Mechanisms of active transport in isolated membrane vesicles. IV. Galactose transport by isolated membrane vesicles from Escherichia coli.

Abstract An inducible transport system for d-galactose is present in isolated membrane vesicles from Escherichia coli ML 3 and ML 35. Like the transport systems for β-galactosides and amino acids, galactose transport is coupled primarily to a membrane-bound d-lactic dehydrogenase. In addition to d-lactate, the transport system is markedly stimulated by ascorbate plus phenazine methosulfate and oxygen. Succinate, dl-α-hydroxybutyrate, and l-lactate partially replace d-lactate, but are less effective than either d-lactate or ascorbate-phenazine methosulfate. ATP and P-enolpyruvate, in addition to a number of other metabolites and cofactors, do not stimulate galactose transport by the vesicles. Galactose transport by isolated membrane vesicles requires oxygen, and is blocked by electron transfer inhibitors and by other general metabolic inhibitors. However, transport is not significantly inhibited by high concentrations of arsenate or oligomycin. "Galactose binding protein" is totally absent from the vesicles, as is a high affinity galactose transport system present in whole cells.


Mechanisms
The results presented in this paper provide preliminary evidence for the concept that the "carriers" of the n-lactic dehydrogenase-coupled transport systems in isolated membrane vesicles from Escherichia co2i may be electron transfer intermediates.
Initial rates of lactose transport and D-&iCtiC dehydrogenase activity respond identically to temperature and both processes have the same activation energy of 8400 cal per mole.
The steady state levels of lactose accumulation at a variety of temperatures represent equilibrium states in which there is a balance between influx and efilux. This balance can be easily influenced by raising or lowering the temperature. Temperature-induced efllux is a saturable process with an apparent affinity constant that is approximately 60 times higher than the affinity constant for influx determined under the same experimental conditions. The apparent maximum velocity of temperature-induced efllux, on the other hand, is the same as that of influx.
Potassium cyanide also induces a saturable efilux phenomenon which has an apparent K, that is much higher than that of the influx process.
Furthermore, p-chloromercuribenzoate inhibits temperature-induced efflux of intramembranal lactose, exchange of external lactose with [14C]lactose in the intramembranal pool, and lactose e&x induced by 2,4-dinitrophenol. Inhibition of these experimental parameters and of D-lactic dehydrogenase by p-chloromercuribenzoate is reversed by dithiothreitol.
Reduction of the respiratory chain between D-lactic dehydrogenase and cytochrome bl is responsible for carriermediated efIlux of lactose. Anaerobiosis, cyanide, and 2heptyl-4-hydroxyquinoline-N-oxide, each of which inhibits electron transfer after cytochrome br, cause marked efIh.tx. Amytal causes slow efflux, and oxamate and p-chloromercuri-* A preliminary report of portions of this work has been submitted for publication (1). $ Postdoctoral fellow of the American Cancer Society (PF 545). Present address, Department of Biochemistry, Baylor College of Medicine, Houston, Texas 77045. benzoate do not cause efflux despite marked inhibition of Dlactic dehydrogenase and the initial rate of lactose transport. Addition of amino acids which are also transported by Dlactic dehydrogenase-coupled transport mechanisms results in little or no inhibition of lactose transport. These findings are discussed in terms of a conceptual working model in which the carriers are depicted as electron transfer intermediates between D-lactic dehydrogenase and cytochrome b,.
The data presented in the first paper in this srries (2) indicate that the site of energy coupling of n-lactic dehydrogenase to active transport lies between the primary dehydrogenase and cytochrome bl, the first cytochrome in the respiratory chain of Escherichia coli. In addition, the possibility that the "carriers" may be electron transfer intermediates between n-lactic dehydrogenase and cytochrome bl TT-as suggested.
The experiments presented in this paper provide further preliminary evidence for a possible electron transfer nature of the transport-specific components of the D-hCtiC dehydrogenaaecoupled transport systems.
Membrane vesicles were prepared from these cells as described previously (5-7).
Transport Studies-Assays for lactose and amino acid uptake were carried out as reported previously (7-Q). Glucuronate, arabinose, galactose, and glucose-6-P uptake studies were performed exactly as described for lactose and amino acid uptake with final concentrations of glucuronate, arabinose, galactose, and glucose-6-P of 0.1,0.3, 0.2, and 0.02 mu, respectively. All other materials used in these experiments were of reagent grade and were obtained from commercial sources.

RESULTS
Effect of Temperature on D-Lactic Dehydrogenase Activity and /3-Galactoside Transport- Fig.  1 represents time courses of lactose uptake by ML 308-225 membrane vesicles at a variety of temperatures.
As shown, initial rates of uptake increase with temperature up to 53", whereas the steady state level of lactose accumulation at 15 min increases from O-18" and then decreases above 18". At 53", membranes take up lactose very rapidly for 15 set and then lose radioactivity such that, by 1 min, approximately 50% of the radioactive lactose that had been accumulated in 15 set is lost. After 1 min, loss of radioactivity continues, but at a much slower rate. These data are similar to studies carried out with whole cells (10). The initial rate of D-lactic dehydrogenase activity increases very slightly as the temperature is raised from 0" to approximately 15", and then increases more markedly above 15" (Fig.  2A).
From 15" to approximately 50", the reaction rate increases essentially linearly, and then decreases abruptly at temperatures exceeding 55". In Fig. 2B, initial rates of lactose transport (16 SEC.) and steady state levels of lactose accumulation (16 MIN.) in the presence of D-lactate are plotted as a function of temperature. The steady state level of lactose accumulated in 15-min incubations exhibits a broad peak, from 15-35" with a maximum at approximately 18". The initial rate of lactose uptake (16 SEC.) is optimal at 50-55", and the initial rates of lactose trans-  The similarity of the effect of temperature on both n-lactic dehydrogenase activity and the initial rate of lactose transport in the presence of u-lactate is further emphasized in Fig. 3. Clearly, the initial rates of transport and n-lactic dehydrogenase activity respond almost identically to temperature. Increments in the log10 of the velocities of both activities from approximately 20-45" are identical, yielding an activation energy of 8400 cal per mole. Furthermore, there are discontinuities in both activities at 10-15" and at 4550".
The small discrepancies in the temperatures at which the discontinuities occur (from 3-5") are within experimental error.
Data given in Fig. 4 show that the steady state levels of lactose accumulated at 25" or 45" represent equilibrium states which can be shifted by changing temperature.
At 45", [i4C]lactose is taken up very rapidly by the vesicles which achieve a steady state level of lactose accumulation of approximately 12 nmoles per mg of membrane protein (corresponding to an intramembranal concentration of about 5.5 mM) within about 3 min. When the temperature is lowered to 25", the membranes accumulate lactose to a steady state level of 31 to 34 nmoles per mg of membrane protein (an intramembranal concentration of 14 to 16 mM) approximately 5 min after the temperature shift.
If the temperature is again raised to 45", the vesicles re-equilibrate at the same steady state level as that observed during the initial incubation at 45" within 2 min.
The temperature dependence of efflux is shown in Fig  The reactions were terminated at the times shown and the samples were assayed as described previously (8,9) linear for at least 1 min at each temperature studied (see temperature shift from 25-45" in Fig. 4). There is a sharp efflux minimum at 18". Thus the relationship between the rate of efflux and temperature is approximately the inverse of the relationship between the steady state level of lactose accumulation and temperature (see Figs. 1 and 2B). Membranes are not irreversibly damaged by I-min incubations at temperatures up to and including 50" (data not shown).
Above 55", there is rapid and irreversible loss of both o-lactic dehydrogenase and transport activities.
Kinetics of E&n-Before presenting kinetic data on temperature-induced efIlux, it is necessary to characterize the kinetics of uptake at 45" as a basis for comparison.
The initial rate of lactose uptake at 45" shows saturation kinetics at each time point studied (Fig. 6, A and B). Since uptake at 45" is so rapid (see Fig. I In Fig. 6B, the data have been plotted by the method of Lineweaver and Burk (12). Each function yields a K, of 0.4 mM. With the data obtained from experiments carried out for 15 set, a V,,, 0 f 50 nmoles per mg of membrane protein per min is obtained.
This value is in good agreement with previous kinetic studies carried out at 25" (9). The V,,, at 45" is higher than that obtained at 25" (approximately 20 nmoles per mg of membrane protein per min at 25" (9)).
The rate of lactose efflux at 45" also exhibits saturation kinetics when studied as a function of intramembranal lactose concentration (Fig. 7A). Vesicles were first loaded with lactose by incubation at 20" in the presence of n-lactate and various concentrations of [YJlactose and subsequently transferred to 45". The data in the inset yield a K, of 25 mM and a V,,, of 50 nmoles per mg of membrane protein per min. Compared to influx at 45", the K, of efflux is approximately 60 times higher, but the V,,,,, is almost identical.
Addition of energy poisons to either membrane vesicles (8) OI  In order to avoid complications due to galactose transport via the p-galactoside transport system, membranes prepared from E. coli ML 3 (y-) grown on galactose as described under "Methods" were used for galactose uptake. These membrane preparations did not transport either lactose or methyl-pthio-n-galactopyranoside in the presence of D-lactate. Studies on glucose-6-P transport were performed with membrane preparations from E. coli GN-2 growu on glucose-6-P as described under "Methods." Membrane vesicles prepared from the organisms described above were t,reated exactly as described in the figure legend to Fig. 9: I. Control-membranes were washed twice with 0.1 1~ potassium phosphate buffer (pH 6.6); II. PCMB-membranes treated with 8.3 X 1O-5 M PCMB as described in Fig. 9 were washed t,wice with potassium phosphate buffer (pH 6.6); III. PCMB-dithiothreitol-membranes treated with 8.3 X 1OV M PCMB were washed twice with 0.1 M potassium phosphate buffer (pH 6.6) containing 1 mM dithiot,hreitol. After washing, the membrane pellets were treat,ed as described in Fig. 9. Uptake studies were carried out for 1 min at 20" as described previously (8,9) with the sugars and amino acids given at the concentrations and specific activities given under "Methods" or reported previously (8,9). vesicles in the presence of n-lactate is sensitive to PCMB3 (Fig. 8), and decreases markedly with PCMB concentrations up to approximately 0.05 mM. NEM is also an effective inhibitor of galactoside transport by membrane vesicles (9) and PCMB inhibits fl-galactoside uptake into whole cells (15).
The inhibitory effect of PCMB on n-lactate-dependent concentrative uptake of lactose is reversed by dithiothreitol.
As described in Fig. 9, membranes were treated with PC'RIB and then washed with solutions to which dithiothreitol either had or had not been added. The time course of lactose uptake by membranes which had been treated with PCMB and subsequently with dithiothreitol is insignificantly different from the control preparation.
On the other hand, the sample which had been treated with PCMB but not dithiothreitol shows marked inhibition of the initial rate of uptake and significant inhibition of the steady state level of lactose accumulation.
On addition of dithiothreitol, the rate and extent of lactose uptake by this preparation are almost doubled.  The inhibitory effect of PCMB on fi-galactoside t,ransport is general for all n-lactic dehydrogenase-coupled transport systems, and, in each case, PCMB inhibition is virtually completely reversed by treating vesicles with dithiothreitol (Table I). Initial rates of transport of lactose, galactose, arabinose, glucuronate, glucose-6-P, proline, glutamic acid, serine, alanine, lysine, tryptophan, and tyrosine are inhibited by PCMB to varying extents. As shown in the last column of Table I, after membrane preparations are washed in phosphate buffer containing dithiothreitol, the inhibition observed is almost completely abolished. Detailed studies on the effects of PCMB on each of these transport systems will be published at a later date.4 The data presented in Table II represent measurements of D-lactate oxidation by the same mernbranc preparations used for most of the transport studies presented in Table I   dithiothreitol, there is complete restoration of activity. Each transport system studied in Table I and n-lactate oxidation is also sensitive to NEM; however, with this sulfhydryl reagent, the effects are not reversed with dithiothreitol.
Addition of PCMB to membrane vesicles first loaded at 20" causes marked inhibition of lactose efflux at 45" (Fig. 10). When dithiothreitol is added just before the temperature shift, however, t,he rate of efflux is indistinguishable from the control. PCMB also inhibits exchange of external lactose with [14C]lactose present in the intramembranal pool (Fig. 11) and lactose efflux induced by the addition of 2,4-dinitrophenol to previously loaded membrane vesicles (Fig. 12). Moreover, PCMB inhibition of both of these phenomena is reversed by addition of dithiothreitol (Figs. 11 and 12).
D-Lactate-stimulated concentrative uptake of lactose and the loss of radioactive lactose evoked by raising the temperature or adding [nC]lactose or 2,4-dinitrophenol are also inhibited by FIG. 12 ( NEM; however, the inhibition observed is not reversed by dithiothreitol. These experiments are not presented in detail here. Effect of Anaerobiosis and Electron Transfer Inhibitors on Lactose Transport-The effects of anaerobiosis and a variety of electron transfer inhibitors on the ability of previously loaded membrane vesicles to retain [W]lactose are shown in Fig. 13. These inhibitors were selected because their sites of inhibition in the respiratory chain of E. coli have been well documented. The samples were gassed with argon for 5 min before the addition of n-lactate and ['*Cllactose by injection through the stopper. The incubations were then continued under argon. At the times indicated, the reactions were terminated and the samples were assayed by methods described previously (8,9). The control samples (hm) contained none of the inhibitors and were incubated under aerobic conditions (room air). HO&NO, 2-heptyl-4-hydroxyquinoline-N-oxide.
On the other hand, amytal is only slightly effective, and PCMB and oxamate do not cause any loss of radioactivity.
The concentrations of each inhibitor used here produce at least 70% inhibition of the initial rate of n-lactic dehydrogenase activity (2) and lactose uptake (Fig. 14). Removal of oxygen from the reaction mixture also blocks n-lactate oxidation and lactose transport (9).
The effects of anaerobiosis and the same electron transfer inhibitors on the time course of lactose uptake in the presence of n-lactate are shown in Fig. 14. The inhibitors were used at the same concentrations as in the experiment presented in Fig. 13. All of the inhibitors, as well as removal of oxygen, markedly inhibit the initial rate of lactose uptake (see also Fig. 9, PCMB). However, with oxamate, amytal, and PCMB (see Fig. 9), although the initial rate of uptake is markedly inhibited, the membranes accumulate significant lactose and begin to approximate the control samples by 15 min. With anaerobiosis, potassium cyanide, or 2-heptyl-4-hydroxyquinoline-N-oxide, inhibition is profound throughout the time course of the incubation.
Effect of Amino Acids on Time Course of Lactose Uptake-The experiments presented in Fig. 15, A  Reaction mixtures containing ML 308-225 membranes were prepared as described in the legend to Fig. 1. In addition, where indicated, the reaction mixtures contained each of the following amino acids @.A.) in the final individual concentrations given: proline, glutamic acid, aspartic acid, lysine, serine, glycine, alanine, threonine, tryptophan, leucine, isoleucine, valine, histidine, tyrosine, phenylalanine, and cysteine. The control samples (NO ADD., O-O) contained no amino acids. At the times shown, the reactions were terminated and the samples were assayed by methods described previously (8,9). A, lactose uptake at 20'; B, lactose uptake at 45".
acid was present at a saturating concentration for its transport system (8) .4 It can be seen that, at either temperature, the presence of amino acids causes only mild inhibition of either the rate or extent of lactose uptake. DISCUSSION The experimental findings presented in this and the previous paper are consistent with the conceptual working model presented in Fig. 16. Detailed experiments with other n-lactic dehydrogenase-coupled transport systems which will be published at a later date show that each system behaves in a manner qualitatively similar to that shown here for the /3-galactoside transport system. In the mechanism presented, the carriers (in this specific case, the M protein (17)(18)(19)(20)) are depicted as electron transfer intermediates which undergo reversible oxidation-reduction. As shown, in the oxidized state, the carrier has a high affinity site for ligand which it binds on the exterior surface of the membrane.
Electrons coming ultimately from n-lactate through one or possibly more flavoproteins reduce a critical disulfide in the carrier molecule resulting in a conformational change. With this conformational change, the affinity of the carrier for its ligand is markedly decreased and ligand is released on the interior surface of the membrane.
The reduced "sulfhydryl" form of the carrier is oxidized by cytochrome bi and electrons then flow through the remainder of the cytochrome chain to reduce molecular oxygen to water.
The reduced form of the carrier can also "vibrate" and catalyze a low affinity, carrier-mediated, non-energy-dependent transport of ligand across the membrane.
Although no direct evidence has been presented which shows unequivocally that the carriers are electron transfer intermediates or that they are the only sulfhydryl-containing components of the respiratory chain between n-lactic dehydrogenase and cytochrome bl, this formulation is consistent with all of the experimental observations presented and is the simplest conception possible.
The concept that the carriers may be electron transfer intermediates is supported primarily by experiments in which the Since only anaerobiosis and those inhibitors which block electron transfer after the site of energy coupling induce efflux (Fig. 13), reduction of the electron transfer chain between n-lactic dehydrogenase and cytochrome bl must be responsible for efflux. The effect of anaerobiosis and the same inhibitors on the time course of lactose uptake (Fig. 14) is also consistent with this interpretation.
Since removal of oxygen or addition of electron transfer inhibitors which inhibit after the site of energy coupling cause reduction of that site, membranes incubated under these conditions manifest profound inhibition of uptake throughout the time course of the experiment. On the other hand, inhibition before the site of energy coupling slows the rate of reduction of the energy-coupling site, but not its rate of oxidation by cytochrome bi. Thus, vesicles incubated under these conditions manifest markedly diminished initial rates of uptake but eventually accumulate significant quantities of lactose.
As shown in this and the previous paper (2), n-lactate oxidation and the n-lactic dehydrogenase-dependent and -independent aspects of P-galactoside transport are inhibited by PCMB and NEM.
Moreover, PCMR inhibition of each of these parameters is reversed by dithiothreitol.
Since the site of energy coupling lies between n-lactic dehydrogenase and cytochrome bl, and the site(s) of inhibition by sulfhydryl reagents is also between the dehydrogenase and cytochrome bl, these data are consistent with the model presented.
It must be emphasized that there may be many sulfhydryl-containing components in the respiratory chain between n-lactic dehydrogenase and cytochrome bl. Furthermore, it is possible that the carriers are not obligatory intermediates in the respiratory chain, but reflect the oxidation-reduction potential at the site of energy coupling in an unknown manner. The mechanism presented is merely the simplest conception that accounts for the data presented.
In the model proposed, for each transport system, there should be a component between n-lactic dehydrogenase and cytochrome bi which has a binding site that is specific for that particular transport substrate. Evidence supporting this prediction is provided by experiments in which lactose transport was studied in the presence of a mixture of amino acids the concentrations of which were more than sufficient to saturate their respective transport systems. Little or no inhibition of either the initial rate of lactose uptake or the steady state level of lactose accumulation was observed under any of the conditions studied.
Other data, the details of which will be published at a later date, indicate that the sum of the V,,,,, values of all of the known n-lactic dehydrogenase-coupled transport systems in a given membrane preparation are comparable to the Vmax of n-lactic dehydrogenase activity in that same membrane preparation.4