Mechanisms of Active Transport in Isolated Membrane Vesicles THE COUPLING OF REDUCED PHENAZINE METHOSULFATE TO THE CONCENTRATIVE UPTAKE OF /?-GALACTOSIDES AND AMINO ACIDS

SUMMARY An artificial electron donor system-ascorbate and phenazine methosulfate-markedly stimulates P-galactoside transport in isolated membrane vesicles from Escherichia coli ML 308-225. Maximal rates of transport are dependent on the presence of both ascorbate and phenazine methosulfate and also oxygen. Moreover, in the presence of phenazine methosulfate, reduced nicotinamide adenine dinucleotide stimulates the initial rate of lactose transport. The effect of ascorbate-phenazine methosulfate is inhibited by removal of oxygen or potassium cyanide, Z-heptyl-4-hydroxyquinoline-N-oxide, p-chloromercuribenzoate, or sodium amytal. Oxamic acid has no significant effect. Ascorbate-phenazine methosulfate reduces the respiratory chain of the vesicles below the potential level of cytochrome bl, but above approximately 80 % of the membrane-bound flavoproteins.


Mechanisms
Maximal rates of transport are dependent on the presence of both ascorbate and phenazine methosulfate and also oxygen.
Moreover, in the presence of phenazine methosulfate, reduced nicotinamide adenine dinucleotide stimulates the initial rate of lactose transport.
The effect of ascorbate-phenazine methosulfate is inhibited by removal of oxygen or potassium cyanide, Z-heptyl-4-hydroxyquinoline-N-oxide, p-chloromercuribenzoate, or sodium amytal. Oxamic acid has no significant effect. Ascorbate-phenazine methosulfate reduces the respiratory chain of the vesicles below the potential level of cytochrome bl, but above approximately 80 % of the membrane-bound flavoproteins. In the presence of ascorbate-phenazine methosulfate, membranes prepared from Escherichia coli, Salmonella typhimurium, Pseudomonas putida, Proteus mirabilis, Bacillus megaterium, and Bacillus subtilis catalyze the concentrative uptake of proline; membranes prepared from Staphylococcus aureus catalyze the concentrative uptake of lysine; and membranes prepared from Micrococcus dentrificans catalyze the concentrative uptake of glutamine.
1yhile the work presented in the first two papers in this series (1, 2) was in progress, Konings and Yrcese (3) presented a preliminary report on L-serine transport in isolated membrane vesicles from Bacillus subtilis.
These workers reported that NADH dehydrogenase is coupled to L-serine transport in membrane preparations from this organism, and, furthermore, that ascorbate-phenazine methosulfate, an artificial electron donor system, could markedly stimulate this transport system. Since many of the properties of the B. to those reported for the Eschericllia coli system (4), it was decided to investigate this artificial electron donor system in the hope that further insight into dehydrogenase-coupled transport systems might be obtained.
Preparation of &1embranes-With the exception of S. aureus, each of the organisms described above was subjected to the same lysozyme-EDTA procedure as that described previously for the preparation of membrane vesicles from E. coli (10,11).
For X. aureus, cells harvested from the logarithmic phase of growth were resuspended in 4.1 ~1 sodium chloride at approximately 5 g, wet weight, per ml. Lysostaphin (Mann) was then added at a concentration of 36 units per ml and the suspension was incubated at 37" for 1 hour. The protoplast suspension was harvested by centrifugation at approximately 20,000 x g for 30 min. All subsequent procedures were the same as those used for the preparation of membrane vesicles from II. coli except that potassium phosphate buffer (pH 7.3) was used throughout (rather than pH 6.6).
Transport &%&es-The methods used to study the uptake of lactose or amino acids were identical with those reported previously (I, 2,5,11,12).
iMaterials-The radioactive amino acids snd lactose used in these experiments were obtained from previously described sources (1,2,5,12). Sodium ascorbate and phenazine methosulfate were obtained from Calbiochem.
All other materials were of reagent grade obtained from commercial sources. The reaction vessels were fitted with perforated rubber stoppers so that each sample could be gassed with oxygen by means of a hypodermic needle inserted through the stopper. Additions to the reaction mixtures were made with a Hamilton microsyringe by insertion through the perforation in the stopper. The samples were incubated at 25", and gassed with pure oxygen for 2 to 3 min before the following additions and throughout the incubation: At the times given, the reactions were terminated and the samples were assayed by methods reported previously (14,12). phenazine methosulfate to isolated membrane vesicles causes marked stimulation of both the initial rate of lactose uptake and the absolute level of lactose accumulation (Fig. 1). In the absence of ascorbate, phenazine methosulfate, or both, vesicles take up lactose at much slower rates, and do not accumulate the galactoside effectively.
The initial rate of lactose uptake with ascorbate-phenazine methosulfate is more than doubled when incubations are carried out in a pure oxygen atmosphere rather than in room air (Fig. 2). Loss of lactose from the intramembranal pool with ascorbatephenazine methosulfate and oxygen after 5 min is due to rapid utilization of ascorbate. As shown, this loss of radioactivity is minimized by incubation in room air (Fig. 2). Although not shown, an increase in oxygen tension above atmospheric levels has no effect on the rate or extent of lactose uptake in the presence of n-lactate under these conditions. It is also noteworthy that the absolute level of lactose accumulation is higher with n-lactate than with ascorbate-phenazine methosulfate.
The rate of oxygen consumption in the presence of vesicles and ascorbate-phenazine methosulfate (1.4 pg atoms of oxygen were incubated with air. per mg of membrane protein) is about 6-fold higher than the rate obtained with n-lactate as donor. Moreover, a significant fraction (about one-third) of the oxygen depletion rate is due to nonenzymatic reaction of reduced phenazine methosulfate with oxygen. This observation shows that the oxygen requirement for lactose transport observed with ascorbate-phenazine methosulfate is due to rate-limiting oxygen diffusion. As the concentration of phenazine methosulfate is increased from 0 to approximately 0.1 mM in the presence of saturating concentrations of ascorbate (20 MM), the rate of lactose uptake increases approximately la-fold. From 0.1 mM to approximately 0.3 mM phenazine methosulfate, the rate is constant, and at concentrations of phenazine methosulfate above 0.3 mM the rate decreases. All attempts to replace phenazine methosulfate with N, N, N', N'tetramethyl-p-phenylenediamine dihydrochloride (13), cytochrome c, ferricyanide, or silicomolybdate (14) were uniformly negative. E$ect of Phenazine Methodfate on D-lactate, Succinaie, and NADH Xtirnulation of Lactose Transport-Lactose uptake in the presence of n-lactate (Fig. 3A) or succinate (Fig. 3B) is slightly inhibited by the addition of phenazine methosulfate. On the other hand, the rate of lactose uptake in the presence of NADH is markedly stimulated by the addition of phenazine methosulfate (Fig. 3C). In the presence of NADH and phenazine methosulfate, the initial rate of lactose uptake is approximately the same as the initial rate of uptake in the presence of n-lactate (compare Fig. 3A with Fig. 3C). As with ascorbate-phenazine methosulfate, lactose uptake in the presence of NADH and phenazine methosulfate increases rapidly for approximately 3 to 5 min. Subsequently, the membranes lose radioactivity such that, by 15 min, there is no significant difference in uptake between membranes incubated with NADH in the presence or absence of phenazine methosulfate. The stimulation of lactose transport by NADH-phenazine methosulfate is markedly inhibited by amytal (data not shown). Membrane samples were assayed for lactose uptake at 25" under oxygen as described in Fig. 1.  Values for reduction are expressed as percentages of controls reduced with dithionite ( Fig. 5) and were based on the spectra shown in Fig. 5 (ascorbate-phenazine methosulfate). The values shown for n-lactate were derived from previously published data (3). Wave length pairs were those suggested by Jones and Redfern (16) and values for the absorption of each component were based on the absorption difference between these wave lengths. Reaction mixtures prepared as in Fig. 1 were assayed for lactose uptake at 25' under oxygen in the presence of ascorbate, phenazine methosulfate, and were assayed in the presence of ascorbate, phenazine methosulfate, and [1-i4C]lactose as described in Fig. 1  Methosulfate-With the exception of oxamic acid, anaerobiosis and the electron transfer inhibitors that block n-lactate oxidation and lactose transport in the presence of n-lactate (1, 2, 5) also inhibit lactose uptake in the presence of ascorbate-phenazine methosulfate (Fig. 4). Moreover, the effects of anaerobiosis, p-chloromercuribenzoate, 2-heptyl-4-hydroxyquinoline-N-oxide, KCN, and amytal shown here are almost identical with those presented for n-lactic dehydrogenase-coupled lactose transport (2). phenazine methosulfate and dithionite is illustrated in Fig. 5. The absorption bands are identified as described previously (1,15). Anaerobic reduction by ascorbate-phenazine methosulfate reveals nearly all available cytochromes, and the cytochrome spectrum observed with ascorbate-phenazine methosulfate is very similar to that obtained with D-lactate (1). However, it is especially significant that the flavoprotein trough at 465 nm is not as deep when the membranes are reduced with ascorbate-phenazine methosulfate as opposed to dithionite (Fig. 5) or n-lactate (1). Quantitative determination of the amounts of each chromophore reduced by ascorbate-phenaeine methosulfate or n-lactate relative to dithionite is shown in Table I. No significant differences are observed for reduction of Soret, cytochrome bi, or cytochrome a2 (cytochrome oxidase) by either of these electron donors. However, ascorbate-phenazine methosulfate reduces only '21% of the membrane-bound flavoprotein as opposed to 63% reduction by n-lactate.
E$ect of Ascorbate-Phenctxine Methosulfate on Amino Acid Transport by Membrane Vesicles Prepared from Various Grampositive and Gram-negative Organisms-The data presented in Fig. 6, A through H, show that ascorbate-phenaeine methosulfate markedly stimulates the uptake of proline by membrane vesicles prepared from E. coli (A), 8. typhimurium (B), P. putida (C), P. mirabilis (D), B. megaterium (G), and B. subtilis (H). In addition, the uptake of glutamine by membranes prepared from M. dcnitri$cans (E) and of lysine by X. aureus membrane vesicles (F) is similarly stimulated by ascorbate-phenazine methosulfate. In each case, ascorbate plus phenazine methosulfate, but not ascorbate or phenazine methosulfate alone, markedly stimulates both the initial rate of uptake and the steady state levels of accumulation of the appropriate amino acid in each membrane preparation studied. The uptake of proline by membrane vesicles prepared from Streptococcus faecalis and Clostridium thermoaceticum was negligible in the presence of either ascorbatephenazine methosulfate, n-lactate, or NADH (data not shown).

DISCUSSION
The results presented in this paper describe an artificial electron donor system which can be coupled to /3-galactoside and amino acid transport in isolated membrane preparations.
Experiments which will be presented in future publications show that ascorbate-phenazine methosulfate can also be coupled to the concentrative uptake of galactose,2 arabinose,' glucuronate,' gluconate,' and glucose-6-P by E. coli membrane vesicles. The initial rate of lactose transport in E. coli ML 308-225 membrane vesicles is stimulated about 3 times better by ascorbate-phenazine methosulfate than n-lactate, the best physiological electron donor for this transport system. This finding is consistent with the observation that reduced phenazine methosulfate is oxidized much faster than n-lactate by the vesicles.
When phenazine methosulfate is added to reaction mixtures containing NADH, there is marked stimulation of lactose transport over that obtained with NADH alone. The observation that phenazine methosulfate does not enhance the effect of Dlactate or succinate is not surprising since phenazine methosulfate is not spontaneously reduced by n-lactate or succinate. This must be due to a kinetic barrier since the reduction potentials are such that the reduction of phenazine methosulfate is thermodynamically favorable. Therefore, the addition of phenazine methosulfate does not increase the rate of oxidation of n-lactate or succinate over that obtained with the appropriate membrane-bound dehydrogenases. Moreover, these dehydrogenases are coupled relatively effectively to the transport system. On the other hand, NADH, like ascorbate, reduces phenazine methosulfate spontaneously, and reduced phenazine methosulfate is able to couple electrons to transport much more effectively than NADH dehydrogenase.
By this means, NADH