The renal drug-drug interaction between famotidine (an H2 receptor antagonist) and probenecid has not been reproduced in rats. We have proposed that this is caused by a species difference in the transport activity by human/rat organic anion transporter (OAT) 3 and the expression of organic cation transporter (OCT) 1 in the rodent kidney. Since monkey OATs (mkOATs) exhibit similar transport activities to human orthologs, it is hypothesized that in vivo studies in monkeys will allow a more precise prediction of renal drug-drug interactions in humans. Famotidine and cimetidine were efficiently taken up by mkOAT3-expressing human embryonic kidney cells (Km, 154 and 71 μM, respectively), and their uptake was strongly inhibited by probenecid (Ki, 3.0–5.7 μM). Quantification of mkOCT1 and mkOCT2 mRNAs in the monkey kidney using real-time reverse transcription-polymerase chain reaction revealed their predominant expression in the liver and kidney, respectively. Crossover studies were conducted in cynomolgus monkeys. Famotidine was given by i.v. administration, with or without probenecid. Probenecid treatment caused a 65% reduction in the renal clearance (0.426 ± 0.079 versus 0.165 ± 0.027 l/h/kg) and a 90% reduction in the tubular secretion clearance (0.275 ± 0.075 versus 0.0230 ± 0.0217 l/h/kg), whereas it had no effect on the renal clearance of cimetidine. In contrast to the species-dependent effect of probenecid, allometric scaling using animal data (rat, dog, and monkey) successfully predicted the renal and tubular secretion clearance of famotidine in humans. These results suggest that monkeys are more appropriate animal species for predicting the renal drug-drug interactions in humans.
The kidney plays an important role in the detoxification of xenobiotics and endogenous waste as well as maintaining the balance of ions and nutrients in the body. Urinary excretion is the major detoxification mechanism of the kidney, and this is governed by glomerular filtration, tubular secretion across the proximal tubules, and reabsorption. Transporters play important roles in the tubular secretion of drugs. Many studies have described the role of multispecific organic anion and cation transporters [organic anion transporter (OAT)/SLC22 and organic cation transporter (OCT)/SLC22] in the renal uptake of drugs (Lee and Kim, 2004; Wright and Dantzler, 2004; Shitara et al., 2005). Both Oct1 (Slc22a1) and Oct2 (Slc22a2) are involved in the renal uptake of organic cations on the basolateral membrane of the proximal tubules in rodents, whereas OCT2 plays a predominant role in human kidney (Lee and Kim, 2004; Wright and Dantzler, 2004). As renal organic anion transporters, two isoforms (Oat1/Slc22a6 and Oat3/Slc22a8) in rodents and three isoforms (OAT1, OAT2/SLC22A7 and OAT3) in humans, have been identified on the basolateral membrane of the proximal tubules (Lee and Kim, 2004; Miyazaki et al., 2004; Wright and Dantzler, 2004).
Probenecid, an antipodagric drug, is a well known inhibitor of organic anion transporters. Drug interactions with probenecid have been reported involving renal excretion in humans resulting in a prolongation of the plasma elimination half-life (Cunningham et al., 1981). For famotidine, an H2 receptor antagonist, its effect has been reported to be species-dependent. The renal secretion clearance of famotidine in humans was significantly reduced by oral coadministration of probenecid (Inotsume et al., 1990), whereas this interaction has not been reproduced in rats, although the plasma concentration of probenecid achieved a similar level to that in clinical studies (Lin et al., 1988). In contrast to famotidine, the renal secretion clearance of cimetidine, another H2 receptor antagonist, in both humans and rats was only slightly (ca. 20%) reduced by coadministration with probenecid (Lin et al., 1988; Gisclon et al., 1989). These results suggest that the contribution of transporters involved in the tubular secretion of famotidine is different between rodents and humans, and the organic anion transporter plays a more important role in humans. We found that the transport activity of famotidine by hOAT3 is greater than that by rOat3 and that the unbound plasma concentration of probenecid is sufficiently higher than its Ki values for rat and human OAT3 (Tahara et al., 2005a). Therefore, we hypothesized that this increases contribution of OAT3, a probenecid-inhibiting fraction, to the renal uptake of famotidine in humans together with the fact that hOCT1 is not present in the kidney.
To predict the possibility of drug-drug interactions in humans, such species differences have to be overcome. The monkey has been used in pharmacological, toxicological, and pharmacokinetic studies by many pharmaceutical companies, and it is recognized as a suitable animal model for the validation of in vitro scaling methods since it is the second nearest species to humans in the evolutionary tree. Ward and Smith (2004a,b) have reported that the monkey provides the most qualitatively and quantitatively accurate predictions of human pharmacokinetics by retrospectively analyzing the pharmacokinetic parameters of 103 nonpeptide xenobiotics in monkeys and humans. In addition, we have demonstrated that there is a good correlation of the transport activities with respect to that of reference compounds between mk- and hOAT3, whereas the correlation was poor between rat and human OAT3 (Tahara et al., 2005b). According to our mRNA quantification, mkOCT1 and mkOCT2 are predominantly expressed in the liver and kidney, respectively. Therefore, we consider that monkeys will be a better animal model for predicting drug-drug interactions involving multiple transporters.
In the present study, we examined the effect of probenecid on the pharmacokinetics of famotidine and cimetidine in cynomolgus monkeys to investigate whether the drug-drug interaction between probenecid and famotidine can be reproduced in monkeys. In addition, the uptake of H2 receptor antagonists (cimetidine, famotidine, and ranitidine) by HEK293 cells expressing mkOAT1 and mkOAT3 and the inhibitory effect of probenecid on the uptake of the H2 receptor antagonists were also examined.
Materials and Methods
Materials. Famotidine and probenecid were purchased from Nacalai Tesque (Kyoto, Japan), and cimetidine and ranitidine were purchased from Sigma-Aldrich (St. Louis, MO). All other chemicals were of analytical grade and commercially available.
Cells and Reagents for in Vitro Studies. In vitro experiments were carried out using HEK293 cells stably transfected to express functional mkOAT1 or mkOAT3 and corresponding control cells transfected with the pcDNA3.2 expression vector. Generation of both cell lines and their characterization are described elsewhere (Tahara et al., 2005b). The cells were maintained in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 μg/ml streptomycin, and 400 μg/ml G418 (Invitrogen) at 37°C with 5% CO2 and 95% humidity on the bottom of a dish. mkOAT1- and mkOAT3-expressing cells were seeded in polylysine-coated 12-well plates (Becton Dickinson, Franklin Lakes, NJ) at a density of 1.2 × 105 cells/well. Cell culture medium was replaced with culture medium supplemented with 5 mM sodium-butyrate 24 h before transport studies to induce the expression of those proteins.
Transport Assay. Transport studies were carried out as described previously (Tahara et al., 2005a). Uptake was initiated by adding medium containing 10 μM of compounds, in the presence or absence of inhibitor, after cells had been washed twice and preincubated with Krebs-Henseleit buffer at 37°C for 15 min. The Krebs-Henseleit buffer consists of 118 mM NaCl, 23.8 mM NaHCO3, 4.83 mM KCl, 0.96 mM KH2PO4, 1.20 mM MgSO4, 12.5 mM HEPES, 5 mM glucose, and 1.53 mM CaCl2 adjusted to pH 7.4. The uptake was terminated at designed times by adding ice-cold Krebs-Henseleit buffer after removal of the incubation buffer. Then, cells were washed twice with 1 ml of ice-cold Krebs-Henseleit buffer. For the determination of the uptake of cimetidine, ranitidine, and famotidine, cells were dissolved in 300 μl of 0.2 N NaOH and kept overnight. Aliquots (150 μl) were transferred to vials after adding 30 μl of 1 N HCl. Aliquots (100 μl) were used for LC-MS quantification as described below. The remaining 10 μl of the aliquots of cell lysate was used to determine the protein concentration by the method of Lowry with bovine serum albumin as a standard. Ligand uptake was given as the cell/medium concentration ratio determined as the amount of ligand associated with cells divided by the medium concentration.
Kinetic Analyses of the Transport Study. Kinetic parameters were obtained using the Michaelis-Menten equation: where v is the uptake rate of the substrate (picomoles per minute per milligram of protein), S is the substrate concentration in the medium (micromolar), Km is the Michaelis-Menten constant (micromolar), and Vmax is the maximum uptake rate (picomoles per minute per milligram of protein). To obtain the kinetic parameters, the equation was fitted to the uptake velocity using a MULTI program (Yamaoka et al., 1981). The input data were weighted as the reciprocals of the observed values, and the Damping Gauss Newton Method algorithm was used for fitting. Inhibition constants (Ki) of several compounds were calculated assuming competitive inhibition using the following equation since the substrate concentration was sufficiently low compared with the Km values. where CL represents the uptake clearance, and the subscript (+inh) represents the value in the presence of inhibitor. I represents the concentration of inhibitor (micromolar). Fitting was performed by an iterative nonlinear least-square method using a MULTI program, and the Damping Gauss Newton Method algorithm was used for fitting (Yamaoka et al., 1981).
Real-Time RT-PCR Analysis. Male cynomolgus monkey liver and kidney was purchased from BOZO Research (Shizuoka, Japan). Total RNA was extracted using the extraction solution of ISOGEN (NIPPON GENE, Tokyo, Japan) according to the manufacturer's protocol. The total RNA was reverse-transcribed using SuperScript First-Strand Synthesis System for RT-PCR kit (Invitrogen) with an oligo(dT)12–18 as primer. Fifty nanograms of the RT reaction mixture was taken for a real-time PCR (SYBR, Green I chemistry) (94°C for 5 min, 94°C for 30 s, 65°C for 30 s, 72°C for 30 s, for 45 cycles; Applied Biosystems ABI PRISM 7700 Sequence Detector system; Applied Biosystems, Foster City, CA) using Ex Taq polymerase (Takara Bio, Kyoto, Japan) with specific primers based on human OCTs (hOCT1 sense primer, 5′-TAAAGATAATGGACCACATCGC-3′; antisense primer, 5′-TATGATGTTTAACCAGTGCAGG-3′, accession no. NM_003057-809; hOCT2 sense primer, 5′-AGTTGCCTATACAGTTGGGCTC-3′; antisense primer, 5′-CAGGGCAGAGTAGAAGAAATCC-3′, accession no. NM_003058-77; hOCT3 sense primer, 5′-GACCAAGGATTTGAGAAAGTTG-3′; antisense primer, 5′-AGGGAATCTGTGGCTCTAGG-3′, accession no. NM_021977-172). All values are expressed as relative luminescence units per 50 ng of total RNA normalized with that of β-actin.
In Vivo Study in Monkeys. Four male cynomolgus monkeys were obtained from Siconbrec Inc. (Manila, Philippines). The mean body weight of the monkeys was 5.7 kg, ranging from 4.7 to 6.5 kg. The four cynomolgus monkeys received a single i.v. dose of famotidine and cimetidine, at a dose of 0.3 and 4 mg/kg, after an oral dose of 10 ml of vehicle as a control phase. Thereafter, a study with a randomized crossover design was conducted at intervals of 1 month. The monkeys received the same amount of famotidine, cimetidine, and 22.5 mg/kg probenecid, that is, 15 mg/kg probenecid 2 h before and 7.5 mg/kg simultaneously with an i.v. dose of famotidine and cimetidine. The study protocol for the animal experiment was approved by Animal Care Committee of Kyowa Hakko Kogyo Co., Ltd (Shizuoka, Japan).
Sample Collection. Blood samples (0.5 ml each) for the determination of famotidine and cimetidine were taken with heparinized syringe at 5, 10, and 15 min and 0.5, 1, 2, 4, and 8 h after the administration of famotidine and cimetidine. Urine samples were collected from 0 to 4, 4 to 8, and 8 to 24 h after dosing. Plasma was separated immediately and kept at –40°C until analysis. A part of the urine sample was stored at –40°C until analysis. The remaining urine samples were discarded after the urine volume had been recorded.
Determination of Protein Binding in Plasma. Plasma (150 μl) obtained at 30 min during the i.v. administration was directly applied to an MPS micropartition device (Millipore Corporation, Bedford, MA). The micropartition device was then centrifuged at 1500g for 15 min, and the unbound cimetidine, famotidine, and probenecid concentration in the filtrate was determined by LC-MS. The free fraction in plasma (fp) was determined as the ratio of the unbound concentration in the filtrate to the total concentration.
Determination of Plasma and Cell Lysate Concentrations. The quantification of cimetidine, famotidine, ranitidine, and probenecid was performed by high-performance liquid chromatography (Alliance 2690; Waters, Milford, MA) connected to a mass spectrometer (ZQ; Micromass, Manchester, UK) (Nagata et al., 2004). Aliquots (100 μl) of plasma and urine containing famotidine, cimetidine, and probenecid were precipitated with 200 μl of methanol containing an internal standard (ranitidine), mixed, then centrifuged, and 25 μl of the supernatants was injected into the LC-MS. High-performance liquid chromatography analysis was performed on a Capcell Pak C18 MG column (3 μm, 4.6 mm i.d., 75 mm; Shiseido, Tokyo, Japan) at room temperature. Elution was performed with a 0 to 90% linear gradient of 10 mM ammonium acetate-acetonitrile over 4 min at 0.8 ml/min. A portion of eluent (split ratio, 1:3) was introduced to the MS via an electrospray interface. Detection was performed by selected ionization monitoring in positive ion mode (m/z, 253; m/z, 315; m/z, 338; and m/z, 286 for cimetidine, ranitidine, famotidine, and probenecid).
The lower limit of quantitation for famotidine and cimetidine was 5 ng/ml in plasma and 10 ng/ml in urine, respectively. The standard concentration of both drugs ranged from 5 to 1000 ng/ml in plasma and 10 to 1000 ng/ml in urine, respectively. The within-day coefficient of variation of both drugs for plasma and urine determinations was less than 10%. Creatinine concentrations in plasma and urine were determined by an enzymatic method (creatinase/sarcosine oxidase/peroxidase) using an AutoAnalyzer 7070 (Hitachi Instruments Service, Tokyo, Japan). Probenecid and H2 receptor antagonists did not interfere with the quantification of creatinine.
Pharmacokinetic Analysis. Plasma concentration time data (Cp) of famotidine and cimetidine were fitted to a two-compartment model using a MULTI program (Yamaoka et al., 1981). The following parameters were calculated whenever appropriate: t1/2α (the distribution half-life, calculated as 0.693/α), t1/2β (the elimination half-life, calculated as 0.693/β); AUC0-∞ (the total area under the plasma concentration time curve extrapolated to infinity, calculated as A/α + B/β); CLp (the plasma clearance, calculated as dose/AUC0-∞); MRT [the mean residence time, calculated as (A/α2 + B/β2)/AUC0-∞)]; Vd,ss (the steady-state distribution volume, calculated as MRT × CLp); Vc [the distribution volume of the central compartment, calculated as dose/(A + B)]; and CLdis (the distribution clearance, calculated as K12 × Vc). The renal clearance (CLrenal) of famotidine and cimetidine was calculated as CLrenal = Ae/AUC0-∞, where Ae is the amount of famotidine and cimetidine excreted in the urine within 24 h. The tubular secretion clearance (CLsec) was calculated as CLsec = CLrenal – fp × GFR, where fp is the unbound fraction of famotidine and cimetidine in plasma, and GFR is the glomerular filtration rate. The percentage of drug excreted in the urine (fe) was calculated as fe = Ae/dose. The creatinine clearance values were used for the values of GFR in this study. The creatinine clearances in cynomolgus monkeys treated with or without probenecid were determined as 0.241 ± 0.021 (0.217–0.252) and 0.189 ± 0.025 (0.169–0.217) l/h/kg, respectively (mean ± S.D., not significant, p > 0.05). The two-tailed paired Student's t test was used for a statistical analysis, and value of p < 0.05 was considered significant.
Uptake of H2 Receptor Antagonists by mkOAT1 and OAT3.Figure 1 shows the time profiles and concentration dependence of the uptake of the H2 receptor antagonists by mkOAT1-, mkOAT-3, and vector-HEK, respectively. As reported previously in hOAT1-HEK (Tahara et al., 2005a), the uptake of cimetidine by mkOAT1-HEK was slightly greater than that by mock cells (7.27 ± 0.20 versus 3.48 ± 0.04 μl/mg protein at 5 min). A slight increase was also observed in the uptake of ranitidine by mkOAT1-HEK (18.8 ± 2.27 versus 11.4 ± 1.11 μl/mg protein at 5 min), but no specific uptake was observed for famotidine. The uptake of cimetidine, famotidine, and ranitidine by mkOAT3-HEK was significantly greater than in vector-HEK (Fig. 1). Since the uptake of cimetidine, ranitidine, and famotidine by mkOAT3-HEK increased linearly up to 3 and 5 min of incubation, the uptake of cimetidine, famotidine, and ranitidine by mkOAT3 at 3 min was used for further characterization.
The concentration dependence of the uptake of the H2 receptor antagonists by mkOAT3-HEK was examined (Fig. 2). Their specific uptake by mkOAT3 consisted of one saturable component. The kinetic parameters are summarized in Table 1. The Km values of the H2 receptor antagonists for mkOAT3 were almost identical; however, the intrinsic transport activity (Vmax/Km) of cimetidine was greater than that of famotidine and ranitidine.
Effect of Probenecid on mkOAT3-Mediated Transport of Famotidine and Cimetidine. The inhibitory effect of probenecid on the mkOAT3-mediated transport of cimetidine and famotidine was examined (Fig. 3). Probenecid strongly inhibited the mkOAT3-mediated transport of cimetidine and famotidine in a concentration-dependent manner. The Ki values of probenecid for cimetidine and famotidine uptake by mkOAT3-HEK were determined to be 5.68 ± 0.78 and 2.97 ± 1.53 μM.
Semiquantitative Real-Time RT-PCR Analysis of Organic Cation Transporters. A method for the detection of monkey OCTs that combines reverse transcription with real-time RT-PCR was developed using specific primers designed from a nucleotide sequence of hOCT1, hOCT2, and hOCT3. The mRNA expression levels of putative mkOCT1, mkOCT2, and mkOCT3 in the liver were 121, 0.0640, and 0.0846, and those in the kidney were 0.0110, 1400, and 0.0879 (relative luminescence units per 50 ng of total RNA). The relative expression level of putative mkOCT1 was more than 10,000-fold higher in the liver than in the kidney, whereas that of putative mkOCT2 was more than 20,000-fold higher in the kidney than in the liver. In contrast, the expression level of putative mkOCT3 mRNA was comparable in the liver and kidney. Using these primer sets, electrophoretic analysis showed a single band of OCT1 [404 base pairs (bp)], OCT2 (430 bp), and OCT3 (419 bp) in monkey liver and kidney as well as human mixed cDNA (data not shown).
Effect of Probenecid on Pharmacokinetics of Famotidine in the Monkeys. The mean plasma concentration time profile of famotidine in cynomolgus monkeys is shown in Fig. 4A. The mean plasma pharmacokinetic parameters are summarized in Table 2. The urinary recovery over the 0- to 24-h collection period and the renal and tubular secretion clearances are included in Table 2. There was a significant difference between the probenecid-treated and untreated groups with regard to the pharmacokinetic parameters of famotidine, such as AUC0-∞, Vd,ss, CLp, CLrenal, and CLsec, but no significant difference in the t1/2α, t1/2β, fp, and fe compared with the controls. The fp and fe in cynomolgus monkeys with or without probenecid treatment were determined as 74.4 ± 9.7 versus 70.3 ± 6.0% and 49.3 ± 13.4 versus 38.5 ± 10.9%. The plasma and renal clearances of famotidine was reduced to 50 and 65% of the control values, and the steady-state distribution volume was also reduced by probenecid. Probenecid completely blocked the renal tubular secretion of famotidine reducing it from 0.275 ± 0.075 to 0.0230 ± 0.0217 l/h/kg and concomitantly increased the total exposure (AUC) of famotidine by approximately 2-fold. The nonrenal clearance of famotidine was reduced from 0.445 ± 0.168 to 0.244 ± 0.029 l/h/kg, although this was not statistically significant (p > 0.05).
Effect of Probenecid on the Pharmacokinetics of Cimetidine in the Monkeys. The mean plasma concentration time profiles of cimetidine in the cynomolgus monkey are shown in Fig. 4B. The mean plasma pharmacokinetic parameters are summarized in Table 3. The urinary recovery over the 0- to 24-h collection period and the renal and tubular secretion clearances are included in Table 3. As observed in Fig. 4B and Table 3, probenecid had very little effect on the pharmacokinetics of cimetidine in the cynomolgus monkeys. Probenecid treatment produced no significant difference in any of the pharmacokinetic parameters of cimetidine. The fp and fe in cynomolgus monkeys with or without probenecid treatment were determined as 79.1 ± 4.8 versus 81.7 ± 2.0% and 37.0 ± 3.9 versus 36.3 ± 3.9%.
Plasma Concentration of Probenecid in the Monkey. The mean plasma concentration time profiles of probenecid in cynomolgus monkeys are shown in Fig. 4C. The maximum (at 1 h) and minimum (at 8 h) plasma concentrations of probenecid were 76.8 ± 10.1 μg/ml (269 μM) and 23.3 ± 2.9 μg/ml (81.8 μM), respectively. Taking the unbound fraction in the plasma (13.1 ± 0.3%) into consideration, the maximum unbound concentration of probenecid (35 μM) in the monkey study was comparable with that observed in the human study (ca. 46 μM) (Inotsume et al., 1990).
Prediction of Renal Clearance of Famotidine by Allometric Scaling. The renal clearance and renal tubular secretion clearance of famotidine in rats (Lin et al., 1987), dogs (Boom et al., 1997), and monkeys were analyzed as a function of species body weight (W) using the simple allometric equation for interspecies scaling and used to predict these parameters in humans. The corresponding allometric equations based on three species data were CLrenal = 0.957 × W0.710 and CLsec = 0.609 × W0.679, respectively. The predicted values of CLrenal and CLsec based on a 70-kg body weight in humans were 19.5 and 10.9 l/h, respectively, which were very similar to the observed values (13.3–18.2 and 8.31–13.2 l/h) (Inotsume et al., 1990; Gladziwa and Klotz, 1993).
In the present study, we examined whether the drug-drug interaction between famotidine and probenecid could be reproduced in monkeys. The transport activities of the H2 receptor antagonists by mkOAT3 were compared with those by hOAT3, and the mRNA expression of the hOCT isoforms in the monkey kidney was quantified. To draw a definite conclusion, the effect of probenecid on the renal clearance of famotidine and cimetidine was examined in cynomolgus monkeys.
Famotidine was transported only by mkOAT3, whereas cimetidine and ranitidine are substrates of mkOAT1 and mkOAT3 (Fig. 1). The Km values for mkOAT3-mediated uptake were similar to those for hOAT3 (Table 1). Previously, it had been shown that the uptake of cimetidine relative to the uptake of benzylpenicillin was similar for mk- and hOAT3 (Tahara et al., 2005b). This also holds true for famotidine uptake. In addition, the relative activity of famotidine exhibited by mkOAT3 was greater than that by rOat3, consistent with previous results in humans (Tahara et al., 2005a). These results support the hypothesis that the transport of the H2 receptor antagonists by mkOAT3 is similar to that by hOAT3, rather than the rodent isoforms (Table 1).
OCT1 is the liver-specific isoform in humans, whereas it is expressed both in rodent liver and kidney (Motohashi et al., 2002; Slitt et al., 2002). This makes the contribution of organic cation transporters to renal uptake greater in rodents than in humans. Quantification of mRNA expression revealed that mkOCT1 and mkOCT2 are predominantly expressed in the liver and kidney, respectively, whereas mkOCT3 is expressed at considerably lower levels in these tissues. This expression patterns are similar to those in humans (Motohashi et al., 2002). Consequently, like human kidney, putative mkOCT2 plays a predominant role in the monkey kidney.
These results suggest that, as far as basolateral transporters are concerned, the monkey OATs and OCTs have similar characteristics to the human orthologs, and this prompted us to perform an in vivo study to obtain conclusive evidence. In monkeys, both famotidine and cimetidine are predominantly excreted into the urine, and the tubular secretion and glomerular filtrate rates that account for their renal clearance, are almost identical (Tables 2 and 3). When probenecid was given orally, the renal and renal tubular secretion clearance were reduced by 65 and 90%, respectively, resulting in a 2.0-fold increase in the AUC (Fig. 4A; Table 2). In addition, the steady-state distribution volume was reduced by 23% by probenecid. This is consistent with the previous findings in humans (Inotsume et al., 1990). It seems that the inhibition of the uptake by tissues, including the kidney, accounts for this effect. On the other hand, the plasma concentration and renal clearance or distribution clearance of cimetidine were not affected by probenecid (Fig. 4B; Table 3). Probenecid achieves a clinically relevant unbound concentration in monkey plasma (11–35 μM), which is sufficient to markedly inhibit mkOAT3, suggesting that the interaction could involve mkOAT3-mediated uptake, at least in part. Taking into account the degree of inhibition of tubular secretion clearance by probenecid, the probenecid-sensitive transporter, mkOAT3, plays a major role in the renal tubular secretion of famotidine, but not cimetidine, in monkeys. These results are consistent with the clinical studies (Gisclon et al., 1989; Inotsume et al., 1990). Consequently, monkeys, rather than rodents, can be used to predict drug-drug interactions involving tubular secretion, particularly when multiple transporters are involved.
The nonrenal clearance of famotidine was smaller in monkeys treated with probenecid than in the control group (Table 2). Although the difference was not statistically significant, it is likely that increasing the number of animals will make this difference statically significant. The presence of the oxidized metabolite of famotidine in human urine suggests that famotidine undergoes hepatic metabolism (Kroemer and Klotz, 1987). Because probenecid had no effect on OCT1, a candidate transporter responsible for the hepatic uptake of famotidine (Tahara eta al., 2005a), the reduction in the nonrenal clearance of famotidine by probenecid may be caused by inhibition of this metabolism. Currently, there is no report showing that probenecid has an inhibitory effect on metabolism, and this should be examined in future studies.
Interspecies scaling has been successfully used to predict human pharmacokinetic parameters from animals based on the concept of allometry (Lin, 1995; Reigner et al., 1997). It has been shown that simple allometric scaling of the renal clearance is a good predictor for drugs, such as methotrexate and several β-lactam antibiotics (Dedrick et al., 1970; Matsushita et al., 1990), although this is not the case for betamipron and enprofylline (Mahmood, 1998). The renal and renal tubular secretion clearances of famotidine in humans were estimated by simple allometric scaling using data from rats, dogs, and monkeys. A good predictability of the absolute values of the renal and renal tubular clearances from animal experiments (rat, dog and monkey) was observed, although the contribution of the renal transporters differs depending on the species. Therefore, good predictability by the allometric scaling cannot exclude the possibility of a species-dependent contribution by basolateral transporters. In particular, for the quantitative prediction of drug-drug interactions in humans, the contribution of transporters should be estimated using human materials. Alternatively, the relative activity factor method can be applied to predict the in vivo contribution of basolateral organic anion transporters using cDNA transfected cells (Hasegawa et al., 2003). In vivo studies in monkeys will further support the prediction of the occurrence of drug-drug interactions in humans. Nagata et al. (2004) found a drug-drug interaction involving rOat3 in the central nervous system (Nagata et al., 2004). Probenecid given as an i.v. constant infusion increased the concentrations of H2 receptor antagonists (also given as an i.v. constant infusion) in the cerebrospinal fluid by inhibiting OAT3-mediated efflux at the choroid plexus. The possibility of drug-drug interactions with probenecid involving efflux transport across the barriers of the central nervous system can be also examined in monkeys in future studies.
In conclusion, the drug-drug interactions between the H2 receptor antagonists (famotidine and cimetidine) and probenecid can be reproduced in monkeys. Hence, a combination of in vitro and in vivo studies could be a useful screening system for evaluating drug-drug interactions involving renal tubular transport in humans.
We thank Takashi Saito and Atsuko Takami (Pharmaceutical Research Institute, Kyowa Hakko Kogyo, Shizuoka, Japan) for helping with the monkey pharmacokinetic study.
- Received August 9, 2005.
- Accepted November 14, 2005.
This work was supported by Health and Labor Sciences Research Grants from the Ministry of Health, Labor, and Welfare for the Research on Advanced Medical Technology and on Regulatory Science of Pharmaceuticals and Medical Devices.
ABBREVIATIONS: OAT, organic anion transporter; OCT, organic cation transporter; h/mk/rOAT, human/monkey/rat OAT; h/mk/rOCT, human/monkey/rat OCT; HEK, human embryonic kidney; LC, liquid chromatography; MS, mass spectrometry; CMD, cimetidine; FMD, famotidine.
- The American Society for Pharmacology and Experimental Therapeutics