- Email: [email protected]
Antidepressive drugs can change the affinity of [3H]imipramine and [3H]paroxetine binding to platelet and neuronal membranes
European Journal of Pharmacology, 119 (1985) 1-8 Elsevier
1
A N T I D E P R E S S I V E D R U G S CAN C H A N G E T H E AFFINITY OF 1 3 H I l M I P R A M I N E AND [ 3 H ] P A R O X E T I N E B I N D I N G T O P L A T E L E T AND N E U R O N A L M E M B R A N E S PER PLENGE * and ERLING T. MELLERUP Psychochemistry Institute, Rigshospitalet, 9 Blegdamsvej, DK-2100 Copenhagen, Denmark Received 9 May 1985, revised MS received 11 September 1985, accepted 17 September 1985
P. PLENGE and E.T. MELLERUP, Antidepressioe drugs can change the affinity of [ 3H]imipramine and [ ~H]paroxetine binding to platelet and neuronal membranes, European J. Pharmacoi. 119 (1985) 1-8. Serotonin transport in synapses and platelets is inhibited by tricyclic antidepressants as well as by more selective transport inhibitors. This inhibition is hypothesized to be of importance for the psychotropic effect, although it is known that some new antidepressants do not possess this transport inhibitory action. We now report that antidepressive drugs can influence the serotonin transport complex in platelets and brain in other ways: [3H]imipramine and [ 3H]paroxetine, which bind with high affinity to the serotonin transport complex, can be dissociated from the complex with velocity constants strongly influenced by the different antidepressants. This effect is not correlated to the inhibitory action of the drugs on serotonin transport. Furthermore the effect is seen in the micromolar range in contrast to the high affinity binding process which takes place in the pico- and nanomolar range. The effects of antidepressants on the dissociation rates of bound ligand make it possible to differentiate between serotonin reuptake inhibitors which appear identical in other assays. Antidepressive drugs can thus be divided into groups which differ from the usual classifications. Serotonin transport complex Antidepressive drugs
[3H]Imipramine Binding
[3H]Paroxetine
I. Introduction Antidepressants are drugs which have an elevating effect on the depressed mood in many manicdepressive patients. However, it is not known how such an effect is brought about. It is known that antidepressants have a number of biological effects, and it is then assumed that one or more of these are related to the psychotropic effect whereas others are assumed to cause side-effects. A wellknown action of tricyclic antidepressants is to inhibit the reuptake of serotonin and noradrenaline across the presynaptic neuronal membranes. This inhibition is often assumed to play a central role for psychotropic effect. Another characteristic which may be related to the inhibition of serotonin uptake is the specific * To whom all correspondence should be addressed. 0014-2999/85/$03.30 © 1985 Elsevier Science Publishers B.V.
Platelets
Brain
high affinity binding of imipramine to binding sites on serotonergic neurons or blood platelets (Langer et al., 1980; Paul et al., 1980). This high affinity binding is specific for imipramine, other tricyclic antidepressants, and several non-tricyclic serotonin uptake inhibitors, whereas various atypical antidepressants have very low affinity for the so-called imipramine binding site (Raisman et al., 1980). A third effect of a number of antidepressants, described in the present work, is a change in the affinity of [3H]imipramine or [3H]paroxetine for their respective binding sites on the serotonin transport complex. The affinity of a radioactive ligand for its binding sites can be expressed by the velocity constant of the dissociation of the ligandbinding site complex, or the half-life of this complex. The half-life can be determined very accurately by following the dissociation of bound ra-
dioactive ligand, after the concentration of nonbound radioactive ligand is reduced. This can be obtained either by extensive dilution (e.g. 50 × ) of the sample or by addition of high amounts of unlabelled ligand. The two methods should theoretically give the same result, unless the high amount of unlabelled ligand has an additional and separate effect on the structure and function of the binding protein, leading to a change in the affinity between ligand and binding site. This seems to be the case for [3H]imipramine binding to platelet membrane binding sites, as Phillips and Williams (1984) have shown that the dissociation of the [3H]imipramine binding sites complex after dilution was influenced by the presence of imipramine. Similarly, Wennogle and Meyerson (1983) have shown that the dissociation of the [3H]imipramine binding site complex was inhibited in the presence of serotonin. In the present work the rates of dissociation of [3H]imipamine or [3H]paroxetine from their respective binding sites on neuronal membranes and platelets (Mellerup et al., 1983; Mellerup et al., 1985) were used as indicators of structural or functional changes in these binding proteins. The effects of several types of antidepressants on the dissociation rates were studied, and the results indicate that there is here a 'new' kind of effect. These effects would divide the known antidepressants into new subgroups quite different from the usual classification.
took place in 8 ml buffer (5 mM Tris, H C I / 5 mM EDTA, pH 7.5) using an Ultra-Turrax homogenizer (5 s high speed). Platelet membranes were washed twice with 8 ml buffer (70 mM TRIS, HCI, pH 7.5) (centrifugation 30000 × g, 10 rain) and finally resuspended in buffer (50 mM Tris, HCI/120 mM NaC1/5 mM KC1, pH 7.5) to a concentration between 0.5 and 1.0 mg protein/ml. The samples were kept at - 8 0 ° C until analysis. 2.2. Rat brain membranes
Rats were decapitated, the brains removed and placed on ice. Whole brain except the brain-stem and cerebellum was homogenized (Ultra-Turrax, 30 s high speed) in 10 ml ice-cold buffer (50 mM Tris, HC1/120 mM NaC1/5 mM KC1, pH 7.5). The homogenate was centrifuged at 30 000 × g for 10 min, and the pellet resuspended in the same buffer and centrifuged twice. After the third resuspension, buffer was added to give a protein concentration of about 4 mg membrane protein/ml (corresponding to about 100 mg brain tissue/ml). The membranes were stored at - 8 0 ° C . 2. 3. Protein analysis
The protein concentrations in the membrane preparations were determined with the Lowry method modified for membrane proteins according to Peterson (1977). 2.4. Drugs
2. M a t e r i a l s a n d m e t h o d s
2.1. Platelet membranes
Platelets were obtained from 500 ml blood, sodium citrate was used as an anticoagulant. Red blood cells and leucocytes were removed by lowspeed centrifugation (200 x g, 10 min, room temperature). All the following steps were performed at 0-4°C. Platelets were precipitated from the platelet-rich plasma by centrifugation at 30000 × g, 10 rain. The platelets were washed twice with 8 ml buffer (50 mM Tris, HCI/150 mM NaC1/20 mM EDTA, pH 7.5), and frozen as a dry pellet at - 80°C. Lysis and homogenization of the platelets
[3H]Imipramine 20 C i / m m o l - ( A - m e r s h a m , England) and [3H]paroxetine 19.6 C i / m m o l (NEN, USA). The drugs used in the dissociation experiments were generously provided by Astra, Ciba Geigy, Ferrosan, Hoechst, Lilly, Lundbeck, Organon, Pfizer, Rhone-Poulenc, Roche and Upjohn. 2.5. Dissociation rate determinations
Membrane suspension, 30-50 ml, (platelets about 0.5 mg protein/ml, rat brain about 1 mg protein/ml) were incubated with [3H]imipramine at 0°C or [3H]paroxetine at 20°C for 2 h. The
[3H]imipramine concentrations were 2 nM in platelet assays and 6 nM in the rat brain assay. The [3H]paroxetine concentration was 1 nM in both the platelet and the rat brain assays. Aliquots (200 /~1) were diluted into 10 ml buffer (50 x ) containing the various drugs. Experiments with [3H]imipramine were carried out at 0°C whereas experiments with [3H]paroxetine were carried out at 20°C. The samples were filtered 2 min to 4 h after the 50 × dilution step, through Whatman G F / F glass-fiber filters and washed with 3 x 5 ml 0°C buffer. All determinations were done in triplicate, and the median values were used to determine the dissociation rates. In each assay the dissociation rate in the presence of imipramine was used as a control. The mean half-life of the [3H]imipramine platelet membrane binding site complex in the presence of imipramine was 60 _+ 3.4 min (S.E.M.). The effects on the dissociation rates of the other drugs were mostly determined in single experiments, using 12-15 different time intervals between dilution and filtration. To determine the non-specific binding, tubes from each dissociation rate determination were heated to 37°C for 3 h (ensuring that all specifically bound ligand had dissociated from the binding site). The samples were then returned to 0°C or 20°C respectively for 20 min before filtration. The filters were dried overnight and counted in Pico-Flour 15 ®.
3. Results
Figure 1 shows the effect of dilution with 200 /~M citalopram, paroxetine, imipramine, or a 50 x dilution with buffer alone on the dissociation rate of the [3H]imipramine platelet membrane binding site complex. It is seen that all three drugs had an effect on the dissociation of the [3H]imipramine binding site complex. Citalopram decreased, and paroxetine increased the dissociation rate and the dilution-induced dissociation of the slow component was abolished by imipramine. Figure 2 shows a dose-response experiment where the dissociation of the [3H]imipramine platelet binding site complex was induced by dilution with various concentrations (between 10 and
CPM 2°°ct 3H-IMIPRAMINE PLATELETS
IOOC
io¢
z0
I
[
1
I
I
2
3
4
HOURS
>
Fig. 1. Dissociation rates of [ 3H]imiprarnine platelet membrane binding site complexes: O 200 /~M citalopram, • 200 ~tM imipramine, zx 200 ~tM paroxetine or X after a 50 x dilution with buffer alone. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after the preincubated membranes had been diluted 50 x with buffer containing the various drugs under investigation.
300 /~M) of paroxetine or citalopram. It is seen that both the paroxetine and the citalopram effects were dose-dependent, and that half maximal effects were obtained with about 80 /~M paroxetine or 15 ~tM citalopram. Serotonin (data not shown), which decreased the dissociation rate, was found to have a half maximal effect at about 120/~M in similar experiments. Figure 3 shows the dissociation rate of the [3 H]imipramine platelet binding site complex when drugs with different effects on the dissociation rate were mixed. Citalopram alone, or mixtures of citalopram and imipramine or citalopram and paroxetine gave essentially the same results namely a very low dissociation rate. This latter effect was also seen when dissociation first took place during 30 rain in 400/~M paroxetine and an equal volume of 400/xM citalopram was then added at this time
CPM 3H-IMIPRAMINE
CPM
PLATELETS
3H- I M I P R A M I N E iq60120 80 •
40
-
eo
PLATELETS
2OOC /u.M
tOOC -
500
io CITALOPRAM
\
\ %
200
\
% . ~ I M I P R A M I NE
PAROXETINE I00
IMI P R A M I N E * C I T A L O P R A M n 50
~M 2C eo
iO I
1 I
I 2
1 3
I 4
)
HOURS
Fig. 2. Dissociation rates of [3 H]imipramine platelet membrane binding site complexes in the presence of 10-300 # M citalopram (the upper part of the figure), or in the presence of 10-30 ~M paroxetine (the lower part of the figure). The concentrations of paroxetine or citalopram corresponding to the individual curves are given to the right of the individual curves. The dotted line represents the dissociation rate in the presence of 10 #M imipramine. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after the preincubated membranes had been diluted 50 × with buffer containing the various drugs under investigation.
(at the arrow) bringing the concentrations of both paroxetine and citalopram to 200 /~M. It is seen that in all cases the citalopram effect dominated and abolished the imipramine and the paroxetine effects. It is also seen that the imipramine effect on the dissociation rate was stronger than the paroxetine effect as a mixture of imipramine and paroxetine produced the same results as imipramine alone. The experiments illustrated in fig. 3 therefore show that if mixtures of drugs are used to induce dissociation the effect of the drug leading to the slowest dissociation rate will dominate. Table 1 shows the effects of various antidepres-
IMIPRAMINE PAROXET/N E ÷ IMIPRAM[NE
=
P A R O X E T I N E ~. C I T A L O P R A M
x
PAROXETINE
#.
I 2 HOURS
t3
I 3
1 4
~.
Fig. 3. Dissociation rates of [ 3H]imipramine platelet membrane binding site complexes in the presence of various combinations of drugs. O 200 /zM citalopram, [] 200 #M citalopram+ 200 t~M imipramine, • 200/~M imipramine, • 200 #M imipramine +200 /~M paroxetine, × 400 #M paroxetine, which was diluted at the arrow with an equal volume of 400 #M citalopram, thus bringing both concentrations to 200 #M, and zx 200 #M paroxetine. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after start of the experiments by a 50 × dilution with buffer containing the various drugs under investigation.
sive drugs and of a few other psychotropic drugs on the dissociation rate of the [3H]imipramine platelet membrane binding site complex. It is seen that the antidepressive drugs could be divided into three main groups. A large group with the same effects as imipramine, a smaller group of drugs which decreased the dissociation rate, and paroxetine which increased the dissociation rate. It is noteworthy that the two closely related serotonin reuptake inhibitors paroxetine and femoxetine had opposite effects on the dissociation rate. Figure 4 shows the effect of 200 # M citalopram, 200 /~M paroxetine, 10 /~M imipramine of 50 × dilution with buffer alone on the dissociation rate of the [3H]imipramine rat brain membrane bind-
CPM
TABLE l
3H-IMIPRAMINE
[3H]Imipramine dissociation rates. Dissociation rate (half-life) of the [3H]imipramine platelet or brain membrane binding site complex, in the presence of various drugs. Platelet membrane suspension (200 #1) was preincubated with 2 nM [3H]imipramine (platelet) or 6 nM [3H]imipramine (brain), and diluted 50 xinto 10 ml buffer, containing the drugs in the concentrations described.
BRAIN
200(
1000
Half-life (rain)
200 ttM 200 #M 200/~M 200/aM 200 ttM
citalopram serotonin sertraline mianserin femoxetine
200/~M 200 #M 200/~M 200 #M 200/tM 200 FM
imipramine chlorimipramine fluoxetine indalpine iproniazid zimilidine
200/tM paroxetine 50 x dilution 200/zM cocain 200/~M tryptophane 200/~M alprazolam 200/tM carbamazepin 200 # M chlorpromazine 200/tM nomifensine
Platelet membranes
Brain membranes
600 240 135 125 70
105 40 25 10
50(3
200
tog
12 .
Range 50-60 min.
32 Two component curve. Range 50-60 min for the fast component.
50
25 16 12
ing site complex. The effects differed from the effects on the platelet binding site as both paroxetine and citalopram reduced the dissociation rate of the brain membrane binding site ligand complex, compared with imipramine or 50 x dilution. The dissociation induced by dilution followed a monoexponential curve and was a little slower than the dissociation induced by 10 /~M imipramine. Table 1 shows the effects of a few drugs on the dissociation rate of the [3H]imipramine rat brain membrane binding site complex. Mianserin reduced the dissociation rate as it did in platelets. The difference between the effects of the closely related drugs paroxetine and femoxetine is also seen. The dissociation rates of the [3H]paroxetine binding site complex from both platelets and rat
20
CITALOPRAM
o
o
PAROXETINE
~,
~"
50 x DILUTION
x
M PRAM NE
•
I 2
I 3
I 4
)
HOURS
Fig. 4. Dissociation rates of [3H]imipramine rat brain membrane binding site complexes: © 200 /zM citaloprarn, • 200 /LM imipramine, zx 200 #M paroxetine or X after a 50 × dilution with buffer alone. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after start of the experiments by a 50 x dilution with buffer containing the various drugs under investigation.
brain membranes were affected by the various drugs. The effects were qualitatively the same in platelets and brain as both paroxetine and citalopram reduced the dissociation rates compared with their values measured after dilution with 10 # M imipramine or buffer alone. However, the effects were much stronger on brain membranes (fig. 5) than on platelet membranes (not shown) as both citalopram and paroxetine stabilized the binding site ligand complex in brain membranes to an extent where no dissociation took place during the 4 h the experiment lasted. This is also seen in table 2 showing the effects on the dissociation rates of v a r i o u s antidepressive drugs in both platelet and rat brain membranes.
CPM
TABLE 2
3H-PAROX ETI NE BRAIN
[3H]Paroxetine dissociation rates. Dissociation rate (half-life) of the [3H]paroxetine platelet or brain membrane binding site complex, in the presence of various drugs. Platelet membrane suspension (200 ffl) was preincubated with 1 nM [3H]paroxetine, and diluted 50 × into 10 ml buffer, containing the drugs in the concentrations described.
200C
tO00
-----~=-"
-
^
5
"-
~
~
a
r,
-
Half-life (min) 500
20C
IOC
5C
1
CITALOPRAM
o
o/
PAROXETINE
A
A
50 x DILUTION
x--x
IMIPRAMINE
=--o
2c
[ I
J
[ 2 HOURS
[ 3
1 4
>
Fig. 5. Dissociation rates of [3H]paroxetine rat brain membrane binding site complexes in the presence of: O 200 ffM citalopram, • 10 /~M imipramine, zx 200 ffM paroxetine or x after a 50 x dilution with buffer alone. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after start of the experiments by a 50 x dilution with buffer containing the various drugs under investigation.
4. Discussion High affinity binding of [3H]imipramine and [3H]paroxetine takes place in platelet and neuronal membranes (Langer et al., 1980; Paul et al., 1980; Mellerup et al., 1983). The binding shows saturability in the low n a n o m o l a r range, which is 10-100 times below the plasma concentration range for therapeutic levels of antidepressive drugs. The binding of [3H]imipramine and its affinity for the binding site m a y depend on the structure of the binding site itself, which in turn may depend on the structure of the protein matrix. M a n y factors influence the protein molecules, e.g. ions, hormones and drugs. Drugs can affect the affinity between [3H]imipramine and the binding sites in two ways, either in a competitive way by interaction at the binding site, or in a non-competitive way by effects on the protein structure and function. In equilibrium conditions the presence of
Platelet membranes
Brain membranes
200/LM citalopram 200 #M paroxetine 200/LM serotonin 200/.t M mianserin 200 ~ M femoxetine 200 ~M irnipramine 200/xM fluoxetine
410 240 220 220 130 100 70
210 oz 540 260 260
50 X dilution 200 ffM alprazolam 10/~M imipramine 200 ~tM indalpine 200/.t M carbamazepin 200/~M zimilidin
Range 60-70 rain.
Range 120-130min.
n a n o m o l a r concentrations of various antidepressive drugs will decrease the affinities of [3H]imipramine and [3H]paroxetine and their respective binding sites because of competitive binding. Competitive reactions will not take place in experiments where the dissociation rate is measured, and the membranes with b o u n d [3H]imipramine or [3H]paroxetine are strongly diluted. A n y change in affinity revealed in such experiments is due to non-competitive interactions. There are reports of similar experiments in which serotonin (Wennogle and Meyerson, 1983), imipramine (Phillips and Williams, 1984) and recently tryptamine (Segonzac et al., 1985) changed the affinity of [3H]imipramine for its binding site non-competitively. The present results show that a large n u m b e r of antidepressants also possess this ability to affect the affinity of the binding protein, with a remarkable drug specificity of this effect. Thus, citalopram strongly increased the affinity of [3H]imipramine and [3H]paroxetine for their respective binding sites in both platelet and neuronal membranes. Paroxetine, as opposed to citalopram decreased the affinity of [3H]imipramine for the platelet binding site to far below the normal value mea-
sured in the absence of unlabeled drug. However, both paroxetine and citalopram increased the affinity between [3H]paroxetine and the platelet binding site. Thus the effect of paroxetine on [3H]imipramine and [3H]paroxetine binding to platelet membranes added further evidence for the earlier finding (Plenge and Mellerup, 1984; Mellerup et al., 1985) that these two drugs have separate binding sites on a common protein complex, probably the serotonin transport mechanism, in the platelet membrane. Paroxetine, like citalopram, strongly increased the affinity between [3H]paroxetine and its binding site in neuronal membranes whereas the affinity for [3H]imipramine was only moderately increased. Thus paroxetine increased the affinity for [3H]paroxetine in both platelet and neuronal membranes, whereas the paroxetine effect on [3 H]imipramine binding was opposite in these two membranes. These results are in agreement with results which indicate that the [3H]paroxetine binding polymer may be the same in platelets and neurons, whereas the [3H]imipramine binding polymers probably are different in these tissues (Mellerup and Plenge, unpublished results). The citalopram-induced increase in affinity between [3H]imipramine and the platelet binding site was dose-dependent, and was seen clearly with 10/~M. Similarly, the decrease in affinity between [3 H]imipramine and the platelet binding site induced by paroxetine was also dose-dependent (fig. 2). The non-competitive action of paroxetine on the structure or action of the [3H]imipramine binding polymer leading to a decrease in affinity for [3H]imipramine, was promptly reversed by the addition of citalopram (fig. 3). Thus, the effect of citalopram could take place in the presence of both paroxetine or imipramine whereas the opposite was not found. Imipramine itself, compared to dilution without addition of unlabelled drug, caused little alteration in the affinity for [3H]imipramine and [3H]paroxetine in platelets or in neuronal membranes. However imipramine changed the biphasic course of the dissociation curve for [ 3H]imipramine bound to platelet membranes to a monophasic course in agreement with the results of Phillips and Williams (1984). Several other drugs were analysed in the dis-
sociation experiments, and it was generally found that drugs devoid of antidepressive activity did not affect the dissociation of [3H]imipramine from platelet membranes. Antidepressive drugs, on the other hand affected the dissociation rate, by either increasing or decreasing the dissociation rate, or by changing the biphasic course to a monophasic one. However, with the relatively small number of drugs tested, it was impossible to evaluate whether these kinetic effects could be of clinical relevance. It cannot be excluded that other drugs or chemical compounds not related to psychopharmacology may influence the [3H]imipramine and [3H]paroxetine binding polymers in a similar way as this possibility has not been tested thoroughly.
Acknowledgements This study was supported by the Danish Medical Research Council and P. Carl Petersens Fund and Dr. Med. Vet. Axel Thomsen and Hustru Martha Thomsens Fund.
References Langer, S.Z., C. Moret, R. Raisman, M.C. Dubocovich and M. Briley, 1980, High-affinity [3H]imipramine binding in rat hypothalamus: association with uptake of serotonin but not of norepinephrine, Science 210, 133. Mellerup, E.T., P. Plenge and M. Engelstoft, 1983, High affinity binding of [3H]paroxetine and [3H]imipramine to hu. man platelet membranes, European J. Pharmacol. 96, 303. Mellerup, E.T., P. Plenge and M. Nielsen, 1985, Size determination of binding polymers for [3H]imipramine and [3H]paroxetine in human platelet membranes, European J. Pharmacol. 106, 411. Paul, S., M.M. Rehavi, P. Skolnick and F.K. Goodwin, 1980, Demonstration of specific 'high-affinity' binding sites for [3H]imipramine of human platelets, Life Sci. 26, 953. Peterson, G.L., 1977, A simplification of the protein assay method of Lowry et al. which is more generally applicable, Anal. Biochem. 83, 346. Phillips, O.M. and C. Williams, 1984, Multiple affinity states for [3H]imipramine binding sites on human platelet membranes, European J. Pharmacol. 100, 247. Plenge, P. and E.T. Mellerup, 1984, Temperature dependence of the binding of [3H]labeled imipramine and [3H]labeled paroxetine to human platelet membrane, Biochim. Biophys. Acta 770, 22.
Raisman, R., M.S. Briley and S.Z. Langer, 1980, Specific tricyclic antidepressant binding sites in rat brain characterized by high-affinity [3H]imipramine binding, European J. Pharmacol. 61, 373. Segonzac, A., R. Raisman, T. Tateishi, H. Schoemaker, P.E. Hicks and S.Z. Langer, 1985, Tryptamine, a substrate for
the serotonin transporter in human platelets, modifies the dissociation kinetics of [3H]imipramine binding: possible allosteric interaction, J. Neurochem. 44, 349. Wennogle, L.P. and L.R. Meyerson, 1983, Serotonin modulates the dissociation of [3H]imipramine from human platelet recognition site, Euroepan J. Pharmacol. 86, 303.
1
A N T I D E P R E S S I V E D R U G S CAN C H A N G E T H E AFFINITY OF 1 3 H I l M I P R A M I N E AND [ 3 H ] P A R O X E T I N E B I N D I N G T O P L A T E L E T AND N E U R O N A L M E M B R A N E S PER PLENGE * and ERLING T. MELLERUP Psychochemistry Institute, Rigshospitalet, 9 Blegdamsvej, DK-2100 Copenhagen, Denmark Received 9 May 1985, revised MS received 11 September 1985, accepted 17 September 1985
P. PLENGE and E.T. MELLERUP, Antidepressioe drugs can change the affinity of [ 3H]imipramine and [ ~H]paroxetine binding to platelet and neuronal membranes, European J. Pharmacoi. 119 (1985) 1-8. Serotonin transport in synapses and platelets is inhibited by tricyclic antidepressants as well as by more selective transport inhibitors. This inhibition is hypothesized to be of importance for the psychotropic effect, although it is known that some new antidepressants do not possess this transport inhibitory action. We now report that antidepressive drugs can influence the serotonin transport complex in platelets and brain in other ways: [3H]imipramine and [ 3H]paroxetine, which bind with high affinity to the serotonin transport complex, can be dissociated from the complex with velocity constants strongly influenced by the different antidepressants. This effect is not correlated to the inhibitory action of the drugs on serotonin transport. Furthermore the effect is seen in the micromolar range in contrast to the high affinity binding process which takes place in the pico- and nanomolar range. The effects of antidepressants on the dissociation rates of bound ligand make it possible to differentiate between serotonin reuptake inhibitors which appear identical in other assays. Antidepressive drugs can thus be divided into groups which differ from the usual classifications. Serotonin transport complex Antidepressive drugs
[3H]Imipramine Binding
[3H]Paroxetine
I. Introduction Antidepressants are drugs which have an elevating effect on the depressed mood in many manicdepressive patients. However, it is not known how such an effect is brought about. It is known that antidepressants have a number of biological effects, and it is then assumed that one or more of these are related to the psychotropic effect whereas others are assumed to cause side-effects. A wellknown action of tricyclic antidepressants is to inhibit the reuptake of serotonin and noradrenaline across the presynaptic neuronal membranes. This inhibition is often assumed to play a central role for psychotropic effect. Another characteristic which may be related to the inhibition of serotonin uptake is the specific * To whom all correspondence should be addressed. 0014-2999/85/$03.30 © 1985 Elsevier Science Publishers B.V.
Platelets
Brain
high affinity binding of imipramine to binding sites on serotonergic neurons or blood platelets (Langer et al., 1980; Paul et al., 1980). This high affinity binding is specific for imipramine, other tricyclic antidepressants, and several non-tricyclic serotonin uptake inhibitors, whereas various atypical antidepressants have very low affinity for the so-called imipramine binding site (Raisman et al., 1980). A third effect of a number of antidepressants, described in the present work, is a change in the affinity of [3H]imipramine or [3H]paroxetine for their respective binding sites on the serotonin transport complex. The affinity of a radioactive ligand for its binding sites can be expressed by the velocity constant of the dissociation of the ligandbinding site complex, or the half-life of this complex. The half-life can be determined very accurately by following the dissociation of bound ra-
dioactive ligand, after the concentration of nonbound radioactive ligand is reduced. This can be obtained either by extensive dilution (e.g. 50 × ) of the sample or by addition of high amounts of unlabelled ligand. The two methods should theoretically give the same result, unless the high amount of unlabelled ligand has an additional and separate effect on the structure and function of the binding protein, leading to a change in the affinity between ligand and binding site. This seems to be the case for [3H]imipramine binding to platelet membrane binding sites, as Phillips and Williams (1984) have shown that the dissociation of the [3H]imipramine binding sites complex after dilution was influenced by the presence of imipramine. Similarly, Wennogle and Meyerson (1983) have shown that the dissociation of the [3H]imipramine binding site complex was inhibited in the presence of serotonin. In the present work the rates of dissociation of [3H]imipamine or [3H]paroxetine from their respective binding sites on neuronal membranes and platelets (Mellerup et al., 1983; Mellerup et al., 1985) were used as indicators of structural or functional changes in these binding proteins. The effects of several types of antidepressants on the dissociation rates were studied, and the results indicate that there is here a 'new' kind of effect. These effects would divide the known antidepressants into new subgroups quite different from the usual classification.
took place in 8 ml buffer (5 mM Tris, H C I / 5 mM EDTA, pH 7.5) using an Ultra-Turrax homogenizer (5 s high speed). Platelet membranes were washed twice with 8 ml buffer (70 mM TRIS, HCI, pH 7.5) (centrifugation 30000 × g, 10 rain) and finally resuspended in buffer (50 mM Tris, HCI/120 mM NaC1/5 mM KC1, pH 7.5) to a concentration between 0.5 and 1.0 mg protein/ml. The samples were kept at - 8 0 ° C until analysis. 2.2. Rat brain membranes
Rats were decapitated, the brains removed and placed on ice. Whole brain except the brain-stem and cerebellum was homogenized (Ultra-Turrax, 30 s high speed) in 10 ml ice-cold buffer (50 mM Tris, HC1/120 mM NaC1/5 mM KC1, pH 7.5). The homogenate was centrifuged at 30 000 × g for 10 min, and the pellet resuspended in the same buffer and centrifuged twice. After the third resuspension, buffer was added to give a protein concentration of about 4 mg membrane protein/ml (corresponding to about 100 mg brain tissue/ml). The membranes were stored at - 8 0 ° C . 2. 3. Protein analysis
The protein concentrations in the membrane preparations were determined with the Lowry method modified for membrane proteins according to Peterson (1977). 2.4. Drugs
2. M a t e r i a l s a n d m e t h o d s
2.1. Platelet membranes
Platelets were obtained from 500 ml blood, sodium citrate was used as an anticoagulant. Red blood cells and leucocytes were removed by lowspeed centrifugation (200 x g, 10 min, room temperature). All the following steps were performed at 0-4°C. Platelets were precipitated from the platelet-rich plasma by centrifugation at 30000 × g, 10 rain. The platelets were washed twice with 8 ml buffer (50 mM Tris, HCI/150 mM NaC1/20 mM EDTA, pH 7.5), and frozen as a dry pellet at - 80°C. Lysis and homogenization of the platelets
[3H]Imipramine 20 C i / m m o l - ( A - m e r s h a m , England) and [3H]paroxetine 19.6 C i / m m o l (NEN, USA). The drugs used in the dissociation experiments were generously provided by Astra, Ciba Geigy, Ferrosan, Hoechst, Lilly, Lundbeck, Organon, Pfizer, Rhone-Poulenc, Roche and Upjohn. 2.5. Dissociation rate determinations
Membrane suspension, 30-50 ml, (platelets about 0.5 mg protein/ml, rat brain about 1 mg protein/ml) were incubated with [3H]imipramine at 0°C or [3H]paroxetine at 20°C for 2 h. The
[3H]imipramine concentrations were 2 nM in platelet assays and 6 nM in the rat brain assay. The [3H]paroxetine concentration was 1 nM in both the platelet and the rat brain assays. Aliquots (200 /~1) were diluted into 10 ml buffer (50 x ) containing the various drugs. Experiments with [3H]imipramine were carried out at 0°C whereas experiments with [3H]paroxetine were carried out at 20°C. The samples were filtered 2 min to 4 h after the 50 × dilution step, through Whatman G F / F glass-fiber filters and washed with 3 x 5 ml 0°C buffer. All determinations were done in triplicate, and the median values were used to determine the dissociation rates. In each assay the dissociation rate in the presence of imipramine was used as a control. The mean half-life of the [3H]imipramine platelet membrane binding site complex in the presence of imipramine was 60 _+ 3.4 min (S.E.M.). The effects on the dissociation rates of the other drugs were mostly determined in single experiments, using 12-15 different time intervals between dilution and filtration. To determine the non-specific binding, tubes from each dissociation rate determination were heated to 37°C for 3 h (ensuring that all specifically bound ligand had dissociated from the binding site). The samples were then returned to 0°C or 20°C respectively for 20 min before filtration. The filters were dried overnight and counted in Pico-Flour 15 ®.
3. Results
Figure 1 shows the effect of dilution with 200 /~M citalopram, paroxetine, imipramine, or a 50 x dilution with buffer alone on the dissociation rate of the [3H]imipramine platelet membrane binding site complex. It is seen that all three drugs had an effect on the dissociation of the [3H]imipramine binding site complex. Citalopram decreased, and paroxetine increased the dissociation rate and the dilution-induced dissociation of the slow component was abolished by imipramine. Figure 2 shows a dose-response experiment where the dissociation of the [3H]imipramine platelet binding site complex was induced by dilution with various concentrations (between 10 and
CPM 2°°ct 3H-IMIPRAMINE PLATELETS
IOOC
io¢
z0
I
[
1
I
I
2
3
4
HOURS
>
Fig. 1. Dissociation rates of [ 3H]imiprarnine platelet membrane binding site complexes: O 200 /~M citalopram, • 200 ~tM imipramine, zx 200 ~tM paroxetine or X after a 50 x dilution with buffer alone. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after the preincubated membranes had been diluted 50 x with buffer containing the various drugs under investigation.
300 /~M) of paroxetine or citalopram. It is seen that both the paroxetine and the citalopram effects were dose-dependent, and that half maximal effects were obtained with about 80 /~M paroxetine or 15 ~tM citalopram. Serotonin (data not shown), which decreased the dissociation rate, was found to have a half maximal effect at about 120/~M in similar experiments. Figure 3 shows the dissociation rate of the [3 H]imipramine platelet binding site complex when drugs with different effects on the dissociation rate were mixed. Citalopram alone, or mixtures of citalopram and imipramine or citalopram and paroxetine gave essentially the same results namely a very low dissociation rate. This latter effect was also seen when dissociation first took place during 30 rain in 400/~M paroxetine and an equal volume of 400/xM citalopram was then added at this time
CPM 3H-IMIPRAMINE
CPM
PLATELETS
3H- I M I P R A M I N E iq60120 80 •
40
-
eo
PLATELETS
2OOC /u.M
tOOC -
500
io CITALOPRAM
\
\ %
200
\
% . ~ I M I P R A M I NE
PAROXETINE I00
IMI P R A M I N E * C I T A L O P R A M n 50
~M 2C eo
iO I
1 I
I 2
1 3
I 4
)
HOURS
Fig. 2. Dissociation rates of [3 H]imipramine platelet membrane binding site complexes in the presence of 10-300 # M citalopram (the upper part of the figure), or in the presence of 10-30 ~M paroxetine (the lower part of the figure). The concentrations of paroxetine or citalopram corresponding to the individual curves are given to the right of the individual curves. The dotted line represents the dissociation rate in the presence of 10 #M imipramine. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after the preincubated membranes had been diluted 50 × with buffer containing the various drugs under investigation.
(at the arrow) bringing the concentrations of both paroxetine and citalopram to 200 /~M. It is seen that in all cases the citalopram effect dominated and abolished the imipramine and the paroxetine effects. It is also seen that the imipramine effect on the dissociation rate was stronger than the paroxetine effect as a mixture of imipramine and paroxetine produced the same results as imipramine alone. The experiments illustrated in fig. 3 therefore show that if mixtures of drugs are used to induce dissociation the effect of the drug leading to the slowest dissociation rate will dominate. Table 1 shows the effects of various antidepres-
IMIPRAMINE PAROXET/N E ÷ IMIPRAM[NE
=
P A R O X E T I N E ~. C I T A L O P R A M
x
PAROXETINE
#.
I 2 HOURS
t3
I 3
1 4
~.
Fig. 3. Dissociation rates of [ 3H]imipramine platelet membrane binding site complexes in the presence of various combinations of drugs. O 200 /zM citalopram, [] 200 #M citalopram+ 200 t~M imipramine, • 200/~M imipramine, • 200 #M imipramine +200 /~M paroxetine, × 400 #M paroxetine, which was diluted at the arrow with an equal volume of 400 #M citalopram, thus bringing both concentrations to 200 #M, and zx 200 #M paroxetine. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after start of the experiments by a 50 × dilution with buffer containing the various drugs under investigation.
sive drugs and of a few other psychotropic drugs on the dissociation rate of the [3H]imipramine platelet membrane binding site complex. It is seen that the antidepressive drugs could be divided into three main groups. A large group with the same effects as imipramine, a smaller group of drugs which decreased the dissociation rate, and paroxetine which increased the dissociation rate. It is noteworthy that the two closely related serotonin reuptake inhibitors paroxetine and femoxetine had opposite effects on the dissociation rate. Figure 4 shows the effect of 200 # M citalopram, 200 /~M paroxetine, 10 /~M imipramine of 50 × dilution with buffer alone on the dissociation rate of the [3H]imipramine rat brain membrane bind-
CPM
TABLE l
3H-IMIPRAMINE
[3H]Imipramine dissociation rates. Dissociation rate (half-life) of the [3H]imipramine platelet or brain membrane binding site complex, in the presence of various drugs. Platelet membrane suspension (200 #1) was preincubated with 2 nM [3H]imipramine (platelet) or 6 nM [3H]imipramine (brain), and diluted 50 xinto 10 ml buffer, containing the drugs in the concentrations described.
BRAIN
200(
1000
Half-life (rain)
200 ttM 200 #M 200/~M 200/aM 200 ttM
citalopram serotonin sertraline mianserin femoxetine
200/~M 200 #M 200/~M 200 #M 200/tM 200 FM
imipramine chlorimipramine fluoxetine indalpine iproniazid zimilidine
200/tM paroxetine 50 x dilution 200/zM cocain 200/~M tryptophane 200/~M alprazolam 200/tM carbamazepin 200 # M chlorpromazine 200/tM nomifensine
Platelet membranes
Brain membranes
600 240 135 125 70
105 40 25 10
50(3
200
tog
12 .
Range 50-60 min.
32 Two component curve. Range 50-60 min for the fast component.
50
25 16 12
ing site complex. The effects differed from the effects on the platelet binding site as both paroxetine and citalopram reduced the dissociation rate of the brain membrane binding site ligand complex, compared with imipramine or 50 x dilution. The dissociation induced by dilution followed a monoexponential curve and was a little slower than the dissociation induced by 10 /~M imipramine. Table 1 shows the effects of a few drugs on the dissociation rate of the [3H]imipramine rat brain membrane binding site complex. Mianserin reduced the dissociation rate as it did in platelets. The difference between the effects of the closely related drugs paroxetine and femoxetine is also seen. The dissociation rates of the [3H]paroxetine binding site complex from both platelets and rat
20
CITALOPRAM
o
o
PAROXETINE
~,
~"
50 x DILUTION
x
M PRAM NE
•
I 2
I 3
I 4
)
HOURS
Fig. 4. Dissociation rates of [3H]imipramine rat brain membrane binding site complexes: © 200 /zM citaloprarn, • 200 /LM imipramine, zx 200 #M paroxetine or X after a 50 × dilution with buffer alone. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after start of the experiments by a 50 x dilution with buffer containing the various drugs under investigation.
brain membranes were affected by the various drugs. The effects were qualitatively the same in platelets and brain as both paroxetine and citalopram reduced the dissociation rates compared with their values measured after dilution with 10 # M imipramine or buffer alone. However, the effects were much stronger on brain membranes (fig. 5) than on platelet membranes (not shown) as both citalopram and paroxetine stabilized the binding site ligand complex in brain membranes to an extent where no dissociation took place during the 4 h the experiment lasted. This is also seen in table 2 showing the effects on the dissociation rates of v a r i o u s antidepressive drugs in both platelet and rat brain membranes.
CPM
TABLE 2
3H-PAROX ETI NE BRAIN
[3H]Paroxetine dissociation rates. Dissociation rate (half-life) of the [3H]paroxetine platelet or brain membrane binding site complex, in the presence of various drugs. Platelet membrane suspension (200 ffl) was preincubated with 1 nM [3H]paroxetine, and diluted 50 × into 10 ml buffer, containing the drugs in the concentrations described.
200C
tO00
-----~=-"
-
^
5
"-
~
~
a
r,
-
Half-life (min) 500
20C
IOC
5C
1
CITALOPRAM
o
o/
PAROXETINE
A
A
50 x DILUTION
x--x
IMIPRAMINE
=--o
2c
[ I
J
[ 2 HOURS
[ 3
1 4
>
Fig. 5. Dissociation rates of [3H]paroxetine rat brain membrane binding site complexes in the presence of: O 200 ffM citalopram, • 10 /~M imipramine, zx 200 ffM paroxetine or x after a 50 x dilution with buffer alone. Ordinate: CPM in sample minus non-specific binding in logarithmic scale. Abscissa: hours after start of the experiments by a 50 x dilution with buffer containing the various drugs under investigation.
4. Discussion High affinity binding of [3H]imipramine and [3H]paroxetine takes place in platelet and neuronal membranes (Langer et al., 1980; Paul et al., 1980; Mellerup et al., 1983). The binding shows saturability in the low n a n o m o l a r range, which is 10-100 times below the plasma concentration range for therapeutic levels of antidepressive drugs. The binding of [3H]imipramine and its affinity for the binding site m a y depend on the structure of the binding site itself, which in turn may depend on the structure of the protein matrix. M a n y factors influence the protein molecules, e.g. ions, hormones and drugs. Drugs can affect the affinity between [3H]imipramine and the binding sites in two ways, either in a competitive way by interaction at the binding site, or in a non-competitive way by effects on the protein structure and function. In equilibrium conditions the presence of
Platelet membranes
Brain membranes
200/LM citalopram 200 #M paroxetine 200/LM serotonin 200/.t M mianserin 200 ~ M femoxetine 200 ~M irnipramine 200/xM fluoxetine
410 240 220 220 130 100 70
210 oz 540 260 260
50 X dilution 200 ffM alprazolam 10/~M imipramine 200 ~tM indalpine 200/.t M carbamazepin 200/~M zimilidin
Range 60-70 rain.
Range 120-130min.
n a n o m o l a r concentrations of various antidepressive drugs will decrease the affinities of [3H]imipramine and [3H]paroxetine and their respective binding sites because of competitive binding. Competitive reactions will not take place in experiments where the dissociation rate is measured, and the membranes with b o u n d [3H]imipramine or [3H]paroxetine are strongly diluted. A n y change in affinity revealed in such experiments is due to non-competitive interactions. There are reports of similar experiments in which serotonin (Wennogle and Meyerson, 1983), imipramine (Phillips and Williams, 1984) and recently tryptamine (Segonzac et al., 1985) changed the affinity of [3H]imipramine for its binding site non-competitively. The present results show that a large n u m b e r of antidepressants also possess this ability to affect the affinity of the binding protein, with a remarkable drug specificity of this effect. Thus, citalopram strongly increased the affinity of [3H]imipramine and [3H]paroxetine for their respective binding sites in both platelet and neuronal membranes. Paroxetine, as opposed to citalopram decreased the affinity of [3H]imipramine for the platelet binding site to far below the normal value mea-
sured in the absence of unlabeled drug. However, both paroxetine and citalopram increased the affinity between [3H]paroxetine and the platelet binding site. Thus the effect of paroxetine on [3H]imipramine and [3H]paroxetine binding to platelet membranes added further evidence for the earlier finding (Plenge and Mellerup, 1984; Mellerup et al., 1985) that these two drugs have separate binding sites on a common protein complex, probably the serotonin transport mechanism, in the platelet membrane. Paroxetine, like citalopram, strongly increased the affinity between [3H]paroxetine and its binding site in neuronal membranes whereas the affinity for [3H]imipramine was only moderately increased. Thus paroxetine increased the affinity for [3H]paroxetine in both platelet and neuronal membranes, whereas the paroxetine effect on [3 H]imipramine binding was opposite in these two membranes. These results are in agreement with results which indicate that the [3H]paroxetine binding polymer may be the same in platelets and neurons, whereas the [3H]imipramine binding polymers probably are different in these tissues (Mellerup and Plenge, unpublished results). The citalopram-induced increase in affinity between [3H]imipramine and the platelet binding site was dose-dependent, and was seen clearly with 10/~M. Similarly, the decrease in affinity between [3 H]imipramine and the platelet binding site induced by paroxetine was also dose-dependent (fig. 2). The non-competitive action of paroxetine on the structure or action of the [3H]imipramine binding polymer leading to a decrease in affinity for [3H]imipramine, was promptly reversed by the addition of citalopram (fig. 3). Thus, the effect of citalopram could take place in the presence of both paroxetine or imipramine whereas the opposite was not found. Imipramine itself, compared to dilution without addition of unlabelled drug, caused little alteration in the affinity for [3H]imipramine and [3H]paroxetine in platelets or in neuronal membranes. However imipramine changed the biphasic course of the dissociation curve for [ 3H]imipramine bound to platelet membranes to a monophasic course in agreement with the results of Phillips and Williams (1984). Several other drugs were analysed in the dis-
sociation experiments, and it was generally found that drugs devoid of antidepressive activity did not affect the dissociation of [3H]imipramine from platelet membranes. Antidepressive drugs, on the other hand affected the dissociation rate, by either increasing or decreasing the dissociation rate, or by changing the biphasic course to a monophasic one. However, with the relatively small number of drugs tested, it was impossible to evaluate whether these kinetic effects could be of clinical relevance. It cannot be excluded that other drugs or chemical compounds not related to psychopharmacology may influence the [3H]imipramine and [3H]paroxetine binding polymers in a similar way as this possibility has not been tested thoroughly.
Acknowledgements This study was supported by the Danish Medical Research Council and P. Carl Petersens Fund and Dr. Med. Vet. Axel Thomsen and Hustru Martha Thomsens Fund.
References Langer, S.Z., C. Moret, R. Raisman, M.C. Dubocovich and M. Briley, 1980, High-affinity [3H]imipramine binding in rat hypothalamus: association with uptake of serotonin but not of norepinephrine, Science 210, 133. Mellerup, E.T., P. Plenge and M. Engelstoft, 1983, High affinity binding of [3H]paroxetine and [3H]imipramine to hu. man platelet membranes, European J. Pharmacol. 96, 303. Mellerup, E.T., P. Plenge and M. Nielsen, 1985, Size determination of binding polymers for [3H]imipramine and [3H]paroxetine in human platelet membranes, European J. Pharmacol. 106, 411. Paul, S., M.M. Rehavi, P. Skolnick and F.K. Goodwin, 1980, Demonstration of specific 'high-affinity' binding sites for [3H]imipramine of human platelets, Life Sci. 26, 953. Peterson, G.L., 1977, A simplification of the protein assay method of Lowry et al. which is more generally applicable, Anal. Biochem. 83, 346. Phillips, O.M. and C. Williams, 1984, Multiple affinity states for [3H]imipramine binding sites on human platelet membranes, European J. Pharmacol. 100, 247. Plenge, P. and E.T. Mellerup, 1984, Temperature dependence of the binding of [3H]labeled imipramine and [3H]labeled paroxetine to human platelet membrane, Biochim. Biophys. Acta 770, 22.
Raisman, R., M.S. Briley and S.Z. Langer, 1980, Specific tricyclic antidepressant binding sites in rat brain characterized by high-affinity [3H]imipramine binding, European J. Pharmacol. 61, 373. Segonzac, A., R. Raisman, T. Tateishi, H. Schoemaker, P.E. Hicks and S.Z. Langer, 1985, Tryptamine, a substrate for
the serotonin transporter in human platelets, modifies the dissociation kinetics of [3H]imipramine binding: possible allosteric interaction, J. Neurochem. 44, 349. Wennogle, L.P. and L.R. Meyerson, 1983, Serotonin modulates the dissociation of [3H]imipramine from human platelet recognition site, Euroepan J. Pharmacol. 86, 303.