Abstract
Seliciclib, a cyclin-dependent kinase inhibitor, is a promising candidate to treat a variety of cancers. Pharmacokinetic studies have shown high oral bioavailability but limited brain exposure to the drug. This study shows that seliciclib is a high-affinity substrate of ATP-binding cassette B1 (ABCB1) because it activates the ATPase activity of the transporter with an EC50 of 4.2 μM and shows vectorial transport in MDCKII-MDR1 cells, yielding an efflux ratio of 8. This interaction may be behind the drug's limited penetration of the blood-brain barrier. ABCB1 overexpression, on the other hand, does not confer resistance to the drug in the models tested. These findings should be considered when treatment strategies using seliciclib are designed.
Introduction
Seliciclib (CYC202, R-roscovitine [2-(1-ethyl-2-hydroxyethylamino)-6-benzylamino-9-isopropylpurine]), a 2,6,9-trisubstituted purine analog, is a second-generation cyclin-dependent kinase (CDK) inhibitor (Meijer et al., 2006). It arrests cellular proliferation and induces apoptosis through molecular interactions with the heterodimers of CDKs and cyclins. It is a potent inhibitor of the human CDK2/cyclin E, CDK1/cyclin B, CDK7/cyclin H, and CDK9/cyclin T1 (Meijer et al., 2006; Okyar and Lévi, 2008). Seliciclib binds to the ATP binding site of the respective kinases in a competitive fashion as shown in structural and kinetic studies (De Azevedo et al., 1997). A few other enzymes such as calmodulin-dependent kinase isoforms, casein kinase 1α, casein kinase 1δ/ε, dual-specificity tyrosine-(Y)-phosphorylation-regulated kinase 1A, elk-related tyrosine kinase 2, extracellular signal-regulated kinase 1, extracellular signal-regulated kinase 2, focal adhesion kinase, and interleukin-1 receptor-associated kinase 4 were also inhibited at micromolar concentrations (Meijer et al., 2006). In addition, seliciclib also bound pyridoxal kinase and reduced the level of pyridoxal phosphate in human erythrocytes (Bach et al., 2005). Seliciclib has displayed activity against human non–small-cell lung, colon, breast, and prostate cancer cell lines in xenografted mouse models as well as against mouse Glasgow osteosarcoma (Iurisci et al., 2006; Meijer et al., 2006). Phase I and II clinical trials have shown adequate drug tolerability (Benson et al., 2007) and recently revealed high activity in patients with chronic lymphocytic leukemia or nasopharyngeal carcinoma (Weingrill et al., 2007; Hsieh et al., 2009).
Limited information exists on the resistance of cancer cells to seliciclib. One study in chronic lymphocytic leukemia in vitro models showed that CD40 stimulation up-regulated antiapoptotic Bcl-xL, A1/Bfl-1, and Mcl-1 proteins and afforded resistance to seliciclib among several agents in various pharmacologic classes (Hallaert et al., 2008).
Although seliciclib has been shown to inhibit ABCB1-mediated transport of rhodamine 123 (Bachmeier and Miller, 2005), the nature of the seliciclib-ABCB1 interaction, namely substrate versus inhibitor, is unknown as yet. The potential role of the MDR transporter ABCB1 in seliciclib resistance has not been evaluated either. A further indication of the possible role of ABCB1 in seliciclib transport stems from the observation of its limited brain uptake, which was estimated to be 25% in adult rats (Vita et al., 2005). In this study, we show that seliciclib is a selective substrate of ABCB1 and discuss how this seliciclib-ABCB1 interaction may affect seliciclib disposition.
Materials and Methods
Chemicals.
Seliciclib (CYC202, R-roscovitine [2-(1-ethyl-2-hydroxyethylamino)-6-benzylamino-9-isopropylpurine]) powder was kindly provided by the Institute de Chimie Organique, Université René Descartes, Paris, France (Hervé Galons). [(2R)-Anti-5-{3-[4-(10,11-difluoromethanodibenzo-suber-5-yl)piperazin-1-yl]-2-hydroxypropoxy}quinoline (LY335979) was synthesized as described previously (Barnett et al., 2004). [3H]NMQ was from Dr. Csaba Tömböly (Biological Research Center, Budapest, Hungary), Ko134 was from Solvo Biotechnology (Szeged, Hungary). Advanced RPMI 1640 (Invitrogen, Carlsbad, CA) was from Csertex Ltd. (Budapest, Hungary). Fetal bovine serum (Lonza, Basel, Switzerland), Dulbecco's modified Eagle's medium (Lonza), and penicillin-streptomycin (Lonza) were purchased from Biocenter Kft. (Szeged, Hungary). The mouse anti-ABCB1 monoclonal antibody C-219 (Abcam) was purchased from Biomarker Kft., and the anti-mouse IgG-horseradish peroxidase (HRP) secondary antibody, a HRP-conjugated species-specific whole antibody was from Sigma Hungary (Budapest, Hungary). Western Lightning Plus-ECL (PerkinElmer Life and Analytical Sciences, Waltham, MA) was from Per-Form Hungary Kft. (Budapest, Hungary). Calcein AM (Invitrogen) was purchased from Invitrogen Hungary (Budapest, Hungary), verapamil, Hoechst 33342, and other chemicals were from Sigma Hungary (Budapest, Hungary).
Cell Lines.
The chronic myeloid leukemia cell line, K562, and its ABCB1-overexpressing variant K562-MDR were received as kind gifts from Professor Balazs Sarkadi (National Blood Transfusion Service, Budapest, Hungary); MDCKII-MDR1, PLB985-BCRP (Kis et al., 2009), and parental cells were kindly provided by Dr. Katalin Német (National Blood Transfusion Service). Cells were maintained in Advanced RPMI 1640 except MDCKII and MDCKII-MDR1 cells, which were in Dulbecco's modified Eagle's medium supplemented with 1 g/l glucose and 1% nonessential amino acids. All media were supplemented with 10% (v/v) heat-inactivated fetal bovine serum, 2 mM l-glutamine, and 100 μg/ml penicillin-streptomycin and were grown under standard conditions (5% CO2 and 37°C).
ATPase Activity.
ABC transporter-overexpressing membrane preparations show vanadate-sensitive ATPase activity that is modulated by interacting compounds. ATPase activity was measured as described previously (von Richter et al., 2009). In brief, the rate of ATP hydrolysis was determined by measuring the liberation of inorganic phosphate using PREDEASY ATPase kits for ABCB1, ABCC1, ABCC2, and ABCG2-HAM from SOLVO Biotechnology (Budapest, Hungary) and used according to the manufacturer's instructions. Membrane vesicles were incubated with various concentrations of test drugs with or without 1.2 mM sodium orthovanadate. ATPase activities were determined as the difference of inorganic phosphate liberation measured in the presence and absence of 1.2 mM sodium orthovanadate, an inhibitor of ABC efflux pumps. Results are presented as vanadate-sensitive ATPase activities.
Vesicular Transport Assay.
The interaction of seliciclib with the transporter was detected as the modulation of the initial rate of NMQ transport into the membrane vesicles. The vesicular transport assay was performed using a PREDIVEZ kit for human ABCB1 according to the manufacturer's recommendations. In brief, membrane fractions containing inside-out membrane vesicles were incubated in a 96-well plate in the presence or absence of ATP using [3H]NMQ as the probe substrate. The transport was stopped by addition of cold washing buffer and consecutive rapid filtration through Millipore B glass fiber filters of a 96-well filter plate (Millipore Corporation, Billerica, MA). Filters were washed five times with 200 μl of ice-cold wash buffer and dried, and the retained radioactivity was measured in scintillation cocktail (Packard UltimaGold; PerkinElmer Life and Analytical Sciences) using a Wallac MicroBeta TriLux liquid scintillation analyzer.
Hoechst Assay.
Hoechst 33342 intercalates DNA, yielding a fluorescent product that can be detected. The presence of ABCG2 in the cell membrane strongly reduces Hoechst 33342 accumulation. Inhibitors of ABCG2 produce an increased rate of accumulation. The Hoechst assay was performed as described earlier (Kis et al., 2009). In brief, accumulation of Hoechst 33342 dye was measured in a fluorometer (Fluoroskan Ascent Type 374; Thermo Labsystems, Helsinki, Finland) at 350 nm (excitation) and 460 nm (emission) by using PLB985-BCRP cells. The cells were preincubated at 37°C in 1× Hanks' balanced salt solution with drugs for 30 min. The Hoechst 33342 dye was added in 50 μl at a final concentration of 12.5 μM. The fluorescence intensities were recorded for 15 min. The positive control measurements to determine 100% inhibition were performed in the presence of 400 nM Ko134, a specific ABCG2 inhibitor (Allen et al., 2002).
Calcein Assay.
Calcein AM penetrates the plasma membrane by passive diffusion. Intracellularly calcein AM is hydrolyzed by endogenous esterases, yielding a fluorescent product, calcein, which can be detected. The presence of ABCB1 in the cell membrane strongly reduces calcein accumulation. Inhibitors of ABCB1 produce an increased rate of accumulation. The calcein assay was performed as described earlier (von Richter et al., 2009). In brief, accumulation of the calcein dye was measured in a fluorimeter (Fluoroskan Ascent Type 374) at 485 nm (excitation) and 538 nm (emission) by using K562-MDR cells. Cells (80,000/well) were incubated in 100 μl of Hanks' balanced salt solution in the presence of the test compound or positive control for 15 min. After the incubation, calcein AM in 100 μl of HBSS was added at a final concentration of 0.25 μM. Fluorescence intensities were recorded for 8 min. The positive control measurements to determine 100% inhibition were obtained in the presence of 60 μM verapamil.
Western Blotting.
The proteins were separated using a 10% polyacrylamide gel and transferred to polyvinylidene difluoride membrane (Immobilon-P; Millipore Corporation) at 350 mA in a transfer buffer composed of 25 mM Tris, 192 mM glycine, and 15% (v/v) methanol, pH 8.3. The membrane was treated with blocking buffer (5% nonfat dry milk powder and 0.5% bovine serum albumin in phosphate-buffered saline with 0.05% Tween 20) for 2 h at room temperature. The membrane was then incubated with the primary antibody, a mouse anti-ABCB1 monoclonal antibody C-219, diluted 1:3000 in blocking buffer for 2 h at room temperature. The membrane was washed three times for 10 min each with phosphate-buffered saline-0.05% Tween 20 at room temperature. It was then incubated with the secondary antibody, anti-mouse IgG-HRP, a horseradish peroxidase-conjugated species-specific whole antibody diluted 1:5000 in blocking buffer for 1 h at room temperature. The membrane was subsequently washed as described above, and immunoreactive bands were visualized with enhanced chemiluminescence.
MDCKII Monolayer Assay.
Transport assays across MDCKII-wt and MDCKII-MDR1 cells were performed described previously (von Richter et al., 2009). Cells were seeded on Millicell 24 (Millipore, Carrigtwohil, Ireland) devices according to the manufacturer's instructions. Seliciclib (5 μM) was added without the ABCB1 inhibitor, LY335979 (1 μM), to the medium at either the basolateral or apical compartment. Samples were taken from the receptor chamber at 15, 30, 60, and 120 min. Concentrations of seliciclib were determined using an Agilent 1100 series high-performance liquid chromatograph equipped with a mass selective detector Quad VL System (Agilent, Waldbronn, Germany). Samples from the 60-min point were used for the apparent permeability coefficient (Papp) calculations.
Cytotoxicity Assay.
Cytotoxicity assays were performed by seeding HL60 (50,000 cells/well), HL60-MDR1 (50,000 cells/well), K562 (50,000 cells/well), K562-MDR (50,000 cells/well), MDCKII-wt (1000 cells/well), and MDCKII-MDR1 (1000 cells/well) in 96-well plates containing the culture medium (200 μl/well). After 24 h, drugs were prediluted in medium and added to the cells at different concentrations as shown in the figures. The cells were further incubated with the drug in a humidified tissue culture chamber (37°C and 5% CO2) for 96 h. Surviving cells were detected by the MTS method (http://www.promega.com). IC50 values were calculated from dose-response curves (i.e., cell survival versus drug concentration) obtained in triplicate experiments.
Data Analysis.
All assays were run in duplicate unless indicated otherwise. The calcein and the Hoechst assays were analyzed using the slope of the curve determined without inhibitors (Rbase), the slope of the curve in the presence of the inhibitor (Rmax), and the slope of the curve determined for any drug at the given drug concentration (Rdrug). The inhibition (percentage) of dye extrusion can be represented by the following formula: IC50 values were derived from these curves.
For calculation of the Papp the following equation was used: where dQ is the amount of test and dT is the incubation time. C0 is the initial concentration of the compound in the donor compartment, and A is the membrane surface area in square centimeters (standard: 0.7). The efflux ratio is given as the Papp B-A/Papp A-B apparent permeability ratio, where A is apical and B is basolateral.
For curve fitting and IC50 calculations GraphPad Prism 4.0 software (GraphPad Software Inc., San Diego, CA) was used. Statistical analysis was performed using an unpaired t test.
Results
To test interactions of seliciclib with efflux transporters present in the blood-brain barrier, ATPase assays were performed using Sf9 membranes overexpressing ABCB1, ABCC1, ABCC2, and ABCG2. Seliciclib activated the vanadate-sensitive ATPase activity of ABCB1 with an EC50 value of 4.2 μM (Table 1) to the level observed in the presence of verapamil, the positive control (Fig. 1A). ABCG2 ATPase was not activated by seliciclib (Fig. 1D). However, the activated ABCG2 as well as the basal vanadate-sensitive ABCG2 ATPase was inhibited by seliciclib (Fig. 1D), albeit at a suprapharmacological concentration range (Table 1). ABCC1 only showed inhibition at very high concentrations (Fig. 1B), whereas ABCC2 did not interact at all (Fig. 1C).
To confirm interactions with ABCB1 a vesicular transport assay using NMQ as a probe and a calcein assay were performed. Both assays showed interaction (Fig. 2, A and B) with IC50 values of 35.5 and 11.5 μM, respectively (Table 1). ABCG2 inhibition by seliciclib was also confirmed in a Hoechst assay (Fig. 2C) with an IC50 of 38 μM (Table 1).
Activation of ABCB1 ATPase by seliciclib indicated that the drug is a transported substrate of ABCB1. To confirm the transport, MDCKII-MDR1 cells were used (Fig. 3). MDCKII-MDR1 cells greatly overexpress human MDR1 (Fig. 3A). Basal calcein fluorescence of MDCKII-MDR1 cells was much lower than basal calcein fluorescence of control cells but reached approximately the same level in the presence of LY335979 (Fig. 3B), an ABCB1-specific inhibitor (Dantzig et al., 2003). In addition, in MDCKII-MDR1 cells unlike in control cells (Fig. 3C), permeability of seliciclib was much greater in the basolateral to apical direction than in the apical to basolateral direction, resulting in an efflux ratio of approximately 8 in MDCKII-MDR1 cells. In the presence of LY335979, the observed efflux ratio in the MDCKII-MDR1 cells was 1.2 (Fig. 3D).
We tested whether ABCB1 overexpression would result in resistance to seliciclib using HL60-MDR1, K562-MDR, MDCKII-MDR1, and control cells. We found that overexpression of ABCB1 did not confer resistance to seliciclib because IC50 values in cytotoxicity tests were not significantly different for MDCKII-wt and MDCKII-MDR1 cells (4.9 ± 1.3 versus 7.1 ± 1.6 μM, respectively), and addition of LY335979 did not affect susceptibility to seliciclib of either cell line (Fig. 4A). In contrast, a marked difference was observed in susceptibility to doxorubicin (Fig. 4B) and paclitaxel (Fig. 4C). No statistically significant difference was observed in IC50 values for seliciclib in K562 and K562-MDR (45.9 ± 5.94 versus 47.1 ± 33.5 μM, respectively) or in HL60 and HL60-MDR1 (12.6 ± 4.6 versus 17.5 ± 7.5 μM, respectively) cells.
Discussion
Although seliciclib was taken up in the brain of rat pups via simple equilibrium diffusion, the brain/plasma AUC ratio was only 25% in adult rat brain (Vita et al., 2005; Sallam et al., 2008). This finding is in line with prior published data on the maturation of the blood-brain barrier, in which ABC transporters play an essential role. In the current study, we tested whether efflux transporters, thought to limit brain exposure to drugs would play a role in limiting seliciclib brain exposure. It is commonly accepted that ABCB1/Abcb1a and ABCG2/Abcg2 are expressed in the luminal membrane of brain microcapillary endothelial cells (Roberts et al., 2008) and limit penetration of substrate drugs (Enokizono et al., 2007). Barrier function of ABCC1/Abcc1 is controversial because it was shown to localize in the luminal membrane in humans (Nies et al., 2004) and more recently was found in the luminal membrane in rat brain microcapillary endothelial cells (Roberts et al., 2008). On the contrary, it is generally accepted that Abcc1 is expressed in the basolateral membrane of murine choroid plexus epithelial cells (Roberts et al., 2008) and protects brain from toxic stimuli (Wijnholds et al., 2000). The function of ABCC2/Abcc2 in the blood-brain barrier is even more controversial. The conclusion from multiple studies suggests that ABCC2/Abcc2 may not play a substantial role in the blood-brain barrier under normal conditions but may get up-regulated and, thus, limit brain penetration of substrate drugs under pathological conditions (Hoffmann et al., 2006).
Our data clearly show that seliciclib is a selective substrate of ABCB1 because it activates the ABCB1 ATPase (Fig. 1A), and it shows an ABCB1-dependent vectorial transport in the MDCKII-MDR1 cells (Fig. 3, C and D). Seliciclib in phase I clinical trials reached a plasma concentration of approximately 10 μM, 90% of which was protein-bound (Benson et al., 2007). This value is approximately 1 log below the EC50 of 4.2 μM for the seliciclib-ABCB1 interaction, a concentration range at which transporters exert significant effects on transcellular permeability of their substrates (Shirasaka et al., 2008). Note that the AUCbrain/AUCplasma value of seliciclib was approximately 1 in the 12-day-old rat pups, whereas it was approximately 0.2 in adult rats (Vita et al., 2005; Iurisci et al., 2006). The blood-brain barrier matures at 3 to 4 weeks postnatally, and it was hypothesized that an immature blood-brain barrier in pups might account for this difference (Sallam et al., 2008). No data comparing Abcb1a expression in brain microvascular endothelial cells in rat pups and in adult rats are available. In cerebellum, however, Abcb1a expression is lower at 11 days than in adult animals (de Zwart et al., 2008). Therefore, it is conceivable that Abcb1a is responsible for lower brain exposure in adult rats. Data demonstrating that seliciclib inhibits Abcb1-mediated rhodamine 123 transport in bovine brain microvascular endothelial cells substantiate this hypothesis (Bachmeier and Miller, 2005).
ABCB1 is also responsible for multidrug resistance to substrate drugs (Szakács et al., 2004). We have not observed ABCB1-mediated resistance to seliciclib-induced cell killing in three different cell lines, K562-MDR1 (Ye et al., 2009), HL-60-MDR1 (Rohlff et al., 1993), and MDCKII-MDR1 (Fig. 3, B–D), which display significant ABCB1-dependent resistance to substrate drugs. Seliciclib is a moderate to high passive permeability compound based on the permeability data of approximately 5 × 10−5 cm/s (Fig. 3D), high bioavailability (Benson et al., 2007), the log P value of 3.244 (Meijer et al., 2006), and the reasonable agreement between the membrane and cellular IC50 values (35.5 versus 11.5 μM). Therefore, the lack of a protective effect of ABCB1 against cell death inflicted by seliciclib may simply be explained by the passive permeability of the drug. In MCF7 cells, seliciclib displays greater cytotoxicity in the ABCB1-overexpressing cells than in the control MCF7 line (Cappellini et al., 2009). This class of compounds has been termed MDR1-inverse because ABCB1 sensitizes cells to cytotoxicity by these compounds (Szakács et al., 2004). Thus, ABCB1 may play a dual role in HL60-MDR1, K562-MDR, and MDCKII-MDR1 cells by reducing intracellular drug concentration and at the same time sensitizing cells to seliciclib. The two opposing effects may effectively cancel each other out in the cell lines used in this study.
Seliciclib also interacts with ABCG2 (Figs. 1 and 2). The interaction does not result in the activation of the ATPase; therefore, seliciclib is a likely inhibitor of the transporter. The observed IC50 values correlate with the low affinity observed for inhibition of ABCG2-mediated hematoporphyrin transport (An et al., 2009). The clinical free drug concentrations under current schedules (Weingrill et al., 2007) are way below the IC50 values for seliciclib-mediated ABCG2 inhibition, rendering this interaction as unlikely to be clinically relevant.
In conclusion, seliciclib is a high-affinity, selective ABCB1 substrate. This interaction is likely to affect disposition of the drug. ABCB1 overexpression, on the other hand, does not confer resistance to seliciclib, making the drug a favorable candidate to treat ABCB1 transporter-overexpressing tumors.
Acknowledgments.
We acknowledge the expert help of Judit Janossy, Ph.D., Katalin Jakab, M.D., and Timea Rosta, M.Sc., in reviewing and preparing the article.
Footnotes
The work was supported by Hungarian National Office for Research and Technology [Grant XTTPSRT1]; and the European Community [Grants LSHB-CT-2004-005137, LSHB-CT-2006-037499, LSHB-CT-2006-518246, LSHGCT-2006-037543, LSSG-CT-2006-037654]. A.O. was a recipient of a postdoctoral fellowship from the Scientific and Technological Research Council of Turkey.
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
doi:10.1124/dmd.110.032805.
-
ABBREVIATIONS:
- CDK
- cyclin-dependent kinase
- ABC
- ATP-binding cassette transporter
- MDR
- multidrug resistance protein
- LY335979
- [(2R)-anti-5-{3-[4-(10,11-difluoromethanodibenzo-suber-5-yl)piperazin-1-yl]-2-hydroxypropoxy}quinoline
- NMQ
- N-methylquinidine
- calcein AM
- calcein acetoxymethylester
- MDCKII
- Madin-Darby canine kidney strain II
- BCRP
- breast cancer resistance protein
- HRP
- horseradish peroxidase
- Papp
- apparent permeability coefficient
- MDCKII-MDR1
- Madin-Darby canine kidney strain II-multidrug resistance protein 1
- PLB985-BCRP
- The human myelomonoblastic leukemia cell line-breast cancer resistance protein.
- Received February 16, 2010.
- Accepted August 10, 2010.
- Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics
References
DMD articles become freely available 12 months after publication, and remain freely available for 5 years.Non-open access articles that fall outside this five year window are available only to institutional subscribers and current ASPET members, or through the article purchase feature at the bottom of the page.
|