ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2015

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Pharmacy 193

In vitro and in silico Predictions ofHepatic Transporter-MediatedDrug Clearance and Drug-DrugInteractions in vivo

ANNA VILDHEDE

ISSN 1651-6192ISBN 978-91-554-9141-3urn:nbn:se:uu:diva-241376

Dissertation presented at Uppsala University to be publicly examined in Sal B41,Biomedicinskt centrum, Husargatan 3, Uppsala, Friday, 27 February 2015 at 09:15 for thedegree of Doctor of Philosophy (Faculty of Pharmacy). The examination will be conductedin English. Faculty examiner: Professor Kim Brouwer (UNC School of Pharmacy ).

AbstractVildhede, A. 2015. In vitro and in silico Predictions of Hepatic Transporter-MediatedDrug Clearance and Drug-Drug Interactions in vivo. Digital Comprehensive Summaries ofUppsala Dissertations from the Faculty of Pharmacy 193. 81 pp. Uppsala: Acta UniversitatisUpsaliensis. ISBN 978-91-554-9141-3.

The liver is the major detoxifying organ, clearing the blood from drugs and other xenobiotics.The extent of hepatic clearance (CL) determines drug exposure and hence, the efficacy andtoxicity associated with exposure. Drug-drug interactions (DDIs) that alter the hepatic CL maycause more or less severe outcomes, such as adverse drug reactions. Accurate predictions ofdrug CL and DDI risk from in vitro data are therefore crucial in drug development.

Liver CL depends on several factors including the activities of transporters involved in thehepatic uptake and efflux. The work in this thesis aimed at developing new in vitro and insilico methods to predict hepatic transporter-mediated CL and DDIs in vivo. Particular emphasiswas placed on interactions involving the hepatic uptake transporters OATP1B1, OATP1B3, andOATP2B1. These transporters regulate the plasma concentration-time profiles of many drugsincluding statins.

Inhibition of OATP-mediated transport by 225 structurally diverse drugs was investigated invitro. Several novel inhibitors were identified. The data was used to develop in silico models thatcould predict OATP inhibitors from molecular structure. Models were developed for static anddynamic predictions of in vivo transporter-mediated drug CL and DDIs. These models rely on acombination of in vitro studies of transport function and mass spectrometry-based quantificationof protein expression in the in vitro models and liver tissue. By providing estimations oftransporter contributions to the overall hepatic uptake/efflux, the method is expected to improvepredictions of transporter-mediated DDIs. Furthermore, proteins of importance for hepatic CLwere quantified in liver tissue and isolated hepatocytes. The isolation of hepatocytes from livertissue was found to be associated with oxidative stress and degradation of transporters andother proteins expressed in the plasma membrane. This has implications for the use of primaryhepatocytes as an in vitro model of the liver. Nevertheless, by taking the altered transporterabundance into account using the method developed herein, transport function in hepatocyteexperiments can be scaled to the in vivo situation. The concept of protein expression-dependentin vitro-in vivo extrapolations was illustrated using atorvastatin and pitavastatin as model drugs.

Keywords: OATP1B1, OATP1B3, OATP2B1, NTCP, drug transporters, human hepatocytes,atorvastatin, pitavastatin, proteomics, sandwich-cultured human hepatocytes, SCHH,mechanistic modeling, in vitro-in vivo extrapolation, transport inhibition, hepatic uptake,hepatocyte isolation, transporter contribution

Anna Vildhede, Department of Pharmacy, Box 580, Uppsala University, SE-75123 Uppsala,Sweden.

© Anna Vildhede 2015

ISSN 1651-6192ISBN 978-91-554-9141-3urn:nbn:se:uu:diva-241376 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-241376)

“Models are not descriptions of reality; they are descriptions of our assumptions about reality.”

Jeremy Gunawardena

Cover artwork by Febe Jacobsson

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Karlgren M., Vildhede A., Norinder U., Wisniewski J.R.,

Kimoto E., Lai Y., Haglund U., Artursson P. (2012). Classifica-tion of Inhibitors of Hepatic Organic Anion Transporting Poly-peptides (OATPs): Influence of Protein Expression on Drug–Drug Interactions. J Med Chem, 55(10): 4740-4763

II Vildhede A., Karlgren M., Svedberg E.K, Wisniewski J.R., Lai Y., Norén A., Artursson P. (2014) Hepatic Uptake of Atorva-statin: Influence of Variability in Transporter Expression on Uptake Clearance and Drug–Drug Interactions. Drug Metab Dispos, 42(7): 1210-1218

III Vildhede A., Wisniewski J.R., Norén A., Karlgren M., Arturs-son P. Comparative Proteomics of Human Liver Tissue and Hepatocytes Reveal Systematic Differences in Proteins Deter-mining Hepatic Drug Exposure. Submitted.

IV Vildhede A., Mateus A., Khan E.K., Lai Y., Karlgren M., Ar-tursson P., Kjellsson M.C. Mechanistic Modeling of Pitava-statin Disposition in Sandwich-Cultured Human Hepatocytes: a Proteomics-Informed Bottom-Up Approach. Manuscript.

Reprints were made with permission from the respective publishers.

Contents

Introduction ................................................................................................... 11 The Human Liver ..................................................................................... 12

Cell Composition and Organization of the Liver ................................. 12 The Hepatocyte .................................................................................... 14

Hepatic Disposition of Drugs ................................................................... 15 Metabolism .......................................................................................... 15 Passive Transmembrane Diffusion ...................................................... 16 Carrier-Mediated Transport ................................................................. 17 Co-Existence of Passive and Active Transport .................................... 21 Drug–Drug Interactions ....................................................................... 22

In vitro Methods for Studying Hepatic Drug Transport and Transporter-Mediated Inhibition .............................................................. 22

Primary Human Hepatocytes ............................................................... 23 Human Hepatoma Cell Lines ............................................................... 24 Transfected Immortalized Cell Lines .................................................. 24 Xenopus Laevis Oocytes ...................................................................... 25 Inverted Membrane Vesicles ............................................................... 25

Predictions of Hepatic Transporter-Mediated Clearance and Drug–Drug Interactions in vivo .......................................................................... 26

In vitro–in vivo Scaling of Hepatic Drug Transport ............................ 26 In vitro–in vivo Scaling of Transporter-Mediated Drug–Drug Interactions .......................................................................................... 28 Mechanistic Pharmacokinetic Modeling ............................................. 30 In silico Predictions of Transport Inhibitors ........................................ 31

Proteomics ................................................................................................ 31 Label-Based Quantification of Protein Abundance ............................. 32 Label-Free Quantification of Protein Abundance ................................ 32

Statins as Model Drugs ............................................................................ 33 Hepatic Disposition of Atorvastatin .................................................... 34 Hepatic Disposition of Pitavastatin ..................................................... 35

Aims of the Thesis ........................................................................................ 36

Methods ........................................................................................................ 37 Compound Selection ................................................................................ 37 Kinetic Characterization of Transporter-Mediated Uptake and Efflux .... 37

In vitro Screening of Transport Inhibition ............................................... 38 Quantifying Isoform Specificity of Inhibitors .......................................... 39 Isolation and Culture of Primary Human Hepatocytes ............................. 39 Characterization of NTCP Contribution to Hepatic Transporter-Mediated Uptake ...................................................................................... 40 Characterization of Drug Disposition in Sandwich-Cultured Human Hepatocytes .............................................................................................. 40 Quantitative Analysis of Protein Abundance by Liquid Chromatography-Tandem Mass Spectrometry ........................................ 41

Targeted Proteomics ............................................................................ 41 Global Proteomics ............................................................................... 41

In vitro–in vivo Scaling of Hepatic Transport and Transporter-Mediated Drug–Drug Interactions ............................................................ 42 Mechanistic Pharmacokinetic Modeling of Hepatic Drug Disposition .... 43 Statistical Analysis ................................................................................... 44

Correlations in Protein Expression ...................................................... 44 Pathway Analysis ................................................................................ 44

Results and Discussion ................................................................................. 45 Identification and Prediction of Inhibitors of the Hepatic Organic Anion Transporting Polypeptide (OATP) Transporters ........................... 45 Prediction of Interindividual Uptake of Atorvastatin and Pitavastatin in Human Liver ............................................................................................ 47 Prediction of Interindividual Differences in Transporter-Mediated Drug–Drug Interactions Involving Atorvastatin ...................................... 50 Comparison of ADME Protein Abundance in Isolated Human Hepatocytes and Human Liver Tissue ...................................................... 52 Prediction of Time-Dependent Pitavastatin Disposition in Sandwich-Cultured Human Hepatocytes .................................................................. 55

Conclusions ................................................................................................... 56

Future Perspectives ....................................................................................... 58

Svensk Populärvetenskaplig Sammanfattning .............................................. 60

Acknowledgements ....................................................................................... 63

References ..................................................................................................... 66

Abbreviations

ABC ATP-Binding Cassette ADME Absorption, Distribution, Metabolism, and Excretion ATP Adenosine Triphosphate AUC Area Under the Plasma Concentration-Time Curve BCRP Breast Cancer Resistance Protein CL Clearance CYP Cytochrome P450 DDI Drug–Drug Interaction HBSS Hank’s Balanced Salt Solution HEK Human Embryonic Kidney IVIVE In Vitro–In Vivo Extrapolation Km Michaelis-Menten Constant LC-MS/MS Liquid Chromatography-Tandem Mass Spectrometry MRP Multidrug Resistance-Associated Protein NTCP Sodium Taurocholate Cotransporting Polypeptide NPC Non-Parenchymal Cell OATP Organic Anion Transporting Polypeptide Pgp P-glycoprotein PSA Polar Surface Area QSAR Quantitative Structure-Activity Relationship SCHH Sandwich-Cultured Human Hepatocytes SLC Solute Carrier TPA Total Protein Approach UGT Uridine 5'-Diphospho-Glucuronosyltransferase UWL Unstirred Water Layer Vmax Maximal Uptake Rate

11

Introduction

For an orally administered drug to achieve its intended systemic effect, it must reach its therapeutic target in sufficiently high levels. Multiple biologi-cal barriers exist that limit the amount of drug that is delivered to the site of action. Examples of such barriers include the intestinal epithelium where absorption occurs and the liver where the drug is eliminated through metabo-lism and excretion. Consequently, not only is the activity of the drug of im-portance for its efficacy, but also its absorption, distribution, metabolism, and excretion (ADME) since these properties govern the amount of drug at the target site. Although this was not fully appreciated twenty to thirty years ago,1, 2 characterization of the ADME properties of a new drug entity is now an integrated part of early drug development along with optimization of its potency. As a result, drug attrition rate due to poor pharmacokinetics has decreased immensely.2

With an increased understanding of the ADME processes that influence the pharmacokinetics of a drug, transport proteins gained attention in the 1990s.3, 4 Since then, numerous drug transporters have been identified.5 Re-search on drug transporters has improved our understanding of their tissue and subcellular localization,5 transport mechanisms,6 substrate and inhibitor specificities,5 genetic polymorphisms,7 and species differences.8

Today, it is well acknowledged that transporters can have a significant role in the absorption, distribution, and excretion of drugs and their metabo-lites. By determining the amount of drug that enters and exits a cell, trans-porters also influence the rate and extent of drug metabolism.9 Being such important determinants of systemic and tissue exposure, transport proteins are possible targets for drug–drug interactions (DDIs) with potential changes in drug safety and/or efficacy. Hence, regulatory agencies such as the Amer-ican Food and Drug Administration and the European Medicines Agency now recommend studies of transporter-mediated drug interactions in their industry guidelines.10, 11 These recommendations direct drug development towards a better understanding of the determinants that account for the pharmacokinetics of new drug entities.10, 11

Transport proteins are expressed in varying abundance in all tissues in the human body including the liver. The elimination of drugs that takes place in the liver has a substantial impact on drug pharmacokinetics. Predictions of hepatic elimination in humans are therefore of crucial importance for as-sessing the probability of success of new drug candidates. Several different

12

methods to predict hepatic clearance (CL) from in vitro studies exist. How-ever, the in vivo elimination of drugs that are dependent on transport proteins for their CL is still generally underpredicted.12, 13 Thus, new improved strate-gies for extrapolating in vitro data to the in vivo situation are needed and have been addressed in this thesis.

The Human Liver The liver is the most important organ for drug elimination. With its strategic anatomical location in the human body, it connects the gastro-intestinal tract to the systemic circulation. Orally administered drugs that are absorbed in the intestine thus have to pass through the liver before they reach the system-ic blood circulation. The elimination that takes place during this step is re-ferred to as the hepatic first pass effect and may have a substantial impact on the bioavailability of the drug (i.e., the fraction of unchanged drug that reaches the systemic circulation upon oral administration).14 Once the drug has reached the systemic blood circulation, the kidneys aid the liver in con-tinued elimination from the body. In addition to the elimination of drugs and other exogenous and endogenous compounds, the liver has several other essential functions, such as bile production and secretion, plasma protein synthesis, and glycogen storage.

Cell Composition and Organization of the Liver Two major types of cells populate the liver, parenchymal and non-parenchymal cells (NPCs). The parenchymal cells, also known as hepato-cytes, represent about 60% of the liver cell population and 80% of the volume of the organ.15 They carry out most of the liver functions and are responsible for the elimination of drugs. Only 6.5% of the liver volume is formed by NPCs.15 Of these, sinusoidal endothelial cells constitute the largest fraction. The remaining NPCs of the liver include hepatic stellate cells (fat- and vitamin A-storing cells), Kupffer cells (liver macrophages), and cholangio-cytes (biliary epithelial cells).15

The human liver has a dual blood supply, where the greater part (~75%) stems from the hepatic portal vein that carries nutrient-rich venous blood from the gastrointestinal tract and spleen. The remaining supply is from the hepatic artery, carrying fresh, oxygenated blood from the aorta. Within the liver, distributing branches of these blood vessels course in parallel and sub-divide repeatedly. The terminal branches of the vessels lead to a lobule, the structural unit of the liver (Figure 1).16 The lobule has a hexagonal shape. At the midpoint of the lobule is the central vein. From here, hepatocytes radiate out in linear cords (plates of hepatocytes just one cell-layer thick) to the pe-riphery of the lobule. Blood from the branches of the hepatic artery and por-

13

tal vein combines in hepatic sinusoids, located in between the hepatic cords, and exits the lobule through the central vein. Most of the hepatocytes are exposed to the sinusoidal vasculature on two sides, allowing efficient ex-change between the blood and the parenchymal cells.16

Liver sinusoidal endothelial cells form the wall of the sinusoid. These cells are highly fenestrated, permitting plasma solutes to move freely into the perisinusoidal space of Disse, the space between the sinusoidal endothelial cells and the hepatocytes.17 Associated with the sinusoids are Kupffer cells and hepatic stellate cells, which reside within the space of Disse. The stellate cells are important for extracellular matrix homeostasis,18 thereby regulating the three-dimensional structure of the liver.

Figure 1. Schematic representation of the microarchitecture of the liver lobule. Blood from the hepatic artery and portal vein combines in the sinusoid and exits through the central vein in the center of the lobule. The hepatocytes form the cords of the lobule. Bile canaliculi are formed between adjacent hepatocytes. Through an interconnected network they transfer bile in the opposite direction of the blood, towards the periphery of the lobule. The sinusoids are lined by highly fenestrated liver sinusoidal endothelial cells (see inset). The hepatic stellate cells reside in the space of Disse between the hepatocytes and the sinusoidal endothelial cells. These cells control the three-dimensional structure of the liver by regulating the turnover of extracellular matrix. Kuppfer cells, which are liver-resident macrophages, are main-ly located within the sinusoids. The figure was drawn after an illustration by DH Adams and B Eksteen, Nat Rev Immunol, 2006.

At the apical membrane of hepatocytes, where adjacent hepatocytes meet, bile canaliculi are formed through invaginations of the plasma membrane.19 The apical membrane is separated from the basolateral membrane by tight junctions and desmosomes. The bile canaliculi form an interconnected net-work that transfers bile in the opposite direction of the sinusoidal blood, i.e., towards the periphery of the lobule. There they join to form bile ducts. The bile ducts are lined by cholangiocytes, which contribute to bile secretion by

14

release of water and bicarbonate.20 The bile ducts carry bile components to the common hepatic duct, which is connected to the gallbladder. The gallbladder, in turn, releases its content into the small intestine.

The Hepatocyte The hepatocyte is a highly polarized cell with two distinct membranes: the basolateral membrane, facing the sinusoids and the neighboring hepatocytes of the cords (intercellular section), and the apical membrane facing the bile canaliculi. Both the basolateral membrane and the apical membrane have microvilli protruding into the space of Disse and the bile canaliculi, respec-tively, increasing the surface area of these membranes. Half of the surface area of a hepatocyte has been reported to be in contact with either the sinus-oid (37%) or the bile canaliculi (13%).21 In another study, 22% and 10% of the hepatocyte surface area was found to be in contact with sinusoids and bile canaliculi, respectively.22

The hepatocytes take up substances destined for the bile via the basolat-eral membrane and excrete them across the canalicular membrane.19 This vectorial transport requires a domain-specific distribution of lipid and pro-tein components (including transport proteins) in the plasma membrane. The polarity is achieved through sorting of the proteins to the different mem-branes and is upheld by tight junctions.23

Due to differences in local blood composition and oxygenation along the sinusoids, the hepatocytes are heterogeneous in function, resulting in the division into three zones (Figure 2).16, 19 Hepatocytes close to the oxygen- and nutrient-rich periportal area are perfused first and reside in zone I. Hepatocytes in the intermediate zone II and distant zone III, close to the central vein, are perfused sequentially with blood already modified by the preceding hepatocytes. Consequently, they receive less nutrients and oxygen. Different metabolic functions are carried out in the various zones. As an example, hepatocytes in zone III have the highest expression of drug metab-olizing enzymes and are the most important for biotransformation of drugs.24

The hepatocytes do not only show functional heterogeneity, but also mor-phological. In general, the hepatocytes are cuboidal in shape with one or more cell nuclei. The fraction of polyploid hepatocytes increases from zone I to zone III.25 In addition, hepatocytes close to the portal vein (zone I) tend to be smaller in size than those close to the central vein (zone III).26

15

Figure 2. Zonation of the liver. Hepatocytes are heterogeneous in appearance and function along the axis of the liver lobule. The functional zonation is a result of a gradual decrease in the concentration of oxygen and nutrients from the periportal (zone I) to the perivenous (zone III) area. The metabolism of drugs and xenobiotics is one of the processes affected by this zonation. Drug metabolism mainly occurs in the highly differentiated hepatocytes located in the perivenous area. These hepato-cytes express the highest levels of drug metabolizing cytochrome P450 (CYP) enzymes.

Hepatic Disposition of Drugs One of the important functions of the hepatocytes is the degradation of en-dogenous and exogenous compounds such as drugs. For this purpose, the liver contains a wide array of metabolizing enzymes that convert lipophilic substances into more water-soluble metabolites. The metabolites can then more easily be excreted via the bile or the kidneys. Because the metabolizing enzymes are located intracellularly, drugs first have to cross the basolateral membrane of the hepatocyte. A drug molecule can pass through the cell membrane by different modes of transport, of which the main ones for small-molecular drugs are passive transmembrane diffusion and carrier-mediated transport. Following uptake across the basolateral membrane, drugs may undergo biotransformation by phase I and phase II metabolizing enzymes. In addition, drugs and their metabolites may be excreted into the bile or trans-ported back into the blood for renal elimination. The various processes con-tributing to hepatic disposition of drugs are described in more detail in the following sections.

Metabolism The hepatocyte is well equipped for its central role in drug elimination through its high expression of drug-metabolizing enzymes. Drug metabolism can be divided into two phases. The first phase includes reactions such as reduction, oxidation, and hydrolysis, which introduce reactive or polar groups into the compound being metabolized. The majority of these reac-tions are catalyzed by enzymes of the cytochrome P450 (CYP) family, ac-counting for around 75% of total drug metabolism.27 CYP3A4, CYP2D6,

16

CYP2C9, and CYP2C19 are the most important for the metabolism of com-mercially available drugs today. Other notable CYPs in drug metabolism are CYP1A2, CYP2B6, and CYP2E1.

If the metabolites being formed during phase I metabolism are sufficient-ly polar, they may be directly excreted into the bile or blood. Many phase I products, however, undergo subsequent phase II metabolism, in which hy-drophilic functional groups such as glutathione or glucuronic acid are conju-gated to the molecules. Phase II metabolism can also occur without preced-ing phase I metabolism. Phase II metabolism is catalyzed by transferases including uridine 5'-diphospho-glucuronosyltransferases (UGTs), glutathione S-transferases (GSTs), and sulfotransferases (SULTs). UGTs are the most important of these conjugating enzymes, accounting for approximately 10% of total drug metabolism.28 The main drug-metabolizing UGT enzymes are UGT2B7, UGT1A1, and UGT1A4.

Passive Transmembrane Diffusion Drugs have to cross the cell membrane to gain access to the cell interior. The unassisted permeation of molecules through the lipid bilayer is generally considered the most important mode of transport by which a drug crosses the cell membrane.29 Some researchers, however, claim that active carrier-mediated uptake is the rule rather than the exception.30

The process of passive transmembrane diffusion involves partitioning in-to/out of the cell membrane and diffusion through the cellular membrane. The fundamental principle of passive permeation is contained in Fick’s first law of diffusion: a substance diffuses from a region of high concentration to a region of low concentration at a rate proportional to the magnitude of the concentration gradient. Passive diffusion is spontaneous and requires no energy. Rather, the driving force is the increase of entropy in the system. The net movement of molecules from a region of high concentration to a region of low concentration can be explained by the random walk (Brownian motion) of individual molecules: it is more likely that any single molecule will move from an area of high concentration to an area of low concentration than in the opposite direction. Therefore, the net movement will be down the concentration gradient until equilibrium is reached.

The diffusion rate across the cell membrane is directly proportional to the permeability coefficient (P), the surface area of the membrane (A), and the concentration gradient (dC/dx), equation 1.31

dx

dCAPv ××= (1)

17

The permeability coefficient associated with the passive diffusion of drug molecules through a cell membrane is temperature-dependent. The lower the temperature, the more rigid the cell membrane becomes and the lower the passive diffusion of molecules across it.32 The decrease in temperature also reduces the motion of the molecules themselves and thereby the permeability constant. Furthermore, membrane fluidity is influenced by the composition of the membrane, which differs for various cell types.33 Besides membrane properties, the physicochemical properties of the drug also influence the rate of passive diffusion. Significant correlations between lipophilicity and mem-brane permeability have been reported in several studies.34 Ionization state of the drug is another important factor, where the uncharged form of a molecule can cross the cell membrane more easily than the charged.35 Partitioning into the hydrophobic interior of the cell membrane and diffusion across it is also benefitted by a low molecular weight.36

Carrier-Mediated Transport Endogenous compounds and drugs that do not readily permeate lipid bi-layers may still be able to cross the hepatic cell membrane to a considerable extent through carrier-mediated transport. This transport mechanism is as-sisted by transporters, which are integral transmembrane proteins that allow for selective passage of molecules. The transport proteins are categorized as either influx or efflux transporters depending on the direction in which they transport substrates across the membrane.37 Influx or uptake transporters mediate transport of molecules into the cell, whereas efflux transporters de-crease intracellular exposure by transporting the compound out of the cell. Some transporters are able to mediate transport in both directions.

There are two forms of carrier-mediated transport, active transport and fa-cilitated diffusion, of which the latter consumes no energy.37 Active transport, on the other hand, requires energy and represents a way to transport molecules against their concentration gradient. Depending on the source of energy, active transport is either categorized as primary or second-ary. Primary active transport relies on the direct hydrolysis of adenosine triphosphate (ATP) for substrate translocation. In contrast, secondary active transport involves the use of another energy source than ATP such as an ion gradient generated by ATP-dependent primary transporters.

The relation between the rate of transport (v) and substrate concentration ([S]) can be described by the Michaelis-Menten equation (eq. 2), originally developed to describe enzyme kinetics.38, 39

[ ][ ]SK

Svv

m +×= max (2)

18

where vmax is the maximal transport rate and Km is the substrate concentration yielding a transport uptake rate that is half of vmax. The maximal transport rate is the product of the total amount of transport protein and the turnover number, i.e., the number of substrate molecules translocated by a transport protein per unit time when the transporter is fully saturated with substrate. Km, on the other hand, is independent of transporter expression. Km is gener-ally thought to describe the binding strength of the transporter-substrate complex, but this only holds true if the dissociation rate of the complex is much faster than the rate of substrate translocation.40

The hepatocytes express several different transport proteins that distribute drugs across the membrane. The most important ones are depicted in Figure 3.

Figure 3. Schematic representation of hepatocytes with selected transport proteins mediating the transmembrane flux of drugs and endogenous substances. Uptake transporters in the basolateral membrane include organic cation transporter 1 (OCT1), organic anion transporter 2 (OAT2), sodium-taurocholate cotransporting polypeptide (NTCP), and three members of the organic anion transporting polypep-tide (OATP) family (OATP1B1, OATP1B3, and OATP2B1). Basolateral efflux transporters are represented by two multidrug resistance-associated proteins (MRP3 and MRP4), while the canalicular membrane contains efflux transporters such as the breast cancer resistance protein (BCRP), bile salt export pump (BSEP), P-glycoprotein (Pgp), and MRP2.

19

Basolateral Uptake Transporters The basolateral membrane of hepatocytes contains a large number of influx transporters belonging to the solute carrier (SLC) family. These transporters mediate the uptake of compounds from the sinusoidal blood. The SLC family includes passive (facilitative) transporters and secondary active transporters. The most important ones for hepatic drug uptake belong to the SLC10, SLC22, and SLCO gene subfamilies.5

The sodium-taurocholate cotransporting polypeptide (NTCP, SLC10A1) is expressed exclusively in the basolateral membrane of hepatocytes. NTCP is the predominant transporter involved in the uptake of bile acids in the liver, mediating electrogenic sodium-dependent transport with a stoichiometry of two sodium ions per bile acid molecule.41, 42 Its substrate specificity is not only restricted to bile acids though. NTCP also accepts various statins (or HMG-CoA reductase inhibitors) as substrates.43-45

Statin and bile acid uptake across the hepatic sinusoidal membrane is also mediated by organic anion transporting polypeptides (OATPs). In contrast to NTCP, the OATP transporters operate in a sodium-independent manner.46 The exact transport mechanism(s) is not fully understood, but it appears to be anion exchange.47, 48 At least three different OATP transporters are ex-pressed in the human liver: OATP1B1 (SLCO1B1), OATP1B3 (SLCO1B3), and OATP2B1 (SLCO2B1).5 Both OATP1B1 and OATP1B3 are liver-specific, while OATP2B1 is expressed in other tissues as well.49 These transporters have a broad and overlapping substrate profile, including drugs such as statins, angiotensin II receptor antagonists, HIV protease inhibitors, fexofenadine, and rifampicin.50-52 OATP-mediated uptake is reported to be the rate-limiting step in the elimination of several drugs.53, 54 In addition, reduced function genetic variants of OATP1B1 and inhibition of OATP-mediated uptake have been associated with increased plasma concentrations of substrate drugs.52

The organic anion transporter 2 (OAT2, SLC22A7) is mainly expressed in the liver, in contrast to the other kidney-localized OAT transporters.55 It ac-cepts a range of structurally diverse compounds as substrates, including allo-purinol, 5-fluorouracil, and paclitaxel.56 Like the OATP transporters, OAT2 is believed to transport substrates through anion exchange.56

The uptake of organic cations in the liver is primarily mediated by the polyspecific organic cation transporter 1 (OCT1, SLC22A1). OCT1 is a facil-itative transporter, preferentially expressed in the liver, which can translocate molecules in both directions across the membrane. The transport by OCT1 is sodium-independent and driven by the membrane potential.57 Preferred sub-strates are low molecular weight and relatively hydrophilic cations such as the anti-diabetic drug metformin and the prototypical substrates tetrae-thylammonium and 1-methyl-4-phenylpyridinium.5 OCT1 is also involved in

20

the transport of the anticancer drug oxaliplatin.58 Other anticancer drugs that are likely OCT1 substrates include irinotecan and paclitaxel.59

Canalicular and Basolateral Efflux Transporters The canalicular membrane of hepatocytes contains several efflux transport-ers that actively excrete exogenous and endogenous compounds and their metabolites into the bile. These include multidrug resistance-associated pro-tein 2 (MRP2, ABCC2), P-glycoprotein (Pgp, ABCB1), breast cancer re-sistance protein (BCRP, ABCG2), and the bile salt export pump (BSEP, ABCB11). A different set of efflux transporters is present in the basolateral membrane, transferring compounds into the blood, e.g., MRP3 (ABCC3) and MRP4 (ABCC4). Both the canalicular and the basolateral efflux transporters belong to the ATP-binding cassette (ABC) family of transporters that direct-ly hydrolyze ATP to transport molecules across the membrane.5

MRP2 is a major xenobiotic efflux pump at the canalicular membrane. It recognizes a wide range of compounds including statins, HIV protease inhibi-tors, anticancer drugs, and conjugated phase II metabolites.60 Like the OATPs, it mainly transports organic anions. The absence of functional MRP2 protein in the liver is associated with conjugated hyperbilirubinaemia resulting from impaired excretion of conjugated bilirubin.61 Besides the hepatocytes, MRP2 is located in the apical membrane of kidney proximal tubules and intestinal epithelia, where it also clears the body from xenobiotics.5

Pgp, also known as multidrug resistance protein 1 (MDR1), is another important transporter for biliary efflux of drugs and their metabolites. It is the most studied of the drug transporters, because of its involvement in multi-drug resistance of tumor cells and its widespread expression in various cell barriers of the body, including the intestine, blood-brain barrier, and kidneys. Pgp has very broad substrate specificity and shares many substrates and in-hibitors with CYP3A4.62 This observation along with the intestinal co-localization has resulted in the hypothesis of a co-operative interplay be-tween Pgp and CYP3A4 in limiting the entry of xenobiotics into the blood circulation.63 The literature is, however, not unified on how these proteins interact to attenuate the oral bioavailability. While some researchers support the hypothesis that transport and metabolism work co-operatively,64-68 others propose that Pgp-mediated efflux may in fact reduce CYP3A4 metabolism in the intestine.69, 70 Substrates of Pgp are generally bulky, hydrophobic, and cationic. Examples of drugs transported by Pgp include digoxin, loperamide, and the anticancer drug doxorubicin.5

BCRP is a half transporter that forms a functional homodimer.71 Like the other efflux transporters, it is expressed not only in the liver but also in other tissues with barrier function, such as the placenta, intestine and blood-brain barrier.72 Substrates of BCRP include statins, anticancer drugs, and sulphate conjugates.5

21

BSEP is not involved in drug transport to the same extent as the other canalicular efflux transporters. Nevertheless, it transports a few drugs, in-cluding pravastatin.73 Instead, BSEP constitutes the rate-limiting step in the secretion of bile salts across the canalicular membrane of hepatocytes. There is no compensatory mechanism for loss of BSEP, manifested by symptoms of cholestasis in individuals with inherited BSEP dysfunction.74 In addition, inhibition of the BSEP transporter by drugs is associated with drug-induced liver injury through intracellular accumulation of bile salts,75 further demon-strating its essential role in bile acid homeostasis.

MRP3 is a sinusoidal efflux transporter expressed by hepatocytes. It me-diates the transport of drugs and their metabolites from the liver into the blood. MRP3 shares 48% amino acid similarity with canalicularly localized MRP2. Like MRP2, it is involved in the transport of glucuronide conjugates and drugs such as methotrexate.76 The expression of MRP3 is markedly in-duced during cholestasis and in individuals with dysfunctional MRP2.77, 78 This indicates that it has a protective role in the setting of impaired canalicu-lar excretion by reducing hepatocellular accumulation of anionic conjugates.

MRP4 is another sinusoidal transporter which shows increased expression levels during cholestasis.79 MRP4 transports a range of drugs, covering anti-viral, cardiovascular, and anticancer drugs.80

Co-Existence of Passive and Active Transport Net transport across a cellular membrane is the sum of passive diffusion and active carrier-mediated transport (Figure 4).29 The relative contribution of each depends on the physicochemical properties and the concentration of the drug. Carrier-mediated transport is more important for drugs with low pas-sive permeability. Hence, active transport is more readily observed for hy-drophilic drugs than for lipophilic ones. Because carrier-mediated transport is saturable, its contribution to the net transport will decrease with increasing drug concentration. The saturable feature is a result of the limited number of carrier proteins in the cell membrane. Because systemic blood concentra-tions typically are low in comparison to the number of transport proteins, saturation of hepatic uptake transporters is not expected in vivo.

22

Figure 4. Drugs must cross the cell membrane to enter the cell interior. The total (net) transport across the cell membrane is the sum of passive diffusion and active transport. Passive transmembrane diffusion is non-saturable and increases in propor-tion to the concentration gradient across the membrane. Active transport, on the other hand, is saturable and characterized by a maximal transport rate (vmax). Satura-tion occurs when the number of permeant molecules exceeds the number of carrier protein binding sites available for transport.

Drug–Drug Interactions Concomitant administration of drugs may cause altered exposure of them, a situation referred to as pharmacokinetic drug-drug interaction (DDI). This is a result of the interference of one of the drugs (perpetrator) with one or more of the determinants accounting for the pharmacokinetics of the other drug (victim drug).81 Processes that can be affected include drug absorption, dis-tribution, metabolism, and excretion. Depending on the direction of change (inhibition or induction), drug exposure will either increase or decrease. In-creased plasma concentration is associated with an increased risk of adverse events, whereas decreased drug levels may result in a loss of drug efficacy. Clinical DDIs can cause severe outcomes. For example, co-administration of cerivastatin and gemfibrozil (an inhibitor of the hepatic uptake of ceri-vastatin) was reported in 12 of 31 cases of fatal rhabdomyolysis in the Unit-ed States before cerivastatin was withdrawn from the market.82

In vitro Methods for Studying Hepatic Drug Transport and Transporter-Mediated Inhibition Several different experimental systems of varying complexity have been developed for the study of transport processes. Each has its advantages and weaknesses. The sections below presents, in order of decreasing complexity, the assay systems commonly used to assess transport function and activity. A subset of the assay systems is depicted in Figure 5.

23

Figure 5. Some in vitro models used to study hepatic drug transport and transporter-mediated inhibition. (A) Primary human hepatocytes cultured between two layers of extracellular matrix (a so-called sandwich configuration) form intact bile canaliculi between the cells which enables vectorial transport studies. SCHH is a complex cell system with expression of most or all transport proteins important for drug disposi-tion in the liver. (B) Immortalized cell lines such as human embryonic kidney (HEK) 293 cells are anchorage-dependent and form monolayers when grown on a solid support. The cell lines can be transfected with one or more transporters of interest for the study of drug uptake and/or efflux. The transfected cell lines repre-sent a simpler assay system than SCHH. (C) Inverted membrane vesicles are derived from transfected cells. They are typically used to study efflux transporters. Because of the inverted configuration, drugs can interact directly with the efflux transporter without having to cross the cell membrane. This is the simplest assay system.

Primary Human Hepatocytes Hepatocytes isolated from human liver tissue are considered the gold stand-ard for studies of drug metabolism and transport activity.83 This experi-mental system is one of the most physiologically relevant ones available today with the advantage of expressing the full complement of drug trans-porters and metabolizing enzymes present in the liver. Isolated hepatocytes are a versatile tool that can be used as both suspensions and plated monolay-ers. Because of the anchorage-dependence of hepatocytes, the cells only survive for a few hours in suspension. Hepatocytes plated in conventional monolayers, on the other hand, survive for about a week. The plated cells are, however, associated with a dedifferentiation over time in culture.84 At-tempts have been made to restore or maintain hepatocyte-specific functions by optimizing the culture conditions. One such way is the use of sandwich-cultures, explained below.

Sandwich-Cultured Human Hepatocytes The in vivo polarity of hepatocytes is lost upon isolation, leading to the ina-bility to assess vectorial transport of compounds. The polarity can be re-established by growing the hepatocytes in a so-called sandwich configura-tion, i.e., between two layers of an extracellular matrix.85 This allows the cells to regenerate tight junctions and to form bile canaliculi-like structures between them. By modulating the tight junctions in the presence or absence of divalent cations (Ca2+ and Mg2+), canalicular efflux can be assessed.86

24

Sandwich-cultured human hepatocytes (SCHH) maintain morphology and function for up to six weeks.87 Culture conditions may, however, alter protein expression of transporters and other liver-specific proteins. While sandwich-cultured rat hepatocytes (SCRH) show significant time-dependent down-regulation of uptake transporters,88, 89 the expression and function in SCHH is better maintained.89-91 In contrast to the uptake transporters, efflux trans-porters are upregulated over time in SCRH.88 SCHH also display increased expression of certain efflux transporters with culture time, whereas others are more stable in their expression.92, 93

Human Hepatoma Cell Lines Because of the scarcity and unpredictable availability of human liver tissue, attempts have been made to generate hepatocyte cell lines from hepatocellu-lar carcinomas as an alternative to isolated primary human hepatocytes. The hepatocyte cell lines also benefit from a more stable phenotype than the pri-mary hepatocytes.

One of the most widely used human hepatocyte cell lines is HepG2. Al-though HepG2 cells are highly differentiated and show many liver-specific functions, they express very low levels of drug-metabolizing CYP enzymes and transporters compared with the expression in primary hepatocytes and liver tissue, respectively.94, 95 The low CYP expression in HepG2 cells is likely a result of altered expression of transcription factors regulating these enzymes.96 This might also explain the weak response to different CYP in-ducers.97 The use of HepG2 cells for drug disposition studies is therefore limited.

In contrast to the HepG2 cell line, differentiated HepaRG cells express phase I and II metabolizing enzymes at levels closer to those measured in freshly isolated human hepatocytes.98-100 Because of the expression of key nuclear receptors involved in CYP induction,98 HepaRG cells show similar induction patterns as primary cultured human hepatocytes.97, 99 In addition, polarized expression of drug transporters has been demonstrated in the HepaRG cell line.101 Although HepaRG cells better reflect primary hepatocytes in their gene expression pattern than HepG2 cells,102 they still generally express lower levels of CYP enzymes, with the exception of CYP3A4.99

Transfected Immortalized Cell Lines Immortalized cell lines can be transfected (stably or transiently) with one or more transport proteins of interest, which enables the study of a single or a limited set of transporters at a time.103-105 This presents a more defined assay system than the more complex primary hepatocytes and hepatic cell lines. Immortalized cell lines that are commonly transfected for transport studies include human embryonic kidney (HEK) 293, Madin-Darby canine kidney

25

(MDCK), and Chinese hamster ovary (CHO) cells. The HEK293 cells have the advantage of being of human origin and of expressing low endogenous levels of drug transporters.106 Transfected cell lines are useful tools for the characterization of substrates and inhibitors of individual transporters; in-formation that can be used to develop quantitative structure-activity relation-ship (QSAR) models.107-109 They can also be used to investigate transport mechanisms57, 110 and differences in transport activity as a result of genetic polymorphisms.111

Most transfected cell lines are grown as monolayers on a solid support. Some cell lines, such as MDCK cells, are polarized, which enables vectorial transport studies. Vectorial assay systems allow for the study of drug flux in two directions: apical-to-basolateral (A-B) and basolateral-to-apical (B-A). These polarized cells are commonly used to study efflux transporters by measuring the B-A/A-B flux ratio.

Xenopus Laevis Oocytes Xenopus laevis oocytes are germ cells surgically harvested from South Afri-can clawed frogs. The Xenopus laevis oocyte system is a valuable expression tool for studies of membrane transporters. The cells are large in relation to other cells (diameter of 1.2 mm), which makes them easy to handle.112 By microinjection of mRNA, the oocytes start to synthesize proteins in high amounts. Because of the injection of each individual cell, a high proportion of cells expressing the transferred genetic material is obtained. The cells can be injected with anything from RNA material coding for a single transporter to total liver mRNA.113 Another great advantage with this expression system is the limited number of endogenous transporters, resulting in a very low background transport activity.112

Inverted Membrane Vesicles Inverted plasma membrane vesicles are prepared from cells or tissue ex-pressing a transport protein of interest. The selection of cell/tissue type in-fluences the percentage of inside-out oriented vesicles obtained from the preparation. This fraction typically varies between 30 and 40%.114 Inverted plasma membrane vesicles are primarily used to study efflux transporters as a complement to cell-based assay systems. The membrane vesicles allow for a direct interaction of substrate with the cytoplasmic side of the efflux trans-porter as a result of the inside-out vesicle orientation. This is a clear ad-vantage over the cell-based assay system in which the substrate first has to pass the cell membrane before it can interact with the efflux transporter. The direct interaction with the transporter is particularly important for determina-tions of transport kinetics and transport mechanisms, which require that changes in assay conditions have a direct impact on the transport process

26

instead of being restricted by the cell membrane barrier. However, transport kinetics of lipophilic drugs that easily permeate cell membranes may be dif-ficult to determine using membrane vesicles because of a high diffusion rate out of the vesicle.115

Predictions of Hepatic Transporter-Mediated Clearance and Drug–Drug Interactions in vivo Accurate predictions of hepatic drug disposition from in vitro data are highly beneficial in drug development to identify candidates with inappropriate in vivo pharmacokinetics at an early stage. In this context, it is not only im-portant to estimate the rate of elimination but also to get an extensive under-standing of the transporters and metabolizing enzymes involved. A charac-terization of major contributors to the hepatic disposition provides an early assessment of the potential for DDIs and interindividual differences in pharmacokinetics due to polymorphic genotypes. The greater the contribu-tion of a single enzyme/transporter, the greater the potential for a clinically relevant DDI.

Drug elimination in pharmacokinetics is usually described by the term clearance (CL).116 Clearance is a measurement relating the elimination rate to drug concentration in the systemic circulation (eq. 3). It can also be inter-preted as the volume of blood that is cleared of drug per unit time.

Eliminationrate = CL × (3) The clearance term has lately been extended to describe distribution process-es, such as passive and carrier-mediated uptake into the liver. Transporter-mediated CL, relating transport rate to substrate concentration, can be de-scribed by equation 4.

[ ]SK

v

m += max

rtransporteCL (4)

In vitro–in vivo Scaling of Hepatic Drug Transport When studying transport processes in vitro, a common aim is to extrapolate the data to the in vivo situation. Because in vitro systems may differ in transport activities compared to the tissue of interest,95 strategies for scaling the data to the in vivo situation need to be established.

Transport data obtained from primary hepatocyte experiments are typical-ly scaled directly to the in vivo situation by multiplying the measured CL by the number of hepatocytes in the liver and by liver weight.54, 117, 118 This scal-

27

ing approach implicitly assumes that the isolated hepatocytes are identical to their in vivo counterpart with regard to transport function.

Transfected cell lines and membrane vesicles, on the other hand, usually differ considerably in transport activity compared to hepatocytes in the liver. Scaling data from these experimental systems therefore requires correction for this difference. This is done on an individual transporter basis. Various methods can be used to assess the difference in transport activity, including activity measurements with transporter-specific substrates and protein ex-pression determinations.119, 120 The data can then be further scaled by the number of hepatocytes in the liver and by liver weight to yield whole liver clearance.

Separating Active and Passive Transport Regardless of the experimental system used, total measured flux across the membrane needs to be separated into the individual transport processes, i.e., active carrier-mediated transport and passive diffusion. Several different approaches exist.

For studies using transfected cell lines, the most common approach is to do parallel studies in mock-transfected control cells.29 If these cells have low endogenous expression of transporters, any uptake will be due mainly to passive diffusion. With knowledge of the passive permeability, this compo-nent can be subtracted from the total transport in the transporter-transfected cells to obtain an estimate of the active transport. The same approach can be used with membrane vesicles although an alternative method also exists. Because efflux transporters directly hydrolyze ATP for the substrate translo-cation, passive permeability in membrane vesicles can be determined by incubations with adenosine monophosphate (AMP) instead of ATP.

Unfortunately, for studies of drug transport in hepatocytes, there are no corresponding control cells with no or low transport activity. Traditionally, incubations at 4°C have been used to assess the passive permeability in these cells, the rationale being that active processes are inhibited at low temperature.121-123 The passive permeability is, however, also affected by temperature. The increase in membrane rigidity at low temperature leads to an underprediction of the passive diffusion. Recently, pan-transporter inhibi-tors have been implemented to measure the passive component of the hepatic uptake.124 A cocktail of inhibitors is another option. The use of inhibitors for quantifying the passive component of the uptake relies on a complete inhibi-tion of the active transport. If this is not achieved, the passive diffusion will be either under- or overestimated depending on whether efflux or uptake transporters are still active. Mechanistic modeling approaches to simultane-ously estimate active and passive processes from in vitro transport data have also been tried.13, 32 However, the use of mathematical models is demanding in that it requires many experimental data points for parameter fitting.

28

Factors to Consider when Scaling Passive Permability and Active Carrier-Mediated Transport A number of factors affect the accuracy of in vitro–in vivo extrapolation (IVIVE) approaches. Below, some factors are discussed that may be of im-portance, but that are not always taken into account.

In any in vitro experimental assay, the total amount of drug added to the incubation is known. It is assumed that this concentration is the one driving both the passive and active transport. However, a compound may bind non-specifically to the incubation vessel, medium components (e.g., proteins if such are added), or the membrane of the biological system, reducing the free concentration of drug available for transport. This non-specific binding is generally higher for more lipophilic compounds.125

Clearance by passive permeability is dependent on the membrane surface area across which transport occurs. The surface area in contact with drug solution should therefore be considered for accurate IVIVE. For example, suspended hepatocytes have a greater surface area exposed to the incubation medium than hepatocytes in intact liver tissue have to the sinusoidal blood.

Transporter-mediated CL is determined by the kinetic parameters Km and vmax. While vmax depends on transporter expression (which usually differs between in vitro systems and hepatocytes in the liver), the Michaelis-Menten constant Km is assumed to be the same in vivo as in vitro. However, the ap-parent Km determination in vitro is affected by the presence of an unstirred water layer (UWL).126 The greater this UWL is, the greater the bias in the Km determination will be. Moreover, if efflux transport kinetics in a cell assay experiment is approximated on the basis of the extracellular concentration instead of the intracellular one, the apparent Km may be very different from the true Km.127

In vitro–in vivo Scaling of Transporter-Mediated Drug–Drug Interactions In vitro–in vivo extrapolations are performed not only to predict drug CL but also to predict potential DDIs. Because the substrate specificities of trans-porters are broad and often overlap with each other, multiple transporters may be involved in the disposition of a drug. The relative contributions of these transporters to the total flux therefore need to be determined for accu-rate estimations of the change in in vivo CL upon inhibition of one or several of them.

Estimating Transporter Contributions Several different approaches to predict transporter contributions have been developed. Many of them originated in the field of drug metabolism, but are now applied to the quantitative characterization of transporter contributions.

29

The relative activity factor (RAF) approach is one such method.120 It esti-mates the relative contributions of uptake transporters to the net active transport of a test compound by using transporter-specific substrates as ref-erence compounds. The RAF value is determined (eq. 5) by relating the up-take clearance of the reference compound in human hepatocytes (CLHep,Reference) to that in an experimental system (CLExp,Reference) expressing a single transporter (e.g., an overexpressing cell line). This RAF value is a measure of the functional activity of the transporter in hepatocytes relative to that in the experimental system.

ReferenceExp,

ReferenceHep,

CL

CLRAF = (5)

The uptake clearance of the test compound in human hepatocytes (CLHep,Test) is subsequently predicted by multiplying the uptake clearance in the experi-mental system (CLExp,Test) by the RAF value (eq. 6).

TestExp,TestHep, CLRAFCL ×= (6)

For many transporters, their specific substrates are unknown, which limits the applicability of the RAF method. Instead, an analogous methodology can be used that relies on transporter expression levels as a substitute of func-tional activity measurements. This method, referred to as the relative expres-sion factor (REF) approach,120 assumes a linear correlation between protein expression and transport activity. By relating the expression in human hepatocytes or liver tissue to that in the experimental system used to study transport function, in vivo CL can be predicted.

Other tools that can be used to estimate transporter contributions include selective gene silencing or gene knock-down in hepatocytes by antisense oligonucleotides and RNA interference.113, 128 These methods reduce trans-porter expression by interfering with the mRNA translation. Transporter contribution can also be estimated by using selective inhibitors. For trans-porters dependent on specific cofactors, e.g., the sodium-dependent uptake transporter NTCP, yet another approach exists. The transporter can be stud-ied in the absence and presence of the cofactor to determine its contribution.44

Predicting Change in Exposure by a Transporter-Mediated Drug–Drug Interaction With knowledge of the relative transporter contributions, predictions of the impact of a transporter-mediated DDI on the in vivo transporter-mediated CL can be obtained using a static model (eq. 7).129

30

[ ]

=

+

+

=n

j

ji

je

KI

f1

,

,I- r,transporte

I r,transporte

1

1

CL

CL (7)

where CLtransporter, +I and CLtransporter, -I is the transporter-mediated CL in the presence and absence of inhibitor respectively, fe,j is the fraction of transport over a given membrane mediated by transport protein j, [I] is the inhibitor concentration, and Ki,j is the inhibition constant for transport protein j. A considerable change in CL (> 2-fold) is expected only if the transport protein contributes to more than 50% of the total transport across the membrane and the inhibitor concentration exceeds the Ki.129

It is the unbound fraction of the drug which exhibits the inhibitory func-tion; hence, the unbound concentration of inhibitor should be used in the DDI predictions. Although the inhibitor concentration at the target site changes over time in vivo, the static model maintains the maximal unbound concentration to avoid false-negative predictions.130 This leads to an over-estimation of the DDI risk. For hepatic uptake transporters, the inhibitor concentration is described by the unbound concentration in the blood, whereas efflux transporters are exposed to the unbound concentration in hepatocytes. Given that the liver is involved in the first pass extraction of drugs, the maximal unbound concentration measured systemically (after passage through the liver) is less than that at the inlet to the liver. Practical aspects, however, limit the measurement of inhibitor concentrations in the hepatic portal vein. Instead, the maximal unbound concentration at the inlet to the liver (Iu,in,max) can be estimated by equation 8.131

h

gaauuu Q

DoseFFkfII

×××+= max,maxin,, (8)

where Iu,max is the observed maximum unbound concentration of inhibitor in the blood, fu is the fraction unbound in blood, ka is the absorption rate con-stant, FaFg represent the intestinal availability (fraction of drug absorbed into the portal vein from the intestinal lumen), and Qh is the hepatic blood flow rate.

Mechanistic Pharmacokinetic Modeling Mechanistic pharmacokinetic models are useful tools in the study of com-plex systems such as the liver and its role in drug disposition. These models are typically developed as a set of mathematical equations that describe the processes underlying the behavior of the system being modeled. A major challenge in mechanistic modeling is to incorporate enough detail to accu-

31

rately predict the output, while limiting the complexity to avoid overly long simulation times. Every model therefore relies on one or more assumptions. A substantial benefit of the mechanistic approach of modeling is that it can simulate the influence of various parameters on the final outcome.132

Mechanistic pharmacokinetic models can be used to predict drug disposi-tion and drug–drug interactions in vivo.133, 134 The dynamic feature of these models allows for the assessment of DDIs on the basis of temporal changes in the free concentration of inhibitor and substrate at the site of interaction. This should provide more realistic predictions of DDI risks than those ob-tained using static models, provided that the dynamic model incorporates enough detail to describe the system properly. Furthermore, the dynamic models allow for simultaneous interplay of multiple processes that influence drug disposition, such as hepatic uptake and metabolic/biliary elimination.

In silico Predictions of Transport Inhibitors Statistical multivariate models can identify correlations between variables in large data sets. This can be exploited to derive QSAR models that predict a given property (e.g., transport inhibition) from a set of molecular de-scriptors.135 The molecular descriptors describe the chemical information contained in the molecule numerically. Examples of molecular descriptors include molecular weight, number of atoms, and maximum positive charge.

The QSAR models are developed using a training set, i.e., a set of mole-cules for which there are experimental data on the property in question. The models are consequently only as good as the data upon which they are based. Once the model is trained, it can be used to make predictions on molecules not included in the training set. Generally, reliable predictions can only be expected for molecules that are similar to those included in the training set.136

Proteomics The extent of hepatic drug elimination is dependent on the abundance and activity of the transporters and enzymes involved in its disposition. Histori-cally, abundance has been assessed by gene expression analysis. However, many reports have shown that protein levels correlate better with activity than mRNA levels do.137-139 This is not surprising given the many processes that influence protein abundance besides mRNA transcription. For example, translation rates vary for different mRNA transcripts and explain much of the differences in protein abundance.140 Translation is also subject to regula-tion by e.g., microRNAs, which can repress the translation of mRNA tran-scripts.141 Since proteins, rather than mRNA, are responsible for carrying out cellular functions, protein abundance measurements are considered useful in

32

estimating cellular activities. Traditional immunological methods are, at best, semi-quantiative. Instead, mass spectrometry has emerged as a more accurate method of quantifying protein abundance. There are two different types of quantitative mass spectrometry methods: label-based and label-free quantification. Both methods rely on the measurement of peptide levels as a surrogate for the protein. The peptides are produced by enzymatic cleavage of the proteins in the sample by different proteases. Trypsin is the most commonly used, but depending on the protein of interest another protease may produce better-suited peptides for the analysis.142

Label-Based Quantification of Protein Abundance The label-based approach of protein quantification, also known as targeted proteomics, focuses on determining the concentration of a specific protein (or set of proteins) of interest. It involves the use of a labeled standard that is spiked into the sample in a known amount. Because the quantification strat-egy involves relative quantification against an internal standard, the accuracy is only as good as the quantification of the standard itself.

The standard can be introduced at different stages of the sample prepara-tion, either as an isotope-labeled protein or isotope-labeled peptide.143 Pro-tein standards are introduced earlier in the workflow (before protein diges-tion), whereas peptide standards are introduced later (during or post-digestion). The earlier the standard is introduced, the better. An early intro-duction can eliminate systematic errors in the sample handling, such as in-complete protein digestion. Purified transport protein standards are, howev-er, expensive and difficult to produce. Hence, the most common approach for targeted transporter quantification is to use peptide standards.144-146 Accu-rate determination of protein levels with this approach requires complete protein digestion since the stoichiometric relationship between the amount of proteolytic peptide and corresponding protein is not established.

Because peptides are quantified as surrogates for protein, selection of suitable target peptides is critical for the label-based quantification. The sur-rogate peptide has to be unique for the target protein and have optimal prop-erties for mass spectrometry analysis. In addition, it is important that the selected peptide does not include any known sites for post-translational mod-ifications or single nucleotide polymorphisms.146, 147

Label-Free Quantification of Protein Abundance While label-based protein quantification with spiked-in standards is restrict-ed to a limited set of target proteins, label-free shotgun proteomics can be used for whole proteome analyses. Shotgun proteomics has historically been used for qualitative analysis of proteins (i.e., protein identification), but more recently label-free protein quantification methods have emerged.148 One such

33

example is the total protein approach (TPA), which has proven to provide accurate estimations of protein abundance.149, 150 The TPA-based quantifica-tion of proteins relies on information obtained directly from the mass spec-trometer readout and does not require any standards. A prerequisite, howev-er, is a sufficient depth of the proteomic analysis since the underlying as-sumption is that the detected proteins constitute the major part of the prote-ome by mass. The number of proteins required for an appropriate analysis depth will vary with cell/tissue type. For example, the plasma proteome is dominated by less than thirty abundantly expressed proteins,151 whereas hepatocytes express thousands of proteins at relatively high levels.152 Like other label-free protein quantification methods, the TPA approach assumes proportionality between measured peptide intensities and protein quantities. Under this assumption, the fractional value of the summed signal intensities of all peptides identifying a particular protein to the total MS-signal de-scribes the protein abundance as a fraction of total mass.

Statins as Model Drugs Statins are a class of widely prescribed drugs to lower cholesterol levels. They mediate their effect through inhibition of 3-hydroxy-3-methyl-glutaryl-coenzyme A (HMG-CoA) reductase, the rate-limiting enzyme in the choles-terol biosynthesis pathway.153 The target is located in the liver, which is also the major eliminating organ of this drug class. Statins are generally well tolerated, but up to 10% of persons experience myopathy.154, 155 These mus-cle symptoms can also, in rare cases, lead to life-threatening rhabdomyolo-sis.156

In 2008, a genome-wide association study discovered that patients with reduced function genetic variants of the hepatic uptake transporter OATP1B1 are at increased risk of statin-induced myopathy.157 This associa-tion is supported by numerous clinical studies that report of increased statin exposure in individuals with reduced function genetic variants of OATP1B1.158-161 Another risk factor for statin-induced myopathy is trans-porter-mediated DDIs resulting in decreased hepatic uptake and, hence, in-creased systemic concentration of the statin.162

Statins are administered either in the active acid form or as the more lipo-philic lactone prodrug, which requires conversion to the active form via hy-drolysis. These two forms co-exist in vivo due to reversible interconversion. Although all statins are substrates of transport proteins, they differ in their interaction patterns and hence, their dependency on specific transporters for hepatic elimination. Some statins are metabolized to a large extent, rendering them vulnerable for metabolic DDIs as well. The section below summarizes the knowledge of hepatic transporters and metabolizing enzymes involved in the pharmacokinetics of the two statins used as model drugs in this thesis,

34

atorvastatin and pitavastatin (Figure 6). Both of these statin drugs are admin-istered in their active acid form.

Figure 6. Molecular structures of atorvastatin and pitavastatin acid.

Hepatic Disposition of Atorvastatin Hepatic clearance of atorvastatin occurs through transporter-mediated uptake followed by metabolism.163 The metabolites are mainly eliminated by excre-tion into bile, while renal secretion represents a minor elimination pathway (< 2%).164

At least four transporters on the basolateral membrane of human hepato-cytes recognize atorvastatin as a substrate. These are OATP1B1, OATP1B3, OATP2B1, and NTCP. At the onset of this work, OATP1B1was known to play an important role in the hepatic uptake as demonstrated by altered pharmacokinetics of atorvastatin in individuals with SLCO1B1 polymor-phism.160 Shitara and colleagues estimated a 47% contribution of OATP1B1 to the total hepatic uptake from the observed change in the area under con-centration-time curve (AUC).165 A similar contribution of OATP1B1 (50%) was reported in a recently published study as predicted from protein abun-dance and activity data.166 The same study predicted OATP1B3-mediated uptake to represent around 13% of the total active uptake of atorvastatin.

Following hepatic uptake, atorvastatin is almost completely metabolized by CYP3A4. Because the lactone form of atorvastatin has a much higher affinity for CYP3A4 than the acid form, metabolism of the lactone form is expected to be the relevant metabolic clearance pathway. The lactone me-tabolites can be interconverted to the active open acid metabolites ortho- and parahydroxylated atorvastatin. Both the parent compound and its two active metabolites can also undergo glucuronidation by UGT1A1 and UGT1A3 followed by subsequent lactonisation.163 In addition, the glucuronidated me-tabolites may undergo deconjugation back to the active metabolites.

Atorvastatin and its metabolites are primarily eliminated into the bile. Transporter knock-down studies in Caco-2 cells indicate that atorvastatin is transported by Pgp, BCRP, and MRP2.167 ATPase activity in membranes expressing MRP2 also suggests MRP2-mediated transport.168 Experiments in a BCRP-expressing cell line confirmed atorvastatin to be a BCRP

35

substrate.169 In addition, clinical studies show that ABCB1 and ABCG2 pol-ymorphisms are associated with increased AUC of atorvastatin, indicating that both Pgp and BCRP are important for the in vivo disposition.169, 170

Hepatic Disposition of Pitavastatin In contrast to atorvastatin, which is heavily metabolized, pitavastatin is mainly eliminated unchanged into the bile. Renal excretion is involved in the elimination to a lesser extent.164 Because of its efficient biliary excretion and high bioavailability compared to most other statins, pitavastatin has been suggested to undergo enterohepatic recirculation.171

Pitavastatin is an in vitro substrate of OATP1B1, OATP1B3, OATP2B1, and NTCP.120, 172, 173 Predictions of the OATP1B1-mediated contribution to the total active uptake in vivo range from 32% to 98%.120, 128, 166 The large (162–191%) increase in pitavastatin AUC in individuals with OATP1B1 polymorphisms compared to those expressing the wildtype variant suggests that OATP1B1 is critical to the in vivo disposition.161, 174-176 In line with this, the in vivo data indicate that 70% of pitavastatin is taken up by OATP1B1.164 OATP1B3 is predicted to play a less important role in the uptake with an estimated contribution of 1.8-25%.120, 166 Early studies of the hepatic uptake of pitavastatin in the presence and absence of sodium suggested that pitavas-tatin is mainly taken up in a sodium-independent manner.120 This indicated that NTCP was not important for the uptake. More recent data, however, assign 29% of the total active uptake across the basolateral membrane of hepatocytes to NTCP.44 The role of OATP2B1 in pitavastatin disposition has not been investigated.

Pitavastatin is metabolized to a very limited extent. The major metabolite formed is pitavastatin lactone, which can be reversibly converted to the par-ent drug. The lactone metabolite is formed by UGT-mediated glucuronida-tion, followed by non-enzymatic conversion of the glucuronidated metabo-lite to the lactone form.177 Pitavastatin is also subjected to oxidative metabo-lism by CYP2C9, producing the minor metabolite 8-hydroxypitavastatin.

Transporter-mediated biliary excretion is the major clearance route for pitavastatin. While Pgp did not interact with the acid form of pitavastatin in one in vitro study,173 other research groups have observed Pgp-mediated transport of pitavastatin in vitro.172, 178 A recent clinical study found an asso-ciation between a single nucleotide polymorphism in ABCB1 and pitavas-tatin AUC,179 validating the role of Pgp in pitavastatin disposition. Pitavas-tatin is also a substrate of BCRP in vitro,173, 178 but its pharmacokinetics is not influenced by genetic variants of BCRP in vivo.175, 176 In addition, MRP2 has been shown to transport pitavastatin in vitro.168, 178 A genetic variant of MRP2 is associated with altered pitavastatin exposure.176

36

Aims of the Thesis

The general objective of this thesis was to provide new predictive in silico and in vitro models of in vivo hepatic transporter-mediated drug clearance and drug–drug interactions. The specific aims were: • To investigate the inhibition overlap of OATP1B1-, OATP1B3-, and

OATP2B1-interacting drugs

• To develop in silico models for the prediction of OATP inhibitors

• To develop static models for the prediction of in vivo transporter-mediated clearance and drug–drug interactions from in vitro experi-ments and protein abundance data

• To map and compare the abundance of transporters and other pro-

teins of importance for drug disposition in human liver tissue and isolated human hepatocytes

• To develop a dynamic mechanistic pharmacokinetic model for the

prediction of hepatic drug disposition from in vitro experiments and protein abundance data

37

Methods

Compound Selection In paper I, the aim was to develop in silico models that could predict OATP inhibitors on the basis of their physicochemical properties. For the model to be applicable to a wide range of drugs, compounds were chosen to represent the chemical space of orally administered drugs. In addition to the drug-likeness and molecular diversity, compounds known to interact with the OATP transporters were included as positive controls and to increase the number of hits for computational modeling. To enable high-throughput screening of OATP inhibition, radioactively labeled compounds (estradiol-17β-glucuronide and estrone-3-sulphate) were used as probe substrates.

Papers I-II investigated the hepatic uptake clearance of atorvastatin in ab-sence and presence of inhibitors. Atorvastatin was chosen as a model drug due to its dependence on multiple transporters (OATP1B1, OATP1B3, OATP2B1, and NTCP) for its hepatic uptake. Although atorvastatin was known to be a substrate for these transporters in vitro, their relative contribu-tion to the uptake in vivo was at the time unknown.

In papers III-IV, pitavastatin was used as a model drug. Unlike atorvas-tatin, which is heavily metabolized in vivo, pitavastatin is mainly excreted unchanged in the bile. Consequently, its hepatic disposition is almost com-pletely dependent on active transport processes, which was the reason for its selection. Paper III characterized the uptake of pitavastatin in human liver and hepatocytes. For a further understanding of hepatic disposition of pitavastatin, the interplay between uptake and efflux in sandwich-cultured human hepatocytes was investigated in paper IV using a mechanistic phar-macokinetic model.

Kinetic Characterization of Transporter-Mediated Uptake and Efflux Papers I-IV used HEK293 cell lines stably transfected with a single trans-porter (OATP1B1, OATP1B3, OATP2B1, or NTCP) to study transporter-mediated uptake of drug compounds and probe substrates. Active transport via the efflux transporters BCRP, Pgp/MDR1, MRP2, MRP3, and MRP4

38

was studied using inverted membrane vesicles derived from HEK293 and Sf9 cells (paper IV).

The kinetic parameters describing transporter-mediated uptake/efflux were determined by measuring initial transport rate at varying substrate con-centrations. In brief, cells or inverted membrane vesicles were incubated with substrate for a short interval. All experiments were performed in the initial linear range of the time-dependent uptake. Transport was stopped by the addition of cold buffer and the cells/membrane vesicles were washed before analysis of accumulated substrate by either scintillation counting (probe substrates) or liquid chromatography-tandem mass spectrometry (LC-MS/MS, drug compounds).

The relation between measured transport rate (v) and substrate concentra-tion ([S]) was described by the Michaelis-Menten equation with the addition of a non-saturable passive diffusion (Pdif) rate component (eq. 9). Using non-linear regression, the maximal transport rate (vmax) and the substrate concen-tration at which the uptake rate is half of vmax (Km) were determined and used to describe transport kinetics.

[ ][ ] [ ]SPSK

Svv dif

m

×++×= max (9)

In vitro Screening of Transport Inhibition In papers I and II, transfected HEK293 cells and isolated human hepatocytes were used to study inhibition of transporter-mediated uptake, respectively. Cells were incubated with a substrate concentration at or below Km in the presence or absence of a potential inhibitor, allowing for the calculation of percent inhibition. In paper I, inhibition was screened at a single concentra-tion of 20 µM. Compounds that inhibited the uptake by at least 50% were classified as inhibitors. In paper II, inhibition was screened at in vivo-relevant concentrations. In addition, concentration-dependent inhibition in HEK293 cells was studied in both papers I and II. The half maximal inhibi-tory concentration (IC50) was determined by fitting the data to equation 10.

[ ] slope Hill)IClog(log 50101

BottomTopBottom control) of (%Transport ×−+

−+=I

(10)

where [I] is the inhibitor concentration and the Hill slope describes the steepness of the curve. A top plateau value of 100% and a bottom plateau value of 0% were used as constraints when fitting the data. This was to allow for estimations of IC50-values despite occasionally incomplete inhibition curves due to solubility limitations of the perpetrator drug. The IC50-values

39

were used to calculate corresponding inhibition constants, Ki, with the rela-tionship described by equation 11 (assuming competitive inhibition).

[ ]m

i

K

SK

+=

1

IC50 (11)

Quantifying Isoform Specificity of Inhibitors The promiscuity (or inversely, the specificity) of the inhibitors in paper II was calculated using the index of inhibition promiscuity introduced by Nath et al.180 This entropy-based metric of drug promiscuity requires the definition of inhibitor potencies (xi) such that xi is higher for more potent inhibitors and always greater than zero. Inhibitor potency was defined by equation 12.

1000

IClog 50−=ix (12)

For pairs of transporters and perpetrator drugs that did not display sufficient inhibition for accurate IC50 determinations, the IC50-value was set to 999 µM. The inhibition promiscuity (Iinh) for each perpetrator drug was subsequently calculated using equation 13.

=

==

−=N

iN

jj

iN

jj

iinh

x

x

x

x

NI

1

11

loglog

1 (13)

where N is the number of transporters in the panel, xi is the inhibitory potency toward transporter i, and xj is the inhibitory potency toward transporter j. The index of inhibition promiscuity can take on any value between 0 (indicating a completely specific inhibitor) and 1 (indicating a completely promiscuous inhibitor with equal potency against all transporters). The specificity was calculated as 1-Iinh.

Isolation and Culture of Primary Human Hepatocytes Non-tumorous excess human liver tissue was obtained from liver resections under ethical approval (ethical approval no. 2009/028 and 2011/037). All donors gave their informed consent. Human hepatocytes were isolated from

40

the liver tissue according to the two-step collagenase perfusion method.83 In brief, the tissue was first perfused with a calcium-free buffer containing the chelating agent EGTA (ethylene glycol tetraacetic acid). Calcium was removed to rapidly disrupt intercellular junctions, resulting in the loss of cell-cell contacts. In the second step, the tissue was perfused with colla-genase to digest the connective tissue of the liver, resulting in the release of dispersed hepatocytes after completed digestion. The crude hepatocyte prep-aration was subsequently filtered to remove undigested material and washed three times by centrifugation to remove dead and damaged cells. The second wash consisted of a Percoll density centrifugation to remove NPCs and en-rich for viable hepatocytes.

Isolated hepatocytes were suspended in Dulbecco’s Modified Eagle’s Medium supplemented with fetal bovine serum, penicillin-streptomycin (PEST), insulin, and dexamethasone. Cells were seeded in 24-well collagen-coated plates at a density of 375,000 viable cells per well and were allowed to attach for 2-3 h at 37°C and 5% CO2 atmosphere. After attachment, medi-um was replaced with Hepatocyte Maintenance Medium supplemented with PEST, insulin-transferrin-selenium, and dexamethasone. For cultures kept longer than 24 h, hepatocytes were overlaid with Matrigel (BD Biosciences, Franklin Lakes, NJ, USA). Sandwich-cultured hepatocytes were maintained in culture for an additional 5 days post-overlay to allow for bile canaliculi formation. Culture medium was replaced daily.

Characterization of NTCP Contribution to Hepatic Transporter-Mediated Uptake The contribution of the sodium-dependent transporter NTCP to the net up-take of atorvastatin and pitavastatin in plated hepatocytes was assessed by comparing uptake in the presence and absence of sodium. The hepatocytes were incubated with 1 µM substrate in sodium-containing or sodium-free (NaCl and NaCHO3 replaced by choline chloride and KHCO3) Krebs-Henseleit bicarbonate buffer. After 2 min incubation, uptake was stopped by addition of cold buffer and the cells were washed. Intracellular accumulation of drug was quantified by LC-MS/MS.

Characterization of Drug Disposition in Sandwich-Cultured Human Hepatocytes Paper IV investigated time-dependent uptake and efflux of pitavastatin in SCHH. In brief, SCHH cells were pre-incubated in the presence of divalent cations, to maintain the integrity of tight junctions, or in their absence, to

41

disrupt the tight junctions and open up the bile canaliculi. Substrate was added and the cells were incubated for a specified time. The incubation me-dium was removed and the cells were washed in cold buffer before analysis of accumulated substrate by LC-MS/MS. The amount of drug accumulated intracellularly was calculated from the incubations in the absence of Ca2+ and Mg2+, whereas net accumulation in cells and bile canaliculi was deter-mined from the incubations in the presence of divalent cations. Substrate depletion from the medium was also measured by taking samples from the medium at representative time points.

Quantitative Analysis of Protein Abundance by Liquid Chromatography-Tandem Mass Spectrometry Transporter abundances were determined by LC-MS/MS. The two different methods are described in more detail below. Both methods rely on the meas-urement of peptide levels as a surrogate for protein abundance. The general workflow applied to the samples consisted of two steps. First, samples were fractionated to enrich the membrane proteins. Second, the proteins were enzymatically cleaved into peptides by trypsin for LC-MS/MS analysis.

Targeted Proteomics The abundance of selected transport proteins in human liver tissue, isolated human hepatocytes, SCHH, HEK293 cell lines, and inverted membrane ves-icles were determined by targeted proteomics. Unique peptides representing each transporter were selected for their favorable LC-MS/MS properties. Stable isotope-labeled peptides with sequences identical to the selected tar-get peptides were synthesized. The tissue and cell samples were lysed and membrane fractions were extracted using the ProteoExtract® Native Mem-brane Protein Extraction Kit (Calbiochem, San Diego, CA, USA). The membrane vesicles, which are already an enriched membrane fraction, were used as they were. The membrane proteins were trypsin-digested into pep-tides. and isotope-labeled peptides were added as internal standards. The peptide samples were analyzed by LC-MS/MS with selective monitoring of the transitions corresponding to the target peptides. The abundance of each transport protein was determined by comparing the peak area of the unla-beled target peptide with that of the heavy isotope-labeled internal standard.

Global Proteomics The abundance of transporters and other membrane proteins in liver tissue and isolated hepatocytes was also determined by untargeted, label- and

42

standard-free LC-MS/MS. Here, ultracentrifugation was used to isolate crude membrane fractions. The membrane proteins were solubilized with a high concentration of SDS followed by depletion of the detergent using the filter-aided sample preparation (FASP) method.181 In brief, SDS was dissoci-ated from proteins by addition of urea and subsequently removed by centrif-ugation within a filtration device. Proteins on the filter were digested with trypsin and the resulting peptides were analyzed by LC-MS/MS. Protein abundance levels were calculated by the total protein approach (TPA), a method developed for protein quantification from in-depth shotgun proteo-mic data. Under the condition that peptide intensity is proportional to protein quantity, the fraction of the MS-signals of all peptides identifying a particu-lar protein to the total MS-signal describes its abundance as a fraction of total mass.

In vitro–in vivo Scaling of Hepatic Transport and Transporter-Mediated Drug–Drug Interactions For scaling in vitro transport capacity (vmax,in vitro) to the in vivo situation, transporter abundance in the in vitro system was compared to that in human liver tissue using targeted proteomics. By assuming that the maximal transport rate is directly related to protein abundance, we used equation 14 to predict the maximal transport rate in vivo. Transporter-mediated drug clear-ance (CLtransporter) was then predicted using equation 15. Scaling factors of 88 mg protein/g liver182 and a liver weight of 1800 g183 were used to extrapolate to whole liver.

vitroinin vitro vmax,

liverabundancer Transporte

abundancer TransporteMTA ×= (14)

[ ]SKm += MTA

CL rtransporte (15)

where MTA is the maximal transport activity or vmax in vivo.

With knowledge of the transporter-mediated CL, predictions of the extent of drug–drug interactions were made using equation 16. The predictions were based on the underlying assumption of competitive inhibition.

[ ] [ ]SK

IK

im +

+×

=+

1

MTACL I r,transporte (16)

43

Mechanistic Pharmacokinetic Modeling of Hepatic Drug Disposition Hepatic drug clearance in vivo is a dynamic parameter that changes over time as a result of alterations in the blood and tissue concentrations of the drug. To capture the dynamic feature of drug distribution, a mechanistic pharmacokinetic model was built in R (www.r-project.org) describing the relation between drug uptake, binding, and efflux in SCHH experiments (Figure 7).

Figure 7. Model used to simulate pitavastatin disposition in sandwich-cultured hu-man hepatocytes (SCHH). The model included three different compartments de-scribing the amount of drug in the medium, cell and bile. Drug distribution was described by non-saturable passive diffusion clearance (CLpassive) and saturable transporter-mediated clearance (CLNTCP, CLOATP1B1, CLOATP1B3, CLOATP2B1, CLMRP3, CLBCRP, CLPgp). Dynamic pulsing of the bile canaliculi, causing release of accumu-lated drug in the bile compartment to the medium, was described by the Kflux param-eter. The change of drug amount in the various compartments over time was de-scribed by the differential equations shown in the figure.

Kinetic parameters describing the transporter-mediated uptake and efflux of pitavastatin were derived by in vitro experiments in cell lines and membrane vesicles overexpressing a single transporter. The maximal transport velocity was scaled to SCHH by the difference in transporter expression using the MTA-based approach described above. Transporter abundance levels in SCHH cultures at day 5 post-overlay, transfected HEK293 cells, and mem-brane vesicles were quantified by targeted LC-MS/MS. Intracellular drug

44

binding was described by the fraction unbound value reported by Menochet and colleagues.184

The impact of changes in transporter activities on simulated time profiles was investigated by varying the maximal transport rate of each transporter 10 fold above and below its experimental value while keeping the other pa-rameters fixed.

Statistical Analysis All experiments were performed in duplicate or triplicate and on at least two separate occasions. Statistically significant differences were assessed by un-paired t-test, one-way or two-way ANOVA with post-hoc test as appropriate. Corrections were made for multiple comparisons.

Correlations in Protein Expression Spearman rank correlation tests were applied in R with the additional pack-age Hmisc to investigate correlations in protein expression levels among drug metabolizing CYP and UGT enzymes. Only those enzymes with quan-tified abundance levels in at least half of the liver and hepatocyte samples were included. The correlation coefficients were adjusted for multiple testing using the Benjamini-Hochberg correction.

Pathway Analysis Proteins with a significant change in abundance for liver tissue and isolated hepatocytes were identified by a two-sample t-test as implemented in Per-seus. A permutation-based false discovery ratio of 0.05 (1000 permutations) was used to define statistical significance. Bioinformatic analysis of the dif-ferently expressed proteins was performed with the Ingenuity Pathway Analysis (IPA) software (http://www.ingenuity.com). The analysis was re-stricted to human experimental observations for liver tissue or hepatocytes. Canonical pathways, from the IPA library of canonical pathways, most sig-nificantly associated with the significantly altered proteins were identified.

45

Results and Discussion

Identification and Prediction of Inhibitors of the Hepatic Organic Anion Transporting Polypeptide (OATP) Transporters In paper I, 225 compounds were screened for inhibition of OATP1B1-, OATP1B3-, and OATP2B1-mediated uptake of probe substrates with the aim of developing predictive in silico models. In total, 91 compounds were classified as inhibitors, characterized by ≥ 50% reduction of the transporter-mediated uptake at a concentration of 20 µM (Figure 8). Over half of these compounds were inhibitors of more than one OATP isoform, revealing a large overlap for these important uptake transporters (Figure 9). Screening identified 4 novel OATP1B1 inhibitors, 16 previously unpublished inhibitors of OATP1B3, and as many as 29 OATP2B1 inhibitors unreported elsewhere. At a concentration of 20 µM, 27, 3, and 9 compounds were identified as selective inhibitors of OATP1B1, OATP1B3, and OATP2B1, respectively. Two selective compounds, pravastatin (OATP1B1-selective) and erlotinib (OATP2B1-selective), were confirmed by assessing their concentration-dependent inhibition of the three OATP transporters. Interestingly, the con-centration-dependent inhibition screen showed that both cyclosporine and rifampicin interacted only weakly with OATP2B1, yet the International Transporter Consortium recommends them as general OATP inhibitors. Cy-closporin and rifampicin should therefore be considered OATP1B inhibitors rather than general OATP inhibitors. Moreover, three compounds commonly used as efflux transporter inhibitors (Hoechst33342, KO143, and MK571) inhibited OATP-dependent uptake as well. MK571 inhibition of the OATP-mediated uptake was even more potent than that reported for the efflux transporters. This emphasizes the difficulty of interpreting transporter inter-actions in complex systems such as hepatocytes, due to the large overlap of interacting drugs. With this in mind, cell lines overexpressing a single trans-porter are useful complements to the more complex cell models by pinpoint-ing the transporter(s) of importance.

46

Figure 8. Interaction of 225 drugs and drug-like compounds with the OATP1B1-, OATP1B3-, and OATP2B1-mediated transport of probe substrate. Compounds were classified as inhibitors if they reduced the transporter-mediated uptake by at least 50% at a concentration of 20 µM. Almost twice as many OATP1B1 inhibitors as OATP1B3 or OATP2B1 inhibitors were identified. Modified from Karlgren et al., J Med Chem, 2012, 55(10):4740-63.

Figure 9. Overlapping inhibitor specificities of the hepatic uptake transporters OATP1B1, OATP1B3, and OATP2B1. There was a large overlap in drug interac-tions, in particular for OATP1B1 and OATP1B3. The majority of OATP1B3 inhibi-tors inhibited OATP1B1 as well, making it difficult to identify OATP1B3-specific inhibitors. Modified from Karlgren et al., J Med Chem, 2012, 55(10):4740-63.

Discriminative multivariate models developed from the inhibition data predicted inhibitors of each OATP with high sensitivity (>75%) and high specificity (>70%), Table 1. High lipophilicity and large polar surface area (PSA) were identified as distinguishing molecular features of OATP inhibi-tors. On the basis of these two properties, an easily interpreted predictive

47

model was developed that could predict inhibitors of any OATP transporter with a sensitivity of 85% for the test set. The finding that OATP inhibition was favored by both high lipophilicity and large PSA, which are negatively correlated molecular features, may be explained by the presence of multiple binding sites on the transport proteins.185, 186

Table 1. Performance of the developed binary classification models.

Correctly classified

Model Type of data set Inhibitors

(%)Non-inhibitors

(%)

OATP1B1 Training 73 76Test 79 73

OATP1B3 Training 81 87Test 92 72

OATP2B1 Training 77 80Test 75 77

Our study was the first to comprehensively analyze inhibition of each he-

patic OATP isoform from a physicochemical molecular perspective. Since then other studies have been published on the same topic, all of them con-firming the importance of lipophilicity and PSA for OATP inhibition.187, 188 The study by De Bruyn and colleagues188 included 151 compounds that were also part of our inhibition screen. When comparing the identified OATP1B1 inhibitors and non-inhibitors among the overlapping compounds, the classi-fication agreed for 79% of them. For OATP1B3, as many as 87% of the overlapping compounds were given the same classification. Discrepancies in the classification may be explained by differences such as the probe sub-strates used or concentration of perpetrator drug in the screens.

Prediction of Interindividual Uptake of Atorvastatin and Pitavastatin in Human Liver Papers I and II identified atorvastatin as an in vitro substrate of OATP1B1, OATP1B3, OATP2B1, and NTCP. Pitavastatin was also found to be trans-ported by all four hepatic transporters (papers III and IV). The kinetics of the transporter-mediated uptake of the two statin drugs was determined in HEK293 cell lines overexpressing each of the transporters. The uptake CL in the liver was predicted by relating the transporter expression in human liver tissue to that in the transfected HEK293 cells. This approach allowed for estimations of the relative contributions of the four transporters to the total active uptake.

48

From measured transporter expression levels in twelve liver-tissue sam-ples, we predicted a two-fold interindividual variability in transporter-mediated uptake of atorvastatin. Despite the low interindividual variability in overall uptake CL, there were large differences in the contribution of the four uptake transporters (Figure 10). Nevertheless, OATP1B1 and OATP1B3 were identified as the major transporters responsible for the up-take of atorvastatin in the liver. The OATP1B1-mediated contribution was estimated to be 53% on average, which is very close to the reported value of 47%, calculated from in vivo AUC changes in individuals with reduced func-tion genetic variants of OATP1B1.165 The good agreement with in vivo data supports our in vitro to in vivo scaling approach. OATP1B3 contribution was predicted to be 35% on average. OATP2B1 and NTCP were found to be less important for the hepatic uptake of atorvastatin than OATP1B1 and OATP1B3. The minor contribution of the sodium-dependent transporter NTCP was confirmed by human hepatocyte experiments in the presence and absence of sodium (Figure 11).

Figure 10. Prediction of interindividual uptake clearance (CLuptake) of atorvastatin by four transporters in the liver. Hepatic uptake CL was predicted from in vitro transport kinetics by correcting for the difference in membrane transport protein expression between in vitro cell models and human liver tissue. Modified from Vildhede et al., DMD, 2014.

49

Figure 11. NTCP-mediated uptake of atorvastatin and pitavastatin in human hepato-cytes. The sodium-dependent, NTCP-mediated contribution to the overall uptake was determined by comparing the uptake in the presence (+Na+) with that in the absence (-Na+) of sodium. NTCP did not contribute significantly to the transport of atorvastatin or pitavastatin. Uptake of the control substrate taurocholate, on the other hand, was significantly less (p < 0.001) in the absence of sodium.

Pitavastatin uptake in human liver was predicted to be very similar for the twelve individuals (Figure 12). Transporter contributions, however, varied to a larger extent. OATP1B1 was identified as the major transporter with 74% contribution on average (range of 49-94%). This estimation agrees with the 70% contribution calculated from clinical data.164 Of the three remaining uptake transporters, OATP1B3 was the second most important with an aver-age contribution of 16%. As for atorvastatin, OATP2B1 and NTCP were predicted to play minor roles in the hepatic uptake of pitavastatin. Uptake studies in the presence and absence of sodium in human hepatocytes con-firmed the negligible contribution of NTCP (Figure 11).

Pitavastatin was more dependent on OATP1B1 for its hepatic uptake than atorvastatin, a finding also supported by clinical data. Individuals with re-duced function genetic variants of OATP1B1 display a greater increase in exposure to pitavastatin than atorvastatin compared to individuals with wild-type OATP1B1.164 Because of its greater dependence on a single transporter, pitavastatin is expected to be more sensitive to transporter-mediated drug–drug interactions than atorvastatin.

50

Figure 12. Prediction of interindividual uptake clearance (CLuptake) of pitavastatin by four transporters in the liver. Hepatic uptake CL was predicted from in vitro transport kinetics by correcting for the difference in membrane transport protein expression between the in vitro cell models and human liver tissue.

Prediction of Interindividual Differences in Transporter-Mediated Drug–Drug Interactions Involving Atorvastatin Paper II studied concentration-dependent inhibition of atorvastatin uptake via OATP1B1, OATP1B3, OATP2B1, and NTCP in transporter-transfected HEK293 cells. The in vitro inhibitory potencies of atazanavir, cyclosporine, gemfibrozil, lopinavir, and rifampicin are summarized in Table 2. All five drugs were found to inhibit OATP1B1-mediated transport with lopinavir demonstrating the highest potency and gemfibrozil the lowest. OATP1B3, OATP2B1, and NTCP were only inhibited by some of the perpetrator drugs (see Table 2).

51

Table 2. Inhibition constants (Ki) of five compounds in interaction with OATP1B1-, OATP1B3-, OATP2B1-, and NTCP-mediated transport of atorvastatin.

OATP1B1 OATP1B3 OATP2B1 NTCP

Compound Ki (µM) Ki (µM) Ki (µM) Ki (µM) Atazanavir 0.42 7.3 >30 >40 Cyclopsorine 0.66 1.3 >13 4.8 Gemfibrozil 58 >420 570 89 Lopinavir 0.34 >4.2 >7.4 >10 Rifampicin 2.2 82 >470 >630

The Ki values were used to predict the extent of transporter-mediated

drug–drug interactions in vivo. Because of the large contribution of OATP1B1 and OATP1B3 to the uptake of atorvastatin, inhibitors of these two transporters were predicted to have a larger effect on in vivo hepatic disposition, and hence drug exposure, than equally potent OATP2B1 and NTCP inhibitors. In addition, the large interindividual variability in trans-porter contribution also affected the extent of drug–drug interactions. Drugs with high OATP1B1 isoform specificity in their inhibition pattern were pre-dicted to cause larger interindividual variability in the remaining atorvastatin uptake than the more general inhibitors (Figure 13).

Figure 13. Interindividual differences in the extent of drug–drug interaction (max/min inhibition) as a function of the isoform specificity of the perpetrator drug. OATP1B1- and OATP1B3-mediated uptake of atorvastatin was predicted to vary considerably between individuals despite a similar summed transporter-mediated uptake. Consequently, the extent of inhibition by the OATP1B1-specific inhibitor lopinavir was associated with a large interindividual variability. Cyclosporine, on the other hand, showed comparable in vitro inhibition potency for OATP1B1 and OATP1B3; it was predicted to inhibit the atorvastatin uptake to a similar degree in all individuals.

52

Comparison of ADME Protein Abundance in Isolated Human Hepatocytes and Human Liver Tissue Isolated human hepatocytes are common tools for the study of hepatic drug disposition, including transport and metabolism. In paper III, the global membrane proteome of isolated human hepatocytes was compared with that of liver tissue. The analysis revealed that drug-metabolizing CYP and UGT enzymes in the intracellular membrane of the endoplasmic reticulum had comparable abundance levels in isolated hepatocytes and liver (Figure 14A-B).

Figure 14. Interindividual protein expression of drug-metabolizing cytochrome P450 (CYP) (A) and UDP-glucuronosyltransferase (UGT) enzymes (B) in human liver (black) and isolated hepatocytes (grey).

53

This allowed for analysis of correlations in protein expression levels be-tween the enzymes (Figure 15). Most UGT enzymes were found to show positive correlations with one another and with CYP2C8 and CYP2C9. This may indicate common regulation factors. Several of the correlations among the UGT enzymes are supported by a meta-analysis on UGT protein abun-dance in the liver,189 while some were only detected in this study. Few CYP enzymes were found to show correlations in abundance. Nevertheless, the significant correlations in our material between CYP2C8 and CYP2C9 (rs = 0.68), between CYP1A2 and CYP2J2 (rs = 0.69), and between CYP1A2 and CYP3A4 (rs = 0.67) are in agreement with the results from a recent meta-analysis.190 The only negative correlation identified was between CYP2C9 and CYP3A7 (rs = - 0.61). CYP3A7 is mainly expressed in fetal liver,191 while CYP2C9 is expressed at much higher levels in adult than fetal liver.192 The negative correlation may thus reflect a change in CYP expression with liver maturation.

Figure 15. Correlations in protein abundance of 24 drug-metabolizing enzymes. Only statistically significant correlations (p < 0.05) are shown. The blue and red colors indicate a positive or a negative correlation in expression, respectively.

In contrast to the metabolizing enzymes, many drug transporters and other integral plasma membrane proteins were significantly less abundant in the isolated hepatocytes than liver tissue (Figure 16). The reduced transporter abundance may explain frequently reported underpredictions of in vivo up-take clearance from hepatocyte experiments. Indeed, we predicted a three-fold lower uptake clearance of both atorvastatin and pitavastatin in freshly isolated hepatocytes than in liver.

54

The transport proteins were not equally affected by the isolation process. OATP1B1, OATP1B3, and NTCP belonged to the drug transporters that were most affected with 3- to 5-fold lower abundance in isolated hepatocytes than liver tissue, whereas OATP2B1 was not affected at all. A possible explanation to this is that the fraction of transporter remaining in the membrane after isola-tion varies from protein to protein depending on the degree of internalization and degradation. This imbalance in functional protein, compared to in vivo, would result in over- or underestimations of the roles of specific transport proteins in drug disposition, with direct implications on DDI predictions. Dif-ferences in transport protein loss between batches may (at least in part) explain the highly variable results that have been published on transporter contribu-tions to the uptake of various compounds. For example, a study on NTCP and the uptake of taurocholate reported an almost 7-fold difference for five differ-ent batches of cryopreserved hepatocytes.193

Figure 16. Interindividual protein abundance of 16 drug transporters in human liver and isolated hepatocytes. MATE1, MRP1, and MRP4 were only detected in the liver samples.

Analysis of pathways associated with proteins whose abundance was signifi-cantly altered in hepatocytes as compared to liver tissue indicated that the hepatocytes were exposed to oxidative stress during the isolation. The hepat-ic exposure to oxidative stress is common to the isolation process as a result of cellular hypoxia-reoxygenation.194 Interestingly, the isolation process was also associated with ubiquitination-dependent proteolysis, providing a possi-ble explanation to the loss of transporters and other plasma membrane-associated proteins.

55

Prediction of Time-Dependent Pitavastatin Disposition in Sandwich-Cultured Human Hepatocytes The methodology developed in paper I, for predicting transporter-mediated drug clearance from kinetics and relative protein abundance, was integrated into a dynamic mechanistic model in paper IV. The model was successfully used to simulate time-dependent uptake and efflux of pitavastatin in three different batches of SCHH (Figure 17). The results showed that functional studies in simple in vitro systems, allowing the study of a single transporter, could be used to simulate drug disposition in more complex systems such as SCHH.

In agreement with the liver predictions, OATP1B1 was identified as the major uptake transporter of pitavastatin with a contribution of 57-87% in the sandwich-cultured hepatocytes. However, in contrast to the liver predictions, NTCP was the second most important transporter instead of OATP1B3. Of the canalicular transporters, BCRP was the major contributor to the biliary efflux with more than 75% contribution across the three SCHH batches.

By altering the transporter activities in the simulations, OATP1B1 was shown to have the largest impact on medium and intracellular drug accumu-lation. Transport activities of the canalicular efflux transporters BCRP and Pgp influenced the amount of drug in bile, but not intracellularly or in the medium.

Figure 17. Comparisons of simulations (lines) and experimental observations (cir-cles) of pitavastatin disposition in three different batches of sandwich-cultured hu-man hepatocytes (SCHH). SCHH (with or without disrupted bile canaliculi) were incubated with pitavastatin at a concentration of 0.1 µM (light grey), 0.6 µM (medi-um grey), or 1 µM (dark grey) for 1-45 min. Note that different scales are used on the y-axis.

56

Conclusions

This thesis developed in vitro and in silico methods for improved predictions of hepatic transporter-mediated drug clearance and drug–drug interactions in vivo. Particular emphasis was placed on interactions involving the hepatic OATP transporters that influence plasma concentration-time profiles of sub-strates such as the commonly administered statins. Several novel inhibitors of the OATP transporters were identified. Discriminative in silico models were developed that could predict OATP inhibitors with high specificity and high sensitivity. These models can be used in early drug discovery to guide compound selection away from unwanted OATP interactions. In addition, a protein abundance-dependent method for prediction of in vivo transporter-mediated clearance and drug–drug interactions was developed. This method can contribute to improved predictions of drug disposition in the human liver and estimations of transporter contributions. Because transporters are pro-miscuous with overlapping substrate profiles, estimations of their contribu-tion to the overall uptake and efflux of new drug entities is important for accurate predictions of DDI risk. In summary, the tools here can contribute to the development of safe and effective drugs as well as to appropriate la-beling of drugs for their safe and effective use. From the work presented herein, it can be concluded that:

• OATP1B1, OATP1B3, and OATP2B1 overlap considerably in their drug interactions as illustrated by the large number of inhibitors in common

• OATP1B1 appears more sensitive to drug–drug interactions than

OATP1B3 and OATP2B1

• Inhibition of the hepatic OATP transporters is governed by and can be predicted from the molecular descriptors lipophilicity and polar surface area

• Quantification of differences in protein expression for in vitro assay

systems and human liver tissue can be used as scaling factors in static and dynamic models to improve predictions of transporter-mediated clearance

57

• Estimations of transporter contributions to the net transport in vivo

are essential for quantitative predictions of drug–drug interactions

• The isolation of hepatocytes from liver tissue is associated with mitochondrial oxidative stress and degradation of transporters and other proteins expressed in the plasma membrane. This has implica-tions for the use of primary hepatocytes as an in vitro model of the liver

58

Future Perspectives

This thesis describes new tools for the prediction of hepatic transporter-mediated clearance and drug–drug interactions. These tools rely on targeted quantifications of transport proteins. Mass spectrometry-based quantification of proteins is an evolving technique anticipated to advance our understand-ing of transporters as determinants of drug pharmacokinetics and possibly pharmacodynamics. Because it is a relatively new technique, methodological improvements can be expected. Such improvements include the use of puri-fied protein standards at known concentrations to address the knowledge gap of protein-peptide stoichiometry. With incomplete protein digestion, surro-gate peptide levels do not reflect the true protein levels. More importantly, digestion efficiency may vary from peptide to peptide and protein to protein. The current lack of protein standards therefore warrants caution when com-paring protein levels determined on the basis of surrogate peptide levels for different studies and different transporters. However, relative expression levels of individual transporters in various samples (as used in this thesis) are reliable as long as sample preparation and digestion conditions are simi-lar across the samples. To improve the accuracy of label-based transporter quantifications, future efforts should focus on the production of transport protein standards of high quality or, alternatively, on optimizing protein di-gestion conditions to maximize the conversion to peptides. An analysis of the current methods could help identify the ones with the most optimal di-gestion conditions for standardization of sample preparation.

An additional research area that needs more focus is the analysis of splice variants and post-translational modifications (PTMs) of proteins. These phe-nomena of protein biosynthesis increase the complexity of the protein space, but have so far received little attention.

The growing body of transporter abundance data will aid in creating vir-tual populations for pharmacologically-based pharmacokinetic modeling purposes. To further improve predictions of liver clearance, models replicat-ing the zonal diversity in the liver are required. It is well-known that the hepatocytes in different zones of the liver express metabolizing enzymes and transporters at varying levels. A thorough characterization of zonal differ-ences in expression levels of proteins determining hepatic clearance would thus be a useful complement to the existing studies on whole liver abun-dance.

59

In the human liver, the transport proteins act in concert with metabolizing enzymes. This highlights the need for a more holistic approach for predicting overall liver clearance and DDI potential. A major challenge for future in-vestigations is to develop mechanistic models capable of describing the transporter-enzyme interplay well enough to allow accurate predictions of drug clearance and DDIs solely from in vitro data. Recently published meth-ods to determine unbound tissue concentrations will be useful in this regard since this is the concentration of drug to which the metabolizing enzymes and efflux transporters are exposed.

Although transport inhibitors can be recognized as such reasonably well from molecular structures, predictions of ligand-protein interactions would be further enhanced by integrating knowledge of the three-dimensional structures of the transport proteins into the modeling. Such structure-based modeling would not only offer information about the potential for inhibition but also the type (competitive or noncompetitive). The latter feature is ad-vantageous given the substrate-dependent inhibition that several hepatic drug transporters display. Structures of human drug transporters are, however, lacking. Hence, efforts to obtain the three-dimensional structures of these proteins by X-ray crystallography or nuclear magnetic resonance spectros-copy are warranted.

60

Svensk Populärvetenskaplig Sammanfattning

Alla organismer, inklusive människan, använder cellmembran som barriärer för att begränsa åtkomsten till cellernas inre. På detta sätt skyddar cellerna sig mot skadliga ämnen som vår kropp utsätts för. Cellmembranet är upp-byggt av ett slags fett, lipider, och av membranproteiner. Lipiderna, framför allt olika fosfolipider och kolesterol, bildar ett dubbellager med en vattenav-visande mitt. Den vattenavvisande egenskapen hos ett cellmembran begrän-sar passagen av hydrofila (vattenlösliga) ämnen, medan ämnen som är hyd-rofoba (fettlösliga) kan passera över membranet enligt lika-löser-lika princi-pen. Vissa membranproteiner bildar kanaler genom cellmembranet, vilka mer eller mindre selektivt transporterar ämnen in eller ut ur cellen. Trans-portproteinerna kan på detta vis t.ex. underlätta ett hydrofilt ämnes åtkomst till cellernas inre miljö eller begränsa passagen av ett hydrofobt ämne genom membranet.

Några av de främmande ämnen som kroppen utsätts för är läkemedel, vil-ket arbetet i denna avhandling handlar om. För att ett läkemedel ska få öns-kad effekt måste det nå sitt målorgan i kroppen i tillräckligt hög mängd. Ef-tersom de flesta läkemedel tas genom att sväljas är det allra första hindret som måste överkommas passage över tarmbarrirärens cellmembran till det förbipasserande blodet. Blodet för sedan läkemedelssubstansen direkt till levern som bland annat har till uppgift att rena blodet från främmande, even-tuellt farliga ämnen. Reningsarbetet utförs av enzymer som finns inne i de leverceller som kallas hepatocyter. Hepatocyterna uttrycker därför en rad olika transportproteiner i sitt membran, vilka hjälper till med att ta upp lä-kemedel från blodet och utsöndra deras nedbrytningsprodukter i gallan eller blodet. Den del av läkemedelsdosen som undgår nedbrytning (metabolism) i levern når blodets systemkretslopp och kan föras vidare till det organ som det är avsett för.

Genom sin roll i nedbrytningen av läkemedel styr levern hur hög nivån av läkemedel i blodet blir. När ett läkemedel ges tillsammans med ett eller flera andra läkemedel kan dess nedbrytning hämmas genom att interaktionen med transportproteinerna eller de metaboliserande enzymen störs. Detta kallas för en läkemedelsinteraktion och får till följd att läkemedlets koncentration i blodet ökar, vilket i sin tur kan leda till biverkningar av mer eller mindre allvarlig grad. Läkemedelsinteraktioner med metaboliserande enzym är ett välstuderat område. På bipacksedeln i läkemedelsförpackningar finns var-ningar för vilka läkemedel som inte bör tas samtidigt på grund av risken för

61

hämmad metabolism. Läkemedelsinteraktioner som involverar transportörer är däremot ett mindre studerat område och har därför undersökts i denna avhandling.

Arbetet i denna avhandling har fokuserat på fyra transportproteiner i le-vern som är involverade i upptaget av läkemedelsgruppen statiner. Statiner används för att minska mängden kolesterol i blodet och därmed risken för hjärt-och kärlsjukdomar. Det är en vanligt förekommande grupp av läkeme-del, framför allt hos äldre. Muskelvärk är en av de vanligare biverkningarna vid statinanvändning. Denna muskelpåverkan kan i värsta fall utvecklas till ett livshotande tillstånd även om detta är mycket ovanligt. Risken för bi-verkningar ökar om statinbehandlingen kombineras med andra läkemedel som ökar nivån av statiner i blodet, t.ex. genom hämning av upptaget till levern.

I första delen av denna avhandling undersöktes förmågan hos 225 läke-medel och liknande ämnen att hämma tre av de transportproteiner i levern som tar upp statiner från blodet. En stor andel av de undersökta ämnena (40%) minskade upptaget via åtminstone ett av de tre transportproteinerna. Med hjälp av dessa experimentella data utvecklades datormodeller som kan användas för att förutsäga om ett nyframtaget läkemedel kommer att tävla med statiner om att tas upp via dessa transportproteiner. Datormodellerna är användbara i ett tidigt skede av läkemedelsutvecklingen för att på ett billigt och enkelt sätt styra utvecklingen bort från oönskade läkemedelsinteraktioner.

I den andra delen av denna avhandling utvecklades en metod som kan an-vändas för att kartlägga hur varje transportprotein bidrar till det samman-lagda upptaget av statiner (eller andra läkemedel) i levern. Metoden bygger på studier av transportfunktionen i speciella cellmodeller där varje protein kan studeras enskilt samt på bestämning av uttrycket av transportproteinerna i dessa cellmodeller och i människans lever. Denna metod applicerades på statinerna atorvastatin och pitavastatin, vilka visade sig vara beroende av främst två transportproteiner för upptaget i levern. Information av denna typ är användbar för att förutsäga hur stor betydelse hämning av ett specifikt transportprotein får på det sammanlagda upptaget i levern. Ju högre bidraget av transportproteinet är till det totala upptaget, desto större påverkan på lä-kemedelskoncentrationen i blodet om detta transportprotein hämmas. Genom att undersöka hur mängden av transportproteinerna varierar mellan olika individer kunde även skillnader mellan olika individer förutsägas.

I den tredje delen av denna avhandling jämfördes uttrycket av tusentals proteiner (inklusive transportproteiner) i levervävnad och i hepatocyter som har isolerats från levervävnad. De isolerade hepatocyterna anses vara den bästa cellmodellen för studier av läkemedelsnedbrytning. Trots detta visade det sig att mängden transportproteiner generellt var lägre i de isolerade hepa-tocyterna än i levervävnad. De läkemedelsmetaboliserande enzymerna som är lokaliserade inne i hepatocyterna uppvisade däremot likvärdiga mängder i hepatocyter och vävnad. Analys av alla de proteiner som mängdmässigt

62

skiljde sig åt mellan de isolerade hepatocyterna och levervävnad pekade på att förlusten av transportproteiner är ett resultat av oxidativ stress som upp-kommer under isoleringen och som leder till nedbrytning av de proteiner som finns i cellmembranet. Resultaten förklarar vanligt förekommande svå-righeter i att förutspå transportörsberoende nedbrytning i levern från försök i hepatocyter. De visar också på fördelen med att använda metoder som tar hänsyn till skillnader i proteinmängd då cellmodeller används för att förut-säga hur mycket nedbrytning som sker i vår lever.

Sammanfattningsvis har arbetet i denna avhandling resulterat i nya dator- och cellbaserade metoder som kan användas till att förutsäga läkemedels-transport i levern samt utvärdera risken för läkemedelsinteraktioner. De framtagna metoderna kan användas för att främja utvecklingen av nya, säkra läkemedel.

63

Acknowledgements

The work presented in this thesis was carried out at the Department of Phar-macy, Faculty of Pharmacy, Uppsala University, Sweden. The studies were financially supported by the Swedish Research Council, the Swedish Fund for Research without Animal Experiments, the Swedish Agency for Innovation Systems, the Lars Hierta Memorial Foundation, and O.E. and Edla Johansson’s Scientific Foundation. IF:s stiftelse, Apotekare C D Carlssons stiftelse, Anna Maria Lundins sti-pendiefond och Leffmans resestipendium tackas för de generösa bidrag som gjort det möjligt för mig att delta i kurser samt åka på internationella konfe-renser.

There are several people who deserve my sincere appreciation and gratitude for their contribution to this thesis. I would like to thank all of you who made this thesis possible in one way or another. My supervisors: Per Artursson för att du antog mig som doktorand i din forskargrupp och för att du i och med detta gav mig en unik möjlighet att lära mig en massa om både mig själv och om den intressanta forskning som innefattar läkemedels-transportörer. Även om vi inte alltid förstår varandra fullt ut så har jag lärt mig oerhört mycket av min tid i din grupp och under din handledning. Jag skulle aldrig vara där jag är idag om det inte var för din strävan efter att bed-riva förstklassig forskning. Jag är väldigt tacksam för att du har tagit dig tid att guida mig på vägen till att bli en självständig forskare. Maria Karlgren för all din hjälp i labbet och för att du introducerade mig för den magiska cellvärlden. Tack för bra samarbete i våra gemensamma projekt och för all stöttning och uppmuntran längs vägen. Ditt systematiska och ef-fektiva arbete är beundransvärt. Vårt gemensamma matintresse har även det varit ett positivt inslag med trevliga middagar på konferensresor och efter accepterat manuskript.

64

Maria Kjellsson för din vägledning under modellerandet. Tack vare dig gick det minst 10 ggr snabbare att få klart modellen än vad det hade gjort om jag hade försökt helt på egen hand. Våra möten var alltid väldigt trevliga och produktiva; även de som inte var planerade, t.ex. när vi stötte på varandra i korridoren på BMC eller på ett träningspass. Ulf Norinder för din insats med den multivariata modelleringen i första ma-nuskriptet. Dina typiskt göteborgshumoristiska skämt var uppskattade även om jag inte alltid var så snabb på att förstå dem. All my other co-authors: Jacek Wisniewski, many thanks for giving us the opportunity to collaborate with you and for the sophisticated method developments over the years of our collaboration. I have also highly appreciated your hospitality during my visits to the proteomics lab at the MaxPlanck Institute. Emi Kimoto and Yurong Lai for your valuable quantitative analysis of transport proteins and helpful comments on the manuscripts that you have co-authored. Ulf Haglund och Agneta Norén för er kirurgiska skicklighet och er positiva inställning till det vetenskapliga arbetet. Utan er insats hade mina studier inte haft samma betydelse. Elin Svedberg för dina experimentella insatser. Du hjälper alltid till när det behövs och är en naturbegåvning i labbet. Tack även för all annan hjälp med administrativa insatser. All the former and present members of the drug delivery group for interest-ing research discussions, useful feedback on my work, and for all the fun in the lab as well as outside of the lab. Work would not have been nowhere near as fun without you. I will truly miss you all. All the former and present PhD students at the Department for all the fun at conferences, after works, dissertation parties, coffee breaks, and events out-side of work. I have particularly enjoyed all the competitive elements of our company, including games, curling, music quizzes and Knally . My office mates Jonas and André. Jonas, tack för att du orkat lyssna på mitt konstanta prat under de år vi delat kontor ihop och för att du alltid har tid för uppmuntrande ord när det är lite tufft och jobbigt. Thank you André for be-ing an excellent research buddy and for taking the time to help me out when needed. I have really enjoyed our research discussions as well as our discus-sions on non-research related topics.

65

My greatest appreciation goes to Terese Bergfors for your invaluable proof-reading and language revision of my thesis and papers II-III. Special thanks also go to Andrea, André, Christine, and Magnus for your thorough proof-reading of this thesis and of paper IV. Stort tack till min familj och mina vänner för att ni alltid finns där. Framför allt vill jag tacka mina föräldrar för att ni alltid har hjälpt till och stöttat mig i allt som jag har gjort. Sist, men absolut inte minst, vill jag tacka min älskade Björn för allt stöd under denna tid. Tack för att du finns där när det är jobbigt och jag behöver få prata av mig, höra några uppmuntrande ord eller få en pussbild skickad till mig.

66

References

1. Prentis RA, Lis Y, Walker SR. Pharmaceutical innovation by the seven UK-owned pharmaceutical companies (1964-1985). Br J Clin Pharmacol 1988. 25(3):387-396.

2. Kola I, Landis J. Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov 2004. 3(8):711-715.

3. Schinkel AH, Wagenaar E, Mol CA, van Deemter L. P-glycoprotein in the blood-brain barrier of mice influences the brain penetration and pharmacological activity of many drugs. J Clin Invest 1996. 97(11):2517-2524.

4. Schinkel AH, Smit JJ, van Tellingen O, Beijnen JH, Wagenaar E, van Deemter L, et al. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs. Cell 1994. 77(4):491-502.

5. Giacomini KM, Huang SM, Tweedie DJ, Benet LZ, Brouwer KL, Chu X, et al. Membrane transporters in drug development. Nat Rev Drug Discov 2010. 9(3):215-236.

6. Aronson PS, Boron WF, Boulpaep EL. Transport of solutes and water, in Medical Physiology: A cellular and Molecular Approach, Editors W.F. Boron and E.L. Boulpaep. 2012, Saunders/Elsevier, Philadelphia, PA, USA. p. 106-146.

7. Chinn LW, Gow JM, Kroetz DL. Polymorphisms of drug transporters and clinical relevance, in Drug Transporters: Molecular Characterization and Role in Drug Disposition, Editors G. You and M.E. Morris. 2007, John Wiley & Sons, Inc., Hoboken, NJ, USA. p. 619-663.

8. Chu X, Bleasby K, Evers R. Species differences in drug transporters and implications for translating preclinical findings to humans. Expert Opin Drug Metab Toxicol 2013. 9(3):237-252.

9. Neve EP, Artursson P, Ingelman-Sundberg M, Karlgren M. An integrated in vitro model for simultaneous assessment of drug uptake, metabolism, and efflux. Mol Pharm 2013. 10(8):3152-3163.

10. European Medicine Agency (EMA), Committee for Human Medicinal Products (CHMP). Guideline on the investigation of drug interactions. 21 June 2012.

11. U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER). Guidance for industry, drug interaction studies—study design, data analysis, implications for dosing, and labeling recommendations. February 2012.

67

12. Poirier A, Funk C, Scherrmann JM, Lave T. Mechanistic modeling of hepatic transport from cells to whole body: application to napsagatran and fexofenadine. Mol Pharm 2009. 6(6):1716-1733.

13. Menochet K, Kenworthy KE, Houston JB, Galetin A. Use of mechanistic modeling to assess interindividual variability and interspecies differences in active uptake in human and rat hepatocytes. Drug Metab Dispos 2012. 40(9):1744-1756.

14. Pond SM, Tozer TN. First-pass elimination. Basic concepts and clinical consequences. Clin Pharmacokinet 1984. 9(1):1-25.

15. Kmiec Z. Cooperation of liver cells in health and disease. Adv Anat Embryol Cell Biol 2001. 161:III-XIII, 1-151.

16. Godoy P, Hewitt NJ, Albrecht U, Andersen ME, Ansari N, Bhattacharya S, et al. Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative hepatocyte sources and non-parenchymal liver cells and their use in investigating mechanisms of hepatotoxicity, cell signaling and ADME. Arch Toxicol 2013. 87(8):1315-1530.

17. Wisse E, Braet F, Luo D, De Zanger R, Jans D, Crabbe E, Vermoesen A. Structure and function of sinusoidal lining cells in the liver. Toxicol Pathol 1996. 24(1):100-111.

18. Hellerbrand C. Hepatic stellate cells--the pericytes in the liver. Pflugers Arch 2013. 465(6):775-778.

19. Klaassen CD, Watkins JB, 3rd. Mechanisms of bile formation, hepatic uptake, and biliary excretion. Pharmacol Rev 1984. 36(1):1-67.

20. Banales JM, Prieto J, Medina JF. Cholangiocyte anion exchange and biliary bicarbonate excretion. World J Gastroenterol 2006. 12(22):3496-3511.

21. Weibel ER, Staubli W, Gnagi HR, Hess FA. Correlated morphometric and biochemical studies on the liver cell. I. Morphometric model, stereologic methods, and normal morphometric data for rat liver. J Cell Biol 1969. 42(1):68-91.

22. Hammad S, Hoehme S, Friebel A, von Recklinghausen I, Othman A, Begher-Tibbe B, et al. Protocols for staining of bile canalicular and sinusoidal networks of human, mouse and pig livers, three-dimensional reconstruction and quantification of tissue microarchitecture by image processing and analysis. Arch Toxicol 2014. 88(5):1161-1183.

23. Bartles JR, Feracci HM, Stieger B, Hubbard AL. Biogenesis of the rat hepatocyte plasma membrane in vivo: comparison of the pathways taken by apical and basolateral proteins using subcellular fractionation. J Cell Biol 1987. 105(3):1241-1251.

24. Jungermann K, Katz N. Functional specialization of different hepatocyte populations. Physiol Rev 1989. 69(3):708-764.

25. LeCluyse EL, Witek RP, Andersen ME, Powers MJ. Organotypic liver culture models: meeting current challenges in toxicity testing. Crit Rev Toxicol 2012. 42(6):501-548.

68

26. Schmucker DL, Mooney JS, Jones AL. Stereological analysis of hepatic fine structure in the Fischer 344 rat. Influence of sublobular location and animal age. J Cell Biol 1978. 78(2):319-337.

27. Wienkers LC, Heath TG. Predicting in vivo drug interactions from in vitro drug discovery data. Nat Rev Drug Discov 2005. 4(10):825-833.

28. Williams JA, Hyland R, Jones BC, Smith DA, Hurst S, Goosen TC, Peterkin V, Koup JR, Ball SE. Drug-drug interactions for UDP-glucuronosyltransferase substrates: a pharmacokinetic explanation for typically observed low exposure (AUCi/AUC) ratios. Drug Metab Dispos 2004. 32(11):1201-1208.

29. Sugano K, Kansy M, Artursson P, Avdeef A, Bendels S, Di L, et al. Coexistence of passive and carrier-mediated processes in drug transport. Nat Rev Drug Discov 2010. 9(8):597-614.

30. Dobson PD, Kell DB. Carrier-mediated cellular uptake of pharmaceutical drugs: an exception or the rule? Nat Rev Drug Discov 2008. 7(3):205-220.

31. Kramer SD, Lombardi D, Primorac A, Thomae AV, Wunderli-Allenspach H. Lipid-bilayer permeation of drug-like compounds. Chem Biodivers 2009. 6(11):1900-1916.

32. Poirier A, Lave T, Portmann R, Brun ME, Senner F, Kansy M, Grimm HP, Funk C. Design, data analysis, and simulation of in vitro drug transport kinetic experiments using a mechanistic in vitro model. Drug Metab Dispos 2008. 36(12):2434-2444.

33. Thompson TE, Huang C. Composition and dynamics of lipids in biomembranes, in Membrane physiology, Editors T.E. Andreoli, J.F. Hoffman, D.D. Fanestil, and S.G. Schultz. 1987, Springer US, New York, NY, USA. p. 25-44.

34. Liu X, Testa B, Fahr A. Lipophilicity and its relationship with passive drug permeation. Pharm Res 2011. 28(5):962-977.

35. Gutknecht J, Walter A. Histamine, theophylline and tryptamine transport through lipid bilayer membranes. Biochim Biophys Acta 1981. 649(2):149-154.

36. Xiang TX, Anderson BD. The relationship between permeant size and permeability in lipid bilayer membranes. J Membr Biol 1994. 140(2):111-122.

37. You G, Morris ME. Overview of drug transporter families, in Drug Transporters: Molecular Characterization and Role in Drug Disposition, Editors G. You and M.E. Morris. 2007, John Wiley & Sons, Inc., Hoboken, NJ, USA. p. 1-10.

38. Michaelis L, Menten ML, Johnson KA, Goody RS. The original Michaelis constant: translation of the 1913 Michaelis-Menten paper. Biochemistry 2011. 50(39):8264-8269.

39. Michaelis L, Menten ML. Die Kinetik der Invertinwirkung. Biochemische Zeitschrift 1913. 49:333-369.

69

40. Berg JM, Tymoczko JL, Stryer L. Enzymes: Basic Concepts and Kinetics, in Biochemistry, 2012, W.H. Freeman & Co Ltd, New York, NY, USA. p. 227-260.

41. Weinman SA. Electrogenicity of Na(+)-coupled bile acid transporters. Yale J Biol Med 1997. 70(4):331-340.

42. Masuda M, Ichikawa Y, Shimono K, Shimizu M, Tanaka Y, Nara T, Miyauchi S. Electrophysiological characterization of human Na(+)/taurocholate cotransporting polypeptide (hNTCP) heterologously expressed in Xenopus laevis oocytes. Arch Biochem Biophys 2014. 562:115-121.

43. Ho RH, Tirona RG, Leake BF, Glaeser H, Lee W, Lemke CJ, Wang Y, Kim RB. Drug and bile acid transporters in rosuvastatin hepatic uptake: function, expression, and pharmacogenetics. Gastroenterology 2006. 130(6):1793-1806.

44. Bi YA, Qiu X, Rotter CJ, Kimoto E, Piotrowski M, Varma MV, Ei-Kattan AF, Lai Y. Quantitative assessment of the contribution of sodium-dependent taurocholate co-transporting polypeptide (NTCP) to the hepatic uptake of rosuvastatin, pitavastatin and fluvastatin. Biopharm Drug Dispos 2013. 34(8):452-461.

45. Vildhede A, Karlgren M, Svedberg EK, Wisniewski JR, Lai Y, Noren A, Artursson P. Hepatic uptake of atorvastatin: influence of variability in transporter expression on uptake clearance and drug-drug interactions. Drug Metab Dispos 2014. 42(7):1210-1218.

46. Hagenbuch B, Meier PJ. Organic anion transporting polypeptides of the OATP/ SLC21 family: phylogenetic classification as OATP/ SLCO superfamily, new nomenclature and molecular/functional properties. Pflugers Arch 2004. 447(5):653-665.

47. Li L, Meier PJ, Ballatori N. Oatp2 mediates bidirectional organic solute transport: a role for intracellular glutathione. Mol Pharmacol 2000. 58(2):335-340.

48. Satlin LM, Amin V, Wolkoff AW. Organic anion transporting polypeptide mediates organic anion/HCO3- exchange. J Biol Chem 1997. 272(42):26340-26345.

49. Hagenbuch B, Stieger B. The SLCO (former SLC21) superfamily of transporters. Mol Aspects Med 2013. 34(2-3):396-412.

50. Stieger B, Hagenbuch B. Organic anion-transporting polypeptides. Curr Top Membr 2014. 73:205-232.

51. Hartkoorn RC, Kwan WS, Shallcross V, Chaikan A, Liptrott N, Egan D, et al. HIV protease inhibitors are substrates for OATP1A2, OATP1B1 and OATP1B3 and lopinavir plasma concentrations are influenced by SLCO1B1 polymorphisms. Pharmacogenet Genomics 2010. 20(2):112-120.

52. Kalliokoski A, Niemi M. Impact of OATP transporters on pharmacokinetics. Br J Pharmacol 2009. 158(3):693-705.

70

53. Watanabe T, Kusuhara H, Sugiyama Y. Application of physiologically based pharmacokinetic modeling and clearance concept to drugs showing transporter-mediated distribution and clearance in humans. J Pharmacokinet Pharmacodyn 2010. 37(6):575-590.

54. Watanabe T, Kusuhara H, Maeda K, Kanamaru H, Saito Y, Hu Z, Sugiyama Y. Investigation of the rate-determining process in the hepatic elimination of HMG-CoA reductase inhibitors in rats and humans. Drug Metab Dispos 2010. 38(2):215-222.

55. Emami Riedmaier A, Nies AT, Schaeffeler E, Schwab M. Organic anion transporters and their implications in pharmacotherapy. Pharmacol Rev 2012. 64(3):421-449.

56. Kobayashi Y, Ohshiro N, Sakai R, Ohbayashi M, Kohyama N, Yamamoto T. Transport mechanism and substrate specificity of human organic anion transporter 2 (hOat2 [SLC22A7]). J Pharm Pharmacol 2005. 57(5):573-578.

57. Zhang L, Schaner ME, Giacomini KM. Functional characterization of an organic cation transporter (hOCT1) in a transiently transfected human cell line (HeLa). J Pharmacol Exp Ther 1998. 286(1):354-361.

58. Zhang S, Lovejoy KS, Shima JE, Lagpacan LL, Shu Y, Lapuk A, et al. Organic cation transporters are determinants of oxaliplatin cytotoxicity. Cancer Res 2006. 66(17):8847-8857.

59. Gupta S, Wulf G, Henjakovic M, Koepsell H, Burckhardt G, Hagos Y. Human organic cation transporter 1 is expressed in lymphoma cells and increases susceptibility to irinotecan and paclitaxel. J Pharmacol Exp Ther 2012. 341(1):16-23.

60. Jedlitschky G, Hoffmann U, Kroemer HK. Structure and function of the MRP2 (ABCC2) protein and its role in drug disposition. Expert Opin Drug Metab Toxicol 2006. 2(3):351-366.

61. Kartenbeck J, Leuschner U, Mayer R, Keppler D. Absence of the canalicular isoform of the MRP gene-encoded conjugate export pump from the hepatocytes in Dubin-Johnson syndrome. Hepatology 1996. 23(5):1061-1066.

62. Kim RB, Wandel C, Leake B, Cvetkovic M, Fromm MF, Dempsey PJ, et al. Interrelationship between substrates and inhibitors of human CYP3A and P-glycoprotein. Pharm Res 1999. 16(3):408-414.

63. Benet LZ. The drug transporter-metabolism alliance: uncovering and defining the interplay. Mol Pharm 2009. 6(6):1631-1643.

64. Gan LS, Moseley MA, Khosla B, Augustijns PF, Bradshaw TP, Hendren RW, Thakker DR. CYP3A-like cytochrome P450-mediated metabolism and polarized efflux of cyclosporin A in Caco-2 cells. Drug Metab Dispos 1996. 24(3):344-349.

65. Cummins CL, Jacobsen W, Benet LZ. Unmasking the dynamic interplay between intestinal P-glycoprotein and CYP3A4. J Pharmacol Exp Ther 2002. 300(3):1036-1045.

71

66. Hochman JH, Chiba M, Nishime J, Yamazaki M, Lin JH. Influence of P-glycoprotein on the transport and metabolism of indinavir in Caco-2 cells expressing cytochrome P-450 3A4. J Pharmacol Exp Ther 2000. 292(1):310-318.

67. Johnson BM, Charman WN, Porter CJ. The impact of P-glycoprotein efflux on enterocyte residence time and enterocyte-based metabolism of verapamil. J Pharm Pharmacol 2001. 53(12):1611-1619.

68. Badhan R, Penny J, Galetin A, Houston JB. Methodology for development of a physiological model incorporating CYP3A and P-glycoprotein for the prediction of intestinal drug absorption. J Pharm Sci 2009. 98(6):2180-2197.

69. Tam D, Sun H, Pang KS. Influence of P-glycoprotein, transfer clearances, and drug binding on intestinal metabolism in Caco-2 cell monolayers or membrane preparations: a theoretical analysis. Drug Metab Dispos 2003. 31(10):1214-1226.

70. Dufek MB, Knight BM, Bridges AS, Thakker DR. P-glycoprotein increases portal bioavailability of loperamide in mouse by reducing first-pass intestinal metabolism. Drug Metab Dispos 2013. 41(3):642-650.

71. Kage K, Tsukahara S, Sugiyama T, Asada S, Ishikawa E, Tsuruo T, Sugimoto Y. Dominant-negative inhibition of breast cancer resistance protein as drug efflux pump through the inhibition of S-S dependent homodimerization. Int J Cancer 2002. 97(5):626-630.

72. Doyle LA, Yang W, Abruzzo LV, Krogmann T, Gao Y, Rishi AK, Ross DD. A multidrug resistance transporter from human MCF-7 breast cancer cells. Proc Natl Acad Sci U S A 1998. 95(26):15665-15670.

73. Hirano M, Maeda K, Hayashi H, Kusuhara H, Sugiyama Y. Bile salt export pump (BSEP/ABCB11) can transport a nonbile acid substrate, pravastatin. J Pharmacol Exp Ther 2005. 314(2):876-882.

74. Noe J, Kullak-Ublick GA, Jochum W, Stieger B, Kerb R, Haberl M, Mullhaupt B, Meier PJ, Pauli-Magnus C. Impaired expression and function of the bile salt export pump due to three novel ABCB11 mutations in intrahepatic cholestasis. J Hepatol 2005. 43(3):536-543.

75. Pedersen JM, Matsson P, Bergstrom CA, Hoogstraate J, Noren A, LeCluyse EL, Artursson P. Early identification of clinically relevant drug interactions with the human bile salt export pump (BSEP/ABCB11). Toxicol Sci 2013. 136(2):328-343.

76. Kool M, van der Linden M, de Haas M, Scheffer GL, de Vree JM, Smith AJ, et al. MRP3, an organic anion transporter able to transport anti-cancer drugs. Proc Natl Acad Sci U S A 1999. 96(12):6914-6919.

77. Konig J, Rost D, Cui Y, Keppler D. Characterization of the human multidrug resistance protein isoform MRP3 localized to the basolateral hepatocyte membrane. Hepatology 1999. 29(4):1156-1163.

72

78. Scheffer GL, Kool M, de Haas M, de Vree JM, Pijnenborg AC, Bosman DK, et al. Tissue distribution and induction of human multidrug resistant protein 3. Lab Invest 2002. 82(2):193-201.

79. Denk GU, Soroka CJ, Takeyama Y, Chen WS, Schuetz JD, Boyer JL. Multidrug resistance-associated protein 4 is up-regulated in liver but down-regulated in kidney in obstructive cholestasis in the rat. J Hepatol 2004. 40(4):585-591.

80. Russel FG, Koenderink JB, Masereeuw R. Multidrug resistance protein 4 (MRP4/ABCC4): a versatile efflux transporter for drugs and signalling molecules. Trends Pharmacol Sci 2008. 29(4):200-207.

81. Hammann F, Drewe J. Data mining for potential adverse drug-drug interactions. Expert Opin Drug Metab Toxicol 2014. 10(5):665-671.

82. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis. N Engl J Med 2002. 346(7):539-540.

83. LeCluyse EL, Alexandre E. Isolation and culture of primary hepatocytes from resected human liver tissue. Methods Mol Biol 2010. 640:57-82.

84. LeCluyse EL. Human hepatocyte culture systems for the in vitro evaluation of cytochrome P450 expression and regulation. Eur J Pharm Sci 2001. 13(4):343-368.

85. LeCluyse EL, Audus KL, Hochman JH. Formation of extensive canalicular networks by rat hepatocytes cultured in collagen-sandwich configuration. Am J Physiol 1994. 266(6 Pt 1):C1764-1774.

86. Swift B, Pfeifer ND, Brouwer KL. Sandwich-cultured hepatocytes: an in vitro model to evaluate hepatobiliary transporter-based drug interactions and hepatotoxicity. Drug Metab Rev 2010. 42(3):446-471.

87. Dunn JC, Tompkins RG, Yarmush ML. Long-term in vitro function of adult hepatocytes in a collagen sandwich configuration. Biotechnol Prog 1991. 7(3):237-245.

88. Tchaparian EH, Houghton JS, Uyeda C, Grillo MP, Jin L. Effect of culture time on the basal expression levels of drug transporters in sandwich-cultured primary rat hepatocytes. Drug Metab Dispos 2011. 39(12):2387-2394.

89. Kotani N, Maeda K, Watanabe T, Hiramatsu M, Gong LK, Bi YA, Takezawa T, Kusuhara H, Sugiyama Y. Culture period-dependent changes in the uptake of transporter substrates in sandwich-cultured rat and human hepatocytes. Drug Metab Dispos 2011. 39(9):1503-1510.

90. Qiu X, Bi YA, Balogh LM, Lai Y. Absolute measurement of species differences in sodium taurocholate cotransporting polypeptide (NTCP/Ntcp) and its modulation in cultured hepatocytes. J Pharm Sci 2013. 102(9):3252-3263.

73

91. Kimoto E, Yoshida K, Balogh LM, Bi YA, Maeda K, El-Kattan A, Sugiyama Y, Lai Y. Characterization of organic anion transporting polypeptide (OATP) expression and its functional contribution to the uptake of substrates in human hepatocytes. Mol Pharm 2012. 9(12):3535-3542.

92. Yang Q, Doshi U, Li N, Li AP. Effects of culture duration on gene expression of P450 isoforms, uptake and efflux transporters in primary hepatocytes cultured in the absence and presence of interleukin-6: implications for experimental design for the evaluation of downregulatory effects of biotherapeutics. Curr Drug Metab 2012. 13(7):938-946.

93. Hoffmaster KA, Turncliff RZ, LeCluyse EL, Kim RB, Meier PJ, Brouwer KL. P-glycoprotein expression, localization, and function in sandwich-cultured primary rat and human hepatocytes: relevance to the hepatobiliary disposition of a model opioid peptide. Pharm Res 2004. 21(7):1294-1302.

94. Wilkening S, Stahl F, Bader A. Comparison of primary human hepatocytes and hepatoma cell line Hepg2 with regard to their biotransformation properties. Drug Metab Dispos 2003. 31(8):1035-1042.

95. Hilgendorf C, Ahlin G, Seithel A, Artursson P, Ungell AL, Karlsson J. Expression of thirty-six drug transporter genes in human intestine, liver, kidney, and organotypic cell lines. Drug Metab Dispos 2007. 35(8):1333-1340.

96. Rodriguez-Antona C, Donato MT, Boobis A, Edwards RJ, Watts PS, Castell JV, Gomez-Lechon MJ. Cytochrome P450 expression in human hepatocytes and hepatoma cell lines: molecular mechanisms that determine lower expression in cultured cells. Xenobiotica 2002. 32(6):505-520.

97. Gerets HH, Tilmant K, Gerin B, Chanteux H, Depelchin BO, Dhalluin S, Atienzar FA. Characterization of primary human hepatocytes, HepG2 cells, and HepaRG cells at the mRNA level and CYP activity in response to inducers and their predictivity for the detection of human hepatotoxins. Cell Biol Toxicol 2012. 28(2):69-87.

98. Aninat C, Piton A, Glaise D, Le Charpentier T, Langouet S, Morel F, Guguen-Guillouzo C, Guillouzo A. Expression of cytochromes P450, conjugating enzymes and nuclear receptors in human hepatoma HepaRG cells. Drug Metab Dispos 2006. 34(1):75-83.

99. Kanebratt KP, Andersson TB. Evaluation of HepaRG cells as an in vitro model for human drug metabolism studies. Drug Metab Dispos 2008. 36(7):1444-1452.

100. Guillouzo A, Corlu A, Aninat C, Glaise D, Morel F, Guguen-Guillouzo C. The human hepatoma HepaRG cells: a highly differentiated model for studies of liver metabolism and toxicity of xenobiotics. Chem Biol Interact 2007. 168(1):66-73.

74

101. Le Vee M, Noel G, Jouan E, Stieger B, Fardel O. Polarized expression of drug transporters in differentiated human hepatoma HepaRG cells. Toxicol In Vitro 2013. 27(6):1979-1986.

102. Hart SN, Li Y, Nakamoto K, Subileau EA, Steen D, Zhong XB. A comparison of whole genome gene expression profiles of HepaRG cells and HepG2 cells to primary human hepatocytes and human liver tissues. Drug Metab Dispos 2010. 38(6):988-994.

103. Konig J, Zolk O, Singer K, Hoffmann C, Fromm MF. Double-transfected MDCK cells expressing human OCT1/MATE1 or OCT2/MATE1: determinants of uptake and transcellular translocation of organic cations. Br J Pharmacol 2011. 163(3):546-555.

104. Ishiguro N, Maeda K, Saito A, Kishimoto W, Matsushima S, Ebner T, Roth W, Igarashi T, Sugiyama Y. Establishment of a set of double transfectants coexpressing organic anion transporting polypeptide 1B3 and hepatic efflux transporters for the characterization of the hepatobiliary transport of telmisartan acylglucuronide. Drug Metab Dispos 2008. 36(4):796-805.

105. Hirouchi M, Kusuhara H, Onuki R, Ogilvie BW, Parkinson A, Sugiyama Y. Construction of triple-transfected cells [organic anion-transporting polypeptide (OATP) 1B1/multidrug resistance-associated protein (MRP) 2/MRP3 and OATP1B1/MRP2/MRP4] for analysis of the sinusoidal function of MRP3 and MRP4. Drug Metab Dispos 2009. 37(10):2103-2111.

106. Ahlin G, Hilgendorf C, Karlsson J, Szigyarto CA, Uhlen M, Artursson P. Endogenous gene and protein expression of drug-transporting proteins in cell lines routinely used in drug discovery programs. Drug Metab Dispos 2009. 37(12):2275-2283.

107. Pedersen JM, Matsson P, Bergstrom CA, Norinder U, Hoogstraate J, Artursson P. Prediction and identification of drug interactions with the human ATP-binding cassette transporter multidrug-resistance associated protein 2 (MRP2; ABCC2). J Med Chem 2008. 51(11):3275-3287.

108. Ahlin G, Karlsson J, Pedersen JM, Gustavsson L, Larsson R, Matsson P, Norinder U, Bergstrom CA, Artursson P. Structural requirements for drug inhibition of the liver specific human organic cation transport protein 1. J Med Chem 2008. 51(19):5932-5942.

109. Chang C, Pang KS, Swaan PW, Ekins S. Comparative pharmacophore modeling of organic anion transporting polypeptides: a meta-analysis of rat Oatp1a1 and human OATP1B1. J Pharmacol Exp Ther 2005. 314(2):533-541.

110. Martinez-Becerra P, Briz O, Romero MR, Macias RI, Perez MJ, Sancho-Mateo C, Lostao MP, Fernandez-Abalos JM, Marin JJ. Further characterization of the electrogenicity and pH sensitivity of the human organic anion-transporting polypeptides OATP1B1 and OATP1B3. Mol Pharmacol 2011. 79(3):596-607.

75

111. Choi MK, Shin HJ, Choi YL, Deng JW, Shin JG, Song IS. Differential effect of genetic variants of Na(+)-taurocholate co-transporting polypeptide (NTCP) and organic anion-transporting polypeptide 1B1 (OATP1B1) on the uptake of HMG-CoA reductase inhibitors. Xenobiotica 2011. 41(1):24-34.

112. Broer S. Xenopus laevis Oocytes, in Membrane Transporters in Drug Discovery and Development, Editor Q. Yan. 2010, Humana Press, New York, NY, USA. p. 295-310.

113. Hagenbuch B, Scharschmidt BF, Meier PJ. Effect of antisense oligonucleotides on the expression of hepatocellular bile acid and organic anion uptake systems in Xenopus laevis oocytes. Biochem J 1996. 316 ( Pt 3):901-904.

114. Keppler D, Jedlitschky G, Leier I. Transport function and substrate specificity of multidrug resistance protein. Methods Enzymol 1998. 292:607-616.

115. Glavinas H, Mehn D, Jani M, Oosterhuis B, Heredi-Szabo K, Krajcsi P. Utilization of membrane vesicle preparations to study drug-ABC transporter interactions. Expert Opin Drug Metab Toxicol 2008. 4(6):721-732.

116. Tozer TN, Rowland M. Disposition following intravenous bolus, in Introduction to Pharmacokinetics and Pharmacodynamics, 2006, Lippincott Williams and Wilkins. p. 63-104.

117. Abe K, Bridges AS, Brouwer KL. Use of sandwich-cultured human hepatocytes to predict biliary clearance of angiotensin II receptor blockers and HMG-CoA reductase inhibitors. Drug Metab Dispos 2009. 37(3):447-452.

118. Umehara K, Camenisch G. Novel in vitro-in vivo extrapolation (IVIVE) method to predict hepatic organ clearance in rat. Pharm Res 2012. 29(2):603-617.

119. Bosgra S, van de Steeg E, Vlaming ML, Verhoeckx KC, Huisman MT, Verwei M, Wortelboer HM. Predicting carrier-mediated hepatic disposition of rosuvastatin in man by scaling from individual transfected cell-lines in vitro using absolute transporter protein quantification and PBPK modeling. Eur J Pharm Sci 2014. 65:156-166.

120. Hirano M, Maeda K, Shitara Y, Sugiyama Y. Contribution of OATP2 (OATP1B1) and OATP8 (OATP1B3) to the hepatic uptake of pitavastatin in humans. J Pharmacol Exp Ther 2004. 311(1):139-146.

121. Komai T, Shigehara E, Tokui T, Koga T, Ishigami M, Kuroiwa C, Horiuchi S. Carrier-mediated uptake of pravastatin by rat hepatocytes in primary culture. Biochem Pharmacol 1992. 43(4):667-670.

122. Ismair MG, Stieger B, Cattori V, Hagenbuch B, Fried M, Meier PJ, Kullak-Ublick GA. Hepatic uptake of cholecystokinin octapeptide by organic anion-transporting polypeptides OATP4 and OATP8 of rat and human liver. Gastroenterology 2001. 121(5):1185-1190.

76

123. Lancon A, Delmas D, Osman H, Thenot JP, Jannin B, Latruffe N. Human hepatic cell uptake of resveratrol: involvement of both passive diffusion and carrier-mediated process. Biochem Biophys Res Commun 2004. 316(4):1132-1137.

124. Bi YA, Kimoto E, Sevidal S, Jones HM, Barton HA, Kempshall S, et al. In vitro evaluation of hepatic transporter-mediated clinical drug-drug interactions: hepatocyte model optimization and retrospective investigation. Drug Metab Dispos 2012. 40(6):1085-1092.

125. Kilford PJ, Gertz M, Houston JB, Galetin A. Hepatocellular binding of drugs: correction for unbound fraction in hepatocyte incubations using microsomal binding or drug lipophilicity data. Drug Metab Dispos 2008. 36(7):1194-1197.

126. Winne D. Unstirred layer, source of biased Michaelis constant in membrane transport. Biochim Biophys Acta 1973. 298(1):27-31.

127. Tachibana T, Kitamura S, Kato M, Mitsui T, Shirasaka Y, Yamashita S, Sugiyama Y. Model analysis of the concentration-dependent permeability of P-gp substrates. Pharm Res 2010. 27(3):442-446.

128. Williamson B, Soars AC, Owen A, White P, Riley RJ, Soars MG. Dissecting the relative contribution of OATP1B1-mediated uptake of xenobiotics into human hepatocytes using siRNA. Xenobiotica 2013. 43(10):920-931.

129. Zamek-Gliszczynski MJ, Kalvass JC, Pollack GM, Brouwer KL. Relationship between drug/metabolite exposure and impairment of excretory transport function. Drug Metab Dispos 2009. 37(2):386-390.

130. Ito K, Chiba K, Horikawa M, Ishigami M, Mizuno N, Aoki J, et al. Which concentration of the inhibitor should be used to predict in vivo drug interactions from in vitro data? AAPS PharmSci 2002. 4(4):E25.

131. Ito K, Iwatsubo T, Kanamitsu S, Ueda K, Suzuki H, Sugiyama Y. Prediction of pharmacokinetic alterations caused by drug-drug interactions: metabolic interaction in the liver. Pharmacol Rev 1998. 50(3):387-412.

132. Watanabe T, Kusuhara H, Maeda K, Shitara Y, Sugiyama Y. Physiologically based pharmacokinetic modeling to predict transporter-mediated clearance and distribution of pravastatin in humans. J Pharmacol Exp Ther 2009. 328(2):652-662.

133. Sjogren E, Bredberg U, Lennernas H. The pharmacokinetics and hepatic disposition of repaglinide in pigs: mechanistic modeling of metabolism and transport. Mol Pharm 2012. 9(4):823-841.

134. Varma MV, Lai Y, Kimoto E, Goosen TC, El-Kattan AF, Kumar V. Mechanistic modeling to predict the transporter- and enzyme-mediated drug-drug interactions of repaglinide. Pharm Res 2013. 30(4):1188-1199.

77

135. Winkler DA. The role of quantitative structure--activity relationships (QSAR) in biomolecular discovery. Brief Bioinform 2002. 3(1):73-86.

136. Guha R. On exploring structure-activity relationships, in In silico Models for Drug Discovery, Editor S. Kortagere. 2013, Human Press, New York, NY, USA. p. 81-94.

137. Hayashi M, Matsumoto N, Takenoshita-Nakaya S, Takeba Y, Watanabe M, Kumai T, et al. Individual metabolic capacity evaluation of cytochrome P450 2C19 by protein and activity in the small intestinal mucosa of Japanese pancreatoduodenectomy patients. Biol Pharm Bull 2011. 34(1):71-76.

138. Sy SK, Ciaccia A, Li W, Roberts EA, Okey A, Kalow W, Tang BK. Modeling of human hepatic CYP3A4 enzyme kinetics, protein, and mRNA indicates deviation from log-normal distribution in CYP3A4 gene expression. Eur J Clin Pharmacol 2002. 58(5):357-365.

139. Ohtsuki S, Schaefer O, Kawakami H, Inoue T, Liehner S, Saito A, et al. Simultaneous absolute protein quantification of transporters, cytochromes P450, and UDP-glucuronosyltransferases as a novel approach for the characterization of individual human liver: comparison with mRNA levels and activities. Drug Metab Dispos 2012. 40(1):83-92.

140. Schwanhausser B, Busse D, Li N, Dittmar G, Schuchhardt J, Wolf J, Chen W, Selbach M. Global quantification of mammalian gene expression control. Nature 2011. 473(7347):337-342.

141. Ambros V. The functions of animal microRNAs. Nature 2004. 431(7006):350-355.

142. Peng M, Taouatas N, Cappadona S, van Breukelen B, Mohammed S, Scholten A, Heck AJ. Protease bias in absolute protein quantitation. Nat Methods 2012. 9(6):524-525.

143. Boja ES, Rodriguez H. Mass spectrometry-based targeted quantitative proteomics: achieving sensitive and reproducible detection of proteins. Proteomics 2012. 12(8):1093-1110.

144. Drozdzik M, Groer C, Penski J, Lapczuk J, Ostrowski M, Lai Y, et al. Protein abundance of clinically relevant multidrug transporters along the entire length of the human intestine. Mol Pharm 2014. 11(10):3547-3555.

145. Prasad B, Evers R, Gupta A, Hop CE, Salphati L, Shukla S, Ambudkar SV, Unadkat JD. Interindividual variability in hepatic organic anion-transporting polypeptides and P-glycoprotein (ABCB1) protein expression: quantification by liquid chromatography tandem mass spectroscopy and influence of genotype, age, and sex. Drug Metab Dispos 2014. 42(1):78-88.

146. Qiu X, Zhang H, Lai Y. Quantitative targeted proteomics for membrane transporter proteins: method and application. Aaps J 2014. 16(4):714-726.

78

147. Prasad B, Unadkat JD. Optimized approaches for quantification of drug transporters in tissues and cells by MRM proteomics. Aaps J 2014. 16(4):634-648.

148. Wong JW, Cagney G. An overview of label-free quantitation methods in proteomics by mass spectrometry, in Methods Mol Biol, Editors S.J. Hubbard and A.R. Jones. 2010, Humana Press, New York, NY, USA. p. 273-283.

149. Wisniewski JR, Ostasiewicz P, Dus K, Zielinska DF, Gnad F, Mann M. Extensive quantitative remodeling of the proteome between normal colon tissue and adenocarcinoma. Mol Syst Biol 2012. 8:611.

150. Wisniewski JR, Rakus D. Multi-enzyme digestion FASP and the 'Total Protein Approach'-based absolute quantification of the Escherichia coli proteome. J Proteomics 2014. 109:322-331.

151. Pernemalm M, Lewensohn R, Lehtio J. Affinity prefractionation for MS-based plasma proteomics. Proteomics 2009. 9(6):1420-1427.

152. Azimifar SB, Nagaraj N, Cox J, Mann M. Cell-type-resolved quantitative proteomics of murine liver. Cell Metab 2014. 20(6):1076-1087.

153. Niemi M. Transporter pharmacogenetics and statin toxicity. Clin Pharmacol Ther 2010. 87(1):130-133.

154. Nichols GA, Koro CE. Does statin therapy initiation increase the risk for myopathy? An observational study of 32,225 diabetic and nondiabetic patients. Clin Ther 2007. 29(8):1761-1770.

155. Joy TR, Hegele RA. Narrative review: statin-related myopathy. Ann Intern Med 2009. 150(12):858-868.

156. Thompson PD, Clarkson P, Karas RH. Statin-associated myopathy. Jama 2003. 289(13):1681-1690.

157. Link E, Parish S, Armitage J, Bowman L, Heath S, Matsuda F, Gut I, Lathrop M, Collins R. SLCO1B1 variants and statin-induced myopathy--a genomewide study. N Engl J Med 2008. 359(8):789-799.

158. Pasanen MK, Neuvonen M, Neuvonen PJ, Niemi M. SLCO1B1 polymorphism markedly affects the pharmacokinetics of simvastatin acid. Pharmacogenet Genomics 2006. 16(12):873-879.

159. Niemi M, Schaeffeler E, Lang T, Fromm MF, Neuvonen M, Kyrklund C, et al. High plasma pravastatin concentrations are associated with single nucleotide polymorphisms and haplotypes of organic anion transporting polypeptide-C (OATP-C, SLCO1B1). Pharmacogenetics 2004. 14(7):429-440.

160. Pasanen MK, Fredrikson H, Neuvonen PJ, Niemi M. Different effects of SLCO1B1 polymorphism on the pharmacokinetics of atorvastatin and rosuvastatin. Clin Pharmacol Ther 2007. 82(6):726-733.

79

161. Deng JW, Song IS, Shin HJ, Yeo CW, Cho DY, Shon JH, Shin JG. The effect of SLCO1B1*15 on the disposition of pravastatin and pitavastatin is substrate dependent: the contribution of transporting activity changes by SLCO1B1*15. Pharmacogenet Genomics 2008. 18(5):424-433.

162. Hedenmalm K, Alvan G, Ohagen P, Dahl ML. Muscle toxicity with statins. Pharmacoepidemiol Drug Saf 2010. 19(3):223-231.

163. Lennernas H. Clinical pharmacokinetics of atorvastatin. Clin Pharmacokinet 2003. 42(13):1141-1160.

164. Elsby R, Hilgendorf C, Fenner K. Understanding the critical disposition pathways of statins to assess drug-drug interaction risk during drug development: it's not just about OATP1B1. Clin Pharmacol Ther 2012. 92(5):584-598.

165. Shitara Y, Maeda K, Ikejiri K, Yoshida K, Horie T, Sugiyama Y. Clinical significance of organic anion transporting polypeptides (OATPs) in drug disposition: their roles in hepatic clearance and intestinal absorption. Biopharm Drug Dispos 2013. 34(1):45-78.

166. Kunze A, Huwyler J, Camenisch G, Poller B. Prediction of organic anion-transporting polypeptide 1B1- and 1B3-mediated hepatic uptake of statins based on transporter protein expression and activity data. Drug Metab Dispos 2014. 42(9):1514-1521.

167. Li J, Volpe DA, Wang Y, Zhang W, Bode C, Owen A, Hidalgo IJ. Use of transporter knockdown Caco-2 cells to investigate the in vitro efflux of statin drugs. Drug Metab Dispos 2011. 39(7):1196-1202.

168. Ellis LC, Hawksworth GM, Weaver RJ. ATP-dependent transport of statins by human and rat MRP2/Mrp2. Toxicol Appl Pharmacol 2013. 269(2):187-194.

169. Keskitalo JE, Zolk O, Fromm MF, Kurkinen KJ, Neuvonen PJ, Niemi M. ABCG2 polymorphism markedly affects the pharmacokinetics of atorvastatin and rosuvastatin. Clin Pharmacol Ther 2009. 86(2):197-203.

170. Keskitalo JE, Kurkinen KJ, Neuvoneni PJ, Niemi M. ABCB1 haplotypes differentially affect the pharmacokinetics of the acid and lactone forms of simvastatin and atorvastatin. Clin Pharmacol Ther 2008. 84(4):457-461.

171. Saito Y. Pitavastatin: an overview. Atheroscler Suppl 2011. 12(3):271-276.

172. Shirasaka Y, Suzuki K, Shichiri M, Nakanishi T, Tamai I. Intestinal absorption of HMG-CoA reductase inhibitor pitavastatin mediated by organic anion transporting polypeptide and P-glycoprotein/multidrug resistance 1. Drug Metab Pharmacokinet 2011. 26(2):171-179.

173. Fujino H, Saito T, Ogawa S, Kojima J. Transporter-mediated influx and efflux mechanisms of pitavastatin, a new inhibitor of HMG-CoA reductase. J Pharm Pharmacol 2005. 57(10):1305-1311.

80

174. Chung JY, Cho JY, Yu KS, Kim JR, Oh DS, Jung HR, et al. Effect of OATP1B1 (SLCO1B1) variant alleles on the pharmacokinetics of pitavastatin in healthy volunteers. Clin Pharmacol Ther 2005. 78(4):342-350.

175. Ieiri I, Suwannakul S, Maeda K, Uchimaru H, Hashimoto K, Kimura M, et al. SLCO1B1 (OATP1B1, an uptake transporter) and ABCG2 (BCRP, an efflux transporter) variant alleles and pharmacokinetics of pitavastatin in healthy volunteers. Clin Pharmacol Ther 2007. 82(5):541-547.

176. Oh ES, Kim CO, Cho SK, Park MS, Chung JY. Impact of ABCC2, ABCG2 and SLCO1B1 polymorphisms on the pharmacokinetics of pitavastatin in humans. Drug Metab Pharmacokinet 2013. 28(3):196-202.

177. Fujino H, Yamada I, Shimada S, Yoneda M, Kojima J. Metabolic fate of pitavastatin, a new inhibitor of HMG-CoA reductase: human UDP-glucuronosyltransferase enzymes involved in lactonization. Xenobiotica 2003. 33(1):27-41.

178. Hirano M, Maeda K, Matsushima S, Nozaki Y, Kusuhara H, Sugiyama Y. Involvement of BCRP (ABCG2) in the biliary excretion of pitavastatin. Mol Pharmacol 2005. 68(3):800-807.

179. Zhou Q, Chen QX, Ruan ZR, Yuan H, Xu HM, Zeng S. CYP2C9*3(1075A > C), ABCB1 and SLCO1B1 genetic polymorphisms and gender are determinants of inter-subject variability in pitavastatin pharmacokinetics. Pharmazie 2013. 68(3):187-194.

180. Nath A, Zientek MA, Burke BJ, Jiang Y, Atkins WM. Quantifying and predicting the promiscuity and isoform specificity of small-molecule cytochrome P450 inhibitors. Drug Metab Dispos 2010. 38(12):2195-2203.

181. Wisniewski JR, Zougman A, Nagaraj N, Mann M. Universal sample preparation method for proteome analysis. Nat Methods 2009. 6(5):359-362.

182. Karlgren M, Vildhede A, Norinder U, Wisniewski JR, Kimoto E, Lai Y, Haglund U, Artursson P. Classification of inhibitors of hepatic organic anion transporting polypeptides (OATPs): influence of protein expression on drug-drug interactions. J Med Chem 2012. 55(10):4740-4763.

183. Davies B, Morris T. Physiological parameters in laboratory animals and humans. Pharm Res 1993. 10(7):1093-1095.

184. Menochet K, Kenworthy KE, Houston JB, Galetin A. Simultaneous assessment of uptake and metabolism in rat hepatocytes: a comprehensive mechanistic model. J Pharmacol Exp Ther 2012. 341(1):2-15.

81

185. Noe J, Portmann R, Brun ME, Funk C. Substrate-dependent drug-drug interactions between gemfibrozil, fluvastatin and other organic anion-transporting peptide (OATP) substrates on OATP1B1, OATP2B1, and OATP1B3. Drug Metab Dispos 2007. 35(8):1308-1314.

186. Shirasaka Y, Mori T, Murata Y, Nakanishi T, Tamai I. Substrate- and dose-dependent drug interactions with grapefruit juice caused by multiple binding sites on OATP2B1. Pharm Res 2014. 31(8):2035-2043.

187. Soars MG, Barton P, Ismair M, Jupp R, Riley RJ. The development, characterization, and application of an OATP1B1 inhibition assay in drug discovery. Drug Metab Dispos 2012. 40(8):1641-1648.

188. De Bruyn T, van Westen GJ, Ijzerman AP, Stieger B, de Witte P, Augustijns PF, Annaert PP. Structure-based identification of OATP1B1/3 inhibitors. Mol Pharmacol 2013. 83(6):1257-1267.

189. Achour B, Rostami-Hodjegan A, Barber J. Protein expression of various hepatic uridine 5'-diphosphate glucuronosyltransferase (UGT) enzymes and their inter-correlations: a meta-analysis. Biopharm Drug Dispos 2014. 35(6):353-361.

190. Achour B, Barber J, Rostami-Hodjegan A. Expression of hepatic drug-metabolizing cytochrome p450 enzymes and their intercorrelations: a meta-analysis. Drug Metab Dispos 2014. 42(8):1349-1356.

191. Hashimoto H, Nakagawa T, Yokoi T, Sawada M, Itoh S, Kamataki T. Fetus-specific CYP3A7 and adult-specific CYP3A4 expressed in Chinese hamster CHL cells have similar capacity to activate carcinogenic mycotoxins. Cancer Res 1995. 55(4):787-791.

192. Johansson M, Strahm E, Rane A, Ekstrom L. CYP2C8 and CYP2C9 mRNA expression profile in the human fetus. Front Genet 2014. 5:58.

193. Shitara Y, Li AP, Kato Y, Lu C, Ito K, Itoh T, Sugiyama Y. Function of uptake transporters for taurocholate and estradiol 17beta-D-glucuronide in cryopreserved human hepatocytes. Drug Metab Pharmacokinet 2003. 18(1):33-41.

194. Elaut G, Henkens T, Papeleu P, Snykers S, Vinken M, Vanhaecke T, Rogiers V. Molecular mechanisms underlying the dedifferentiation process of isolated hepatocytes and their cultures. Curr Drug Metab 2006. 7(6):629-660.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Pharmacy 193

Editor: The Dean of the Faculty of Pharmacy

A doctoral dissertation from the Faculty of Pharmacy, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofPharmacy. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Pharmacy”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-241376

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2015