Abstract
Organic anion–transporting polypeptides (OATPs) mediate the liver uptake and hence plasma clearance of a broad range of drugs. For rosuvastatin, a cholesterol-lowering drug and OATP1A/1B substrate, the liver represents both its main therapeutic target and its primary clearance organ. Here we studied the impact of Oatp1a/1b uptake transporters on the pharmacokinetics of rosuvastatin using wild-type and Oatp1a/1b-null mice. After oral administration (15 mg/kg), intestinal absorption of rosuvastatin was not impaired in Oatp1a/1b-null mice, but systemic exposure (area under the curve) was 8-fold higher in these mice compared with wild-type. Although liver exposure was comparable between the two mouse strains (despite the increased blood exposure), the liver-to-blood ratios were markedly decreased (>10-fold) in the absence of Oatp1a/1b transporters. After intravenous administration (5 mg/kg), systemic exposure was 3-fold higher in Oatp1a/1b-null mice than in the wild-type mice. Liver, small intestinal, and kidney exposure were slightly, but not significantly, increased in Oatp1a/1b-null mice. The biliary excretion of rosuvastatin was very fast, with 60% of the dose eliminated within 15 minutes after intravenous administration, and also not significantly altered in Oatp1a/1b-null mice. Rosuvastatin renal clearance, although still minor, was increased ∼15-fold in Oatp1a/1b-null males, suggesting a role of Oatp1a1 in the renal reabsorption of rosuvastatin. Absence of Oatp1a/1b uptake transporters increases the systemic exposure of rosuvastatin by reducing its hepatic extraction ratio. However, liver concentrations are not significantly affected, most likely due to the compensatory activity of high-capacity, low-affinity alternative uptake transporters at higher systemic rosuvastatin levels and the absence of efficient alternative rosuvastatin clearance mechanisms.
Introduction
Organic anion–transporting polypeptides [human: OATP, gene: SLCO (solute carrier organic-anion); rodents: Oatp, gene: Slco] form a superfamily of transmembrane transporters that mediate the cellular uptake of structurally diverse endogenous and exogenous compounds (Hagenbuch and Meier, 2004). With wide and overlapping substrate specificities and expressed in tissues important for pharmacokinetics (liver, small intestine, and kidney), the OATP1A/1B subfamilies are thought to have an important role in drug absorption, distribution, and elimination. Based on tissue distribution and amino acid sequence, there are no straightforward orthologs between mouse and human members of these subfamilies. OATP1A/1B subfamilies contain three human members (OATP1A2, -1B1, and -1B3) but at least five mouse members (Oatp1a1, -1a4, -1a5, -1a6, and -1b2) (Hagenbuch and Meier, 2003). Human OATP1B1 and OATP1B3 are predominantly expressed in the hepatic sinusoidal membrane and are thought to play a key role in the hepatic uptake and plasma clearance of drugs. Several low-activity polymorphic variants of human OATP1B1 have been associated with decreased transport activity and increased plasma levels, and hence with toxicity of statins (cholesterol-lowering drugs) (reviewed in Kalliokoski and Niemi, 2009). In addition, a previous study revealed that Rotor syndrome is caused by a complete simultaneous deficiency in the OATP1B1 and OATP1B3 genes (van de Steeg et al., 2012). Although Rotor syndrome is very rare (∼1 in 106 individuals), individuals with complete deficiencies in either OATP1B1 or OATP1B3 alone likely exist at a much higher frequency in various populations.
Rosuvastatin is one of the most efficacious 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins), and widely used in the treatment of hypercholesterolemia. Its high potency in inhibiting cholesterol synthesis is mainly due to liver-selective distribution of rosuvastatin (Nezasa et al., 2002a; Olsson et al., 2002). Rosuvastatin has very low passive membrane permeability, and with limited metabolism, its disposition is mediated almost entirely by uptake and efflux transporters. Rosuvastatin can be transported in vitro by multiple hepatic uptake transporters, e.g., OATP1B1, OATP1B3, and OATP2B1, and a bile acid uptake transporter in the liver, the sodium-taurocholate cotransporting polypeptide (NTCP) (Ho et al., 2006; Kitamura et al., 2008; Choi et al., 2011). In human hepatocytes, OATP1B1, OATP1B3, and NTCP are the predominant uptake transporters, with OATP1B1 and/or OATP1B3 accounting for ∼55% of the rosuvastatin uptake, both having a high affinity and high capacity, while NTCP accounts for ∼35%, having high capacity but lower affinity for rosuvastatin (Ho et al., 2006). The ATP-binding cassette (ABC) efflux transporters ABCG2 and ABCC2 are responsible for the biliary excretion of rosuvastatin in humans, as demonstrated by in vitro and in vivo studies (Kitamura et al., 2008; Hu et al., 2010; Jemnitz et al., 2010).
Patients carrying polymorphic variants of SLCO1B1 exhibit increased rosuvastatin plasma concentrations (Lee et al., 2005; Pasanen et al., 2007; Choi et al., 2008; Hua et al., 2011), but evidence regarding a correlation between SLCO1B1 genotype and therapeutic response is equivocal. Some studies find no correlation between polymorphic variants of SLCO1B1 and cholesterol-lowering efficacy of rosuvastatin (Romaine et al., 2010; Sirtori et al., 2011), whereas others do observe an association between these factors (Chasman et al., 2012). These findings raise the question of how reduced hepatic uptake by OATP proteins affects systemic and liver concentrations of rosuvastatin.
Several single (Oatp1b2, Oatp1a1, Oatp1a4) and combined knockout mouse models (Oatp1a/1b knockout mice) are available and have proved very useful in elucidating the in vivo physiologic and pharmacological functions of OATP1A/1B (reviewed in Iusuf et al., 2012b). First, Ose et al. (2010) showed that Oatp1a4 can transport rosuvastatin across the blood-brain barrier, but only upon in situ injection into the brain. Using Oatp1b2 knockout mice, a small-scale study showed that mouse Oatp1b2 might contribute to the liver uptake of rosuvastatin after intravenous administration, although only the liver-to-plasma ratio was significantly decreased in comparison with wild-type mice, whereas the plasma or liver concentrations were not significantly affected (Degorter et al., 2012).
In the present study, we aimed to obtain an in-depth understanding of the in vivo role of Oatp1a/1b transporters in the oral absorption and hepatic uptake of rosuvastatin. For this we used the Oatp1a/1b knockout mouse model (Slco1a/1b−/− mice, lacking all Oatp1a and -1b transporters) (van de Steeg et al., 2010). We compared the disposition of rosuvastatin in Oatp1a/1b knockout and wild-type mice after oral and intravenous administration.
Materials and Methods
Animals.
Animals were housed in small groups in a temperature-controlled environment with a 12-hour light/dark cycle. They received a standard diet (AM-II; Hope Farms, Woerden, The Netherlands) and acidified water ad libitum. All mouse experiments were approved by the Animal Experiments Review Board of The Netherlands Cancer Institute (Amsterdam), complying with Dutch legislation and in accordance with European Directive 86/609/EEC and the GlaxoSmithKline Policy on the Care, Welfare and Treatment of Animals. Male or female wild-type and Slco1a/1b−/− (Oatp1a/1b knockout) mice (van de Steeg et al., 2010) of comparable genetic background (>99% FVB) between 9 and 14 weeks of age were used as indicated.
Chemicals and Reagents.
Rosuvastatin calcium salt [(3R,5S,6E)-7-[4-(4-Fluorophenyl)-6-(1-methylethyl)-2-[methyl(methylsulfonyl)amino]-5-pyrimidinyl]-3,5-dihydroxy-6-heptenoic Acid Calcium Salt] was from Sequoia Research Products (Pangbourne, UK); other chemicals (dimethyl sulfoxide, bovine serum albumin) were from Sigma-Aldrich (St. Louis, MO), isoflurane (Forane) was from Abbott Laboratories (Queenborough, UK), and disodium EDTA was from LeoPharma BV (Breda, The Netherlands).
Drug Analysis.
Concentrations of rosuvastatin in blood, organs [homogenized in 1:10 volumes of ice-cold 4% (w/v) bovine serum albumin], and bile (diluted 100 times with human blank plasma) were determined by liquid chromatography–tandem mass spectrometry analysis as described (Hobbs et al., 2012).
Blood and Tissue Pharmacokinetic Experiments.
Rosuvastatin was dissolved in dimethylsulfoxide and diluted with saline (to 1 or 1.5 mg/ml) for administration of dose levels of 5 mg/kg i.v. or 15 mg/kg orally to mice. The maximum concentration of dimethylsulfoxide in the final solution was 2%. Ten microliters per gram of body weight was administered via oral gavage (n = 5–7 for each group), and 5 μl/g body weight was used for administration in the tail vein of mice. At different time points, EDTA-blood (via cardiac puncture) was sampled under isoflurane anesthesia. Mice were then sacrificed by cervical dislocation and tissues (liver without gall bladder) were isolated. After oral administration of 15 mg/kg to male mice, portal vein blood samples were taken prior to cardiac puncture. Blood samples were diluted 1:1 with water and then stored at −20°C until analysis.
Biliary Excretion of Rosuvastatin.
Gall bladder cannulations and collection of bile in male wild-type and Slco1a/1b−/− mice (n = 7) were performed as described (van Herwaarden et al., 2003). At the end of the experiment, blood and tissue samples were isolated and treated as described above.
Urinary and Fecal Excretion of Rosuvastatin.
A mass balance study was performed with Rucco Type M/1 stainless steel metabolic cages (Valkenswaard, The Netherlands). Mice (n = 5) received rosuvastatin orally (15 mg/kg) or intravenously (5 mg/kg). Urine and feces were collected in a 0- to 24-hour fraction after the drug administration, followed by isolation of blood and tissue samples as described above. For female mice, rosuvastatin was only given orally (15 mg/kg), and at different times after administration (7.5, 15, 30, 60, and 120 minutes) blood samples were isolated from the tail vein. After collecting the urine and feces for 24 hours, mice were sacrificed as described above.
RNA Isolation, cDNA Synthesis, and Reverse-Transcription Polymerase Chain Reaction.
RNA isolation from mouse kidney and subsequent cDNA synthesis and reverse-transcription polymerase chain reaction were performed as described (van Waterschoot et al., 2008). Specific primers (QIAGEN, Hilden, Germany) were used to detect expression levels of the following mouse Oatp1a genes: Oatp1a1, Oatp1a4, and Oatp1a6.
Pharmacokinetic Calculations and Statistical Analysis.
When variances were not homogeneous, the data were log-transformed to obtain normal distribution and equal variances. The two-sided unpaired Student’s t test was used throughout the study to assess the statistical significance of differences between two sets of data. Results are presented as mean ± S.D. Differences were considered to be statistically significant when P < 0.05. Averaged blood concentrations for each time point were used to calculate the area under the blood concentration–versus–time curve (AUC) from t = 0 to the last sampling time point by the linear trapezoidal rule; S.E. was calculated by the law of propagation of errors (Bardelmeijer et al., 2000). Results of the AUC measurements are presented as mean ± S.E.M. We calculated the ratio between plasma exposure after i.v. and oral administration, corrected for dose levels: (AUCoral/AUCi.v.) ⋅ (Dosei.v./Doseoral). The apparent hepatic extraction ratio was calculated as E = [1 – (AUCoral systemic/AUCoral portal vein)] (Gridelli et al., 1986).
Pharmacokinetic Modeling.
The modeling software Phoenix WinNonlin 6.1 (Tripos, L.P., St. Louis, MO) was used. Noncompartmental analysis was performed using blood data from i.v. and orally dosed mice from both strains. As the study design involved composite sampling, the "sparse" sampling function was used to maximize the contribution of the data from each mouse at each sample time. After oral administration, we calculated the renal clearance based on the amount of rosuvastatin recovered in the urine over 24 hours after oral administration, corrected for AUCsystemic extrapolated to infinity (AUC0–∞) and for individual mouse body weight. After i.v. administration, we calculated both renal and nonrenal clearance based on the amount of rosuvastatin recovered in the urine or the feces, corrected for AUCsystemic extrapolated to infinity (AUC0–∞) and for individual mouse body weight.
Results
Oatp1a/1b Transporters Are Not Essential for the Intestinal Absorption of Rosuvastatin.
Rosuvastatin is administered orally to patients, but not much is known about the transporters that facilitate its intestinal absorption. In wild-type mice, Oatp1a4 and Oatp1a5 are expressed in the small intestine (van de Steeg et al., 2012), where they might theoretically facilitate the absorption of various drugs. Therefore, we first investigated a possible role of Oatp1a uptake transporters in absorption of rosuvastatin across the intestinal wall using wild-type and Oatp1a/1b knockout mice. We measured rosuvastatin portal vein blood concentrations at various time points after oral administration (15 mg/kg). Oral absorption was very rapid, with the highest blood rosuvastatin concentrations observed at the earliest technically feasible time point, 5 minutes after dosing (Fig. 1A). The portal vein blood concentrations were modestly increased in Oatp1a/1b knockout mice (Fig. 1A; Table 1), indicating that Oatp1a transporters are not essential for the intestinal absorption of rosuvastatin. The modest increase in the portal vein blood concentrations in Oatp1a/1b knockout mice likely reflects the higher systemic blood concentrations (see below; Fig. 1, B and C).
Increased Systemic Exposure of Rosuvastatin in Oatp1a/1b-Null Mice after Oral Administration.
We also determined the systemic blood concentrations after oral administration of 15 mg/kg rosuvastatin to wild-type and Oatp1a/1b knockout mice. Rosuvastatin blood concentrations were markedly increased in Oatp1a/1b knockout mice in comparison with wild-type mice (Fig. 1, B and C), with an 8.2-fold-higher blood AUC5–240 (Table 1). Although liver concentrations were not significantly different between the mouse strains (Fig. 1, D and E), liver-to-blood ratios were at least 10-fold decreased in Oatp1a/1b-null mice at most time points, indicating a partially impaired liver uptake in the absence of Oatp1a/1b transporters (Fig. 1F).
The liver represents both the therapeutic target and the main clearance organ for rosuvastatin. Therefore, the extraction capacity of the liver is an important parameter to assess. Assuming that portal vein blood concentrations represent the amount of rosuvastatin before entering the liver and systemic blood concentrations represent the amount of drug escaping the uptake in the liver, we calculated the apparent hepatic extraction ratio E = [1 − (AUCoral systemic/AUCoral portal vein)] (Table 1). This approach assumes there is little alternative clearance (e.g., metabolic or renal) and tissue distribution of rosuvastatin outside of the liver, as was previously described for the rat (Nezasa et al., 2002a). In wild-type mice, rosuvastatin distributes almost exclusively to the liver after oral administration, with a very high apparent hepatic extraction ratio (0.93). Interestingly, in the absence of Oatp1a/1b uptake transporters this ratio dropped to 0.72, indicating a diminished efficacy in hepatic uptake in Oatp1a/1b knockout mice (Table 1).
Increased Systemic Exposure of Rosuvastatin in Oatp1a/1b-Null Mice after Intravenous Administration.
To further increase our understanding of how Oatp1a/1b transporters modulate the liver uptake of rosuvastatin, we performed a pharmacokinetic study upon i.v. administration of rosuvastatin (5 mg/kg). Similar to the oral administration experiment, the systemic exposure of rosuvastatin was markedly higher in Oatp1a/1b knockout mice in comparison with wild-type mice (Fig. 2, A and B), with a 3.1-fold-higher blood AUC5–240 (Table 1). Again, liver concentrations were not significantly different between the two mouse strains (Fig. 2, C and D), whereas liver-to-blood ratios were significantly and substantially reduced (5- to 10-fold) at most time points from 15 minutes on in Oatp1a/1b knockout mice (Fig. 2E), indicating a partially impaired hepatic uptake.
For the small intestinal wall (tissue) and small intestinal content concentrations and tissue-to-blood ratios of rosuvastatin, we observed very similar results as for the liver (Fig. 3, A–D). These results would be in line with the liver concentrations and liver-to-blood ratios: the substantial percentage of the dose of rosuvastatin (∼15%) found in the small intestinal wall early after administration (Fig. 3A) may reflect extensive enterohepatic circulation of rosuvastatin, assuming rapid hepatobiliary excretion in the intestine (see below). The kidney concentrations of rosuvastatin were also increased in Oatp1a/1b knockout mice, most likely reflecting the increased systemic exposure. There were no significant differences in kidney-to-blood ratios between the strains, suggesting that there is no important role of Oatp1a/1b transporters in the uptake of rosuvastatin into the kidney (Fig. 4, A–C).
Effect of Oatp1a/1b Transporters on the Biliary Excretion of Rosuvastatin.
We investigated the effect of Oatp1a/1b deficiency on biliary elimination of rosuvastatin after i.v. administration (5 mg/kg) to mice with a cannulated gall bladder and ligated common bile duct. The bile flow was not different between the two mouse strains (∼1.5 μl/min/g of liver). Biliary excretion of rosuvastatin was very rapid in both strains, with ∼60% of the dose being excreted in the first 15 minutes (Fig. 5A). In the first 30 minutes there was no significant difference between the two strains of mice, and only from 30 minutes on there was a slightly higher biliary output of rosuvastatin in Oatp1a/1b knockout mice (Fig. 5A), possibly reflecting slightly higher liver concentrations (e.g., Fig. 2C). Additionally, from 30 minutes after dosing, biliary excretion of rosuvastatin was much slower in both strains than in the first 15 minutes after i.v. administration (Fig. 5B). In this experiment enterohepatic circulation of rosuvastatin is interrupted due to ligation of the common bile duct, blocking possible recharging of the liver with rosuvastatin reabsorbed from the intestinal lumen, and thus continued biliary excretion. Note that the mRNA expression of Abcc2, one of the canalicular efflux transporters responsible for the biliary excretion of rosuvastatin, is somewhat lower in the Oatp1a/1b knockout mice, while expression of Abcg2 is not changed (van de Steeg et al., 2012).
There were no significant differences in the blood, liver, and small intestinal tissue concentrations in gall bladder–cannulated wild-type and Oatp1a/1b knockout mice 60 minutes after dosing (Supplemental Fig. 1). In the small intestinal content we observed significantly higher levels of rosuvastatin in Oatp1a/1b knockout mice compared with wild-type mice (1.1% ± 0.8% versus 0.4% ± 0.2% of dose; P < 0.05) (Supplemental Fig. 1). It is notable that only a small fraction of rosuvastatin was found back in the small intestinal wall and lumen. In the context of a ligated common bile duct, rosuvastatin can only reach the small intestinal lumen via direct intestinal excretion from the blood possibly mediated by Abcg2 or Abcc2, whose mRNA expression levels in the small intestine are similar in both strains (van de Steeg et al., 2012). Note that the amount of rosuvastatin directly excreted from the blood (Supplemental Fig. 1, B and C) is far lower than that excreted via the bile and probably reabsorbed via the small intestinal wall (Fig. 3A). Taken together, these data suggest that rosuvastatin undergoes extensive enterohepatic circulation.
Effect of Oatp1a/1b Transporters on the Urinary and Fecal Excretion of Rosuvastatin.
Next, we performed a mass balance experiment over 24 hours after i.v. (5 mg/kg) or oral (15 mg/kg) administration of rosuvastatin to male wild-type and Oatp1a/1b knockout mice. In line with the similar and high percentages of the dose excreted in the bile after i.v. administration (Fig. 5A), the dose recovered in the feces was nearly 60% in the wild-type mice and slightly, albeit significantly, lower in the Oatp1a/1b-null mice (57.3% ± 6.7% versus 49% ± 4.1% of dose; P < 0.05) (Fig. 6A). The amount of rosuvastatin recovered in the urine was 3-fold higher in the Oatp1a/1b knockout mice (20.4% ± 3% versus 5.9% ± 2.5% of dose) (Fig. 6A), probably reflecting the 3-fold-higher systemic exposure of rosuvastatin after i.v. administration (Table 1) and the diminished renal reabsorption of rosuvastatin in the Oatp1a/1b-null mice (see below; Table 2). The recovery after i.v. administration was ∼70%, possibly because upon i.v. administration rosuvastatin can distribute more extensively to other compartments in the body, from which rosuvastatin may be released only after 24 hours after administration.
After oral administration to male mice, the total rosuvastatin recovery was close to 100% of the dose (Fig. 6B), indicating very limited metabolism of this drug in mice. In line with the high apparent hepatic extraction ratio after oral administration (Table 1), the percentage of the dose recovered in the feces in wild-type mice was very high (∼100%), while it was reduced to ∼86% in Oatp1a/1b knockout mice, albeit not significantly (Fig. 6B). The amount of rosuvastatin recovered in the urine of wild-type mice was very low (0.10% ± 0.04% of dose), whereas in the absence of Oatp1a/1b transporters it was about 100-fold higher (10.7% ± 4.7%) (Fig. 6B). As a consequence, the renal clearance of rosuvastatin was 15.5-fold increased, from 1.3 ± 0.6 to 20.1 ± 9.0 ml/min/kg (P < 0.01), in the male Oatp1a/1b knockout mice (Table 2).
Role of Oatp1a/1b Transporters in the Renal and Nonrenal Clearance of Rosuvastatin.
The increased renal clearance of rosuvastatin in the Oatp1a/1b knockout mice might be explained if one or more of the Oatp1a/1b proteins in the kidney played a role in the tubular reabsorption of glomerularly filtrated or otherwise renally secreted rosuvastatin. It has been demonstrated that only Oatp1a1 and Oatp1a6, and to a lesser extent Oatp1a4, are significantly expressed in the male kidney (Cheng et al., 2005). Interestingly, Oatp1a1 was hardly expressed in the female kidney. If Oatp1a1 is primarily responsible for renal rosuvastatin reabsorption, the renal clearance in female wild-type mice should be higher than in male wild-type mice, and female Oatp1a/1b knockouts should show little increase in clearance. To test whether this was the case, we performed an oral systemic exposure and mass balance study with 15 mg/kg rosuvastatin in female mice (Fig. 7). Similar to results obtained in male mice (Fig. 1B), systemic blood concentrations were highly increased in female Oatp1a/1b knockout mice (Fig. 7A). Most of the rosuvastatin was recovered in the feces, with similar levels in wild-type and Oatp1a/1b-null mice (∼75% of dose) (Fig. 7B), whereas the amount in the urine was 19-fold higher in the female Oatp1a/1b knockout mice in comparison with wild-type controls (18.9% ± 3% versus 1% ± 0.3% of dose). Importantly, when comparing the male versus female mice, we observed that the amount of rosuvastatin in the urine of female wild-type mice was 9-fold higher than in the male wild-type mice (Fig. 7C), and it was 1.7-fold higher in the female Oatp1a/1b-null mice than in the male Oatp1a/1b-null mice (Fig. 7C).
Subsequent calculation of the renal clearances (Table 2) showed that renal clearance in wild-type females was 15 times higher than that in wild-type males, but not different from that in Oatp1a/1b knockout males. Moreover, the renal clearance was not significantly increased in female Oatp1a/1b knockout compared with wild-type female mice (Table 2). These results are consistent with a role of Oatp1a1 in renal rosuvastatin reabsorption in male mice. Reverse-transcription polymerase chain reaction analysis of Oatp1a1, Oatp1a4, and Oatp1a6 expression in kidney of our FVB strain wild-type mice (Supplemental Fig. 2) confirmed that Oatp1a1 was far more highly expressed in male than in female kidney (∼5000-fold), whereas Oatp1a4 was not differentially expressed, and Oatp1a6 only slightly (∼2-fold) more in male mice than in female mice. Collectively, the data suggest that Oatp1a1 plays a role in the renal reabsorption of rosuvastatin and thus diminishes its renal clearance.
After i.v. administration, the renal clearance was only 2-fold increased in Oatp1a/1b knockout male mice in comparison with wild-type mice and the nonrenal clearance was 2-fold decreased (Table 2). This reflects the decreased renal reabsorption of rosuvastatin in the male Oatp1a/1b knockout mice (see above; Table 2). The renal clearance accounted for ∼10% of the total clearance in the male wild-type mice and for ∼30% in the Oatp1a/1b knockout mice (Table 2).
Pharmacokinetic Modeling.
We further performed a limited noncompartmental modeling of the pharmacokinetic data using the sparse sampling function. The results are presented in Table 3. It is noteworthy that values for blood AUC after i.v. administration calculated using the pharmacokinetic software are much higher than the AUC values observed from t = 5 until t = 240 minutes, calculated using the linear trapezoidal rule (Table 1 versus Table 3). This discrepancy is mainly due to the extrapolation of the blood concentration data to t = 0 minutes. Below we discuss only the data from Table 3.
The ratio between blood exposure after oral versus i.v. administration of rosuvastatin was increased in the Oatp1a/1b knockout mice (from 0.011 to 0.058), most likely as a consequence of the impaired first-pass uptake in the liver of these mice after oral administration (Table 3).
After i.v. administration we observed a decrease, albeit modest, in the total clearance (from 20.3 to 11.8 ml/min/kg). Finally, the half-life of rosuvastatin after i.v. administration was almost 2-fold higher (26.2 minutes in wild-type mice versus 49.9 minutes in Oatp1a/1b knockout mice) (Table 3).
The pharmacokinetic parameters (total clearance and exposure) we obtained after oral and i.v. administration in the wild-type mice were in general agreement with previous studies, although there were some differences in half-life values, probably due to the different genetic backgrounds of the mice and the dosages used (Peng et al., 2009).
Discussion
Here we show that Oatp1a/1b uptake transporters are not essential for the intestinal absorption of rosuvastatin after oral administration, but that they strongly affect rosuvastatin systemic exposure after oral and i.v. administration. Interestingly, the strong increase (8-fold) in systemic exposure in Oatp1a/1b-null mice is not accompanied by a significant decrease in liver exposure, or in biliary excretion of rosuvastatin after i.v. administration, but the hepatic extraction ratio is markedly decreased in the absence of Oatp1a/1b transporters. The major pharmacokinetic impact of Oatp1a/1b transporters on rosuvastatin therefore occurs through their hepatic uptake activity. We also show that renal clearance of rosuvastatin, although small compared with hepatic clearance, is gender-dependent and might be affected by the different expression levels of Oatp1a1 in the kidney of male versus female wild-type mice.
It has been proposed that OATP1A/Oatp1a transporters can mediate the intestinal absorption of many drugs, including statins. Despite being quite polar, rosuvastatin was very rapidly and efficiently absorbed, with both portal vein and systemic blood concentrations highest at the earliest feasible sampling time point (5 minutes) in both wild-type and Oatp1a/1b knockout mice. Although Oatp1a/1b uptake transporters are clearly not essential for the intestinal uptake of rosuvastatin, given its polarity (logP = 1.92), it is almost certain that other uptake transporters must be involved. One candidate could be mouse Oatp2b1, since several studies have shown that its human ortholog OATP2B1 can transport rosuvastatin in vitro (Ho et al., 2006; Kitamura et al., 2008; Varma et al., 2011). The contribution of OATPs in the oral absorption of rosuvastatin was also investigated in an in vivo study in pigs in which gemfibrozil (an OATP inhibitor) was coadministered with rosuvastatin (Bergman et al., 2009). However, despite high concentrations of gemfibrozil, enough to efficiently inhibit OATP1A2 and OATP2B1 in the small intestine, the intestinal absorption of rosuvastatin was not affected (Bergman et al., 2009). Additional studies are therefore required to establish the transporters responsible for the intestinal uptake of rosuvastatin.
Interestingly, our results suggest that Oatp1a1 in the kidney might play a role in the renal reabsorption of rosuvastatin, although the contribution of renal clearance to the systemic clearance of rosuvastatin is small in wild-type mice. We observed a gender-dependent difference in the renal clearance of rosuvastatin, which has also been described in rats for perfluorooctanoic acid, a potentially toxic chemical and substrate of Oatp1a1, which is mainly eliminated renally (Yang et al., 2010). Therefore, besides their predominant role in mediating hepatic clearance of drugs, Oatp1a transporters might affect the renal clearance of drugs as well.
Systemic exposure of rosuvastatin after oral administration was 7- to 14-fold higher in the absence of Oatp1a/1b transporters. Oatp1a1, Oatp1a4, and Oatp1b2, present in the basolateral membrane of hepatocytes in mice, likely mediate the same function(s) as human OATP1B1 and OATP1B3. Therefore, our results are in line with data from patients carrying low-activity genetic polymorphic variants of OATP1B1. These variants are associated with increased plasma levels of rosuvastatin (reviewed in Hua et al., 2011), but it seems that the magnitude of effects varies between different ethnic groups. For example, Korean individuals with the low-activity variant *15/*15 (two copies of the 521T>C allele) had 1.7-fold-higher rosuvastatin AUCs than the control group (Choi et al., 2008). In a study comparing white and Asian subjects, the variant *15/*15 was associated with higher rosuvastatin AUC only in the white subjects, and not in the Asian ones (Lee et al., 2005). Similarly, in a Finnish population, a slightly higher systemic exposure to rosuvastatin was observed in carriers of the OATP1B1 521T>C variant (Pasanen et al., 2007). In addition, documented drug-drug interactions between rosuvastatin and OATP1B inhibitors further support the importance of OATP1B1 in the systemic exposure of rosuvastatin. In humans, it was shown that after repeated administrations of oral gemfibrozil (at plasma concentrations that mainly inhibit OATP1B1) the plasma AUC of rosuvastatin was 1.8-fold higher (Schneck et al., 2004). Coadministration with cyclosporine led to 7- to 10-fold-higher rosuvastatin AUCs (Simonson et al., 2004). The net effect of OATP1B inhibition by cyclosporine is difficult to estimate, as cyclosporine can also inhibit various other influx and efflux transporters involved in the pharmacokinetics of rosuvastatin, e.g., NTCP and/or ABCC2 and ABCG2. A similar pronounced effect of coadministration of cyclosporine and rosuvastatin was seen in a study with pigs (Bergman et al., 2009).
Despite the markedly increased systemic exposure of rosuvastatin in Oatp1a/1b knockout mice, the liver concentrations were not significantly reduced in these mice. However, the liver-to-blood ratios were markedly decreased, indicating an impaired liver uptake. A recent small-scale study in Oatp1b2-null mice showed that at 30 minutes after i.v. administration, liver-to-plasma ratios of rosuvastatin were 2.7-fold lower than in wild-type mice, whereas the plasma concentrations were not significantly different between the two mouse strains (Degorter et al., 2012). We observed 10-fold-lower liver-to-blood ratios in Oatp1a/1b-null mice at the same time point, and 7-fold-increased blood concentrations. This indicates that, in addition to Oatp1b2, hepatic Oatp1a1 and/or Oatp1a4 also play an important role in liver uptake of rosuvastatin.
As previously mentioned, liver (and bile) concentrations of rosuvastatin were mostly not significantly altered by the absence of Oatp1a/1b transporters, in spite of the strongly increased blood exposure. This surprising finding can be explained by the intrinsic properties of rosuvastatin. Although our data show a very high hepatic extraction ratio (0.93) of rosuvastatin in wild-type mice after oral administration, this dropped only to 0.72 in Oatp1a/1b knockout mice (Table 1), indicating that in the knockouts there is still a very substantial hepatic uptake of rosuvastatin. Considering the high and similar amount of rosuvastatin taken up in wild-type and knockout liver (Fig. 1E), the modest decrease in hepatic extraction is sufficiently offset by the higher portal vein concentrations. Small differences in liver concentration can be easily lost in the experimental variation and thus not become obvious. However, any small decrease in liver exposure theoretically translates into a much larger (4-fold) increase in the fraction of rosuvastatin that “escapes” the liver [from 0.07 to 0.28 (1 − apparent hepatic extraction ratio)] (Table 1) and thus ends up in the systemic circulation. Note that the estimated total amount of rosuvastatin in the systemic circulation represents <0.5% of the dose (at 1 μg/ml) (Fig. 1B), an amount that is negligible compared with the ∼20% of the dose found in the liver over the first 15–60 minutes (Fig. 1E). It is therefore not surprising that a relatively big change in systemic blood concentrations can occur with little impact on the liver concentrations. This idea is also supported by a physiologically based pharmacokinetic model described for pravastatin (Watanabe et al., 2009). This model predicts that for a predominantly hepatically cleared drug, a diminished hepatic uptake (like in the absence of Oatp1a/1b transporters) leads to a substantial increase in the systemic exposure, while the liver exposure is not so much affected, especially for drugs that have a negligible renal clearance. In our study, renal clearance of rosuvastatin accounts for maximally 1% of total clearance in the wild-type mice versus 25% in the Oatp1a/1b-null mice after oral administration (Table 2). This is in line with data from humans (Martin et al., 2003) and rats (Nezasa et al., 2002b), which also exhibit a low renal clearance of rosuvastatin.
The still substantial hepatic uptake of rosuvastatin in Oatp1a/1b knockout mice, albeit at higher blood concentrations, indicates that alternative transporters for rosuvastatin can partially compensate for the loss of Oatp1a/1b transporters. We hypothesize that already shortly after administration, reduced liver uptake due to the absence of Oatp1a/1b transporters results in a substantial increase in rosuvastatin blood concentrations. This high blood concentration allows continued substantial uptake of rosuvastatin into the liver via low-affinity but high-capacity alternative transporters. Nevertheless, the rescue provided by these alternative transporters is partial, as the systemic blood concentrations remain markedly increased in the Oatp1a/1b-null mice over at least 4 hours after administration. The most obvious candidate as an alternative transporter in mouse or human, but not rat, would be NTCP/Ntcp, which has been described to facilitate cellular uptake of rosuvastatin in vitro (Ho et al., 2006; Kitamura et al., 2008). Using double-expressing oocytes of wild-type and/or polymorphic variants of OATP1B1 and NTCP, it was shown that reduced rosuvastatin uptake by OATP1B1*15 can be masked in the presence of NTCP, suggesting that NTCP can rescue OATP1B1 loss of function in vitro (Choi et al., 2011). Another transporter that might compensate for the loss of Oatp1a/1b function is OATP2B1 (human) or Oatp2b1 (mouse/rat), which is also present in the basolateral membrane of hepatocytes and can mediate rosuvastatin transport in vitro (Ho et al., 2006; Kitamura et al., 2008; Varma et al., 2011).
Previously, we showed that mouse Oatp1a/1b uptake transporters control the hepatic uptake of pravastatin (Iusuf et al., 2012a). Similar to rosuvastatin, absence of Oatp1a/1b resulted in a substantial increase in systemic exposure of pravastatin both after oral and i.v. administration. However, in contrast to rosuvastatin, pravastatin liver exposure was reduced 2-fold in the Oatp1a/1b-null mice, and the impact of Oatp1a/1b transporters was very obvious in the 8-fold-decreased biliary excretion of pravastatin after i.v. administration (Iusuf et al., 2012a). Although similar in hydrophilicity (pravastatin, logP = 1.65; rosuvastatin, logP = 1.92), rosuvastatin has a higher affinity to distribute to the liver than pravastatin (Nezasa et al., 2002a, 2003). Indeed, rosuvastatin has a higher apparent hepatic extraction ratio (0.93 in wild-type versus 0.72 in Oatp1a/1b-null mice) (Table 1) than pravastatin (0.88 in wild-type versus 0.52 in Oatp1a/1b-null mice) (Iusuf et al., 2012a). In addition, rosuvastatin appears to be more substantially transported by alternative transporters such as Ntcp and/or Oatp2b1 compared with pravastatin, resulting in a somewhat less pronounced increase in the systemic exposure in the Oatp1a/1b knockout model. Nevertheless, hepatic uptake of both compounds is still substantial in the absence of Oatp1a/1b transporters, indicating that the uptake transporters involved have an appreciable redundancy and are capable of relatively efficient clearance even when the main disposition mechanism has been compromised.
In conclusion, Oatp1a/1b uptake transporters determine the systemic exposure of rosuvastatin, without substantially affecting its liver exposure after bolus administration. Whether this is also true after chronic administration of rosuvastatin in patients remains to be seen. Our findings are clinically relevant for individuals with low-activity polymorphic variants of OATP1B1 and heterozygous carriers of the various full-deficiency mutations of OATP1B1 and/or OATP1B3 (Pasanen et al., 2008; van de Steeg et al., 2012). These individuals, when treated with rosuvastatin, might be at risk of developing myopathy, the major systemic side effect of statins. On the other hand, the therapeutic effect of rosuvastatin in the liver might be less affected.
Acknowledgments
The authors thank Marion Ludwig for technical assistance.
Authorship Contributions
Participated in research design: Iusuf, Kenworthy, Schinkel.
Conducted experiments: Iusuf, van Esch, Hobbs, Wagenaar.
Contributed new reagents or analytic tools: Taylor, Wagenaar.
Performed data analysis: Iusuf, Hobbs.
Wrote or contributed to the writing of the manuscript: Iusuf, Hobbs, Kenworthy, van de Steeg, Schinkel.
Footnotes
- Received August 15, 2012.
- Accepted January 29, 2013.
This work was supported by the Dutch Cancer Society [Grant 2007-3764].
This article has supplemental material available at molpharm.aspetjournals.org.
Abbreviations
- ABC
- ATP-binding cassette
- AUC
- area under the curve
- NTCP
- sodium-taurocholate cotransporting polypeptide
- OATP/Oatp
- organic anion–transporting polypeptides
- SLCO/Slco
- solute carrier organic-anion
- Copyright © 2013 by The American Society for Pharmacology and Experimental Therapeutics