as the facilitated diffusion reactions catalyzed by the carrier in the absence of .... reflect carrier-mediated movements: these were therefore used for comparison.
Vol. 263, No.20, Issue of July 15,pp. 9663-9667.1988
THEJOURNALOF BIOLOGICAL CHEMISTRY 0 19R8 by The American Society for Biochemistry and Molecular Biology, Inc.
Printed
in
U.S.A.
Melibiose Permeaseof Escherichia coli CHARACTERISTICS OF CO-SUBSTRATES RELEASE DURING FACILITATED DIFFUSION REACTIONS* (Received for publication, December 28,1987)
Martine Bassilana, Thierry Pourcher, and Gerard Leblanc$ From the Laboratoire Jean Maetz, Departementde Biologie du Commissariat a 1’Energie Atomique, Station Marine, B. P. 68, 06230 Villefranche-sur-Mer,France
The mechanism of melibiose symport by the meli- the coupled cation (4-5). It was then shown that the sugar biose permease of Escherichia coli was investigatedby binding characteristics of the melibiose permease (6) as well further analyzing the Na+ (H+ or Li+)-coupledfacili- as the facilitated diffusion reactions catalyzed by the carrier tated diffusion reactions catalyzed by the carrier in in theabsence of energy ( 7 )differ greatly according to whether de-energized membrane vesicles, with particular em- the coupled cation is H+, Na+, or Li’. Careful examination of phasis on the reaction of sugar exchange at equilib- the differences in the kinetic properties of the melibiose rium. It is firstshown that melibiose exchange at equi- permease as afunction of the coupled cation should therefore librium proceeds without concomitant movement of afford insight intothe role of the coupled cation in the Na+, i.e. the coupled cation is kineticallyoccluded dur- mechanism of co-transport of melibiose. ing themelibiose exchange reaction. These results proWe have indeed recently reported ( 7 ) that the facilitated vide further experimental support for the model of Na+ diffusion properties of the melibiose permease observed in desugar co-transport of the physiological substrate melibiose previously suggested (Bassilana, M., Pourcher, energized membrane vesicles, i.e. melibiose influx or efflux down a sugar concentration gradient, exchange at equilibrium, T., and Leblanc, G . (1987) J. Biol. Chem. 262,1686516870) in which: 1) the mechanisms of co-substrate and counterflow activity, are considerably changed by perbinding to (or release from) the carrier are ordered muting the three coupling cations (H+, Na+, and Li+). Strikprocesses on both the outer (Na+ first, sugar last) and ingly, the permease catalyzes a significant H+, but negligible inner membrane surfaces (sugar first, Na+ last) and Na+- or Li+-coupledmelibiose influx down a sugar concentragive rise toa mirror-type model; 2) release of Na+ from tion gradient. Also, exchange at equilibrium or counterflow the carrieron the inner membrane surface is very slow activity is seen when the permease functions as aH+ or Na+and rate-limiting for carriercycling but is fast on the melibiose symport, but not when Li’ is the coupled cation. opposite side, contributing to the asymmetrical func- An asymmetrical functioning of the melibiose permease was, tioning of the permease. On the other hand, analysis of however, observed with all three cations. Cross-comparison theexchange of identicalsugars (homologous ex- of the facilitated diffusion properties during H+-, Na+-, or change) and different sugaranalogs (heterologous ex- Li+-melibiose symport activities led us to suggest a kinetic change) indicates that the overall rate of sugar ex- model of cation-sugarco-transport. It was proposed that change reactioncoupled to Na+ or Li+ is limited by the cycling of the permease giving rise to inward movement of corate of one (or more) partial step(s) associated with the substrates in de-energized conditions is rate-limited by release inflow of co-substrates and most probably by the rate of the co-substrates from the carrier intothe cytoplasm rather of sugar release into the intravesicular medium. It is than by a reduced rate of translocation of the loaded or empty proposed that the variability of the facilitateddiffusion reactions catalyzed by the carrier in the presence of carrier. different coupled cations and/or sugar analogs reflects Moreover, the apparent contradiction raised by the observariations in the rate of co-substrate release from the vation that the rate of net entry of melibiose during a Na+coupled influx reaction down a sugar concentration gradient carrier on the innermembrane surface. is negligible, whereas a 20-fold greater unidirectional influx of melibiose is measured during sugar exchange at equilibrium or counterflow has been explained by adding the two following assumptions to the model: (i) release of the co-substrates on The a-galactoside permease of Escherichia coli, or melibiose’ permease, is particularly interesting because it the inner surface is a sequenced process (sugar first, Na+ last) facultatively co-transports H+, Na+, or Li+ depending on the and (ii) the rate of release of the cation is much slower than substrate orcationic environment (1-3). It was, however,soon the rate of release of the sugar. Thus, net influx down a noticed that thekinetic parameters of the energy-dependent, concentration gradient is small because the overall reaction melibiose transport reaction vary according to theidentity of is rate-limited by the rate of release of Na+ on the inner surface. On the other hand, the assumption that the sugar * This work was supported in part by the Centre National de la leaves the carrier first on the inner surface makes the exRecherche Scientifique UA 638 Associbe au Commissariat 1 1’Energie change of sugar possible even when the cation does not Atomique. The costs of publication of this article were defrayed in dissociate from the carrier. If this interpretation is correct, part by the payment of page charges. This article must therefore be one should not expect Na’ movements during the reaction of hereby marked “advertisement” in accordance with 18U.S.C. Section melibiose exchange at equilibrium. 1734 solelyto indicate this fact. The aim of the present study was to analyze further the $ To whom correspondence should be addressed. ’ The abbreviations used are: melibiose, 6-O-a-D-galactosyl-~-glu-properties of facilitated diffusion reactions and in particular cose; TMG, methyl-P-D-thiogalactopyranoside; aMG, 1-0-methyl-a- the reaction of exchange at equilibrium mediated by the D-galactopyranoside;NEM, N-ethylmaleimide. melibiose permease in the presence of Na+, H’, or Li’ in
9663
9664
Melibiose Permease of E. coli
membrane vesicles from E. coli. W e first examined how the exchange reaction in spite of t h e slow rate of release of Na' movements of Na+ catalyzed by the permease during Na'the from the carrier onthe inner surface. A reduced rate of Na' coupledreactions of facilitateddiffusionconformto that release should, however, impair Na' movements during the expected from the abovemodel of melibiose co-transport. We Na+ melibiose exchange reaction. T o verify this prediction, then studied the dependence of the Na' and Li+-melibiose an analysis of concomitant "Na+ and [3H]melibiose fluxes exchange and counterflow reactions on the chemical structure undertaken in memduring effluxand exchange reactions was of the sugar. Finally,the p H dependence of the H+-melibiose brane vesicles (Fig. 1).In these experiments, membrane vesThe results icles were previously equilibrated either with 10 mM *'Na+ efflux and exchangereactionswasexamined. provide further insight into the characteristics of the partial and 20 mM unlabeled melibiose or with 10 mM unlabeled Na' steps involved in the release of the co-substrates from the and 20 mM [3H]melibiose. F o r efflux determinations, these of the membrane. permease on the inner surface vesicles were diluted in sugar-free media either lacking or containing 10 mM Na+ ions; in conditions of exchange, the EXPERIMENTALPROCEDURES outer solution contained 10 mM Na+ and 20 mM unlabeled Growth of Cells and Preparation of Membrane Vesicles-The E. coli sugar. It has previously been demonstrated (7) that the fracRAll strain used throughout this study has a temperature-stable, tions of cation and sugar outward fluxes sensitive to NEM inducible melibiose transport activity but lacks a-galactosidase and reflect carrier-mediated movements: these were therefore used has a deleted lacy gene (8).Cells were grown in minimal medium in for comparison. Fig. 1 shows that in the presence or absence the presence of 10 mM melibiose as previously described (5). Rightside out membrane vesicles were prepared by the method described of outer Na' ions, the rate of Na' efflux is similar t o that of by Kaback (9),resuspended in 0.1 M potassium phosphate (pH 6.6) melibiose efflux. The Na+/sugar stoichiometry value of 1/1 observed inthese Na+-coupledefflux reactions is in agreement and stored in liquid nitrogen until use. Flux Assays-For determinations of Na+-coupled sugar efflux or with previous data (7); the results also demonstrate that the exchange, thawed membrane vesicles were washed and equilibrated Na+/sugar stoichiometry value of the efflux reaction is indein 0.1 M potassium phosphate, 10 mM MgS04 buffered solution (pH pendent of the presenceof external Na' ions. Incidentally, it 6.6) containing 10 mM Na+ and then concentrated to about 30 mg of c a n be noted that t h e rates of sugar and Na+ efflux were membrane protein/ml. Unless otherwise stated, carbonyl cyanide p trifluoromethoxyphenylhydrazone and monensin were added to give reduced tothe same extent when the outer Na+ concentration final concentrations of 10 and 0.75pM, respectively. Aliquots of is increased from less than 20 I.LM t o 10 mM. The transinhibitory effect of outer Na' ions on the coupled effluxes of concentrated [3H]melibiose(2-5 mCi/mmol) or ["CITMG (1.5 mCi/ mmol) were then added to a final concentration of20mM, and the Na' and melibiose possibly reflects a decrease in the number membrane suspensions were left to equilibrate at room temperature of unloaded carriers cycling across the membrane as a result for 40 min and then overnight at 4 'C. Sugar efflux and exchange of the formation of binary Na' carrier complexes which are reactions were initiated by diluting 2.5-pl aliquots of sugar-loaded vesicles in 2 ml of 0.1 M potassium phosphate, 10 mM MgS04 buffer unable to cross the membrane. When the sugar and Na+ outward fluxes are determined in (pH 6.6) lacking both Na+ and sugar or containing 10 mM Na+ (for efflux measurements); the dilution medium used for exchange meas- conditions of exchange at equilibrium, the carrier-mediated urementscontained Na+ andan equimolar concentration of the outward movementof Na' is no longer equal the to sugar flux desired unlabeled sugar. At given times after dilution, samples were (Fig. 1). In thisparticular experiment, the Na+/melibiose filtered and the filters (Whatman GF/F) washed once and assayed for radioactivity. For determinations of Na+ movements coupled to w either sugar efflux or exchange, membrane vesicles were previously loaded with 20 mM unlabeled sugar and "Na+ (10 mM, 3 mCi/mmol) in medium lacking monensin. The rate of "Na+ outward movement was monitored by the filtration technique as described above. Determinations of Li+-coupled sugar efflux or exchange rates were performed according to the same experimental protocol except that the media contained 10 mM Li+ ions. H+-coupled melibiose efflux and exchange were measured as described above in media containing no Li+ and less than 20 p~ Na+ ions. In the experiments in which the effect of pH was studied, both the equilibration and dilution media were adjusted to the desired pH. Protein Determinations and Flux Rates Calculation-Protein was measured as described by Lowry et al. (10) with serum albumin as standard. Rates of sugar efflux and exchange as well as outward Na+ movements mediated by the permease were computed from the rate 0 10 10 [Na of internal label decrease and corrected for passive permeability 0 0 20 [SI0 components by performing parallel experiments on membrane vesicles previously incubated with 1 mM N-ethylmaleimide for 30 min at FIG. 1. Na*/melibiose stoichiometry during Na+-coupled room temperature. melibiose efflux and exchange reactions mediated by the meMaterials-[3H]Melibiose (2.7 Ci/mmol) was tritiated by catalytic libiose permease. Concentrated membrane vesicles (30 mg memexchange in the Service des Molkules Marquies (Commissariat a brane protein/ml) incubated in 0.1 M potassium phosphate buffer 1'Energie Atomique, France). "NaC1 (carrier-free) and [I4C]TMG(30 (pH 6.6) were loaded either with 10 mM unlabeled Na+ and 20mM mCi/mmol) were from Amersham Corp. and C. E. A. (France), [3H]melibiose (2.5 mCi/mmol) or with 10 mM '*Na+ (3 mCi/mmol) respectively. Stock solutions of carbonyl cyanide p-trifluoromethox- and unlabeled melibiose. Rates of [3H]melibioseoutward movement yphenylhydrazone (Boehringer) and monensin were prepared in (open columns) and "Na+ (hatched columns) were measured during a MezSOand EtOH. N-Ethylmaleimide (Behring Diagnostics) and 2,4- Na+-coupled melibiose efflux reaction by diluting the loaded vesicles dinitrophenol (Prolabo) solutions were prepared shortly before use. in sugar-free, Na+-free medium (left-hand columns) or in sugar-free, Na+-containing medium (middle columns). In conditions of Na+RESULTS coupled melibiose exchange reaction (right-hand columns) the diluNa+ Movements During Na+ Melibiose or Nu' TMG Efflux tion medium contained both Na+ andmelibiose. Abscissa,concentrainmillimolar of Na+ ([Na+],) andmelibiose ([SI,)in the diluting and Exchange at Equilibrium-The sequence of release of t h e tions medium. Ordinate, NEM-sensitive components of'*Na+ and [3H] two co-substrates from the carrier the on inner surface (sugar melibiose outward movements calculated from the rates of intravesicfirst, Na+ last) proposed in the kinetic model for Na+-meliular radioactivity decrease monitored by filtration and expressed in biose transport should allow the system to promote a sugar nmol/mg protein. min.
1 +I0
Melibiose Permease of E. coli coupling ratio was less than 0.2. In six similar experiments the Na+/melibiose stoichiometry was found to vary between 0 and 0.2. It is important to note that the rate of melibiose outward movement during the exchange reaction is about the same as that in efflux conditions in the presence of external Na+ ions. This observation demonstrates that exchange of melibiose during the Na+-coupled exchange reaction occurs without concomitant movement of the coupled cation. TMG, a galactose derivative with p conformation, is COtransported with either Na+ or Li+ but not with H+ by the melibiose permease (1, 11).As in the case of melibiose, the rate of Na'-coupled TMG influx down a sugar concentration gradient is reduced its V,, is about 4 nmol/mg protein. min as compared with less than 1 for the Na+-melibiose influx (not shown). Also, the rates of downhill Na+-coupled TMG efflux and exchange at equilibrium do not differ by more than a factor of two (11).It was, therefore, of interest to measure the relative rates of Na+ and sugar fluxes during the Na+coupled TMG efflux and exchange reactions. Fig. 2 first shows that the Na+/sugar stoichiometry during the Na+-coupled TMG efflux reaction is approximately 1/1 in both the presence and absence of external Na' ions. Furthermore,a transinhibitory effect of outer Na+ions on Na+-coupledTMG efflux is observed. Moreover, Na+ movements during TMGTMG exchange do not have a 1/1relationship with the sugar movement, again suggesting dissociation of the flows of sugar and coupled cation duringthe sugar exchange reaction. These properties are very similar to those described above for Na+coupled melibiose flows during efflux or exchange reactions. The calculated Na+/TMG stoichiometry is near 0.4, i.e. at least twice as much as that calculated during melibiose exchange (compare with Fig. 1).This fact, in turn, implies that the rate of release of the coupled Na+ during the exchange reaction varies as a function of the chemical structure of the exchanged sugar. Homologous and Heterologous Exchange Reactions by the Melibiose Permease-In a previous study of the facilitated diffusion properties of the melibiose permease, we reported that thepermease catalyzes melibiose exchange and counterflow when coupled to Na+ but not when coupled to Li' (7). W
,
10 0
1o
This observation was tentatively explained by assuming that the sugar can be released from the Na' ternary complex but not from that of Li+ on the inner surface of the membrane. Comparative analysis of the Li+- or Na+-dependent sugar exchange reaction involving exchange of similar sugars (homologous exchange) or different sugars (heterologous exchange) provides useful information in this connection. In one series of experiments, membrane vesicleswere loaded or ["c] with 10 mM Na' ions and either 20 mM [3H]melibio~e TMG; aliquots of each batch of loaded vesicles were then diluted in media containing 10 mM Na+ and either 20 mM unlabeled melibiose or unlabeled TMG. Similar experiments were performed in the presence of 10 mM Li+. The results illustrated in Fig. 3 show first that in the presence of a saturable concentration of co-substrates, the rate of homologous Na'-coupled TMG exchange is faster than that of the homologous Na+-coupled melibiose exchange. It can also be seen that themelibiose carrier mediates significant homologous Li+-coupled TMG exchange when homologous Li+-coupledmelibiose exchange activity is negligible (Fig. 4). The overall rate of the Na+ (orLi+)-coupledexchange reaction is thus dependent on the chemical structure of the exchanged sugar. Second, Fig. 3 indicates that in the presence of Na+, the rate of exchange of internal melibiose is faster with external TMG than with external melibiose; similarly, the rate of exchange of internal TMG is decreased when the external TMG is substituted for melibiose. These data indicate that the overall rate of the sugar exchange reaction is controlled by one (or more) steps linked to the inflow of the externally exchangeable sugar. Even more pronounced effects of external sugar substitution on the sugar exchange reaction are observed in the presence of Li+ ions. Thus, Li+-coupledmelibiosei, + TMGOut exchange is at least six times faster than Li+-coupled melibiosei, + melibioseoutexchange; also, exchange of internal TMG proceeds at a greatly reduced rate when the external exchangeable sugar is melibiose instead of TMG. It is also interesting to note in Fig. 4 that the a analog 1-O-methyl-aD-galactopyranoside (aMG) considerably increases the rate W
I
0 0
9665
ern
[Na+ lo [SI,
U
h
MEL, TMG,
TMG, MEL,
FIG. 3. Na+-coupledhomologous and heterologous exchange reactions catalyzed by the melibiose permease. Left-hand colFIG. 2. Na+/TMG stoichiometry during Na+-coupled TMG umn, concentrated membranevesicles equilibrated in 0.1 M potassium efflux and exchange reactions mediated by the melibiose per- phosphate, 10 mM MgS04 and 10 mM Na+ medium (pH 6.6) and mease. "Na+ outward movements (hatched columns)and ["CITMG loaded with 20 mM [3H]melibiose (Mel;,) were diluted in the same outward movements (open columns) were measured as described in buffered medium containing either 20 mM unlabeled melibiose (hoFig. 1with membranevesicles loaded with 20 mM labeled or unlabeled mologous exchange, open column) or 20 mM unlabeled TMG (heterTMG. The Na+/TMG stoichiometry was measured in conditions of ologous exchange, hatched column). Right-hand columns, same experNa+-coupled efflux in the absence of external Na+(left-hand columns) iment as above using ["CITMG-loaded vesicles (TMG;,) diluted in or in the presence of 10 mM Na+ ions (middle columns); right-hand media containing either TMG (homologous exchange) or melibiose columns, conditions of Na+-coupled TMG exchange. Abscissa, con- (heterologous exchange). S,, in the abscissa corresponds to the excentrations in millimolar of external Na+([Nu+],)and external TMG changeable external sugar. Ordinate, NEM-sensitive components of ([SI,). Ordinate, NEM-sensitive componentsof "Na+ [14C]TMGout[3H]melibiose or ["CITMG outward movements estimated as deward movements. scribed in Fig. l.
20
Melibiose Permease of E. coli
9666
membrane on these two facilitated diffusion reactions. In these experiments, membrane vesicles were loaded with 10 (or 40) mM melibiose in Na+-freemedia at a given pH; efflux and exchange were measured in media with the same composition and pH but in the absence or presence of unlabeled melibiose. Fig. 5 shows first that the rate of melibiose exchange remains low and almost constant over a pH range of 5.5-7.5. On the other hand, the melibiose efflux rate, which is 3-fold higher than the exchange rate and pH-insensitive above pH 6.6, drops sharply as the pHis lowered from 6.6 to 5.5. At this latter pH,melibiose effluxand exchange rates are of comparable amplitude. One is tempted to establish a parallel between the pH-dependent profile of melibiose efflux MEL TMG aMG TMG MEL aMG inhibitionandtitration of the cation binding site of the FIG. 4. Li+-coupledhomologous and heterologous exchange carrier, the pK, of which has been suggested to be about 6.3 reactions catalyzed by the melibiose permease. Concentrated (7). As in the case of Na+-coupled melibiose efflux inhibition membrane vesicles equilibrated in 0.1 M potassium phosphate, 10 mM by external Na+ described in the first part of this study, MgSO,, and 10 mM Li' (pH 6.6) wereloaded with 20 mM [3H] melibiose (Mek, left-hand columns) or 20 mM [14C]TMG (TMGi,, reduction of H+-coupled melibiose efflux at acidic pH could right-hand columns) and diluted in the same buffered medium con- correspond to anincrease in the rateof reprotonation of the taining the sugar indicated below each column at a final concentration unloaded carrier on the external surface of the membrane, of 20 mM. Rates of sugar exchange were calculated as indicated in thus preventing cycling of a fractionof the permeases. InsenFig. 1. sitivity of the melibiose exchange reaction to pH changes could be explained by assuming that protons are notreleased 60 from the carrier during this reaction. A similar interpretation has already been proposed to explain the absence of a pH .-c effect on the lactose exchange reaction catalyzed by the lacE tose permease of E. coli (12). 1
DISCUSSION
The experiments described in the present paper provide additional insight into the catalytic role of the melibiose permease of E. coli and place particular emphasis on the properties of the reaction of substrate exchange at equilibrium catalyzed by the permease in de-energized conditions. Two PH major conclusions can be drawn from the experimental findFIG. 5. Influence of pH on the rates of H+-coupledmelibiose ings reported above: (i) exchange of the physiological subefflux and exchange catalyzed by the melibiose permease. Concentrated membrane vesicles previously equilibrated in 0.1 M strate melibiose by the carrier in the presence of Na+ (or H+) potassium phosphate, 10 mM MgSO4 solutions containing less than proceeds without concomitant movements of the coupling 20 p~ Na' ions and adjusted to a given pH in the range 5.5-7.5 were cation and (ii) the rate of sugar exchange not only depends loaded with [3H]melibioseat a final concentrationof 10 mM. Aliquots on the chemical identity of the coupled cation but is also a of vesicle suspensions were diluted in media of composition and pH function of the chemical structure of the co-transported sugar. similar to thatof the equilibration solution lacking melibiose (efflux) It is of interest to analyze these resultsin terms of the kinetic or supplemented with 10 mM of the sugar. Melibiose efflux and model of Na+-melibiose co-transport briefly described in the exchange rates were calculated as indicated in Fig. 1: 0, melibiose introduction and illustrated in more details in Fig. 6. This efflux rates; 0 , melibiose exchange rates. model, suggested by previous analysis of the sugar-binding of Li+-coupled exchange of internal melibiose, behaving as characteristics of the melibiose permease (6) and studyof the the methyl derivative TMG, on the other hand and as does facilitated diffusion properties of the melibiose carrier (7), the (Y dissacharide melibiose, aMG strongly decreases the rate displays two essential characteristics. In thefirst place, bindof Li+-coupled exchange of internal TMG. The differences OUT IN were of similar magnitude in the presence of Na+ as coupled cation. It is clear from these data that theglucosyl moiety of cc melibiose as well as its spatial location in the (Y position are essential for interaction of the transported sugar with the carrier. The existence of substantialrates of TMGi, + TMG,, melibiosein + TMGout, or melibiosei, --.$ (YMG,,~ C Na' Ne+C exchange coupled to Li+ indicates that association of Li+ with the carrier does not hamper translocation of the ternary complex. It thus seems more likely that the absence of homologous Li+/melibiose exchange is the result of an infrequent dissociation of melibiose from the carrier on the inner surface of the membrane. FIG. 6. Partial reactions of the melibiose carrier cycle durEffect of pH on H+-coupled Melibiose Fluxes-When the ing Na+-coupled melibiose transport in de-energized memmelibiose permease functions as a H+-symport, the rate of brane vesicles. C represents the melibiose carrier and Naf and Mel melibiose efflux down a concentrationgradient is greater than correspond to sodium and melibiose. Note that therate of Na+ release the rate of exchange at equilibrium (7). Fig. 5 illustrates the from the carrier on the inner side of the membrane is slower than effects of varying the pH medium facing both sides of the the rate of all other partial reactions.
tr., Na+l
Melibiose Permease of E. coli ing of the co-substrates on the outer face of the membrane (6) and their release on the inner side are ordered reactions ( 7 ) ; the respective sequences of co-substrate binding (Na+ first, melibiose last) and release (melibiose first, Na+ last) give rise to a mirror-type model. In addition, the rate of Na+ release on the inner surface is postulated to be many times slower than thatof the release of melibiose. It is immediately apparent from the characteristics of the above model that in conditions of Na+-melibiose exchange at equilibrium, the sequence and relative rates of release of the cosubstrates on the inner surface of the membrane together enhance the probability of reforming ternary complexes with internal melibiose; this happens at theexpense of binary Na+ carrier complexes which otherwise tend to predominate because of the slow rate of Na' release. Sugar exchange could then proceed by the stages shown in the lower branch of the model, bypassing the stepof Na+ release. In thiscase, kinetic occlusion of the coupled cation should occur during the sugar exchange reaction. The absence of Na+ movements noted during the Na'-coupled melibiose exchange reaction (Fig. 1) provides experimental evidence for this deduction. A reduced rate of release ofNa' from the carrier on the inner surface is in striking contrast to thesubstantial rate of release of Na+ on the outer surface. A rapid Na' release outside is indeed suggested by the high rate of Na+-coupled melibiose efflux down the co-substrateconcentration gradients (Fig. 1).It is tempting to suggest that theasymmetrical functioning of the melibiose permease previously outlined, i.e. net influx down a sugar concentration gradient much slower than net efflux down a sugar concentration, is consecutive to the asymmetrical rate of Na+ release from the carrier on the inner andouter surface of the membrane. In agreement with previous evidence ( 7 ) ,these observations further indicate that the kinetic model of Na+-melibiosetransport proposed in Fig. 6 convincingly describes the properties of facilitated diffusion catalyzed by the melibiose permease in the presence of the physiological sugar melibiose and Na+ions. The observation that the Na+/TMG stoichiometry is only 0.4 during the TMG exchange reaction (Fig. 2) further indicates that the coupled flows of co-substrates are dissociated during this reaction. It is important to note, however, that the rate of Na+ outward movement is significantly faster during TMG exchange than during melibiose exchange (Fig. 1).Since a higher rate of Na+ flow during exchange implies a higher rate of Na+ release from the carrier (and in particular from the inner cationic site of the permease), it follows that the rate of release of the coupled cation from the permease varies with the chemical structure of the sugar and more precisely depends on the type of interaction of the sugar with the carrier. Analysis of the homologous and heterologous exchange reactions (Fig. 3 and 4) clearly shows that the overall rate of sugar exchange mediated by the melibiose permease is controlled by one (or more) steps linked to the inflow of the externally exchangeable sugar. At a saturating concentration of sugar and coupled cation, the limiting partial reaction could alternatively be a translocation of the ternary complex or a release of the co-substrates from the carrier on the inner
9667
membrane surface. Either explanation would account for the inability of the melibiose carrier to mediate Li+-melibiose exchange (Ref. 7 and Fig. 4). It is apparent from the data shown in Fig. 4, however, that the melibiose carrier does promote a Li+-coupled exchange reaction in the presence of external TMG or aMG. This latter observation eliminates the possibility that association ofLi' with the carrier, in itself, prevents the Li+ ternary complexes from crossing the membrane. A more likely explanation is that the rate of dissociation of sugars from the Li+ ternary complex varies in relation to thechemical structure of the co-transported sugar and is therefore controlled by the strength of the interaction of the sugar with the carrier in the ternarycomplex. From the above considerations, it would appear that release of co-substrates from the carrier on the inner surface of the membrane displays three characteristics: first, the rate of release of the coupled cation varies as a function of the strength of interaction of the sugar with the carrier; second, as can be logicallyexpected, the rateof sugar release depends on its interaction with the carrier. In addition, the early observation (Ref. 7 , see also Fig. 4) indicating that melibiose can be exchanged in thepresence of Na+ but not when coupled to Li+ suggests that the extent towhich the cation interacts with the carrier affects release of the sugar. A hypothesis based on these observations, and in general on the variability of the facilitated diffusion properties of the melibiose carrier in thepresence of different sugar analogs or different coupled cations, is that therelease characteristics of the co-substrates into the intravesicular medium varies according to the conformation of the ternary complexes, conformation resulting from coordinated interactions between the carrier and the two co-substrates. Acknowledgments-We express ourgratitude to Drs. I. C. West and M. G. P. Page for helpful discussions, to R. Lemonnier and P. Lahitette for excellent technical assistance, and A. Giovagnoli for typing the manuscript. REFERENCES 1. Tsuchiya, T., and Wilson, T. H.(1978) Membr. Biochem. 2 , 6379 2. Wilson, D. M., and Wilson, T. H.(1986) in Ion Gradient-coupled Transport, Institut National de la Santi. et dela Recherche eds) Mbdicale Symposium 26 (Alvarado,F., and Van Os, C. H., pp. 97-104, Elsevier Scientific Publishing Co. Inc., New York 3. Tsuchiya, T.,Wilson, D. M., and Wilson, T. H.(1985) Ann. N . Y. Acad. Sci. 466, 342-349 4. Tanaka, K., Niiya, S., and Tsuchiya, T. (1980) J. Bacterid. 141, 1031-1036 5. Bassilana, M., Damiano-Forano, E., and Leblanc, G. (1985) Bwchem. Bwphys. Res. Commun. 1 2 9 , 626-631 6. Damiano-Forano, E., Bassilana, M., and Leblanc, G. (1986) J. Bwl. Chem. 261,6893-6899 7. Bassilana, M., Pourcher, T.,and Leblanc, G. (1987) J. Bwl. Chem. 262,16865-16870 8. Lopilato, J., Tsuchiya, T., and Wilson, T.H.(1978) J. Bacteriol. 1 3 4 , 147-156 9. Kaback, R. H.(1971) Methods Enzymol. 22,99-120 10. Lowry, 0.H.,Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 11. Cohn, D. E., and Kaback, H.R. (1980) Biochemistry 19, 42374243 12. Kaczorowski, G . J., and Kaback, H. R. (1979) Biochemistry 18, 3691-3697
