Characterization of the Respiration-dependent Na ' Pump in the Marine Bacterium Vibrio alginolyticus *

The respiration-dependent Na' pump (Tokuda, H. and Unemoto, T. (1981) Biochem Bwphys. Res. Commun. 102,265-271) is examined in detail under various conditions using the cation-loaded Vibrio alginolyticus. The Na' pump can extrude Na+ against its electrochemical gradient and generate a membrane potential (inside negative) in the presence of a proton conductor, carbonylcyanide m-chlorophenylhydrazone (CCCP). As a result, a passive uptake of H+ occurs that leads to the generation of ApH (acidic inside) of similar magnitude to that of the membrane potential. Anoxia or a respiratory inhibitor, 2-heptyl-4-hydroxyquinoline-N-oxide, inhibits the Na' extrusion, membrane potential generation, and H+ uptake in the presence of CCCP while these activities resume immediately when oxygen or an artificial electron donor, N,N,N',N'-tetramethyl-pphenylenediamine, is supplied. The Na+ pump is independent of ATP since arsenate drasticallif decreases the level of intracellular ATP but has no effects on the Na+ extrusion and membrane potential generation. The Na+ pump has a pH optimum at about 8.5 to 9.0 and the generation of membrane potential, extrusion of Na', and uptake of H+ in the presence of CCCP are not observed at acidic pH. At alkaline pH, Na+ markedly stimulates the generation of membrane potential and rates of oxygen consumption by K'-loaded cells. Such results strongly indicate that the Na' pump is indispensable to energetics of V. alginolyticus under alkaline conditions. We conclude that K alginolyticus is able to pump out not only H+ but also Na' as an immediate result of electron transport at alkaline pH. We also discuss the possible roles of the primary Na' pump under Na+-rich environments.

The respiration-dependent Na' pump (Tokuda, H. and Unemoto, T. (1981) Biochem Bwphys. Res. Commun. 102,265-271) is examined in detail under various conditions using the cation-loaded Vibrio alginolyticus. The Na' pump can extrude Na+ against its electrochemical gradient and generate a membrane potential (inside negative) in the presence of a proton conductor, carbonylcyanide m-chlorophenylhydrazone (CCCP). As a result, a passive uptake of H+ occurs that leads to the generation of ApH (acidic inside) of similar magnitude to that of the membrane potential. Anoxia or a respiratory inhibitor, 2-heptyl-4-hydroxyquinoline-N-oxide, inhibits the Na' extrusion, membrane potential generation, and H+ uptake in the presence of CCCP while these activities resume immediately when oxygen or an artificial electron donor, N,N,N',N'-tetramethyl-pphenylenediamine, is supplied. The Na+ pump is independent of ATP since arsenate drasticallif decreases the level of intracellular ATP but has no effects on the Na+ extrusion and membrane potential generation. The Na+ pump has a pH optimum at about 8.5 to 9.0 and the generation of membrane potential, extrusion of Na', and uptake of H+ in the presence of CCCP are not observed a t acidic pH. At alkaline pH, Na+ markedly stimulates the generation of membrane potential and rates of oxygen consumption by K'-loaded cells. Such results strongly indicate that the Na' pump is indispensable to energetics of V. alginolyticus under alkaline conditions. We conclude that K alginolyticus is able to pump out not only H+ but also Na' as an immediate result of electron transport at alkaline pH. We also discuss the possible roles of the primary Na' pump under Na+-rich environments.
A number of papers have been presented concerning the generation of an electrochemical potential of H' (a proton motive force) across bacterial membranes and its roles in energy-dependent reactions (1)(2)(3)(4). From these reports, it is no doubt that the proton motive force, as predicted by the chemiosmotic hypothesis of Mitchell (5,6), plays a central role in energetics of various bacteria. Many active transport systems have been shown to be driven by the proton motive force (H' symport system). Although a certain number of transport systems in nonhalophilic bacteria (see references cited in Ref. 16) exceptionally utilize the electrochemical potential of Na+ as an immediate driving force (Na' co-transport system), the generation of the Na' electrochemical po-* 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.
$ To whom all correspondence should be addressed. tential via a Na+/H+ antiport system is yet secondary to the generation of the proton motive force in these systems (7,8). However, since most of the reported results have been obtained with nonhalophilic bacteria, such an exclusive H' economy may not be applicable to halophilic bacteria. Indeed, transports of all amino acids in the extreme halophile Halobacterium halobium are driven by the Na+ electrochemical potential (9) which can be generated not only by the Na+/H+ antiport system (10) but also by a light-dependent primary Na' pump, halorhodopsin (11,12). In bacteria, halorhodopsin is the fiist example which is able to generate membrane potential by flux of ion(s) other than protons.
Marine bacteria require Na' for their growth. Fein and MacLeod (13), Drapeau et al. (14), and Niven and MacLeod (15) reported that uptake of all the solutes examined in Alteromonas haloplanktis was Na+-dependent. We have also found that the accumulations of AIB' (16), 19 amino acids,2 and sucrose3 by V. alginolyticus, another marine bacterium, were driven by the Na' electrochemical potential. Furthermore, in our recent report (17), it was shown that V. alginolyticus could generate A# and take up AIB in the presence of CCCP. The examination of Na' extrusion revealed that such a A$ may be generated by the primary electrogenic Na' extrusion system. The system appeared to be dependent on respiration and had a pH optimum at alkaline region. Since the presence of such a Na+ pump is unique and may provoke the question about the fundamental concepts of energetics in halophilic bacteria, it is necessary to characterize the Na' pump in detail. In this paper, the activities of Na' pump were examined under various conditions and it was concluded that Na+ is pumped out by V. alginolyticus as a direct result of electron transport.

EXPERIMENTAL PROCEDURES
Growth of Cells-The marine bacterium V. alginolyticus 138-2 was grown aerobically at 37 "C on a synthetic medium (18) containing 0.3 M NaCl and 1% glycerol as a sole source of carbon. The cells were harvested at the late logarithmic phase of growth by centrifugation at 4 "C.
Preparation of cells loaded with various monovalent cations was performed as described (16) using DEA-HC1, pH 8.5, containing 0.4 M desired cation as a chloride salt. After the second treatment with DEA buffer, the cells were washed with and resuspended in a specified buffer containing 0.4 M salt and kept on ice until use.
The intracellular concentration of loaded cation was about 0.4 M and that of Na' in K' -or Li'-loaded cells was negligible (16).
Determination of AI) andhpH-Generations of AI) (negative inside) and ApH (acidic inside) were examined at room temperature from the equilibrium distribution of ["HITPP' and ['4C]methylamine, respectively. The distribution of these probes was determined by flow dialysis (3) and filtration.
Flow dialysis was performed as described (17,18) with a sample flow rate of 1 ml/min. Radioactivity was continuously monitored by a Flow-One radioactivity monitor (Radiomatic Instruments and Chemical Co., Tampa, FL) using liquid scintillator (flow rate, 4 ml/ min) for "H or solid scintillator for I4C. Counts accumulated for 1 min were printed out and plotted after the correction for background. Oxygenated buffer was pumped from the lower chamber of the flow dialysis cell and the upper chamber was kept under a stream of oxygen. Where cited, nitrogen gas was employed instead of oxygen.
In some experiments, AI) and ApH were determined by fdtration. Cells were preincubated at 25 "C for 5 min in 50 p1 of 50 llul specified buffer containing 0.4 M NaCl and 20 mM glycerol. The assay was started by the addition of radioactive probe. At time intervals, the uptake was terminated by addition of 2 ml of 0.4 M NaCl at room temperature and by filtration with cellulose acetate fdters (Schleicher and Schull, pore size 0.45 pm). The filters were washed once with 2 ml of 0.4 M NaCl and radioactivities were determined.
Steady state concentration gradients of 'I'PP' and methylamine were calculated by using a value of 3.3 p1 of internal water space/mg of cell protein (18). AI) was determined from the Nernst equation, AI) = 59 log[TPP+]i,/[TPP+],,,. Internal pH was calculated from the distribution of methylamine (pK = 10.62) as described (1). ApH (pHi, -pH,,t, in units), the chemical potential of H' (59ApH, in mV), and Ap (A$ -59ApH, in mV) were then calculateci. Flow dialysis and filtration gave essentially the same values of A+ and ApH.
Determination of ATP-Intracellular ATP was determined by the luciferin/luciferase method as described (19).
Energization with TMPD-Where specified, the energization was performed with TMPD under the stream of oxygen. In order to determine the optimum concentration of TMPD, the generation of AI) and uptake of AIB by the cells treated with HQNO were assayed in the presence of various concentrations of TMPD. The best results were obtained with 0.14 m~ TMPD in the absence of ascorbate. Although ascorbate added with TMPD prevented the development of blue color due to the oxidized form of TMPD, the generation of AI) and uptake of AIB were not stimulated by ascorbate over the concentration of 1 to 20 m~. TMPD was therefore added alone. The generation of A+ and uptake of AIB were stable for at least 20 min under such conditions.
Respiration-dependent H' Flow across Membranes-Na'-or K'loaded cells were resuspended in 2 ml of weakly buffered 0.4 M salt solution containing 20 m~ glycerol. The cell suspensions were incubated at 25 "C in a water-jacketed vessel (Radiometer, TTA6O titration assembly) under the continuous flow of nitrogen. Medium pH was monitored by a glass electrode (G204OC) attached to a pH meter (Radiometer, PHM 84) with a calomel electrode (K701) as a reference. Since the calcmel electrode contained a secondary salt bridge which was filled with a saturated CH,COOLi, K' leakage from the electrode was negligible. Such a precaution was necessary since K' induces the extrusion cf Na' (16,17) and H+ (18). Medium pH was readjusted with 0.1 N NaOH or HC1 and the incubation was continued until medium pH became near constant. Oxygen pulse was performed by addition of 100 pl of air-saturated 0.4 M salt solution and pH change was recorded by a Hitachi 056 recorder.
Oxygen Consumptton-Rate of oxygen consumption was determined at 25 "C with an oxygen electrode (Yellow Springs Instrument Co., Yellow Springs, OH) attached to a Hitachi 056 recorder.
Other Determinations-Protein was determined as described by Lowry et ul. (20) by using bovine serum albumin as a standard. Intracellular concentrations of cations (Na', K' , and Li') were analyzed as previously reported (16).

RESULTS
Effects of Membrane-permeable Amines and CCCP on the Generation of A+-The first indicaticn of the presence of a primary ion pump that extrudes ion(s) other than H' was obtained when the generations of A+ (negative inside) and ApH (acidic inside) by Na+-loaded V. alginolyticus wsre monitored from the distribution of [?H]TPP+ and ['4C]methylamine, respectively, by flow dialysis (Fig. 1). Although CCCP almost completely collapsed A# generated in 50 m~ DEA-HC1, pH 8. Such effects of amines were expected since these amines were thought to permeate membranes in their unprotonated form and alkalinize the cytoplasm by their protonation. These results forced us to draw the unexpected conclusion that the A# observed at pH 8.5 in the presence of CCCP is generated by a pump which is extruding io&) other than H' . When membranes become permeable to H' by CCCP, such a A# causes the passive accumulation of H' until the magnitude of ApH reaches the magnitude of A+ and then the net influx of H' ceases. On the other hand, H' influx mediated by CCCP is continuous in the presence of membrane-permeable amines since these amines collapse ApH. This is the reason why the combined addition of CCCP and amines is necessary for the depolarization of A+. These results also indicated that the ion which is responsible for the generation of CCCP-insensitive A+ should be the one present in much higher concentration than H' . Cation-specificity in the Generation of CCCP-insensitive A+-V. alginolyticus loaded with Li' or K+ was prepared and assayed for the generation of A+ in 50 mM Tricine buffer (pH 8. indicated that Na' is essential for the generation of a maximum magnitude of A+ whether CCCP is present or not and that the Na' pump is functioning to generate A+ at pH 8.5. As discussed in a later section (Fig. 6) and in a previous paper (17), the examination of Na' extrusion at pH 8.5 clearly demonstrated the presence of the primary Na' pump. The fact that Li'-loaded cells generated a considerable magnitude of A+ even in the presence of CCCP may indicate that Li' can partly substitute for Na' in the generation of A+.
Effect of External pH on the Generation of CCCP-insensitive A+ a n d CCCP-dependent ApH-The generation of A+ by Na'-loaded cells was examined at pH 6.0 and 8.5 in the presence of various concentrations of CCCP (Fig. 3). Although A+ at pH 8.5 was hardly changed by up to near 50 p~ CCCP, A+ at pH 6.0 was almost completely collapsed by 2 p~ CCCP. When the respiration-dependent H' extrusion was examined at pH 8.5, CCCP added at 2 p~ completely abolished the H' extrusion. Hence, the sensitivity of membranes to CCCP was not altered by external pH. A+ and ApH in the presence of 10 p~ CCCP were determined over the pH range of 6.0 to 9.0 (Fig. 4). As shown by closed circles, the generation of CCCPinsensitive A+ was pH-dependent and had a pH optimum at (open circles). The filtration method was employed to detect a low level of accumulation which was under a detection limit of flow dialysis. Although a ApH (alkaline inside) in membrane vesicles of E. coli ML 308-225 could not be detected by the filtration method using dimethyloxazolidine-2,4-dione as a pH indicator (21), the filtration method with methylamine as a pH probe was able to monitor ApH (acidic inside) in whole cells of V. alginolyticus since magnitude of ApH determined at pH 8.5 and 9.0 by filtration and flow dialysis gave essentially the same results (for example, compare the results shown by open circles in Fig. 1B with the values at pH 8.5 presented in Fig. 4). As shown, the generation of CCCP-dependent ApH exhibited a similar pHdependence to that of the generation of CCCP-insensitive A+. As a result of ApH generation, Ap in the presence of CCCP was very small (less than -30 mV) over the pH range tested (triangles).
Requirement for Respiration in the Generation of CCCPinsensitive A$ and CCCP-dependent ApH-The results shown in Fig. 5A represent the flow dialysis experiments performed under the stream of nitrogen instead of oxygen. AJ, (closed circles) observed after addition of cell suspensions (kept under air) was unstable (about -159 mV at maximum) and decreased gradually if no additions were made. Furthermore, CCCP added to the anaerobic cell suspensions instantaneously collapsed this A$. After the complete collapse of A#, switching of the gas to oxygen immediately caused the generation of A$ (-154 mV) in the presence of CCCP. On the other hand, ApH, shown by open circles, was small, if any, before the switching of gas and was not affected by CCCP. When A$ was generated by aerobiosis, the concomitant generation of ApH (-123 mV) was also observed.
As shown by closed circles in Fig. 5B, A# detected after additions of cell suspensions and HQNO was rather stable and calculated to be about -160 mV. This may be due to the fact that about 10% of oxygen consumption is insensitive to HQNO and, moreover, may indicate that HQNO does not have an ionophoric activity. In any event, A# generated under such conditions was collapsed by CCCP. The inhibition of oxygen consumption in V. alginolyticus by HQNO occurs at the site before cytochrome c and can be overcome by the addition of artificial electron donor TMPD, which probably donates electrons to cytochrome c (22). As expected from such data, the addition of TMPD led to the generation of A# of about -145 mV in the presence of HQNO and CCCP, whereas a large ApH (-130 mV) was not generated until the reduced TMPD was added to the cells treated with HQNO and CCCP (open circles). Although results were omitted from the figures, if CCCP was not added, oxygen or TMPD did not lead to the generation of ApH.
From these results and the result that cyanide inhibits the generation of CCCP-dependent ApH (17), it is clear that the generation of CCCP-insensitive A$ with a concomitant generation of ApH requires continuous respiration.
The Intracellular Level of ATP and the Generation of A+Although it seemed unlikely from the results described above that ATP was required for the generation of CCCPinsensitive A+, such a possibility was examined by determining ATP level and comparing it with A# under various conditions (Table I). CCCP added alone had effects on neither the ATP level nor A# determined in the presence of 20 mu glycerol. It may be noteworthy that in the absence of glycerol, the ATP level was decreased by CCCP to about half of control level. Although arsenate did not collapse the CCCP-insensitive A$ at all, ATP level was significantly reduced and only about 20% of the level was found after the addition of arsenate. Furthermore, the level of ATP under the presence of CCCP and HQNO was not affected by TMPD and remained in a high level, on the contrary, TMPD was necessary to the generation of A# under such conditions. These results demonstrate that the intracellular ATP is not essential for the generation of A# by the Na+ pump.
Energetics of Naf Extrusion at Alkaline pH-If the Naf pump is truly responsible for the generation of CCCP-insensitive A$, the extrusion of Na' must be shown under the conditions where the CCCP-insensitive A# was generated. In our previous paper (17), it was shown that Na' extrusion at pH 8.5 in the presence of glycerol required K' as a counter ion and was not inhibited by CCCP, while the strong inhibition of Na' extrusion by CCCP was observed at pH 6.5. These results let us propose that the extrusion of Na' at pH 8.5 was a primary process and responsible for the generation of CCCPinsensitive A$. Energetics of Na' extrusion was further examined to confirm and extend the proposal (Fig. 6).
When respiration was inhibited by HQNO, CCCP-insensitive Na' extrusion did not take place even in the presence of K' (Fig. 6A, open circles). While TMPD added to the cells treated with HQNO failed to induce the CCCP-insensitive extrusion of Na' at pH 6.5 (Fig. 6A, triangles), the extrusion of Na+ at pH 8.5 in the presence of CCCP immediately took place upon the addition of TMPD (Fig. 6A, closed circles).
AS expected from the result that membrane-permeable amines in combination with CCCP collapsed A$, the addition ion and induced a bulk extrusion of Na' (Fig. 6B, closed  circles). Compared to control (Fig. 6B, open circles), more than half of intracellular Na' was extruded for 5 min after the addition of DEA and CCCP. Although results were omitted from the figure, a single addition of DEA or CCCP caused less than 10% of Nat extrusion for 1 min and the levels of intracellular Na+ at 5 min were about 80% in both cases. Effects of arsenate on the CCCP-insensitive Na+ extrusion were examined in such systems. Arsenate affected neither the rate nor the extent of Na+ extrusion induced by the addition of DEA and CCCP (Fig. 6B, triangles). These results show complete agreement with the results of A$ generation in the  respectively. Control experiments (0) were performed in the absence of sodium arsenate, DEA, and CCCP. The level of ""a' retained by cells was determined at given times by the filtration method as described in Fig. 3. The values obtained were corrected for background radioactivity which was determined with samples boiled for 5 min prior to assay. The results are given in percentage of radioactivity at 0 time. The radioactivities of 6650, 3950, and 6900 cpm were obtained at 0 time with the cells assayed at pH 8.5 and 6.5 in A and the cells assayed in B, respectively. presence of CCCP and confirm that the extrusion of Na' is a primary process driven solely by electron transport. It should be pointed out that K'-dependent extrusion of Na+ was significantly inhibited by arsenate (results not shown). This presumably represents the requirement of ATP for K' transport system, but not for the Na' pump, which is absolutely necessary to the K'-dependent Na' extrusion (16). It is also notable that the extrusion of Na' at pH 6.5 in the presence of CCCP was not induced by the addition of TMPD (Fig. 6 A ) or DEA (not shown). The Na+/H' antiport system, driven by the proton motive force, seems to be the only way to extrude Na' at acidic pH.
Respiration-dependent Flux of Protons across the Membrane-Anaerobic suspensions of V. alginolyticus were pulsed with oxygen under various conditions and the external pH was monitored to examine the flux of H' (Fig. 7 ) . When Na'loaded cells were assayed at pH 6.5, the pulse caused a transient acidification of external medium (A) which indicated that H' was extruded on respiration. If the cells were treated with CCCP prior to the pulse, such an acidification was completely inhibited (B). When the assay was performed at pH 8.5 (C), a slight alkalinization occurred after the pulse and then it was followed by the acidification which was slower than that at pH 6.5. On the other hand, the assay performed at pH 8.5 with the CCCP-treated cells (D) gave the result which was strikingly different from that at pH 6.5. The oxygen pulse to such cells led to the marked alkalinization of external medium which indicated that H' was taken up on respiration. In order to characterize the electrogenic feature of H' flux at pH 8.5 in the presence and absence of CCCP, a membranepermeable cation TPP' was included in the assay medium. The acidification by the control cells in the presence of TPP' ( E ) was significantly faster than that in the absence of TPP+ (C). This indicates that the extrusion of H+ occurs against its electrical potential and that TPP+, as a counter ion, stimulates 0.. 1  Such a result clearly indicates that the uptake of H' by the CCCP-treated cells occurs down the electrical potential and that TPP' inhibits the uptake of H' by collapsing A$ generated by the respiration-dependent Na' pump. When K'loaded cells instead of Na+-loaded cells were assayed at pH 8.5, the alkalinization by CCCP-treated cells was not significant ( H ) although the acidification by control cells was observed ( G ) . On the contrary, the K'-loaded cells treated with CCCP did take up H' when the assay was performed in 0.4 M NaCl (results not shown).
Effects of Na' on the Rates of Oxygen Consumption by V. alginolyticus-Since K' stimulated the oxygen consumption as reported previously (18), effects of Na' on the rates of oxygen consumption by K'-loaded cells were determined in the presence of excess KC1 over the pH range of 6.0 to 9.0 (Fig. 8). The rates determined in the absence of Na' increased almost linearly between pH 6.0 and 7.5 and then, above this pH, decreased as the increase in external pH. On the other hand, the rates obtained in the presence of 0.2 M each of NaCl and KC1 (closed circles) showed continuous increase as the increase in pH. As a result, the marked stimulation by Na' was observed at alkaline pH. Although detailed results are not presented, these results show good agreement with the effects of pH and Na' on the generation of A+. The stimulation of A$ by Na' was not detected at acidic pH but became evident at external pH above 7.5 as shown by lower patterns in Fig. 2.

DISCUSSION
The results presented in this paper confirmed and extended our previous proposal that V. alginolyticus possesses a unique system to generate the proton conductor-insensitive A$. The generation of such a A$ was dependent on Na' (Fig. 2). The examination of Na' extrusion revealed that the process at pH 8.5 was a primary transport system and was dependent on respiration (Fig. 6). From the results shown in Figs. 5 and 7, where the generation of A$ and uptake of H+ in the presence of CCCP were demonstrated by the supply of oxygen to anaerobic cell suspensions, it is unequivocal that V. alginolyticus does generate A# without the net extrusion of H'. Therefore, the generation of such a A# cannot be ascribed to any fluxes of H' but should be so to the electrogenic extrusion of Na+. Both the generation of A# and extrusion of Na' in the presence of CCCP showed maximum activity at alkaline pH (Figs. 4 and 6) and were dependent on respiration (Figs. 5 and 6). A possible involvement of ATP in the phenomena discussed above was excluded since the fluctuation of ATP level was not related to A# generation (Table I) and arsenate did nothing to the extrusion of Na' in the presence of CCCP (Fig.  6).
It is our conclusion that V. alginolyticus is able to pump out Na+ as well as Hf by its respiratory chain (Fig. 9). At alkaline pH, the electrogenic extrusion of Na' in addition to H' occurs to generate A#. When the membrane becomes permeable to H' by the action of CCCP, A# due to H' collapses but that due to Na+ exists since the concentration of Na' is several orders of magnitude higher than that of H+. Therefore, the addition of CCCP causes the accumulation of H' which continues until the magnitude of ApH becomes similar to that of A#. The membrane-permeable amines added in high concentration scavenge H' accumulated inside the cells and make the continuous influx of H+ possible. Thus, CCCP with a help of such amines is able to collapse A# generated by the Na+ pump. Although questions still remain whether or not the extrusion of H' at acidic pH and the extrusion of H' plus Na' at alkaline pH are performed by the same respiratory chain and whether or not the participation of the additional components at alkaline pH is necessary for the extrusion of Na', it is likely that at least one "Na+ site" exists after cytochrome c of the respiratory chain since TMPD causes the extrusion of Na' (Fig. 6) with concomitant generation of A# (Fig. 5) in the presence of CCCP. Detailed molecular mechanisms of the respiration-dependent Na+ pump, including the possibility of other Na+ sites, are currently under examinations.
It is of interest that both H. halobium (11,12) and V. alginolyticus live in Na+-rich environments and have developed devices by which energy is directly converted to the Na' electrochemical potential across the memebranes. Furthermore, it has recently been demonstrated that Klebsiella aerogenes that requires Na' for the anaerobic growth on citrate also possesses an inducible Na' pump (23)(24)(25). However, the mechanisms involved are totally different among these bacteria. H. halobium conserves light energy as the Naf electrochemical potential by means of the light-dependent Na' pump halorhodopsin. K. aerogenes obtains the energy from the decarboxylation of oxaloacetate and V. alginolyticus utilizes redox energy for the generation of Na' electrochemical potential by the respiration-dependent Na' pump. Although, so far reported, the proton motive force plays a central role in energetics of nonhalophilic bacteria, the results presented here and those with H. halobium seem to indicate that it is Pump in V. alginolyticus 10013 the Na+ electrochemical potential, or the sodium motive force, that plays a central role in energetics of halophilic bacteria. Consistent with this idea, transports of solutes in these bacteria are all driven by the Na' electrochemical potential. A# generated by the Na' pump may be able to drive ATP synthesis since a red mutant of H. halobium lacking bacteriorhodopsin synthesizes ATP upon illumination (26, 27) although the ion which actually passes through the ATPase complex is likely to be H' . It should be pointed out that the proton motive force in V. alginolyticus also plays an indispensable role in energetics, especially at acidic pH. Since the respiration-dependent Na+ pump has a low activity at acidic pH, it must be the proton motive force that drives the generation of the Na+ motive force via the Na+/H' antiport system and the synthesis of ATP at acidic pH.
The Na' pump seems to give advantages to V. alginolyticus living in Na'-rich environments. First of all, under the conditions where a passive influx of Na+ is significant, the extrusion of Na+ as an immediate result of electron transport may be more economical than the extrusion of Na' by the Na+/H' antiport system energized by the proton motive force. Furthermore, the driving force for solutes uptake can be obtained directly from an oxidation of substrate as in the case of nonhalophilic bacteria which generate the proton motive force and take up many solutes by H+ symport systems. There may be an additional advantage in having the primary Na' pump at alkaline pH. As far as reported, the cytoplasmic pH in microorganisms is regulated near neutral and becomes even more acidic than external pH under alkaline conditions. The proposed mechanism concerning the regulation of cytoplasmic pH involves the Na+/H' antiport system which enables cells to generate A# and ApH in opposite directions (28). However, since such a system may consume more energy a t alkaline pH than it does at acidic pH (as), the generation of A# by the Na+ pump, which does not cause the alkalinization of cytoplasm, seems to be economically superior under alkaline and Na+-rich environments to the generation of A# by the proton pump, which inevitably causes the alkalinization.