INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

The ADP/ATP carrier (AAC) is the most abundantly occurring transporter of the inner mitochondrial membrane and is responsible for the membrane potential-driven exchange of ATP versus ADP between the matrix space and the cytosol (Klingenberg, 1972, 1980). This process is the last step of oxidative phosphorylation. The function and the electrogenicity of the carrier have been investigated in detail mainly by flux measurements and binding studies on mitochondria, submitochondrial particles, and purified AAC reconstituted into liposomes (Klingenberg, 1985). By these techniques only steady-state transport studies were possible.

More recently we succeeded in measuring directly the currents generated by the ATP and ADP transport (Brustovetsky et al., 1996, 1997). Proteoliposomes were adsorbed to a planar lipid bilayer, and after a light-induced concentration jump of ATP (ADP) from the corresponding caged nucleotide, the resulting currents were recorded. These experiments directly suggested the electrogenicity of transport by the AAC. The six possible transport modes were investigated: ADP_ex-ATP_in, ATP_ex-ADP_in, ADP_ex-ADP_in, ATP_ex- ATP_in, ATP_ex-0_in, and ADP_ex-0_in. In the ATP_ex-ADP_in heteroexchange mode, net negative charge is transported into the liposomes, and in the ADP_ex-ATP_in mode it is transported out of the liposomes. However, transient currents could also be measured in the homoexchange modes (ADP_ex-ADP_in, ATP_ex-ATP_in) and with unloaded liposomes (ATP_ex-0_in, ADP_ex-0_in).

All of the electrical measurements made so far have been carried out by the adsorption of proteoliposomes to a planar lipid bilayer and applying nucleotide concentration jumps by photolysis of a caged substrate with a high-pressure mercury lamp (pulse duration 125 ms). Yet, there were two main points that could not be answered. First, which part of the electrical signal is due to the electrogenic transport of a nucleotide by the AAC and which part is due to the electrogenic release of the nucleotide from the caged nucleotide in the binding site? Second, what are the respective kinetics of the electrogenic ATP and ADP translocations?

To answer the first question we used a rapid solution exchange technique based on solid supported membranes that allows concentration jump experiments to be carried out without caged substrates. The second question could be solved by using a UV laser pulse (pulse duration 10 ns) instead of a UV lamp for the release of the caged nucleotides, so that the time resolution could be improved down to ~500 µs. To investigate the steps of ATP and ADP transport in detail with an experimental system that is as simple as possible, we used unloaded liposomes (ATP_ex-0_in and ADP_ex-0_in). This system allows one to study a half-cycle of the ATP/ADP exchange. For the first time it is possible to give an estimation of the reaction rates of the electrogenic transport of ATP and ADP by the AAC. A transport model based on these results is proposed.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

BLM experiments

Electrical currents of the AAC were measured by adsorption of AAC-containing proteoliposomes to a black lipid membrane (BLM) as described elsewhere (Brustovetsky et al., 1996). The black lipid membranes, with an area of 1.3 mm², were formed in a Teflon cell filled with an electrolyte containing 100 mM NaCl and 20 mM 2-(N-morpholino)ethanesulfonic acid (MES) (pH 6.2) (if not otherwise stated). The high buffer concentration was used to avoid the fast protonation of the bilayer, which can also yield a transient current. Each compartment of the Teflon cell has a volume of 1.5 ml. The membrane-forming solution contained 1.5% (wt/vol) diphytanoylphosphatidylcholine and 0.025% (wt/vol) octadecylamine in n-decane to obtain a positively charged membrane surface (Dancshazy and Karvaly, 1976). Membrane formation was controlled by eye, and the capacitance of the membrane was determined (for further details see Bamberg et al., 1979). The temperature was kept at 24°C, and because there is virtually no UV light absorption by the BLM and the liposomes, the local temperature is not influenced by the laser flash.

The purified AAC from bovine heart mitochondria was reconstituted into liposomes (95% phosphatidylcholine, 5% cardiolipin) with a final protein concentration of 0.4 mg/ml and a protein-to-lipid weight ratio of 0.02 (Krämer, 1986; Brustovetsky et al., 1996, 1997). Fifteen microliters of the proteoliposome-containing suspension was added to the compartment of the cuvette opposite the light source. Under stirring, the adsorption of the liposomes to the BLM took ~90 min. Then caged ATP or caged ADP (P³-1-(2-nitrophenyl)ethyl ester of ATP or ADP) was added to the suspension.

The caged nucleotides were irradiated by a UV light pulse of an excimer laser (wavelength 308 nm, pulse duration 10 ns), which was focused on the BLM. Approximately 0.8 mm² of the BLM was exposed to the laser beam. The average radiant exposure on the BLM surface was 150 mJ/cm², which yields an  (fraction of nucleotide released from caged analog) of 26%. As was shown in earlier publications, no significant temperature change (T < 1°C) that could influence the current signal is caused by the illumination of the BLM (Knoll and Stark, 1977). This is supported by the fact that in the experiments presented in this publication, at varying radiant exposures in the range of one order of magnitude, no effect on the kinetics of the carrier was observed.

The transient current generated by the AAC was amplified (10⁹-fold), filtered (1 kHz with a first-order low-pass filter), and recorded with a digital oscilloscope. Because of the capacitive coupling of the AAC to the electrical circuit, an additional exponential function with a system time constant ₀ is expected in the signal (Bamberg et al., 1979; Borlinghaus et al., 1987): ₀ = (C_m + C_p)/(G_m + G_p) (C_m/p = specific capacitance of the BLM/liposomes, G_m/p = specific conductivity of the BLM/liposomes).

The fraction of released ATP was determined with a luciferin-luciferase assay as described previously (Nagel et al., 1987). For the fraction of released ADP from caged ADP, ADP was first converted to ATP by creatine phosphokinase and phosphocreatine, and then the ATP concentration was determined by the luciferin-luciferase assay. For both caged ATP and caged ADP the amount of released nucleotide was the same for a given radiant exposure.

The time constants for the ATP and ADP release were measured spectroscopically as described by Walker et al. (1988). For this purpose the formation and decay of an aci-nitro compound during the release of the caged nucleotide were observed at a wavelength of 406 nm. The values of the time constants at different pH and Mg²⁺ concentrations are given in Tables 1 and 2; these are in the same range as the values found by Walker et al. for the release of ATP. The release rates depend on the Mg²⁺ concentration and on the pH. At high Mg²⁺ concentration the release of the nucleotides was slower. This effect was more pronounced in the case of ATP, probably because of the higher binding affinity for Mg²⁺ (pK 4.06 (ATP), pK 3.17 (ADP); Martell and Smith, 1974). At varying pH the release rate was increased by a factor of 10^pH, which is in agreement with theory (Walker et al., 1988). The activation energies for the release of ATP and ADP were 60 kJ/mol and 57 kJ/mol, respectively.

Fast solution exchange on a solid supported membrane

In addition to the bilayer experiments, we applied a fast fluid injection technique, which allows the measurement of electrical currents generated by the AAC, using a solid supported membrane (SSM) (Pintschovius and Fendler, 1999).

The advantage of this technique is that concentration jumps of an arbitrary substrate (i.e., ADP or ATP) can be performed with a time resolution of up to 10 ms without the use of caged substrates.

Solid supported membranes consist of an alkanethiol monolayer on a gold electrode combined with an additional layer of phospholipid on top. The resulting double layer can be used in a way similar to that of the BLM. The electrical properties of the SSM and the BLM are similar. The capacitance of the SSM was typically 400 nF/cm², and the conductance 100 nS/cm². The high conductivity of the SSM compared to the BLM (3 nS/cm²) is probably due to defect structures on the large surface of the SSM (4-5 mm²). The SSM is positioned in a plexiglass cuvette with an inner volume of 17 µl, which allows a fast exchange of the solutions.

Furthermore, the concentration rise over the SSM had to be taken into account. The nucleotide concentrations were corrected as described by Pintschovius and Fendler (1999), so that the resulting concentration was c_korr.  0.5c₀.

The AAC-containing proteoliposomes were adsorbed to the SSM. The carrier was activated by a rapid solution exchange over the SSM. The rapid solution exchange was performed by driving two syringes with a perfusor pump. The two syringes contained the activating and nonactivating solutions, and the fluid flow was controlled by an electrical valve. The resulting electrical current was amplified by an operational amplifier and a low-noise amplifier 10⁹-10¹⁰-fold, filtered (300 Hz, low-pass), and recorded. The carrier was activated via solution exchange for 2 s from t = 6 s to t = 8 s. The short peaks at t = 6 s and t = 8 s are artefacts and are caused by the turning on and off of the electrical valve. At t < 6 s and t > 8 s nonactivating solution was flowing through the mixing chamber.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

Unloaded liposomes (ATP_ex-0_in, ADP_ex-0_in) were used to investigate half-cycles of the ATP/ADP exchange. Two techniques were applied to perform ATP or ADP concentration jumps: a fast solution exchange on an SSM and the photolytic release of nucleotides with the BLM method. The purpose of the SSM measurements was 1) to determine in a control experiment whether ATP and ADP translocation or binding, respectively, are electrogenic and to investigate the influence of the photolytic release of these nucleotides; 2) to investigate the behavior of the AAC on addition and removal of the nucleotides (on-response and off-response); and 3) to compare with the results obtained by the BLM experiments. The BLM technique was used to investigate the kinetic properties of the AAC by laser-induced nucleotide concentration jumps.

Current measurements on solid supported membranes

ADP and ATP Transport and inhibition

View larger version (30K): [in this window] [in a new window]   FIGURE 1   SSM experiment: ADP and ATP concentration jump and inhibition of the AAC transport activity. (A) Activating solution: 100 mM NaCl, 20 mM MES (pH 6.2), 100 µM ADP; nonactivating solution: 100 mM NaCl, 20 mM MES (pH 6.2). (B) Activating solution: 100 mM NaCl, 20 mM MES (pH 6.2), 100 µM ADP, 1 µM CAT, 1 µM BKA; nonactivating solution: 100 mM NaCl, 20 mM MES (pH 6.2). (C) Activating solution: 100 mM NaCl, 20 mM MES (pH 6.2), 100 µM ATP; nonactivating solution: 100 mM NaCl, 20 mM MES (pH 6.2). (D) Activating solution: 100 mM NaCl, 20 mM MES (pH 6.2), 100 µM ATP, 4 mM MgCl2; nonactivating solution: 100 mM NaCl, 20 mM MES (pH 6.2), 4 mM MgCl2. T = 24°C.

Transported charge

Dependence on the ATP concentration

View larger version (13K): [in this window] [in a new window]   FIGURE 2   SSM experiment: peak current of an ATP on-response at increasing ATP concentration; for conditions see Fig. 1. The ATP concentration is corrected according to Pintschovius et al. (1999). The half-saturating concentration is 60 µM.

Current measurements on black lipid membranes

In the BLM experiments the transport through AAC is activated via photolysis of caged ATP or caged ADP. With this method ~20-30% of ATP or ADP is released with a time constant of ~1 ms (pH 6.2) after the laser flash (Tables 1 and 2).

ADP and ATP transport and inhibition by Mg²⁺

View larger version (20K): [in this window] [in a new window]   FIGURE 3   BLM experiment. (A) 100 mM NaCl, 20 mM MES (pH 6.2). Upper trace: 400 µM caged ATP; time constants: 1 = 1.0 ms, 2 = 6.4 ms, 3 = 30 ms. Lower trace: 400 µM caged ATP, 2 mM MgCl2. (B) 100 mM NaCl, 20 mM MES (pH 6.2). Upper trace: 200 µM caged ADP; time constants: 1 = 1.8 ms, 2 = 11 ms. Lower trace: 200 µM caged ADP, 4 mM MgCl2; T = 24°C; fraction of released nucleotide () = 26%. frp, fast-release phase; srp, slow-release phase; tp, transport phase. (C and D) Inhibition of the carrier with 5 µM CAT and BKA.

pH dependence

Dependence on ATP and ADP concentration

View larger version (28K): [in this window] [in a new window]   FIGURE 4   BLM experiment: nucleotide dependence of the peak current and the time constants. Conditions: 100 mM NaCl, 20 mM MES (pH 6.2), T = 24°C. , Caged nucleotide concentration increased from 5 µM to 2 mM,  = 26%; , caged nucleotide 2 mM,  decreased from 30% to 0%. (A) ATP dependence of the peak current. (B) ATP dependence of 2 (decay of transport phase). (C) ATP dependence of 3 (decay of transport-phase). (D) ADP dependence of the peak current. (E) ADP dependence of 1 (decay of transport phase + slow-release phase). (F) ADP dependence of 2 (decay of transport phase + slow-release phase).

View larger version (21K): [in this window] [in a new window]   FIGURE 5   BLM experiment (see Fig. 4): corresponding amplitudes of i for ATPex-0in (A and B) and ADPex-0in (C and D).

Activation energy of the reactions

View larger version (26K): [in this window] [in a new window]   FIGURE 6   BLM experiments: activation energies of 2, 3 for ATP (A: 500 µM caged ATP) and 1, 2 for ADP (B: 500 µM caged ADP) concentration jumps. Conditions: 100 mM NaCl, 20 mM MES (pH 6.2); temperature 19-34°C,  = 26%.

Inhibition of the AAC by high anion concentration

View larger version (19K): [in this window] [in a new window]   FIGURE 7   BLM experiment: inhibition of an ADPex-0in signal with MgCl2. (A) Transient current traces. (B) Peak current of transport phase + slow-release phase at different MgCl2 concentrations; 100 mM NaCl, 20 mM MES (pH 6.2), 200 µM caged ADP; T = 24°C;  = 26%.

View larger version (32K): [in this window] [in a new window]   FIGURE 8   BLM experiment: Inhibition of the nucleotide transport at high chloride concentrations. (A) ADPex-0in: 200 µM caged ADP. (B) ATPex-0in: 100 µM ATP. (Inset) Larger time scale; 100 mM NaCl, 20 mM MES (pH 6.2); T = 24°C;  = 26%.

View larger version (15K): [in this window] [in a new window]   FIGURE 9   BLM experiments: normalized peak current of the transport phase + slow-release phase of an ADPex-0in signal at different chloride and gluconate concentrations. Upper trace: 100-300 mM Na gluconate, 20 mM MES (pH 6.2), 200 µM caged ADP; T = 24°C;  = 26%. Lower trace: 100-300 mM NaCl, 20 mM MES (pH 6.2), 200 µM caged ADP; T = 24°C;  = 26%.

Loading of empty liposomes with ADP and ATP

View larger version (35K): [in this window] [in a new window]   FIGURE 10   BLM experiments: loading of empty liposomes with ADP and ATP. 100 mM NaCl, 20 mM MES (pH 6.2); T = 24°C. (A) Lower trace: 500 µM caged ATP; upper trace: 5th flash after the addition of 5 units/ml hexokinase and 1 mM glucose. (B) Upper trace: 500 µM caged ADP; lower trace: 5th flash after the addition of 10 units/ml creatine phosphokinase and 1 mM phosphocreatine. (Inset) Schematic model of the loading process (for an explanation see the Discussion).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

ATP and ADP translocation is electrogenic

In previous BLM experiments transient currents induced by ATP and ADP concentration jumps were reported. However, in the case of the ATP_ex-0_in and ADP_ex-0_in transport modes it was not quite clear whether electrogenic binding, electrogenic transport, or both are observed in the transient currents. Now we have further evidence for the electrogenic transport of ATP and ADP.

The experiments shown in Fig. 10 strongly support the proposed ATP and ADP uniport in the case of unloaded liposomes. The transport of nucleotides into the liposomes could be proved on the addition of hexokinase + glucose (ATP_ex-0_in) and creatine phosphokinase + phosphocreatine (ADP_ex-0_in), respectively, to the solution. The effect of hexokinase was already described for the reloading of ADP-loaded liposomes in the ATP_ex-ADP_in exchange mode (Brustovetsky et al., 1997).

Here, in the case of initially unloaded liposomes, the shape of the transient current changes on the addition of hexokinase and glucose from a simple ATP_ex-0_in signal to an ATP_ex-ADP_in signal. The prerequisite for the observation of this heteroexchange signal is the loading of the liposomes with ADP, which in turn requires an initial loading with ATP. This loading with ADP can be explained as follows (see scheme in Fig. 10). After its release from the caged analog, ATP is transported into the liposomes. Because of the low concentration of hexokinase and glucose, the ATP in the solution is converted into ADP relatively slowly, i.e., over the next few minutes. Then the generated ADP in the solution can exchange with the ATP that was translocated into the liposomes, so that the liposomes become loaded with ADP. Now the ATP that was transported out of the liposomes is also converted into ADP, so that after a while there is only ADP present. With the next flash ATP is released, and it can exchange with the ADP inside the liposomes. Consequently, only by assuming that ATP is initially transported into the liposomes in a uniport mode can the loading of the liposomes with ADP in the presence of hexokinase and glucose be explained. In a similar way a net transport of ADP into empty liposomes was shown with creatine phosphokinase and phosphocreatine (Fig. 10 B). Taking into account the results from the SSM experiments (Fig. 1, A and C), an electrogenic ATP and ADP transport is postulated.

An estimation of the internal nucleotide concentration after several flashes yielded a value of 10 µM. This amount of internal ATP or ADP is sufficient to supply the carrier with nucleotides for approximately one turnover. As can be seen in Fig. 10, the amount of transported charge after several flashes is about twice as much as in the completely unloaded state of the liposomes.

For the BLM experiments it is desirable that the liposomes are again nucleotide free before the next flash. This can be accomplished by the following mechanism. After the laser flash only a small fraction of the total caged nucleotide is converted into free nucleotide in the cuvette (~1%). Stirring between the flashes therefore causes a 100-fold dilution of the nucleotide in the vicinity of the membrane (Nagel et al., 1987). Because of the low nucleotide concentration in the solution, the nucleotides that were transported into the liposomes are transported out of the liposomes, yielding virtually nucleotide-free liposomes before the next flash. The maximum amount of internal ATP or ADP after several flashes is in the range of 10 µM (see above). Except for the experiment depicted in Fig. 10, all experiments were carried out in the absence of hexokinase or creatine phosphokinase. Consequently, after several flashes the liposomes were loaded with small amounts of the nucleotide that was released from its caged analog. Because the homoexchange modes (ATP_ex-ATP_in, ADP_ex-ADP_in) yield exactly identical results in terms of kinetics compared to the uniport modes (Brustovetsky et al., 1996, and unpublished data), this loading effect is negligible.

The depletion of internal nucleotide after the flash is experimentally supported by the SSM experiments. As shown in Fig. 1, an on-response as well as an off-response is seen because of electrogenic ATP and ADP transport, respectively, whereas the sign of the transported charge was directly opposite in the two cases. The amount of charge that is transported during the on- and off-responses is approximately the same (off-response  80-90% of the on-response). Calculations of the ATP or ADP concentration inside the liposomes after one half-cycle of transport activity yielded a value of ~100 µM. Consequently, the internal nucleotide concentration is sufficient for the rebinding and transport of nucleotides out of the liposomes after removal of the nucleotides outside.

The on-response and off-response of the signals (ATP_ex-0_in and ADP_ex-0_in) could be blocked by the specific inhibitors CAT and BKA (Fig. 1 B). The sign of the transient current corresponds to the transport of negative charge into the liposomes for the ATP on-response (Fig. 1 C) and of net positive charge into the liposomes for the ADP on-response (Fig. 1 A).

In connection with the transported charge, the question arises if only the protonated nucleotides (ATPH³, ADPH²), the unprotonated nucleotides (ATP⁴, ADP³), or both are transported by the AAC. From previous investigations it was suggested that the unprotonated species are exclusively transported, because the pH dependence of the ADP/ATP exchange is similar to the pH dependence of the fraction of deprotonated nucleotides (Pfaff and Klingenberg, 1968; Brustovetsky et al., 1997). In addition, the absence of H⁺ movement accompanying ATP/ADP exchange supports the strict adherence to ATP⁴ and ADP³ (Wulf et al., 1978).

Further experimental evidence in support of these findings is given by the comparison of transported charge at different pH values in our investigations. Because of the pK of ATP (pK 6.51) and ADP (pK 6.41) for the binding of the first proton, one would expect a relative difference in the ratio of transported charge (Q_ATP/Q_ADP) in the two uniport modes at pH 6.2 and pH 7.4, which are below and above the pK of the nucleotides. In the SSM experiments we found that within the same experiment (i.e., the same SSM), in the ATP translocation step approximately twice as much charge is transported compared to the ADP translocation (Q_ATP/Q_ADP = 2/1), independent of the pH (pH 6.2 and pH 7.4). Considering the fact that this ratio is not dependent on the pH, we assume that only the unprotonated nucleotides ATP⁴ and ADP³ are transported. Under this assumption it is concluded that an equivalent of 3.3 countercharges resides in the nucleotide binding site and is cotransported with the nucleotides. This is in agreement with Brustovetsky et al. (1996), who proposed that an equivalent of 3.5 positive countercharges is cotransported with the nucleotides.

Assignment of the transient current phases

The translocation of ATP by the AAC can be blocked by addition of Mg²⁺ (Fig. 1), because MgATP is not transported by the carrier. This allows the differentiation between translocation and other charge movements for the transient signal of the BLM experiments (Fig. 3, A and B). In the presence of Mg²⁺ a charge movement that is consequently not due to the transport of nucleotides can clearly be identified. Therefore the difference in the signal with and without Mg²⁺ corresponds to the transport of the nucleotides (transport phase). The sign of the current in the transport phase is in agreement with the direction of the current in the SSM experiment. In the case of ATP_ex-0_in, the transport phase represents the translocation of negative charge into the liposomes and in the case of ADP_ex-0_in of positive charge into the liposomes.

On inhibition of the transport phase by Mg²⁺, two phases remain unaffected in both cases (ATP_ex-0_in, ADP_ex-0_in), namely the fast-release phase and the slow-release phase. The fast-release phase corresponds to a negative charge movement toward the inside of the liposomes or a positive charge movement in the opposite direction. It is too fast to be resolved. It might be due to the release of a proton (k > 10⁵ s¹) in the photolytic nucleotide release reaction (Walker et al., 1988). The slow-release phase corresponds to the movement of positive charge toward the binding site, and the amount of transported charge is equal to or smaller than that in the fast-release phase. From the pH dependence of _1,Mg²⁺ (Tables 1 and 2) and the comparison with spectroscopic data it can be concluded that _1,Mg²⁺ represents the release of the nucleotides, which is accompanied by a charge movement. In addition, the local pH in the binding site has to be similar to the pH in the solution, because the release of the nucleotides is proposed to take place in the binding site (see below), and the time constant for nucleotide release at pH 6.2 is equal to the time constant for nucleotide release in the bulk solution. This fact suggests that with respect to H⁺ the binding site is in free exchange with the bulk phase. Because the pK of the fixed charges of the carrier is not known, it is not clear whether there is an effect on the local pH from this side.

After inhibition with CAT/BKA, both the fast-release phase and the slow-release phase are blocked, and only the laser artefact remains (Fig. 3 C and D). CAT binds selectively to the AAC in the cytosolic state (Fig. 11 A) and cannot permeate the membrane. BKA, however, can permeate the membrane and binds to the matrix state of the carrier (Weidemann et al., 1970; Klingenberg, 1985). CAT and BKA block the corresponding binding site, so that nucleotides or caged nucleotides can no longer bind.

View larger version (42K): [in this window] [in a new window]   FIGURE 11   (A) Inhibition by CAT and BKA. BKA can permeate the membrane and inhibits the carrier in the matrix state (m-state), whereas CAT cannot permeate the membrane and inhibits the carrier in the cytosolic state (c-state). The two states symbolize the orientation of the binding site with respect to the matrix space and the cytosol in the native system. (B) Release-in-site mechanism for ATP (ADP is analogous). (1) binding of caged ATP, (2) laser flash, (3) release of ATP from prebound caged ATP in the binding site, (4) conformational transition of the carrier, (5) release of ATP into the liposome. (C) Schematic model of ADP and ATP translocation in the BLM experiments. *, caged nucleotide that has absorbed a photon.

At high chloride concentration (see Fig. 8) the fast-release phase can be blocked in both cases (ATP_ex-0_in, ADP_ex-0_in), and the slow-release phase only in the case of ADP_ex-0_in. The reason why the slow-release phase in the case of ATP_ex-0_in is not inhibited is unclear. A tentative explanation could be that in the binding site the Cl activity is reduced because of the presence of a negative charge due to ATP binding.

By applying concentration jumps with different ATP concentrations, a half-saturating value of 60 µM was found in the SSM experiment (Fig. 2). However, according to our assumption that only the unprotonated fraction of ATP (~30%, pK 6.51; Martell and Smith, 1974) is transported by the carrier, the half-saturation concentration for ATP has to be corrected to ~20 µM. This is still a relatively high value compared to literature values (3 µM; Klingenberg, 1976) and to BLM measurements with nucleotide-loaded liposomes (unpublished results), which yield values in the range of K_0.5  10 µM for the ATP binding. The difference in the affinity might be due to different experimental conditions, for example, the absence of nucleotides inside the liposomes in the steady-state experiments.

Transport model

For the interpretation of the data it has to be assumed that caged ATP binds to the carrier, which is in line with previous studies (Brustovetsky et al., 1996). On the basis of our results and referring to Fig. 11 C, we propose the following model for nucleotide binding and transport by the AAC: caged ATP C_o  caged ATP* C_o  ATP C_o  C_i + ATP. Caged ATP is bound to the carrier, and according to the data the dissociation rate is probably slow. After the laser flash, ATP is released from the prebound caged ATP in the binding site and is transported through the carrier without exchanging with the solution. For ADP the same mechanism is applicable. We call this mechanism release in site.

In Fig. 11 C the reaction pathways for ADP and ATP translocation in the case of the BLM experiment are schematically depicted. It is similar to a model already proposed for the NaK-ATPase (Fendler et al., 1994). However, for the NaK-ATPase rapid exchange of the nucleotides in the binding site and in solution was assumed. Caged ATP* represents caged ATP that has already absorbed a photon and will release the nucleotide with a time constant of ~1 ms at pH 6.2 (see Materials and Methods).

The first and most obvious interpretation of the nucleotide dependence of the peak current (Fig. 4) would be the competitive binding of nucleotides and caged nucleotides to the AAC as has already been described for the NaK-ATPase (Fendler et al., 1993; Nagel et al., 1987). This is supported by the fact that caged nucleotides bind to the AAC binding site with a affinity similar to that for nucleotides but are not transported by the carrier (Brustovetsky et al., 1997). Under these assumptions a nucleotide concentration-dependent relaxation time should be obtained. However, no ATP or ADP dependence of the time constants (₂, ₃ and ₁, ₂) was observed in the experiments.

Therefore, an alternative interpretation involving the release-in-site mechanism is proposed (Fig. 11 B). In this mechanism competitive binding of caged ATP and ATP has no effect on the initial reaction after the flash, which can be explained as follows. No ATP-dependent time constant was found, so that it has to be concluded that nucleotide binding is not a rate-limiting process under the given experimental conditions. Therefore, at increasing caged ATP concentration (Fig. 4, filled circles), the peak current can only be increased by increasing the number of activated AAC molecules. A hyperbolic fit to the caged ATP dependence of the peak currents yields half-saturating concentrations of K_0.5(cATP) = 104 µM and K_0.5(cADP) = 78 µM (Fig. 4, A and D). The release-in-site mechanism also explains the linear behavior of the peak current at varying  and constant caged nucleotide concentration (Fig. 4, A and D, open circles). At a caged nucleotide concentration of ~2 mM, the peak current is in saturation. All of the available binding sites are assumed to be oriented toward the outside of the liposomes. At decreasing  and constant caged nucleotide concentration, the number of activated carrier molecules in the first instance is proportional to , yielding a linear nucleotide dependence. A slight curvature of these data might indicate overlapping activation from the bulk solution, with competition between nucleotides and caged nucleotides.

Following this interpretation, nucleotides and caged nucleotides compete for the binding site. However, this competitive binding has no influence on the peak current, because the peak current is determined by the initial half-cycle of the reaction, which is initiated by release in site. This does not rule out the possibility that competitive binding is of importance in the steady state.

Kinetics of the ATP and ADP translocation

Because the time resolution of the rapid solution exchange technique on the basis of the SSM is significantly below the time resolution that is achieved by the use of caged compounds in the BLM technique (10 ms compared to 1 ms), only the BLM data were used for the analysis of the AAC kinetics. For the assignment of the time constants the following criteria were applied: ATP dependence of the time constants and the corresponding amplitudes, pH dependence, and temperature dependence. The general approach to the assignment of certain time constants to the nucleotide translocation is by exclusion of all other possible assignments.

ATP translocation

pH

1

1

ADP translocation

1

1

1

1

1

The function of the AAC is symmetrical

After inhibition with CAT, which blocks only the cytosolic site (c-site) of the carrier (Fig. 11 A; Brustovetsky et al., 1996), in an ATP_ex-0_in experiment the peak current is reduced to ~50%, implying that the AAC is probably randomly incorporated into the liposomes, i.e., that 50% are inside-out and right-side-out oriented. The values for the time constants ₂ and ₃ (results not shown) after inhibition with CAT are in agreement with the values without CAT. From this we conclude that there is no significant kinetic difference between the two transport directions, and as a first approximation, the AAC can be regarded as symmetrical.

Inhibition of the AAC by anions

In Figs. 7 A and 8 A it is demonstrated that all three phases of the transient current in the case ADP_ex-0_in can be blocked at high chloride concentrations (200 mM). Fig. 9 further proves that it is an anionic effect because two anions, chloride and gluconate, with the same charge inhibit with different efficiencies. This is in agreement with the observation that caged ATP binds more efficiently to the AAC in a gluconate medium than in a chloride medium (data not shown). In the case of ATP_ex-0_in the fast-release phase was significantly reduced at 300 mM Cl, but the slow-release phase could still be observed. The reason for this finding is unclear.

Inhibition of the AAC carrier by anions could be explained by the shielding of the positive charges in the carrier-binding site, which are believed to be responsible for nucleotide binding. Similar results were found for the NaK-ATPase by Nørby and Esmann (1997).

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

In the system ATP_ex-0_in and ADP_ex-0_in, ATP and ADP are electrogenically transported into the liposomes by the AAC. The charge movements in the system ATP_ex-0_in and ADP_ex-0_in have opposite directions. This supports the model of an equivalent of 3.3 positive countercharges at the binding site of the AAC, resulting in 0.7 and +0.3 net transported charges in the cases ATP_ex-0_in and ADP_ex-0_in, respectively. The transport of the nucleotides could be inhibited by CAT/BKA and by Mg²⁺. A reaction rate constant for the ATP translocation of 160 s¹ and a lower limit for the ADP translocation of 400 s¹ were determined. The function of the carrier was found to be approximately symmetrical. The release of the nucleotides could be observed in the transient electrical signal. As a model we propose that the nucleotides are released in the binding site from the prebound caged nucleotides and then translocated by the carrier (here for ATP): caged ATP C_o  caged ATP* C_o  ATP C_o  C_i + ATP. The nucleotide binding and transport by the AAC can be blocked by anions.