The Poorly Membrane Permeable Antipsychotic Drugs Amisulpride and Sulpiride Are Substrates of the Organic Cation Transporters from the SLC22 Family

Abstract

Variations in influx transport at the blood-brain barrier might affect the concentration of psychotropic drugs at their site of action and as a consequence might alter therapy response. Furthermore, influx transporters in organs such as the gut, liver and kidney may influence absorption, distribution, and elimination. Here, we analyzed 30 commonly used psychotropic drugs using a parallel artificial membrane permeability assay. Amisulpride and sulpiride showed the lowest membrane permeability (P_e < 1.5 × 10⁻⁶ cm/s) and will require influx transport to penetrate the blood-brain barrier and other physiological barriers. We then studied the uptake of amisulpride and sulpiride by the organic cation transporters of the SLC22 family OCT1, OCT2, OCT3, OCTN1, and OCTN2 Amisulpride was found to be transported by all five transporters studied. In contrast, sulpiride was only transported by OCT1 and OCT2. OCT1 showed the highest transport ability both for amisulpride (CL_int = 1.9 ml/min/mg protein) and sulpiride (CL_int = 4.2 ml/min/mg protein) and polymorphisms in OCT1 significantly reduced the uptake of both drugs. Furthermore, we observed carrier-mediated uptake that was inhibitable by known OCT inhibitors in the immortalized human brain microvascular endothelial cell line hCMEC/D3. In conclusion, this study demonstrates that amisulpride and sulpiride are substrates of organic cation transporters of the SLC22 family. SLC22 transporters may play an important role in the distribution of amisulpride and sulpiride, including their ability to penetrate the blood-brain barrier.

Electronic supplementary material

The online version of this article (doi:10.1208/s12248-014-9649-9) contains supplementary material, which is available to authorized users.

INTRODUCTION

In order to reach their therapeutic target, psychotropic drugs have to cross the blood-brain barrier. The paracellular diffusion of compounds through the blood-brain barrier is restricted due to the presence of tight junctions between the endothelial cells that constitute the brain capillaries. The transcelullar diffusion of compounds is limited by insufficient membrane permeability or by the presence of efflux transporters. Some small molecules, for example nutrients, overcome this limitation by using influx transporters to enter the brain (1). Membrane influx transporters are known to be important for the distribution of drugs both in the brain and in other organs relevant for drug metabolism and elimination (2). However, the current knowledge about influx transport of psychotropic drugs is very limited.

Most psychotropic drugs have a relatively high LogD_7.4. Therefore, according to the Biopharmaceutics Drug Disposition Classification System (3), their cellular uptake should not depend on carrier-mediated transport. However, psychotropic drugs that have low membrane permeability will likely require a membrane transporter to enter the brain. Indeed, drugs with lower membrane permeability have a lower brain-to-plasma ratio of the unbound drug. Amisulpride and sulpiride have a brain-to-plasma ratio below 0.1, in contrast to highly lipophilic psychotropic drugs, which have a brain-to-plasma ratio close to 1 (4,5).

Psychotropic drugs are mainly weak organic bases. They are positively charged at the physiological pH of 7.4, and some of them have been reported to inhibit organic cation transporters of the SLC22 family (6,7). Therefore, organic cation transporters from this family may also mediate the penetration of psychotropic drugs through the blood-brain barrier.

The SLC22 family is a large group of polyspecific transporters that includes the five organic cation transporters: OCT1 (SLC22A1), OCT2 (SLC22A2), OCT3 (SLC22A3), OCTN1 (SLC22A5), and OCTN2 (SLC22A5) (2,8). OCT1 is strongly expressed in the sinusoidal membrane of human hepatocytes and is responsible for the hepatic uptake of several weak basic drugs (9–13). OCT2 is mainly expressed on the basolateral membrane of epithelial cells in the kidney proximal tubules, being responsible for the renal elimination of several drugs (14). OCT1 and OCT2 have been both proposed to be expressed at the blood-brain barrier (15). The anticonvulsant lamotrigine has been shown to be a substrate for OCT1 in immortalized human brain endothelial cells (16). Additionally, several psychotropic drugs have been shown to be inhibitors of OCT1 and OCT2 (6,7). OCT3, also known as extraneuronal monoamine transporter (EMT), is expressed ubiquitously and is known to transport monoamines as well as other organic cations (9,17,18). OCTN1 and OCTN2 are expressed in several tissues important for drug distribution, including the kidney (both OCTN1 and OCTN2). Importantly, OCTN2 is also expressed at the blood-brain barrier (17). OCTN1 and OCTN2 transport zwitterionic substrates like ergothioneine and l-carnitine, respectively, but are also able to transport organic cations (18–20).

In this study we used an artificial membrane assay (parallel artificial membrane permeability assay (PAMPA)) to identify psychotropic drugs with low membrane permeability. These drugs may require influx transporters to penetrate the blood-brain barrier. Therefore, we used transporter-overexpressing HEK293 cells to characterize in detail the ability of organic cation transporters from the SLC22 family to transport the poorly permeable psychotropic drugs amisulpride, sulpiride, tiapride, and sultopride. Finally, we studied the uptake of amisulpride and sulpiride in hCMEC/D3 cells, an immortalized human brain microvascular endothelial cell line commonly used as a model for the blood-brain barrier.

MATERIALS AND METHODS

Materials

Radiolabeled ³H-Sulpiride, with a specific activity of 80.6 Ci/mmol, was obtained from Hartmann Analytic (Braunschweig, Germany). Acetonitrile and methanol were obtained from Merck (Darmstadt, Germany). Sultopride and the internal standard sumatriptan-d6 were obtained from Santa Cruz Biotechnology (Heidelberg, Germany). All other non-radiolabeled chemicals were obtained from Sigma-Aldrich (Taufkirchen, Germany). PAMPA 96-well plates were obtained from BD Biosciences (Heidelberg, Germany). The 96-well Costar® UV plates were obtained from Corning (NY, USA). HEK293 T-REx™ cells, the Flp-In™ system, the pcDNA5.1 plasmid, the pCR-XL-TOPO plasmid, Hank’s buffered salt solution (HBSS), media, and cell culture additives were obtained from Life Technologies (Darmstadt, Germany). Cultrex® rat collagen I was obtained from R&D Systems (Wiesbaden-Nordenstadt, Germany), and EndoGRO^TM-MV (SMCE-004) cell culture media was obtained from Merck. Restriction enzymes were obtained from Thermo Fischer Scientific (Waltham, MA, USA).

Parallel Artificial Membrane Permeability Assay

The PAMPA was performed using PAMPA 96-well plates (BD Biosciences) according to the manufacturer’s instructions. Briefly, 300 μl of phosphate buffered saline (PBS, pH 7.4) solution with increasing drug concentrations was pipetted into the donor wells and 200 μl of PBS (pH 7.4) into the acceptor wells. Each concentration was measured in duplicate. After incubation at room temperature for 5 h, the drug concentrations from both the donor and acceptor wells were determined by UV absorption in a TECAN Ultra Microplate Reader (TECAN, Crailsheim, Germany). The following wavelengths were used: 230 nm for amitriptyline, citalopram, clomipramine, doxepine, fluoxetine, haloperidol, methylphenidate, nortriptyline, and paroxetine; 260 nm for desipramine, fluphenazine, imipramine, lamotrigine, melperon, milnacipram, perphenazine, and tranylcypromine; and 280 nm for amisulpride, clozapine, duloxetine, flupentixol, mirtazapine, olanzapine, paliperidone, perazine, promethazine, quetiapine, risperidone, sulpiride, sultopride, and tiapride.

Membrane permeability (P_e) was calculated as follows:

$$P_{\mathrm{e}}=-ln\left[1-{\mathrm{C}}_{\mathrm{A}}/\left(1-\left(0.3\ast C_{\mathrm{D}}+0.2\ast C_{\mathrm{A}}\right)/0.3\ast C_0\right)\right]/\left[A\times \left(1/V_{\mathrm{D}}+1/V_{\mathrm{A}}\right)\times t\right]$$

where C₀ is the initial concentration in the donor well and C_D and C_A are the end concentrations in the donor and acceptor wells, respectively. A represents the area of each well (0.3 cm²), V_D is the volume of the donor well (0.3 ml), V_A is the volume of the acceptor well (0.2 ml) and t is the time of incubation in seconds (18,000 s = 5 h). The assays were performed using at least two different concentrations for each drug and average P_e values were calculated.

HEK293 Cell Lines Overexpressing OCT1, OCT2, OCT3, OCTN1, and OCTN2

HEK293 cells overexpressing the human OCT1 gene and its variants have been generated from HEK293 T-REx™ cells using the Flp-In™ system as described in detail before (21). HEK293 cells overexpressing human OCT3, and the prcCMV::OCT2 plasmid, were kind gifts from Prof. Koepsell and Dr. Gorboulev (University of Würzburg, Germany). HEK293 cell lines overexpressing the human OCT2, OCTN1, and OCTN2 genes were generated using the Flp-In™ system following the procedure used for OCT1. Briefly, the OCT2 gene was cut from the prcCMV::OCT2 plasmid using BamHI and NotI, and subsequently ligated into the pcDNA5.1 plasmid, which had been cut with the same enzymes. The OCTN1 and OCTN2 open reading frames were amplified from human kidney cDNA by PCR. The cDNA was synthesized from total kidney RNA as previously described (11). The primers used for the amplification of OCTN1 were 5′-AAGTTTCGGATCCGCAGTGGGAAGCATGCGGGACTA-3′ and 5′-ATTTCAGCGGCCGCAACGAATTTCTCCACAGG-3′, and for the amplification of OCTN2 were 5′-GCTCTGTAAGCTTCTGAGGGCGGCATGCGGGACTA-3′ and 5′-AGCTGGCGATATCTGGCAAGACAGTCTTTCC-3′. Artificial restriction sites for BamHI, NotI, EcoRV, and HindIII were integrated into the primers (underlined in the primer sequences). The OCTN1 fragment was originally cloned into pCR-XL-TOPO and then re-cloned into the pcDNA5.1 plasmid using the BamHI and NotI sites. The OCTN2 fragment was cut with EcoRV and HindIII and ligated into the pcDNA5.1 plasmid, which was previously cut with the same restriction enzymes. The obtained constructs were verified by sequencing of the whole reading frame. OCTN1 carried the major leucine₅₀₃ allele of the common Leu503Phe polymorphism. The OCT2, OCTN1, and OCTN2 constructs were chromosomally integrated into HEK293 T-REx™ cells using the Flp-In™ system, according to the manufacturer’s instructions. Genomic integration was confirmed by PCR and sequencing, and transporter overexpression was validated by real-time reverse transcriptase-PCR. The activity of the HEK293 cell lines overexpressing OCT2, OCTN1, and OCTN2 was tested using the model substrates MPP⁺ and TEA⁺ (data available on request).

Uptake Measurements in SLC22-Overexpressing HEK293 Cells

To measure the uptake of sulpiride, 5 × 10⁵ HEK293 cells were cultured in a single well of 12-well plates precoated with poly-d-lysine (1–4 kDa) for 3 days to reach complete confluence. All uptake measurements were performed at pH 7.4. The cells were washed with 1 ml 37°C HBSS. The reaction was started by adding 400 μl 37°C HBSS containing sulpiride. The concentrations of sulpiride used varied between 5 and 500 μM. A mixture of ³H-labeled and non-radiolabeled sulpiride in ratio of 1:4,000 was used. The reaction was stopped after exactly 2 min by adding 2 ml ice-cold HBSS. The uptake was performed for 2 min for all cell lines in order to compare the uptake among the cell lines expressing different transporters. The time was chosen in accordance with previous studies on the organic cation transporters of the SLC22 family (10,21–26). The cells were washed twice with 2 ml ice-cold HBSS and lysed with 500 μl of a 0.1 N NaOH solution containing 0.1% SDS. Four hundred microliters of the cell lysates were supplemented with 9 ml liquid scintillator (AQUASAFE 500 Plus^TM from Zinsser Analytics, Frankfurt-am-Main, Germany) and the accumulated substrate was quantified in a liquid scintillation counter (Beckman LS5000TD from Beckman Coulter, Krefeld, Germany). Counts per minute were normalized to the total protein amount, as determined using the bicinchoninic acid assay (BCA method, (27)).

To measure the uptake of amisulpride, Petri dishes with a diameter of 100 mm (BD Falcon, Heidelberg, Germany) were precoated with poly-d-lysine, 9 × 10⁶ cells were plated and incubated for 48 h to reach confluence. Prior to the uptake measurements, cells were washed with 10 ml 37°C HBSS. The uptake was initiated by adding 5 ml 37°C HBSS supplemented with amisulpride. The concentrations of amisulpride used varied between 5 and 200 μM. After 2 min, the reaction was stopped by adding 20 ml ice-cold HBSS and the cells were washed twice with 20 ml ice-cold HBSS. The cells were transferred to a 2 ml tube and an aliquot was kept for protein quantification. The remaining cells were lysed in 1 ml of a mixture of 80% acetonitrile and 20% 30 mM potassium dihydrogen phosphate buffer (pH 6.5). The cell debris was pelleted by centrifugation (10 min at 16,000 g, 4°C). The supernatant was evaporated to dryness under a nitrogen stream at 40°C. The dry residue was reconstituted in 200 μl of 30 mM dihydrogen phosphate buffer (pH 5.0) and used for the HPLC quantification as described below. Intracellularly accumulated amisulpride was quantified by HPLC and normalized to the total protein content of the sample determined by the BCA method.

To measure the uptake of sultopride and tiapride, 5 × 10⁵ HEK293 cells were cultured in a single well of 12-well plates for 3 days to reach complete confluence. The cells were washed with 1 ml 37°C HBSS. The uptake was started by adding 400 μl 37°C HBSS containing 5 μM of sultopride or tiapride and stopped after exactly 2 min by adding 2 ml ice-cold HBSS. The cells were washed twice with 2 ml ice-cold HBSS and lysed with 500 μL of a lysis solution (80% acetonitrile, 20% water). The cell lysate was centrifuged (16,000 g, 10 min) and the supernatant was diluted 1:10 in 0.1% formic acid. The intracellular amounts of sultopride and tiapride were quantified by LC-MS/MS.

Uptake Measurements in HCMEC/D3 Cells

The immortalized human brain endothelial cell line hCMEC/D3 (28,29) was propagated in EndoGRO^TM-MV medium on surfaces coated with rat collagen I. For uptake measurements, 10⁶ HCMEC/D3 cells were plated in a single well of a 6-well plate and cultured for 2 days to reach complete confluence. The cells were washed with 2 ml 37°C HBSS.

To measure the uptake of sulpiride, the reaction was started by adding 1 ml 37°C HBSS containing 1, 5, or 25 μM sulpiride. We used a mixture of ³H-labeled and non-radiolabeled sulpiride in a molar ratio of 1:400. The reaction was stopped after 2 min by adding 5 ml ice-cold HBSS. The cells were washed twice with 5 ml ice-cold HBSS and lysed with 1 ml of a 0.1 N NaOH solution containing 0.1% SDS. Eight hundred microliters of the cell lysates was mixed with 15 ml liquid scintillator (AQUASAFE 500 Plus^TM from Zinsser Analytics). The amount of sulpiride was quantified in a liquid scintillation counter (Beckman LS5000TD from Beckman Coulter) and normalized to the total protein content of the sample.

To measure the uptake of amisulpride, the reaction was started by adding 1 ml 37°C HBSS containing 1, 5, or 25 μM of amisulpride. The reaction was stopped after exactly 2 min by adding 5 ml ice-cold HBSS. The cells were washed twice with 5 ml ice-cold HBSS and lysed with 1 ml of a lysis solution (80% acetonitrile, 20% water) containing 10 ng/ml of sumatriptan-d6, which was used as internal standard. The cell lysate was centrifuged (16,000 g, 10 min) and the supernatant was diluted 1:10 in 0.1% formic acid and used for the quantification of amisulpride by LC-MS/MS.

The uptake measurements at 4°C were performed after preincubating the cells for 15 min on ice and the uptake reaction was carried out in ice-cold HBSS buffer.

HPLC Quantification of Amisulpride

Amisulpride was quantified as described by Kudris et al. with modifications (30). A LaChrom system (Merck Hitachi, Darmstadt, Germany) was used with an interface (D-7000, Merck Hitachi), a pump (L-7100, Merck Hitachi), an autosampler (L-7200, Merck Hitachi), a fluorescence detector (L-7400, Merck Hitachi), and a degasser (L-7614, Merck Hitachi). Separation was carried out at 40°C on a LiChrospher 100 RP-18e (5 μm, 4 × 12.50 mm) with a LiChrospher100 CN guard column (5 μm, both from Merck). Isocratic elution was performed with a mobile phase of 88% 30 mM potassium dihydrogen phosphate buffer (pH 6.5, adjusted with triethylamine) and 12% acetonitrile at a flow rate of 1.0 ml/min. Metoclopramide was used as internal standard. Fluorescence detection was performed using excitation and emission wavelengths of 274 and 370 nm, respectively. The peaks of amisulpride and metoclopramide were detected with retention times of 9 and 12 min, respectively, and were quantified using the peak areas. We used calibrator concentrations of 1, 5, 10, 25, and 50 ng/ml, and the limit of quantification of amisulpride was 1 ng/ml.

LC-MS/MS Quantification of Tiapride, Sultopride, and Amisulpride

The LC-MS/MS quantification of tiapride, sultopride, and amisulpride was performed based on previously described methods (31,32) with modifications. Briefly, 10 μl of the sample were injected into a chromatographic system comprised of a PerkinElmer/Sciex HPLC coupled with a API4000 tandem mass spectrometer (Applied Biosystems, Darmstadt, Germany). The separation was achieved in a Brownlee SPP RP-Amide column (4.6 × 100 mm, 2.7 μm particle size; Perkin Elmer, Waltham, MA, USA) with a Phenomenex SecurityGuard Standard precolumn (C18, ODS, 4 mm × 2 mm, KJO-4282, Phenomenex, Aschaffenburg, Germany). The mobile phase consisted of 0.1% formic acid, 6.9% acetonitrile, and 1.1% methanol. Under these conditions, the retention times of tiapride, sultopride, and amisulpride were 6.6, 10.7, and 16.0 min, respectively. The mass transitions used for the quantification of tiapride, sultopride, and amisulpride were m/z 329.4 > 256.1, 355.5 > 227.3, and 370.2 > 242.1, respectively.

Prediction of Drug Chemical Properties, Data, and Statistical Analyses

LogD_7.4 and pKa values of the drugs analyzed were estimated based on their chemical structures using ADMET Predictor 5.5 (Simulations Plus, Lancaster, CA, USA). The Michaelis-Menten constant (K_M) and maximal transport rates (V_max) was determined by a non-linear regression fitted to the Michaelis-Menten equation using SigmaPlot 12 (Systat Software Inc., Erkrath, Germany).

The Student’s t test was used for two independent group comparisons. ANOVA was used for multiple group comparisons followed by post hoc pair-wise comparisons using Tukey’s HSD test. Spearman’s correlation was used to analyze the correlation between the estimated LogD_7.4 and measured membrane permeability (P_e) values. The statistical analyses were performed with IBM SPSS Statistics version 21.0 (IBM Corporation, Ehningen, Germany).

RESULTS

Membrane Permeability of Psychotropic Drugs

Most psychotropic drugs are weak bases and positively charged at physiological pH. However, a fraction of the drug exists in a non-charged form and can permeate the cell membrane. We have studied the membrane permeability of 30 psychotropic drugs at pH 7.4 using a PAMPA assay (Fig. 1) Amisulpride and sulpiride were the least membrane permeable of the drugs tested (P_e = 0.36 ± 0.1 × 10⁻⁶ cm/s and P_e = 1.19 ± 0.3 × 10⁻⁶ cm/s, respectively). In comparison, the most membrane permeable drug was doxepine, with P_e of 24.9 ± 1.6 × 10⁻⁶ cm/s. Only 26% of the variability in the drug membrane permeabilities could be explained by variation in the in silico predicted logD_7.4 (r² = 0.26, p = 0.004, Fig. 1).

Fig. 1.

Comparison of the carrier-independent membrane permeabilities of commonly used psychotropic drugs. The membrane permeability (P _e) was measured using a PAMPA assay (BD Biosciences). Shown are means and standard error of the means of at least two independent experiments. In each experiment two different drug concentrations were measured and each measurement was performed in duplicate, resulting in eight or more data points per drug. Upper left hand panel shows the correlation between experimentally determined P _e and the predicted logD_7.4 of the drugs. The logD_7.4 values were estimated based on the chemical structures using ADMET Predictor version 5.5 (Simulations Plus, Lancaster, CA, USA)

Amisulpride and sulpiride are weak bases with pKa of 9.0 and 8.8, respectively (Fig. 2). At the physiological pH of 7.4, 98% of the amisulpride and 96% of the sulpiride molecules will be positively charged. The low membrane permeability and the weak basic chemical properties of amisulpride and sulpiride suggest that both drugs may require carrier-mediated influx transport to penetrate cell membranes and may be good substrates for the organic cation transporters from the SLC22 family.

Fig. 2.

Chemical structures of amisulpride and sulpiride. The pKa values were estimated using ADMET Predictor version 5.5 (Simulations Plus, Lancaster, CA, USA)

Amisulpride and Sulpiride as Substrates of SLC22 Cation Transporters

We compared the ability of the organic cation transporters from the SLC22 family to mediate the cellular uptake of amisulpride and sulpiride. First, we incubated amisulpride at a single concentration of 5 μM with HEK293 cells overexpressing OCT1, OCT2, OCT3, OCTN1, or OCTN2 (Fig. 3a). The experiments were also performed in the presence of known OCT inhibitors. TBA⁺ was used to inhibit OCT1 and OCT2 uptake (9), irinotecan was used to inhibit OCT3 uptake (33), and l-carnitine was used to inhibit OCTN1 and OCTN2 uptake (34).

Fig. 3.

Transport of amisulpride by the organic cation transporters of the SLC22 family. a Cellular uptake of amisulpride at a single concentration of 5 μM in cells overexpressing OCT1, OCT2, OCT3, OCTN1, OCTN2, and the control cell line (transfected with the empty expression vector). The uptake was inhibited either by 1 mM TBA⁺ (OCT1 and OCT2), 250 μM irinotecan (OCT3), or 500 μM l-carnitine (OCTN1 and OCTN2). b and c Concentration dependence of amisulpride uptake by OCT1, OCT2, and OCT3 (b) and by OCTN1 and OCTN2 (c). The results are separated between panels b and c for clarity of presentation. d and e Transporter-mediated uptake of amisulpride. The transporter-mediated component of the total cellular uptake was estimated by subtracting the uptake measured in the control cell lines from the uptake in cell lines overexpressing the corresponding transporter. Shown are the means and standard error of the means of three independent experiments. *p < 0.05 and **p < 0.01, as calculated by the Student’s t test

Amisulpride was transported by all the SLC22 transporters tested, and the inhibitors used were able to significantly inhibit amisulpride transport (Fig. 3a). The strongest transport was observed in cells overexpressing OCT1, followed by OCTN2 (7.1-fold and 5.9-fold increase compared to the control cells transfected with empty pcDNA5, respectively). In cells overexpressing OCTN1, OCT3, and OCT2, the increase was 3.2-fold, 2.9-fold, and 2.2-fold, respectively. The transport by OCT1 and OCT2 was completely inhibited by TBA⁺ (p < 0.01, Student’s t test, Fig. 3a), and the transport by OCT3 was completely inhibited by irinotecan (p < 0.05, Student’s t test, Fig. 3a). The transport of amisulpride by OCTN1 and OCTN2 was inhibitable by l-carnitine (p < 0.05, Student’s t test, Fig. 3a). None of the inhibitors used showed an influence on the uptake of amisulpride in the control cells.

Next, we determined the concentration dependence of the amisulpride uptake using concentrations ranging between 5 and 200 μM. In all the cells overexpressing a transporter, the concentration-dependent uptake of amisulpride was higher than in the control cells (Fig. 3b, c). In order to measure only the transporter-mediated amisulpride uptake, the uptake in the control cells was subtracted from the uptake in the cells overexpressing a transporter (Fig. 3d, e).

The transporter-mediated amisulpride uptake showed typical Michaelis-Menten kinetics (Fig. 3d, e). OCT1 showed the highest affinity (K_M of 31.3 ± 5.4 μM; Table I), while the remaining transporters showed comparatively moderate affinities for the uptake of amisulpride (K_M > 150 μM). All transporters showed similarly low capacity for amisulpride uptake (V_max < 200 pmol min⁻¹ mg protein⁻¹). Therefore, with the highest intrinsic clearance of 1.9 ml min⁻¹ mg protein⁻¹, OCT1 was the best transporter for amisulpride within the organic cation transporters from the SLC22 family.

Table I.

Kinetics of Amisulpride and Sulpiride Transport by Organic Cation Transporters of the SLC22 Family

	K M [μM]	V max [pmol min−1 mgprotein −1]	CLint (V max/K M) [ml min−1 mgprotein −1]
Amisulpride
OCT1	31.3 ± 5.4	59.6 ± 6.4	1.9
OCT2	167.9 ± 32.1	99.2 ± 18.8	0.6
OCT3	191.9 ± 6.1	162.4 ± 28.6	0.8
OCTN1	179.9 ± 20.1	78.8 ± 13.6	0.4
OCTN2	185.3 ± 68.0	168.8 ± 34.8	0.9
Sulpiride
OCT1	259.7 ± 5.4	1081.4 ± 188	4.2
OCT2	187.2 ± 21.6	439 ± 4.7	2.3

Sulpiride uptake was increased by 2.5-fold in cells overexpressing OCT1, and 1.8-fold in cells overexpressing OCT2, compared to the control cells (Fig. 4a). The uptake increase was completely inhibited by TBA⁺ (p < 0.01, Student’s t test). No increase in sulpiride uptake in relation to the control was observed in cells overexpressing OCT3, OCTN1, and OCTN2.

Fig. 4.

Transport of sulpiride by the organic cation transporters of the SLC22 family. a Cellular uptake of sulpiride at a single concentration of 5 μM in cells overexpressing OCT1, OCT2, OCT3, OCTN1, OCTN2, and the control cell line (transfected with the empty expression vector). The uptake was inhibited either by 1 mM TBA⁺ (OCT1 and OCT2), 250 μM irinotecan (OCT3), or 500 μM l-carnitine (OCTN1 and OCTN2). b Concentration dependence of sulpiride uptake by OCT1 and OCT2. c Transporter-mediated uptake of sulpiride. The transporter-mediated component of the total cellular uptake was estimated by subtracting the uptake measured in the control cell lines from the uptake in cell lines overexpressing the corresponding transporter. Shown are the means and standard error of the means of three independent experiments. *p < 0.05 and **p < 0.01, as calculated by the Student’s t test

A concentration-dependent increase in the uptake of sulpiride was observed in the OCT1 and OCT2 overexpressing cell lines, in comparison to the control cells (Fig. 4b). Similarly to amisulpride, the transport of sulpiride showed Michaelis-Menten kinetics (Fig. 4c). Both OCT1 and OCT2 showed low affinity (K_M > 150 μM), but high capacity for the uptake of sulpiride (V_max > 400 pmol min⁻¹ mg protein⁻¹; Table I). The calculated sulpiride intrinsic clearance by OCT1 and OCT2 were 4.3 and 2.4 ml min⁻¹ mg protein⁻¹, respectively. These values are higher than the values obtained for the transport of amisulpride by OCT1 and OCT2, suggesting higher capacity of OCT1 and OCT2 to transport sulpiride than amisulpride.

Effect of OCT1 Polymorphism on the Uptake of Amisulpride and Sulpiride by OCT1

The OCT1 gene SLC22A1 is highly polymorphic. Approximately 10% of Caucasians are homozygous carriers and 40% are heterozygous carriers of a loss-of-function OCT1 allele (12). Our results show that among the organic cation transporters of the SLC22 family, OCT1 is the transporter with the highest intrinsic clearance for both amisulpride and sulpiride (Table I). Therefore, it is important to study how genetic variants in the OCT1 gene affect amisulpride and sulpiride transport. We measured the uptake of amisulpride and sulpiride by OCT1 in cells overexpressing the known common functional variants of OCT1. The uptake of both drugs was significantly lower in cells expressing the loss-of-function OCT1 variants OCT1*2 to *6, in comparison to the wild-type OCT1*1 variant (post hoc Turkey-HSD test p < 0.001; Fig. 5). These results indicate that uptake of amisulpride and sulpiride, in organs where OCT1 plays a role in the distribution, may be reduced in individuals carrying loss-of-function OCT1 polymorphisms.

Fig. 5.

Loss-of-function OCT1 polymorphisms and their effect on the uptake of amisulpride and sulpiride. a The haplotype combinations of the five most frequent loss-of-function polymorphisms and their frequencies in Caucasians. The haplotype combinations were obtained from Tzvetkov et al. (11). b and c HEK293 cells overexpressing the wild-type OCT1*1 (black bars), the loss-of-function variants OCT1*2-*6 (gray bars) and the control cells transfected with the empty vector (white bars) were incubated for 2 min with 5 μM of amisulpride (b) or sulpiride (c). Uptake is shown as a percentage of the uptake in OCT1*1. The means and standard error of the means of three independent experiments are shown. ***p < 0.001 for comparing the variant OCT1 alleles with the OCT1*1 allele (post hoc Turkey-HSD after significant ANOVA test for differences among the groups)

Uptake of Amisulpride and Sulpiride in the hCMEC/D3 Endothelial Cell Line

Next, we analyzed the uptake of amisulpride and sulpiride in brain microvascular endothelial cells. We used hCMEC/D3 cells, a well-characterized immortalized human brain microvascular endothelial cell line (29). We observed concentration- and temperature-dependent uptake of amisulpride and sulpiride (Fig. 6a, b). This indicates that a carrier-mediated uptake mechanism is present for both drugs.

Fig. 6.

Uptake of amisulpride and sulpiride in human brain microvascular endothelial cells. Uptake of amisulpride (a, b) and sulpiride (c, d) was measured in the immortalized human brain microvascular endothelial cell line hCMEC/D3. Panels a and b show the concentration-dependent uptake of amisulpride and sulpiride. Uptake of 1, 5, and 25 μM of amisulpride (a) and sulpiride (b) was measured at 37°C (closed circles) and 4°C (open circles) for 2 min. Panels c and d show the inhibition of the uptake of amisulpride and sulpiride. The uptake of 5 μM amisulpride (c) or 5 μM sulpiride (d) was inhibited with 1 mM of MPP+ or verapamil. Shown are the means and standard error of the means of two or more independent experiments. *p < 0.05 and ***p < 0.001 for comparison with the uptake without inhibitors at 37°C (post hoc Turkey-HSD after significant ANOVA test for differences among the groups)

Furthermore, the uptake of amisulpride and sulpiride was inhibitable by MPP⁺ (Fig. 6c, d), a model organic cation and competitive inhibitor of OCT1, OCT2, and OCT3 (8). MPP⁺ inhibited 45% of the temperature-dependent fraction of the amisulpride uptake (the difference between the uptake at 37 and 4°C) and 71% of the temperature-dependent fraction of the sulpiride uptake. Verapamil inhibited the temperature-dependent uptake of sulpiride by 63%, but did not affect the uptake of amisulpride. These experiments strongly suggest that the uptake of amisulpride and sulpiride in hCMEC/D3 cells is a carrier-mediated process partially mediated by organic cation transporters of the SLC22 family.

Tiapride and Sultopride are not Transported by the Cation Transporters of the SLC22 Family

Tiapride and sultopride, two psychotropic drugs with high structural similarities to amisulpride and sulpiride, also showed low membrane permeability in the PAMPA assays (P_e < 1.5 × 10⁻⁶ cm/s, Fig. 1). We compared the ability of the organic cation transporters from the SLC22 family to mediate the cellular uptake of tiapride and sultopride. In contrast to amisulpride and sulpiride, the uptake of tiapride and sulpiride was not increased in any of the HEK293 cells overexpressing the organic cation transporter and was not decreased by inhibitors specific for the different transporters tested (Fig. 7). Therefore, we concluded that tiapride and sultopride are not substrates of the organic cation transporters of the SLC22 family.

Fig. 7.

Transport of tiapride and sultopride by the organic cation transporters of the SLC22 family. Cellular uptake of tiapride (a) and sultopride (b) at a single concentration of 5 μM in cells overexpressing OCT1, OCT2, OCT3, OCTN1, OCTN2, and the control cell line (transfected with the empty expression vector). The uptake was inhibited either by 1 mM TBA⁺ (OCT1 and OCT2), 250 μM irinotecan (OCT3), or 500 μM l-carnitine (OCTN1 and OCTN2)

DISCUSSION

We have identified amisulpride and sulpiride as drugs with low membrane permeability that will need carrier-mediated transport to enter cells and penetrate the blood-brain barrier. We demonstrated that amisulpride can be transported by OCT1, OCT2, OCT3, OCTN1, and OCTN2, while sulpiride can only be transported by OCT1 and OCT2. SLC22 transporters are expressed at the blood-brain barrier and at other physiological barriers important for drug absorption, distribution, metabolism, and elimination (8,9,35). Therefore, SLC22 cation transporters may be important for the distribution of amisulpride and sulpiride, including their ability to reach their targets in the brain.

Amisulpride and sulpiride are psychotropic drugs in the class of substituted benzamides that are used to treat schizophrenia (36–38). Both drugs are selective dopamine receptor D2 blockers and need to pass the blood-brain barrier in order to reach their site of action. Although amisulpride and sulpiride are low permeability drugs, previous to this study, it was not known how they penetrate the blood-brain barrier. Here, we clearly demonstrate that amisulpride is a substrate of the organic cation transporters of the SLC22 family. To our best knowledge, this is the first study analyzing influx transport of amisulpride.

Sulpiride has been previously suggested as a substrate for PEPT1 (39). However, PEPT1 is not expressed at the blood-brain barrier (35,40). Here, we showed that sulpiride is a substrate of OCT1 and OCT2. Our data is in line with previously reported ability of sulpiride to inhibit OCT2 (7) and suggested involvement of OCTs in the uptake of sulpiride in Caco-2 cells (39).

Previous studies have reported mRNA expression of OCT1, OCT2, OCT3, OCTN1, and OCTN2 (15,16,19,35) and protein expression of OCT1, OCT2, and OCT3 (15,35) at the blood-brain barrier. Furthermore, OCT1 and OCT2 have been suggested to be able to take up lamotrigine, MPTP, and MPP⁺ in the brain (15,16). Taken together with the ability of OCTs to transport amisulpride and sulpiride reported here, it could be suggested that OCT1, OCT2, and other organic cation transporters from the same family may be substantially involved in the transport of amisulpride and sulpiride in the brain.

We also performed transport measurements in the human brain microvascular endothelial cell line hCMEC/D3, a well-characterized model of the blood-brain barrier (28,29). We observed concentration- and temperature-dependent uptake for both amisulpride and sulpiride (Fig. 6), suggesting the presence of carrier-mediated uptake in human brain endothelial cells. Furthermore, the uptake of both amisulpride and sulpiride was inhibited by MPP⁺, a well-known inhibitor of OCT1, OCT2, and OCT3 (8). Taken together with the previously reported functional expression of OCT1 in the hCMEC/D3 cells (16), our data strongly supports the involvement of OCT1 or another organic cation transporter from the SLC22 family in the uptake of amisulpride and sulpiride through the blood-brain barrier.

In contrast to MPP+, verapamil, another known inhibitor of OCT1 and OCT2, did not change the rates of amisulpride uptake into hCMEC/D3 cells (Fig. 6c). This may be a result of simultaneous inhibition of the OCT-mediated influx and MDR1-mediated efflux by verapamil (41). MDR1 is known to transport amisulpride (42) and is functionally expressed in hCMEC/D3 cells (43–45). Therefore, verapamil may inhibit both the uptake and the efflux of amisulpride in hCMEC/D3 cells, resulting in a lack of a net change in the intracellular amisulpride levels (Fig. 6d). In contrast, sulpiride is not a substrate of MDR1 (46). Therefore, verapamil may only inhibit the influx of sulpiride and lead to reduced intracellular sulpiride in hCMEC/D3 cells, as we observed (Fig. 6d). These results indicate possible interplay between influx and efflux processes on uptake through the blood-brain barrier. However, one should account that the levels of expression of OCTs and MDR1 may vary between hCMEC/D3 cells and the HBMECs constituting the blood-brain barrier (20,43). Therefore, the net uptake observed in hCMEC/D3 cells cannot be directly extrapolated to the human blood-brain barrier.

The intrinsic clearance of amisulpride by OCT1 is 2.5-fold higher than the intrinsic clearance by OCTN2, suggesting a dominant role of OCT1 in transport through the blood-brain barrier (Table I). However, one should account for the higher expression of OCTN2 at the blood-brain barrier(35). Furthermore, OCT1 and OCTN2 may be localized on the different sides of the blood brain barrier. OCT1 and OCT2 have been previously show to be localized at the luminal side of the brain microvascular endothelial cells (15) and may mediate the uptake of both amisulpride and sulpiride from the blood into the endothelial cells. In bovine BMECs, OCTN2 is expressed at the basolateral membrane of the endothelium (47). OCTN2 can transport drugs in both directions across the plasma membrane (8). Therefore, OCT1 and OCT2 may both be involved in the uptake from the blood, and OCTN2 may be involved in the basolateral efflux transport of amisulpride from the intracellular space of the capillary endothelial cells into the brain.

The chemical structures of amisulpride and sulpiride are similar (Fig. 2). However, the small structural differences between them seem to account for distinct substrate specificity and transport kinetics (Table I). Specifically, our experiments show that amisulpride is a substrate of OCT1, OCT2, OCT3, OCTN1, and OCTN2, and sulpiride is a substrate of OCT1 and OCT2 only (Figs. 3 and 4 and Table I).

It is interesting to observe that although OCTs transport amisulpride and sulpiride, none of the OCTs transport tiapride or sultopride (Fig. 6). These results suggested that structural differences between these drugs may play a role in substrate recognition by the organic cation transporters of the SLC22 family. One such difference may be the amine moieties near the aromatic ring, which are present in sulpiride and amisulpride, but missing in tiapride and sultopride (Supplemental figure 1).

The maximal plasma concentrations after oral administration of 50 mg amisulpride or 100 mg sulpiride are 0.13 and 0.29 μM, respectively (48,49). These values are much lower than the K_M for the uptake of amisulpride and sulpiride by the transporters studied (Table I), meaning that transport in vivo will not be saturated.

Besides the distribution to the brain, SLC22-mediated transport may allow amisulpride and sulpiride to cross several other pharmacologically relevant barriers. The renal clearances of amisulpride and sulpiride are 330 and 223 ml/min, respectively (48,49) indicating that tubular secretion plays substantial role in their renal elimination. OCT2 is strongly expressed in proximal tubules (8), and it is probable that OCT2 is responsible for the significant tubular secretion of both drugs. Interestingly, in the prescribing information of amisulpride and sulpiride, there are several known drug-drug interactions listed, including quinidine, verapamil, or imipramine. All these drugs are known inhibitors of OCT2 (9). OCT1 is proposed to be expressed on the apical membrane in the gut and, in addition to PEPT1, might play a role in sulpiride absorption (50). OCT1 is the major organic cation transporter in human hepatocytes (11,26). Amisulpride and sulpiride are not metabolized in the liver, but biliary excretion may account for up to 20% of their clearance (48,49). Hence, influx transport by SLC22 transporters in kidney and intestine (and potentially in liver) may be an important determinant of the plasma concentrations of both drugs.

OCT1 is highly genetic polymorphic. Five of the six common OCT1 alleles are known to lead to reduced and even absent activity in close to 9% of Caucasians. Here, we present in vitro data demonstrating a strong reduction of the OCT1-mediated uptake of amisulpride and sulpiride by the five most common OCT1 loss-of-function alleles (Fig. 5). This adds amisulpride and sulpiride to the relatively limited number of clinically relevant drugs shown to be substrates of OCT1 and whose pharmacokinetics may be affected by common OCT1 polymorphisms (10–13,51,52). The role of OCT1 in the transport of amisulpride and sulpiride through the blood-brain barrier and in other organs needs confirmation in in vivo models before concluding about the importance of OCT1 polymorphisms in the treatment of psychiatric patients with amisulpride or sulpiride.

An involvement of other influx transporters in the uptake of amisulpride and sulpiride into the brain cannot be excluded. Among the other members of the SLC22 family, SLC22A17 and SLC22A23 were reported to be expressed at the blood-brain barrier (35,53). However, the rat isoforms of both transporters are not able to transport typical organic cations (53) and data about the human isoforms is still missing. On the other hand, SLC22A16 (OCT6) is known to transport organic cations, but data on its expression at the blood-brain barrier is missing (18). Beyond the SLC22 family, PMAT, an organic cation transporter from the SLC29 family (SLC29A4), is also expressed in the brain and in immortalized rat BMEC cells (54) and may be involved in the uptake of amisulpride and sulpiride.

CONCLUSION

Our study shows that the psychotropic drugs amisulpride and sulpiride have only limited membrane permeability and require carrier-mediated influx transport to reach the brain. In addition, we could show carrier-mediated transport of amisulpride and sulpiride in an immortalized HBMEC cell line that was inhibitable by the OCT inhibitor MPP+. According to our in vitro data, amisulpride is a substrate of all organic cation transporters of the SLC22 family tested. Therefore, amisulpride may penetrate the blood-brain barrier via transport by OCT1, OCTN2, and possibly OCTN1, OCT2, and OCT3. In comparison, sulpiride may only use OCT1 and possibly OCT2 to penetrate the blood-brain barrier. OCT2 may also be important for the renal elimination of both drugs. Further work is required to clarify the clinical implications of these findings, with respect to therapy efficacy, genetic variation, and potential drug-drug interactions.