Energization of Active Transport by Escherichia COG*

SUMMARY An activated membrane state necessary for the active transport of certain amino acids, carbohydrates, and cations by Escherichia coli can be generated by either oxidative energy or phosphate bond energy. Membrane fragment vesicles from E. coli couple oxidative but not phosphate. bond energy to the transport of proline. The vesicles show an O,-dependent transport yet lack meas-ureable capacity for oxidative phosphorylation. Such transport is arsenate-resistant, consistent with the transport being independent of oxidative phosphorylation. Although vesicles possess ATPase activity, added ATP does not stimulate proline transport. With intact E. coli cells, incubation with high arsenate and low phosphate drastically lowers intracellular ATP and P-enolpyruvate levels in the presence or the absence of Op. Aerobic proline transport is uninhibited, but in contrast, anaerobic transport is sharply reduced. These results show that intact cells can use energy either from oxidations or from phosphorylations to drive active proline transport. Similarly, active accumulation of leucine, &methylthiogalactoside, Rb+ (as a K+ analogue), and a-methylglucoside occurs aerobically in the presence of high arsenate. Uncouplers of oxidative phosphorylation block the use of energy for transport both from oxidations and from ATP, suggesting that a similar or identical energy-conserving membrane state is formed from


Energization
of Active Transport by Escherichia COG* (Received for publication, February 10, 1972) SUMMARY An activated membrane state necessary for the active transport of certain amino acids, carbohydrates, and cations by Escherichia coli can be generated by either oxidative energy or phosphate bond energy.
Membrane fragment vesicles from E. coli couple oxidative but not phosphate. bond energy to the transport of proline. The vesicles show an O,-dependent transport yet lack measureable capacity for oxidative phosphorylation.
Such transport is arsenate-resistant, consistent with the transport being independent of oxidative phosphorylation. Although vesicles possess ATPase activity, added ATP does not stimulate proline transport.
With intact E. coli cells, incubation with high arsenate and low phosphate drastically lowers intracellular ATP and P-enolpyruvate levels in the presence or the absence of Op. Aerobic proline transport is uninhibited, but in contrast, anaerobic transport is sharply reduced.
These results show that intact cells can use energy either from oxidations or from phosphorylations to drive active proline transport. Similarly, active accumulation of leucine, &methylthiogalactoside, Rb+ (as a K+ analogue), and a-methylglucoside occurs aerobically in the presence of high arsenate.
Uncouplers of oxidative phosphorylation block the use of energy for transport both from oxidations and from ATP, suggesting that a similar or identical energy-conserving membrane state is formed from either energy source.
Transport ofi a-methyl glucoside by intact E. coli is only partially inhibited by the high arsenate-low phosphate incubation, and further, part of the transported sugar still appears intracellularly in the phosphorylated form.
The glucose-6-P level remains high in presence of arsenate and this or other intracellular phosphate compounds may be involved in phosphorylation of the transported sugar by an unknown mechanism.
Low concentrations of iodoacetate nearly completely block the a-methylglucoside transport with but little effect on the 02-driven proline transport. tablished since Lipmann elaborated the significance of *4TP made it likely that ATI' could serve as an energy donor for transport.
Factors known to affect transport seemed consistent with this view, including the well recognized sensitivity of transport to inhibitors of oxidative phosphorylat,ion. Furthermore, in accord with such possibilities, oxygen uptake required for the transport of a galactoside molecule into E. coli was estimated as equivalent to that required for synthesis of one ATP molecule (1).
Certain observations, however, suggested t.he possibility that active transport by microorganisms might be driven oxidatively without intervening phosphorylatiorts. Several years ago, Halvorsou and Cowie showed that arsenate would inhibit protein synthesis by yeast, but not inhibit the uptake of phenylalanine (2). An arsenate block of ATP production, but not of membrane energization by oxygen uptake, would explain their data. Also, following earlier demonstrations (3,4), considerable evidence has accumulated showing that 02 uptake by mitochondria or light absorption by rhloroplasts can drive the active transport of ions without intermediary formation of ATP. Mechanistic similarities of the energy-transducing processes in mitochondrial and chloroplast membranes with those of bacterial membranes appear plausible.
In other related studies, Harold et al. more recently demonstrated that uncouplers of oxidative phosphorylation inhibit anaerobically driven transport by Streptococcus faecalis and Escherichiu coli (5, 6). These observations point to the format'ion from ATP of an energized membrane state coupled to transport processes. A unifying hypothesis that emerges is that, in general, either oxidation-reduction reactions or ATP cleavage can serve to energize the membrane, and this high energy membrane state can be coupled to an active transport system (7,8).
The objective of the experiments reported in this paper was to assess in various ways the possibility that accumulation of amino acids, sugars, and ions by E. coli might be directly coupled either to oxygen uptake or to substrate level phosphorylations. A preliminary report of our results indicating that this is indeed the case has appeared (9). Subsequent reports from our laboratory (lo), by Kaback et al. (11,12), by Hirata et al. (13), and by Konings and Freese (14) have presented additional evidence favoring the concept of direct oxidative energization of active transport. tained from Amersham-Searle. /3-Methyl[14C]thiogalactoside was purchased from New England Nuclear Corp. 86Rb was obtained from Oak Ridge National Laboratory.
Luciferin and luciferase were obtained from DuPont, lysozgme from Worthington, and all other enzymes from Sigma. ATP and hDP were purchased from Boehringer and P-enolpyruvate was obt,ained from Calbiochem.
Preparation of Cells-E. coli ML 308-225 (lac i-z-y+a+) was grown at 30" either in 65 inM potassium phosphate, pH 7, with 7 mM ammonium sulfate and 0.4 mM MgSO, (Medium A) or in 50 mm Tris-Cl, pH 7.6, supplemented with 0.5 rnhf potassium phosphate, 7 mM amn1onium sulfate, and 0.01 volume of a stock solution containing per liter the following salts: 50 g of potassium chloride, 42.8 of magnesium chloride, 5 g of sodium citrate, 1 g of ferrous chloride, 1 g of calcium chloride, 1 g of manganous chloride, 1 g of ammonium molybdate, and 0.2 g of cobalt chloride. The carbon source utilized was glucose at 20 mM, unless otherwise specified. Logarithmically growing cells obta.ined from IO-fold dilutions of overnight cultures were quickly chilled and pelleted.
The pellet was washed three times in ice-cold 25 ITIM Tris-Cl, pH 7.6, and resuspended in the same medium to an A650 of 1.0 (0.23 mg of bacterial protein per ml).
Cells were used for the reported experiments irnmediately after resuspension, although only slight loss in transport capacity was observed during storage for up to 6 hours at 0". Transport Assa?g with Intact Cells -Cells resuspended as described were incubated at 30" for 10 min with 80 pg of chloramphenicol per ml and the desired supplements before adding transport solute. For kinetic determiuations of soiute uptake, the radioactive solute was added to a small volunle of cells, permitting adequate oxygen diffusion, yet of sufficient size for convenient removal of a.11 aliquots from the same suspension. Five hundred-microliter total volumes of cells were used, and uptake of solute was followed by diluting 50.~1 aliquots into 5.ml volumes of temperature-equilibrated, transport solut,e-free incubation solution foliowed by rapid filtration on 0.45-p Millipore filters. Dilution and filtration took less than 6 s for completion.
Radioactivity retained by Millipore filters was counted with a Nuclear Chicago liquid scintillation detector after dissolving the filter in Bray's solution, and the specific uptake was calculated after subtracting background counts retained by filters when sampling was performed in the absence of bacteria.
To measure the apparent steady state level of solute accumulation, a single aliquot was assayed as above, 10 min after the addition of radioactive material.
Various points give assurance that the amino acids transported were accumulated in unchanged form. Paper chromatography of cell extracts essent,ially as described by Piperno and Oxender (15) showed that intracellular proline and leucine accumulated during the transport assays are not chemically modified or incorporated into protein.
In studies of carbohydrate transport, the nonmetabolizeable glucose and lactose analogues, a-methylglucoside and fi-thiomethylgalactoside, respectively, were used routinely.
Furthermore, lactose is not metabolized by E. co/i MC308225, a strain which lacks P-galactosidase. Hence, lactose rather than the analogue was ocrasionally employed to assay transport.
In addition, we not.ed that lactose is not phosphorplated during transport as determined by ion exchange chromatography.
*Membrane Vesicle Preparation and Transport Assay-Mernbrane vesicles were prepared as described by Kaback (16), washed two times in the buffer of choice, resuspended in the same to final concentration near 1 mg per ml, and stored at liquid Nz temperature.
To measure transport, vesicles were thawed quickly, brought to 10 mM MgSO4, and incubated for 10 min at 30" with agitation and in the presence of desired addition.
Sampling was done as described for illtact cells. Protein was measured by the Lowry procedure, with bovine serum albumin as a standard.
Xeasuremenl of ATP and P-Enolpyruvate -*4TP was measured by a luciferin-luciferase assay using a DuPont Iliometer. P-Enolpyruvate was measured by the amount of AT1 formed from ,4DP by pyruvate kinase.
To measure ATP, 2 ml of the sarnple t,o be assayed were rapidly mixed with 2 ml of ice-cold 1.2 N HC104 and chilled for 30 min. After centrifugation the decanted supernatant was neutralized, then recentrifuged for clarification.
With the exception of 4-fold dilution of the suggested luciferin-luciferase conccnt,ration, the ATP assay was performed according to the manufacturer's instructions.
The solution was maintained at 25" for 2 min, then chilled to 0". Total ATP present before and after the addition of pyruvate kinase was used to obtain the P-cnolpyruvate co~icentration.
Standard curves for ATP and 1'.enolpyruvate indicated that, the assays were linear for at least three orders of magnitude variation in concentrations. Intcr1lal standards showed the absence of mediumquenching effects on the assay. Measurement of Vesicle Cnpacity Jar Phosphorylation---In duplicate or replicate experiments utilizing a low phosphate medium, membrane vesicles at 1.3 mg per ml were suspended in a medium containing 50 mM glycylglycine, pH 7.2, 10 m&I magnesium sulfate, 1 ELM potassium phosphate, 10 m&t lithium r)-lac:tate, I mM glucose, and 0.05 mg per ml of hexokinase at 0". The suspension wa.s brought quickly to 30" and incubated for 1 min. Carrier-free "Pi was added to give a specific act,ivity of 100,000 cpm per nmole. At various times, 0.4.ml aliquots wrre removed and quickly mixed with 0.6 1111 of 0.5 N HClO+ The precipitate was removed, and the supernatant was miscd with 1 ml of 1.57: ammonium molybdate tetrahydrate in 0.5 s H,SO., at 0". The solution was extracted three times with 2-ml portions of reagent grade C-met.hylpeutarrone. 0.5-ml aliquots of the lower, aqueous layer were dried and counted. Subsequent experiments were essentially identical, but 10 II~I l)otassiunl phosphate (final specific activity 321'i of 100 cpm per umolc) replaced the glycylglycine and low phosphate of the initial czperiments, glucose concentration was increased to 40 I~IM, and ,4Dl' was added to give a final concentration of 2 m&f. To measure experimental backgrourld similar incubations wcr~ made in the presence of 1 In&f 2,4-dinitrophenol. RESUI TS J h'xperiments with Uembra?le T'esicles 0,.linked Transport-For meaningful correlations with Lresicle phosphorylative activity, t.he transport capacit.y of the vesicles needed to be established.
The results given in Fig. 1 show the maximum capacity for active transport of proline by the vesicles used. Transport was stimulated markedly by added I)-lactate, in harmony with results by Kaback and Milner (11) nud was prevented by anaerobic conditions.
When glucose was substituted for nlact'ate, no stimulation of transport or of O2 uptake Fro. 1 i/cjl). Proline uptake by vesicles in phosphate buffer. Membrane vesicles at 1 mg per ml, suspended in 50 mM potassium phosphate, pH B.F, and 10 mnl magnesium sulfate, were assayed for aerobic proline transport as described under "Experimental Procednre." D-Lactate (10 mM) was present where indicated 10 min prior to addition of 10 mM j3Hjproline.
For anaerobic upt,ake measurements, 2 ml of a vesicle suspension in each of four Thunberg tubes were evacrlatcd with a vacuum pump for 60 s at 0". The anaerobic vesicles were incuhat.ed at 30" for 10 min in the presence of 10 mM lithium I)-lactate, 10 mM magnesium sulfate, and 50 mnc potassium phosphate.
[gH]Proline from the side arm was mixed with the vesicles to give a final concentration of 10 pM. At 1, 2, 5, and 10 min following mixing, a different tube was opened and 50 ~1 of sample were quickly withdrawn, diluted into 5 ml of 30" incubation medium lacking proline, and filtered. Sampling took less than 15 s.
FIG. 2 (center). Proline uptake by vesicles in arsenate buffer. Membrane vesicles at 1 mg per ml, suspended in 50 m&r potassium arsenate, pH 6.6, and 10 mM magnesium sulfate, were assayed for proline transport as described under "Experimental Procedlire." u-Lactate at 10 rn~ was added during preliminary incubation to the sample indicated, and j3H]proline was added to 10 PM.

Escherichia coli grown in l&Tedium A and washed as described under "Experimental
Procedure" were resuspended in Tris-Cl buffer, pH 7.6, with the indicated concentrations of arsenate. Following incubation at 30" for 10 min, 10 pM ['%]leucine was added and incubation was continued for 10 min furt.her. Cell suspensions were rapidly mixed with an equal volume of ice-cold 1.0 N HClO, and filtered after chilling for 10 min. Maximum counts ta.ken ttp were 5300. Similar results were cbtained with proline in place of leucine.
KBS observed, indicating the lack of whole cells in the vesicle preparatior,. TABLS I Transport of leucine, glutamate, and alanine by vesicles also wa,? tested. Of these, only transport of alanine was stimulated bv n-lactate, although glutamate was actively accumulated. L&&e transport was very low and not stimulated by D-lactate. Stimulation of all these transport systems by D-lactate has been reported (11) in contrast to the present findings.
Small or negligible capaciiy of melnbrane vesicles for 2,4-dilailrophenol-sensitive uptake of Pi Transport-competent vesicles were incubated for 10 min in presence of n-lactate and 32P, under conditions favorable to oxidat.ive phosphorylation as described under "Experiment.al Procedure." Pi uptake is reported as the difference between sample wit,hout or with 1 mm 2,4-dinitrophenol present.
Masimum capacity for proline accumulation varied with the buffer used. As shown i n Fig. I detect possihle osi&t.ive l'ilosl,florS'latioil capacity in vesicles used for trsnsport studies, a conventional hcxokinase trap as employed in mitochondrial and bacterial particle systems was used. The disappearancc of pcrchloric acid-soluble 321'i in the presence of glucose, hexokinnse, 2nd menlbranc vesicles (see "Experimental Procedure") ins followed under various conditions using 50 rnhf glycylglycine or IO rnnr potassium phosphate buffers. 7 able I surnmarizc.s the results of t.hese experiments.
assay. The vesicles did, however, show a weak capacity for phosphorylation by inorganic phosljhate in the presence of high 321'i levels in the medium that appeared sensitive to 1 mM 2,4dinitrophenol.
This phosphorylation was difficult to measure accurately because it represented only about 0.15%. of the phosphatc of the media. Thus, little or no 2,4-dinitrophenol-~ensitive uptxkc was obsrrvcd.
The nature of the phosphorylated product was not determined. As noted in the table, even if all the observed uptake in the presence of high phosphate were by oxidative phosphorylation, P:O ratio would be low. In addition, with only 1 pM added Pi, under conditions similar to those of the transport assay, phosphorylatiou was only 1 7c or less of the capacity for proline transport.
The nlcJSt irnportattt facet shown by the data of Table I is that under collditiolls permitting active proline transport, but where the I'; present ilk the medium is very low, no detectable phosphorylation by inorganic phosphate occurred in a sensitive ATP und P-Enolpyruvafe Levels-Although n-lactate can stimulate proline tlsnsyort markedIy, it has practically no effect on the small amount of XTP and P-enolpyruvate detected in vesicles by the sensitive fluorometric assay. Table 11 shows that. lvhile proline accumulation was increased by 1.9 nmoles In the absence of Mg++, ATPase activity was decreased by more than 957,. This membrane ATPase activity appears to have no relation to the active transport of proline as added ATP did not stimulate the proline uptake capacity of the vesicles. The apparent lack of ATP coupling in vesicles distinguishes this system from intact cells as will be discussed later.
Arsenate-insensitive Proline Transport-When arsenate, a secondary inhibitor of oxidative phosphorylation (171, is present in high concentrations, vesicles still actively accumulate proline. As can be seen in Fig. 2, vesicles washed and suspended in 50 mM arsenate exhibit lactate-stimulated transport, accumulating proline to the same maximum level obtained by vesicles suspended in phosphate.
The presence of arsenate makes localized Pi uptake highly unlikely.
Arsenate-resistant transport is consistent with active transport being independent of oxidative phosphorylation.

Experiments with Intact Cells
The results with membrane vesicles suggested that amino acid transport by bacteria could be driven by 02 uptake without intervening phosphorglation.
Means of assessing such a possibility with whole cells thus seemed important.
Various approaches were devised to examine the nature of energy coupling in intact cells.
Arsenate Block of Phosphate Activation-The results of Fig. 3 confirm that arsenate is an effective metabolic inhibitor for an ATP-driven process in intact E. coli cells under conditions of our experiments.
In uivo protein biosynthesis, dependent upon a normal ATP supply, was strikingly sensitive to the addition of exogenous arsenate, even at concentrations as low as lo+ tir. Arsenate can thus enter the cells used readily, compete with the residual intracellular phosphate, and sufficiently perturb enei'gy metabolism to eliminate protein synthesis.
Since many anabolic and catabolic processes are known to be sensitive to modest changes in energy charge (18), assessing the effectiveness of arsenate as an inhibitor of intracellular Pi uptake required direct measurement of AT1 and P-enolpyruvate levels after arsenate addition. Fig. 4 illustrates that arsenate quickly and drastically reduced both intracellular ATP and Penolpyruvate.
Repeated measurements with perchloric acid extracts established routine ATP and P-enolpyruvate levels in arsenate-treated cells at 90 to 99°C less than the levels found when Pi replaced arsenate.
At a symposium, we reported that incubation with high arsenate a,ncl low phosphat,e apparemly did not reduce intracellular P-enolpyruvate levels (10). This report is now known to be in error.
In part, assay difficulties arose because extraction of cells with hot aqueous ethanol, 50';; (v/v), did not inactivate adenylate kinase.
Arsenate-insensifive Acfiz~e Transporf-Comrary to its in hibitory effect on protein biosynthesis and phosphorylation, arsenate did not interfere with cellular proline transport.
As shown in Fig. 5, addition of arsenate did not reduce and, indeed, somewhat increased the capacity of cells to accumulate proline. Although the intracellular ATP decreased 98% to a level of 0.15 nmole per mg of bacterial protein, the maximum steady state accumulation of proline reached a level of 13.8 nmoles per mg of bacterial protein.
The rate of proline upta.ke was approximately the same in the presence and absence of arsenate.
An additional important point showed by this experiment is that a relatively long exposure to arsenate prior to testing transport did not lower the total accumulation.
Arsenate-insensitive transport is not restricted to proline or other amino acids. Table III shows that various classes of transport ligands are accumulated in the presence of high arsenate. The lactose analogue, p-methylthiogalactoside, and t,he glucose analogue, ol-methylglucoside, as well as proline and leutine were accumulated against a gradient in the relative absence of intracellular phosphate bond energy. Variable sensitivity among the transport system exists, although, with the maximal inhibition of leucine and a-methylglucoside transport frequently exceeding 50$$. The continued transport of various solutes did not result because of insufficient concentrations of arsenate; data for cy-methylglucoside are shown in Fig. 6.
Efects of Arsenate, NaF, and Zodoacefafe on mMethylglu~oside Transport-The observation that cr-methylglucoside could be accumulated by cells having almost no P-enolpyruvate is somewhat surprising, as the transport of hexoses in h'. coli has been documented extensively as proceeding via a P-enolpyruvat,e phosphotransferase complex (19). Furthermore, as seen in Fig.  7, cells treated with arsenate appear to establish an intracellular pool of a-methylglucoside phosphate although very little Penolpyruvate is present. In addition, cells mainta,ined in arsenate medium for extended periods of time contain a near normal level of glucose-B-P, as determined by assay with TPN+ aud glucose-6-P dehydrogenase.
Further atternpts to inhibit cu-methylglucoside transport by reducing intracellular P-enolpyruvate lower than the level obtained by adding arsenate have been unsuccessful. Sodium fluoride, an inhibitor of the enolase reaction, had no effect on (Ymethylglucoside transport when added at concentrations as high as 100 InM.
Differential inhibition of cu-methylgluroside and proline transport was obtained with iodoacetate.
Low concentrations of iodoacetate are known to inhibit glyceraldehyde 3-phosphate dehydrogenase and thus glucose breakdown.
The inhibition of cr-methylglucoside but not of proline or lactose transport is shown in Fig. 8 In contrast to the other transport systems tested, Rb+ transport was completely abolished by exposing glucose-grown cells to arsenate as shown in Fig. 9. 2,4-IXnitrophenol at 1 mM also blocked accumulation of Rb+. L1dditional interesting aspects of the Rb+ system are its responses to the carbon source of the growth medium.
As seen in Fig. 9 Table  IV shows that when a known uncoupler of oxidative phosphorylation, 2,4-dinitrophenol, is added to cells suspended in arsenate, proline is no longer accumulated.
O2 Requirement 01 Arsenate-insensitive Transport-As illustrated in Fig. 10, another significant characteristic of the arsenateinsensitive proline transport, in addition to being abolished by uncouplers, is that it requires OZ. Anaerobiosis eliminates the capacity of cells to couple transport oxidatively.
Accordingly, although proline transport is insensitive to the addition of arsenate under aerobic conditions, it is nearly abolished by arsenate under anaerobic conditions. Neither glucose nor n-lactate stimulated transport under these conditions. A significant point is that the residual ATP measured after anaerobic incubation in presence of arsenate was only 0.25 to 0.5 nmole per mg of bacterial protein.
Thus arsenate reduced ATP approximately the same in cells maintained aerobically or anaerobically.
This observation gives additional evidence against the possibility that aerobic, arsenate-insensitive proline transport reflects insufficient ATP reduction. Because anaerobic cells will accumulate proline in a process sensitive to the addition of arsenate, phosphate bond energy At 30 s, 5 min, and 10 min, l-ml aliquots were withdrawn and 0.5 ml was pipetted into 10 ml of Tris-Cl at. 37", and 0.5 ml was pipetted into 10 ml of ice-cold 37 rnnl BaBrz in 80% ethanol.
The sample in Tris buffer was filtered rapidly through a Millipore filter and washed with another 10 ml of Tris, while the BaBr?-treated sample was filtered after 10 min at 0" and the precipitate was washed with 10 ml of ice-cold 8O7o et.hanol.
The 14C retained in filt.ration wit.hout prior BaBrPethanol precipitation was taken as a measure of total intracellular a-methylglncoside plus a-methylglucoside phosphate, and the 14C in the BaBr2-ethanol precipitate was taken as a measure of intracellular a-methylglucoside phosphate. The data of Fig. 11A illustrate the sensitivity of anaerobic proline transport to 2,4-dinitrophenol. Azide, also an uncoupler, had a similar effect.
As can be seen from Fig. llB, the peculiar and as yet unexplained stimulation of oc-methylglucoside transport in response to 2,4-dinitrophenol (see Table  III  The transport of glucose only partially satisfies the predictions, reflecting now1 aspects of glucose transport by E. coli. Inhibition of oxidative input is readily achieved by establishing anaerobic conditions. As noted in Fig. 10 proline uptake by intact cells proceeds at a near-normal level in the absence of osygen Vesicles, on the other hand, do not transport well anaerobically, even in the presence of ATP (see Fig. 1). Possibly a coupling factor or factors have been lost during the vigorous disruption and estensire washing required for membrane vesicle preparation.
The inability of vesicles to use ,4TP or other phosphates to stimulate transport of proline, although a significant ATPase activity is present (25), has been noted earlier (9, 11).
Ulocking phosphorylative input is more difficult than blocking osidative input, and requires careful assessment.
In our first approach, sensitive measurement of membrane vesicle capacity for osidative phosphorylation in presence of n-lactate and low Pi concentration revealed little or no uptake of added Y'i (Table  I). In the low phosphate medium, (1 FM added Pi) as used for transport assay, the molar capacity of the vesicles to produce ATE' was at least two orders of magnitude less than the molar rapacity to accumulate proline.
With much more Pi present (10 mnr), a small 2,4-dinitrophenol-sensitive uptake was observed (Table I) producing sufficient ATP to account stoichiometrically for the proline transport observed in osmotically prepared vesicles. The important result reported here is that n-lactate oxidation in our system greatly stimulates transport under conditions where capacity for Pi uptake is negligible.
Thus in the vesicles, oxidative input for active transport may function near normally, but capacity for phosphorylative input from ATP is negligible. Another approach to blocking phosphorylative input is to eliminate the intracellular energy-rich phosphates, rather than to disrupt the membrane coupling sites. The known uncoupling action of arsenate made this P; analogue a likely reagent for reducing intracellular energy-rich ph0sphat.e pools. However, the relatively poor ability of arsenate to compete with Pi and other factors known to modify arsenate inhibition (29), as well as the inability of arsenate to enter certain strains of E. coli (30), could reduce the effectiveness of arsenate.
Adequate assessment of arsenate effects under our conditions thus seemed essential.
The results show that incubation with high arsenate and low Pi concentrations markedly reduces intracellular ATP and P-enolpyruvate levels (Fig. 4). Indeed, metabolically available levels may be reduced even more drastically than indicated by our data because perchloric acid extraction will also remove firmly but noncovalently bound nucleotides that might be present.
Loss of phosphate bond energy is also indicated by the complete inhibition of protein synthesis by low concentrations of arsenate.
However, even with little or no phosphorylative capacity retained, proline transport proceeds normally, as predicted by the model given in Fig. 13. Under such conditions, leucine, P-methylthiogalactoside, Rb+, and ol-methylglucoside uptake all occur at nearly half or more of the level found in the absence of arsenate.
Partial loss of some transport capacities in presence of arsenate does not appear to result from incomplete disruption of phosphorylations; increasing concentration and time of arsenate exposure has limited effect on arsenate-insensitive transport (see Fig. 6).
The differential effects of arsenate on anaerobic and aerobic transport are instructive. Arsenate allows normal proline transport in aerobic cells, but compietely eliminates proline accumulation ill anaerobic cells, although the same low level of ATP is found under each condition (see Fig. 10). This provides additional strong evidence that the transport dependent on O2 is not occurring via ATP generation.
'l'he arsenate-irlsellsitive process requires oxygen and the oxygen-independent process requires generation of phosphate bond energy. In harmony with the model of Fig. 13, either oxidative or phosphorylative Energy is a minimal requirement for energizing transport.
Results with the hTI'ase inhibitor I\', N-dicyclohexylcarbodiimide are it 1 complete accord with this conclusion (see Fig. 12).
Transport dependent on phosphorglative input and transport dependent on oxida.tive input are ea.ch sensitive to the action of uncouplers.
Although the molecular interpretation of uncoupler action remains an area of considerable speculation, unrouplers functionally appear to prevent the formation of, or cause the dissipation of, high energy membrane states. Thus, the action of uncouplers in eliminating amino acid, fl-mcthylthiogalactoside, and Rbf transport under a variety of conditions gives strong support for the requirement of a high energy membrane state to couple energy input with many transport systems. 1)espite the considerable experimental evidence favoring the exclusive participation of the P-cnolpyruvate phosphotransferase system in accumulating certain hesoses in several bacterial species (31, 32), experiments showing the uptake wmethylglucoside with drastically lowered intraceilulnr P-enolpyruvnte and ATP levels (see Figs. 4,6,7,and Table III) suggest consideration of the possibility of ar-methylglucoside entry in the free form followed by phosphorylation with an intracellular donor other than P-enolpyruvate.
For example, phosphoryl transfer from glucose-6-P to various hexoses can be catalyzed by an enzyme preparation from E. coli (33). Present information leaves unresolved, however, the source of the phosphate for the formation of the phosphorylated oc-methylglucoside. Cells transferred to arsenate media still have considerable amounts of various intracellular phosphate present, and one or more of these by transfer reactions might be involved.
The mechanism by which oxidative uptake is coupled to active transport is unknown.
Similar or identical energized compounds or states of membranes appear to be involved in both oxidative phosphorylation and in active transport. The means of coupling oxidative energy to transport thus joins oxidative phosphorglation in presenting the problem of determining the nature of high energy membrane states.