Abstract
The breast cancer resistance protein (BCRP/ABCG2) is an ATP-binding cassette drug efflux transporter that extrudes xenotoxins from cells in intestine, liver, mammary gland, and other organs, affecting the pharmacological and toxicological behavior of many compounds, including their secretion into the milk. The purpose of this study was to determine whether three widely used fluoroquinolone antibiotics (ciprofloxacin, ofloxacin, and norfloxacin) are substrates of Bcrp1/BCRP and to investigate the possible role of this transporter in the in vivo pharmacokinetic profile of these compounds and their secretion into the milk. Using polarized cell lines, we found that ciprofloxacin, ofloxacin, and norfloxacin are transported by mouse Bcrp1 and human BCRP. In vivo pharmacokinetic studies showed that the ciprofloxacin plasma concentration was more than 2-fold increased in Bcrp1–/– compared with wild-type mice (1.77 ± 0.73 versus 0.85 ± 0.39 μg/ml, p < 0.01) after oral administration of ciprofloxacin (10 mg/kg). The area under the plasma concentration-time curve in Bcrp1–/– mice was 1.5-fold higher than that in wild-type mice (48.63 ± 5.66 versus 33.10 ± 4.68 min · μg/ml, p < 0.05) after i.v. administration (10 mg/kg). The milk concentration and milk/plasma ratio of ciprofloxacin were 2-fold higher in wild-type than in Bcrp1–/– lactating mice. We conclude that Bcrp1 is one of the determinants for the bioavailability of fluoroquinolones and their secretion into the milk.
Quinolone antimicrobial drugs are widely used because of their broad spectrum and intense bactericidal activity. They are developed for oral and parenteral use in the treatment of bacterial diseases, including severe systemic infections (Brunner and Zeiler, 1988). Most quinolone antibacterial drugs are rapidly absorbed from the intestine, with a bioavailability of close to 90%, and then penetrate well into most body tissues and fluids. However, some fluoroquinolones have been reported to undergo efflux, which can explain the low bioavailability of some of them; for instance, the bioavailabilities of ciprofloxacin and norfloxacin are 50 to 80% and 30 to 40%, respectively (Sörgel et al., 1989; Lamp et al., 1992). In addition, at least 10% of i.v. administered ciprofloxacin is eliminated by intestinal secretion (Rohwedder et al., 1990). Only 1% of the dose is excreted into the bile (Parry et al., 1988). Some studies indicated that intestinal elimination of ciprofloxacin was not mediated by P-glycoprotein (ABCB1) (Griffiths et al., 1993; Cavet et al., 1997; Dautrey et al., 1999), one of the most important members of the ATP-binding cassette group of transporters, which is often involved in restricting the bioavailability of drugs. Accordingly, several groups (Lowes and Simmons, 2002; Michot et al., 2004) have recently shown that ciprofloxacin is not a substrate of this transporter or of MRP2 (ABCC2), another ABC transporter. However, the pharmacokinetics of ciprofloxacin was suggested to involve one or more active intestinal or hepatobiliary transport mechanisms in rats (Dautrey et al., 1999). Norfloxacin and ofloxacin have also been shown to be subject to active efflux (Cao et al., 1992; Rabbaa et al., 1996). Active secretory mechanisms common to all fluoroquinolones have been suggested (Griffiths et al., 1993, 1994), as well as competition between fluoroquinolones at transporter sites (Rabbaa et al., 1996). However, the precise mechanisms involved in the pharmacokinetics of ciprofloxacin and other fluoroquinolones remain to be clarified.
Several adverse effects have been reported with the use of quinolones, including nausea, diarrhea, vomiting, dizziness, headache, insomnia, and phototoxicity (Stahlmann, 2002). Arthropathy (condrotoxicity and tendopathies) in immature animals has been observed after administration of all quinolones tested so far. Although the significance of this undesirable side effect for humans remains unclear, the toxic potential has led to the contraindication of quinolones in children and women during pregnancy and lactation. Despite the extensive use of fluoroquinolones, little is known about their secretion into human breast milk. More accurate data on the passage of these compounds into breast milk will allow for the establishment of better recommendations for their use in lactating women (Dan et al., 1993).
Breast cancer resistance protein (BCRP/ABCG2), another member of the ABC family of transporters, is apically expressed and mediates the active and outward transport of a range of anticancer drugs, dietary compounds, food carcinogens, and antibiotics (van Herwaarden et al., 2003; Burger et al., 2004; Merino et al., 2005a). BCRP is found not only in tumor cells but also in a variety of normal tissues such as intestine, liver, brain, and mammary gland. Several in vivo studies indicated that Bcrp1 mediates the hepatobiliary excretion of its substrates and limits oral bioavailability and fetal penetration (Jonker et al., 2000; van Herwaarden et al., 2003). Induced expression of Bcrp1/BCRP in the lactating mammary gland and an important role of this ABC transporter in the active secretion of several xenobiotics into milk have been demonstrated (Jonker et al., 2005). Because fluoroquinolone breast milk concentrations are higher than serum concentrations (Giamarellou et al., 1989), we hypothesized that the pharmacokinetics and milk secretion of these fluoroquinolones could be regulated by BCRP in vivo.
In this study, we have demonstrated that three fluoroquinolones (ciprofloxacin, ofloxacin, and norfloxacin) are transported in vitro by Bcrp1/BCRP using polarized cell lines and that this transport results in altered pharmacokinetics and milk secretion of ciprofloxacin in Bcrp1–/– mice.
Materials and Methods
Animals. Mice were housed and handled according to institutional guidelines complying with European legislation. Animals used were male or lactating female Bcrp1–/– and wild-type mice, all of >99% FVB genetic background between 9 and 15 weeks of age. Animals were kept in a temperature-controlled environment with a 12-h light/12-h dark cycle and received a standard diet (Panlab, Barcelona, Spain) and water ad libitum.
Chemicals. Ofloxacin, norfloxacin, and oxfendazole were from Sigma Chemical (Steinheim, Germany); ciprofloxacin was from Fluka Chemie (Buchs, Switzerland); isoflurane (Isovet) was from Schering-Plough (Madrid, Spain); oxytocin (Oxiton) was from Ovejero (León, Spain); and Ko143 was prepared as described previously (Allen et al., 2002). All other compounds used were reagent grade.
Cells and Tissue Culture. The polarized canine kidney cell line MDCK-II was used in the transport assays. Human BCRP- and murine Bcrp1-transduced MDCK-II subclones were prepared as described previously (Jonker et al., 2000; Pavek et al., 2005). The cells were cultured in Dulbecco's modified Eagle's medium supplied with GlutaMAX (Life Technologies, Inc.) and supplemented with penicillin (50 units/ml), streptomycin (50 μg/ml), and 10% (v/v) fetal calf serum (MP Biomedicals) at 37°C in the presence of 5% CO2. The cells were trypsinized every 3 to 4 days for subculturing.
Transport Assays. Transport assays were carried out as described earlier (Merino et al., 2005a) with minor modifications. Cells were seeded on microporous membrane filters (3.0 μm pore size, 24 mm diameter; Transwell 3414; Costar, Corning, NY) at a density of 1.0 × 106 cells/well. Cells were grown for 3 days, and medium was replaced every day. Transepithelial resistance was measured in each well using a Millicell-ERS ohmmeter (Millipore, Bedford, MA); wells registering a resistance of ≥200 Ω, after correction for the resistance obtained in control blank wells, were used in the transport experiments. The measurement was repeated at the end of the experiment to check the tightness of the monolayer. Two hours before the start of the experiment, medium at both the apical and basolateral sides of the monolayer was replaced with 2 ml of Optimem medium (Life Technologies, Inc.), without serum and either with or without 1 μM Ko143. The experiment was started (t = 0) by replacing the medium in either the apical or basolateral compartment with fresh Optimem medium, either with or without 1 μM Ko143 and containing 10 μM fluoroquinolone. Cells were incubated at 37°C in 5% CO2 and 100-μl aliquots were taken at t = 2 and 4 h and stored at –20°C until the time of analysis. The appearance of fluoroquinolone in the opposite compartment was measured by HPLC as described below and is presented as the fraction of total fluoroquinolone added at the beginning of the experiment.
Accumulation Assays. In vitro accumulation assays were carried out as described previously (Pavek et al., 2005). In brief, 5 μM mitoxantrone was used as fluorescent substrate, and fluoroquinolones were used as inhibitors at different concentrations. Cells were cultured in 12-well plates (20 × 103 cells/well) in complete medium for 36 h to subconfluence. Medium was aspirated, and cells were preincubated in prewarmed Optimem medium with or without fluoroquinolone for 60 min before addition of 5 μM mitoxantrone. Accumulation of mitoxantrone was allowed for 1 h at 37°C. Cells were washed with ice-cold phosphate-buffered saline and trypsinized. Collected cells were sedimented and resuspended in phosphate-buffered saline with 2.5% fetal calf serum. Relative cellular accumulation of mitoxantrone was determined by flow cytometry using a FACSCalibur cytometer. Excitation and emission wavelengths for mitoxantrone were 488 and 650 nm, respectively. The fluorescence of the accumulated substrate in tested populations of at least 5000 cells was quantified from histogram plots using the median of fluorescence (MF). BCRP inhibition increases accumulation of mitoxantrone in Bcrp1/BCRP-transduced cells and thus increases MF. Possible background fluorescence of all tested fluoroquinolones was checked in appropriate channels, but the fluorescence was negligible in all cases. Flow cytometry data were processed and analyzed using WinMDI version 2.8 software.
Pharmacokinetic Experiments. For oral administration of ciprofloxacin (10 mg/kg), 10 μl of drug solution (appropriate concentration in 5% glucose and 1% NaOH 1 M)/g b.wt. were dosed by gavage into the stomach. For i.v. administration of ciprofloxacin (10 mg/kg), 5 μl of drug solution (appropriate concentration in 5% glucose and 1% NaOH 1 M)/g b.wt. were injected into the tails of mice lightly anesthestized with isoflurane. Blood was collected by orbital bleeding after anesthesia with isoflurane. Heparinized blood samples were centrifuged immediately at 3000g for 15 min, and plasma was collected and stored at –20°C until the time of HPLC analysis. Four to nine animals were used for each time point.
Milk Secretion Experiments. For milk experiments, pups ∼10 days old were separated from the mother ∼4 h before the experiment was started. To stimulate milk secretion, oxytocin (200 μl of 1 IU/ml solution) was administered s.c. to lactating dams. Ciprofloxacin (5 mg/kg) was injected into the tail vein, and milk was collected at t = 10 min from the mammary glands by gentle vacuum suction. Blood samples were collected by orbital bleeding under anesthesia with isoflurane. Milk and plasma were stored at –20°C until the time of HPLC analysis. Three animals were used for each group.
HPLC Analysis. The conditions for HPLC analysis of quinolones were modified on the basis of previously published methods (Idowu and Peggins, 2004; Marazuela and Moreno-Bondi, 2004). To each 100-μl aliquot of sample, 10 μl of a 12.5 μg/ml oxfendazole solution were added as an internal standard in a 1.5-ml reaction tube. The mixture was vortexed vigorously, and 300 μl of dichloromethane were added. Samples were shaken for 60 s, and the organic and water phases were separated by centrifugation at 5000g for 5 min and evaporated to dryness under a nitrogen stream. The samples were resuspended in 100 μl of methanol and injected into the HPLC system. Samples from the transport assays were not processed, and 100 μl of the culture media were directly injected into the HPLC system. The system consisted of a Waters 600 pump, a Waters 717plus Autosampler, and a Waters 2487 UV detector. Separation of the samples was performed on a reversed-phase column (AQUA C18 polar encapped, 5-μm particle size, 250 mm × 4.6 mm). The composition of the mobile phase was 25 mM orthophosphoric acid (pH 3.0)-acetonitrile (77:23). The flow rate of the mobile phase was set to 1.0 ml/min. UV absorbance was measured at 278 nm. Peak area ratios (fluoroquinolone/oxfendazole) were used for comparison with the standard curve. The integration was performed using the software Millennium32 (Waters). Standard samples in the appropriate drug-free matrix were prepared, yielding a concentration range from 0.08 up to 10 μg/ml.
Pharmacokinetic Calculations and Statistical Analysis. The two-sided unpaired Student's t test was used throughout to assess the statistical significance of differences between the two sets of data. Results are presented as means ± S.D. Differences were considered to be statistically significant when p < 0.05. The area under the plasma concentration-time curve from t = 0 to the last sampling point was calculated by the linear trapezoidal rule.
Results
In Vitro Transport of Fluoroquinolones. We used the polarized canine kidney cell line MDCK-II and its subclones transduced with murine Bcrp1 and human BCRP cDNAs to test the possible role of murine Bcrp1 and human BCRP in the in vitro transport of ciprofloxacin, ofloxacin, and norfloxacin. The parental and transduced cell lines were grown to confluent polarized monolayers on porous membrane filters, and vectorial transport of the fluoroquinolones (10 μM) across the monolayers was determined. In the MDCK-II parental cell line, apically and basolaterally directed translocation were similar (Figs. 1A, 2A, and 3A). In the Bcrp1-transduced MDCK-II cell lines, apically directed translocation was highly increased and basolaterally directed translocation drastically decreased for the three compounds (Figs. 1B, 2B, and 3B). When the selective Bcrp1 inhibitor Ko143 was used (Allen et al., 2002), the Bcrp1/BCRP-mediated transport was completely inhibited (Figs. 1, C and E, 2, C and E, and 3, C and E), resulting in a vectorial translocation pattern equal to that of the MDCK-II parental cell line. In the human BCRP-transduced cell line, the difference in the directional transport was lower than that in the murine Bcrp1-transduced cell line (Figs. 1D, 2D, and 3D). This may be due to an effectively lower level of the human protein in the cell line used, as also suggested by the nearly 5-fold smaller reduction in mitoxantrone accumulation in these cells (Fig. 4). Alternatively, human BCRP may have reduced transport efficiency for all of these substrates. These results thus show efficient transport of ciprofloxacin, ofloxacin, and norfloxacin by murine Bcrp1 and moderate transport by human BCRP in the cell lines used.
Mitoxantrone Accumulation Studies. To further characterize the interactions of the fluoroquinolones with Bcrp1/BCRP, the ability of these compounds to reverse the reduced mitoxantrone accumulation in murine Bcrp1- and BCRP-expressing cell lines was tested in flow cytometry experiments (Fig. 4). BCRP inhibition with Ko143 increases the accumulation of mitoxantrone in Bcrp1- and BCRP-transduced cells and thus increases the MF to levels similar to those in the parental cells. The addition of fluoroquinolones (ciprofloxacin, ofloxacin, and norfloxacin) did not affect the accumulation of mitoxantrone (5 μM) in Bcrp1- and BCRP-transduced cells, indicating that these compounds were not effective Bcrp1/BCRP inhibitors at 100 to 200 μM.
Plasma Pharmacokinetics of Ciprofloxacin in Bcrp1–/–andWild-Type Mice. To assess whether the in vitro Bcrp1-mediated transport of fluoroquinolones was also relevant in vivo, we studied ciprofloxacin pharmacokinetics in Bcrp1–/– and wild-type male mice. We determined the plasma concentration of ciprofloxacin after its oral and intravenous administration (10 mg/kg) in both types of mice (Fig. 5). The plasma concentration was more than 2-fold increased in Bcrp1–/– compared with wild-type mice (1.77 ± 0.73 versus 0.85 ± 0.39 μg/ml, p < 0.01) 15 min after oral administration (Fig. 5A). For i.v. administration (Fig. 5B), the area under the plasma concentration-time curve of the Bcrp1–/– was 1.5-fold higher compared with that in the wild-type mice (48.63 ± 5.66 versus 33.10 ± 4.68 min · μg/ml, p < 0.05). Bcrp1 thus appears to be both an important determinant for the oral availability and the elimination of ciprofloxacin.
Secretion of Ciprofloxacin into Milk in Bcrp1–/–and Wild-Type Mice. To test whether Bcrp1 plays a role in the secretion of fluoroquinolones into milk, ciprofloxacin (10 mg/kg) was administered i.v. to lactating Bcrp1–/– and wild-type females, and 10 min after administration, milk and blood were collected. The data obtained from the analysis of milk and plasma concentrations are shown in Fig. 6. The concentration of ciprofloxacin was 2-fold lower in the milk of Bcrp1–/– mice (2.19 ± 0.13 versus 4.44 ± 0.84 μg/ml, p < 0.01). Ciprofloxacin appears to be actively transported into the milk of mice by Bcrp1, since the milk-to-plasma ratio for wild-type mice was 2-fold higher compared with that for Bcrp1–/– lactating females (3.08 ± 0.92 versus 1.59 ± 0.08, p < 0.05).
Discussion
In this study, we show that three widely used fluoroquinolone antibiotics (ciprofloxacin, ofloxacin, and norfloxacin) are transported by Bcrp1/BCRP and that murine Bcrp1 affects the oral availability, pharmacokinetics, and milk secretion of ciprofloxacin. As seen with some other transported Bcrp1/BCRP substrates (A. H. Schinkel and P. Pavek, unpublished data), these fluoroquinolones are not effective inhibitors of Bcrp1/BCRP.
Our results show efficient in vitro transport of ciprofloxacin, ofloxacin, and norfloxacin by murine Bcrp1 and more moderate transport by human BCRP. These apparent differences in efficiency of transport between the murine and human homologues (Figs. 1, 2, and 3) may be due to a lower effective expression level of the human BCRP construct compared with Bcrp1 in the cell lines used, as the prototype BCRP substrate mitoxantrone was also less affected in the BCRP- than in the Bcrp1-expressing cell line (Fig. 4). However, we cannot exclude the possibility that there are differences in the affinity/selectivity of BCRP and Bcrp1 for substrates, as has been hypothesized by Mizuno et al. (2004).
An additional pathway for secretion of ciprofloxacin in Caco-2 cells, which is distinct from both P-glycoprotein and MRP2, has been suggested (Lowes and Simmons, 2002). BCRP is a good candidate responsible for the ciprofloxacin secretion in Caco-2 cells, since this cell line also expresses functional BCRP (Xia et al., 2005).
Our in vivo data suggest that Bcrp1 restricts ciprofloxacin oral bioavailability and contributes to overall ciprofloxacin elimination (Fig. 5). From the present study, we cannot exclude the presence of additional transporters for these fluoroquinolones. Basolateral transport mechanisms for ciprofloxacin in intestine and liver have been reported (Griffiths et al., 1994). However, the most obvious apical ABC transporters P-glycoprotein and MRP2 do not appear to transport ciprofloxacin (Lowes and Simmons, 2002; Michot et al., 2004). The fact that other fluoroquinolones, such as grepafloxacin, are transported by MDR1 and MRP2, demonstrates that multiple ABC transporters could contribute to the overall transepithelial secretion of fluoroquinolones (Lowes and Simmons, 2002). In any case, BCRP seems to be the secretory mechanism common to at least the three fluoroquinolones tested in this study.
Xenobiotic transport in the mammary gland can have tremendous clinical, toxicological, and nutritional implications. Passive and active transport processes are involved in the transfer of many drugs into breast milk. In vitro models of passive drug transfer identify drug protein binding, drug ionization, and fat partitioning as the critical determinants of passive drug transfer (Fleishaker et al., 1987; Begg and Atkinson, 1993). However, recently, an important role of Bcrp1 in the active secretion and concentration of several drugs and carcinogenic xenotoxins into milk has been established (Jonker et al., 2005; Merino et al., 2005a; van Herwaarden et al., 2005). Our data clearly demonstrate that the presence of Bcrp1 in the mammary gland is also an important factor in the transfer of fluoroquinolones into the milk, as indicated by the 2-fold higher milk/plasma ratio of wild-type compared with Bcrp1–/– mice (Fig. 6). Extrapolating our data to other species, we find that the ciprofloxacin milk/plasma ratio for wild-type mice we measured (3.08 ± 0.92) is close to the value obtained in other species, such as rabbits (3.61) and humans (4.67) (Gardner et al., 1992; Aramayona et al., 1996) and higher than the predicted milk/plasma ratio of 1.5 that physicochemical principles would suggest in human milk (Agatonovic-Kustrin et al., 2002). This value is similar to the ciprofloxacin milk/plasma ratio for Bcrp1–/– mice (1.59 ± 0.08). All this information suggests that our findings may be relevant to other species, including humans.
Milk secretion and pharmacokinetics of another antibiotic, nitrofurantoin, have been recently shown to be highly affected by Bcrp1 (Merino et al., 2005a). The results so far suggest that this transporter could be a very important factor in the study and evaluation of contamination of milk with antibiotics and residues and in the safety of antibiotic application during lactation.
The data presented here may already have other important biological and therapeutic significance. The expression and function of BCRP are variable, depending on different factors, such as coadministered drugs or natural products (Sugimoto et al., 2005), gender (Merino et al., 2005b), or genetic polymorphisms (Sparreboom et al., 2004). Fluoroquinolones are barely metabolized (Montay et al., 1994), so their potential to be affected in their pharmacokinetics by coadministered BCRP inhibitors may have significant consequences in vivo. The reciprocal situation, inhibition of BCRP by fluoroquinolones, is less probable, since our mitoxantrone accumulation results suggest that fluoroquinolones do not efficiently inhibit Bcrp1/BCRP in vitro (Fig. 4). However, we cannot completely exclude this possibility when substrates distinct from mitoxantrone would be used. In any case, all of these factors could modify the pharmacokinetics and milk secretion of fluoroquinolones and thus their local toxicity, their effect on the newborn, and even their bactericidal activity. Fluoroquinolone antibiotic activity is concentration-dependent and suboptimal concentrations due to variations (either inter- or intraindividual) in BCRP activity could be associated with insufficient eradication of microorganisms and potentially rapid emergence of resistance (Michot et el., 2004). Finally, elevated concentrations of the fluoroquinolones could increase the incidence of adverse effects not only in adults but also in infants.
Footnotes
-
This work was supported by Grant AGL2003-03888 (to A.A.) and by a Juan de la Cierva grant (to G.M.) from the Ministerio de Ciencia y Tecnología, Spain.
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.105.008219.
-
ABBREVIATIONS: MRP2, multidrug-resistance-associated protein 2; BCRP, breast cancer resistance protein; Ko143, 3-(6-isobutyl-9-methoxy-1,4-dioxo-1,2,3,4,6,7,12,12a-octahydropyrazino[1′,2′:1,6]pyrido[3,4-b]indol-3-yl]-propionic acid tert-butyl ester; MDCK, Madin-Darby canine kidney; HPLC, high performance liquid chromatography; MF, median of fluorescence.
- Received November 4, 2005.
- Accepted January 20, 2006.
- The American Society for Pharmacology and Experimental Therapeutics