Different Mechanisms of Energy Coupling for the Active Transport of Proline and Glutamine in Escherichia Coli

Proc. Nat. Acad. Sci. USA Vol. 70, No. 5, pp. 1514-1518, May 1973

Different Mechanisms of Energy Coupling for the Active Transport of Proline and Glutamine in Escherichia coli (inhibitors/mutants/ATP/energized membrane state/starvation)

EDWARD A. BERGER Section of Biochemistry, Molecular and Cell Biology, Wing Hall, Cornell University, Ithaca, New York 14850 Communicated by Leon A. Heppel, March 15, 1973 ABSTRACT The ability of either glucose or D-lactate to the ATPase inhibitor NN-dicyclohexylcarbodiimide (4), to energize active transport of amino acids in E. coli was as well as the loss of thiomethyl- studied in starved cells blocked at specific sites of energy from cyanide-resistant metabolism. Proline uptake could be driven by either galactoside uptake in an ATPase mutant (5). oxidative or substrate-level processes. The oxidative path- Energy for active transport can thus be derived indepen- way was sensitive to cyanide but not to arsenate, and dently from either respiration or ATP hydrolysis. The ob- operated normally in a mutant deficient in the Ca, Mg- servations (4, 7) that uptake driven by either pathway is dependent ATPase. The substrate-level pathway, which was active with glucose but not with D-lactate as the car- sensitive to uncouplers of oxidative phosphorylation has led bon source, was sensitive to arsenate but not to cyanide, several workers (4-8) to propose that a high-energy membrane and required a functional ATPase. Uncouplers prevented state is the immediate energy donor for bacterial transport, the utilization of energy for proline uptake by either path- though other models have recently been formulated (3, 9). way. Since this conclusion is based upou studies with only a few Energy coupling for glutamine uptake was quite different. The oxidative pathway was sensitive to cyanide and uptake systems, it is necessary to identify the energy donors uncouplers and, in contrast with proline, required an for the active transport of other metabolites before any active ATPase. The glycolytic component was resistant to generalizations can be applied. cyanide and uncouplers, and functioned normally in the In this study, the ability of different carbon sources to ATPase mutant: Arsenate abolished glutamine transport provide energy for active transport was investigated in energized by either pathway. The results suggest that proline transport is driven starved E. coli cells blocked at specific sites of energy me- directly by an energy-rich membrane state, which can be tabolism. By the judicious choice of energy sources and in- generated by either electron transport or ATP hydrolysis. hibitors, it is possible to distinguish whether respiration per se, Glutamine uptake, on the other hand, is apparently the energized membrane state, or phosphate-bond energy is driven directly by phosphate-bond energv formed by way of oxidative or substrate-level phosphorylations. the obligatory energy donor for a particular system. For example, glucose can provide ATP by either oxidative phos- Early investigations of bacterial active transport have im- phorylation, which requires both electron transport and a plicated the high-energy phosphate bond in the energy- functional Ca,Mg-ATPase, or by the substrate-level phos- coupling process (1, 2). Recent evidence, however, suggests phorylations of glycolysis, which require neither process. that the role of ATP is probably indirect. The extensive Glucose can also give rise to an energy-rich membrane state studies of Kaback and his colleagues (3) demonstrated that by two pathways: through hydrolysis of glycolytic ATP by membrane vesicles incapable of oxidative phosphorylation can ATPase in the presence or absence of respiration, or through still use respiration to drive the uptake of a wide variety of the oxidations of the respiratory chain, which may occur in amino acids and sugars. Klein and Boyer recently showed (4) the absence of ATPase. Alternatively, D-lactate is oxidized that aerobic proline transport in intact cells of Escherichia directly by a membrane-bound dehydrogenase coupled to the coli is retained under conditions where intracellular ATP and cytochrome chain (3), and can provide energy only in the phosphoenolpyruvate levels are drastically reduced by presence of electron transport. The synthesis of ATP re- arsenate. Furthermore, a functional Ca,Mg-dependent quires an active ATPase, whereas the generation of the ATPase is not required for aerobic accumlation of thiomethyl- energized membrane state does not. galactoside (5) or proline (4, 6), confirming that transport The results presented here confirm the suggestions of can proceed independently of oxidative phosphorylation. others (4, 6) that proline uptake is driven by the energized While it is clear that the respiration-linked uptake of certain membrane state. Glutamine transport, on the other hand, is substrates does not involve the formation or use of high-energy driven directly by phosphate-bond energy formed by either phosphates, these same transport systems can apparently be oxidative phosphorylation or glycolysis. energized by an alternate nonoxidative mechanism that MATERIALS AND METHODS uses ATP. E. coli is able to accumulate various substrates anaerobically (4, 7), and it has been shown for proline that Bacterial Strains. E. coli ML 308-225 and its derivative anaerobic uptake is abolished by arsenate (4). The essential DL-54 were the generous gifts of Dr. Robert D. Simoni. role of the Ca, Mg-ATPase in the use of ATP for transport is DLh54 is missing more than 95% of the Ca,Mg-ATPase inferred from the sensitivity of anaerobic proline accumulation activity and is unable to grow on carbon sources that require oxidative phosphorylation for ATP formation (6). Abbreviation: FCCP, carbonyl cyanide-p-trifluoromethoxy- Chemicals. L- [ U_14C]proline and - [ U-'4C]glutamine were phenylhydrazone. purchased from New England Nuclear Corp. The isotopes 1514 Downloaded by guest on October 1, 2021 Proc. Nat. Acad. Sci. USA 70 (1973) Energy Coupling Mechanisms for Active Transport 1515

were diluted with nonradioactive amino acids to final specific the endogenous rates of proline uptake in unstarved cells activities of 20-25 Ci/mol. 2,4-Dinitrophenol was purchased were so high as to partially or completely mask the effects from Sigma. Carbonyl cyanide-p-trifluoromethoxyphenyl- of glucose or D-lactate. Glutamine uptake in the absence of hydrazone (FCCP) was a gift from Dr. Efraim Racker. All added carbon source was not as high, suggesting that the two other compounds were analytical grade. transport systems might be using different energy stores. Several starvation methods were examined for their ability Growth of Cells. Bacteria were grown in a synthetic, phos- to deplete endogenous energy reserves. Vigorous overnight phate-buffered minimal medium (10) supplemented with 11 aeration of the cells at 370 in the presence or absence of thio- mM glucose as a carbon source. Optical density at 600 nm methylgalactoside failed to reduce endogenous rates of was monitored with a Gilford spectrophotometer model 240 proline transport significantly. Incubation with a-methyl- (OD600 of 1 corresponded to 109 cells per ml). Cultures were glucoside in the presence of sodium azide (14) was more innoculated to an OD of about 0.2 and were allowed to grow effective, though the inhibitory effects of azide on transport at 370 on a gyratory shaker for about two generations. The could not be completely reversed by washing. The most cells were harvested by centrifugation, washed twice with satisfactory results were obtained with the uncoupler dini- minimal medium at 230, and suspended in 20 ml of minimal trophenol. Table 1B shows that incubation of ML 308-225 medium per g wet weight of cells. This suspension was used with 5 mM dinitrophenol for 10 hr at 370, followed by ex- directly for transport experiments with unstarved cells. tensive washing, reduced the endogenous rates of proline Starvation of Cells. Washed cells were suspended in minimal uptake sufficiently to permit large stimulations by added medium containing 5 mM dinitrophenol at a density of 1 g energy sources. Furthermore, the transport rates in the (wet weight) of cells per 200 ml. The suspension was incubated presence of glucose and D-lactate were very similar to those at 370 with shaking for 10 hr for ML 308-225, or for 1 hr. with observed with unstarved cells (Table 1A). DL-54. The cells were then isolated by centrifugation, washed DL-54 was much more sensitive to this starvation procedure three times in minimal medium, and suspended in 20 volumes than the wild type. Incubations as short as 30 min greatly of minimal medium per g of cells for transport assays. reduced endogenous rates of proline uptake in this strain, while they had little effect on ML 308-225. If the dinitro- Transport Assay. A reaction flask containing cells (100-200 phenol treatment was continued beyond 2 hr with the mutant, ,ug of protein), chloramphenicol (80,4g/ml), and inhibitors- endogenous uptake rates were reduced below detectable where designated-was incubated at 370 for 5 min. 11 mM limits, but the rates with added energy sources also began Glucose or 10 mM D-lactate were added and the incubation to decline. A 1-hr incubation was therefore chosen for D1h54, was continued for 10 min. The flask was brought to room since as shown in Table 1B, endogenous proline transport temperature and the transport reaction was initiated by the was adequately depleted and the rates of glucose- and D- addition of labeled amino acid to a concentration of 10 AM. lactate-supported uptake of both proline and glutamine were The final volume of the mixture was 0.5 ml. At various times, similar to those in unstarved cells. 0.2-ml aliquots were withdrawn, filtered on 25-mm nitro- The dinitrophenol starvation procedure thus permitted cellulose filters (0.45 Mm, Matheson-Higgins), and washed assessment of the ability of different energy sources to stimu- with 10 ml of a solution containing 0.01 M Tris HCl (pH late transport, and starved cells were used in the remainder 7.3)-0.15 M NaCl-0.5 mM MgCl2 (11) at 230. The filters of the experiments. were dried and counted in 7.5 ml of a solution of 15 g of 2,5-diphenyloxazole and 0.2 g of 1,4-bis[2-(4-methyl-5- Transport in DL-54. The membrane-bound Ca, Mg-ATPase phenyloxazolyl) ]benzene dissolved in 3.8 liters of toluene. is believed to catalyze the reversible interconversion of ATP To insure linearity of initial rates of uptake, 15-sec and 30-sec and the energized membrane state (15). Table 1B shows that time points were taken, and the 15-sec points were used for loss of this activity by mutation had marked effects on trans- the calculations. Uptake values are expressed as nmol/min port, but only under certain conditions. In wild-type cells, per mg of cellular protein. D-lactate supported a relatively high rate of glutamine up- For studies on the effects of arsenate, starved cells were take, while in DL-54 this uptake was greatly impaired. D- washed three times with 25 mM Tris * HCl (pH 7.3) at 40, and TABLE 1. Aminoacid transport in unstarved and starved cells of suspended at 23° in a minimal medium (12), modified by replacing the potassium phosphate with 0.1 M Tris * HCl ML 308-225 and DL-54 (pH 7.0) (referred to as phosphateless medium). Transport Initial rate of Initial rate of assays were performed as described above, except that the proline uptake glutamine uptake reaction flask contained phosphateless medium instead of the (nmol/min per mg) (nmol/min per mg) normal minimal medium, and the cells were incubated at 370 ML ML with arsenate for 15 min before addition of the carbon source. Amino acid Addition 308-225 DL-54 308-225 DL-54 Protein . ssay. The content of cell suspensions was protein A. Unstarved cells determined by a micromodification of the method of Lowry Proline None 2.26 2.37 0.37 0.13 et al. (13), with bovine-serum albumin as a standard. Glucose 5.78 2.42 5.52 5.69 RESULTS 1)-Lactate 3.74 2.83 2.62 0.24 B. Starved cells Starvation of the Cells. To study transport driven by a None 0.11 0.67 0.08 0.11 particular carbon source, endogenous energy stores must be Proline Glucose 4.05 2.41 5.78 5.20 sufficiently low so that the bulk of the measured uptake is -Lactate 3.29 3.19 4.21 0. 29 supported by the added compound. In Table 1A, I show that Downloaded by guest on October 1, 2021 1516 BiochemistryP Berger Proc. Nat. Acad. Sci. USA 70 (1973) Mg-ATPase, since DL-54 displayed a comparable level of Praline - DL - 54 glucose-dri ven glutamine uptake in the presence of cyanide. 3 This result supports the view that phosphate-bond energy 2 can directly drive glutamine transport. The Effects of Uncouplers. Uncouplers of oxidative phos-

' O phorylation dissipate the energized membrane state and would YOc 10 20 30 be to inhibit those transport systems that use Oe o thus expected _3 E this state directly as a source of energy. Systems driven 0 Glutomine - DL - 54 5'. directly by phosphate-bond energy should be strongly in- <0a 4 hibited under conditions where oxidative phosphorylation is 3 - the primary source of ATP, but should be relatively resistant 2 where substrate-level phosphorylation is the major source. As shown in Fig. 2, proline uptake driven by glucose or D- lactate was abolished by the uncoupler FCCP in both ML 0 20 30 K) 20 30 308-225 and DL-54. Glutamine uptake with D-lactate as the KCN (mM) energy source was also quite sensitive. As expected, however, FIG. 1. The effects of cyanide on aminoacid uptake in ML glucose was able to support a substantial rate of glutamine 308-225 and DL-54. Additions: A A, none; @-- , glucose; transport in DL-54 in the presence of high levels of FCCP, O-O, D-lactate. again implicating a direct role of phosphate-bond energy in this uptake system. FCCP had more drastic effects on glucose- Lactate-driven proline transport, however, was nearly identical drive glutarnine uptake in ML 308-225 than in DL-54. This in the two strains. Phosphorylation is thus necessary for finding likely reflects the ability of uncouplers to enhance the glutamine, but not for proline, uptake.* With glucose as the hydrolytic activity of the ATPase, a phenomenon known to carbon source, glutamine transport in DL-54 was nearly identi- occur in both mitochondria (17) and chloroplasts (18). cal to that in the wild type (Table 1B). Energy derived from As shown in Fig. 3, dinitrophenol produced results similar to glycolysis can thus drive glutamine uptake in the absence of those with FCCP. Furthermore, cyanide had relatively little the Ca, Mg-ATPase. Glucose-driven proline transport was effect on the dinitrophenol-resistant glutamine uptake in only partially reduced in DI54 (Table 1B). DL-54 (data not shown). In the remaining experiments, the effects of various meta- bolic inhibitors were examined to see whether they support The Effects of Arsenate. Incubation of E. coli cells with ar- the interpretation that phosphate-bond energy can directly senate causes a drastic reduction of the intracellular ATP drive glutamine transport, and to distinguish whether respira- (and phosphoenolpyruvate) levels (4). Transport systems tion per se or the energized membrane state derived from it that use high-energy phosphates directly as a source of energy (or from ATP hydrolysis) is the primary energy donor for should be severely inhibited by arsenate, whereas those driven proline uptake. by the energy-rich membrane state should be relatively resistant, unless this state is derived from ATP hydrolysis. As The Effects of Cyanide. Inhibitors of respiration would be shown in Fig. 4, glutamine uptake was nearly completely in- expected to block uptake processes under conditions where energy is derived from electron transport, but should have relatively little effect on transport supported by substrate- level phosphorylations. Fig. 1 shows that this was indeed the 4 4 - Proline - DL - 54 case. Cyanide abolished D-lactate-driven proline uptake in I Proline ML308-225 both the wild type and the ATPase mutant, while glucose- 3

driven proline uptake in ML 308-225 was only partially 2 2 sensitive. Furthermore, cyanide completely inhibited glucose- supported proline transport in DL-54. Thus, in contrast with < 08 findings in membrane vesicles (3), proline uptake in whole 2 cells can occur in the presence of high levels of cyanide. This a o 2 4 6 8 2 4 6 8 process requires a nonoxidative pathway for ATP formation C0 c (glycolysis), as well as a functional Ca, ltg-ATPase for gen- Eo Glutomine - ML308-225 eration of the energized membrane state from ATP. The

need for the ATPase for energy coupling in the absence of 4~ electron transport has been noted for the uptake of proline (4) and f-galactosides (5).

As shown in Fig. 1, D-lactate-driven glutamine uptake was it n, _ , completely inhibited by cyanide. Glucose, on the other hand, 2 4 6 8 2 4 6 8 supported a substantial rate of cyanide-resistant glutamine FCCP (1±M) transport. Utilization of this energy did not require the Ca, FIG. 2. The effects of FCCP on aminoacid uptake in ML 308- 225 and DL-54. Since the stock solution of FCCP contained 95% * Since growth on glucose is known to repress oxidative phos- ethanol, ethanol was added to each uptake flask to a final con- phorylation in certain strains of E. coli (16), the experiments centration of 4%. This concentration of ethanol alone had no reported here were repeated with glycerol-grown cells with effect on either proline or glutamine transport. Additions: essentially identical results. *-*, glucose; O-O, D-lactate. Downloaded by guest on October 1, 2021 Proc. Nat. Acad. Sci. USA 70 (1973) Energy Coupling Mechanisms for Active Transport 1517

hibited by arsenate in both the wild type and the ATPase 4 - I h Proline - ML 308 -p25 Proline - DL - 54 mutant, with either glucose or 1-lactate as the energy source. 3 The effects on proline transport were much less severe (Fig. 3 4). With glucose as the energy source, a significant portion of 2 2 proline uptake in ML 308-225 was inhibited by arsenate, whereas in DL-54, which presumably cannot form the energized membrane state from ATP, arsenate had no effect. c a d Furthermore, glucose-driven proline uptake in the wild type 02 0-4 ML3u | 02 Q4 06 08 1.0 in the presence of arsenate was nearly completely abolished 8 < Glutomine - MIL309-S25 Shilamine - DL - 54 OzxE by cyanide (Fig. 5). This result confirms the suggestion that zZ Eo the cyanide-insensitive component of proline transport in 6 6 ML 308-225 (see Fig. 1) is energized by the hydrolysis of 41 4 glycolytic ATP. Klein and Boyer reached similar conclusions of the basis of the oxygen requirement for arsenate-resistant 2 2 proline accumulation (4). As expected, arsenate had only Oi minor effects on -lactate-driven proline uptake (Fig. 4). Q2 0.4 0.6 0.8 -.0 02 0.4 Q6 0.8 1.0 DINITROPHENOIL (mM) DISCUSSION FIG. 3. The effects of dinitrophenol on aminoacid uptake in The data presented here are consistent with the scheme shown ML 308-225 and DL-54. Additions: *-*, glucose; O-O, in Fig. 6. Energy for proline transport can be derived from n-lactate. either electron transport or glycolysis. The oxidative pathway is sensitive to cyanide and uncouplers, does not require the reasonable explanation is that the cyanide- and uncoupler- Ca, Mg-ATPase, and is resistant to arsenate. The cyanide- sensitive component of glucose-driven glutamine uptake in resistant component is abolished by both uncouplers and DL-54 arises from a quantity of ATP formed by oxidative arsenate, and requires an active ATPase. The two pathways phosphorylation via residual ATPase activity in the mutant. converge in the presence of the Ca, Mg-ATPase to generate Indeed, n-lactate is able to stimulate glutamine transport the energized membrane state that can drive proline uptake. in DL-54 slightly (Table 1). If the ATP requirements for In contrast, glutamine transport is apparently driven di- maximum uptake are small, it is possible that a low rate of rectly by phosphate-bond energy, which can be formed by oxidative phosphorylation could provide enough ATP to either oxidative or substrate-level processes. The oxidative support a substantial level of glutamine transport with glu- pathway is sensitive to cyanide and uncouplers, but in con- cose, but not with n-lactate, as the carbon source. trast to proline, requires the Ca, Mg-ATPase. The glycolytic Simoni and Shallenberger (6) have reported that in mem- portion is resistant to cyanide and uncouplers, and does not brane vesicles; the transport of proline and alanine driven require the ATPase. Arsenate prevents the utilization of by n-lactate is markedly reduced in DL-54. These findings energy for glutamine uptake by either pathway, presumably led Hong and Kaback (9) to suggest that this strain might by preventing the synthesis of ATP. possess a secondary or polar mutation affecting an energy- A major portion of glucose-supported glutamine uptake transfer coupling locus, an interpretation that seems un- in DL-54 is sensitive to both cyanide (Fig. 1) and uncouplers likely in view of my finding that intact cells of DL-54 show (Figs. 2 and 3). It is unlikely that this sensitivity represents normal lactate-supported proline transport. An alternative side effects of these reagents on the formation of glycolytic explanation for the defect in vesicles is provided by the recent ATP, since the inhibitions plateau with increasing inhibitor concentrations, and the effects are observed with the chemi- cally dissimilar reagents, cyanide, FCCP, and dinitrophenol. 100 Nor does this resistant uptake seem to reflect a component of transport driven directly by the energy-rich membrane state, since D-lactate-driven glutamine uptake in DL-54 is b:g 0 extremely low under conditions where this carbon source can 4 n 4 Gk5wnt5A.*w-5o 64 240 support a full level of proline transport (Table 1B). A more Glucose An, 1 5 10 15 KCN (ff) FIG. 4 (left). The effects of sodium arsenate on aminoacid uptake in ML 308-225 and DL-54. Assays were performed in phosphateless medium. Open bars, no arsenate; solid bars, 0.5 mM arsenate. FIG. 5 (right). The effects of cyanide on the arsenate-resistant portion of glucose-driven proline uptake in ML 308-225. Assays were performed in phosphateless medium. After the cells were incubated at 370 with 0.5 mM sodium arsenate for 10 min, the designated level of KCN;3 was added and the incubation was FIG. 6. A scheme of energy flow for the active transport of continued for 5 min. Glucose was then introduced and the in- glutamine and proline. "I" Represents the energized membrane cubation was allowed to proceed for an additional 10 min before state. initiation of the transport reaction with labeled proline. Downloaded by guest on October 1, 2021 1518 Biochemistry: Berger Proc. Nat. Acad. Sci. USA 70 (1973)

experiments of Bragg and Hou (19). From studies of aerobic strain NR 70, another ATPase mutant, is markedly stimulated and ATP-driven transhydrogenase reactions, these workers by N,N-dicyclohexylcarbodiimide. I have made similar observa- tions in DL-54 (unpublished information), consistent with the concluded that in addition to its catalytic functions, the Ca, hypothesis that the transport defects in vesicles from this strain Mg-ATPase of E. coli also performs a structural role in stabi- are due to loss of the Ca,Mg-ATPase (and its structural function) lizing the high-energy intermediate. Furthermore, the modi- during vesicle preparation. fied ATPase of DL-54 appeared to be more readily lost from These experiments were performed in the laboratory of Dr. the membrane than is the wild-type enzyme. The loss of this L. A. Heppel, who was a constant source of encouragement and structural molecule during the preparation of membrane advice. Dr. R. D. Simoni kindly donated the bacterial strains vesicles could readily account for the differences in D-lactate- used in this study. I thank Dr. D. B. Wilson, Dr. G. Schatz, Dr. driven proline uptake between vesicles and whole cells. I E. Racker, and Mr. E. Hertzberg for their valuable discussions have repeated many of the transport experiments described and critical appraisals of this manuscript. The expert technical assistance of Miss Amy Shandell and Mrs. Anat Bromberg is here with another ATPase mutant, AN 120, and its parent, gratefully acknowledged. This research was supported by Grant AN 180 (20), and obtained very similar results. GB-27396X from the National Science Foundation and Grant The difference in the mode of energy transduction for glu- AM 11789-05 from the National Institutes of Health. tamine and proline uptake raises the possibility that these 1. Scarborough, G. A., Rumley, M. K. & Kennedy, E. P. two uptake processes proceed by entirely different molecular (1968) Proc. Nat. Acad. Sci. USA 60, 951-958. mechanisms. Mitchell has suggested (21) that an electro- 2. Schachter, D. & Mindlin, A. J. (1969) J. Biol. Chem. 244, chemical potential created by the extrusion of protons during 1808-1816. 3. Kaback, H. R. (1972) Biochim. Biophys. Acta 265, 367-416. either respiration or ATP hydrolysis can serve as the driving 4. Klein, W. L. & Boyer, P. D. (1972) J. Biol. Chem. 247, force for the transport of nonelectrolytes in bacteria. The 7257-7265. finding that entry of lactose into E. coli is accompanied by 5. Schairer, H. U. & Haddock, B. A. (1972) Biochem. Biophys. the influx of protons (22, 23) lends support to this concept. Res. Commun. 48, 544-551. A potential gradient model could equally apply to proline 6. Simoni, R. D. & Shallenberger, M. K. (1972) Proc. Nat. Acad. Sci. USA 69, 2663-2667. transport, though alternative hypotheses are certainly pos- 7. Pavlosova, E. & Harold, F. M. (1969) J. Bacteriol. 98, 198- sible. Similarly, several models could explain how phosphate- 204. bond energy drives glutamine uptake. Direct phosphorylation 8. Kashket, E. R. & Wilson, T. H. (1972) J. Bacteriol. 109, of the carrier molecule, a mechanism that has been elabo- 784-789. 9. Hong, J. S. & Kaback, H. R. (1972) Proc. Nat. Acad. Sci. rated for sodium uptake in erythrocytes (24), is an attractive USA 69, 3336-3340. hypothesis that is currently being explored. E. coli MOX19, 10. Tanaka, S., Lerner, S. A. & Lin, E. C. C. (1967) J. Bacteriol. which is unable to transport substrates of the phosphoenol- 93, 642-648. pyruvate-phosphotransferase system (25), has normal glu- 11. Anraku, Y. (1968) J. Biol. Chem. 243, 3128-3135. tamine uptake. 12. Davis, B. D. & Mingioli, E. S. (1950) J. Bacteriol. 60, 17-28. 13. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. Another question that obviously arises is the identity of J. (1951) J. Biol. Chem. 193, 265-275. the energy donors for the myriad of different permeases in 14. Koch, A. L. (1971) J. Mol. Biol. 59, 447-459. bacteria. In general, it has been observed (26) that aminoacid 15. Harold, F. M. (1972) J. Bacteriol. Rev. 36, 172-230. uptake systems in E. coli fall into at least two broad cate- 16. Hempfling, W. P. (1970) Biochem. Biophys. Res. Commun. 41, 9-15. gories, those whose activities are sharply reduced by osmotic 17. Lardy, H. A. & Wellman, H. (1952) J. Biol. Chem. 195, shock and those that are more tightly associated with the 215-224. plasma membrane. The glutamine permease is a well-charac- 18. McCarty, R. E. & Racker, E. (1968) J. Biol. Chem. 243, terized shock-releasable system whose activity depends upon 129-137. a periplasmic binding protein (27, 28). By contrast, proline 19. Bragg, P. D. & Hou, C. (1973) Biochem. Biophys. Res. Commun. 50, 729-736. uptake is resistant to shock (26), and the corresponding 20. Butlin, J. D., Cox, G. B. & Gibson, F. (1971) Biochem. J. binding protein is extracted from the membrane only with 124, 75-81. difficulty (29). Membrane vesicles display very active uptake 21. Mitchell, P. (1970) "Organization and control in proof proline, but not of glutamine (30). Preliminary evidence karyotic and eukaryotic cells," XXth Symp. Soc. Gen. from this laboratory indicates that other Microbiol. pp. 121-166. aminoacid transport 22. West, I. C. (1970) Biochem. Biophys. Res. Commun. 41, 655- systems associated with shock-releasable binding proteins 661. derive energy directly from the high-energy phosphate bond, 23. West, I. C. & Mitchell, P. (1972) Biochem. J. 127, 56P. whereas other tightly-bound systems are coupled to the ener- 24. Post, R. L. (1968) in Regulatory Functions of Biological gized membrane state. In addition, I have observed (un- Membranes, ed. Jarnfelt, J. (Elsevier, Amsterdam), pp. 311-340. published information) that the tightly-bound aminoacid 25. Fox, C. F. & Wilson, G. (1968) Proc. Nat. Acad. Sci. USA permeases are irreversibly inactivated by the sulfhydryl rea- 59, 988-995. gent N-ethylmaleimide, consistent with findings in membrane 26. Heppel, L. A., Rosen, B. P., Friedberg, I., Berger, E. A. & vesicles (3). Shock-releasable systems are much less sensitive Weiner, J. H. (1972) "The molecular basis of biological to this inhibitor. The difference in behavior of various transport," Miami Winter Symposia 3, 133-156. systems 27. Weiner, J. H., Furlong, C. E. & Heppel, L. A. (1971) Arch. towards osmotic shock, which has long been a source of confu- Biochem. Biophys. 142, 715-717. sion and controversy for workers in this field (26), may thus 28. Weiner, J. H. & Heppel, L. A. (1971) J. Biol. Chem. 246, reflect a fundamental mechanistic difference between these 6933-6941. two classes of uptake systems. 29. Gordon, A. S., Lombardi, F. J. & Kaback, H. R. (1972) Proc. Nat. Acad. Sci. USA 69, 358-362. Note Added in Proof. B. P. Rosen has found (personal com- 30. Lombardi, F. J. & Kaback, H. R. (1972) J. Biol. Chem. 247, munication) that proline uptake in membrane vesicles from 7844-7857. Downloaded by guest on October 1, 2021