold of those of the cells in KNY medium, and the increase in the amounts of Na ϩ -ATPase was finally saturated (Fig. 6, closed circles). Gramicidin D, which renders the cell membrane unselectively permeable to monovalent cations such as H ϩ , Na ϩ , or K ϩ , brought about the same effect as monensin. Gramicidin D (5 g/ml) enhanced the amounts of the A plus B subunits only 5-fold in Na ϩ -depleted medium. However, in the presence of 50 mM Na ϩ , gramicidin D enhanced the amounts of A and B subunits by 17-fold (data not shown). In parallel with the amounts of these subunits, the Na ϩ -stimulated ATPase activity of the membranes of the cells grown in NaTY medium and in the same medium containing gramicidin D were 0.11 and 0.42 mols/min/mg of protein, respectively. Although the proton gradient of growing cells was dissipated by the ionophore, the increase in the amounts of Na ϩ -ATPase was limited without addition of NaCl, suggesting that the effect of the size of the proton gradient was negligible for the induction of Na ϩ -ATPase. Thus, the amount of Na ϩ -ATPase is probably regulated by [Na ϩ ] in . The effect of monensin on the amounts of mRNA for the ntp operon was also examined by Northern blotting (Fig. 7). In this experiment, the probe, a 2.5-kilobase pair HindIII-HindIII DNA fragment that corresponds to the ntpA to ntpD genes in the ntp operon (Fig. 7, shaded box) was used, and the total RNA fraction was prepared from the cells grown in four different media: KNY medium, KNY medium supplemented with 5 g/ml monensin, KNY medium supplemented with 100 mM NaCl, and KNY medium supplemented with both 100 mM NaCl and monensin. The amount of mRNA for the ntp operon was negligible in KNY medium (Fig. 5B, lane  1), and it was slightly increased by supplementation with NaCl (lane 2) or monensin (lane 3). Then, the mRNA of this operon was increased remarkably by supplementation with both NaCl and monensin. The relative amount of mRNA in these fractions corresponded to the values of the amounts of the A and B subunits in the cells grown in the same media as shown in Fig.  6. All of these results suggest that the Na ϩ -ATPase level of E. hirae is regulated by [Na ϩ ] in . Transcription of the ntp operon is stimulated by an increase in [Na ϩ ] in as the signal. DISCUSSION Although the ion specificity of E. hirae Na ϩ -ATPase has been reported in a previous paper (8), the data were based on the characteristics of the enzyme activity with Na ϩ -stimulated ATP hydrolytic activity that was too low. Ion specificity of the ATP hydrolytic activity by this enzyme was reexamined. The ATP hydrolytic activity was stimulated by not only NaCl but LiCl; the apparent K m values for Na ϩ and Li ϩ were about 3 and 2 mM, respectively. KCl, CsCl, and ammonium chloride did not stimulate activity. As the sodium-pumping activity via the Na ϩ -ATPase in whole cells was also induced by supplementa-2 T. Murata, I. Yamato, K. Igarashi, and Y. Kakinuma, unpublished results.
FIG. 6. Effect of [Na ؉ ] out on the amounts of the A and B subunits of Na ؉ -ATPase. Cell culture was started in KNY medium, and at A 600 ϭ 0.1, various concentrations of NaCl and monensin (5 g/ml) were added. After two generations (at A 600 ϭ 0.4), cells were harvested and the amounts of the A plus B subunits in total cell proteins were immunochemically measured as described under "Materials and Methods." Open circles, plus NaCl; closed circles, plus NaCl and monensin.
FIG. 7. Northern blotting. Cells were grown on KNY medium, and at A 600 ϭ 0.1, 100 mM NaCl and/or monensin (5 g/ml) was added to the medium. After two generations (A 600 ϭ 0.4), cells were harvested, and the total RNA was extracted from these cells as described under "Materials and Methods." Northern blotting was carried out using a DNA fragment (Fig. 5A, shaded box)  tion with 150 mM LiCl into Na ϩ -limited medium, Li ϩ is also the signal for induction of the Na ϩ -ATPase (data not shown). These results suggest that an increase in the transcription of the E. hirae Na ϩ -ATPase operon specifically responds to the coupling ions sodium and lithium. It is another interesting question whether the deficiency in [K ϩ ] in at high pH participates in the regulation of Na ϩ -ATPase operon, since the ntpJ gene, the tail-end cistron of the ntp operon, encodes a component of KtrII K ϩ uptake system working at high pH (19). We have not obtained concrete data for or against this possibility because of the difficulty of depletion of both Na ϩ and K ϩ from the complex medium; this bacterium does not grow in minimal medium. Development of a synthetic medium for the growth of E. hirae at high pH is the urgent requirement to answer this question.
We also observed a decrease in the Na ϩ -ATPase level when E. hirae cells grown in media containing 0.5 M NaCl were transferred to a Na ϩ -limited medium. The amount of the A and B subunits of Na ϩ -ATPase decreased by one-half of the original level within 1 h after the medium change. In this period, the cell mass increased by about 2-fold. Therefore, assuming that E. hirae ceased to synthesize the enzyme in Na ϩ -limited medium, the decrease in Na ϩ -ATPase level may be explained by a dilution of the enzyme during cell division, although further investigation is required.
It is noteworthy that the Escherichia coli nhaA gene encoding the Na ϩ /H ϩ antiporter is highly expressed in high concentrations of Na ϩ or Li ϩ and in alkaline medium (30), suggesting that the nhaA gene may be regulated by the internal Na ϩ (5, 31). The nhaR gene, downstream of the nhaA gene, has been reported as a positive regulator for expression of the nhaA gene (32); it would be interesting to investigate whether NhaR is a Na ϩ -sensor. There are several palindromic sequences in the promotor region of the ntp operon (12), which may be the regulatory element(s) for the expression of this operon. It is hard to expect that sodium ions directly interact with the promoter region of the ntp operon. Transcription of the Na ϩ -ATPase operon is probably regulated by a [Na ϩ ] in -sensor (protein)-mediated system. Just upstream of the ntp operon, there is the ntpR gene, which encodes a 27-kDa hydrophilic protein having a putative helix-turn-helix motif. At first, we expected this gene to be the regulatory gene for the ntp operon and knocked it out by gene disruption. However, the ntpR-disrupted strain did not show any difference in induction of the Na ϩ -ATPase. 3 Our attempt to detect a sodium-specific response in a gel retardation assay with cell lysates and the plasmid harboring the DNA region upstream of the operon has been so far unsuccessful. We will isolate regulatory mutants of the operon induction as another approach for identifying the regulatory protein.
Recently it was pointed out that transcription of the napA gene, the Na ϩ /H ϩ antiporter gene of E. hirae, may be stimu-lated by an increase in [Na ϩ ] out or [Li ϩ ] out (33). The Na ϩ -ATPase-defective mutant Nak1 (34) did grow in NaTY medium containing more than 0.5 M NaCl at low pH (data not shown). These results suggest the possibility that induction of Na ϩ -ATPase may be compensated for by an elevation in the antiporter activity under this condition. Thus, the internal Na ϩ concentration of E. hirae is probably regulated by an interplay between these dual extrusion systems at the transcriptional level.