Advertisement
Research ArticleArticle

Extrapolation of In Vivo Hepatic Clearance from In Vitro Uptake Clearance by Suspended Human Hepatocytes for Anionic Drugs with High Binding to Human Albumin: Improvement of In Vitro-to-In Vivo Extrapolation by Considering the “Albumin-Mediated” Hepatic Uptake Mechanism on the Basis of the “Facilitated-Dissociation Model”

Soo-Jin Kim, Kyeong-Ryoon Lee, Seiji Miyauchi and Yuichi Sugiyama
Drug Metabolism and Disposition February 2019, 47 (2) 94-103; DOI: https://doi.org/10.1124/dmd.118.083733

Abstract

We investigated whether human serum albumin (HSA) in suspended human hepatocytes would affect the uptake clearance of anionic drugs with high binding to HSA and improve the extrapolation of in vivo hepatic clearance from in vitro uptake clearance by the hepatocytes via the “albumin-mediated” hepatic uptake mechanism. The uptake clearances for total forms (PSinf) and for unbound forms (PSu,inf) of 11 anionic drugs [all of which were organic anion-transporting polypeptide (OATP) substrates] were determined with suspended human hepatocytes in varying concentrations of HSA. The fraction of unbound drugs (fu) was determined using an equilibrium dialysis at the various HSA concentrations. The PSinf values decreased with increasing concentrations of HSA, whereas the unbound uptake clearances (PSu,inf(+) = PSinf/ fu) in the presence of HSA increased substantially, thus demonstrating the “albumin-mediated” hepatic uptake mechanism. The relationships between PSinf and HSA concentration were well described by the previously proposed facilitated-dissociation model, in which the drug–albumin complex interacts with the cell surface, enhancing the dissociation of the complex and providing unbound drug for hepatic uptake. Furthermore, the PSu,inf (+) values in in vivo conditions (at 5% HSA) were predicted from those obtained in isolated hepatocytes on the basis of the facilitated-dissociation model, revealing compatibility with the overall hepatic intrinsic clearance in vivo. We conclude that the “facilitated-dissociation” model is useful for describing the “albumin-mediated” hepatic uptake phenomenon of OATP drugs and to predict hepatic uptake clearance in vivo.

Introduction

It is very important that the pharmacokinetic features of new chemical entities in humans be adequately predicted during early stages of drug discovery and development. The prediction of human hepatic clearance is crucial because the liver is the major organ responsible for elimination of a variety of endogenous and exogenous chemicals via metabolism and/or biliary excretion. Currently, human-derived reagents, such as human hepatic microsomes, cytosols, and suspended hepatocytes, are available for in vitro studies, greatly improving the quantitative in vitro-to-in vivo extrapolation (IVIVE) for hepatic clearance values (Watanabe et al., 2011; Izumi et al., 2017). Compared with anion drugs, anionic drugs with high plasma binding tend to have poor predictive accuracy for human hepatic clearances, because they require mechanistic elucidation and quantitative improvement (Watanabe et al., 2011; Izumi et al., 2017).

Poulin et al. (2016) retrieved the in vitro and in vivo hepatic clearances of several drugs with high albumin binding and found that the conventional IVIVE procedure whose basis is the “free drug” hypothesis largely underestimated the in vivo hepatic clearances. Poulin et al. proposed an adjustment procedure to obtain the unbound drug concentration under in vivo conditions by considering differences in pH between intracellular and extracellular spaces and in albumin concentrations between the plasma and liver, as well as pH-partition mechanisms. This adjustment procedure improved the predictive accuracy of hepatic clearances for drugs that bind avidly to albumin.

We recently demonstrated that organic anion-transporting polypeptide (Oatp/OATP) substrates were taken up by primary cultured rat hepatocytes, and human suspended hepatocytes in the presence of albumin, to a much greater extent than could be expected on the basis of their unbound concentration and the “free drug” hypothesis (Miyauchi et al., 2018). The enhancement of the hepatic uptake (often referred to as “albumin-mediated” hepatic uptake) has been described in the “facilitated-dissociation” model, in which the drug–albumin complex interacts with the cell surface, enhancing the dissociation of the complex and providing unbound drug for hepatic uptake (Tsao et al., 1988a). Considering the “albumin-mediated” hepatic uptake, IVIVE for hepatic uptake clearance improves prediction accuracy. Using an isolated perfused rat liver system, Bounakta et al. (2018) and Poulin et al. (2017) also demonstrated that the IVIVE method considering the “albumin-mediated” hepatic uptake mechanism improved the predictive accuracy of in vivo hepatic clearance compared with their empirical IVIVE that considered pH and albumin concentration differences between the plasma and liver. Together, these results provide solid evidence supporting a kinetic model for describing the “albumin-mediated” hepatic uptake phenomenon for a drug avidly bound to albumin, which is important for robust IVIVE.

In the present study, our first objective was to investigate the effect of human serum albumin (HSA) on the uptake clearances of 11 drugs known as OATP substrates using suspended human hepatocytes. Our second objective was to improve the prediction of quantitative IVIVE for hepatic uptake clearance values by considering “albumin-mediated” hepatic uptake mechanisms on the basis of the “facilitated-dissociation” model.

Materials and Methods

Materials.

Pitavastatin calcium salt, atorvastatin calcium trihydrate, rosuvastatin (RSV) calcium salt, fluvastatin sodium salt, and pravastatin (PRV) sodium salt were purchased from Wako Pure Chemicals (Kyoto, Japan). Cerivastatin sodium salt and repaglinide were purchased from LKT Laboratories (St. Paul, MN). Glibenclamide (GLB), nateglinide, and valsartan were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Bosentan was purchased from Toronto Research Chemicals Inc. (North York, ON, Canada). Human plasma and HSA (fatty acid free) were purchased from MilliporeSigma (St. Louis, MO). All other reagents used were of analytical or high-performance liquid chromatography grade.

Unbound Fraction in Human Plasma and HSA.

Plasma protein binding of 11 compounds was determined in 100% human plasma and six different concentrations (0%, 0.125%, 0.25%, 0.5%, 1%, and 5% or 0, 18.8, 37.6, 75.2, 150, and 752 μM) of HSA by equilibrium dialysis with a Rapid Equilibrium Dialysis (RED) Device (Thermo Fisher Scientific, Rockford, IL). Various concentrations of HSA were prepared with transporter uptake buffer (pH 7.4). The final concentration of compounds in human plasma or HSA samples was 3 μM. Aliquots (100 μl) of each sample and phosphate-buffered saline or transporter uptake buffer (300 μl) were added to the equilibrium dialysis device. The plate was put into a 37°C incubator and shaken gently for 20 hours. Aliquots of samples (25-μl) and 50-μl aliquots of buffers were then taken and mixed with acetonitrile containing an internal standard. The mixture was vigorously mixed and centrifuged at 4°C. An aliquot (5 μl) of each supernatant was injected into a liquid chromatography–tandem mass spectrometry (LC–MS/MS) system (Supplemental Table S1).

Uptake Studies in Suspended Human Hepatocytes.

The uptake rate of all compounds was determined using pooled cryopreserved human hepatocytes (Lot TFF, 20 mixed-gender donors) (BioIVT, Baltimore, MD) by a centrifugal filtration technique described previously (Hirano et al., 2004). Immediately prior to the uptake experiment, the cryopreserved hepatocytes (1 ml suspension) were thawed at 37°C, then suspended in 50 ml of prewarmed cryopreserved hepatocyte recovery media thawing medium (APSciences, Inc., Columbia, MD) and centrifuged (100 g) for 2 minutes at 4°C, followed by removal of the supernatants. The hepatocytes were gently resuspended in ice-cold Krebs-Henseleit buffer to give 2.0 × 106 viable cells/ml on ice. The uptake studies were initiated by adding an equal volume of prewarmed buffer containing two sets of drugs: 1) pitavastatin, atorvastatin, RSV, fluvastatin, cerivastatin, and PRV; 2) glibenclamide, valsartan, repaglinide, bosentan, and nateglinide, with six different concentrations of HSA (0%, 0.125%, 0.25%, 0.5%, 1%, and 5%), added to the cell suspension after a 5-minute preincubation at 37°C, resulting in a final substrate concentration of 3 μM in 1 × 106 cells/ml. The reported Michaelis constant for the transporter (Km) value of 11 drugs ranged from 0.18 to 76.7 μM (Supplemental Table S2). The final free concentration of substrate in the presence of HSA would still be lower than 3 μM.

After incubation for 0.25 and 1.25 minutes, the reactions were terminated by separating the cells from the substrate solution. The hepatic uptake study was conducted using an oil-spin method described previously in detail (Miyauchi et al., 2018). The concentrations of medium and cells were analyzed by LC–MS/MS (Supplemental Table S1). Our results for the uptake clearances in the absence of HSA were comparable with reported values (Supplemental Table S2), indicating that the competition among drugs in the cassette mixture was minimal in our hepatic uptake experimental conditions.

Kinetic Analysis.

The unbound fraction (fu) for each drug in steady-state conditions using equilibrium dialysis was calculated using the following equation:Embedded Image(1)where n, [Alb] and Kd represent the number of binding sites on the albumin, the HSA concentration (micromolar), and the dissociation constant (micromolar) between drug and HSA, respectively. The n values for some OATP1B1 substrates such as statins, antidiabetic drugs, and anti-HIV drugs have been reported to be 0.87–1.1 (Seedher and Kanojia, 2008; Gulati et al., 2009; Shi et al., 2017). For the simplicity of eq. 1, the n value was assumed to be unity. The molecular concentration of HSA was calculated using 66,500 as the molecular weight.

The initial uptake rates were obtained from the slopes of the time courses within 0.25–1.25 minutes using linear regression analysis. The uptake clearances for total (PSinf) and unbound forms (PSu,inf) of each drug were calculated by dividing the initial uptake velocity by the total and unbound drug concentrations in the incubation buffer at the various concentrations of HSA.

In our previous study, the “albumin-mediated” hepatic transport phenomenon was analyzed using the “facilitated-dissociation” model, in which the interaction between the cell surface and the drug–albumin complex enhances the dissociation of drug from the complex (Tsao et al., 1988a). A facilitation mechanism has been postulated to the effect that the interaction of the drug–albumin complex with the surface of hepatocytes induces a conformational change in albumin, resulting in enhanced dissociation of drug from the complex, providing unbound drug near the surface. Therefore, the uptake of highly albumin-bound drugs by hepatocytes includes the pathway of unbound drugs and the additional pathway of drugs dissociated from drug–albumin complexes near the surface. We assumed that unbound albumin and drug-bound albumin compete for the same binding site(s) on the surface of hepatocytes with the same affinity, which is the dissociation constant (Kd,m) between albumin and the surface of hepatocytes. The fraction of albumin bound to the surface of the hepatocytes (λ) is expressed as:Embedded Image(2)where Embedded Image represents the capacity of albumin-binding sites on the surface of hepatocytes. The equations for the uptake rate (v) of drug derived by Tsao et al. (1988a) are as follows:Embedded Image(3)Embedded Image(4)Substituting eqs. 1 and 2 into eq. 4 yieldsEmbedded Image(5)where Embedded Image represents the total concentration of drug; PSinf, PSu,inf, and PSb,inf represent the uptake clearance for the total drug, unbound drug, and unbound drug dissociated from the drug–albumin complex at the surface, respectively. On the basis of the “facilitated-dissociation” model, the uptake clearance of unbound drug (PSinf (free)) and the uptake clearance of unbound drug associated from the drug–albumin complex (PSinf (bound)) at the cell surface were estimated from the first and second terms of the right-hand side of eq. 5, respectively, as follows.Embedded Image(6)Embedded Image(7)In eq. 7, Embedded Image represents the parameter of the clearance capacity for the uptake of “facilitated-dissociation” drugs, and is designated as Embedded Image. The unbound uptake clearance in the presence of HSA, (PSu,inf (+)), was calculated using the following equation modified from eqs. 1 and 5.Embedded Image(8)This equation implies that PSu,inf (+) shows saturation as [Alb] increases. The Embedded Image values of each compound were estimated using eq. 1, and PSu,inf, Kd,m, and VB,max were obtained with eq. 5 using nonlinear least-squares fitting software (Napp, version 2.31, The University of Tokyo Hospital, Japan) (Hisaka and Sugiyama, 1998).

Extrapolation to In Vivo Hepatic Clearances from the In Vitro Uptake Clearance by Suspended Hepatocytes.

The in vitro unbound hepatic uptake clearance (microliters per 106 cells per milliliter) was scaled up to in vivo (milliliters per minute per kilogram body weight) using the following physiologic scaling factors: 1.2 × 108 cells/g liver and 25.7 g liver/kg body weight (Davies and Morris, 1993; Miyauchi et al., 1993).

In vivo hepatic overall intrinsic clearance, representing the elimination of unbound drug (Embedded Image) from circulating blood, was calculated from the reported in vivo hepatic clearance (Supplemental Table S3) using the dispersion model (eqs. 9–12) (Roberts and Rowland, 1986) and the well-stirred model (eqs. 9 and 13) (Pang and Rowland, 1977).Embedded Image(9)For the dispersion model:Embedded Image(10)Embedded Image(11)Embedded Image(12)For the well-stirred model:Embedded Image(13)Where CLh,B, Qh, DN, and fB are the hepatic clearance, hepatic blood flow rate (20.7 ml/min per kilogram body weight) (Davies and Morris, 1993), the dispersion number (0.17) (Roberts and Rowland, 1986), and the unbound fraction in the blood, respectively. For almost all of the drugs used, it has been suggested that hepatic uptake is a rate-determining step for hepatic elimination (Watanabe et al., 2010; Izumi et al., 2017). According to the extended clearance concept, in vivo hepatic intrinsic clearance for 11 OATP substrates could be well accounted for by the in vitro hepatic uptake clearance, regardless of the involvement of hepatic metabolism.

Results

Unbound Fractions in Human Plasma and HSA.

We determined the unbound fraction (fp) of 11 drugs in human plasma. These values were used for the calculation of blood unbound fraction as fp/RB (where RB is the blood partitioning; Table 2). The fu values were determined in the presence of HSA at six concentrations (0%, 0.125%, 0.25%, 0.5%, 1%, and 5%). The fp values for almost all drugs except RSV and PRV were less than 0.05 (0.000787–0.0308), whereas those for RSV and PRV were 0.134 and 0.563, respectively. The fp values were very close to fu values in the presence of 5% HSA, a normal albumin concentration (Table 2). The Kd values of drugs were obtained by nonlinear least-squares fitting using eq. 1 (Fig. 1). For nine of the drugs, these were 0.490–27.1 μM, and for RSV and PRV they were 94.5 and 667 μM, respectively (Table 1). Most of the drugs used here are highly bound to albumin in plasma (bound form >95%), whereas two drugs, RSV and PRV, showed low protein binding, with fp values of 0.134 for RSV and 0.563 for PRV (Fig. 1; Table 1). These results were consistent with the fp values reported previously (Colussi et al., 1997; Hatorp, 2002; Blanchard et al., 2005; Watanabe et al., 2010, 2011), indicating that the major binding protein in the plasma is indeed albumin. The recovery of the measurement for 11 drugs in the uptake experiments was more than 92%, irrespective of HSA in the buffer. The nonspecific adsorption to the wall of test tubes was assumed to be minimal.

Fig. 1.
Fig. 1.

The unbound fraction (fu) of 11 drugs in HSA at 0%, 0.125%, 0.25%, 0.5%, 1%, and 5%. The filled circles represent the observed HSA fu values (mean ± S.D., n = 4) in the absence and presence of HSA, which concentrations were 18.8, 37.6, 75.2, 150, and 752 μM. The solid lines represent the fitted line estimated from eq. 1. (A) and (B) show sets A and B, respectively. Key: PTV, pitavastatin; ATV, atorvastatin; FLV, fluvastatin; CRV, cerivastatin; GLB, glibenclamide; VST, valsartan; RPG, repaglinide; BOS, bosentan; NTG, nateglinide.

View this table:
TABLE 1

 Unbound fractions in plasma and 5% HSA, and the estimated contribution of “albumin-mediated” uptake of 11 drugs based on Tsao’s model

Data are presented as the mean or mean ± S.D.

Effect of Albumin on Uptake Clearances by Isolated Human Hepatocytes.

The PSinf (hepatic uptake clearance for total drug) values of drugs at various concentrations of HSA (0.125%, 0.25%, 0.5%, and 1%) were determined from the initial slopes of the uptake volume (Supplemental Fig. S1). However, the PSinf value in the presence of 5% HSA could not be determined robustly because of the avid binding of these drugs to albumin, resulting in no significant time-dependent uptake volume by hepatocytes over 0.25–1.25 minutes. The PSu,inf (+) (the hepatic unbound uptake clearance in the presence of the albumin) value was estimated by dividing the PSinf value by the fu value at the various concentrations of HSA, whereas the PSu,inf value was obtained in the absence of HSA. The PSinf values of almost all drugs except PRV were decreased in association with increases in the concentration of HSA, whereas the PSu,inf (+) values were increased substantially (Figs. 2 and 3; Supplemental Fig. S2), thus demonstrating that these 10 drugs with high binding to the albumin show “albumin-mediated” hepatic uptake. Nevertheless, PRV with its low protein binding to albumin (fp = 0.563) did not show any change in the PSinf and PSu,inf (+) values, irrespective of the concentration of HSA (Fig. 3; Supplemental Fig. S2). Furthermore, the relationships between the PSinf values and HSA concentrations for the 10 drugs except for PRV were simultaneously fitted to eq. 5 on the basis of the “facilitated-dissociation” model (Tsao et al., 1988a), and the estimated Embedded Image value was 45.2 ±13.0 μM (±calculated S.D.). The PSu,inf values obtained in the absence of HSA were very close to those of PSu,inf estimated in the fitting analysis by the “facilitated-dissociation” model. These estimated parameters are summarized in Table 1. The PSinf values in the presence of a normal physiologic albumin concentration (5%) were predicted according to eq. 5 by the “facilitated-dissociation” model. The predicted PSinf values were much greater than those obtained on the basis of the “free drug” hypothesis (Fig. 3). The contributions of “albumin-mediated” uptake clearance (PSinf (bound)) to total uptake clearance for these 10 drugs were estimated to be 55–98%, according to eqs. 6 and 7 (Table 1).

Fig. 2.
Fig. 2.

The uptake clearances for total form (Embedded Image) (A) and the unbound uptake clearance in the presence of HSA [Embedded Image] (B) of pitavastatin (PTV) in suspended human hepatocytes. The filled circles, solid lines, and broken lines represent the observed or calculated values (mean ± S.D., n = 8), the fitted line by Tsao’s model, and the theoretical line with the “free drug” theory as a basis, respectively.

Fig. 3.
Fig. 3.

The observed uptake clearances for total form (Embedded Image) of clinical OATP substrates in suspended human hepatocytes. The filled circles, solid lines, and broken lines represent the observed Embedded Image (mean ± S.D., n = 8), the fitted line by Tsao’s model (eq. 5), and the theoretical line with the “free drug” theory as a basis, respectively. (A) and (B) represent sets A and B, respectively. Key as in Fig. 1. Notes: 1) When fitting for RSV by Tsao’s model, only Embedded Image and Embedded Image were estimated with a fixed Embedded Image value (Table 1). 2) The fitting by Tsao’s model was not performed for PRV.

Extrapolation to In Vivo Hepatic Clearances from In Vitro Uptake Data.

The PSu,inf values of 10 drugs were extrapolated to the PSu,infvitro values (ml/min/kg body weight). As shown in Table 2, the fB ⋅ PSu,infvitro values whose basis was the “free drug” hypothesis were underestimated substantially compared with the fB ⋅ CLh,u,int,allvivo (the in vivo unbound overall hepatic intrinsic clearance) values obtained from hepatic clearance by the dispersion model (Fig. 4A) and the well-stirred model (Fig. 5A). The PSu,inf (+) values for 10 drugs at 5% HSA estimated on the basis of the “facilitated-dissociation” model (eq. 8) were also extrapolated to the in vivo value per body weight (PSu,infvitro, ml/min/kg) (Table 2). The fB ⋅ PSu,infvitro values at 5% HSA showed similarities to the fB ⋅ CLvivoh,u,int,all values estimated from the hepatic clearance regardless of the dispersion or well-stirred model (Figs. 4B and 5B; Table 2). The “albumin-mediated” uptake factors or enhancement of the hepatic uptake clearance (R) in the physiologic albumin concentration range were estimated as the ratios of PSu,inf (+) values at 5% HSA to Embedded Image values in the absence of HSA; the estimated R values were 2.44–63.8 (Table 2). As shown in Figs. 4B and 5B, the “albumin-mediated” uptake factor improved the prediction of IVIVE, but the relationship between fB ⋅ PSu,infvitro and fB ⋅ CLh,u,int,allvivostill showed some underestimation in the IVIVE from the in vitro hepatic uptake clearance. The ratios of fB ⋅ CLvivoh,u,int,all whose basis was the dispersion and well-stirred models to fB ⋅ PSvitrou,inf at 5% HSA were 0.63–5.57 and 0.63–7.11, respectively, and the average value was 2.44 and 3.19, respectively (Table 2). Furthermore, the fB ⋅ PSu,infvitro values at 5% HSA were corrected by the mean value (2.44 and 3.19) designated as scaling factor (SF) , giving a better relationship between fB ⋅ PSvitrou,inf ⋅ SF and fB ⋅ CLvivoh,u,int,all within a 3-fold range of difference regardless of estimated models of fB ⋅ CLvivoh,u,int,all (Fig. 4C; Table 2).

View this table:
TABLE 2

 Kinetic parameters for drug uptake in suspended hepatocytes

Data are presented as means.

Fig. 4.
Fig. 4.

Comparison between the uptake clearance (Embedded Image) at 0% HSA (A), the predicted Embedded Image at 5% HSA (B), and the adjusted Embedded Image by SF (C) in suspended human hepatocytes and the observed hepatic overall intrinsic clearance (Embedded Image) by the dispersion model. The Embedded Image at 5% HSA was predicted by Tsao’s model, the adjusted Embedded Image at 5% HSA was calculated by the predicted Embedded Image, and SF is 2.44, the average value of the ratios of Embedded Image to Embedded Image at 5% HSA. The dashed and dotted lines represent the line of unity and the lines of the 1:5 and 5:1 correlations, respectively. Key as in Fig. 1.

Fig. 5.
Fig. 5.

Comparison between the uptake clearance (Embedded Image) at 0% HSA (A), the predicted Embedded Image at 5% HSA (B), and the adjusted Embedded Image by SF (C) in suspended human hepatocytes and the observed hepatic overall intrinsic clearance (Embedded Image) by the well-stirred model. The Embedded Image at 5% HSA was predicted by Tsao’s model, the adjusted Embedded Image at 5% HSA was calculated by the predicted Embedded Image, and SF is 3.19, the average value of the ratios of Embedded Image to Embedded Image at 5% HSA. The dashed and dotted lines represent the line of unity and the lines of the 1:5 and 5:1 correlations, respectively. Key as in Fig. 1.

Discussion

Human-derived reagents, such as hepatic microsomes, cytosols and suspended hepatocytes, are now available for in vitro studies (Chiba et al., 2009; Shitara et al., 2013), resulting in success in the prediction of hepatic clearances for new chemical entities at early stages of drug discovery and development. Here, we investigated whether the in vitro hepatic uptake clearance by suspended human hepatocytes could be extrapolated quantitatively to estimate the in vivo hepatic clearances for 11 drugs known as OATP substrates.

Previously, in vitro hepatic uptake clearance with cryopreserved human hepatocytes showed good prediction of in vivo hepatic clearance for some OATP substrates (Watanabe et al., 2010, 2011; Kimoto et al., 2012). Nevertheless, the predictions of human hepatic clearances in some anionic drugs with high binding to albumin were poor, showing underestimation of IVIVE for hepatic uptake clearances. Consistent with this notion, Figs. 4A and 5A show the underestimation of IVIVE for the hepatic uptake clearances regardless of estimated models of fB ⋅ CLh,u,int,allvivo. We hypothesized that “albumin-mediated” hepatic uptake would result in poor predictions of IVIVE because uptake studies in human hepatocytes have usually been conducted in the absence of serum albumin. Recently, using primary cultured rat hepatocytes and suspended human hepatocytes, we demonstrated that IVIVE for hepatic clearance for Oatp/OATP substrates with high protein binding was robustly improved by an “albumin-mediated” hepatic uptake mechanism considered in the analysis (Miyauchi et al., 2018), thereby suggesting that the underestimation in IVIVE could be attributed to an “albumin-mediated” hepatic uptake mechanism. In the present study, we further investigated the effect of HSA on the hepatic uptake clearance of 11 anionic OATP substrates by using suspended human hepatocytes in the absence and presence of various concentrations of HSA, and the improvement in the prediction of quantitative IVIVE for hepatic uptake clearance values by considering “albumin-mediated” hepatic uptake mechanisms that used the “facilitated-dissociation” model as a basis.

The PSinf values of 11 drugs by human hepatocytes were determined in the presence of HSA, and the PSu,inf (+) values were obtained by dividing the PSinf by fu for the various concentrations of HSA. Except for PRV, all drugs showed decreases in the PSinf values along with increases in HSA concentration (Figs. 2 and 3), whereas there was a substantial increase in PSu,inf (+) value (Fig. 2B; Supplemental Fig. S2), which invalidated the “free drug” hypothesis, suggesting that “albumin-mediated” hepatic uptake for various OATP substrates with high protein binding was also observed in human hepatocytes. Nevertheless, PRV with low protein binding to albumin (fp = 0.563) did not show any change in the PSinf and PSu,inf (+) values, irrespective of the concentrations of HSA (Fig. 3; Supplemental Fig. S2).

In the present study, the PSu,inf (+) was composed of the uptake clearances for unbound drug and unbound drug dissociated from the drug–albumin complex near the cell surface (eq. 8). If “albumin-mediated” uptake is involved in hepatic uptake, the PSu,inf (+) value would be greater than the PSu,inf value by the uptake clearance of unbound drug dissociated from the drug–albumin complex. It should be noted that PSu,inf (+) is a hybrid parameter accounting for the “albumin-mediated” uptake clearance attributable to the hepatic uptake of unbound drug dissociated from the drug–albumin complex.

Previously, the hepatic uptake of a number of endogenous and exogenous ligands and OATP substrates has been shown to exhibit the kinetics of “albumin-mediated” hepatic transport (Weisiger et al., 1981, 1984; Forker et al., 1982; Forker and Luxon, 1983; Tsao et al., 1986, 1988b; Pond et al., 1992; Fujino et al., 2018; Miyauchi et al., 2018). We have analyzed the “albumin-mediated” hepatic uptake transport, which uses the “facilitated-dissociation” model as a basis, in which the interaction of the drug–albumin complex with the surface of hepatocytes enhances the dissociation of the drug from the complex and provides unbound drug that is more available for hepatic uptake (Tsao et al., 1988a). The relationships between the PSinf values and HSA concentrations in all 10 drugs except PRV were simultaneously fitted to eq. 5. Embedded Image, a drug-independent parameter showing the dissociation constant between the albumin (unbound albumin or drug-bound albumin) and the surface of hepatocytes, was estimated to be 45 μM regardless of the drug tested, which is consistent with results reported in other studies (Weisiger et al., 1981; Forker and Luxon, 1983). VB,max, a drug-dependent parameter showing the clearance capacity for the uptake of “facilitated-dissociation” drug, was estimated at various values (Table 1). The PSinf values at the normal blood concentration of HSA (5%) are a key parameter of the quantitative IVIVE of hepatic uptake clearances; unfortunately, these values could not be determined robustly because the drugs bind avidly to the albumin, resulting in no time-dependent uptake volume by hepatocytes during 0.25–1.25 minutes in the presence of 5% HSA. To predict IVIVE quantitatively for hepatic uptake clearances by isolated hepatocytes, we therefore estimated PSinf in the presence of 5% HSA according to eq. 5 on the basis of the “facilitated-dissociation” model (Table 2). The PSinf values with the “facilitated-dissociation” model as a basis were higher than values whose basis was the “free drug” hypothesis (Fig. 3; Table 2). The contributions of “albumin-mediated” uptake to the total uptake clearance of 10 drugs were 54%–98%, whereas rosuvastatin with its relatively low protein binding property (0.134 of fp) exhibited the minimum value for “albumin-mediated” hepatic uptake contribution (Table 1). In light of these findings, albumin facilitates the hepatic uptake of the bound form of OATP substrate drugs with high protein binding (approximately >90%), a major pathway of hepatic uptake in the normal range of albumin concentration in the blood, thereby violating the “free drug” hypothesis. Therefore, we believe that the “albumin-mediated” hepatic uptake mechanism may improve the prediction accuracy for IVIVE.

Regardless of estimated models of fB ⋅ CLvivoh,u,int,all, the fB ⋅ PSvitrou,inf values estimated from in vitro PSu,inf (+) values at 5% HSA correlated better with the fB ⋅ CLvivoh,u,int,all values estimated from the hepatic clearances (Fig. 4B; Fig. 5B; Table 2), suggesting that the underestimation of IVIVE for the hepatic uptake clearances could be adjusted sufficiently by “albumin-mediated” uptake factors, the ratio of fB ⋅ PSvitrou,inf at 0% HSA to that at 5% HSA (R), or by enhancement of the hepatic uptake clearance in the normal range of albumin concentrations; the R values were estimated to be 2.44–63.8 (Table 2). Nevertheless, the relationship between fB ⋅ PSvitrou,inf and fB ⋅ CLvivoh,u,int,all continued to show slight underestimation of IVIVE from the in vitro hepatic uptake clearances, although the “albumin-mediated” uptake factor improved the prediction. The ratios of fB ⋅ CLvivoh,u,int,all with the dispersion and well-stirred models as bases to fB ⋅ PSvitrou,inf at 5% HSA were 0.63–5.57 and 0.63–7.11, respectively, and the mean values were 2.44 and 3.19, respectively (Table 2). The ratios of the in vivo-to-in vitro for hepatic clearance exhibited substrate dependency. Although drugs are only taken up by OATP1B1, the affinities for OATP1B1 of each drug ranged widely from low to high (the reciprocal of the Km values, summarized in Supplemental Table S2). Furthermore, the relationship between the values of “albumin-mediated” uptake factors (R) and the affinities of drug for OATP1B1 exhibited a tendency by which the higher the affinity for the transporter, the more albumin effectively enhanced uptake (Miyauchi et al., 2018). Furthermore, the fB ⋅ PSvitrou,inf values at 5% HSA were corrected by the mean value designated as SF, resulting in a better relationship between fB ⋅ PSvitrou,inf ⋅ SF and fB ⋅ CLvivoh,u,int,all within a 3-fold range of differences (Fig. 4C; Fig. 5C; Table 2) for both the dispersion and well-stirred models.

Alternatively, the underestimation of IVIVE for hepatic uptake clearance also could be partially explained by multifactorial mechanisms, including physiologic factors such as pH (Poulin and Haddad, 2015) and albumin concentration in the interstitial fluids (Poulin et al., 2016), experimental conditions (Badolo et al., 2011), and kinetic models describing the liver disposition (Iwatsubo et al., 1996). As shown in Figs. 4B and 5B, some discrepancy remained between fB ⋅ CLh,u,int,allvivo and fB ⋅ PSu,infvitro, and the SF values could be made up with various factors described above. The comprehensive review by Bowman and Benet (2018) has investigated the plausible mechanisms involved in the SF value of IVIVE for hepatic uptake clearance. Determining how these factors are involved in the SF values requires further investigation.

Recently, we reported an improvement in the accuracy of prediction of intrinsic hepatic clearance for three antidiabetic drugs that are substrates of both OATP1Bs and CYP2Cs by considering the “albumin-mediated” uptake in the in vitro measurements of OATP1B-mediated hepatic uptake and CYP2C-mediated metabolism (Fujino et al., 2018). Therefore, it is important to note that a kinetic model for describing the “albumin-mediated” hepatic uptake phenomenon for a drug avidly bound to albumin is useful for its robust IVIVE.

In conclusion, albumin facilitates the hepatic uptake of anionic drugs known as OATP substrates with high protein binding to hepatocytes, and the predicted PSu,inf (+) values at 5% HSA estimated by the “facilitated-dissociation” model provide a better extrapolation for in vivo hepatic intrinsic clearances. Furthermore, the ratios of fB ⋅ CLh,u,int,allvivo to fB ⋅ PSvitrou,inf at 5% HSA or the “albumin-mediated” factors are applicable and useful for quantifying IVIVE from the hepatic uptake clearances by isolated human hepatocytes.

Authorship Contributions

Participated in research design: Kim, Lee, Miyauchi, Sugiyama.

Conducted experiments: Kim, Lee.

Contributed new reagents or analytic tools: Kim, Lee, Miyauchi, Sugiyama.

Performed data analysis: Kim, Lee, Miyauchi, Sugiyama.

Wrote or contributed to the writing of the manuscript: Kim, Lee, Miyauchi, Sugiyama.

Footnotes

    • Received July 31, 2018.
    • Accepted November 27, 2018.

  • ↵1 S.-J.K. and K.-R.L. contributed equally to this work.

  • ↵2 Current affiliation: Drug Evaluation Center, R&D Institute, CJ HealthCare, Majang-myeon, Icheon, Korea.

  • ↵3 Current affiliation: Laboratory Animal Resource Center, Korea Research Institute of Bioscience and Biotechnology, Cheongwon-gu, Cheongju, Korea.

  • This study was supported by Grant-in-Aid for Scientific Research (C) and (S) from the Japanese Ministry of Education, Culture, Sports, Sciences, and Technology [Grant 26460044] and [Grant 24229002], respectively.

  • https://doi.org/10.1124/dmd.118.083733.

  • ↵Embedded ImageThis article has supplemental material available at dmd.aspetjournals.org.

Abbreviations

CLvivoh,u,int,all
the in vivo unbound overall hepatic intrinsic clearance
fp
plasma-unbound fraction
fu
HSA-unbound fraction
HSA
human serum albumin
IVIVE
in vitro-to-in vivo extrapolation
OATP
organic anion transporting polypeptide
PSb,inf
the hepatic uptake clearance for the unbound drug dissociated from the drug–albumin complex at the surface
PSinf
the hepatic uptake clearance for total drug
PSu,inf
the hepatic uptake clearance for unbound drug
PSu,inf (+)
the hepatic unbound uptake clearance in the presence of the albumin
PRV
pravastatin
R
hepatic uptake clearance
RSV
rosuvastatin
SF
scaling factor

References

  1. ↵
    1. Badolo L,
    2. Trancart MM,
    3. Gustavsson L, and
    4. Chesné C
    (2011) Effect of cryopreservation on the activity of OATP1B1/3 and OCT1 in isolated human hepatocytes. Chem Biol Interact 190:165–170.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Blanchard N,
    2. Alexandre E,
    3. Abadie C,
    4. Lavé T,
    5. Heyd B,
    6. Mantion G,
    7. Jaeck D,
    8. Richert L, and
    9. Coassolo P
    (2005) Comparison of clearance predictions using primary cultures and suspensions of human hepatocytes. Xenobiotica 35:1–15.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Bounakta S,
    2. Bteich M,
    3. Mantha M,
    4. Poulin P, and
    5. Haddad S
    (2018) Predictions of bisphenol A hepatic clearance in the isolated perfused rat liver (IPRL): impact of albumin binding and of co-administration with naproxen. Xenobiotica 48:135–147.
    OpenUrl
  4. ↵
    1. Bowman CM and
    2. Benet LZ
    (2018) An examination of protein binding and protein-facilitated uptake relating to in vitro-in vivo extrapolation. Eur J Pharm Sci 123:502–514.
    OpenUrl
  5. ↵
    1. Chiba M,
    2. Ishii Y, and
    3. Sugiyama Y
    (2009) Prediction of hepatic clearance in human from in vitro data for successful drug development. AAPS J 11:262–276.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Colussi DM,
    2. Parisot C,
    3. Rossolino ML,
    4. Brunner LA, and
    5. Lefèvre GY
    (1997) Protein binding in plasma of valsartan, a new angiotensin II receptor antagonist. J Clin Pharmacol 37:214–221.
    OpenUrlPubMed
  7. ↵
    1. Davies B and
    2. Morris T
    (1993) Physiological parameters in laboratory animals and humans. Pharm Res 10:1093–1095.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Forker EL and
    2. Luxon BA
    (1983) Albumin-mediated transport of rose bengal by perfused rat liver. Kinetics of the reaction at the cell surface. J Clin Invest 72:1764–1771.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Forker EL,
    2. Luxon BA,
    3. Snell M, and
    4. Shurmantine WO
    (1982) Effect of albumin binding on the hepatic transport of rose bengal: surface-mediated dissociation of limited capacity. J Pharmacol Exp Ther 223:342–347.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Fujino R,
    2. Hashizume K,
    3. Aoyama S,
    4. Maeda K,
    5. Ito K,
    6. Toshimoto K,
    7. Lee W,
    8. Ninomiya SI, and
    9. Sugiyama Y
    (2018) Strategies to improve the prediction accuracy of hepatic intrinsic clearance of three antidiabetic drugs: application of the extended clearance concept and consideration of the effect of albumin on CYP2C metabolism and OATP1B-mediated hepatic uptake. Eur J Pharm Sci 125:181–192.
    OpenUrl
  11. ↵
    1. Gulati A,
    2. Boudinot FD, and
    3. Gerk PM
    (2009) Binding of lopinavir to human α1-acid glycoprotein and serum albumin. Drug Metab Dispos 37:1572–1575.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Hatorp V
    (2002) Clinical pharmacokinetics and pharmacodynamics of repaglinide. Clin Pharmacokinet 41:471–483.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Hirano M,
    2. Maeda K,
    3. Shitara Y, and
    4. Sugiyama Y
    (2004) Contribution of OATP2 (OATP1B1) and OATP8 (OATP1B3) to the hepatic uptake of pitavastatin in humans. J Pharmacol Exp Ther 311:139–146.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Hisaka A and
    2. Sugiyama Y
    (1998) Analysis of nonlinear and nonsteady state hepatic extraction with the dispersion model using the finite difference method. J Pharmacokinet Biopharm 26:495–519.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Iwatsubo T,
    2. Hirota N,
    3. Ooie T,
    4. Suzuki H, and
    5. Sugiyama Y
    (1996) Prediction of in vivo drug disposition from in vitro data based on physiological pharmacokinetics. Biopharm Drug Dispos 17:273–310.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Izumi S,
    2. Nozaki Y,
    3. Komori T,
    4. Takenaka O,
    5. Maeda K,
    6. Kusuhara H, and
    7. Sugiyama Y
    (2017) Comparison of the predictability of human hepatic clearance for organic anion transporting polypeptide substrate drugs between different in vitro-in vivo extrapolation approaches. J Pharm Sci 106:2678–2687.
    OpenUrl
  17. ↵
    1. Kimoto E,
    2. Yoshida K,
    3. Balogh LM,
    4. Bi YA,
    5. Maeda K,
    6. El-Kattan A,
    7. Sugiyama Y, and
    8. Lai Y
    (2012) Characterization of organic anion transporting polypeptide (OATP) expression and its functional contribution to the uptake of substrates in human hepatocytes. Mol Pharm 9:3535–3542.
    OpenUrlCrossRef
  18. ↵
    1. Miyauchi S,
    2. Masuda M,
    3. Kim SJ,
    4. Tanaka Y,
    5. Lee KR,
    6. Iwakado S,
    7. Nemoto M,
    8. Sasaki S,
    9. Shimono K,
    10. Tanaka Y, et al.
    (2018) The phenomenon of albumin-mediated hepatic uptake of organic anion transport polypeptide substrates: prediction of the in vivo uptake clearance from the in vitro uptake by isolated hepatocytes using a facilitated-dissociation model. Drug Metab Dispos 46:259–267.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Miyauchi S,
    2. Sawada Y,
    3. Iga T,
    4. Hanano M, and
    5. Sugiyama Y
    (1993) Comparison of the hepatic uptake clearances of fifteen drugs with a wide range of membrane permeabilities in isolated rat hepatocytes and perfused rat livers. Pharm Res 10:434–440.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Pang KS and
    2. Rowland M
    (1977) Hepatic clearance of drugs. II. Experimental evidence for acceptance of the “well-stirred” model over the “parallel tube” model using lidocaine in the perfused rat liver in situ preparation. J Pharmacokinet Biopharm 5:655–680.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Pond SM,
    2. Davis CKC,
    3. Bogoyevitch MA,
    4. Gordon RA,
    5. Weisiger RA, and
    6. Bass L
    (1992) Uptake of palmitate by hepatocyte suspensions: facilitation by albumin? Am J Physiol 262:G883–G894.
    OpenUrlPubMed
  22. ↵
    1. Poulin P,
    2. Bteich M, and
    3. Haddad S
    (2017) Supplemental analysis of the prediction of hepatic clearance of binary mixtures of bisphenol a and naproxen determined in an isolated perfused rat liver model to promote the understanding of potential albumin-facilitated hepatic uptake mechanism. J Pharm Sci 106:3207–3214.
    OpenUrl
  23. ↵
    1. Poulin P,
    2. Burczynski FJ, and
    3. Haddad S
    (2016) The role of extracellular binding proteins in the cellular uptake of drugs: impact on quantitative in vitro-to-in vivo extrapolations of toxicity and efficacy in physiologically based pharmacokinetic-pharmacodynamic research. J Pharm Sci 105:497–508.
    OpenUrl
  24. ↵
    1. Poulin P and
    2. Haddad S
    (2015) Albumin and uptake of drugs in cells: additional validation exercises of a recently published equation that quantifies the albumin-facilitated uptake mechanism(s) in physiologically based pharmacokinetic and pharmacodynamic modeling research. J Pharm Sci 104:4448–4458.
    OpenUrl
  25. ↵
    1. Roberts MS and
    2. Rowland M
    (1986) A dispersion model of hepatic elimination: 1. Formulation of the model and bolus considerations. J Pharmacokinet Biopharm 14:227–260.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Seedher N and
    2. Kanojia M
    (2008) Reversible binding of antidiabetic drugs, repaglinide and gliclazide, with human serum albumin. Chem Biol Drug Des 72:290–296.
    OpenUrlPubMed
  27. ↵
    1. Shi JH,
    2. Wang Q,
    3. Pan DQ,
    4. Liu TT, and
    5. Jiang M
    (2017) Characterization of interactions of simvastatin, pravastatin, fluvastatin, and pitavastatin with bovine serum albumin: multiple spectroscopic and molecular docking. J Biomol Struct Dyn 35:1529–1546.
    OpenUrl
  28. ↵
    1. Shitara Y,
    2. Maeda K,
    3. Ikejiri K,
    4. Yoshida K,
    5. Horie T, and
    6. Sugiyama Y
    (2013) Clinical significance of organic anion transporting polypeptides (OATPs) in drug disposition: their roles in hepatic clearance and intestinal absorption. Biopharm Drug Dispos 34:45–78.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Tsao SC,
    2. Sugiyama Y,
    3. Sawada Y,
    4. Iga T, and
    5. Hanano M
    (1988a) Kinetic analysis of albumin-mediated uptake of warfarin by perfused rat liver. J Pharmacokinet Biopharm 16:165–181.
    OpenUrlPubMed
  30. ↵
    1. Tsao SC,
    2. Sugiyama Y,
    3. Sawada Y,
    4. Nagase S,
    5. Iga T, and
    6. Hanano M
    (1986) Effect of albumin on hepatic uptake of warfarin in normal and analbuminemic mutant rats: analysis by multiple indicator dilution method. J Pharmacokinet Biopharm 14:51–64.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Tsao SC,
    2. Sugiyama Y,
    3. Shinmura K,
    4. Sawada Y,
    5. Nagase S,
    6. Iga T, and
    7. Hanano M
    (1988b) Protein-mediated hepatic uptake of rose bengal in analbuminemic mutant rats (NAR). Albumin is not indispensable to the protein-mediated transport of rose bengal. Drug Metab Dispos 16:482–489.
    OpenUrlAbstract
  32. ↵
    1. Watanabe T,
    2. Kusuhara H,
    3. Maeda K,
    4. Kanamaru H,
    5. Saito Y,
    6. Hu Z, and
    7. Sugiyama Y
    (2010) Investigation of the rate-determining process in the hepatic elimination of HMG-CoA reductase inhibitors in rats and humans. Drug Metab Dispos 38:215–222.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Watanabe T,
    2. Kusuhara H,
    3. Watanabe T,
    4. Debori Y,
    5. Maeda K,
    6. Kondo T,
    7. Nakayama H,
    8. Horita S,
    9. Ogilvie BW,
    10. Parkinson A, et al.
    (2011) Prediction of the overall renal tubular secretion and hepatic clearance of anionic drugs and a renal drug-drug interaction involving organic anion transporter 3 in humans by in vitro uptake experiments. Drug Metab Dispos 39:1031–1038.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Weisiger R,
    2. Gollan J, and
    3. Ockner R
    (1981) Receptor for albumin on the liver cell surface may mediate uptake of fatty acids and other albumin-bound substances. Science 211:1048–1051.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Weisiger RA,
    2. Zacks CM,
    3. Smith ND, and
    4. Boyer JL
    (1984) Effect of albumin binding on extraction of sulfobromophthalein by perfused elasmobranch liver: evidence for dissociation-limited uptake. Hepatology 4:492–501.
    OpenUrlPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleArticle

IVIVE for “Albumin-Mediated” Hepatic Uptake Clearance

Soo-Jin Kim, Kyeong-Ryoon Lee, Seiji Miyauchi and Yuichi Sugiyama
Drug Metabolism and Disposition February 1, 2019, 47 (2) 94-103; DOI: https://doi.org/10.1124/dmd.118.083733
Twitter logo Facebook logo Mendeley logo

Jump to section

Related Articles

Cited By...

More in this TOC Section

Similar Articles

Advertisement