Antiepileptic drugs (AEDs) do not effectively treat 30–40% of patients with epilepsy. Export of AEDs by P-glycoprotein (Pgp, ABCB1, or MDR1), which is overexpressed in the blood–brain barrier in drug-resistant patients, may be a mechanism for resistance to AEDs. For most recently approved AEDs, whether they are transported by Pgp is unknown. We investigated whether a new AED, lacosamide (LCM), is a substrate of human Pgp.
LLC-PK1 and MDCKII cells transfected with the human MDR1 gene were used to determine the substrate status of LCM in concentration equilibrium transport assays (CETAs). An equal concentration of drug was initially loaded in both the apical and basal chambers, and the concentration in both chambers was measured up to 4 h. The experiments were repeated in the presence of the Pgp inhibitors verapamil and tariquidar. Caco-2 assays were used to determine the intrinsic permeability and efflux ratio of LCM as well as its potential to inhibit digoxin, a Pgp substrate.
Lacosamide was transported by MDR1-transfected cells from basolateral to apical sides. The efflux of LCM could be completely blocked by verapamil or tariquidar. In Caco-2 assays, LCM showed high permeability without a significant efflux ratio; it did not inhibit digoxin, a Pgp substrate.
Although LCM is a substrate of Pgp in CETA, Caco-2 data demonstrated that passive diffusion should play a major role in the overall disposition of LCM. The critical role of Pgp should be addressed in vivo.
Although >20 antiepileptic drugs (AEDs) have been developed, epilepsy remains resistant to AED treatment in about 30–40% of patients, and the mechanisms of pharmacoresistance are still unclear. P-glycoprotein (Pgp), encoded by the ABCB1 or MDR1 gene, transports xenobiotics across the blood–brain barrier (BBB) from the basolateral to the apical direction. It is expressed on the luminal membrane of capillary endothelium in the brain and is overexpressed in epileptic foci (Kwan & Brodie, 2005; Loscher & Potschka, 2005). Expression of Pgp is greater in drug-resistant than in drug-responsive patients (Tishler et al., 1995; Dombrowski et al., 2001). Some studies indicated that several AEDs are substrates or inhibitors of Pgp (Weiss et al., 2003; Baltes et al., 2007a,b; Luna-Tortos et al., 2008; Zhang et al., 2010, 2011), including phenytoin, phenobarbital, topiramate, lamotrigine, levetiracetam, oxcarbazepine, eslicarbazepine acetate, and the drug metabolites carbamazepine-10,11-epoxide, and S-licarbazepine (Luna-Tortos et al., 2008, 2009; Zhang et al., 2010, 2011). These findings support the hypothesis that Pgp may play an important role in drug-resistant epilepsy (Kwan et al., 2011).
Continued efforts to develop new drugs to treat epilepsy have resulted in the approval in recent years of new generation AEDs, including lacosamide (LCM). Whether this drug is also subject to Pgp transport is largely unknown. No evidence has been reported to indicate the substrate status of LCM in animal models or epilepsy patients. Therefore, using the concentration equilibrium transport assay (CETA), we studied whether LCM is a substrate of Pgp, which may help in understanding the molecular basis of pharmacoresistance. In addition, the well-established Caco-2 model was used to assess the permeability of LCM as well as its role in the transport of a Pgp substrate.
Materials and Methods
LCM was obtained from 3B Pharmachem International Co., Ltd. (Wuhan, China) (>99% purity) for CETA assays in Hong Kong and from UCB Pharma (Brussels, Belgium) (100% purity) for Caco-2 assays in Belgium. Verapamil was from Alexis Biochemicals (San Diego, CA, U.S.A.). Tariquidar was kindly provided by Dr. Kenneth To (The Chinese University of Hong Kong, Hong Kong), who received it as a kind gift from Dr Susan Bates (Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, NIH). Verapamil and MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, a yellow tetrazole) were dissolved in water, and other drugs were dissolved in DMSO (dimethyl sulfoxide, <0.1% in final solution). Acetonitrile (Labscan Asia, Bangkok, Thailand), ethanol (TEDIA Company, Inc., Fairfield, OH, U.S.A.), and methanol (TEDIA Company, Inc.) were high pressure liquid chromatography (HPLC) grade. All other reagents were at least analytical grade.
Wild-type LLC-PK1 (LLC-WT) and MDCKII (MDCK-WT) cells, and LLC-MDR1 and MDCK-MDR1 cells transfected with the human MDR1 gene, were obtained from The Netherlands Cancer Institute and were cultured as described (Zhang et al., 2010, 2011). All the cell lines were used within 10 passages after receipt. Six-well transwells (0.4 μm, polycarbonate membrane, 24 mm insert; Transwell, Corning, NY, U.S.A.) were used for the transport studies. Cells (2 × 106 MDCK cells or 1.5 × 106 LLC cells) were seeded on the transwells as described previously (Zhang et al., 2005; Luna-Tortos et al., 2008), and were grown in the relevant medium (MDCK: Dulbecco's modified Eagle's medium [DMEM], 10% fetal bovine serum [FBS], 100 U/ml penicillin-streptomycin; LLC: Medium 199, 10% FBS, 100 U/ml penicillin-streptomycin) at 37°C with 5% CO2 for 5 days. The medium was changed every day.
Caco-2 cells were obtained from In Vitro Technologies (IVT), Inc. (Baltimore, MD, U.S.A.) Cultures were delivered in a ready-to-use 24-well format, complete with the appropriate medium and buffers. The cells were plated on Transwell membrane supports (0.33 cm2 surface area) and were delivered at room temperature. The transport experiments were conducted after a preincubation period of 24 h at ca. 37°C in a humidified atmosphere of 95% air and 5% CO2.
Cytotoxicity of LCM to LLC and MDCK cells
Potential cytotoxicity of AEDs was tested by MTT assay as described previously (Zhang et al., 2010, 2011). Cells at a density of 1.5 × 104 cells/well were seeded in 96 well plates and cultured for 48 h. After withdrawing the culture medium, 150 μl transport buffer with LCM was added to each well and incubated for 4 h at 37°C in 5% CO2. Then 20 μl of 5 mg/ml MTT was added to each well, followed by incubation for another 4 h at 37°C in 5% CO2. The buffer was removed and replaced with 200 μl DMSO in each well. The absorbance of each well in the plate was recorded at a wavelength of 590 nm on a microplate reader (Benchmark; BioRad, Hercules, CA, U.S.A.).
Expression level of Pgp in LLC and MDCK cell lines
To quantify MDR1 mRNA levels, RNA was isolated by Trizol (Invitrogen, Carlsbad, CA, U.S.A.). Reverse transcription was performed using a QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Real time polymerase chain reaction (PCR) was performed with FastStart Universal SYBR Green Master (Roche, Mannheim, Germany) in a Roche LightCycler 480, by using the following primers: MDR1, 5′-CCCATCATTGCAATAGCAGG-3′ and 5′-TGTTCAAACTTCTGCTCCTGA-3′; β-Actin, 5′-CCTCTATGCCAACACAGTGC-3′ and 5′-ACATCTGCTGGAAGGTGGAC-3′.
To detect the location and protein level of Pgp, immunofluorescent staining was performed as described previously (Zhang et al., 2011). Briefly, cells grew on cover slips for 48 h to about 80% density. After rinsing once with cold phosphate-buffered saline (PBS), the cells were fixed in 4% paraformaldehyde for 15 min. The slips were washed twice with cold PBS for 5 min each. Then the slips were blocked by 10% FBS in PBS at room temperature for 30 min, followed by diluted Pgp antibody JSB-1 (1:50; Enzo Life Science, Farmingdale, NY, U.S.A.) in PBS with 0.5% bovine serum albumin [BSA] at 4°C overnight. Cells were rinsed with PBS twice and incubated in secondary antibody (1:300, Alex 488-conjugated immunoglobulin G [IgG]; Invitrogen) in PBS with 0.5% BSA in the dark at room temperature for 30 min. Finally, cells were washed by PBS twice, and slides were mounted to observe them on a fluorescent microscope.
Rhodamine-123 uptake assay
To confirm the functional activity of P-gp in MDR1-transfected cell lines, we used the Rhodamine-123 (Rho123) uptake assay as described previously (Kimchi-Sarfaty et al., 2007; Zhang et al., 2011). Briefly, 5 × 105 cells were collected, rinsed once with PBS, and diluted in 1 ml DMEM. 1 μg/ml Rho123 was added to cell suspensions and incubated for 20 min at 37°C with 5% CO2. The cells were rinsed once with cold PBS, and resuspended in 500 μl cold PBS for flow cytometry assay. Flow cytometry analysis of Rho123 fluorescence was performed with the BD FACSAriaII (Becton Dickinson, San Jose, CA, U.S.A.) at 585 nm.
Validation of LLC and MDCK cell monolayers
The validation of LLC and MDCK cell monolayers was performed as described previously (Zhang et al., 2010, 2011). Briefly, the integrity of the monolayer was monitored by measuring the transepithelial electrical resistance (TEER) with an epithelial volt/ohm meter (World Precision Instruments, Sarasota, FL, U.S.A.). Only the monolayers with TEER > 100 Ω cm2 (subtracting the background value of a transwell) before and after the experiment were used for the transport assay. The integrity of monolayers was also verified by atenolol (paracellular marker) and propranolol (transcellular marker). Only the monolayers with integrity values comparable to published data were used (Crespi et al., 2000; Wang et al., 2008; Thiel-Demby et al., 2009).
Concentration equilibrium transport assay
The concentration equilibrium transport assay was performed as described in previous publications (Zhang et al., 2010, 2011). Briefly, the volumes on the apical and basolateral sides of cell monolayers were 2 and 2.7 ml, respectively. LCM was initially added to both sides of the monolayer at 5 μg/ml. Aliquots of 100 μl apical and 130 μl basolateral samples, which did not affect the hydrostatic pressure on the cell monolayers, were collected at various time points of drug exposure (30, 60, 90, 120, 180, and 240 min). In order to test the effect of Pgp inhibition on AED transport, the above AED transport assays were also conducted in the presence of verapamil (100 μm) or tariquidar (2 μm). The collected samples were stored at −20°C until analysis.
Transport experiment in Caco-2 assay
Before and after each experiment, the physical integrity of the monolayer was investigated by measuring the TEER. Shipping medium was read and recorded as background (blank). The TEER values were measured after the cells had equilibrated at room temperature for 20–30 min.
Transport experiments were conducted under nonsterile conditions. Before initiating an experiment, the basolateral transport medium (BM) was adjusted to pH 7.4 and the apical transport medium (AM) and the rinse medium were adjusted to pH 6.5. Prior to use, the cells were washed twice in rinse medium and equilibrated for 1 h at ca. 37°C in the incubator. In all experiments, the volume in the apical chamber was 100 μl and in the basolateral chamber 600 μl. Fifty microliter aliquots from the apical side and 100 μl from the basolateral compartment were taken for analysis. For determination of the total dosed radioactivity, 100 μl from each dosing solution was counted in duplicate. As an internal quality control of each cell batch, the monolayer integrity and the tightness of cell–cell junctions were assessed by investigating the apparent coefficients of permeation of mannitol (passive paracellular transport) and propranolol (passive transcellular transport) in the A > B (apical to basolateral) direction. The functional expression of Pgp in each cell batch was controlled by measuring the bidirectional transport of vinblastine in the absence and presence of the Pgp modulator verapamil. During the transport studies, the plates were maintained in the incubator at ca. 37°C in a humidified atmosphere of 5% CO2.
LCM was quantified by using high performance liquid chromatography with UV detection (HPLC/UV) as described in our previous study (Zhang et al., 2010, 2011). A Thermo Hypersil BDS C18 column (5 μm pores, 250 × 4.6 mm inner diameter) (Thermo Scientific, Waltham, MA, U.S.A.) was used. The HPLC/UV method was validated, and the interday and intraday root mean square deviations were calculated. The lowest limit of quantification was detected. For Caco-2 assays, 14C-LCM was used, and total radioactivity was determined by liquid scintillation counting.
In the CETA, the data are presented as the percentage of the drug loading concentration in either apical or basolateral chamber versus time, as described previously (Zhang et al., 2010, 2011). At various time intervals, differences in drug concentration between the two chambers of each well are compared, and differences in drug concentration between the two chambers in wild-type (WT) cells are compared with those from MDR1-transfected cells.
For MTT assays, percentage survival was calculated according to the formula: (mean of drug treatment OD − mean of blank OD)/(mean of control OD − mean of blank OD) × 100%. LCM was considered to be safe if cells exhibited 80% survival. For real time PCR, relative expression levels of MDR1 mRNA in LLC, LLC-MDR1, MDCK, and MDCK-MDR1 cells were scaled to the mean relative expression level of β-Actin, which was defined as 1. For Rhodamine-123 uptake, fluorescence values for LLC-MDR1, LLC-WT, and MDCK-MDR1 cells were scaled to the median fluorescence value of MDCK-WT cells, which was defined as 100.
Values are shown as means ± standard error of the mean (SEM). Significant differences between two groups or more than two groups were calculated by Student's t-test or one-way analysis of variance (ANOVA), respectively, with p <0.05 considered as significant.
The apparent coefficient of permeation was calculated using the following equation:
where volume is in ml and incubation time is in seconds.
A concentration of 5 μg/ml LCM was examined because it is within the range of steady-state plasma concentrations in human subjects (Cross & Curran, 2009). The cytotoxicity of 5 μg/ml LCM was tested by MTT assay. LCM was not toxic to the four cell lines for at least 240 min. Integrity of monolayers of the cell lines was verified by testing that the apparent permeability values (Papp) of propranolol and atenolol were within the range of 15 × 10−6 to 40 × 10−6 cm/s and 0.5 × 10−6 to 1.5 × 10−6 cm/s, respectively, which were comparable to those previously published (Crespi et al., 2000; Wang et al., 2008; Thiel-Demby et al., 2009).
The expression of human Pgp in MDR1 transfected and nontransfected cell lines was measured by real-time PCR and immunofluorescent staining. Pgp messenger RNA (mRNA) levels in MDR1-transfected LLC and MDCK cell lines were similar (p =0.43) and were significantly higher than in their respective wild-type cell lines (LLC: p =0.02; MDCK: p =0.01) (Fig. 1A). The expression and intracellular location of Pgp were detected by immunofluorescent staining, confirming that Pgp expression levels in MDR1-transfected cell lines were significantly higher than in wild-type cell lines (Fig. 1B). Exogenous Pgp protein was located largely at the membrane (Fig. 1B). The activity of human Pgp in MDR1-transfected and nontransfected cell lines was detected by Rho123 (a standard substrate of Pgp) uptake assay. The density of Rho123 in MDCK-MDR1 and LLC-MDR1 cells was lower than in their respective wild-type cells (Rho123 density: MDCK-WT = 100, MDCK-MDR1 = 29, LLC-WT = 92, LLC-MDR1 = 20) (Fig. 1C), which confirmed the functional activity of Pgp in MDR1-transfected cell lines.
In the concentration equilibrium transport assay, 5 μg/ml LCM was applied to LLC-WT and LLC-MDR1 cell lines. For LLC-MDR1 cells, the concentration of LCM in the apical side was significantly higher than that in the basolateral side from 60 min of drug exposure onward (Fig. 2). For LLC wild-type cells, there was no significant difference between apical and basolateral sides (Fig. 2). After adding the Pgp inhibitors tariquidar (2 μm) (Fig. 2) or verapamil (100 μm) (data not shown), the efflux of LCM from basolateral to apical sides was almost completely blocked, indicating that LCM was transported by Pgp.
To confirm the transport profile of LCM for Pgp, we performed the concentration equilibrium transport assay using 5 μg/ml LCM on MDCK-WT and MDCK-MDR1 cell lines. MDCK-MDR1 cells transported LCM to a similar degree as did LLC-MDR1 cells (Fig. 2). A significant difference in LCM concentrations between apical and basolateral sides of monolayers was detected from 90 min of drug exposure onward for MDCK-MDR1 cells, but not for MDCK-WT cells. When the Pgp inhibitor tariquidar (2 μm) was added, the transport of LCM was almost completely inhibited (Fig. 2), confirming that LCM is a substrate of Pgp.
To complete the investigations, LCM was also tested in another in vitro model where the impact of passive diffusion on the overall permeability is taken into account. To this aim, the transport of LCM across the Caco-2 monolayer expressing functional Pgp was assessed in both the A > B and the B > A direction at two concentrations: 125 μm (31 μg/mL) and 1 mm (250 μg/mL). Representative plots are given in Fig. 3. The cumulative transport data showed that LCM is transported across the Caco-2 monolayer in a linear time-dependent fashion in both directions.
Papp values for the transport of LCM are shown in Table 1. Mean Papp values for the transport of LCM at 125 μm (31 μg/ml) and 1 mm (250 μg/ml) in the A > B direction were 163.9 ± 14.6 and 156.9 ± 6.8 nm/s, respectively. In the B > A direction, the Papp values were determined to be 212.3 ± 23.8 and 217.7 ± 28.2 nm/s at 125 μm and 1 mm, respectively. The B > A / A > B transport ratios at 125 μm (31 μg/ml) and 1 mm (250 μg/ml) were found to be 1.3 and 1.4, respectively.
Table 1. Papp values of LCM in the Caco-2 model (mean ± SD, n = 3)
Papp values (nm/s)
A > B
B > A
125 μm (31 μg/ml)
164 ± 15
212 ± 24
1 mm (250 μg/ml)
157 ± 7
218 ± 28
To investigate the potential role of Pgp in the permeability of LCM, its transport (125 μm and 1 mm) in the absence and presence of verapamil (100 μm) was investigated in both directions. Papp values (A > B and B > A) were not significantly impacted by the presence of verapamil (data not shown).
To further explore the role of LCM in the transport of a Pgp substrate, the transport of 3H-digoxin over the course of 240 min in the absence and presence of LCM (over a concentration range from 10 μm to 3 mm) was determined. As shown in Fig. 4, LCM had no significant effect on 3H-digoxin transport at concentrations up to 3 mm.
Drug resistance in the treatment of epilepsy is a serious problem. The efflux transport of AEDs from the brain by Pgp might be involved (Kwan et al., 2011). There is in vitro and in vivo evidence that a number of established and newer AEDs are transported by Pgp (Zhang et al., 2012). Several drugs have recently been developed and marketed for the treatment of drug-resistant epilepsy, including LCM. To determine whether the efficacy of this novel agent might also be limited by efflux transport by Pgp, we set out to determine its substrate status using a recently established in vitro assay system (Luna-Tortos et al., 2008).
After validating the CETA assay—by demonstrating that cell monolayers were intact, that MDR1 cell lines expressed functional Pgp (Fig. 1), and that LCM was stable and nontoxic—we demonstrated that LCM at a clinically relevant concentration was transported by Pgp in both LLC-MDR1 and MDCK-MDR1 cells, and that the transport could be blocked by the Pgp inhibitors verapamil and tariquidar. There was no transport by wild-type cells, indicating that LCM is a substrate of human Pgp, as has been reported for some of the AEDs (Luna-Tortos et al., 2008; Zhang et al., 2011).
However, in the well-established Caco-2 assay (Artursson et al., 2001; van Breemen & Li, 2005), LCM showed a high permeability in both directions with efflux ratios <1.5, suggesting that Pgp does not play a major role in the permeability of LCM. According to U.S. Food and Drug Administration (FDA) guidance, LCM would not be considered as a Pgp substrate since its efflux ratio is lower than 2 (Food & Drug Administration, 2012). In addition, using the same model, LCM at concentrations up to 3 mm showed no potential to modulate the transport of digoxin, a well-known Pgp substrate, demonstrating at least the low affinity of LCM for Pgp (Km > 3 mm). This finding was confirmed in a clinical drug–drug interaction study (SP844) that demonstrated the lack of effect of LCM on the pharmacokinetics of digoxin (Thomas et al., 2009).
Although we cannot rule out that LCM is a Pgp substrate using the classical Caco-2 model, the results in this model highly suggest that passive diffusion is the major component in the overall disposition of LCM and that the active transport mediated by Pgp should play a minor role, if any. Although the outcomes of CETA and Caco-2 assays differ, we do not believe they are contradictory. Indeed, the objectives are different. CETA mainly aims to identify whether a compound is a Pgp substrate, whereas Caco-2 investigates the mechanisms (active efflux vs. passive diffusion) driving membrane permeability. This explains why a modest Pgp substrate having a high permeability will not be identified as a Pgp substrate in the classical Caco-2 model, since in that case the passive diffusion component will predominate over the active transport, resulting in the absence of a significant efflux ratio.
However, in vivo studies should be performed to clearly identify the role of Pgp compared to passive diffusion in the overall disposition of LCM. If the disposition of LCM is shown to be limited by Pgp, preclinical and clinical studies that coadminister a Pgp inhibitor with LCM to improve its efficacy may be warranted.
Cell lines were provided by Prof. P. Borst, The Netherlands Cancer Institute. Financial support was provided by UCB Pharma. The authors would like to thank K. Hansen and U. Scharfenecker for their contributions in the Caco-2 assay.
Hugues Chanteux is employed by UCB Pharma, which sells lacosamide. Larry Baum and Patrick Kwan received a research contract from UCB Pharma to conduct this study. The remaining authors have no conflicts of interest. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.