Anaerobic Transport in Escherichia cob Membrane Vesicles

Anaerobic lactose and/or amino acid transport by membrane vesicles prepared from Escherichia coli ML 308-225 can be coupled to at least four electron transfer systems: a-glycerol-P-dehydrogenase:nitrate reductase, formate dehydrogenase:nitrate reductase, a-glycerol-P dehydrogenase:fumarate reductase, and formate dehydrogenase:fumarate reductase. Vesicles contain one or more of these electron transfer systems depending on the growth conditions of the parent cells. a-Glycerol-P dehydrogenase and fumarate reductase are present only in vesicles prepared from cells grown in the presence of glycerol or fumarate, respectively. Formate dehydrogenase and nitrate reductase activities, on the other hand, are present in vesicles from cells grown on a variety of media. a-Glycerol-P and formate are able to drive aerobic transport in vesicles prepared from anaerobically grown cells, indicating coupling between aerobic and anaerobic electron transfer systems.

Anaerobic lactose and/or amino acid transport by membrane vesicles prepared from Escherichia coli ML 308-225 can be coupled to at least four electron transfer systems: a-glycerol-P-dehydrogenase:nitrate reductase, formate dehydrogenase:nitrate reductase, a-glycerol-P dehydrogenase:fumarate reductase, and formate dehydrogenase:fumarate reductase. Vesicles contain one or more of these electron transfer systems depending on the growth conditions of the parent cells. a-Glycerol-P dehydrogenase and fumarate reductase are present only in vesicles prepared from cells grown in the presence of glycerol or fumarate, respectively.
Formate dehydrogenase and nitrate reductase activities, on the other hand, are present in vesicles from cells grown on a variety of media. a-Glycerol-P and formate are able to drive aerobic transport in vesicles prepared from anaerobically grown cells, indicating coupling between aerobic and anaerobic electron transfer systems.
Active transport of P-galactosides and amino acids, as well as a variety of other metabolites.
is coupled to electron transfer in cytoplasmic membrane vesicles isolated from Escherichia coli and a number of other organisms (l-4). In E. coli and Salmonella typhimurium membrane vesicles, the transport systems are coupled primarily to the oxidation of D-lactate or reduced phenazine methosulfate via a membrane-bound cytochrome chain with oxygen as the terminal acceptor (l-4). Recent experiments demonstrate that virtually all of the vesicles isolated from E. coli ML 308-225 catalyze active transport (5). Moreover, using antibodies directed against D-lactate dehydrogenase and calcium, magnesium-stimulated ATPase, it has been shown that these enzymes are localized on the inner surface of the vesicle membrane (6, 7). These and other findings (l-4, 8, 9) indicate that few, if any, of the vesicles are inverted, and, as such, support previous observations (l-4, 10, 11) which are consistent with the contention that the energy-coupling site for transport in E. coli vesicles is localized in a segment of the respiratory chain between D-lactate dehydrogenase and cytochrome b,. The carriers, however, are not electron transfer intermediates (12), and are present in the membrane in large excess relative to D-lactate dehydrogenase (13). Recent evidence indicates that generation of a membrane potential is involved in the transport mechanism (4,14,15 Previous studies (l-4, 16) indicate that ATP does not play a role in active transport in membrane vesicles. However, other evidence (17-24) is consistent with the hypothesis that glycolytically generated ATP is able to drive active transport in whole cells under anaerobic conditions.
Recently, Konings and Kaback (25) demonstrated that anaerobic P-galactoside transport in whole cells and membrane vesicles from E. coli ML 308-225 is coupled to the oxidation of a-glycerol-P with fumarate as an anaerobic electron acceptor or to the oxidation of formate with nitrate as an anaerobic electron acceptor. In addition, Butlin (20) and Rosenberg et al. (26) have shown that mutants of E. coli which are deficient in calcium, magnesium-stimulated ATPase (uncA) are able to catalyze active transport of serine and phosphate under anaerobic conditions in the presence of fumarate as an electron acceptor.
The results presented in this paper demonstrate that in addition to P-galactosides, anaerobic amino acid transport is also driven by electron transfer in isolated membrane vesicles from E. coli ML 308-225. In addition, it is demonstrated that the P-galactoside and amino acid transport systems may be coupled to at least four distinct anaerobic electron transfer systems--a -glycerol-P dehydrogenase:fumarate reductase, cu-glycerol-P dehydrogenase:nitrate reductase, formate dehydrogenase:fumarate reductase, and formate dehydrogenase:nitrate reductase.  (29). Membrane vesicles were prepared and isolated by the modified procedure described previously (25). Purified vesicles were resuspended in 50 mM potassium phosphate (pH 6.6) to a protein concentration of 5 to 10 mg/ml, and samples of 0.5 to 1.0 ml were frozen rapidly and stored in liquid nitrogen.
assaying the production of nitrite as described by Showe and DeMoss (34 Transport Assays-Transport under aerobic and anaerobic conditions was assayed as described previously (29,30) with the exception that oxygen-free nitrogen rather than argon was used for the anaerobic assays.  Cells were harvested at the end of exponential growth and resuspended in 50 mM potassium phosphate (pH 6.6) containing 50 Kg/ml of chloramphenicol, and assayed without exogenous electron donors or acceptors. Transport assays were carried out at 25" as described previously (25, 30) and under "Methods." fumarate catalyze anaerobic lactose transport when supplied with three combinations of electron donors and acceptors--cuglycerol-P and fumarate, cu-glycerol-P and nitrate, and formate and nitrate (Fig. 2, top). Very little or no stimulation is observed with formate and fumarate, or with cu-glycerol-P, formate, nitrate, or fumarate alone. However, as reported previously (25), in the absence of added electron donors or acceptors, relatively high endogenous uptake is observed.
When the same vesicles are assayed under aerobic conditions (Fig. 2, micldle), the effect of the electron donors and acceptors is quite different. As shown, neither fumarate nor nitrate has a significant effect on lactose transport with Lu-glycerol-P as the electron donor and oxygen as acceptor.
With formate as electron donor, however, nitrate and especially fumarate enhance the rate and extent of lactose uptake over that observed with oxygen alone as the terminal acceptor. As shown in the bottom of Fig. 2, glycerol-fumarate vesicles do not catalyze amino acid transport effectively under anaerobic conditions. Slight stimulation of uptake is observed with a-glycerol-P or formate as electron donors and nitrate as electron acceptor. Moreover, under aerobic conditions, amino acid uptake by these vesicles is also relatively small (data not shown).
Glycerol-Nitrate Vesicles-Membrane vesicles prepared from E. coli ML 308-225 grown anaerobically with glycerol as carbon source and nitrate as electron acceptor catalyze amino acid transport under anaerobic conditions in the presence of formate and nitrate (Fig. 3, top) much lower rates and extents of uptake. Addition of 01glycerol-P or formate in the absence of an electron acceptor results in only a slight increase in amino acid transport under these conditions, and addition of fumarate or nitrate in the absence of an electron donor has no significant effect. When amino acid uptake by these vesicles is assayed under aerobic conditions (Fig. 3, middle), significant stimulation is observed in the presence of u-glycerol-P and formate. Moreover, addition of fumarate or nitrate as electron acceptors in the presence of oxygen has no significant effect. Anaerobic lactose uptake in these vesicles is very low and no stimulation is observed with any combination of electron donors and acceptors tested (Fig. 3, bottom). This finding is consistent with experiments carried out with whole cells grown under these conditions (Fig. 1).

Glucose-Fumarate
Vesicles-Although cells grown on glucose and fumarate transport lactose moderately well (Fig. l) 3. Uptake of lactose and amino acids under anaerobic and aerobic conditions in membrane vesicles of anaerobically grown E. coli ML 308-225 on glycerol-nitrate medium. Transport assays and additions were performed as described in legend of Fig. 2. under anaerobic conditions with any of the electron donors or acceptors when they are added alone or in combination (Fig. 4,  top). The reason for this inconsistency is not apparent. Under aerobic conditions, however, the vesicles exhibit moderate activity with formate as electron donor (Fig. 4, middle), and addition of fumarate or nitrate has no additional effect. No stimulation of aerobic lactose uptake in these vesicles is observed with a-glycerol-P in the presence or absence of fumarate or nitrate.
Formate is also the only effective electron donor for aerobic amino acid uptake in these vesicles, and under these conditions, nitrate or fumarate produces no additional stimulation of amino acid uptake (middle panels).
Anaerobic lactose transport is also catalyzed by these vesicles in the presence of formate and nitrate, but the absolute amount of lactose transported is relatively low (bottom panels). As shown for amino acid transport, all other electron donors and acceptors alone or in various combinations have no significant effect on lactose uptake.
Amino acid uptake by these vesicles under anaerobic conditions is mildly stimulated by formate in the presence of fumarate, and even less so by formate in the presence of nitrate (Fig. 4, bottom). a-Glycerol-P with fumarate or nitrate as acceptors stimulates anaerobic amino acid uptake slightly or not at all, and no significant uptake is observed with electron donors or acceptors alone.

Enzyme Assays and Coupled Activities
Glucose-Nitrate Vesicles-As shown in Fig. 5, vesicles prepared from cells grown anaerobically with glucose and nitrate The data presented in Table I  vesicles, and lowest in glucose-fumarate vesicles. Anaerobic cu-glycerol-P dehydrogenase activity is present only in vesicles prepared from glycerol-grown cells, and higher activity is observed when the cells are grown with fumarate as opposed to nitrate as an electron acceptor. Similarly, fumarate reductase activity is observed only when the cells are grown on fumarate, and activity is best when glycerol is used as the carbon source.
Membrane vesicles prepared from cells grown anaerobically on all of the media described exhibit coupling between formate dehydrogenase and nitrate reductase (Table  II). Coupled activity between these enzymes is highest in vesicles prepared from cells grown on glucose and nitrate, intermediate in glycerol-fumarate and glycerol-nitrate vesicles, and lowest in glucose-fumarate vesicles. Coupling between formate dehydrogenase and fumarate reductase is not present in any of the vesicle preparations, a finding which is consistent with observations demonstrating that none of the vesicles catalyzes anaerobic lactose or amino acid transport in the presence of formate and fumarate. Significant coupling between cyglycerol-P dehydrogenase and nitrate reductase is observed only in vesicles from cells grown on glycerol. It is surprising, moreover, that this activity is better when the cells are grown with fumarate as an electron acceptor as opposed to nitrate. Finally, coupling between a-glycerol-P dehydrogenase and fumarate reductase is observed only in vesicles from cells grown on glycerol with fumarate as an anaerobic electron acceptor, a finding which is consistent with the observations that only these vesicles exhibit both a-glycerol-P dehydrogenase and fumarate reductase activity (Table I).
The data presented in Table III demonstrate that formate is oxidized by all of the vesicle preparations (although formate oxidation is quite low in glucose-fumarate vesicles), a finding which is consistent with observations demonstrating that all of the vesicles exhibit formate dehydrogenase activity (Table I) and that formate drives lactose or amino acid transport under aerobic conditions. cY-Glycerol-P is oxidized only by glycerolfumarate and glycerol-nitrate vesicles, both of which contain oc-glycerol-P dehydrogenase (Table I) and exhibit cu-glycerol-Pdriven lactose or amino acid transport under aerobic and anaerobic conditions. These results confirm the observation that oxygen can serve as a terminal electron acceptor in vesicles from certain anaerobically grown cells.
Effect of Ascorbate-Phenazine Methosulfate, None of these electron donors stimulates anaerobic lactose or amino acid transport in vesicles prepared from cells grown anaerobically under any of the conditions described in this paper, even in the presence of nitrate or fumarate (data not shown). However, as shown in Fig. 6, when the vesicle preparations used in these experiments are assayed under aerobic conditions, many of these electron donors drive lactose or amino acid uptake effectively. Thus, lactose transport and amino acid transport are markedly stimulated by ascorbate-phenazine methosulfate in glycerol-fumarate vesicles, and glycerol-nitrate, glucose- FIG. 6. Aerobic lactose or amino acid uptake by membrane vesicles of E. coli ML 308-225, anaerobically grown on different media as indicated in the panel headings of the figure. Transport assays were carried out at 25" as described previously (25,30). Where indicated, NADH, lithium n-lactate, and sodium succinate were added at final concentrations of 10 mM. Sodium ascorbate and phenazine methosulfate where indicated, were added at final concentrations of 20 and 0.1 mM, respectively, 0, NADH; 0, ascorbate-phenazine methosulfate; A, succinate; V, n-lactate; A, no additions. 6797 fumarate, and glucose-nitrate vesicles, respectively. Interestingly, n-lactate does not drive lactose or amino acid transport aerobically except in glycerol-nitrate vesicles. Moreover. NADH is able to drive lactose or amino acid uptake reasonably well in all of these vesicles. Finally, it is noteworthy that succinate drives lactose transport at the same rate as ascorbate-phenazine methosulfate in glycerol-fumarate vesicles, but is much less effective than ascorbate-phenazine methosulfate for amino acid transport in the other vesicle preparations.

DISCUSSION
The data presented in this paper confirm and extend previous studies (25) on active transport under anaerobic conditions in isolated membrane vesicles from E. coli. As shown, lactose and/or amino acid transport can be coupled to at least four electron transfer systems which do not require oxygen as the terminal electron acceptor. These anaerobic electron transfer systems are cu-glycerol-P dehydrogenase: fumarate reductase, a-glycerol-P dehydrogenase:nitrate reductase, formate dehydrogenase:nitrate reductase, and formate dehydrogenase:fumarate reductase.
The Lu-glycerol-P dehydrogenase:fumarate reductase system has been studied in detail by Miki and Lin (38), and the involvement of a b-type cytochrome(s) was suggested by Konings and Kaback (25). The formate dehygrogenase:nitrate reductase system has been studied extensively by  and others (27,(40)(41)(42)(43).
Anaerobic growth on glycerol induces an anaerobic 01. glycerol-P dehydrogenase (32,44), and addition of fumarate as an electron acceptor induces fumarate reductase (33, 45). The synthesis of fumarate reductase is inhibited by the presence of nitrate in the growth medium (46). Alternatively, anaerobic growth on glucose with nitrate as an electron acceptor induces formate dehydrogenase and nitrate reductase (27). In addition, the latter two enzymes require selenium and molybdate for maximal activity (28,47). Nitrate reductase has been shown to accept electrons from a cytochrome of the b-type (39,42), and addition of nitrate to the growth medium inhibits the biosynthesis of a soluble formate dehydrogenase which is involved in the hydrogenlyase pathway (40). a-Glycerol-P dehydrogenase:fumarate reductase is present only in vesicles prepared from cells grown anaerobically on media containing glycerol and fumarate. Such vesicles catalyze anaerobic lactose transport in the presence of cu-glycerol-P and fumarate.
The formate dehydrogenase:nitrate reductase system is present in vesicles prepared from cells grown anaerobically on media containing glucose and nitrate as reported previously (25), but also in vesicles prepared from cells grown anaerobically on glycerol and nitrate, glucose and fumarate, and glycerol and fumarate.
Not only are the appropriate enzymes present, but addition of formate and nitrate drives anaerobic lactose and/or amino acid transport in all of the vesicle preparations except those prepared from cells grown on glucose and fumarate. The reason for the lack of activity in the latter preparation may be due to the relatively low level of nitrate reductase present in these vesicles. In any case, it is apparent that significant levels of formate dehydrogenase and nitrate reductase are produced in cells grown anaerobically under each of the conditions described. It should be emphasized, however, that yeast extract was added to all of the media, and it is possible that this supplement contains the factors necessary for the synthesis of these enzymes.
The oc-glycerol-P dehydrogenase:nitrate reductase system is present only in glycerol-fumarate and glycerol-nitrate vesicles, and this anaerobic electron transfer system is able to drive both lactose and amino acid transport under anaerobic conditions. Formate dehydrogenase and fumarate reductase activities are observed in glucose-fumarate and glycerol-fumarate vesicles, but coupling between these enzymes could not be demonstrated.
It is apparent from these studies that formate dehydrogenase and/or ol-glycerol-P dehydrogenase are also able to drive active transport in vesicles prepared from anaerobically grown cells with oxygen as the terminal electron acceptor. This finding suggests that there must be some degree of coupling between anaerobic and aerobic electron transfer chains in the vesicle membrane.
In contrast to vesicles prepared from aerobically grown cells where n-lactate and reduced phenazine methosulfate are the best electron donors with respect to active transport, these electron donors do not drive transport under anaerobic conditions.
Similarly, NADH and succinate are not able to drive transport under anaerobic conditions. Surprisingly, however, when vesicles prepared from anaerobically grown cells are assayed under aerobic conditons, o-lactate is no longer the best physiological electron donor for active transport, and succinate or NADH functions more effectively in this capacity.
Since recent experiments provide convincing evidence that the generation of a membrane potential is intimately involved in active transport (4, 14, 48) 1 it seems quite possible that variations in the efficiency of different electron donors to drive active transport under anaerobic or aerobic conditions may reflect variations in their ability to generate the appropriate membrane potential. Measurement of membrane potentials with triphenylmethylphosphonium and rubidium (in the presence of valinomycin) (4, 49) 1 under the conditions described here is currently in progress.