Modification of ligand binding to membranes by a soluble acceptor

BIOL PSYCHIATRY 1986;21:883-888

883

Modification of Ligand Binding to Membranes by a Soluble Acceptor Alpha- 1-Acid Glycoprotein Attenuates 3H-Imipramine Binding to Cerebral Membranes Amiram I. Barkai, Miron Baron, Sharon Kowalik, and Thomas B. Cooper

The effect of individual plasma proteins on the binding of ‘H-imipramine {;‘H-IMI) was investigated, using rat cerebral membranes as the binding site source. Addition of alpha1 -acid glycoprotein (aAGLP) to an incubation medium containing 4 nM ‘H-IMI resulted in a concentration-dependent inhibition of ‘H-IMI binding, with an ICse value of 0.53 q. Albumin (0.7 mu) and gamma globulin (3 t.uu) had no apparent efSect. Scatchard analyses of ‘H-IMI binding in the presence and absence of 0.5 w aAGLP revealed that the inhibition of “H-IMI binding was associated with an increased I(d, with no appreciable change in B,,. The inhibitory action of platelet-free plasma (PFP) on ‘H-IMI binding to cerebral membranes was eliminated after treatment of PFP with specific antibodies to aAGLP. It is suggested that the inhibition of ‘H-IMI binding to cerebral membranes by PFP is due, at least partially, to the presence of aAGLP in the plasma. The observed inhibitory effect is consistent with a competition for the ligand between the membrane binding site and the soluble protein acceptor. These findings may explain the serum effect confounding radioreceptor assays of various drugs.

Introduction The high-affinity binding site for 3H-imipramine (3H-IMI) in brain and platelets (Briley et al. 1979; Raisman et al. 1979; Paul et al. 1980a; Rehavi et al. 1980) has been shown to be closely related to serotonin (5HT) uptake, and it possibly functions as a modulator of the 5-HT recognition site (Paul et al. 1981; Rehavi et al. 1981; Sette et al. 1981). Several investigators have suggested that the 3H-IMI binding site is acted upon by an endogenous ligand that has a regulatory role in the function of the 5-HT uptake system (Langer and Raisman 1983; Barbaccia et al. 1984; Paul et al. 1984). Thus, Langer et al. (1984) have indicated that 6-methoxy-tetrahydro+-carboline may be considered to be a

From the New York State Psychiatric Institute, Columbia University College of Physicians and Surgeons, Department of Psychiatry, New York, NY. Supported in part by NIMH Grants MH 33690 (A.B.) and MH 00176 (M.B.). Address reprint tqttests to Dr. Amiram I. Bark& New York State Psychiatric Institute. Columbia University College of physicians and Surgeons, Department of Psychiatry, 722 West 168th Sheet, New York, NY 10032. Received November 15. 1985; revised March 3, 1986.

0

1986

Society of Biological

Psychiatry

0006-3223/86/$03.50

884

A.1. Barkai et al.

BIOL PSYCHIATRY 1986;21:883-888

putative endogenous ligand of the ‘H-IMI binding site. Barbaccia et al. (1984) have confirmed that several tetrahydro-B-carbolines interact with the ‘H-IMI binding site, but in addition, they isolated from rat brain an endogenous effector, endocoid, which appeared to have a high-pressure liquid chromatography (HPLC) elution profile different from that of tetrahydro-B-carbolines. Paul et al. (1984) reported the presence of inhibitory components in blood that inhibit both 5-HT uptake and “H-IMI binding in synaptosomes and platelets. However. at least part of the observed inhibitory activity in their studies could be reduced by enzymatic proteolysis. Inhibition of ‘H-IMI binding to its high-affinity binding sites has been demonstrated after the addition of drug-free plasma to a radioreceptor assay for tricyclic antidepressants (Paul et al. 198Ob) or to a ‘H-IMI binding assay (Barkai et al. 1986). The latter study has demonstrated that the activity of the endogenous plasma substance (or substances) that inhibits ‘H-IMI binding to rat cerebral membranes did not differ in depressed patients and normal subjects. Furthermore, the inhibitory activity was found to be associated with plasma proteins. Preliminary observations indicated that the plasma component that inhibited ‘H-IMI binding in these experiments did not act by competition with ‘H-IMI for binding sites. The observed inhibition appeared to be a result of competition for ‘H-IMI between the cerebral membranes and an endogenous acceptor present in the plasma. The possibility that an endogenous soluble acceptor may inhibit “H-IMI binding to its membrane binding site is interesting, as such acceptors may function as modulators of ligand-receptor interactions. A modulatory mechanism of this kind may not be limited to the ‘H-IMI binding site and could possibly be operative in other ligand-receptor systems. The purpose of the present study was to investigate whether or not identifiable plasma proteins that have been shown by other investigators to bind imipramine can inhibit ‘l-1IMI binding to rat cerebral membranes. In addition, it was of interest to examine whether or not such inhibitory activity is consistent with that expected after the addition of a soluble IMI acceptor to the ‘H-IMI binding assay.

Methods Preparation

of Rut Bruin Membranes

Male Sprague-Dawley rats (200400 g) were decapitated and the brain quickly removed. The cerebellum and lower brain stem were dissected away on ice and the remaining brain tissue was homogenized in 50 volumes of ice-cold assay buffer (50 mM Trizma HCl containing 120 IIIM NaCl and 5 mM KCI. pH 7.5) using a polythron homogenizer (Brinkman Instruments, Inc., Westbury, NY; small probe; setting 10; 30 set). The homogenate was centrifuged at 43,000 x g for 10 min at 4°C and the supernatant discarded. The pellet was washed twice in 50 volumes assay buffer, and the final pellet was resuspended in assay buffer to obtain a concentration of 20 mg tissue/ml. The membranes were frozen at - 80°C until used.

Determination

of Protein Effect on ‘H-IMI Binding

Specific binding of “H-IMI was obtained as the difference between total and nonspecific binding at the ‘H-IMI concentration of 4 nM. Each assay for total binding was performed with 200 pg membrane protein and varying amounts of the studied plasma protein in a

Ligand Binding Modified by Soluble Acceptor

BIOL PSYCHIATRY 1986;21:883-888

885

final volume of 250 p.1. Membrane protein concentration was determined according to Lowry et al. (195 1). Tubes used to determine nonspecific binding contained desipramineHCl (DMI) at a final concentration of 100 FM. The tubes were incubated for a period of 60 min at 0°C (on ice) and then filtered rapidly (Whatman GF/F filters; Whatman Laboratory Products, Inc., Clifton, NJ) under vacuum. The filters were than washed three times with 5 ml cold assay buffer, dried under vacuum, and transferred to counting vials for measurement of radioactivity. Values representing specific binding in tubes containing the studied protein were expressed as a percentage of the mean value for specific binding obtained in the absence of the soluble protein. The following plasma proteins were used: albumin (bovine serum, RIA grade), gamma-globulin (human), and alpha- 1-acid glycoprotein (aAGLP, human). These proteins were purchased from Sigma Chemical Co., St. Louis, MO. The protein effect on 3H-IMI binding was evaluated following the addition of increasing amounts of the protein to the assay mixture. The protein concentration required to inhibit 50% of 3H-IMI binding at 4 nM (I&,) was used, when applicable, to describe inhibition of ‘HIMI binding by the protein. When the effect of the protein on the apparent density (B,,,) or dissociation constant (Kd) of 3H-IMI binding was studied, a constant amount of the protein (usually equal to its average amount in 10 ~1 plasma) was added to the incubation mixture containing 200 pg membrane protein, assay buffer, and 3H-IMI in final concentrations ranging from 0.5 to 8 nM. DMI (100 PM) was added to assay tubes used to determine nonspecific binding. B max and Kd were determined by Scatchard analysis. A binding assay with identical 3HIMI concentrations, but without the added protein, was performed in the same run to provide control values for B,,, and Kd.

Removal of Individual Proteins from Platelet-Free Plasma Platelet-free plasma (PFP) was obtained as described previously (Barkai et al. 1986). Briefly, human blood (20 ml) was collected in 100 pl of 10% EDTA. Platelet-rich plasma (PRP) was prepared by centrifugation (Baron et al. 1983). PFP was obtained from PRP after sedimentation of the platelets by centrifugation at 20,000 x g for 10 min at 4°C. Removal of aAGLP from the PFP was accomplished by addition of antiserum to aAGLP (goat antiserum; Atlantic Antibodies Inc., Scarborough, ME; Beckman number 1.2) and centrifugation at 43,000 X g for 1 hr at 4°C to precipitate the antigen-antibody complex. Albumin was removed by a similar procedure using antiserum to albumin (goat antiserum; Atlantic Antibodies Inc.; Beckman number 4.8).

Results Modijication of 3H-IMI Binding by Individual Plasma Proteins Alpha-l-Acid Glycoprotein. When different amounts (2-200 Fg) of aAGLP were added to the incubation mixture containing 4 nM of 3H-IMI in a final volume of 250 ~1, the specific binding of 3H-IMI decreased with increasing aAGLP concentrations. Plotting of bound 3H-IMI against the logarithm of the final concentration of aAGLP indicated an inhibition-type relationship, with an I&, value of 23 Fg/ml (Figure 1). Albumin. When the 3H-IMI assay contained 4 nM ‘H-IMI, 200 pg membrane protein, and varying concentrations of bovine serum albumin (16-480 mg/ml), there appeared to be no influence of the added albumin on the 3H-IMI specific binding (Figure 1).

886

A I. Barkai et al

BIOL PSYCHIATRY 1986;21:883-888

y

Albumm

Globulrn

--5&---w

uv Acid

Glycoprole~n

Figure I. Effect of plasma proteins on ‘H-IMI binding to rat cerebral membranes. lnhibition of ‘H-IMI binding is seen in the presence of aAGLP. A nonlinear regression analysis of the experimen-

points yielded an IC3,1value of 23 kg/ml or 0.53 FM for this protein. tal

Gamma-Globulin. Increasing concentrations of y-globulin in the assay mixture (2CrSOo Fg/ml) had no effect on the specific binding of ‘H-IMI at 4-nM ligand concentration (Figure 1). Protein EfSect on B,,, and Kd When specific 3H-IMI binding was assayed at different ligand concentrations in the presence and absence of the studied proteins and the results analyzed by Scatchard plots, it was found that both albumin and gamma-globulin had no effect on the apparent density (B,,,) or dissociation constant (Kd). Addition of aAGLP, however, resulted in a significant increase of K,, without apparent effect on B,,, (Table I).

Effect

qf Antibody-Treated

PFP

Treatment of PFP with specific antibodies to aAGLP resulted in a decrease of the PFP inhibitory activity of ‘H-IMI binding. Treatment with antibodies to albumin showed no significant change in the inhibition of ‘H-IMI binding to the cerebral membranes (Table 2).

Discussion The results presented here indicate that the previously observed inhibition of ‘H-IMI binding to cerebral membranes following addition of drug-free plasma was due, at least partially, to the presence of aAGLP in the plasma. The soluble aAGLP can act as an

Table I. Effect of Plasma Proteins on 3H-IMI Binding to Rat Cerebral Membranes’ Final

Protein added None (control) Albumin al-Acid glycoprotein y-Globulin

concentration (w/ml)

II

40 0.036 0.4

7 3 4 3

K,/

B”,,X (fmollmg protein) 283 252 265 311

2 -t + -t

32 28 40 52

(nM) 2.52 2.81 4.87 3.1

2 t 5 f

0.34 0.41 0.53h 0.42

“Protein solutions were prepared at average concentrations found in human plasma,and 10 &I of the solution was added to the’H-IMI binding assay. A volume of 10 ~1 plasma was found previously to significantly inhibit ‘H-IMI bmding to cerebral membranes (Barkai et al. 1986). Values for B,,, and Kd represent mean + SD. hSignilicantly different from control @ i 0.01)

Ligand Binding Modified by Soluble Acceptor

BIOL PSYCHIATRY 1986;21:883-888

887

Table 2. Effect of Treatment with Antisera to Individual Plasma Proteins on the Inhibition of ‘H-IMI Binding to Cerebral Membranes by PFP” Substance 3H-IMI

added to

binding assay

PFP 10 CL1 30 CL1 PFP treated with antiserum to aGLP 10 PI

30 ps PFP treated with antiserum to albumin 10 CL1 30 kl ‘Treatment with a )H-IMI ‘A control ‘p < 0.01

3H-IMI

binding

(% control)b 60 + I’ 35 I 5’

92 ? 6 83 5 8

62 2 6’ 43 2 8’

of PPP with antiserum to precipitate the individual protein is described in the text. Binding was determined concentration of 4 no. run was performed with each assay, and the data were analyzed by matched-pair t-test. compared with control.

acceptor that competes with the cerebral membranes for 3H-IMI binding and thus inhibits the ligand binding to the membranes. This partial inhibition appeared to be a result of an effect consistent with a coman increased Kdr with no appreciable change in B,,petition between a receptor and an acceptor for the ligand (Boeynaems and Dumont 1980). Although other plasma proteins, such as albumin or gamma-globulin, have the ability to bind IMI (Bickel 1975; Javaid et al. 1983), the binding affinity of aAGLP to this drug is very high compared with that of other proteins (Javaid et al. 1983). This higher affinity might explain the observed inhibitory effect of aAGLP, but not of albumin or gammaglobulin, under the present experimental conditions. A similar effect of aAGLP on specific binding in the neuroleptic radioreceptor assay (NRRA) has been described recently by Ko et al. (1985). These authors have suggested that the major inter- and intraindividual variation in the effect of drug-free plasma on spiroperidol binding in NRRA (Mailman et al. 1984) may be at least partially attributed to the presence of various aAGLP concentrations in different plasma samples. The observation that the inhibition by plasma of 3H-IMI binding is no longer apparent after addition of antiserum against aAGLP indicates that such antibodies may be useful in radioreceptor assays that may be confounded by varying concentrations of aAGLP. The demonstration that a soluble protein may act as an acceptor and thus affect 3HIMI binding to its membrane-bound receptor indicates that soluble acceptors could play an important role in the regulation of receptor-ligand interactions. The presence of a soluble acceptor might be expected to attenuate the binding of the ligand to its membranebound receptor by modifying the apparent binding affinity. Although there is currently no evidence for a physiological role of soluble acceptors in modulating receptor-neurotransmitter interactions, a search for possible effects of specific soluble proteins in such systems may be warranted.

The authors wish to thank Mr. Nicholas Rienzi for his technical

assistance.

888

BIOL PSYCHIATRY 1986:21:883-88X

A.I. Barkai et al.

References Barbaccia NL, Karum F, Costa E (1984): Tryptolines and imipramine recognition sites. Abstracts of the ACNP Annual Meeting,

December

10-14.

1984, p 8.

Barkai AI, Baron M, Kowalik S, Cooper TB (1986): Inhibition of ‘H-imipramine from depressed and normal subjects. Ps~chiurr?; Res (in press).

binding by plasma

Baron M, Barkai AI. Gruen R. Kowalik S, Quitkin F (1983): .‘H-imipramine platelet binding sites in unipolar depression. Biol Psychiuty 18: 1403-1409. Bickel MH (1975): Binding of chlorpromazine and imipramine to red cells. albumin lipoproteins and other blood components. J Pharm Pharmacol 271733-738. Boeynaems JM, Dumont JE ( 1980): Outlines ofReceptor Biomedical. pp 163-l 68.

Theory. New York: Elsevier/North

Holland

Briley MD, Raisman R. Langer SZ (1979): Human platelets possess high affinity binding sites for ‘H-imipramine. Eur J Pharmacol 58:347-348. Javaid IJ, Hendricks K. Davis JM ( 1983): Alpha-l -acid glycoprotein involvement in high affinity binding of tricyclic antidepressants to human plasma. Biochem Pharmacol 32: I 149-l 153. Ko GN, Korpi ER, Linnoila M ( 1985): On the clinical relevance and methods of quantification plasma concentrations of neuroleptics. J C/in Psyhopharmacol 5:253-262. Langer SZ. Raisman R ( 19X3): Binding of -‘H-imipramine and ‘H-desipramine for studies in depression. Nr,rcrophurma~olo~~ 22:407-1113.

as biochemical

ot

tool5

Langer SZ. Lee CR, Segonzac A. Tafeishi T, Esnaud H, Shoemaker H, Winblad B ( 1984): Possible endocrine role of the pineal gland for 6-methoxytetrahydro-B-carboline, a putative endogenous neuromodulator of the ‘H-imipramine recognition site. Eur J Pharmacol 102:379-380. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (I95 1): Use of the Folin phenol reagent in protein determination. J Bid Chem 193:265-275. Mailman RB, Dehaven DL, Halpern EA, Lewis MH (1984): Serum effects confound the neuroleptic radioreceptor assay. Life Sci 34: 1057-I 064. Paul SM, Rehavi M. Skolnick P, et al (1980a): Demonstration of specific high affinity binding sites for ‘H-imipramine on human platelets. Life Sci 26:953-959. Paul SM, Rchavi M. Hulihan B. Skolnick P. Goodwin FK (1980b): A rapid and sensitive radtorcceptor assay for tertiary amine tricyclic antidepressants. Commun Psyhopharmucol4:487494. Paul SM. Rehavi M. Skolnick P. Ballenger JC, Goodwin FK (1981): Depressed patients have decreased binding of tritiated imipramine to platelet serotonin “transporter.” Arch Gen Psychiatn 38:13lS-1317.

Paul SM. Angel I, Goldman ME. Skolnick P (1984): Characterization of endogenous modulators of serotonin uptake and ‘H-imiprdmine binding in rat and human plasma. Abstracts of the ACNP Annual Meeting. December 10-14, 1984. Raisman R, Briley M. Langer SZ (1979): Specific tricyclic antidepressant Nature (Land) 281:148-150.

binding sites in rat brain.

Rehavi M, Paul SM. Skolnick P. ct al (1980): Demonstration of specific high affinity binding sites for ‘H-imipraminc in human brain. Life .Gi 26:2273-2279. Rehavi M. lttah Y, Rice RC, et al (1981): 2-Nitro-imipramine. A selective irreversible inhibitor of ‘H-serotonin uptake and ‘H-imipramine binding in platelets. Biochem Biophvs Res Common 99:954-959.

SetteM, Raisman R, Briley M, Langer SZ (1981): Localization of tricyclic antidepressant sites on serotonin nerve terminals.

J Neurochem

37:4042.

binding