Relationship between brain serotonin transporter binding, plasma concentration and behavioural effect of selective serotonin reuptake inhibitors

Materials

[³H]Paroxetine (795.5 GBq mmol⁻¹) was purchased from Dupont-NEN Co. Ltd. (Boston, MA, U.S.A.). Fluvoxamine maleate was purchased from Tocris (U.K.). Fluoxetine hydrochloride, paroxetine hydrochloride and sertraline hydrochloride were kindly donated by Eli Lilly pharmaceuticals (Greenfield, IN, U.S.A.), GlaxoSmithKline pharmaceuticals (West Sussex, England) and Pfizer Inc. (Groton, CT, U.S.A.), respectively. All other drugs and materials were obtained from commercial sources.

Animals

Male ICR strain mice at 6–8 weeks of age (Japan SLC Inc., Shizuoka, Japan) were housed 10 per cage in the laboratory with free access to food and water, and were maintained on a 12 h dark/light cycle in a room with controlled temperature (24±1°C) and humidity (55±5%). This study was conducted according to guidelines approved by the Experimental Animal Ethical Committee of University of Shizuoka.

Measurements of specific binding of [³H]paroxetine

Mice were fasted for 16 h before the administration of drugs, and received orally fluvoxamine (6.91–69.1 μmol kg⁻¹), fluoxetine (2.89–28.9 μmol kg⁻¹), paroxetine (0.27–2.67 μmol kg⁻¹) and sertraline (2.92–29.2 μmol kg⁻¹) dissolved in distilled water. At 0.25–48 h after the drug administration, mice were exsanguinated by taking the blood from descending aorta under light anesthesia with diethyl ether, and brain was perfused with 0.9% NaCl from the aorta. Then, the whole brain tissue was removed. The plasma from mouse blood was isolated by centrifugation, and stored at –80°C until the drug concentration was determined. The brain tissue was homogenised in 19 volumes of 50 mM Tris-HCl buffer (pH 7.4) containing 120 mM NaCl and 5 mM KCl with a Polytron homogeniser, and the homogenate was centrifuged at 40,000 × g for 15 min. The pellet was resuspended in 24 volumes of the buffer. All steps for the tissue preparation were performed at 4°C. The binding assay for SERT in brain homogenates from mice was performed by using [³H]paroxetine, as previously described (Habert et al., 1985). Briefly, the brain homogenates (approximately 400 μg protein) were incubated with six different concentrations (0.03, 0.1, 0.3, 0.5, 1.0, 2.0 nM) of [³H]paroxetine for 2 h at 20°C in 50 mM Tris-HCl buffer (pH 7.4) containing 120 mM NaCl and 5 mM KCl. The reaction was terminated by rapid filtration (Cell Harvester, Brandel Co., Gaithersburg, MD, U.S.A.) through Whatman GF/B glass fibre filters, and filters were rinsed three times with 2 ml of ice-cold buffer. Tissue-bound radioactivity was extracted from filters overnight in scintillation fluid (2 l of toluene, 1 l of Triton X-100, 15 g of 2,5-diphenyloxazole, 0.3 g of 1,4-bis[2-(5-phenyloxazolyl)]benzene) and it was determined in a liquid scintillation counter. Specific binding of [³H]paroxetine was determined experimentally from the difference between counts in the absence and presence of 10 μM fluoxetine. All assays were conducted in duplicate. Every binding experiment was performed using fresh tissues. Protein concentration was measured according to the method (Lowry et al., 1951) using bovine serum albumin as standard.

Determination of SSRI and its active metabolite in plasma

The concentration of SSRIs and norfluoxetine in mouse plasma was determined by the column-switching HPLC with clomipramine as an internal standard, according to the previous method (Meineke et al., 1998) with some modification. Briefly, 400 μl of plasma was utilised, and 50 μl of internal standard (final concentration 500 ng ml⁻¹) and 50 μl of 2 M NaOH were added to the plasma. Samples were then vortex-mixed and consequently extracted into 2 ml of n-heptan/3-methyl-1-butanol (985/15, v v⁻¹) on a reciprocating shaker for 20 min. Then, the upper organic layer was transferred and dried under a gentle stream of nitrogen. The residue reconstituted in 110 μl of mobile phase, and it was analysed by HPLC. The HPLC system consisted of two pumps (model LC-9A, Shimadzu, Tokyo, Japan), a six-port switching valve (model FCV-2AH, Shimadzu, Tokyo, Japan) controlled by the use of a system controller (model SCL-6B, Shimadzu, Tokyo, Japan), an injector (SIL-6B, Shimadzu, Tokyo, Japan) and a variable wavelength UV detector (Hitachi, Tokyo, Japan). The UV detector was set at 254 nm (fluvoxamine), 225 nm (fluoxetine and norfluoxetine), 295 nm (paroxetine) and 214 nm (sertraline). Integration of peak area was performed by a computing integrator (model Chromatopac C-R4A, Shimadzu, Tokyo, Japan). Chromatography was performed on a reversed-phase. Column I was a Shim-pack SPC-RP3, 9 μm polyvinylalcohol resin, 30 × 4 mm inside diameter, pretreatment column (Shimadzu, Tokyo, Japan). Column II was a Shim-pack CLC-CN(M), 5 μm cyanopropyl group, 250 × 4.6 mm inside diameter, analytical column (Shimadzu, Tokyo, Japan). The columns were maintained at 50°C. The mobile phase consisted of 0.01 M phosphate buffer (pH 6.8), acetonitrile and methanol (100 : 123 : 40). The sensitivity limits of plasma assay were 5 ng ml⁻¹ (fluvoxamine, fluoxetine and norfluoxetine) and 2.5 ng ml⁻¹ (paroxetine and sertraline).

Behavioural tests

Mice received orally fluvoxamine (69.1 μmol kg⁻¹), fluoxetine (28.9 μmol kg⁻¹), paroxetine (2.67 μmol kg⁻¹) and sertraline (29.2 μmol kg⁻¹), and control animals received vehicle. At 0.25–48 h after the administration, mice were tested for the marble-burying test. Briefly, an open cubic transparent plastic box (22.5 × 33.8 × 14.0 cm) was used, and 20 clean glass marbles (15 mm diameter) were evenly spaced (5 cm apart) on sawdust (5 cm deep) as previously described (Njung'e & Handley, 1991; Ichimaru et al., 1995). Mice were placed into the cubic box, and the number of marbles buried at least two-thirds were counted for 30 min. The total locomotor activity of animals for 30 min was also counted by an activity sensor (NS-AS01, Neuroscience Inc., Tokyo, Japan).

Analysis of binding data and estimation of SERT-binding activity

Analysis of binding data was performed as described previously (Yamada et al., 1980). The apparent dissociation constant (K_D) and maximal number of binding sites (B_max) for [³H]paroxetine were estimated by Rosenthal analysis of the saturation data (Rosenthal, 1967). The SERT-binding activity (percent increase in K_D values by each SSRI) was calculated by the equation: K_D (drug)/K_D (control) × 100, where K_D (control) and K_D (drug) are K_D values for specific [³H]paroxetine binding in the brains of vehicle- and drug-treated mice, respectively.

The ability of SSRIs and norfluoxetine to inhibit specific [³H]paroxetine (0.3 nM) binding in vitro was estimated by IC₅₀ values, which are the molar concentrations of unlabelled drugs necessary for displacing 50% of specific binding (estimated by log probit analysis). The inhibition constant, K_i, was calculated from the equation, K_i=IC₅₀/(1+L/K_D), where L equals concentration of the radioligand.

Pharmacokinetic analysis

The pharmacokinetic parameters of each drug were estimated with noncompartmental methods as previously described (Caccia et al., 1990; Yoon et al., 2000). The terminal elimination rate constant (β) was estimated from the least-square regression slope of terminal log-linear plasma concentration data points. The half-life (t_1/2) of the elimination phase was calculated as ln 2/β. The total area under the plasma concentration versus time curve (AUC) was calculated by the trapezoidal rule, with extrapolation to the infinity by the terminal slope. The first moment of the plasma concentration versus time profile (AUMC) was determined similarly after multiplying each concentration by its time. Apparent total clearance (CL/F) and volume of distribution (Vd/F) were calculated by the equations:

in which F indicates bioavailability. The t_1/2 and AUC of SERT-binding activity were also calculated as described above.

Statistics

The results were reported as mean±s.e.m. Statistical analysis of the data was performed by one-way analysis of variance followed by Dunnett's test for multiple comparisons. A value of P<0.05 was considered significant.

Effects of oral administration of SSRIs on specific [³H]paroxetine binding in mouse brain

In the in vitro binding experiment, fluvoxamine (1–100 nM), fluoxetine (1–100 nM), norfluoxetine (1–100 nM), paroxetine (0.1–10 nM) and sertraline (1–100 nM) inhibited specific [³H]paroxetine binding in mouse brain homogenates in a concentration-dependent manner and their K_i values were 5.52±0.82, 10.8±1.9, 8.72±1.30, 0.54±0.04 and 3.39±1.25 nM (n=3–4), respectively.

The values of K_D and B_max for [³H]paroxetine binding were determined in brains of mice at 0.25–48 h after oral administration of fluvoxamine, fluoxetine, paroxetine and sertraline (Table 1). Following oral administration of fluvoxamine at doses of 6.91, 23.0 and 69.1 μmol kg⁻¹, there were dose-dependent increases in K_D without a change in B_max for brain [³H]paroxetine binding at 0.5 h after the administration, compared to the control values (Table 1). The enhancements at 23.0 (0.5 h) and 69.1 (0.25–4 h) μmol kg⁻¹ of this drug were statistically significant and maximal (4.8- and 23.1-fold, respectively) at 0.5 or 1 h later.

Similarly, there were significant (3.2-, 7.0- and 5.7-fold, respectively) increases in K_D for [³H]paroxetine in the mouse brain at 1, 4 and 12 h after oral administration of fluoxetine (8.68 μmol kg⁻¹). Higher dose (28.9 μmol kg⁻¹) of this drug brought about greater (5.5-, 10.4-, 19.6-, 22.7- and 12.7-fold, respectively) increases in K_D at 0.5, 1, 4, 12 and 24 h after the administration, and enhanced significantly (17–33%) B_max at 1–24 h. The maximal effect by fluoxetine tended to be seen at 4 or 12 h later.

Oral administration of paroxetine at relatively low dose (0.80 and 2.67 μmol kg⁻¹) caused dose-dependent increases in K_D for brain [³H]paroxetine binding, and the extents of enhancement were statistically significant at 4 h (0.80 μmol kg⁻¹) and 1, 4 and 12 h (2.67 μmol kg⁻¹) after the oral administration. The maximal increases (5.0- and 15.1-fold, respectively) at each dose were seen at 4 h later. The significant decreases (29.7 and 28.3 %, respectively) of B_max were observed at 4 and 12 h after the oral administration of paroxetine (2.67 μmol kg⁻¹).

Sertraline at oral doses of 2.92, 8.75 and 29.2 μmol kg⁻¹ brought about dose-dependent increases in K_D without a change in B_max for brain [³H]paroxetine binding at 0.5–12 h later, and the enhancement (2.5-, 5.6- and 14.8-fold, respectively) at each dose was maximal at 4 h later.

Based on the linear regression analysis of oral doses of SSRIs versus their increased rates of K_D for brain [³H]paroxetine binding (Figure 1), oral doses of fluvoxamine, fluoxetine, paroxetine and sertraline, which increased K_D for [³H]paroxetine binding by 10-fold, were 31.3, 14.3, 1.74 and 18.7 μmol kg⁻¹, respectively.

Figure 2a illustrates the time courses of enhancement of K_D for brain [³H]paroxetine binding in mice after oral administration of fluvoxamine (69.1 μmol kg⁻¹), fluoxetine (28.9 μmol kg⁻¹), paroxetine (2.67 μmol kg⁻¹) and sertraline (29.2 μmol kg⁻¹). The t_1/2 of elimination phase and AUC of the enhancement were largest for fluoxetine, followed by sertraline and paroxetine ≫ fluvoxamine (Table 2).

Plasma concentration of SSRIs

Plasma concentrations of SSRIs and norfluoxetine were measured in mice after the oral administration of each SSRI (Figure 2b). The plasma concentration of fluvoxamine after the oral administration of 69.1 μmol kg⁻¹ peaked at 1 h (453 nM) and decreased with the time. Fluoxetine undergoes extensive metabolic conversion, leading to the active metabolite norfluoxetine and other nonactive metabolites (Hiemke & Härtter, 2000). After the oral administration of fluoxetine (28.9 μmol kg⁻¹) in mice, plasma concentrations of fluoxetine and norfluoxetine peaked at 4 h (450 nM) and 12 h (439 nM), respectively, and norfluoxetine decreased more slowly than fluoxetine. In the case of oral administration of paroxetine (2.67 μmol kg⁻¹) and sertraline (29.2 μmol kg⁻¹) in mice, the plasma concentrations peaked at 4 h later (36.9 nM and 103 nM, respectively), and decreased with the time. The time courses of rise and decline of plasma concentrations of SSRIs were reasonably parallel to those of increases of K_D for brain [³H]paroxetine binding (Figure 2a and b).

Pharmacokinetic parameters were estimated from the plasma concentrations of SSRIs and norfluoxetine in mice. As shown in Table 2, the t_1/2 value for plasma concentration was largest in norfluoxetine, followed by fluoxetine, sertraline > paroxetine > fluvoxamine. The AUC values were much greater in norfluoxetine and fluoxetine than other SSRIs. The CL/F values for fluvoxamine, paroxetine and sertraline were greater (17.7-, 2.5- and 7.0-fold, respectively) than that of fluoxetine, and Vd/F value for sertraline was the greatest.

Behavioural effects

The marble-burying behaviour was significantly (66.7, 69.4 and 36.7%, respectively) suppressed at 0.25, 1 and 4 h after the oral administration of fluvoxamine (69.1 μmol kg⁻¹) compared with the control values (Figure 3a), and it was recovered to the control value at 12 h. At 1, 4 and 24 h after oral administration of fluoxetine (28.9 μmol kg⁻¹), there was significant (31.5, 84.6 and 18.6%, respectively) attenuation of marble-burying behaviour (Figure 3b). In the case of paroxetine (2.67 μmol kg⁻¹), significant (40.3 and 46.6%, respectively) suppression of marble-burying behaviour was also observed at 1 and 4 h after the oral administration, and the behaviour was recovered to the control value at 12 and 24 h later (Figure 3c). Similarly, oral administration of sertraline (29.2 μmol kg⁻¹) attenuated significantly (18.8, 75.5 and 23.3%, respectively) the marble-burying behaviour at 1, 4 and 12 h later (Figure 3d). Thus, the maximal suppression by SSRIs was observed at 0.25 and 1 h (fluvoxamine), 4 h (fluoxetine), 1 and 4 h (paroxetine) and 4 h (sertraline), respectively.

On the other hand, the oral administration of SSRIs at the doses that suppressed significantly the marble-burying behaviour had no change in locomotor activity in mice (Figure 4).