Substrate and inhibitor binding and translocation ... - Semantic Scholar

Nervous System Membrane Proteins

15. Radian, R. & Kanner, B. 1. (1985) J. Biol. Chem. 260, 11859-11865 16. Radian, H.. Bendahan, A. & Kanner, €3. I. (1986) J. Biol. Chem. 261, 15437- 15441 17. Danbolt. N. C.. Pines, G. & Kanner. B. I. (1990) Biochemistry 29.6734-6740 18. Lopez-Corcuera, B., Kanner, B. 1. & Aragon, C. (1989) Biochim. Biophys. Acta 983.247-252 19. Shouffani, A. & Kanner, B. 1. (1990) J. 'Biol. Chem. 265,6002-6006 20. Kanner, B. I., Keynan, S. & Radian, K. (1989) Biochemistry 28.3722-3727 21. Radian, K., Ottersen. 0. I,., Storm-Mathisen, J., Castel. M. & Kanner, €3. I. (1990) J. Neurosci. 10, 13 19- 1330 22. Bowery, N. G., Jones, G. P. & Neal, M. J. (1976) Nature (London)264,28 1-284

23. Neal, M. J. & Bowery, N. G. (1977) Brain Res. 138, 169- 174 24. Schon, F. & Kelly, J. S. (1975) Brain Res. 86,243-257 25. Guastella, J., Nelson, N., Nelson, H., Czyzyk, L., Keynan, S., Miedel, M. C., Davidson, N. C., Lester, H. A. & Kanner, B. I. (1990) Science 249,1303-1306 26. Mandel, M., Moriyama. Y., Hulmes. J. D.. Pan, Y. C. E., Nelson, H. & Nelson, N. (1988) Proc. Natl. Acad. Sci. U.S.A. 85.5521-5524 27. Snutch, T. P.. Leonard, J. P., Gilbert, M. M., Lester, H. A. & Davidson, N. (1990) Proc. Natl. Acad. Sci. U S A . in the press

Received 30 July 1990

Substrate and inhibitor binding and translocation by the platelet plasma membrane serotonin transporter Cynthia J. Humphreys, David Beidler and Gary Rudnick Department of Pharmacology, Yale University School of Medicine, PO Box 3333, New Haven, CT 065 10-8066, U.S.A.

Summary The Na+ and CI- dependence of imipramine binding and dissociation were determined in platelet plasma membrane vesicles. Equilibrium imipramine binding affinity depends on Na+ binding to two non-interacting, low-affinity sites. Binding of a single CI- ion also enhances imipramine affinity. Imipramine dissociation is inhibited by Na and CI-, indicating that both ions can bind after imipramine. Of the two Na+ ions required for imipramine binding, only one is involved in slowing imipramine dissociation, indicating that imipramine binding makes the two Na+ ions non-equivalent. The initial rate of imipramine association is strongly Na+-dependent, suggesting that Na+ binds prior to imipramine. C1-, however, affects imipramine dissociation but not association. Thus, while Na+ and CI- can bind either before or after imipramine, kinetic considerations impose a most likely binding order of first Na+, then imipramine and finally Cl-. We have confirmed and extended these conclusions using serotonin exchange and efflux measurements. Efflux of radioactivity from vesicles preloaded with [ 3H]serotonin is stimulated by both external K and external unlabelled serotonin. K acts to accelerate a step that is rate-limiting for net efflux but that does not involve Na+, C1- or sero+

+

+

tonin translocation. Unlabelled serotonin accelerates radioactivity efflux by exchanging with intravesicular label. This serotonin exchange requires external CI-, but not external Na+. These results suggest that first Na+, then serotonin and finally CI- bind from the external medium. Although serotonin exchange requires external C1-, internal CIis not required. These results suggest that translocation does not disturb the spatial order of bound substrates, which dissociate internally in a first-infirst-out order.

Introduction The serotonin transporter is responsible for terminating the action of serotonin released from nerve terminals, and for serotonin accumulation by platelets [ I l l . Inhibitors of this process, such as imipramine, are clinically useful in the treatment of depression. Moreover, platelets and post-mortem brain tissue from depressed patients transport serotonin more slowly and bind less imipramine than controls [2, 8, 141. The serotonin transporter also represents the most sensitive known target in the brain for cocaine inhibition [9]. Serotonin translocation across the plasma membrane requires cotransport of Na+ and CI- and countertransport of K + [6, 101. Previous results have suggested that

1991

95

Biochemical Society Transactions

serotonin, Na+ and CI- are all translocated in a single step, and that K + is translocated in a separate step of the catalytic cycle [6]. Imipramine binding is Na -dependent and competitive with serotonin, as expected if binding occurs at the normal substrate-binding site of the transporter. The lack of detectable imipramine transport [ 123 suggests that imipramine binding could be used as a probe of the steps leading to formation of the transporter complex with serotonin, Na+ and CI-. Previous studies using porcine platelet plasma membrane vesicles demonstrated a requirement for more than one Na+ ion for maximal imipramine binding, although a single Na+ ion is apparently co-transported with serotonin [ 131. The ability to use imipramine as a probe of substrate binding to the serotonin transporter has led us to examine the binding interactions between Na+, CI-, serotonin and imipramine. In this paper we describe experiments using human platelet plasma membranes at 25°C which examine the positive co-operativity in the binding of Na+, CIand imipramine. Predictions from these experiments concerning substrate-binding order were confirmed by equilibrium exchange experiments with serotonin. +

96

Results 6inding order for ions, imipramine and serotonin Working under the hypothesis that imipramine binding is analogous to serotonin binding prior to transport, we examined the ionic dependence of imipramine binding. Our general model of serotonin transport predicts that Na+, CI- and serotonin all bind to. the transporter prior to translocation. To test this model we measured the Na+ and CI- dependence of equilibrium imipramine affinity, and of association and dissociation rates. In the absence of Na+, no specific imipramine binding is detectable. At 50 meq/l Na+, low-affinity binding is detected, and as "a+] increases from 50 to 300 meq/l the affinity increases while the maximal binding remains constant. The effect of Na+ on imipramine affinity ( 1/KD) is shown in Fig. 1(a). It is clear from the sigmoid shape of the plot that more than one Na+ ion participates in the binding process. We then examined the CI- concentration dependence of imipramine affinity in the presence of high Na+. In the absence of CI-, low-affinity binding is still detectable. It is still Na+-dependent and blocked by serotonin [13]. The results of a series of experiments to measure the effect of CI-

Volume 19

Fig. I Effect of N a + and CI- on imipramine affinity (a) Na+ dependence. Equilibrium imipramine binding t o platelet plasma membrane vesicles was measured as described previously [I31 at 0-300 meq/l Na+ and 0.25-30 n~-[~H]imipramine. The lines are drawn using K, values determined from a simultaneous nonlinear regression fit of all the experimental binding data using the model described in Fig. 3. (b) CI- dependence. Binding was measured at 0-300 meqA CI- and 0.25-45 nM[3H]imipramine. Each point is the reciprocal of an apparent K, for imipramine determined from the direct binding data. 0.4 - ( a ) 3

13 -X

'E

v

0.2-

$ 0.0

I r / o 0 100 200 300 0 "a'],

meq/l

.

o

200 LCl'], meq/l 100

f

300

on affinity is shown in Fig. l ( b ) . They indicate simple hyperbolic saturation as predicted if only one CI - ion participated in imipramine binding. Using the SCoP computer program, we simultaneously fit all the equilibrium binding data at various Na+ and CI- concentrations. The simplest model consistent with the equilibrium binding data predicts that imipramine affinity is enhanced by binding of two Na+ ions with equal affinity to independent sites. The model predicts that at least one of the Na+ ions must bind prior to imipramine. Chloride binds more tightly to the transporterimipramine complex, according to this analysis. The lines in Fig. 1 represent the calculated variation of equilibrium binding affinity as a function of Na+ and CI-. The model is clearly capable of explaining the variation of imipramine affinity as a function of Na+ and CI-. The imipramine association rate depends on Na+ in the same sigmoidal fashion as equilibrium affinity (not shown). The strong Na+ dependence of association confirms the model's prediction that Na+ binds before imipramine. We were surprised to observe that, although CI- increases equilibrium imipramine affinity, it does not stimulate (in fact it inhibits slightly) the imipramine association rate. These results imply that the predominant binding sequence is Na+, imipramine and finally CI-.

Nervous System Membrane Proteins

Although this sequence may predominate, it is not strictly ordered, as shown by imipramine dissociation kinetics (not shown). The imipramine dissociation rate is inhibited by Na+, indicating that at least one Na+ ion can dissociate before imipramine. The inhibition by Na+ saturates with binding of a single Na+ ion, as if imipramine dissociation increases when one of the two required Na+ ions dissociates from the transporter. Like Na+, C1- also slows down imipramine dissociation, but high CI - does not completely inhibit, indicating that imipramine can occasionally dissociate prior to C1-. Order of ion binding in serotonin transport and exchange

These results provide information about the binding order for Na+, imipramine and CI-. As a competitive inhibitor of serotonin transport, imipramine is likely to bind to the serotonin site on the transporter. The dependence of imipramine affinity on Na+ and CI- [13], therefore, suggests that the transporter also forms a quaternary complex with serotonin, Na+ and C1- prior to transport. There are few examaples of multisubstrate transport proteins where the order of substrate and ion binding is established. The best studied system is the Na+dependent glucose transporter for which Hopfer & Groseclose [3] proposed that first Na+ and then glucose bind to the external face of the transporter. The order of binding was based on equilibrium exchange studies (which indicated an obligatory binding order) and the observation by Aronson [ 11 that binding of the competitive inhibitor phlorizin required external Na . More recent studies on phlorizin binding carried out by Moran et d. [51 indicate many similarities with the present work on imipramine binding by the serotonin transporter. In both cases two Na+ ions must bind for maximal inhibitor binding. Na+ also accelerates the rates of both imipramine association with the serotonin transporter and phlorizin association with the glucose transporter [ S ] . Thus, Na+ ion is likely to bind to the unliganded form of each transporter. In the case of the serotonin transporter, the "a+] dependence of the imipramine association rate mirrors that of equilibrium binding of about 280-380 (Fig. l), with an apparent meq/l for two equivalent Na+ ions. This close correspondence indicates that the Na+ dependence of equilibrium binding is primarily determined by the Na+ requirement for association. If imipramine binding reflects the same process as serotonin binding prior to transport, then the binding model of first Na+, then serotonin and +

finally CI- ,makes certain predictions about equilibrium serotonin exchange. The ability of external, unlabelled serotonin to stimnulate efflux of intravesicular [3H]serotonin requires external C1- , as expected if C1- must dissociate prior to the exchange of unlabelled for labelled serotonin in the external medium (Fig. 2a). Na+ could presumably remain bound during the isotopic exchange, and Fig. 2(b) shows that a portion of the exchange is independent of external Na+. In a symmetrical mechanism, internal Na+ and CI- would be expected to have similar effects. However, internal C1- has little effect on serotonin exchange (not shown), suggesting that the order of binding is different on the two sides of the membrane and that Na+, serotonin and CI- are transported in a firston-first-off order. These results clearly differentiate between the roles of Na+ and CI- in serotonin transport. Moreover, they reflect the similarity between serotonin and imipramine interaction with the transporter. Fig. 2 Effect of external N a + and CIexchange

on serotonin

( a ) CI- dependence. Vesicles which had accumulated

[3H]serotoninwere diluted into 200 mM-NaCI containing 10 mdithium phosphate buffer, pH 6.7, I mM-MgSO, in which NaCl was replaced to the indicated extent with sodium isothionate to maintain Na+ and isotonicity. Efflux was measured 30 s after dilution into serotoninfree medium (not shown). In parallel experiments, efflux was measured in the same media containing 0.5 ,UMserotonin. Exchange is the increase in efflux which results from addition of external serotonin. Values represent averages of three measurements and the error bars represent standard deviations. ( b ) Na+ dependence. Similar measurements were made in solutions of varying Na+ concentrations formed by replacing NaCl with LiCI.

0

50 [Cl], meq/l

I00

00

100

200

3001

"a+ J, meq/l

1991

97

Biochemical Society Transactions

98

The results imply that external serotonin binds to the transporter before CI- in an ordered process and that they move across the membrane in a single file. In contrast, external Na' appears to bind to the free transporter. External CI- is absolutely required for net influx [4,71 and internal C1- is required for efflux [ 7 ] .Measurements of serotonin exchange, however, illustrate the basic asymmetry of the CI- effect. Imipramine binding measurements provide evidence that CI- binds after imipramine, presumably on the vesicle exterior. The CI- requirements for serotonin exchange suggest that serotonin and CI- bind in an ordered sequence. Serotonin exchange requires external C1- (Fig. 2) apparently reflecting the obligate loss of the CI- ion transported out with [3H]serotonin prior to replacement of ['Hlserotonin with external unlabelled serotonin. This CI - requirement is conspicuously absent on the vesicle interior. Exchange is at least as rapid in the absence of added internal CI- as in its absence. We interpret these results as evidence for an ordered sequence of first serotonin and then CIbinding externally which is reversed internally. Internally, CI- evidently remains bound while unlabelled serotonin dissociates and ['Hlserotonin binds. An obvious interpretation of ordered CI- and serotonin binding is that the serotonin transporter displays 'glide symmetry' as defined originally by Hopfer & Groseclose [3]. By this interpretation external serotonin and CI- would bind in a narrow pore which, by conformational change, would be exposed to the cytoplasmic membrane face without altering the spatial order of ligands (Fig. 3). Hence, Fig. 3

Model for serotonin transport The thick black line separates the two forms of the transporter which bind and release substrates on the two sides of the membrane. S represents serotonin.

they would dissociate in a 'first-in-first-out' order. The lack of an absolute external Na+ requirement for serotonin exchange (Fig. 2) implies that ['Hlserotonin can dissociate and be replaced with external unlabelled serotonin without Na+ dissociation. Na+ also dramatically inhibits efllux (not shown), an effect we attribute to a second Na+ ion binding to the serotonin-transporter complex. Fig. 3 shows a schematic diagram of the entire transport cycle of the serotonin transporter. Starting with the free transporter in the lower left, the first steps are the sequential binding of external Na+, serotonin and CI-. As depicted in the diagram, binding of each ligand alters the conformation of the transporter so as to create the binding site for the next ligand. When all the sites are occupied, the transporter undergoes a further conformational change which allows access of the bound ligands to the cell interior (Fig. 3, upper right). Because this conformational change does not involve significant movement of the bound solutes, the order in which they dissociate to the interior is the opposite of the external dissociation order. When all of the solutes have dissociated, the transporter binds an internal K + ion (Fig. 3, upper left). This permits a conformational change that allows K + to dissociate to the external medium and finally generates the free transporter to which external solutes bind. 1. Aronson, I). S. (1978) J. Membr. Hiol. 4 2 , s 1-98 2. Hriley, M. S., Langer, S. %.. Kaisman, K.. Sechter, I). &

Zarifian, E. (1 980) Science 209,303-305 3. Hopfer, U. & Groseclose, K. (1980) J. Hiol. Chem. 255,4453-4462 4. Lingiaerde, 0.. Jr. (1971) Acta I'hysiol. Scand. 81, 75-83 5. Moran, A., Davis. I,. J. & Turner, K.J. (1988) J. Hiol. Chem. 263. 187-192 6. Nelson, P. J. & Kudnick. G. (1979) J. Hiol. Chem. 254, 10084- 10089 7. Nelson, P. J. & Kudnick, G. (1982) J. Hiol. Chem. 257, 615 1-61 55 8. Perry, E. K., Marshall, E. F., Blessed, G., Tornlinson, B. E. & Perry, K. H. (1983) Br. J. Psychiat. 142, 188-192 9. Ritz, M. C., Lamb, K. J., Goldberg, S. K. & Kuhar, M. J.(1987)Science237,1219-1213 10. Rudnick, G. (1977) J. Hiol. Chem. 252,2170-2174 11. Sneddon, J. M. (1973) Prog. Neurobiol. 1, 15 1-198 12. Talvenheimo, J., Nelson, P. J. & Kudnick, G. (1979) J. Hiol. Chem. 254,4631-4635 13. Talvenheimo. J.$ Fishkes, H.. Nelson. P. J. & Kudnick, G. (1983) J. Biol. Chem. 258,6115-6119 14. Tuomisto. J. & Tukiainen, E. (1976) Nature (London) 262,596-598

a'" M

11

Received 20 August 1990

Volume 19