ARCHIVES
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 281, No. 1, August 15, pp. 13-20, 1990
Cooperative Regulation of the Na+/H+-Antiporter in Halobacterium halobium by ApH and A4 Naoyuki Murakami and Tetsuya Konishi’ Department of Radiochemistry-Biophysics, Niigata College of Pharmacy, 5-13-2 Kamishin’ei, Niigata 950-21, Japan
Received December 20,1989, and in revised form April 2,199O
The contributions of the transmembrane pH gradient (ApH) and electrical potential (A& to the A&+-driven Na+ efflux (mediated by the N,N’-dicyclohexylcarbodiimide-sensitive Na+/H+-antiporter) were investigated in membrane vesicles of Halobacterium hdobium. Kinetic analysis in the dark revealed that two different Na+-binding sites are located asymmetrically across the membrane: One, accessible from the external medium, stimulation of Na+ efflux) of has a Kd (half-maximal about ~50 m&f, and the NaC binding to the site is a prerequisite for the antiporter activation by A&+. The other cytoplasmic site is the NaC transport site. The K, for the cytoplasmic Na+ decreased as the ApH inessentially constant creased, while the V,, remained in the presence of defined A#J (140 mV). On the other hand, AI$ elevation above the gating potential (-100 changes in the K,,, in the mV) increased the V,, without presence of a fixed ApH. It was also noted that the K,,, value in the absence of A4 was completely different from and far higher than that observed in the presence of A4 (>lOO mV), indicating the existence of two distinct conformations in the antiporter, resting and A4 gated; the latter state may be reactive only to ApH. On the basis of the present data and the previous data on the pH effect (N. Murakami and T. Konishi, 1989 Arch. Biochem. Biophye. 271, 515-523), a model for the ApH-A# regulation of the antiporter activation is proposed. 0 1990 Academic Press, Inc.
Halobacterium hulobium, an extremely halophilic bacterium, has a unique light-energy transduction mechanism. Two retinoid proteins, bacteriorhodopsin (bR) and halorhodopsin, act as light-driven electrogenic pumps for H+ and Cl-, respectively (1, 2). The primary energy stored by these pumps is buffered by other ions (3). In this respect, Na+ is extremely important to this r To whom correspondence
should be addressed.
0003-9861/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
organism not only because it lives in a high NaCl environment but also because Na+ acts as the coupling ion for nutrient uptake (4). A Na+ circulation mechanism was first proposed by Lanyi and MacDonald (5) in halobacterial envelope vesicles in which a A&-driven Na+/ H+-antiporter plays an important role in transforming ApH established by bR into ApNa.2 The same Na+/H+ antiport mechanism was suggested by another group (6), but the molecular features of the antiporter remain unclear. Our recent studies (7-12) have demonstrated several unique aspects of the role of the halobacterial antiporter and its Ain+ coupling mechanism. For example, the antiporter mediates apparent A~H-AI$ transformation (7,8), a process that is DCCD sensitive (7-9). The antiporter is driven by Abn+, but the constituents of Abn+, ApH and A4, regulate separate steps of the antiporter activation. For instance, halobacterial antiporter is gated solely by A$ (lo-12), in contrast to the Escherichia coli system which is gated by ApH (13, 14). Another interesting feature is that the antiporter facilitates Na+ translocation only from the cytoplasm to the external medium without regard to the direction or the size of ApNa: Reversal of the direction of ApH and ApNa does not stimulate Na+ influx (12). In this respect, the halobacterial antiporter differs markedly from that of nonhalophilic cells, which transports Na+ to either side of the membrane depending on the direction of ApH or ApNa imposed (S-18). These findings indicate that H. hulobium, which lives in a high saline environment, has developed the Na+/H+-antiporter as a A&+ pump for Na+. Therefore, understanding the mechanism of Ain+ activation of the halobacterial Na+/H+-antiporter
2 Abbreviations used: ApNa, bR, bacteriorhodopsin; transmembrane sodium gradient; ApH, transmembrane pH gradient; A& transmembrane electrical potential; Afin+, electrochemical potential of proton; DCCD, NJ’-dicyclohexylcarbodiimide; Hepes; 4-(2-hydroxyethyl)-1-piperazineethanesuphonic acid, Pipes, 1,4-piperazinediethane sulfonic acid; Eppes, 4-(2-hydroxyethyl)-l-piperazinepropane sulfonic acid; Mes, 4-morpholineethanesulfonic acid. 13
Inc. reserved.
14
MURAKAMI
AND
KONISHI
should prove valuable for the general understanding of the coupling mechanism of the secondary pump to A;c,+ . In the present investigation, we studied the effects of cytoplasmic and medium [Na+] on the Na+ efflux kinetics under various ApH and A$ conditions in the dark, in an effort to understand the kinetic roles of ApH and A4 in the antiporter-mediated Na+ efflux in halobacterial membrane vesicles. EXPERIMENTAL
PROCEDURES
Preparation of membrane vesicles. Membrane vesicles were prepared from H. halobium RiMi by the freeze-thaw method, as previously described (7). The vesicles obtained were oriented more than 95% right-side-out, as determined by the menadione reductase assay (19). The vesicles were washed twice with 3 M KCl, then dialyzed against the same medium to remove as much intravesicular Na+ as possible. To change the intravesicular Na+ composition, the vesicles were suspended in a loading buffer containing Na+ at final concentration ranging from 30 to 480 mM and were stored at 4°C for 3-4 days to allow equilibration of ions. The loading buffer consisted of 3 M salt (KC1 + NaCl) and a combination of Good buffers (20 mM each of Mes, Pipes, Hepes, and Eppes) (pH 5.0-8.0). The Na+-loaded vesicles were concentrated by centrifugation at 17,000g for 60 min and resuspended in a minimal volume of loading buffers of the same salt composition (usually 30-50 mgprotein/ml). Protein was determined by the method of Lowry et al. (20) using bovine serum albumin as the standard. Na+ efiux measurement. For loading **Na+ in the vesicles, an aliquot of “NaCl (carrier free) was added to the concentrated vesicle preparation (5 &i/ml, 4.0 X 105-2.5 X 10’ cpm/pmol Na+), and the suspension was stored in a refrigerator for 2-4 days to allow equilibrium distribution of “Na+ in the vesicles. The pH of the vesicle suspension was checked before each experiment. The pH gradient (ApH, outside acid) was imposed by diluting the vesicles in a lOO- to 200-fold excess volume of a dilution buffer consisting of 3 M (choline-Cl + KCl) buffered with a combination of Good buffers (20 mM each of Mes, Pipes, Hepes, and Eppes) at a defined more acidic pH (pH jump). Inside-negative A$J was generated by the valinomycin-K+ method. The size of the AI$ was adjusted by appropriately changing the KC1 concentration in the dilution buffer and was calculated according to the equation A4 (mV) = 59 X log[
(K+)&(K+M.
To study the effect of varying the external Na+ concentration, NaCl in the dilution buffer was isoosmotically substituted for choline-Cl. Na+ efflux was determined as a time-dependent decrease of the intravesicular z2Naf content after the vesicle dilution. Typically, “Na+ extrusion from the vesicles was measured as follows: A 50-~1 aliquot (1.5-2.5 mg of protein) of the concentrated vesicle suspension was preincubated for 10 min with 5 pM valinomycin, then rapidly diluted in lo-20 ml of dilution buffer. The initial rate of “Na+ extrusion (V& was calculated from the linear phase of the semilogarithmic plot of z’Na+ remaining in the vesicbs. The reaction mixture was magnetically stirred with a magnetic bar and maintained at 20°C. At a specified time, a l-ml aliquot of the reaction mixture was taken into a filtration apparatus and rapidly filtered through a presoaked cellulose nitrate filter (Toyo Roshi, pore size 0.45 pm) under reduced pressure. The vesicles trapped on the filter were immediately washed five times with 2 ml of a cold loading buffer. The radioactivity remaining on the filter was counted with a liquid scintillation counter (Aloka, LSC700) using 10 ml of Scintisol EX-H (Dotite). Nonspecific binding of “Na+ to the vesicle membrane was estimated from the amount of radioactivity retained on the filter after filtration of the 22Na+-loaded vesicles preincubated with 1 pM gramieidin for 1 h in the dilution buffer. Non-
0
1
234567 min
FIG. 1. Time course of Ai&+-driven Na+ efflux. Vesicles were loaded with buffers (pH 7.0) which contained 2.91 M KC1 and 90 mM nNaC1. The vesicles were diluted in K+-free isotonic 3 M choline-cl (pH 4.5) (0) or 3 M NaCl (pH 4.5) (m) in the presence of 5 pM valinomycin to generate inside-negative A4 (140 mV) simultaneously with ApH (2.5 units). The 2zNa+ remaining in the vesicles was determined by the filtration method described under Experimental Procedures. specific binding was no more than 2% of the total radioactivity on the filter and was corrected for all Na+ efflux data.
placed
Muterids. 22NaC1 (carrier free) was obtained from CEA (France). Valinomycin and gramicidin were purchased from Sigma. Good buffers were purchased from Dotite, and other chemicals were obtained from Wako Chemical Co., Ltd. (Japan).
RESULTS
Effects of External and Cytoplasmic Nat on the Nat Eflux Kinetics The effects of external ([Na+],) and cytoplasmic Nat concentration ([Na+]J on the Ajin+-coupled Na+ efflux were studied. Figure 1 shows a typical profile of Ai.&+driven 22Na+ efflux from the vesicles which were preloaded with 90 mM 22NaCl. When the 22Na+-loaded vesicles were diluted in 3 M choline-Cl in the presence of A4 (140 mV) and ApH (2 units), “Na+ efflux followed firstorder kinetics, although the apparent “Nat efflux leveled off within a few minutes when the vesicles were diluted in 3 M NaCl instead of choline-Cl. A passive Na+ influx occurs depending on [Na+], and the rate increased linearly up to about 40 nmol . (mg protein. min)-’ when the Na+-depleted vesicles were diluted in isotonic media containing increased concentrations of Nat up to 3 M NaCl (Fig. 2). Thus, the apparent slowdown of the Na+ extrusion after 2 min can be attributed to the dilution of the intravesicular radioactive Na+ due to the influx of extravesicular cold Na+. Nevertheless, the slope was almost linear even in 3 M NaCl as the dilution medium within a time range from 0 to 2 min, indicating a small contribution from the back-flux on the initial phase of Na+ extrusion. Thus, the initial linear phase of the Nat extrusion (V,,) was taken as a topic for further analysis.
Na+/H+-ANTIPORTER
2
1
0 F5 5 0P406080 TIME(min)
i? 20L!L0
1 2 CNaCil,,
3 (M)
FIG. 2. Dependence of the passive Na+ influx on the medium Na+ concentration. A O.l-ml aliquot of the vesicles (3.5 mg as protein) loaded with 3 M KC1 in 5 mM Pipes (pH 6.8) was suspended in 12 ml of buffer containing “NaCl (replaced isoosmotically with choline-Cl) and 5 mM Pipes (pH 6.8). A O.&ml aliquot of the reaction mixture was taken and the vesicles were collected by filtration. Intravesicular *‘Na+ was determined by liquid scintillation counting. The influx rate was calculated from the slope of the uptake curve and corrected by the initial specific activity of external “Na+ and an intravesicular volume of vesicle protein (3 pl/mg protein). (A) Time course of Na+ influx at various [Na+],. (B) Effect of Na+ concentration on influx rate (slope = 14 nmol Na+ . (mg protein. min. M)-‘).
The [Na+]i dependence of V,, obtained with 3 M choline-cl and 3 M NaCl as a dilution medium is shown in Fig. 3. The [Na+]i dependence exhibits a saturation profile under both conditions. Since the saturation levels of VN, were essentially identical under both conditions, it is clear that the Na+ efflux occurs actively and that the ApNa is not a direct driving force for the Na+ efflux. However, the [Na+]i dependence of V,, obtained in 3 M NaCl and 3 M choline-Cl as the dilution medium was different at the low [Na+]i. In 3 M choline-Cl, V,, increased sigmoidally with regard to [Na+]i, while V,, showed a monotonic increase in 3 M NaCl. This difference cannot be explained by the effect of reversed ApNa as mentioned above. Since the [Na+], varied with the [Na+]i imposed under the present experimental conditions where the vesicles equilibrated in different Na+ concentrations were diluted in a choline-cl-based medium, it was expected that the external Na’ would directly affect the Na+ efflux kinetics. To confirm this, the [Na’jO dependence of V,, at a fixed [Na+]i (30 mM) was studied (Fig. 3, inset). Even though inwardly directed ApNa increased the V,, increased with an increase in [Na+lo and showed a saturation profile. One possible explanation for the [Na+],-dependent increase in Na+ efflux is that the passive influx of Na+ from the external medium elevates [Na+]i , thereby accelerating Na+ efflux. This was, however, negligible because the stimulative effect of increased [Na+]i was, for example, less than 1% at [Na+]i = 60 mM when
IN Halobacterium
15
hulobium
calculated from the passive Na+ influx (Fig. 2) and the resulting change in the specific activity of 22Na+ inside the vesicles. 22Na+-Na+ exchange might be another factor affecting the apparent decrease in 22Na+ inside. To examine this possibility, 22Naf-Na+ exchange was determined under conditions where Naf-loaded vesicles were suspended in the medium containing the same 22NaC1 concentration as the intravesicular Na+ in the absence of both ApH and A4 at pH 7.0. This revealed that the 22Na+-Na+ exchange rate increased only from 1.6 to 3 nmol Na+ . (mg protein. min))’ when the Naf concentration in the medium increased from 0.1 to 3 M. It was also noted that an increase in A4 did not enhance the 22Na+Naf exchange without an increase in ApH or vice versa (see also Fig. 2 of Ref. (12)), indicating a minor contribution of 22Na+-Na+ exchange to the observed V,, . Therefore, it is concluded that there is a Na+-binding site on the external side of the membrane and the Na+ binding to the site determines the V,,. Since the 3 M choline-Cl used as a dilution medium contains at least 10 mM Na+, the [Na+], required for the half-maximal stimulation (Kd) of VNawas calculated to be no higher than 50 mM. To understand the [Na+]i dependence of V,,, all of the VN, data in Fig. 3 were analyzed according to Hanes-
0
4cxl
al0 I
FIG. 3.
Na’Ii,
600
1
(mM)
Effect of Na+ concentration in the cytoplasmic or external medium on the Na+ efflux kinetics. Vesicles equilibrated with the loading buffer (pH 7.0) which contained a combination of Good buffers (20 mM each of Mes, Pipes, Hepes, Eppes) and 30-486 mM 22NaC1 (replaced isoosmotically with KC1 to give a final concentration of 3 M) were diluted in 3 h4 NaCl (0) or 3 M choline-Cl (0) as the dilution medium at pH 5.0. The initial rate of Na+ extrusion (V,,) was calculated from the linear phase of a semilogarithmic plot of residual %Na+ in the vesicles. Inset: Effect of external Na+ concentration on the Nat efflux. The vesicles equilibrated with the loading buffer (pH 7.0) which contained 30 mM 22NaC1, 2.97 M KCl, and Good buffers were diluted in the dilution medium (pH 5.0) with various Na+ concentrations (replaced isoosmotically with choline-Cl). The vertical bars indicate the standard error.
16
MURAKAMI
[ Na+l,
(mW
AND
KONISHI
CNa+li,
(mM)
FIG. 4. Dependence of ApH-induced Na+ efflux on cytoplasmic Na+ concentration. Vesicles contain 30-480 mM “NaCl (replaced isoosmotitally by KCl) and a combination of 20 mM each of Mes, Hepes, Pipes, Eppes, at pH 5.0-7.0. The vesicles were incubated for 10 min in the and then diluted in 3 M choline-Cl buffered at pH 4.0 or 5.0 to generate Aq5 (140 mV) at various ApH’s. (A) Initial presence of 5 pM valinomycin, rate of Na+ extrusion (Vn.) vs cytoplasmic Na+ concentration [Na+]i. (B) Hanes-Woolf plot ([Na+]i/Vn, vs [Na+]J.
Woolf plots ([Na+]i/V vs [Na+]i). A straight line was obtained within the [Na+]i range 90 to 480 mM. Since the [Na’10 effect on the V,, was negligible in the above [Na+]i range, the kinetic parameters must be attributed to the cytoplasmic Na+ site(s). A replacement of 3 M choline-Cl with 3 M NaCl as the dilution medium caused little change in both K,,, for Na+ and V,,, (data not shown), indicating that the Na+ efflux kinetics are not affected by ApNa. The pH Gradient Modulates
the K,,, for Na+
We have shown previously (12) that the halobacterial Na+/H+-antiporter has two distinct H+-dissociable groups: One is accessible from the cytoplasmic and the other from the external side of the membrane. The apparent pK, for the external group (pK, 4.6) was independent of the size of A);iH+ imposed, but the apparent pK, of the cytoplasmic site was shifted to alkaline by elevating ApH. Thus, we have predicted a regulatory H+-binding site on the cytoplasmic side, and the protonation modulates the affinity for the transport Na+. For further characterization of the cytoplasmic Na+ transport site, we studied the effects of [Na+]i on the kinetics of ApH-dependent Na+ efflux at a fixed A$J (140 mV). As it was shown previously that pH, determines primarily the number of Ajin+ reactive intermediates of the antiporter (12), the effect of ApH was studied at a fixed pH,. All VNa’s at different ApH’s followed saturation profiles with respect to [Na+]i (Fig. 4A). Data plotted as Hanes-Woolf plots (Fig. 4B) yielded straight lines as a function of [Na+]i (>90 mM), suggesting the exis-
tence of another Na+-binding site accessible from the cytoplasmic side. Furthermore, the results indicated that ApH decreased only the K,,, for cytoplasmic Na+ without a significant change of V,,,,,. To analyze the ApH effect on the K, for transport Na+ and the V,,,, kinetic parameters at constant pH, and A$ were replotted as a function of ApH imposed (Fig. 5). It is apparent that the K, for transport Na+ decreased with increases in ApH, but the change in V,,, was quite small. These results support our previous observations that V,,,,, of the ApHdependent Na+ efflux is determined primarily by the extent of protonation of the external H+-dissociable group, and that the ApH-modulated pKc, shift in the cytoplasmic regulatory H+-binding site is responsible for the affinity change at the transport Na+ site (12). It could be argued that the ApH-dependent apparent K, shift for Na+ resulted from a simple competition between internal Na+ and H+ for binding to a site on the transporter, because ApH elevations increase pHi at constant pH,. However, this may not explain the fact that ApH decreases both internal K,, for H+ and K,,, for Na+ at their binding sites simultaneously. Membrane Potential
Increases the V,, for Na+ Eflux
To evaluate the underlying mechanism of A&dependent gating of the Na+/H+ exchange, we studied the effect of elevated A4 on the Na+ efflux kinetics. Figure 6A shows the VNa’s as a function of [Na+]i in the presence of ApH = 2 at a fixed pH, of 5.0. VNa showed a saturation profile under all A4 conditions. It was also noted that in the low [Na+]i range the V,, was slightly inhib-
Na+/H+-ANTIPORTER
IN H&bacterium
17
halobium
i
APH FIG. 5. plotted
Effects of ApH on the kinetic parameters. Kinetic parameters obtained by varying as a function of ApH imposed. Conditions were the same as those described in Fig. 4.
ited regardless of the presence or absence of a large A$, as shown in Fig. 4A, indicating that the Na+ binding to the external Na+ site is essential for activation of the antiporter by A&+ (or A~#J) as mentioned above. The [Na+]i dependence of VN, in the presence of various A&s was analyzed by a Hanes-Woolf plot (Fig. 6B). The results again indicated the binding of Na+ to a single transport site. Increased A4 decreased the slopes of the lines, reflecting an increased V,,,, but the K, for Naf remained essentially constant as long as pH, and ApH were fixed. Kinetic parameters obtained were plotted as a function of A4 imposed (Fig. 7). At any pHo’s, V,,,
0
2al
FIG. 6.
(mM)
A4 (140 mV) and pH, were
sharply increased at around 100 mV of A+, which is identical to the threshold A4 previously reported (10-12). Although we cannot exclude the possibility that external K+ itself inhibits Na+/H+ exchange, independent of its effects on A& such a possibility may be small because the Na+/H+ exchange activity could be observed by illumination in the Naf-loaded vesicles suspended either in 3 M NaCl or in 3 M KC1 medium (7,8). It is also notable that the K,,,‘s at the A&s below 100 mV were completely different from those obtained above at the larger AI$‘s, although the Km gap became unclear at acidic pH, (pH, 4.0). Since the apparent Km for Na+ is also dependent on
-m
ml
INa’Ii”
ApH at a specified
0
603
[ N?Iin
EM)
Dependence of A&-induced Na+ efflux on cytoplasmic Na+ concentration. Vesicles containing 30-480 mM xxNaC1 (replaced isoosmotitally with KCl) and a combination of 20 mM each of Mes, Hepes, Pipes, Eppes (pH 7.0) were diluted in the dilution medium (pH 5.0) at various K+ concentrations (replaced isoosmotically with choline-Cl) to generate various A#. (A) Initial rate of Na+ extrusion (V,.) vs cytoplasmic Na+ concentration [Na+h. (B) Hanes-Woolf plot ([Na+]JV,, vs [Na+]J.
18
MURAKAMI
300
AND
KONISHI
Km
I 200,
WHO= 4.0
p\=4.0
E
G
100
0
-i
i
py= 5.0
pH,=5.0
100
3
(mV) FIG. 7. Effects of A@ on kinetic were plotted against A+ imposed.
parameters
of Na+ efflux. Kinetic
parameters
pH, (acidic pH, increases K,) in the presence of A4 (Fig. 5A), the A&induced K, shift is possibly compensated by the pH, effect at pH, 4.0. In addition, Km in the absence of A$ increased more with increased ApH (not shown). These results indicate that two distinct states such as resting and gated conformational states are possible with or without threshold A& Since the Km remained constant above the threshold A+, it is suggested that A4 plays two distinct roles in the antiporter activation. First is the gating in which the A$ of approx 100 mV induces a conformational transition of the antiporter from resting to a ApH-reactive active state (gated state). A subsequent effect of A4 is as a driving force to facilitate ApH-dependent Na+/H+ exchange, for example, by modulating the number of active intermediates without changing the Na+ affinity for the cytoplasmic transport site as reflected in the increased V,,, with a constant Km above the threshold A&
were obtained by elevating
always larger than unity. These results strongly support the aforementioned idea that there are high- and lowaffinity Na+-binding sites and the Na+ binding to the high-affinity (external) site is a prerequisite for the A4
TABLE
Hill coefficient
LOW [Na+]i
Dilution medium
PH,
ApH
A$ (mV)
3 M Choline-Cl
4.0
4 3 2 1 3 2 2 2 2 2
-140 -140 -140 -140 -140 -140 -125 -110 -60
1 2
-140 -140
Site
The Na+ efflux data were further analyzed by Hill plotting, which is routinely used to characterize ligand binding to the receptor site (21). Under all conditions the lines obtained broke at around 90 mM [Na+]i except under the 3 M NaCl condition and in the absence of A& Hill coefficients calibrated from the least-squares fitting of the slopes above and below the breaking point are summarized in Table I. Above 90 mM [Na+]i, the Hill coefficients are almost unity under all conditions tested and are the same as that under the 3 M NaCl condition and that in the absence of A4, but below 90 mM [Na+]i where the [Naflo effect is significant, the coefficients are
I
Hill Coefficients Obtained under Several Na+ Efflux Conditions
5.0
Hill Analysis of the Na’-Binding
A# at a specified ApH (2 units) and
3 M NaCl
5.0
(lOOmV
ApH -A’4
coOperati\
“A W Sensing Vahre”
FIG. 8. Model for the cooperative regulation of the Na+/H+-antiporter by ApH and A& (See details under Discussion).
is difficult to conclude that A$ sequentially affects a certain intermediate with a fixed K, that is formed as a consequence of the effect of the ApH and increases its amount, because there is no additivity of the ApH and A+ effects on V,,. It is thus considered that A4 acts at the H+-binding and releasing steps at the regulatory H+ site as a force such as an electrical field, to increase the turnover of the process cooperatively with ApH. Model of Cooperative Regulation
by ApH and A4
Our present and earlier studies on the mechanism of regulation (10-12) tend to rule out purely kinetic models of Na+/H+ exchange activity. Although the precise mechanism of Na+ and H+ transport and whether the antiporter is a single peptide remain unclear, the ApHA$ cooperativity in antiporter activation is reasonably explained in the following model, as shown in Fig. 8. The initial step for antiporter activation is the binding of Na+ and Hf to the corresponding external binding sites. The H+-Na+ bound intermediate is then activated by A@ of approx 100 mV to the gated state and becomes reactive to A,&+. Voltage gating phenomena have been studied in detail in a simpler channel peptide such as alamethicin (22) or porin (23), where the results suggest that gating arises from the changes in the oligomeric structure in response
20
MURAKAMI
to an applied electrical field. Such a possibility may also exist in the halobacterial Na+/H+-antiporter. The subsequent Ajin+-coupled Na’/H+-exchange process is more likely mechanochemical, where the chemical process (H+ binding) is transduced directly into the force for Na+ translocation in the presence of A& probably through a protein conformational change. In order to explain the cooperative feature of ApH and AC#J in Na+ efflux stimulation, we suggest that a peptide segment carring the regulatory H+-binding site is a “A&sensing valve.” The unprotonated conformation restricts the accessibility of the cytoplasmic Na+ to the transport site even if Ad is present. Once the site is protonated, the valve becomes A&sensitive, and the A&driven conformational transition would be expected to expose the transport Na+ site. The Hf-binding site exposed to the medium would release H+ immediately, probably because of its inherently high pK,. It then becomes A& insensitive and returns to its original conformation to expel the bound Na+ to the external medium. In this model, the turnover rate of the protonationdeprotonation of the regulatory H+ site determines the Na+ efflux. Hence, the rate of H+ supply to the site is rate-limiting on the Na+ efflux as long as Ac$is constant, while the A#+assisted conformational change of the valve is rate-limiting when the rate of H+ supply is constant. As we showed previously (12), the pK, change at the regulatory H+ site gives rise to a pK, difference across the membrane because the pK, at the external H+ site (pK, 4.6) remains constant under the conditions. The pKa difference would be a direct driving force for the transmembranous supply of H+ to the regulatory H+ site, thus acting as a modulator of the turnover rate of the “A&sensing flip-flop valve” under a constant electric field. This is actually reflected in the observed ApHdependent pK, shift at the regulatory H+ site and the Km change at the Na+-transport site. It is difficult to attribute the observed pK,, shift which is not saturable over 23 units (12) to a single H+-dissociable group. Such a large pK, change could be explained only as an increased turnover of the protonation-deprotonation process at the Hf-binding site. This could also explain our observation that Ac$ does not affect the K,,, but only the V,,,,,. Recently, Na+/H+-antiporters have been extensively studied in nonhalophilic cells and several molecular features were clarified (24-30). It is of interest to explore the molecular similarity between the halophilic and the nonhalophilic antiporters, because their kinetic and structural properties are quite different from those that we revealed in the halobacterial Na+/H+-antiporter. It is probable that the halobacteria modified its antiporter into a sophisticated device for Na+ extrusion to adapt to the extremely high saline environment. The mechanism we proposed may enable the halobacterial antiporter to expel cytoplasmic Na+ against a large Na+ gradient and function as a Ajin+-driven Na+ pump.
AND
KONISHI
ACKNOWLEDGMENTS We thank Dr. Lester Packer for his helpful comments in revising the manuscript. This study was supported by a grant-in-aid for scientific research in the priority area of “Bioenergetics” to T.K. from the Ministry of Education, Science and Culture of Japan.
REFERENCES 1. Oesterhelt,
D., and Stoeckenius,
W. (1973) Proc. Natl. Acad. Sci.
USA. 70,2853-2857. 2. Schobert, B., and Lanyi, J. K. (1982) J. Biol. Chem. 257,10,30610,313. 3. Skulachev, V. P. (1978) FEBS L&t. 87,71-179. 4. Lanyi, J. K. (1978) Microbial. Rev. 42,682-706. 5. Lanyi, J. K., and MacDonald, R. E. (1976) Biochemistry 16,46084614. 6. Eisenbach, M., Cooper, B. S., Garty, H., Johnstone, R. M., Rottenberg, H., and Caplan, S. R. (1977) Biochim. Biophys. Acta 465, 599-613. 7. Murakami, N., and Konishi, T. (1985) J. B&hem. (Tokyo) 98,
897-907. 8. Murakami, N., and Konishi, 9. Murakami, N., and Konishi, 10. Murakami, N., and Konishi, 231-236. 11. Konishi, T., and Murakami, N., and Konishi, 12. Murakami,
T. (1987) B&hem. Znt. 15,215-221. T. (1988) Biochimie 70.819-826. T. (1988) J. Biochem. (Tokyo) 103,
N. (1988) FEBS L.&t. 226,270-274. T. (1989) Arch. Biochem. Biophys.
271,515-523. 13. Bassilana, M., Damiano, E., and Leblanc, G. (1984) Biochemistry 23.5288-5294. 14. Nakamura, T., Hsu, C., and Rosen, B. P. (1986) J. Biol. Chem.
261,678-683. 15. Krulwich, 16. Grinstein,
T. A. (1983) Biochim. Biophys. Acta 726,245-264. S., Goety, J. D., and Rothstein, A. (1984) J. Gen. Phys-
iol. 84,585-600. 17. Aronson, P. S. (1985) Annu. Rev. Physiol. 47,545-560. 18. Ehrenfeld, J., Cragoe, E. J., and Harvey, B. J. (1987) Pfluegers
Arch. 409,200-207. 19. Lanyi, J. K. (1969) J. Biol. Chem. 244,4168-4173. 20. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 49,105-121. 21. Changeux, J. P., Thiery, J., Tung, Y., and Kittel, C. (1967) Proc.
Natl. Acad. Sci. USA 57,335-341. 22. Fox, R. O., and Richards, F. M. (1982) Nature (London) 300,325330. 23. Mauro, A., Blake, M., and Labarca, P. (1988) Proc. Natl. Acad. Sci. USA 85,1071-1075. S., Rothstein, A., Sarkadi, B., and Gelfand, E. W. 24. Grinstein, (1984) Amer. J. Physiol. 246, C204-C215. 25. Aronson, P. S., and Igarashi, P. (1986) in Current Topics in Membranes and Transport (Aronson, P. S., and Boron, W. F., Eds.), Vol. 26, pp. 57-75, Academic Press, San Diego. 26. Nord, E. P., Goldfarb, D., Mikhail, N., Moradeshagi, P., Hafezi, A., Vaystub, S., Gragoe, E. J., Jr., and Fine, L. G. (1986) Amer. J. Physiol. 250, F539-550. 27. Grinstein, S. (Ed.) (1988) Na+/H+ Exchange, CRC Press, Boca Raton, FL. 28. Huot, S. J., Cassel, D., Igarashi, P., Cragoe, E. J., Jr., and Slayman, C. W. (1989) J. Biol. Chem. 264,683-686. 29. Padan, E., Maisler, N., Taglicht, D., Karpel, R., and Schuldiner, S. (1989) J. Biol. Chem. 264,20,297-20,302. 30. Sardet, C., Franchi, A., and Pouyssegur, J. (1989) Cell 56, 271280.
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 281, No. 1, August 15, pp. 13-20, 1990
Cooperative Regulation of the Na+/H+-Antiporter in Halobacterium halobium by ApH and A4 Naoyuki Murakami and Tetsuya Konishi’ Department of Radiochemistry-Biophysics, Niigata College of Pharmacy, 5-13-2 Kamishin’ei, Niigata 950-21, Japan
Received December 20,1989, and in revised form April 2,199O
The contributions of the transmembrane pH gradient (ApH) and electrical potential (A& to the A&+-driven Na+ efflux (mediated by the N,N’-dicyclohexylcarbodiimide-sensitive Na+/H+-antiporter) were investigated in membrane vesicles of Halobacterium hdobium. Kinetic analysis in the dark revealed that two different Na+-binding sites are located asymmetrically across the membrane: One, accessible from the external medium, stimulation of Na+ efflux) of has a Kd (half-maximal about ~50 m&f, and the NaC binding to the site is a prerequisite for the antiporter activation by A&+. The other cytoplasmic site is the NaC transport site. The K, for the cytoplasmic Na+ decreased as the ApH inessentially constant creased, while the V,, remained in the presence of defined A#J (140 mV). On the other hand, AI$ elevation above the gating potential (-100 changes in the K,,, in the mV) increased the V,, without presence of a fixed ApH. It was also noted that the K,,, value in the absence of A4 was completely different from and far higher than that observed in the presence of A4 (>lOO mV), indicating the existence of two distinct conformations in the antiporter, resting and A4 gated; the latter state may be reactive only to ApH. On the basis of the present data and the previous data on the pH effect (N. Murakami and T. Konishi, 1989 Arch. Biochem. Biophye. 271, 515-523), a model for the ApH-A# regulation of the antiporter activation is proposed. 0 1990 Academic Press, Inc.
Halobacterium hulobium, an extremely halophilic bacterium, has a unique light-energy transduction mechanism. Two retinoid proteins, bacteriorhodopsin (bR) and halorhodopsin, act as light-driven electrogenic pumps for H+ and Cl-, respectively (1, 2). The primary energy stored by these pumps is buffered by other ions (3). In this respect, Na+ is extremely important to this r To whom correspondence
should be addressed.
0003-9861/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
organism not only because it lives in a high NaCl environment but also because Na+ acts as the coupling ion for nutrient uptake (4). A Na+ circulation mechanism was first proposed by Lanyi and MacDonald (5) in halobacterial envelope vesicles in which a A&-driven Na+/ H+-antiporter plays an important role in transforming ApH established by bR into ApNa.2 The same Na+/H+ antiport mechanism was suggested by another group (6), but the molecular features of the antiporter remain unclear. Our recent studies (7-12) have demonstrated several unique aspects of the role of the halobacterial antiporter and its Ain+ coupling mechanism. For example, the antiporter mediates apparent A~H-AI$ transformation (7,8), a process that is DCCD sensitive (7-9). The antiporter is driven by Abn+, but the constituents of Abn+, ApH and A4, regulate separate steps of the antiporter activation. For instance, halobacterial antiporter is gated solely by A$ (lo-12), in contrast to the Escherichia coli system which is gated by ApH (13, 14). Another interesting feature is that the antiporter facilitates Na+ translocation only from the cytoplasm to the external medium without regard to the direction or the size of ApNa: Reversal of the direction of ApH and ApNa does not stimulate Na+ influx (12). In this respect, the halobacterial antiporter differs markedly from that of nonhalophilic cells, which transports Na+ to either side of the membrane depending on the direction of ApH or ApNa imposed (S-18). These findings indicate that H. hulobium, which lives in a high saline environment, has developed the Na+/H+-antiporter as a A&+ pump for Na+. Therefore, understanding the mechanism of Ain+ activation of the halobacterial Na+/H+-antiporter
2 Abbreviations used: ApNa, bR, bacteriorhodopsin; transmembrane sodium gradient; ApH, transmembrane pH gradient; A& transmembrane electrical potential; Afin+, electrochemical potential of proton; DCCD, NJ’-dicyclohexylcarbodiimide; Hepes; 4-(2-hydroxyethyl)-1-piperazineethanesuphonic acid, Pipes, 1,4-piperazinediethane sulfonic acid; Eppes, 4-(2-hydroxyethyl)-l-piperazinepropane sulfonic acid; Mes, 4-morpholineethanesulfonic acid. 13
Inc. reserved.
14
MURAKAMI
AND
KONISHI
should prove valuable for the general understanding of the coupling mechanism of the secondary pump to A;c,+ . In the present investigation, we studied the effects of cytoplasmic and medium [Na+] on the Na+ efflux kinetics under various ApH and A$ conditions in the dark, in an effort to understand the kinetic roles of ApH and A4 in the antiporter-mediated Na+ efflux in halobacterial membrane vesicles. EXPERIMENTAL
PROCEDURES
Preparation of membrane vesicles. Membrane vesicles were prepared from H. halobium RiMi by the freeze-thaw method, as previously described (7). The vesicles obtained were oriented more than 95% right-side-out, as determined by the menadione reductase assay (19). The vesicles were washed twice with 3 M KCl, then dialyzed against the same medium to remove as much intravesicular Na+ as possible. To change the intravesicular Na+ composition, the vesicles were suspended in a loading buffer containing Na+ at final concentration ranging from 30 to 480 mM and were stored at 4°C for 3-4 days to allow equilibration of ions. The loading buffer consisted of 3 M salt (KC1 + NaCl) and a combination of Good buffers (20 mM each of Mes, Pipes, Hepes, and Eppes) (pH 5.0-8.0). The Na+-loaded vesicles were concentrated by centrifugation at 17,000g for 60 min and resuspended in a minimal volume of loading buffers of the same salt composition (usually 30-50 mgprotein/ml). Protein was determined by the method of Lowry et al. (20) using bovine serum albumin as the standard. Na+ efiux measurement. For loading **Na+ in the vesicles, an aliquot of “NaCl (carrier free) was added to the concentrated vesicle preparation (5 &i/ml, 4.0 X 105-2.5 X 10’ cpm/pmol Na+), and the suspension was stored in a refrigerator for 2-4 days to allow equilibrium distribution of “Na+ in the vesicles. The pH of the vesicle suspension was checked before each experiment. The pH gradient (ApH, outside acid) was imposed by diluting the vesicles in a lOO- to 200-fold excess volume of a dilution buffer consisting of 3 M (choline-Cl + KCl) buffered with a combination of Good buffers (20 mM each of Mes, Pipes, Hepes, and Eppes) at a defined more acidic pH (pH jump). Inside-negative A$J was generated by the valinomycin-K+ method. The size of the AI$ was adjusted by appropriately changing the KC1 concentration in the dilution buffer and was calculated according to the equation A4 (mV) = 59 X log[
(K+)&(K+M.
To study the effect of varying the external Na+ concentration, NaCl in the dilution buffer was isoosmotically substituted for choline-Cl. Na+ efflux was determined as a time-dependent decrease of the intravesicular z2Naf content after the vesicle dilution. Typically, “Na+ extrusion from the vesicles was measured as follows: A 50-~1 aliquot (1.5-2.5 mg of protein) of the concentrated vesicle suspension was preincubated for 10 min with 5 pM valinomycin, then rapidly diluted in lo-20 ml of dilution buffer. The initial rate of “Na+ extrusion (V& was calculated from the linear phase of the semilogarithmic plot of z’Na+ remaining in the vesicbs. The reaction mixture was magnetically stirred with a magnetic bar and maintained at 20°C. At a specified time, a l-ml aliquot of the reaction mixture was taken into a filtration apparatus and rapidly filtered through a presoaked cellulose nitrate filter (Toyo Roshi, pore size 0.45 pm) under reduced pressure. The vesicles trapped on the filter were immediately washed five times with 2 ml of a cold loading buffer. The radioactivity remaining on the filter was counted with a liquid scintillation counter (Aloka, LSC700) using 10 ml of Scintisol EX-H (Dotite). Nonspecific binding of “Na+ to the vesicle membrane was estimated from the amount of radioactivity retained on the filter after filtration of the 22Na+-loaded vesicles preincubated with 1 pM gramieidin for 1 h in the dilution buffer. Non-
0
1
234567 min
FIG. 1. Time course of Ai&+-driven Na+ efflux. Vesicles were loaded with buffers (pH 7.0) which contained 2.91 M KC1 and 90 mM nNaC1. The vesicles were diluted in K+-free isotonic 3 M choline-cl (pH 4.5) (0) or 3 M NaCl (pH 4.5) (m) in the presence of 5 pM valinomycin to generate inside-negative A4 (140 mV) simultaneously with ApH (2.5 units). The 2zNa+ remaining in the vesicles was determined by the filtration method described under Experimental Procedures. specific binding was no more than 2% of the total radioactivity on the filter and was corrected for all Na+ efflux data.
placed
Muterids. 22NaC1 (carrier free) was obtained from CEA (France). Valinomycin and gramicidin were purchased from Sigma. Good buffers were purchased from Dotite, and other chemicals were obtained from Wako Chemical Co., Ltd. (Japan).
RESULTS
Effects of External and Cytoplasmic Nat on the Nat Eflux Kinetics The effects of external ([Na+],) and cytoplasmic Nat concentration ([Na+]J on the Ajin+-coupled Na+ efflux were studied. Figure 1 shows a typical profile of Ai.&+driven 22Na+ efflux from the vesicles which were preloaded with 90 mM 22NaCl. When the 22Na+-loaded vesicles were diluted in 3 M choline-Cl in the presence of A4 (140 mV) and ApH (2 units), “Na+ efflux followed firstorder kinetics, although the apparent “Nat efflux leveled off within a few minutes when the vesicles were diluted in 3 M NaCl instead of choline-Cl. A passive Na+ influx occurs depending on [Na+], and the rate increased linearly up to about 40 nmol . (mg protein. min)-’ when the Na+-depleted vesicles were diluted in isotonic media containing increased concentrations of Nat up to 3 M NaCl (Fig. 2). Thus, the apparent slowdown of the Na+ extrusion after 2 min can be attributed to the dilution of the intravesicular radioactive Na+ due to the influx of extravesicular cold Na+. Nevertheless, the slope was almost linear even in 3 M NaCl as the dilution medium within a time range from 0 to 2 min, indicating a small contribution from the back-flux on the initial phase of Na+ extrusion. Thus, the initial linear phase of the Nat extrusion (V,,) was taken as a topic for further analysis.
Na+/H+-ANTIPORTER
2
1
0 F5 5 0P406080 TIME(min)
i? 20L!L0
1 2 CNaCil,,
3 (M)
FIG. 2. Dependence of the passive Na+ influx on the medium Na+ concentration. A O.l-ml aliquot of the vesicles (3.5 mg as protein) loaded with 3 M KC1 in 5 mM Pipes (pH 6.8) was suspended in 12 ml of buffer containing “NaCl (replaced isoosmotically with choline-Cl) and 5 mM Pipes (pH 6.8). A O.&ml aliquot of the reaction mixture was taken and the vesicles were collected by filtration. Intravesicular *‘Na+ was determined by liquid scintillation counting. The influx rate was calculated from the slope of the uptake curve and corrected by the initial specific activity of external “Na+ and an intravesicular volume of vesicle protein (3 pl/mg protein). (A) Time course of Na+ influx at various [Na+],. (B) Effect of Na+ concentration on influx rate (slope = 14 nmol Na+ . (mg protein. min. M)-‘).
The [Na+]i dependence of V,, obtained with 3 M choline-cl and 3 M NaCl as a dilution medium is shown in Fig. 3. The [Na+]i dependence exhibits a saturation profile under both conditions. Since the saturation levels of VN, were essentially identical under both conditions, it is clear that the Na+ efflux occurs actively and that the ApNa is not a direct driving force for the Na+ efflux. However, the [Na+]i dependence of V,, obtained in 3 M NaCl and 3 M choline-Cl as the dilution medium was different at the low [Na+]i. In 3 M choline-Cl, V,, increased sigmoidally with regard to [Na+]i, while V,, showed a monotonic increase in 3 M NaCl. This difference cannot be explained by the effect of reversed ApNa as mentioned above. Since the [Na+], varied with the [Na+]i imposed under the present experimental conditions where the vesicles equilibrated in different Na+ concentrations were diluted in a choline-cl-based medium, it was expected that the external Na’ would directly affect the Na+ efflux kinetics. To confirm this, the [Na’jO dependence of V,, at a fixed [Na+]i (30 mM) was studied (Fig. 3, inset). Even though inwardly directed ApNa increased the V,, increased with an increase in [Na+lo and showed a saturation profile. One possible explanation for the [Na+],-dependent increase in Na+ efflux is that the passive influx of Na+ from the external medium elevates [Na+]i , thereby accelerating Na+ efflux. This was, however, negligible because the stimulative effect of increased [Na+]i was, for example, less than 1% at [Na+]i = 60 mM when
IN Halobacterium
15
hulobium
calculated from the passive Na+ influx (Fig. 2) and the resulting change in the specific activity of 22Na+ inside the vesicles. 22Na+-Na+ exchange might be another factor affecting the apparent decrease in 22Na+ inside. To examine this possibility, 22Naf-Na+ exchange was determined under conditions where Naf-loaded vesicles were suspended in the medium containing the same 22NaC1 concentration as the intravesicular Na+ in the absence of both ApH and A4 at pH 7.0. This revealed that the 22Na+-Na+ exchange rate increased only from 1.6 to 3 nmol Na+ . (mg protein. min))’ when the Naf concentration in the medium increased from 0.1 to 3 M. It was also noted that an increase in A4 did not enhance the 22Na+Naf exchange without an increase in ApH or vice versa (see also Fig. 2 of Ref. (12)), indicating a minor contribution of 22Na+-Na+ exchange to the observed V,, . Therefore, it is concluded that there is a Na+-binding site on the external side of the membrane and the Na+ binding to the site determines the V,,. Since the 3 M choline-Cl used as a dilution medium contains at least 10 mM Na+, the [Na+], required for the half-maximal stimulation (Kd) of VNawas calculated to be no higher than 50 mM. To understand the [Na+]i dependence of V,,, all of the VN, data in Fig. 3 were analyzed according to Hanes-
0
4cxl
al0 I
FIG. 3.
Na’Ii,
600
1
(mM)
Effect of Na+ concentration in the cytoplasmic or external medium on the Na+ efflux kinetics. Vesicles equilibrated with the loading buffer (pH 7.0) which contained a combination of Good buffers (20 mM each of Mes, Pipes, Hepes, Eppes) and 30-486 mM 22NaC1 (replaced isoosmotically with KC1 to give a final concentration of 3 M) were diluted in 3 h4 NaCl (0) or 3 M choline-Cl (0) as the dilution medium at pH 5.0. The initial rate of Na+ extrusion (V,,) was calculated from the linear phase of a semilogarithmic plot of residual %Na+ in the vesicles. Inset: Effect of external Na+ concentration on the Nat efflux. The vesicles equilibrated with the loading buffer (pH 7.0) which contained 30 mM 22NaC1, 2.97 M KCl, and Good buffers were diluted in the dilution medium (pH 5.0) with various Na+ concentrations (replaced isoosmotically with choline-Cl). The vertical bars indicate the standard error.
16
MURAKAMI
[ Na+l,
(mW
AND
KONISHI
CNa+li,
(mM)
FIG. 4. Dependence of ApH-induced Na+ efflux on cytoplasmic Na+ concentration. Vesicles contain 30-480 mM “NaCl (replaced isoosmotitally by KCl) and a combination of 20 mM each of Mes, Hepes, Pipes, Eppes, at pH 5.0-7.0. The vesicles were incubated for 10 min in the and then diluted in 3 M choline-Cl buffered at pH 4.0 or 5.0 to generate Aq5 (140 mV) at various ApH’s. (A) Initial presence of 5 pM valinomycin, rate of Na+ extrusion (Vn.) vs cytoplasmic Na+ concentration [Na+]i. (B) Hanes-Woolf plot ([Na+]i/Vn, vs [Na+]J.
Woolf plots ([Na+]i/V vs [Na+]i). A straight line was obtained within the [Na+]i range 90 to 480 mM. Since the [Na’10 effect on the V,, was negligible in the above [Na+]i range, the kinetic parameters must be attributed to the cytoplasmic Na+ site(s). A replacement of 3 M choline-Cl with 3 M NaCl as the dilution medium caused little change in both K,,, for Na+ and V,,, (data not shown), indicating that the Na+ efflux kinetics are not affected by ApNa. The pH Gradient Modulates
the K,,, for Na+
We have shown previously (12) that the halobacterial Na+/H+-antiporter has two distinct H+-dissociable groups: One is accessible from the cytoplasmic and the other from the external side of the membrane. The apparent pK, for the external group (pK, 4.6) was independent of the size of A);iH+ imposed, but the apparent pK, of the cytoplasmic site was shifted to alkaline by elevating ApH. Thus, we have predicted a regulatory H+-binding site on the cytoplasmic side, and the protonation modulates the affinity for the transport Na+. For further characterization of the cytoplasmic Na+ transport site, we studied the effects of [Na+]i on the kinetics of ApH-dependent Na+ efflux at a fixed A$J (140 mV). As it was shown previously that pH, determines primarily the number of Ajin+ reactive intermediates of the antiporter (12), the effect of ApH was studied at a fixed pH,. All VNa’s at different ApH’s followed saturation profiles with respect to [Na+]i (Fig. 4A). Data plotted as Hanes-Woolf plots (Fig. 4B) yielded straight lines as a function of [Na+]i (>90 mM), suggesting the exis-
tence of another Na+-binding site accessible from the cytoplasmic side. Furthermore, the results indicated that ApH decreased only the K,,, for cytoplasmic Na+ without a significant change of V,,,,,. To analyze the ApH effect on the K, for transport Na+ and the V,,,, kinetic parameters at constant pH, and A$ were replotted as a function of ApH imposed (Fig. 5). It is apparent that the K, for transport Na+ decreased with increases in ApH, but the change in V,,, was quite small. These results support our previous observations that V,,,,, of the ApHdependent Na+ efflux is determined primarily by the extent of protonation of the external H+-dissociable group, and that the ApH-modulated pKc, shift in the cytoplasmic regulatory H+-binding site is responsible for the affinity change at the transport Na+ site (12). It could be argued that the ApH-dependent apparent K, shift for Na+ resulted from a simple competition between internal Na+ and H+ for binding to a site on the transporter, because ApH elevations increase pHi at constant pH,. However, this may not explain the fact that ApH decreases both internal K,, for H+ and K,,, for Na+ at their binding sites simultaneously. Membrane Potential
Increases the V,, for Na+ Eflux
To evaluate the underlying mechanism of A&dependent gating of the Na+/H+ exchange, we studied the effect of elevated A4 on the Na+ efflux kinetics. Figure 6A shows the VNa’s as a function of [Na+]i in the presence of ApH = 2 at a fixed pH, of 5.0. VNa showed a saturation profile under all A4 conditions. It was also noted that in the low [Na+]i range the V,, was slightly inhib-
Na+/H+-ANTIPORTER
IN H&bacterium
17
halobium
i
APH FIG. 5. plotted
Effects of ApH on the kinetic parameters. Kinetic parameters obtained by varying as a function of ApH imposed. Conditions were the same as those described in Fig. 4.
ited regardless of the presence or absence of a large A$, as shown in Fig. 4A, indicating that the Na+ binding to the external Na+ site is essential for activation of the antiporter by A&+ (or A~#J) as mentioned above. The [Na+]i dependence of VN, in the presence of various A&s was analyzed by a Hanes-Woolf plot (Fig. 6B). The results again indicated the binding of Na+ to a single transport site. Increased A4 decreased the slopes of the lines, reflecting an increased V,,,, but the K, for Naf remained essentially constant as long as pH, and ApH were fixed. Kinetic parameters obtained were plotted as a function of A4 imposed (Fig. 7). At any pHo’s, V,,,
0
2al
FIG. 6.
(mM)
A4 (140 mV) and pH, were
sharply increased at around 100 mV of A+, which is identical to the threshold A4 previously reported (10-12). Although we cannot exclude the possibility that external K+ itself inhibits Na+/H+ exchange, independent of its effects on A& such a possibility may be small because the Na+/H+ exchange activity could be observed by illumination in the Naf-loaded vesicles suspended either in 3 M NaCl or in 3 M KC1 medium (7,8). It is also notable that the K,,,‘s at the A&s below 100 mV were completely different from those obtained above at the larger AI$‘s, although the Km gap became unclear at acidic pH, (pH, 4.0). Since the apparent Km for Na+ is also dependent on
-m
ml
INa’Ii”
ApH at a specified
0
603
[ N?Iin
EM)
Dependence of A&-induced Na+ efflux on cytoplasmic Na+ concentration. Vesicles containing 30-480 mM xxNaC1 (replaced isoosmotitally with KCl) and a combination of 20 mM each of Mes, Hepes, Pipes, Eppes (pH 7.0) were diluted in the dilution medium (pH 5.0) at various K+ concentrations (replaced isoosmotically with choline-Cl) to generate various A#. (A) Initial rate of Na+ extrusion (V,.) vs cytoplasmic Na+ concentration [Na+h. (B) Hanes-Woolf plot ([Na+]JV,, vs [Na+]J.
18
MURAKAMI
300
AND
KONISHI
Km
I 200,
WHO= 4.0
p\=4.0
E
G
100
0
-i
i
py= 5.0
pH,=5.0
100
3
(mV) FIG. 7. Effects of A@ on kinetic were plotted against A+ imposed.
parameters
of Na+ efflux. Kinetic
parameters
pH, (acidic pH, increases K,) in the presence of A4 (Fig. 5A), the A&induced K, shift is possibly compensated by the pH, effect at pH, 4.0. In addition, Km in the absence of A$ increased more with increased ApH (not shown). These results indicate that two distinct states such as resting and gated conformational states are possible with or without threshold A& Since the Km remained constant above the threshold A+, it is suggested that A4 plays two distinct roles in the antiporter activation. First is the gating in which the A$ of approx 100 mV induces a conformational transition of the antiporter from resting to a ApH-reactive active state (gated state). A subsequent effect of A4 is as a driving force to facilitate ApH-dependent Na+/H+ exchange, for example, by modulating the number of active intermediates without changing the Na+ affinity for the cytoplasmic transport site as reflected in the increased V,,, with a constant Km above the threshold A&
were obtained by elevating
always larger than unity. These results strongly support the aforementioned idea that there are high- and lowaffinity Na+-binding sites and the Na+ binding to the high-affinity (external) site is a prerequisite for the A4
TABLE
Hill coefficient
LOW [Na+]i
Dilution medium
PH,
ApH
A$ (mV)
3 M Choline-Cl
4.0
4 3 2 1 3 2 2 2 2 2
-140 -140 -140 -140 -140 -140 -125 -110 -60
1 2
-140 -140
Site
The Na+ efflux data were further analyzed by Hill plotting, which is routinely used to characterize ligand binding to the receptor site (21). Under all conditions the lines obtained broke at around 90 mM [Na+]i except under the 3 M NaCl condition and in the absence of A& Hill coefficients calibrated from the least-squares fitting of the slopes above and below the breaking point are summarized in Table I. Above 90 mM [Na+]i, the Hill coefficients are almost unity under all conditions tested and are the same as that under the 3 M NaCl condition and that in the absence of A4, but below 90 mM [Na+]i where the [Naflo effect is significant, the coefficients are
I
Hill Coefficients Obtained under Several Na+ Efflux Conditions
5.0
Hill Analysis of the Na’-Binding
A# at a specified ApH (2 units) and
3 M NaCl
5.0
(lOOmV
ApH -A’4
coOperati\
“A W Sensing Vahre”
FIG. 8. Model for the cooperative regulation of the Na+/H+-antiporter by ApH and A& (See details under Discussion).
is difficult to conclude that A$ sequentially affects a certain intermediate with a fixed K, that is formed as a consequence of the effect of the ApH and increases its amount, because there is no additivity of the ApH and A+ effects on V,,. It is thus considered that A4 acts at the H+-binding and releasing steps at the regulatory H+ site as a force such as an electrical field, to increase the turnover of the process cooperatively with ApH. Model of Cooperative Regulation
by ApH and A4
Our present and earlier studies on the mechanism of regulation (10-12) tend to rule out purely kinetic models of Na+/H+ exchange activity. Although the precise mechanism of Na+ and H+ transport and whether the antiporter is a single peptide remain unclear, the ApHA$ cooperativity in antiporter activation is reasonably explained in the following model, as shown in Fig. 8. The initial step for antiporter activation is the binding of Na+ and Hf to the corresponding external binding sites. The H+-Na+ bound intermediate is then activated by A@ of approx 100 mV to the gated state and becomes reactive to A,&+. Voltage gating phenomena have been studied in detail in a simpler channel peptide such as alamethicin (22) or porin (23), where the results suggest that gating arises from the changes in the oligomeric structure in response
20
MURAKAMI
to an applied electrical field. Such a possibility may also exist in the halobacterial Na+/H+-antiporter. The subsequent Ajin+-coupled Na’/H+-exchange process is more likely mechanochemical, where the chemical process (H+ binding) is transduced directly into the force for Na+ translocation in the presence of A& probably through a protein conformational change. In order to explain the cooperative feature of ApH and AC#J in Na+ efflux stimulation, we suggest that a peptide segment carring the regulatory H+-binding site is a “A&sensing valve.” The unprotonated conformation restricts the accessibility of the cytoplasmic Na+ to the transport site even if Ad is present. Once the site is protonated, the valve becomes A&sensitive, and the A&driven conformational transition would be expected to expose the transport Na+ site. The Hf-binding site exposed to the medium would release H+ immediately, probably because of its inherently high pK,. It then becomes A& insensitive and returns to its original conformation to expel the bound Na+ to the external medium. In this model, the turnover rate of the protonationdeprotonation of the regulatory H+ site determines the Na+ efflux. Hence, the rate of H+ supply to the site is rate-limiting on the Na+ efflux as long as Ac$is constant, while the A#+assisted conformational change of the valve is rate-limiting when the rate of H+ supply is constant. As we showed previously (12), the pK, change at the regulatory H+ site gives rise to a pK, difference across the membrane because the pK, at the external H+ site (pK, 4.6) remains constant under the conditions. The pKa difference would be a direct driving force for the transmembranous supply of H+ to the regulatory H+ site, thus acting as a modulator of the turnover rate of the “A&sensing flip-flop valve” under a constant electric field. This is actually reflected in the observed ApHdependent pK, shift at the regulatory H+ site and the Km change at the Na+-transport site. It is difficult to attribute the observed pK,, shift which is not saturable over 23 units (12) to a single H+-dissociable group. Such a large pK, change could be explained only as an increased turnover of the protonation-deprotonation process at the Hf-binding site. This could also explain our observation that Ac$ does not affect the K,,, but only the V,,,,,. Recently, Na+/H+-antiporters have been extensively studied in nonhalophilic cells and several molecular features were clarified (24-30). It is of interest to explore the molecular similarity between the halophilic and the nonhalophilic antiporters, because their kinetic and structural properties are quite different from those that we revealed in the halobacterial Na+/H+-antiporter. It is probable that the halobacteria modified its antiporter into a sophisticated device for Na+ extrusion to adapt to the extremely high saline environment. The mechanism we proposed may enable the halobacterial antiporter to expel cytoplasmic Na+ against a large Na+ gradient and function as a Ajin+-driven Na+ pump.
AND
KONISHI
ACKNOWLEDGMENTS We thank Dr. Lester Packer for his helpful comments in revising the manuscript. This study was supported by a grant-in-aid for scientific research in the priority area of “Bioenergetics” to T.K. from the Ministry of Education, Science and Culture of Japan.
REFERENCES 1. Oesterhelt,
D., and Stoeckenius,
W. (1973) Proc. Natl. Acad. Sci.
USA. 70,2853-2857. 2. Schobert, B., and Lanyi, J. K. (1982) J. Biol. Chem. 257,10,30610,313. 3. Skulachev, V. P. (1978) FEBS L&t. 87,71-179. 4. Lanyi, J. K. (1978) Microbial. Rev. 42,682-706. 5. Lanyi, J. K., and MacDonald, R. E. (1976) Biochemistry 16,46084614. 6. Eisenbach, M., Cooper, B. S., Garty, H., Johnstone, R. M., Rottenberg, H., and Caplan, S. R. (1977) Biochim. Biophys. Acta 465, 599-613. 7. Murakami, N., and Konishi, T. (1985) J. B&hem. (Tokyo) 98,
897-907. 8. Murakami, N., and Konishi, 9. Murakami, N., and Konishi, 10. Murakami, N., and Konishi, 231-236. 11. Konishi, T., and Murakami, N., and Konishi, 12. Murakami,
T. (1987) B&hem. Znt. 15,215-221. T. (1988) Biochimie 70.819-826. T. (1988) J. Biochem. (Tokyo) 103,
N. (1988) FEBS L.&t. 226,270-274. T. (1989) Arch. Biochem. Biophys.
271,515-523. 13. Bassilana, M., Damiano, E., and Leblanc, G. (1984) Biochemistry 23.5288-5294. 14. Nakamura, T., Hsu, C., and Rosen, B. P. (1986) J. Biol. Chem.
261,678-683. 15. Krulwich, 16. Grinstein,
T. A. (1983) Biochim. Biophys. Acta 726,245-264. S., Goety, J. D., and Rothstein, A. (1984) J. Gen. Phys-
iol. 84,585-600. 17. Aronson, P. S. (1985) Annu. Rev. Physiol. 47,545-560. 18. Ehrenfeld, J., Cragoe, E. J., and Harvey, B. J. (1987) Pfluegers
Arch. 409,200-207. 19. Lanyi, J. K. (1969) J. Biol. Chem. 244,4168-4173. 20. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 49,105-121. 21. Changeux, J. P., Thiery, J., Tung, Y., and Kittel, C. (1967) Proc.
Natl. Acad. Sci. USA 57,335-341. 22. Fox, R. O., and Richards, F. M. (1982) Nature (London) 300,325330. 23. Mauro, A., Blake, M., and Labarca, P. (1988) Proc. Natl. Acad. Sci. USA 85,1071-1075. S., Rothstein, A., Sarkadi, B., and Gelfand, E. W. 24. Grinstein, (1984) Amer. J. Physiol. 246, C204-C215. 25. Aronson, P. S., and Igarashi, P. (1986) in Current Topics in Membranes and Transport (Aronson, P. S., and Boron, W. F., Eds.), Vol. 26, pp. 57-75, Academic Press, San Diego. 26. Nord, E. P., Goldfarb, D., Mikhail, N., Moradeshagi, P., Hafezi, A., Vaystub, S., Gragoe, E. J., Jr., and Fine, L. G. (1986) Amer. J. Physiol. 250, F539-550. 27. Grinstein, S. (Ed.) (1988) Na+/H+ Exchange, CRC Press, Boca Raton, FL. 28. Huot, S. J., Cassel, D., Igarashi, P., Cragoe, E. J., Jr., and Slayman, C. W. (1989) J. Biol. Chem. 264,683-686. 29. Padan, E., Maisler, N., Taglicht, D., Karpel, R., and Schuldiner, S. (1989) J. Biol. Chem. 264,20,297-20,302. 30. Sardet, C., Franchi, A., and Pouyssegur, J. (1989) Cell 56, 271280.
