Reversible Effects of Chaotropic Agents on the Proton Permeability Of

Proc. Nat. Acad. Sci. USA Vol. 72, No. 9, pp. 3387-3391, September 1975 Biochemistry

Reversible effects of chaotropic agents on the proton permeability of Escherichia coli membrane vesicles* (active transport/membrane potential/ft-galactosides/amino acids/lipophilic cations/carbodiimides) LEKHA PATEL, SHIMON SCHULDINER, AND H. R. KABACK The Roche Institute of Molecular Biology, Nutley, New Jersey 07110 Communicated by Severo Ochoa, June 16, 1975

ABSTRACT Extraction of E. coli ML 308-225 membrane lize D-LDH from E. coli membrane vesicles, and enzyme vesicles with chaotropic agents causes the vesicles to become solubilized in this manner has been used to reconstitute ac- specifically permeable to protons. As a result, .the vesicles no longer generate a membrane potential, interior negative, and tive transport in vesicles prepared from mutants defective in they do not catalyze respiration-dependent lactose or proline D-LDH activity. All attempts to reconstitute the extracted transport. Treatment of the extracted vesicles with various vesicles, however, were unsuccessful presumably because of carbodiimides decreases the permeability of the vesicle a physiologic alteration unrelated to the loss of D-LDH. The membrane to protons, causing them to regain their ability to observations presented here deal with the nature of this generate a membrane potential. By this means, active trans- physiologic alteration and its reversal by carbodiimides. port is completely reactivated. Exposure of the vesicles to carbodiimides prior to extraction with chaotropic agents makes transport activity impervious to the effects. of the chaotropes. METHODS Bacterial membrane vesicles catalyze active transport of Growth of Cells and Preparation of Membrane Ves- many different solutes by a respiration-dependent mecha- icles. E. coli ML 308-225 (i-z-y+a+) was grown on mini- nism that does not involve the generation or utilization of mal medium A with 1.0% sodium succinate (hexahydrate) as ATP or other high-energy phosphate compounds (1-4). In the sole carbon source, and membrane vesicles were pre- Escherichia coli vesicles, most of these transport systems are pared as described (10, 29). S. aureus U-71 vesicles were also coupled primarily to the oxidation of D-lactate or reduced prepared as described (30). phenazine methosulfate (PMS) via a membrane-bound re- Transport was assayed as described (31). [1-'4C]Lactose spiratory chain with oxygen or, under appropriate condi- (22 mCi/mmol) was used at a final concentration of 0.4 mM; tions (5-7), fumarate or nitrate as terminal electron accep- [U-'4C]proline (232 mCi/mmol) at a final concentration of tor. Recent experiments demonstrate that essentially all of 8.3 gM; and [3H]triphenylmethylphosphonium (114 mCi/ the vesicles catalyze transport (8), and that D-lactate dehy- mmol) at a final concentration of 0.4 mM. [3H]Triphenyl- drogenase (D-LDH) and Ca++,Mg++-stimulated ATPase are methylphosphonium uptake was assayed using Millipore both located on the inner surface of the vesicle membrane Cellotate filterst. (9, 10). These and other findings (1-4, 11-13) demonstrate Extraction of Membrane Vesicles with Chaotropic that essentially none of the vesicles is inverted. Agents. Suspensions of vesicles containing approximately 2.0 Chemiosmotic phenomena, as postulated by Mitchell mg of membrane protein per ml (final concentration) in (14-16), play an important role in respiration-linked active 0.05 M potassium phosphate (pH 6.6) were adjusted to a transport (4, 17-22).t Some of the evidence is derived from given concentration of a particular chaotropic agent by add- studies of certain mutants that are defective in the membra- ing an aliquot of a concentrated solution of the chaotrope. nous Ca++,Mg++-stimulated ATPase complex (23-25). In The samples were incubated at 0° for 5-10 min, and centri- certain instances (18, 26, 27), these mutants and vesicles de- fuged at approximately 40,000 X g for about 30 min. The rived from them exhibit a pleiotropic transport defect which supernatants were removed for protein determinations; the is related to an increase in the permeability of the cell mem- pellets were washed once in 0.1 M potassium phosphate (pH brane to protons. Moreover, the alteration in proton perme- 6.6), and resuspended to approximately 4.0 mg of protein ability and the defect tn active transport can be cured by per ml in 0.1 M potassium phosphate (pH 6.6). Aliquots of treatment with N,N'-dicyclohexylcarbodiimide (DCC). these suspensions were then used directly for protein deter- Previous experiments from this laboratory (28) demon- minations and transport assays. Control preparations were strate that chaotropic agents such as guanidine-HCl solubi- treated in an identical manner, except that the chaotropic Abbreviations: PMS, phenazine methosulfate; D-LDH, 1-lactate agent was omitted. dehydrogenase; DCC, N,N'-dicyclohexylcarbodiimide; EDC, 1- Discontinuous Polyacrylamide Slab Gel Electrophore- ethyl-3-(3-dimethylaminopropyl)carbodiimide; EDCMI, 1-ethyl- sis was carried out in sodium dodecyl sulfate (32). 3-(3-dimethylaminopropyl)carbodiimide [14C]methiodide; CMC, Oxygen Consumption was determined with a Clark oxy- 1-cyclohexyl-3-[2-morpholyl-(4)-ethyllcarbodiimide; CMCMI, 1- gen electrode as described (33). cyclohexyl-3-[2-morpholyl-(4)-ethyllcarbodiimide [14C]methiodide; Materials. Enzyme grade guanidine-HCI and urea were DiPC, diisopropylearbodiimide; and CCCP, carbonylcyanide m- chlorophenylhydrazone. obtained from Schwartz/Mann Biochemicals. Reagent grade * This is paper XXIX in the series "Active Transport in Isolated sodium perchlorate and sodium thiocyanate were purchased Bacterial Membrane Vesicles". Paper XXVIII is S. Schuldiner and from Fisher Scientific Co. H. R. Kaback, submitted for publication. [1-'4C]Lactose was obtained from Amersham-Searle, and t S. Schuldiner and H. R. Kaback, submitted for publication. [U-'4C]proline from New England Nuclear. Triphenyl- 3387 Downloaded by guest on September 30, 2021 %-"'0088 Biochemistry: Patel et al. Proc. Nat. Acad. Sci. USA 72 (1975)

120

2 2 0IE 2 I e in uj so I.- -C cc t: I.- n E 60 uj I..- 0 IF1

[CHAOTROPE] (M) FIG. 1. Effect of chaotropic agents on E. coli ML 308-225 membrane vesicles. Vesicles were extracted with given concentrations of urea, guanidine-HCl, sodium thiocyanate (SCN), or sodium perchlorate (C104) as described in Methods. Rates of lactose uptake (0-*) were determined for 30 sec under oxygen in the 0 2.5 5.0 presence of 20 mM sodium ascorbate, 0.1 mM PMS, and 0.4 mM TIME (minutes) [1-14Cllactose (22 mCi/mmol) (31). 0-----0, protein solubilized FIG. 2. Reactivation of lactose transport in guanidine-UCI- by various concentrations of urea or guanidine.HCl. extracted vesicles by DCC. Membrane vesicles were extracted with 1.0 M guanidine-HCl as described in Methods (0). Where indicated (4), samples of this preparation were incubated with 70 IAM methylphosphonium bromide was labeled with tritiumt. DCC for 30 min at 250 prior to assaying lactose transport. Lactose DCC was obtained from Calbiochem and N,N'-dicyclo- transport was assayed at 250 under oxygen in the presence of as- hexyl[I4C]carbodiimide (45 mCi/mmol) was synthesized by corbate and PMS as described previously (31) and in Methods. (@) The Radiation Synthesis Group of Hoffmann-LaRoche, Inc. Lactose uptake by untreated control vesicles. Addition of 70 AM under the direction of Dr. A. Leibmant. I-Ethyl-3-(3-di- DCC had no significant effect on either the rate or extent of lac- methylaminopropyl)carbodiimide (EDC) and 1-ethyl-3-(3- tose uptake by the control preparation. dipthylaminopropyl)carbodiimide ['4C]methiodide ides extracted with 5.0 M urea, 1.0 M sodium perchlorate, (EDCMI) [17 mCi/mmol] (34); 1-cyclohexyl-3-[2-morpho- or 0.5 M sodium thiocyanate (data not shown), and with 1.0 linyl-(4)-ethyl]carbodiimide (CMC) and 1-cyclohexyl-3-[2- mM EDC, 2.0 mM CMC, or 3.0 mM DiPC (see below). Al- morpholinyl-(4)-ethyl]carbodiimide [3H]methiodide though higher concentrations of chaotropes result in some (CMCMI) [80 mCi/mmol] (35); and diisopropylcarbodiim- degree of irreversible inactivation, even after extraction ide (DiPC) were graciously contributed by Dr. James Ofen- with 8.0 M urea or 2.5 M guanidine-HCl, the transport activ- gand of The Roche Institute of Molecular Biology. In each ity of the vesicles can be restored to 20-40% of the control case, the radioactive carbodiimides reactivated active trans- activity by exposure to carbodiimides. Although data will port in membrane vesicles extracted with chaotropic agents. not be shown, it should be emphasized that the reactivating Non-water-soluble carbodiimides were dissolved in ethanol, effect of the carbodiimides cannot be reversed by washing and the final concentration of ethanol in the reaction mix- the vesicles extensively. Similar results were obtained when tures did not exceed 1.0%. proline transport was assayed in ML 308-225 vesicles and All other materials were reagent grade obtained from when serine transport was assayed in Staphylococcus aureus commercial sources. vesicles extracted with 1.0 M guanidine.HCl. D-Lactate-dependent transport is also reactivated with RESULTS carbodiimides after extraction with guanidine-HCl or urea, Effect of Chaotropic Agents on Active Transport. Ex- but in this case, restoration to only about 40-50% of the con- traction of ML 308-225 membrane vesicles with increasing trol activity is observed. This is as expected since approxi- concentrations of urea, guanidine-HGl, sodium thiocyanate, mately 50-60% of the membrane-bound D-LDH activity is or sodium perchlorate leads to drastic inactivation of ascor- removed from the vesicles under these conditions (28). Urea bate-PMS-driven lactose transport (Fig. 1). Inactivation is or guanidine-HCl-extracted vesicles are completely reconsti- dependent upon the concentration of the particular chaotro- tuted for D-lactate-dependent transport by exposure to pic agents and with urea and guanidine-HCl, inactivation is DCC, followed by treatment with purified D-LDH (36). correlated reasonably well with the loss of protein from the When vesicles are treated with carbodiimides prior to ex- vesicles. With 5.0 M urea or 1.0 M guanidine.HCl, transport traction with urea or guanidine-HCl, ascorbate-PMS-driven is virtually completely inactivated and approximately 15- lactose transport is not significantly affected by extraction 20% of the membrane protein is extracted. with chaotropic agents (data not shown). It should be em- Reactivation of Active Transport by Carbodiimides. phasized though that extraction of carbodiimide-treated ves- Strikingly, the inactivating effects of these harsh denaturants icles with urea or guanidine-HCl solubilizes the same on active transport are completely reversed by treating the amount of protein and D-LDH activity as observed with un- extracted vesicles with an appropriate concentration of any treated vesicles. These results indicate that carbodiimide- of four carbodiimides tested. As shown in Fig. 2, guanidine- reactive sites in the membrane are accessible to these re- HCl-extracted vesicles exposed to 70 ,uM DCC for 30 min agents even before exposure to the chaotropes. exhibit about a 10-15% increase in specific activity relative Treatment of vesicles with phospholipase A (bee venom), to control vesicles. Similar results were obtained with ves- Triton X-100, pyridine, or high concentrations of lithium or potassium chloride also results in inactivation of active trans- $ A manuscript describing the synthesis is in preparation (C. W. port, but these effects are not reversed by carbodiimides. Perry, W. Burger, and A. Leibman). Concentration Dependence of Carbodiimide Reactiva- Downloaded by guest on September 30, 2021 Biochemistry: Patel et al. Proc. Nat. Acad. Sci. USA 72 (1975) 3389

- .D 030 X - a: E a- 3?--. C 7 E - 01 [CARBODIIMIDE] (mM) 230 E reactiva- 03 FIG. 3. Concentration dependence of carbodiimide 31~~~~~~~s tion. Membrane vesicles were extracted with 1.0 M guanidine.HCl E as described in Methods. Samples were incubated with the designated carbodiimide at the concentration given for 30 min at 25°. 0 At that time, the rate of lactose uptake was measured for 30 sec in 0 9 20 30 40 0 l0 20 30 40 50 Time (minutes) the presence of ascorbate and PMS as described previously (31) and in Methods. FIG. 4. Time course of carbodiimide reactivation of lactose transport and incorporation of labeled carbodiimides. ML 308-225 tion. Each carbodiimide tested reactivates lactose transport, membrane vesicles were extracted with 1.0 M guanidine-HCl as described in Methods. Samples of these vesicles were incubated with although the concentrations required for optimal reactiva- 70 AM DCC, 3.0 mM DiPC, 1.0 mM EDC, or 2.0 mM CMC for the tion vary considerably (Fig. 3). Maximal rates of transport times indicated, and lactose transport was assayed for 30 sec with are observed with 0.07 mM DCC, 1.0 mM EDC, 2.0 mM ascorbate and PMS as described previously (30) and in Methods CMC, and 3.0 mM DiPC, and in most cases, supraoptimal (0). (0) Incorporation of radioactive carbodiimides by guanidine- concentrations result in diminished activity. Although the HCl-extracted ML 308-225 vesicles. Incorporation of EDC['4C]MI was determined reason for the variation in optimal concentrations for reacti- (17 mCi/mmol) and CMC[3H]MI (80 mCi/mmol) at final concentrations of 1.0 mM and 2.0 mM, respectively, by vation is not apparent, it seems unlikely that it is directly re- rapid filtration and Millipore filtration as described for transport lated to differences in hydrophobicity. Although DCC reac- (31), except that the filters were washed five times with 2.0-ml ali- tivates at the lowest concentration, DiPC, another hydropho- quots of 0.1 M LiCl and Millipore Cellotate filters (0.5 jm) were bic carbodiimide, exhibits an optimum that is higher than used. Incorporation of [14C]DCC was determined in the following those observed with EDC and CMC, both of which are manner: 100-Al reaction mixtures containing 50 mM potassium water soluble. Treatment of guanidine.HCl-extracted ves- phosphate (pH 6.6), 10 mM magnesium sulfate, 70 AM ['4CJDCC (45 mCi/mmol), and about 0.2 mg of membrane protein were incu- icles with various concentrations of oligomycin does not bated for the times shown. The samples were diluted with 2.0 ml of cause reactivation, nor does treatment with dimethylsuberi- ethanol (100%) and centrifuged. The supernatants were discarded midate (up to 5 mM). The latter reagent demonstrably cross- and the pellets resuspended in 100 Ml of 0.1 M potassium phos- links vesicle proteins (data not shown). phate (pH 6.6). Aliquots (50 Al) of these suspensions were counted Time Course of Reactivation and Carbodiimide Incor- in a liquid scintillation spectrometer. The data were corrected for to addition poration. With each carbodiimide tested, a marked lag control samples which were diluted with ethanol prior of [14C]DCC and centrifugation. phase is observed in the time course of reactivation (Fig. 4). With DCC, the lag is approximately 10 min; with DiPC, approximately 5 min; with EDC, approximately 2 min; and imides. Oxidation of D-lactate or reduced PMS by E. coli with CMC, approximately 15 min. In each case, subsequent membrane vesicles generates a membrane potential, interior to the lag, transport activity increases dramatically to a max- negative, which results from extrusion of protons from the imum which remains constant over the time course of the vesicles (17, 19, t). Evidence has also been presented (4, 18- experiments. Treatment of the vesicles with any one of the 22, t) which indicates that the membrane potential is the carbodiimides for the time period characteristic of its lag driving force for respiration-linked active transport. When phase, followed by washing of the vesicles, completely abol- ML 308-225 vesicles are extracted with guanidine.HCl, they ishes the lag phase when the vesicles are incubated a second are unable to generate a membrane potential, interior nega- time with a different carbodiimide (data not shown). This tive, as judged by their inability to catalyze the uptake of the observation suggests that each of the carbodiimides reacts lipophilic cation triphenylmethylphosphonium (Fig. 5). If with the same site(s) in the membrane. the extracted vesicles are exposed to DCC or EDC, their When the vesicles are incubated with radioactive carbodi- ability to generate a potential in the presence of reduced imides, a lag phase is not observed in the time course of in- PMS is completely restored. These effects cannot be attrib- corporation (Fig. 4). Thus, with DCC, EDCMI, and uted to a reversible defect in the respiratory chain, since re- CMCMI, incorporation increases linearly with time, reach- duced PMS is oxidized at comparable rates by each of the ing a plateau prior to or at the same time that maximal vesicle preparations (data not shown). transport activity is observed. Although data will not be pre- Since any defect resulting in increased ion permeability sented, the time course of incorporation of the carbodi- could abolish the ability of the vesicles to generate and/or imides into vesicles that were not extracted with chaotropic maintain an electrical potential, the permeability of the ves- agents is essentially the same as that shown, a result which is icles to various ions was investigated. No significant alter- consistent with the conclusion that carbodiimide reactive ation in the permeability to 22sodium or Wrubidium is ob- sites within the vesicles are accessible prior to extraction. It is served when the vesicles are extracted with 1.0 M guani- also noteworthy that vesicles treated with these carbodi- dine.HCI or extracted and subsequently treated with DCC imides exhibit disc gel electrophoretic patterns that are iden- (Table 1). Preliminary observations (not shown) also indicate tical to guanidine-HCI-extracted vesicles (data not shown). that the permeability of the vesicle membrane to sulfate and Mechanism of Inactivation of Active Transport by arabinose is not affected by these treatments. In contrast, Chaotropic Agents and Its Reactivation with Carbodi- marked differences in proton permeability are observed Downloaded by guest on September 30, 2021 3390 Biochemistry: Patel et al. Proc. Nat. Acad. Sc. USA 72 (1975)

C '5-fo

.5 FIG. 6. Effect of chaotropic agents and DCC on proton permeability of ML 308-225 membrane vesicles. Membrane vesicles were extracted with 1.0 M guanidine-HCl as described in Methods (B), and a portion of the extracted vesicles was incubated with 70 uM 0 2.5 5.0 DCC for 30 min at 250 (C). These preparations and untreated con- TIME (minutes) trol preparations (A) were centrifuged and washed once with 0.66 FIG. 5. Reactivation of triphenylmethylphosphonium M KCl containing 2.0 mM potassium phosphate (pH 6.6). The pel- (TPMP+) uptake in guanidine.HCl-extracted vesicles by carbodi- lets were resuspended in the same medium to a protein concentration of 27.5 mg/ml, and valinomycin was added to a final concen- imides. Membrane vesicles were with 1.0 extracted M guanidine. tration of 15 MuM. Rates of proton equilibration were measured HCl as described in Methods (0). Where of as indicated, samples this follows: 1.4-ml aliquots of the suspensions were transferred to a preparation were incubated with 1.0 mM EDC (v) or 70 MAM DCC chamber and continually stirred with a small magnetic bar. The (A) for 30 min at 25° prior to assaying triphenylmethylphosphon- pH was monitored with a pH electrode (Radiometer GK 2321C) ium uptake. was as- [3H]Triphenylmethylphosphonium uptake connected to a Radiometer pH meter (model 26), and the signal sayed at 25° under oxygen in the presence of ascorbate and PMS was recorded in a double channel Corning recorder (model 845). as described previouslyt and in Methods. (0) Triphenylmethyl- The experiment was started by rapid addition of 90 nmol of HCl phosphonium uptake by untreated control vesicles. Addition of 70 (arrows). Where indicated, CCCP was added to the reaction mix- ,MM DCC or 1.0 mM EDC had no significant effect on either the tures at a final concentration of 25 MM. rate or extent of triphenylmethylphosphonium uptake by the control preparation. crude preparations of coupling factor (40), however, have (Fig. 6). When control vesicles are pulsed with acid, the rate been completely unsuccessful. Furthermore, since 60-80% of equilibration of protons across the vesicle membrane ex- of the cellular Ca++,Mg++-stimulated ATPase activity is lost hibits a half-time of about 1-2 min, and is increased dramat- during the preparation of membrane vesicles (10), it is ap- ically when the proton conductor carbonylcyanide m-chlor- parent that most of the catalytic subunit of the ATPase com- ophenylhydrazone (CCCP) is added. With guanidine.HCl- plex, at least, is absent from the vesicles even before they are extracted vesicles, the time course of equilibration is virtual- extracted. Two other important points should be mentioned: ly identical to that observed with untreated vesicles in the (i) carbodiimide-reactive sites are available prior to extrac- presence of CCCP, and it is not altered by the proton con- tion of the vesicles with chaotropic agents; and (u) the ef- ductor. Finally, when the guanidine.HCl-extracted vesicles fects reported here are not specific for DCC, and do not ap- are exposed to DCC, their behavior reverts to that of the control preparation. Table 1. Permeability of ML 308-225 vesicles to sodium and rubidium DISCUSSION Efflux rate The results presented in this paper resemble observations (t½2)* (sec) made with mitochondrial (37, 38), chloroplast (39), and E. coli (25, 40) membrane particles depleted of ATPase. They Treatment 22Na "6Rb also resemble observations made with vesicles prepared from certain mutants of E. coli which are defective in the 1. None 27 111 membrane ATPase complex (18, 23-27). In these instances, 2. Guanidine-HCl-extractedt 20 90 energy-linked transhydrogenase, oxidative phosphorylation, 3. Guanidine-HCl-extracted + DCCt 27 66 photophosphorylation, and/or respiration-linked active Designated preparations of membrane vesicles were adjusted to transport are impaired, but can be restored by exposure of approximately 40 mg of membrane protein per ml in 50 mM potassium to or DCC. The studies of phosphate (pH 6.6). Aliquots (100 Ml) of these suspensions the preparations oligomycin were made up to final concentrations of 50 mM 22NaCl (25 MCi/Ml) Mitchell and his associates (41, 42) and others (43, 44) indi- or 40 mM 86RbCl (26.7 MCi/Ml), and equilibrated overnight at 0°. cate that removal of ATPase enhances the proton permeabil- Aliquots of the suspensions (2 Ml) were then rapidly diluted 100-fold ity of their mitochondria and chloroplasts, and that oligomy- into 50 mM potassium phosphate (pH 6.6) containing either 50 cin restores their proton impermeability. Similarly, vesicles mM NaCl or 40 mM RbCl, respectively, and the samples were as- prepared from certain ATPase mutants of E. coli exhibit a sayed at various times by rapid dilution and Millipore filtration (31). After correcting for radioactivity adsorbed to the filters, the pleiotropic defect in respiration-linked transport which can results were plotted as the logarithm of the residual radioactivity be cured by treatment with DCC (18, 26, 27). These effects against time, and the t1/2 values were taken directly from these were also attributed to alterations in the proton permeability plots. In each case, the loss of radioactivity followed pseudo first- of the vesicle membrane. order kinetics. No evidence has been presented here which indicates that * tl/2 denotes the time required for loss of 50% of the radioactivity the effect of chaotropic agents is related specifically to the present at zero time. t ML 308-225 vesicles extracted with 1.0 M guanidine.HCl as de- removal or inactivation of a component of the ATPase com- scribed in Methods. plex, although this is certainly a possibility. Attempts to t Guanidine-HCl-extracted vesicles were incubated with 70 4M reactivate extracted vesicles with chaotropic extracts or with DCC for 30 min at 25° prior to equilibration with 22Na or 86Rb. Downloaded by guest on September 30, 2021 Biochemistry: Patel et al. Proc. Nat. Acad. Sci. USA 72 (1975) 3391

pear to be directly related to the hydrophobicity of the car- 11. Altendorf, K. H. & Staehelin, L. A. (1974) J. Bacteriol. 117, bodiimides. 888-899. With regard to the chemistry of carbodiimide-induced 12. Konings, W. N., Bisschop, A., Voenhuis, M. & Vermeulen, C. A. (1973) J. Bacteriol. 116,1456-1465. reactivation, very little is known, except that it is highly un- 13. Rosen, B. P. & McClees, J. S. (1974) Proc. Nat. Acad. Sci. USA likely that these reagents catalyze intermolecular crosslink- 71,5042-5046. ing in this system. Given these results and the known carbox- 14. Mitchell, P. (1966) Biol. Rev. (Camrridge) 47,445-502. yl reactivity of these compounds (45, 46), it seems reason- 15. Mitchell, P. (1973) J. Bioenergetics 4,63-91. able to postulate that the carbodiimides may react with car- 16. Harold, F. M. (1972) Bacteriol. Rev. 36, 172-230. boxyl groups in certain membrane proteins that are involved 17. Hirata, H., Altendorf, K. & Harold, F. M. (1973) Proc. Nat. in proton permeability. Whether these proteins form a car- Acad. Sci. USA 70,1804-1808. boxyl-lined channel through the membrane or function as 18. Altendorf, K., Harold, F. M. & Simoni, R. D. (1974) J. Biol. proton carriers cannot be answered at present, although it Chem. 249,4587-4593. seems intuitively unlikely that a channel could exist that 19. Altendorf, K., Hirata, H. & Harold, F. M. (1975) J. Biol. would transmit protons but not sodium or potassium. In any Chem. 250, 1405-1412. 20. Schuldiner, S., Kerwar, C. K., Weil, R. & Kaback, H. R. (1975) case, the availability of carbodiimide-resistant E. coli mu- J. Biol. Chem. 250,1361-1370. tants that lack a specific protein (47) should help to provide 21. Rudnick, G., Weil, R. & Kaback, H. R. (1975) J. Biol. Chem. a meaningful experimental approach to these problems. 250, 1371-1375. In addition to providing supportive evidence for the es- 22. Rudnick, G., Weil, R. & Kaback, H. R. (1975) J. Biol. Chem. sential role of chemiosmotic phenomena in the mechanism 250, in press. of respiration-linked active transport, these studies are re- 23. Simoni, R. D. & Shallenberger, M. K. (1972) Proc. Nat. Acad. markable for their implications with respect to the nature of Sci. USA 69,2663-2667. the molecular machinery involved in these transport sys- 24. Rosen, B. P. (1973) Biochem. Biophys. Res. Commun. 53, tems. Not only are the f,-galactoside and proline carriers im- 1289-1296. pervious to the effects of these harsh denaturants, but the in- 25. Nieuwenhuis, F.J.R.M., Kanner, B. I., Gutnick, D. L., Postma, P. W. & Van Dam, K. (1973) Biochim. Blophys. Acta 325, teractions between most of the components of the respira- 62-71. tory chain are not affected. These conclusions are particular- 26. Rosen, B. P. (1973) J. Bacteriol. 116, 1124-1129. ly impressive in view of the fact that approximately 20% of 27. Van Thienen, G. & Postma, P. W. (1973) Biochim. Biophys. the membrane protein and 50-60% of the D-LDH are solu- Acta 323, 429-440. bilized. Essentially all of the D-LDH in these vesicles is lo- 28. Reeves, J. P., Hong, J.-s. & Kaback, H. R. (1973) Proc. Nat. cated on the inner surface of the vesicle membrane (10). Acad. Sci. USA 70, 1917-1921. Finally, these results may be relevant to the development 29. Kaback, H. R. (1971) in Methods in Enzymology, ed. Jakoby, of other vesicle systems for the study of active transport. It is W. B. (Academic Press, New York), Vol. XXII, pp. 99-120. conceivable that some vesicle systems may be nonfunctional 30. Short, S. A. & Kaback, H. R. (1974) J. Biol. Chem. 249, because the membrane becomes permeable to protons dur- 4275-4281. 31. Kaback, H. R. (1974) in Methods in Enzymology, eds. ing preparation. In this context, carbodiimides may prove to Fleischer, S. & Packer, L. (Academic Press, New York), Vol. be generally useful in studies of active transport in other iso- XXXI, pp. 698-709. lated membrane systems. 32. Ames, G. F.-L. (1974) J. Biol. Chem. 249,634-644. 33. Barnes, E. M., Jr. & Kaback, H. R. (1971) J. Biol. Chem. 246, We are indebted to The Radiation Synthesis Group of Hoff- mann-LaRoche and Dr. A. Leibman for synthesizing [14C]DCC, 5518-5522. and to Dr. James Ofengand for providing EDC[14C]MI and 34. Sheehan, J. C., Cruickshank, P. A. & Boshart, G. L. (1961) J. CMC[3H]MI. Org. Chem. 26,2525-2528. 35. Brostoff, S. W. & Ingram, V. M. (1970) Biochemistry 9, 2372-2376. 1. Kaback, H. R. (1972) Biochim. Biophys. Acta 265,367-416. 36. Short, S. A., Kaback, H. R. & Kohn, L. D. (1974) Proc. Nat. 2. Kaback, H. R. & Hong, J.-s. (1973) in CRC Critical Reviews Acad. Sci. USA 71,1461-1465. in Microbiology, eds. Laskin, A. I. & Lechevalier, H. (CRC 37. Lee, C. P. & Ernster, L. (1965) Biochem. Biophys. Res. Com- Press, Ohio), Vol. 2, pp. 333-376. mun. 18, 523-529. 3. Kaback, H. R. (1973) in Bacterial Membranes and Walls, ed. 38. Racker, E. & Horstman, L. L. (1967) J. Biol. Chem. 242, Leive, L. (Marcel Dekker, New York), Vol. I, pp. 241-292. 2547-2551. 4. Kaback, H. R. (1974) Science 186,882-892. 39. Fessenden-Raden, J. M. (1969) J. Biol. Chem. 244, 6662- 5. Konings, W. N: & Kaback, H. R. (1973) Proc. Nat. Acad. Sci. 6667. USA 70,3376-3381. 40. Bragg, P. D. & Hou, C. (1973) Biochem. Biophys. Res. Com- 6. Boonstra, J., Huttunen, M. T., Konings, W. N. & Kaback, H. mun. 50, 729-736. R. (1975) J. Biol. Chem., in press. 41. Mitchell, P. (1967) Fed. Proc. 26, 1370-1379. 7. Konings, W. N. & Boonstra, J. (1975) in Current Topics in 42. Scholes, P., Mitchell, P. & Moyle, J. (1969) Eur. J. Biochem. 8, Membranes and Transport, eds. Kleinzeller, A. and Bronner, 450-454. F., in press. 43. Hinkle, P. & Horstman, L. L. (1971) J. Biol. Chem. 246, 8. Short, S. A., Kaback, H. R., Kaczorowski, G., Fisher, J., Walsh, 6024-6028. C. T. & Silverstein, S. (1974) Proc. Nat. Acad. Sci. USA 71, 44. Uribe, E. G. (1972) Biochemistry 11, 4228-4235. 5032-5036. 45. Kurzer, F. & Douraghi-Zadeh, K. (1967) Chem. Rev. 67, 9. Short, S. A., Hawkins, T., Kohn, L. D. & Kaback, H. R. (1975) 107-152. J. Biol. Chem. 250,4285-4290. 46. Hoare, D. G. & Koshland, D. E. (1967) J. Biol. Chem. 242, 10. Short, S. A., Kaback, H. R. & Kohn, L. D. (1975) J. Biol. 2447-2453. Chem. 250,4291-4296. 47. Fillingame, R. H. (1975) Fed. Proc. 34, Abs. no. 2025. Downloaded by guest on September 30, 2021