L. G. Abood, F. Takeda, and N. Salem, Jr.

Center for Brain Research and Department of Biochemistry, University of Rochester Medical Center, Rochester, New York, 14642.

 

INTRODUCTION

 

The molecular nature and configuration of the opiate receptor is a problem under intensive investigation in a number of laboratories (1,2,3). Our interest in the problem developed as a result of a finding of others (4) that a proteolipid fraction from brain may be responsible for stereospecific opiate binding observed in various preparations of brain tissue. Upon examining various lipids, proteolipids, and proteins from membranes derived from brain, it was found that phosphatidyl serine (PS),t the major acidic lipid in brain, exhibited stereospecific opiate binding (5). Since the Kd's for the opiate-PS complex and the opiate-tissue complex differed by three orders of magnitude, it was recognized that the binding to PS was nonspecific and different from the high affinity binding to membrane preparations. What interested us was the possibility that PS, in the form of a complex with a membranous protein, may be an important component of the opiate receptor.

 

In an effort to test this hypothesis a study was undertaken on the effect of added phospholipids on opiate binding. It had been known for many years that exogenous lipids were capable of exchanging with endogenous ones with no apparent alteration in their natural functional or structural characteristics (6). It was demonstrated that the addition of PS to suspensions of synaptic membranes significantly enhanced both high and lower affinity binding, the Kd's without lipid being 1.0 x 10-9 M and 5.7 x 10-9 M and those with lipid being 5.0 x 10-10 M and 3.8 x 10-7 M respectively (7). Another observation of interest was that exogenous PE was inhibitory to opiate binding. The present report describes studies aimed at extending the initial findings on the PS-enhancement and to attempt to understand the role of PS and other phospholipids in opiate binding.

 

MATERIALS AND METHODS

 

The procedures for the preparation of synaptic membranes from rat brain are described elsewhere (7). The technique for measuring the binding of 3H-DHM was essentially that of Pert and Snyder (3) with slight modifications (7). Other details concerning materials and methods are described in the legends of details and in the text.

 

RESULTS AND DISCUSSION

 

ENHANCEMENT OF OPIATE BINDING BY VARIOUS PHOSPHOLIPIDS

 

A number of phospholipids associated with most biological membranes were tested for their enhancement effect of the binding of 3H-DHM to a membrane preparation of rat brain. The three acidic lipids, PS, PA, and PI, produced a 35%, 29%, and 28% increase respectively when added at a concentration of 100 ug/m1 (Table 1). PC was without any effect, while PE resulted in a 10% decrease in binding. In a previous study it had been reported that a commercial preparation (from egg yolk and 70% pure) of PA was inactive (7). However, the present sample of synthetic dipalmitoryl PA (Sigma Chemical Co.), which was over 98% pure, proved to be active. As described elsewhere (7) other lipids such as cholesterol, diglycerides, and cerebrosides were without effect, while some enhancement (10%) was observed with cerebroside sulfates.

 

Preparation of synaptic membrane preparation and experimental conditions for measurement of stereospecific opiate binding described elsewhere (7). Lipids (100 ug/2 mg membrane protein) were homogenized with membranes and incubated 15 minutes prior to addition of 3H -DNM (0.05 1Ci/1.2 ml) incubation medium containing 50 mM Tris, pH 7.5 and 10-7 M of either levorphanol or dextrorphan. Binding measured after filtration through GF/B glass fiber filters. Results expressed in counts/minutes (cpm), representing difference in radioactivity in presence of dextrorphan and levorphanol.

 

ENHANCEMENT BY VARIOUS SPECIES OF PS WITH VARYING FATTY ACID COMPOSITION

 

Recently, Salem et al. (8) have succeeded in separating and purifying the various species of PS from rat brain. At least four distinct species were identified, and their fatty acid composition is described elsewhere (8). In an effort to determine whether the fatty acid profile of PS influenced the opiate binding of neural membranes, the various species were tested for their ability to enhance the binding of 3H-DHM to membrane preparations (Table 2). The 18:1, and 18:0-PS were as effective as the PS mixture in enhancing opiate binding, while the 18:0, 22:4 and 18:0, 22:6-PS were less active. As will be described in a subsequent publication the degree of unsaturation in the aliphatic group is an important factor in the lipid enhancement of opiate binding. Although the 22:6 species is the major PS of synaptic membranes, some 18:1-PS is also present. The implications of these findings are presently under investigation.

 

Lipids were prepared from bovine brain by argentation thin layer chromatography as described elsewhere (8).

 

EFFECT OF VARIOUS PHOSPHOLIPASES ON OPIATE BINDING

 

One approach to determining the role of lipids in opiate binding has involved the use of lipases. It has been reported that 3H-naloxone binding to rat brain homogenates is inhibited completely by a phospholipase A from V. russelli, 40% by phospholipase C (Clostridium welchii), and unaffected by phospholipase C (cabbage) (9). Although these observations suggest that phospholipids are involved in opiate binding, their interpretation is somewhat difficult. For example, the lysophospholipids formed by phospholipase A are strong detergents, which are themselves inhibitory. Another problem is the purity of the enzymes, which may contain proteases as well as various toxins.

 

We have been able to essentially confirm the findings of Pasternak and Snyder (9) utilizing a similar array of enzymes. After treatment of membranes with phospholipase C (B. cereus) opiate binding was reduced 45%, and although PS and PA were able to enhance binding after enzymic treatment, the activity was still below the control (Table 3). A combination of phospholipase D and C resulted in an 84% inhibition of opiate binding, while the addition of either PS or PA produced only a slight enhancement (Table 4). The effect of the lipases, therefore, appears to be irreversible, possibly because of a configurational change in the receptor caused by the lipase or other enzyme.

Membranes were incubated with 0.5 unit of phospholipase C (Bacillus cereus) per mg membrane protein for one hour at 35° in 50 mM CaC12 - 50 mM Tris buffer, pH 7.5. Membranes were centrifuged at 100,000 x g, washed once with 50 mM Tris - 1 mM EDTA, pH 7.5. Aliquots of membrane preparation were homogenized with 100 pg of each lipid and pre-incubated 15 minutes before addition of 3H-DHM and 10-7 M of either levorphanol or dextrorphan. The lipase caused a 45% inhibition in absence of added PS.

 

In an effort to determine the extent of enzymic degradation of the membranes by phospholipase C, an analysis was made of the membrane lipids. Even after exposure of the membranes for two hours at a concentration of 2 units/mg membrane protein, there was no detectable change in the lipid content, utilizing quantitative thin layer chromatography (5). Another possibility is that impurities present in the enzyme were responsible for the inhibition, for example, a nonenzymic inhibitor is present in a phospholipase C prepared from Cl. welchii (Sigma Chemical Co.). In any case, the findings to date with the lipases are difficult to interpret.

 

Experimental conditions same as in Table 1 except for the addition of phospholipase A (Vipera russelli) at concentration of 2 units/mg membrane protein.

 

EFFECT OF DETERGENTS ON OPIATE BINDING

 

In attempting to solubilize the opiate receptor, a variety of detergents were employed and found to be inhibitory to varying degrees. The membrane preparation was assayed for stereospecific opiate binding after being exposed to a 0.01% solution of the detergent. The nonionic detergent triton X-100 produced a 60% inhibition of opiate binding at a concentration of 0.01%, while the three cationic detergents, CTAB, CTC, and FSC, at the same concentration resulted in a 39%, 21%, and 15% inhibition respectively (Table 5). With the anionic detergent, SDS, a 25% inhibition occurred. Upon the addition of PS to the detergent-treated membranes, a marked increase (40%) in binding was observed with CTAB, while little or no effect was observed with the other detergents. The detergent-solubilized material was analyzed for opiate binding after removal of the excess detergent of Sephadex G-25 columns, and only with the CTAB-solubilized material was activity detectable (data not shown).

 

Results are expressed as % differences in stereospecific binding of 3H-DHM.

 

RESTORATION OF OPIATE BINDING WITH VARIOUS LIPIDS AFTER CATIONIC DETERGENTS

 

In an effort to understand the mechanism by which cationic detergents were inhibiting opiate binding, various lipids were tested for their ability to restore binding (Table 6). After membranes were exposed to 0.01% CTAB for 30 minutes, they were centrifuged at 100,000 x g, washed with 50 mM Tris, pH 7.5, and then re-centrifuged. (As discussed bearlier, this preparation has only 60% of the binding activity of CTAB untreated.) Of the various lipids tested only the two acidic phospholipids, PA and PS, were effective in restoring opiate binding. The PS-ester, although enhancing opiate binding of untreated membranes, was without effect, as were PE and PC.

 

On the basis of these results there are two possible mechanisms for the effect of cationic detergents on opiate binding. The detergent may be interacting either with an anionic site on the protein component of the receptor or with the associated PS. Acidic lipids, by combining with CTAB, may, thereby, permit the receptor to assume its natural configuration. Another possibility is that the CTAB occupies an anionic site, either on the protein or lipid, and its removal by an acidic lipid frees the site for the opiate. Since the PS-ester is ineffective after CTAB-treatment but not before, it might seem more likely that the acidic lipids were acting by removal of the CTAB, rather than by activation of sites not occupied by CTAB. As will be discussed, however, the PS-ester may still retain an anionic charge.

Microsomes were treated with 0.01 CTAB and washed before exposure to lipids (0.1 mg/ma). Results are expressed as counts/minute, using 10-7 M dextrorphan (d) or levorphanol (1).

 

ENHANCEMENT OF OPIATE BINDING BY AN ESTER OF PS

 

In order to determine whether the -000- group of PS was involved in the enhancement effect of the lipid, W. Hoss in this laboratory prepared an ester of PS from bovine brain PS, which has the following structure:

As is shown in Table 1, the PS-ester produced a significant enhancement effect on opiate binding. This finding suggests that the -000 group is not essential to opiate binding, and that the PS-ester is able to interact with the opiate receptor complex similarly to PS. Another possible explanation is that the PS-ester undergoes enzymic hydrolysis in the vicinity of the receptor and is converted to PS itself. It has nnot been possible to demonstrate any hydrolysis of the PS-ester by the membrane preparation under the experimental conditions for opiate binding. As is discussed, however, the number of membranous PS molecules involved in receptor binding is extremely small. Therefore, the possibility of hydrolysis cannot be excluded. Another explanation for the effectiveness of the PS-ester is that the -PO- is still available for interaction with either the opiate or protein, since the -C=0 groups of the ester are capable of forming H bonds with the -NH2 group.

 

In the studies dealing with the ability of phospholipids to restore opiate binding after inhibition with CTAB, PS was effective while the PS-ester was not (Table 6). Such a difference supports the argument that the -000 group of PS is required for interaction of the bound CTAB, and although it favors the hypothesis that the PS-ester was not hydrolyzed, it is still conceivable that the extent of hydrolysis was insufficient to overcome the effect of CTAB. Nevertheless, the evidence to date favors the hypothesis that the PS-ester itself enhances opiate binding; therefore, the molecular configuration of the endogenous receptor complex has not been altered by the presence of the ester.

 

EFFECT OF CROSS-LINKING OF AMINO GROUPS ON OPIATE BINDING

 

By means of the reagent difluorodinitrobenzene (DFDNB), the amino groups of PS and, to a lesser extent, PE can be cross-linked to neighboring phospholipids and proteins within the erythrocyte membrane (10). In our laboratory it has been .recently demonstrated that synaptic membranes undergo the same kinds of lipid-lipid and protein-lipid interactions observed for erythrocyte ghosts (10). After the exposure of neural membranes to 100 pM DFDNB stereospecific opiate binding of 3H-DHM was decreased to 74% of the control (Table 7). Furthermore, upon the addition of PS to the DFDNB-treated membranes, there is no enhancement of opiate binding.

 

One interpretation of these observations is that the cross-linking of phospholipids to one another and to proteins resulted in either a partial structural alteration in the receptor complex or in restricting the availability of PS in the vicinity of the complex. The fact that exogenous PS did not enhance binding after cross-linking would indicate that either PS could not penetrate the membrane or gain access to the binding site, possibly because of the inability to exchange with endogenous phospholipids. It has been shown (10) that the erythrocyte membrane becomes refractory to hypotonic lysis after exposure to cross-linking reagents. Furthermore, the reagents tend to decrease the permeability of the membrane to anions (11).

Stereospecific 3H -BUM binding was measured to synaptic membranes treated with 100 ILM DNDFB in 50 mM NaHCO3, pH 8.5 for one hour at 28° C, centrifuged at 100,000 x g, washed with 50 mM Tris, pH 7.5, and assayed for opiate binding.

 

Preliminary experiments with the amino-crosslinking reagent, diflurodinitrobenzene indicate that PS is preferentially associated with protein in the synaptic membrane. When the covalently reacting reagent is incubated with a nerve ending preparation at pH 8.5 and room temperature, the lipid extract contains relatively large amounts of labelled PE and smaller amount of labelled PS. Similar results have been obtained in the red blood cell membrane (10). Quantitative analysis shows that 84% of the phospholipid crosslinked to protein is PS; the remaining 16% is PE. Further experiments are needed to establish a protein-PS complex in the synaptic membrane, however.

 

Other lines of evidence also suggest an interaction of PS with proteins. PS is known to activate enzymes such as Na+K+ ATPase (12) and tyrosine hydroxylase (13). It is also interesting to note that brain PS is more effective in Na+K+ ATPase activation than is PS prepared from egg yolk (12). Circular dichroism experiments have shown that PS can change the conformation of basic polypeptides (14) and nuclear magnetic resonance studies of intact bovine retinal rod outer segment membranes have suggested that a polyunsaturated species of PS is associated with protein, mainly rhodopsin (15).

 

EFFECT OF Ca2+ ON OPIATE BINDING

 

Ca2+ has been shown to interfere with opiate binding (3) as well as the pharmacological effects of opiates (16). Furthermore, morphine inhibits the phospholipid-facilitated transport of Ca2+ (17) and results in a depletion of brain Ca2+ in vivo which is reversed by naloxone (18). From such observations it would appear that the pharmacological action of the opiates is related to that of Ca2+. One possibility is that Ca2+ by combining with PS or other phospholipids, may either compete directly with the opiates for an anionic site on the receptor or modify the configuration of the receptor.

 

In order to examine this possibility, the enhancement of opiate binding by PS was compared with that occurring in the presence of the Ca2+ salt of PS. Only a slight enhancement (9%) was observed with Ca2+-PS (Table 8). When 1 mM Ca2+ was present, the addition of PS failed to produce any enhancement of 3H-DHM binding. It is difficult to determine from these studies whether Ca2+ inhibits opiate binding by occupying an anionic site on the receptor independently of the lipid, or whether it interferes with the action of the phospholipid component by complexing with PS or other acidic lipids.

 

POSSIBLE MECHANISMS FOR PS ENHANCEMENT OF OPIATE BINDING

 

Since PS by itself is known to interact stereospecifically' with opiates, a question to consider is whether it can still do so as a normal constituent of the membrane. A comparison of the Kd's for opiate binding of PS alone and membrane fragments reveals a difference of three orders of magnitude. Consequently, the binding characteristics of the two systems must be very different. There are, however, similarities between the two, including pH optimum and inhibition by divalent cations (e.g., Ca2+ and Mg2+) and by high ionic strength. It is likely that the molecular interactions existing in the membrane between PS and proteins are perturbed by the opiates and other exogenous ligands and that the molecular complementarity of the ligand involves a PS-protein complex. If it is assumed that PS is a component of the opiate binding site, then only a very small percentage of the total endogenous PS participates in binding.

Experimental conditions were the same as described in Table 1 except that membranes were prepared in absence of EDTA.

 

If the amount of PS in whole rat brain is assumed to be about 10 pmoles/g, then the number of PS molecules would be about 6 x 1018/g. From a Scatchard analysis the number of opiate binding sites/g of whole rat brain was estimated at 1013 (19). Since the number of PS molecules exceeds the binding sites by 5 to 6 orders of magnitude, it must be concluded that the determining factor in binding is the number of specific proteins or other macromolecular species of which PS is a component.

 

In attempting to interpret the results of the various phospholipases on opiate binding, it is essential to first determine whether they are due to a removal of PS from the membrane or to a general effect on membrane structure produced by the chemical alteration of the phospholipids. The lysophospholipids resulting from exposure to phospholipase A are known to be strong detergents and their presence in the vicinity of the opiate receptor may be altering its configuration. If such were the case, then the experiment with phospholipase A could not be used to support the argument that phospholipids are part of the receptor complex. Although the results with phospholipase C are more supportive of the role of phospholipids in opiate binding, they do not exclude the possibility of indirect effects resulting from altered lipids.

 

In an effort to determine the extent of the phospholipid degradation resulting from exposure to phospholipase C and D, analyses were made of the lipid content of the membranes. With phospholipase D, which had no effect on opiate binding, no alteration in lipid content of the membranes was detectable even after prolonged exposure to relatively high concentrations of the enzyme. Although phospholipase C had markedly inhibited opiate binding, the alteration in the lipid composition was barely detectable, being less than 1-2%. When a combination of phospholipase A and C was used, the extent of hydrolysis was less than 5%. It appears that neural membranes are considerably more resistant to the combination of the two lipases than were erythrocyte membranes, where the hydrolysis was virtually complete (20). Despite the slight degree of lipid degradation by phospholipase C, the results may be particularly significant, since only a small fraction of the total membrane PS may be associated with the opiate binding site.

 

SUMMARY

 

A study was undertaken to investigate the possible role of phospholipids in opiate binding to brain membrane preparations, particularly with regard to phosphatidyl serine (PS). Upon the addition of lipids to suspensions of synaptic membranes it was found that PS and phosphatidic acid enhanced the stereospecific binding of 3H-dihydromorphine as much as 50%. Phosphoinositides were less effective, while other lipids were inactive. After partly inhibiting opiate binding by exposure of membranes to phospholipase C, it was still possible to enhance binding with PS, but not to restore it, whereas after treatment with phospholipase A + C, no enhancement occurred. Since the ethoxymethyl acetate ester of PS was as effective as PS in enhancing binding, it was concluded thdt the -COO- group was not directly involved in opiate binding; however, the -PO- group may be involved. Inhibition produced by the cationic detergent cetyltrimethylammonium bromide was completely reversible with PS and other acidic phospholipids, but not with the PS-ester. In the presence of 1 mM Ca2+ PS produced no enhancement. It is concluded that PS may be an essential component of the opiate receptor.

 

REFERENCES

 

1 Goldstein, A., Lowney, L.I. and Pal, B.K.: Stereospe-

cific and nonspecific interactions of the morphine congener levorphanol in subcellular fraction of rat

brain. Proc. Nat. Acad. Sci. 68:1742-1747 (1971).

2 Simon, E. J., Hiller, J.M. and Edelman, I.: Stereospe-

cific binding of the potent narcotic analgesic 3H-ethorphine to rat brain homogenate. Proc. Nat. Acad. Sci. 70:1947-1949 (1973).

3. Pert, C.B. and Snyder, S.H.: Properties of opiate-receptor binding in rat brain. Proc. Nat. Acad. Sci. 70:2243-2247 (1973).

4. Lowney, L.I., Schulz, K., Lowery, P.J. and Goldstein, A.: Partial purification of an opiate receptor from mouse brain. Science 183:749-753 (1974).

5. Abood, L.G. and Hoss, W.: Stereospecific morphine absorption to phosphatidyl serine and other membranous components of brain. Eur. J. Pharmacol. 32:66-75 (1975).

6. Wirtz, K.W.A. and Zilversmith, D.B.: Exchange of phospholipids between liver mitochondria and microsomes in vitro. J. Biol. Chem. 243:3596-3602 (1968).

7. Abood, L.G. and Takeda, F.: Enhancement of stereospecific opiate binding to neural membranes by phosphatidyl serine. Eur. J. Pharmacol., in press (1976).

8. Salem, N., Jr., Abood, L.G. and Hoss, W.: Separation of brain phosphatidyl serines according to degree of unsaturation by argentation thin layer chromatography. Submitted for publication, Anal. Biochem. (1976).

9. Pasternak, G.W. and Snyder, S.H.: Opiate receptor binding: effects of enzymatic treatments. Mot. Pharmacol. 10:183-193 (1974).

10. Gordesky, S.E., Marinetti, G.V. and Love, R.: The reaction of chemical probes with the erythrocyte membrane. J. Mem. Biol. 20:111-132 (1975).

11. Knauf, P.A. and Rothstein, A.: Chemical modification of membranes. J. Gen Physiol. 58:190-223 (1971).

12. Tanaka, R.: Comparison of lipid effects of e-Mg2+- activated p-nitrophenyl phosphatase and Na+-e-Mg2+- activated ATPase of membrane. J. Neurochem. 16:13011307 (1969).

13. Lloyd, T. and Kaufman, S.: The stimulation of partially purified bovine caudate tyrosine hydroxylase by phosphatidyl serine. Biochem. Biophys. Res. Comm. 59:1262-1269 (1974).

14. Bach, D., Rosenhiek, K. and Hiller, R.: Interaction of basic polypeptides with phospholipid vesicles: conformational studies. Eur. J. Biochem. 53:265-269 (1975).

15. Millett, F., Hargrave, P.A. and Raftery, M.A.: Natural abundance 13C-NMR spectra of the lipid in intact bovine retinal rod outer segment membranes. Biochem. 12:35913592 (1973).

16. Kakunaga, T., Kaneto, H. and Hano, K.: Pharmacological studies on analgesics--VII. Significance of the calcium ion in morphine analgesia. J. Pharmacol. Exp. Ther. 153:134-143 (1966).

17. Mule, S.J.: Morphine and the incorporation of 32orthophosphate in vivo into phospholipids of the guinea pig cerebral cortex, liver and subcellular fractions. Biochem. Pharmacol. 19:581-593 (1970).

18. Cardenas, H.L. and Ross, D.H.: Morphine-induced calcium depletion in discrete regions of rat brain. J. Neurochem. 24:487-493 (1975).

19. Yamamura, H.I. and Snyder, S.H.: Muscarinic cholinergic binding in rat brain. Proc. Nat. Acad. Sci. 71:17251729 (1974).

20. Verkleij, R.F.A., Zwaal, B., Roelofsen, B., Cimfurius, P., Kastelijn, D. and van Deenen, L.L.M.: The asymmetric distribution of phospholipids in the human red cell membrane. Biochim. Biophys. Acta 323:178-193 (1973).