Proc. Nat. Acad. Sci. USA Vol. 68, No. 8, pp. 1742-1747, August 1971 Stereospecific and Nonspecific Interactions of the Morphine Congener Levorphanol in Subcellular Fractions of Mouse Brain AVRAM GOLDSTEIN, LOUISE I. LOWNEY, AND B. K. PAL Department of Pharmacology, Stanford University School of Medicine, Stanford, California 94305 Communicated by Marshall Nirenberg, May 21, 1971 ABSTRACT A method is described for analyzing the antagonists, and therefore presumably cannot enter the re- association of the opiate narcotic levorphanol with brain tissue into three components: nonsaturable, saturable ceptor sites. We take advantage of this by using radioactive nonspecific, and saturable stereospecific. The method may levorphanol [a synthetic D(-) congener of morphine] and its be of general applicability for the study of the interaction nonradioactive enantiomer, dextrorphan, in the following of drugs with body tissues. In mouse brain the stereo- way (Fig. 1). The association of ['H levorphanol (or [14C]- specific binding of levorphanol represents only 2% of the levorphanol) with a given tissue fraction is measured under total association of drug with tissue, and it was found only in certain membrane fractions. The material responsible three conditions: for the stereospecific binding might be the opiate receptor. A, ['H ]Levorphanol alone is present. It will participate in all the possible kinds of interaction. Pharmacologic action of a drug presupposes interaction of B, The system is first incubated with a 100-fold excess of drug molecules with tissue receptors. Progress has been made nonradioactive dextrorphan. Then [8H Ilevorphanol, at the in identifying and isolating some drug receptors (1, 2), espe- same concentration as in A, should be largely blocked from cially where specificity and affinity happened to be very high entering the nonspecific saturable sites. But its nonsaturable (3), or where a specific site-directed label could be attached associations ("trapped and dissolved") should not be hindered irreversibly (4, 5). For most drugs, however (including the because here each isomer behaves independently. The dif- opiate narcotics), direct measurement of localization and ference A minus B measures nonspecific saturable binding. binding in tissues is unlikely to yield useful information about C, The system is first incubated with a 100-fold excess of cellular or subcellular receptor sites, until two methodologic nonradioactive levorphanol. Then ['HIlevorphanol should be obstacles are understood and surmounted. blocked from entering all the saturable sites, but should still participate in the nonsaturable associations. The difference 1. Specificity B minus C measures stereospecific binding, which could be Many drugs interact nonspecifically with a wide variety of due to drug receptors. ['H ]Levorphanol that remains as- tissue components. Localization of drug in a particular organ, sociated with the tissue fraction in condition C measures non- subcellular particle, or macromolecule does not necessarily saturable associations. (The results here and in condition B imply a site of action there. are, of course, corrected for 1/101 of the ['H levorphanol We distinguish between nonsaturable and saturable inter- that will still occupy saturable sites.) actions. Nonsaturable interactions are of two kinds. First, particles surrounded by an osmotic membrane (e.g., synap- NONSPECIFIC STEREOSPECIFIC tosomes) or having a spongy matrix can contain trapped drug TRAPPED AND DISSOLVED BINDING BINDING in aqueous solution. Second, membranes will contain dis- solved drug in amounts determined by the lipid/water parti- tion coefficient and the ambient aqueous concentration. The I ] jJ mere finding that drug molecules are associated with some sub- cellular fractions cannot be interpreted as drug "binding". Nonspecific saturable interaction arises through ionic bonds, hydrogen bonds, and hydrophobic forces. Cationic drugs like the opiate narcotics can be expected to interact nonspecifically B with anionic groups of proteins, nucleic acids, phospholipids, A .I X i L sphingolipids, and mucopolysaccharides-interactions that are likely to be pharmacologically irrelevant. The problem is how to sort out the nonsaturable and nonspecific saturable C interactions in order to measure a relatively small amount of A specific saturable interaction, at the receptor sites, where drug i A i i the chain of events that leads to the charac- binding triggers FIG. 1. Displacement of levorphanol from various types of teristic pharmacologic effect. binding in tissue. Solid symbols represent radioactive levorphanol, The opiate narcotics display an extraordinary degree of open symbols indicate a 100-fold excess of nonradioactive dextror- stereospecificity. Whereas the D(-) compounds are pharma- phan [the inert L(+) isomer] in B or the same excess of non- cologically active, the L(+) isomers are neither agonists nor radioactive levorphanol in C. 1742 Downloaded by guest on September 26, 2021 Proc. Nat. Acad. Sci. USA 68 (1971) Interaction of Levorphanol with Brain Tissue 1743 2. Reversibility with Hyamine hydroxide (Packard) or ethanol, as appropriate Most pharmacologic effects are readily reversible, whence it is for solubilization. Counting efficiencies were determined with supposed that the corresponding drug-receptor interactions 'H20 or [14C]benzoate as internal standard. Protein was de- are also reversible. Dilution of a reversible system promotes termined by the method of Lowry et al. (12). ligand dissociation, and this may occur to a significant extent Major subfractions during manipulation of tissue fractions prior to analysis. All procedures were carried out at 2-5oC. The homogenate The problem is analogous to the "dilution effect" whereby an was centrifuged at 1000 X g for 10 min. The sediment was re- enzyme inhibited reversibly in vivo recovers activity if diluted prior to assay (6). suspended twice in sucrose-Tris with a Vortex mixer and centrifuged as before, to obtain the crude nuclear fraction. Tissues are usually homogenized in several volumes of The supernatant fluids and washes from the crude nuclear aqueous medium; but even if no water is added (7), the mixing fraction were combined and centrifuged for 20 min at 12,000 of intracellular and interstitial fluids might cause dilution X g, and the pellet was washed once as above, to obtain artifacts. Fractionation methods also entail dilution. In the crude mitochondrial fraction. The microsomal and density-gradient centrifugation, as in simple sedimentation, soluble fractions were obtained by combining the supernatant drug molecules will dissociate from particles as they move and wash from the crude mitochondrial fraction and centri- down the tube, to be collected with components that band fuging for 1 hr at 105,000 X g. at higher levels, which will lead to spurious results. In molec- ular-sieving techniques, progressive dilution occurs down the Subfractionation of nuclear fraction length of the column. Acid precipitation of soluble macro- To isolate nuclei we used the method of Lovtrup-Rein and molecules releases bound ligands to an unpredictable extent. McEwen (13), without detergent. The nuclear fraction was Seeing no meaningful way to assess the stereospecific resuspended in 2 M sucrose (10 ml/g brain tissue) and cen- binding of a drug in vivo, we decided to study the capacity of trifuged for 30 min at 93,000 X g. The supernatant fluid, in- tissue fractions to bind in vitro. The essential precaution was to cluding a parchment-like floating layer, was removed, mixed prevent dissociation of reversibly-bound drug by maintaining vigorously, and centrifuged again. The floating material ambient drug concentrations constant throughout all proce- ("floated membranes") was collected and resuspended in dures. Particulate material was fractionated (or simply 0.32 M sucrose-Tris. The pink jelly-like pellets containing sedimented) centrifugally in the presence of the appropriate whole nuclei were pooled and resuspended in 0.32 M sucrose- radioactive and nonradioactive drugs (conditions A, B, and Tris. In another procedure we obtained membranes by sub- C), and the excess radioactivity associated with the bands and jecting 1 or 2 ml of the crude nuclear suspension to three 20-sec pellets was measured. For soluble and some particulate frac- bursts in the MSE-Mullard ultrasonic disintegrator (19-mm tions, molecular sieving columns were used. Here the method probe, 9.5-mm tip, 00C) in a 5 X 1.2 cm cellulose nitrate of Hummel and Dreyer (8) suited our requirements perfectly, tube. No whole nuclei were visible microscopically after this the column being first equilibrated with the appropriate drugs. treatment. The nuclear membranes (including nucleoli) were Equilibrium dialysis was suitable for measuring the binding diluted in 0.32 M sucrose-Tris and sedimented for 1 hr at capacity of soluble fractions. 105,000 X g. METHODS Subfractionation of mitochondrial fraction Male Swiss-Webster mice (25-30 g) were used. Usually, 3-6 Nerve-ending fractions were isolated from the mitochondrial brains were pooled. After decapitation, whole brains or por- fraction by discontinuous gradient centrifugation (14). The tions thereof were weighed quickly and homogenized in 10 mitochondria were resuspended in 0.32 M sucrose-Tris (1 ml volumes of cold 0.32 M sucrose in 0.01 M Tris 1HCl, pH 7.0, per brain) containing 10 gM CaC12, and layered on a dis- with a loose-fitting motor-driven pestle, exactly as described continuous sucrose gradient