Methanogenesis and the K+ Transport System Are Activated by Divalent Cations in Ammonia-treated Cells of Methanospirillurn hungatei*

We describe a K+ transport system in Methanospi- rillum hungatei cells depleted of cytoplasmic K* via an ammonia/K+ exchange reaction (Sprott, G . D., Shaw, K. M., and Jarrell, K. F. (1984) J. Biol. Chem. 259, 12602-12608). Ammonia-treated cells contained low concentrations of ATP and were unable to make CH4 or to transport “‘Rb+. All of these properties were restored by CaC12, MgCl2, or MnCl,, and not by CoC12 or NiC12. The Rb’ transport system had a K,,, of 0.42 and V,,, of 29 nmol/min.mg; K+ inhibited competi-tively. Both Hz and C02 were required for appreciable transport, whereas air, valinomycin, or nigericin were potent inhibitors. The influx of Rb+ was electrogenic and associated with proton efflux, producing a ApH (alkaline inside) in acidic media. In the absence of K+ (or Rb+), the activation of CH, synthesis by Mg2+ pro- duced little change in the cytoplasmic pH, showing that methanogenesis did not elicit a net efflux of protons. The pH optimum for transport was in the range 6.0-7.3 where the transmembrane pH gradient would con- tribute minimally to the proton motive force. Protonophores at pH 6.3 caused a partial decline in CH4 synthesis and the ATP content and dramatically col-lapsed


7.3
where the transmembrane pH gradient would contribute minimally to the proton motive force. Protonophores at pH 6.3 caused a partial decline in CH4 synthesis and the ATP content and dramatically collapsed Rb+ transport. These and other inhibitor experiments, coupled with the fact that the Rb+ gradient was too large to be in equilibrium with the proton motive force alone, suggest a role for both ATP and the proton motive force in Rb' transport. Also, a role for K+ in osmoregulation is indicated.
During growth via Hz-dependent COz reduction, methanogenic bacteria exhibit a wide species variation in their cytoplasmic K+ contents, but all accumulate K+ to concentrations above those in the medium (1-3). This ability to accumulate K+ is typical of eubacteria and Halobacterium species (see Ref. 4), being linked to functions of enzyme activation, osmoregulation (5), and pH homeostasis (6)(7)(8)(9)(10)(11)(12)(13). In the case of the methanogen grouping of archaebacteria, little is known about either the function or transport of this important cation (4). It may be relevant that Methanobacterium thermoautotrophicurn maintains a cytoplasmic K+ concentration (3) which corresponds to the 1 M KC1 required for maximal activation of the partially purified hydrogenase (14).
* This is National Research Council of Canada Publication 24503.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom correspondence should be addressed. During the reduction of COz to CH4, methanogenic bacteria appear to maintain their cytoplasmic pH slightly acidic (pH 6.6-6.8) and to generate an electrical potential (inside negative) of -120 to -200 mV (15)(16)(17)(18)(19)(20). Growth is usually conducted in media buffered near p H 6.8 where the membrane potential is the predominant, or sole, component of the pmf.' An ATP pool is maintained in the range of 1-23 nmol/mg protein (4, 21), being especially large in Msp. hungatei (21). These factors may be relevant to the bioenergetics of K+ movement, since activity of the low affinity, constitutive systems of Escherichia coli (22) and Streptococcus faecalis (9) requires both ATP and the pmf.
A dramatic ammonia/K+ exchange reaction occurs in Msp. hungatei, Ms. barkeri, E. coli, and Bacillus polymyxa (20). We postulated that following diffusion inward of the NH, species, a K+/H+ antiport activity is triggered in a futile attempt to prevent large increases in the cytoplasmic pH (20). A K+/H+ antiport model has been used to explain the loss of cytoplasmic K+ when weak alkylamines are added to Vibrio alginolyticus or E. coli (23). Because the K'/H+ antiporter activity may react differently, or be absent, in different bacteria, this model seems preferred to a simple leak mechanism in explaining the ineffectiveness of ammonia to cause K+ efflux from several other methanogens (20). The exchange reaction is likely to affect growth, especially in media of alkaline pH, although it remains unclear whether organisms exhibiting the exchange would be more, or less, sensitive to ammonia toxicity.
The purpose of the present study was to investigate K' transport in Msp. hungatei pretreated with ammonia to deplete the cytoplasmic K+. During this approach, we found that methanogenic activity was lost, but could be recovered by the addition of Caz+ or M2+. Cells so activated were then able to make ATP and to transport K+. Methanogenic activity did not result in net H+ efflux unless K+ was being transported.
Organism and K+ Depletion-Msp. hungatei strain GP1 was obtained from Patel et al. (24). The conditions for the growth of this organism in a prereduced, defined salts medium were described pre- The abbreviations used are: pmf, proton motive force; Msp., Methanospirillum; Ms., Methanosarcinu; Mb., Methanobacterium; TPMP+, triphenylmethylphosphonium ion; Hepes, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid TCS, 3,3',4',5-tetrachlorosalicylanilide; CCCP, carbonyl cyanide m-chlorophenylhydrazone; pH,, pH of the medium. viously (20). Cells in the late logarithmic to early stationary growth phase were harvested anaerobically (1, 4) in sealed centrifuge tubes, containing COz/H, (1:4, v/v). Each pellet (3-4 mg dry weight/lO ml of growth medium) was washed once in 0.1 M Hepes buffer adjusted to pH 6.5 with NaOH (30 mM). Anaerobiosis was maintained throughout by using buffer solutions prereduced with a stream of HZS (cysteine was not used) and monitored with resazurin (20). Buffers were dispensed by syringe and stored until needed in 1-liter growth bottles under a positive pressure of Nz. K+ depletion was achieved by resuspending each cell pellet into 5 ml of the wash buffer and injecting approximately 80 mM NH,OH, to obtain a pH, of 8.0-8. 4. The ammonia/potassium exchange reaction (20) was allowed to proceed for 15 min at ambient temperature. Each cell suspension was then washed twice in prereduced Hepes or Tris-HC1 buffer (5 ml/tube), prepared as above as 0.02 or 0.1 M. In some experiments it was necessary to remove contaminating cations from the buffer. In such cases, 1 liter of 0.05 M Trizma (Tris base) was treated for 30 min by stirring with 10 g of Dowex 50W-X8 resin in the H' form. The base was then adjusted to pH 7.5 with HCI, prereduced with H2S, and dispensed under an N, atmosphere. This procedure decreased the MgZ+ concentration from approximately 15 p M to less than 0.3 MM. Unless stated otherwise, methanogenesis was activated in an atmosphere of COz/H, (1:4, v/v) by incubating for 1 h in buffer containing 10 mM MgCl2.
=Rb+ Transport-Cell suspensions in serum bottles received 0,free "RbC1 (1.0 mM unless stated otherwise) to begin the reaction. Samples of 0.5 ml were removed by syringe (previously flushed with CO,/H,) and filtered, using 0.45-pm cellulose acetate filters. Each filter with adhering cells was washed with 0.1 M LiCI, prepared aerobically. No difference in label retention was noted in cases where the prereduced buffer was compared to aerobic LiCl as wash fluid. Filters were counted in water by liquid scintillation spectrometry. Following each analysis, 0.1-ml aliquots of each reaction mixture were counted for calculation of specific activity.
Proton Motiue Force-Measurements of the chemical potential using butyric acid and the membrane potential with TPMP+ in the presence of tetraphenylboron have been described (20). The internal space measured before (3) was used. In the case of the chemical potential where a centrifugation method was used, the total (extracellular plus intracellular) H 2 0 space was determined with urea (20).
Cation Measurements-Cell samples of 3-4 mg (dry weight) were filtered as before (20) and washed with 10 ml of 0.1 M LiCl solution. The filters were placed in small glass Petri dishes, and cations were extracted with 0.2 ml of hot 1-butanol (1). The cations K', Ca2+, and MgZ+ in 2.5 ml of H20 were quantitated by atomic absorption spectrometry.
Ammonia Extraction-Cell samples, filtered and washed as above, were placed in glass Petri dishes containing 2.5 ml of H,O plus 0.2 ml of 1-butanol. Following heating at 92 "C for 30 min, the volume was taken to 2.5 ml, debris removed by centrifugation, and ammonia measured colorimetrically (25).
ATP Pools-ATP was extracted from Msp. hungatei with cold HClO, and analyzed as described before (21).
Methane Synthesis-CH, was quantitated in 0.2-ml aliquots of head space gas by gas-liquid chromatography (20).

RESULTS
Cution Activation of =Rb' Transport-Msp. hungatei depleted of K' by ammonia treatment (20) lost the ability to reaccumulate the ion. However, transport activity was restored with sigmoidal transport kinetics if MgZ+ was included in the transport buffer, indicating that the divalent ion was an activator. K+-depleted cells exposed to MgZ+ for 1 h prior to initiating transport displayed typical Michaelis-Menten kinetics, achieving steady state within 30-60 min (Fig. 1). The concentration gradient was 300-fold, and the cytoplasmic concentration of Rb+ was very similar to the original K' concentration prior to depletion. Air, which is a potent inhibitor of methanogenesis (4), prevented transport. In K'-loaded cells, the uptake of =Rb' was much diminished. It is noteworthy that K' transport in S. fuecalis is subject to feedback regulation (9), a property which seems typical of bacteria.
The activation of Rb' transport by Mg2+ can be explained through an activation of methanogenesis. Ammonia-treated Cells of Msp. hungatei were K'-depleted by incubation with ammonia, washed twice in the transport buffer (pH 6.5), and activated by incubating for 60 min with 10 mM MgC12. K+-loaded cells were treated similarly except for the omission of the ammonia treatment. A gas atmosphere of COz/Hz (1:4, v/v) was used throughout the experiment. In the case of air inhibition, K+-depleted cells were incubated under an atmosphere of air from the point where MgC1, was added. At zero time, "RbCl was added to a final concentration of 1.0 mM, 1.48 X 10' cpm/pmol. Transport was assayed by filtration. cells produced little CH4, had lost most of their ATP pool, and retained a K+ content of only 5 mM (Fig. 2). Upon incubation with low concentrations of M$' , methanogenesis was restored even though the Mg2' (and Ca2+) content of the cells had been little influenced by the ammonia treatment. We stress that K' and Rb' were absent for those curves showing recovery of CH4 and ATP synthesis. Also, 10 mM MgC1, had only small stimulatory effects on CH, synthesis in K'-loaded, untreated cells (Table I).
Various divalent ions were tested for their ability to restore CH, synthesis and Rb' transport in K+-depleted cells. Methanogenesis and Rb' transport in the absence of divalent ions were 6.6 and 2.5 nmol/min.mg, respectively. Exposure of the cells for 1 h to 10 mM of M$+, Ca2+, and Md' resulted in activities of 217, 185, and 131 nmol of CH,/min.mg and 23, 25, and 15 nmol Rb+/min'mg. Fez+, Co2+, and Ni" chlorides were relatively ineffective.
Sodium can stimulate methanogenesis in several methanogens, including Msp. hungatei (26). However, we show here that CH, synthesis is recovered to rates similar to the nondepleted control cells in the absence of added Na' (Figs. 2 and 3). In these cases, contaminating cations in the buffer had been removed with an exchange resin and glassware acid cleaned. We conclude that exogenous Na+ need not be provided for CH4 synthesis to occur in ammonia-treated cells of Msp. hungatei when incubated with 10 mM MgC12. The K,,, for Na' in methanogenesis decreases if the medium contains low amounts of K+ (27). Thus, this apparent discrepancy with claims that Na' is required for CH, synthesis in Msp. hungatei (26) may be explained, at least in part, by the absence of environmental K' in our studies. In the presence of MgCi2, 25 mM NaCl had no influence on methanogenesis becoming inhibitory a t 50 mM. Sodium seemed to have some ability to activate CH, synthesis in the absence of M F , but this may be attributed to the contamination by Mg2' of the NaCl solution (15 pM Mg' present in 50 mM NaCl solutions).  , v/v) atmosphere. The exchange reaction was allowed to proceed for 15 min in the presence of 35 mM NH,OH. Removal of the ammonia and MgClz was achieved by washing the cells twice in 5-ml amounts of Dowex-treated Tris-HC1 buffer. Each cell pellet (3.05 mg, dry weight) was resuspended in 5 ml of Tris-HC1 buffer and transferred to 60-ml serum bottles for CH4 analysis. The rate of CH, synthesis was calculated from the linear response occurring between 20 and 60 min. At the end of the 60-min period, each reaction mixture was filtered to determine the amount of cytoplasmic K' . Filtrates were of pH 6.6. to ammonia prevented the loss of methanogenic activity, even though the degree of K+ efflux was hardly affected (Table I).
Transport Kinetics-In K+-depleted cells, Rb+ was transported with a K , of 0.42 mM and Vmax of 29 nmol/min.mg dry weight (Fig. 4). K+ inhibited Rb' transport competitively; with 1.0 mM KC1 included in the transport assay, the K , increased to 2.0 mM and Vmax was 31 nmol/min.mg. No preference for K+ over Rb' was seen, since a 45% inhibition in the initial rate of =Rb+ transport resulted when both ions were 1.0 mM. K+-associated Ion Movements-The ammonia/K+ exchange reaction (20) was excluded as a mechanism for Rb' uptake by the necessity for CH, synthesis to drive uptake (previous section) and by quantitative ammonia measurements. Following K+ depletion, and washing to remove the cytoplasmic ammonia, the cell-associated ammonia concentration was low and remained quite constant during Rb' uptake (Fig. 5).
The K' transport system is electrogenic, resulting in a large decrease in the membrane potential during Rb+ uptake (Fig.  6). Since there is nonspecific binding of [3H]TPMP+ by the filter and cells, it was necessary to determine that the TPMP' released during Rb' transport indicated a decline in the membrane potential and not release of the nonspecific component. This was done by comparing the time-dependent efflux of [3H]TPMP+ from cells transporting Rb' to K+loaded cells where Rb' uptake is much less (Fig. 1). It is evident that RbCl when added to K+-loaded cells caused relatively small effects (Fig. 6). Although uptake is clearly electrogenic, the apparent complete quenching of the membrane potential in cells actively transporting Rb+ is not certain because of the inaccuracy in measuring small membrane potentials.
An electrogenic uptake of K' (or Rb') is associated with the formation of a chemical potential (Table 11) FIG. 5. The K+ transport system is not coupled to ammonia efflux. A series of cell pellets (3.86 mg, dry weight) were K+-depleted with ammonia and washed twice in 0.1 M Hepes (pH 6.5) containing 10 mM MgC12. Each pellet was resuspended in 5 ml of Me-containing buffer and transferred from the centrifuge tube to a 60-ml serum bottle. A gas phase of C02/H2 (1:4, v/v) was used throughout. Cell suspensions were incubated for 1 h at 35 "C (100 rpm). To duplicate suspensions, =Rb+ (0.387 X lo6 cpmlpmol) was injected and samples (0.5 ml) filtered from zero time to 60 min. In the case of ammonia measurements, an entire reaction mixture (5 ml) was filtered at various times, extracted, and ammonia measured colorimetrically. plasmic pH, even though CH4 synthesis was fully activated by M2'.
In separate experiments, we found the methanogenic rate of K+-depleted cells to remain constant and unaltered by the injection of KC1 (2 mM) to initiate K' transport. the pH of the medium was 6.0-7.2 (Fig. 7 ) . A precipitous decline in transport at alkaline pH values closely followed the methanogenic rate reported before (20), although Rb' uptake was less sensitive to acidic pH. Inhibitor Effects-The most potent inhibitors of Rb' uptake were valinomycin, nigericin, and protonophores. Nigericin (20 p~) inhibited initial transport rates by 83% and valinomycin (20 p~) inhibited by 73% (data not shown). Inhibition of methanogenesis by gramicidin is prevented by the outer wall layers (28), explaining the inability of gramicidin (20 Mg/ml) to inhibit Rb' transport in this study. Monensin (20 p~) inhibited neither methanogenesis at pH 6.9 (28) nor Rb' K" Transport Are Activated by Divalent Cations uptake at pH 6.3. Inhibition of methanogenesis by monensin is pH-dependent in Mb. thermoautotrophicum (26) and Mb. bryantii (16). This pH effect is unlikely to explain the lack of inhibition noted above, since at pH 6.2 with marine isolate Methanococcus voltae monensin inhibits CH, synthesis, abolishes the pH gradient (inside alkaline), and lowers dramatically the ATP ~0 0 1 .~ Rather, the lack of inhibition by monensin in Msp. hungatei supports our conclusion that Rb' uptake and methanogenesis (Figs. 2 and 3) do not require added Na' . These results tend to exclude any obligatory coupling between Rb' and Na' movements.
Inhibition of Rb' uptake by valinomycin leads to the conclusion that Rb' (K+) is not in equilibrium with the membrane potential, as discussed for S. faecalis (9). Measurements of Rb' uptake following 90-min exposures to the label indicated a concentration gradient of about 2300-fold, in the range of 2-10 FM substrate. Distributions of TPMP' were much lower, in the 120-to 180-fold concentration range.
The inhibition of CH, synthesis by protonophores was less acute than inhibitions to Rb' transport (Fig. 8). Since current evidence favors that both CH, synthesis (4,29,30) and Rb' transport depend on the pmf, Rb' transport may be gated, as suggested for Alteromonus haloplanktis (7). In this case, the magnitude of the pmf would be sufficient to maintain a decreased rate of CH, synthesis, but insufficient for Rb' transport.
In Saccharomyces frugilis, CCCP is thought to inhibit sugar transport by direct interaction with the carrier (31). To test this possibility, Rb' (50 FM) was allowed to accumulate to a steady state value of 48 nmol/mg. Inhibitors were then added and efflux measured after 15 min (not shown). In the case of CCCP (40 FM), only 4 nmol of %Rb+/mg was found to efflux, K. F. Jarrell and G. D. Sprott, unpublished information. while TCS (20 PM) caused the efflux of 20 nmol of =Rb+/mg. Since efflux is likely to occur through the carrier, it seems that TCS does not interfere with carrier activity, while for CCCP it is less clear. Attempts to measure the effects of protonophores on =Rb'/K' exchange were not successful, because the exchange was slow and incomplete (Fig. 9).
Acetylene inhibits CH, synthesis in Msp. hungatei with the result that ATP pools decline (21). Incubation of K+-depleted cells with 116 PM dissolved acetylene for 1 h resulted in a decline in the CH4 synthesis rate to 4%, a loss of 50% of the ATP pool, but retention of the membrane potential (146 mV). These *cells transported Rb' at 35% of the rate found for uninhibited cells, providing correlative evidence for a role for ATP (or a product of ATP) in transport.
Osmoregulation-Glucose does not penetrate the cytoplasm of a number of methanogens (1, 3) including Msp. hungatei (I). Changes in the osmotic strength of the medium through the inclusion of glucose (up to 100 mM) resulted in an increase in the steady state Rb' accumulation, but little change in either initial transport rates or CH4 synthesis. Further increases in glucose concentration fully inhibited CH, synthesis and caused severe inhibitions in Rb' transport ( Fig. 10). Previously, we noted that large increases in the osmotic strength of the medium cause K' to efflux from Msp. hungatei (20).

DISCUSSION
Exposure of Msp. hungatei to ammonia at alkaline pH results in an ammonia/K+ exchange (20) and in the loss of methanogenic activity. Cells washed free of ammonia recover their ability to make CH, when exposed to Ca", M$+, or Mn2+. This dramatic enhancement of CH4 synthesis upon exposure to divalent ions does not occur in K'-loaded cells, showing that ammonia was responsible for the activity loss. T o explain these effects, two alternative models should be considered. The loss of methanogenic activity and K+ efflux occur with a similar dependency on the pH of the medium (20) but may be considered in the first model to represent two separate effects of ammonia: the first effect to increase K' / H' antiporter activity by increasing the cytoplasmic pH, and the second to displace from the cell surface divalent ions critical to CH, synthesis. This model can explain the lack of K' efflux from certain bacteria treated with ammonia, if these organisms lack the K+/H+ antiporter (20). The second model predicts that displacement of divalent ions from the cytoplasmic membrane by ammonia is the primary event to cause a concomitant efflux of K' and loss of methanogenic activity.
In support of this model, divalent ions, notably M$+, are known to decrease the permeability to K' of mitochondrial (32), erythrocyte (33), and yeast (34) membranes. This second model is largely discounted, however, by the finding that in the presence of MgC12 ammonia still induced K' to efflux, but methanogenic activity was retained (Table I). This finding is consistent with the first model and our previous interpretation of ammonia/K+ exchange (20), pointing to a second effect of ammonia to displace divalent ions necessary for CH, synthesis. Activation by divalent ions shows the site is exposed to the medium, because the ammonia treatment caused little change in the content of cytoplasmic M$+ or Ca2+. The decline in intracellular ATP concentration (Fig. 2 ) may explain the loss of methanogenesis, or vice versa, if a tight coupling between CH, and ATP synthesis occurs in this methanogen, as it does in Ms. barkeri (18,30). ATPase inactivation seems unlikely because activation by divalent ions occurs near the cell surface. Activation of a membrane-bound methanogenic enzyme is possible, since in Msp. hungatei part of the methanogenic apparatus may be exposed to the medium (20). Also, the ether-linked membrane lipids (35) could be the activation site. In either event, the importance of the cytoplasmic membrane in methanogenesis from CO, is emphasized (4, 29, 30).
Once depleted of K' by ammonia treatment, Msp. hungatei cells had lost not only their ability to make CHI, but also the ATP pool was diminished and Rb' transport was of low activity. Activation of methanogenesis by M$+ resulted in net ATP synthesis. Under these conditions, Rb+ was taken up by the K+ transport system to a steady state concentration approximating the K' content of nondepleted cells. Similarly, in starved cells of Ms. barkeri, the onset of methanogenesis upon methanol addition led to ATP synthesis (30). Since acetate is a central intermediate in pathways of intermediary metabolism in methanogens and our culture of Msp. hungatei cannot form acetate from CO, (36), it is likely that methanogenesis is linked to ATP synthesis via a chemiosmotic mechanism rather than by substrate phosphorylation. This agrees with most previous predictions (4, 29, 30).
Little net movement of protons is associated with methanogenic activity in K+-depleted cells in an acidic buffer lacking Na' and K+ (Table 11). This is reminiscent of the need for electrogenic K' influx to serve as a counterion to allow increased rates of H' ejection and establishment of ApH in S. faecalis (37), E. coli (38), or V. alginolyticus (10). In most cases, K' influx is associated with increases in the glycolytic (37) or respiratory (10, 39) rates. In contrast, no change in the methanogenic rate is seen in Msp. hungatei during the uptake of K' . Thus, the present results are consistent with a decreased energy demand for proton efflux to occur (passive proton efflux is not excluded), as a result of electrogenic K+ (or Rb') influx. The possibility of an ATPase exchanging H' / K' as hypothesized for E. coli (40) should be considered, but a 1:1 stoichiometry is excluded by the electrogenic nature of K' uptake. Blaut and Gottschalk (30) suggested that methyl coenzyme M reduction is linked to proton translocation in Ms. barkeri metabolizing methanol, based on the ability of an uncoupler to relieve the inhibition to CH, synthesis caused by N,N" dicyclohexylcarbodiimide. In Mb. thermoautotrophicum a net proton influx during methanogenesis in alkaline media may be interpreted in favor of a model based on internal vesicles (41) or attributed to secondary cation/H+ antiporters. The mechanism of pmf formation in methanogens is beyond the scope of this study and remains unresolved. However, interpretations based on internal membranes are excluded by the absence of significant numbers of these structures from Msp. hungatei, grown as in this study (42).
Archaebacteria include the methanogens, extreme halophiles, and thermoacidophiles and represent a third line of evolutionary descent separate from the eucaryotic and eubacterial lines (43). Here we offer for comparison the first data on K' transport in a methanogen. In spite of the early divergence in evolution, the methanogen system resembles in several aspects the low affinity K+ transport system of the eubacteria reported on. First, the influx of K' may be subject to feedback regulation (9) because of the enhanced rates of influx observed following depletion of cytoplasmic K+ (Fig.   1). Inhibitor data indicate a requirement for both ATP (or a product of ATP) and the pmf to achieve gradients as high as 2300-fold, as found for the low affinity K' transport system of certain eubacteria (9,22). Also, K+ uptake in Msp. hungutei is electrogenic and associated with the formation of a ApH in acidic media. K' influx is linked to the efflux of H' or Na' in several eubacteria (10, 37, 38). A role for cytoplasmic K' as an osmoregulator, as in E. coli (5), is implied from the increase in cytoplasmic K' concentrations at steady state when the osmotic pressure of the medium is increased (Fig. 10). In Mycoplasma mycoides, Na+-dependent K' influx may be catalyzed by a Na',K'-ATPase, typical of higher organisms (44). However, in another eubacterium (13) and Msp. hungatei, K' transport does not depend on Na' movements.