In Renal Organic Cation Secretion Bethzaida

Multidrug And Toxin Extrusion's (MATE) Role in Renal Organic Cation Secretion

Item Type text; Electronic Dissertation

Authors Astorga, Bethzaida

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 01/10/2021 03:57:28

Link to Item http://hdl.handle.net/10150/205470 1

The Role of Multidrug And Toxin Extrusion (MATE) in Renal Organic Cation Secretion

Bethzaida Astorga

A Dissertation Submitted to the Faculty of the

PHYSIOLOGICAL SCIENCES

GRADUATE INTERDISCIPLINARY PROGRAM

In Partial Fulfillment of the Requirements for the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2011

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Dissertation Committee, we certify that we have read the dissertation prepared by: Bethzaida Astorga entitled: The Role of Multidrug And Toxin Extrusion (MATE) in Renal Organic Cation Secretion and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor of Philosophy

______Date: 11/29/11 Stephen H. Wright

______Date: 11/29/11 Nathan J. Cherrington

______Date: 11/29/11 William H. Dantzler

______Date: 11/29/11 Nicholas A. Delamere

______Date: 11/29/11 Alexander Simon

Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

______Date: 11/29/11 Dissertation Director: Stephen H. Wright

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED: Bethzaida Astorga

ACKNOWLEDGEMENTS

To all of the faculty members that made the last 7 years in the Physiological Science Program interesting, ‘in a good way’. To the members of my committee, Dr. Nathan Cherrington, Dr. William Dantzler, Dr. Nick Delamere, Dr. Alex Simon and Dr. Stephen Wright. A special thanks to Dr. Dantzler, who always took time to be a second mentor to me. Dr. Scott Boitano for always making it to all the graduate student activities and for facilitating so many connections in the field. Dr. Cindy Rankin for always encouraging me to continue on and for helping me decide to continue my education with the Physiological Sciences Graduate Program. Also, Dr. Heddwen Brooks for introducing me to many people in the world of renal physiology and for being a role model for ‘women in science’. Dr. Tom Pennabaker for teaching me immunohistochemistry and for always being willing to help me on the microscope.

To all my fellow graduate students and colleagues that made my time in graduate school so special. Thank you to Dr. Ryan Pelis who always made things in the lab fun and ‘doable’; and for his continued mentorship. Dr. Yod Dangpraprai who was there to commiserate with me and to always point out the positive (and for listening to my political rants). Mark Morales, who deserves at least 50% of this degree for helping me with so many of my experiments and for listening to me complain for hours on end when things were not working. Theresa Wunz, who took the role of lab manager to a whole new level with her ‘no-nonsense’ attitude. She taught me almost everything I know about cell culture, keeping an orderly lab notebook and most importantly how to be the “squeaky wheel”. Xiaohong Zhang made everything about molecular biology look and sound easy, and while I will never be as good as she is at it, I certainly appreciate all the time she spent explaining her tricks to me. Thank you to Kristen Evans for being a good friend and a great lab technician. And finally a very special thank you to (soon to be Dr.) Laura Wright and Lori Cole who I met in this program and who are now my lifelong friends.

To those I met while in Los Angeles at the University of Southern California, thank you so much for your continued friendship and for a great year in the Programs in Biomedical & Biological Sciences (PIBBS). Thank you to Dr. Alicia McDonough and Dr. Alan Yu for kindly taking me into your labs and for your continued support. Thank you to University of Arizona (Bear Down!) for making me proud to be a Wildcat! And to the Physiological Sciences Graduate Program for taking me back after my ‘sabbatical’ at the University of Southern California. Especially, Holly Lopez for ‘taking care of business’.

Finally to my family, Elizabeth Astorga, Dad (Octavio Astorga) and Grandmother (Manuela Orozco) for being patient and for not asking me “So, how long before you graduate?”, anymore. I could not be more grateful to my Mother, Evadina Lomerson, whose prayers “for me to finally finish!” always brightened up my day (especially while writing this dissertation). To my Grandfather, Manuel Orozco, whose own sacrifices have always been a source of motivation, pride, strength and ambition. Also, to my incredible husband, Kevin Yarbrough, who has not only listened and consoled me but also encouraged and been proud of me. His support has been invaluable.

And last but certainly not least, Dr. Stephen Wright, who has guided me in this long and at times frustrating journey. I’ve learned things about science, life, academia, obscure Sci-fi novels and movies, the 60’s and 70’s and membrane protein kinetics that I am certain I could not have learned anywhere else. In the summer of 2004 I walked into his lab thinking that “I would give research a try”, 7 years later I am leaving with great memories, friendships and with the certainty that it would not have been this fun and unique anywhere else. Thanks Steve.

TABLE OF CONTENTS

LIST OF FIGURES ...... 9

LIST OF TABLES ...... 11

ABSTRACT ...... 12

CHAPTER 1: INTRODUCTION ...... 14

1.1 Background ...... 14

1.2 Organic Cations ...... 14

1.3 Cellular Physiology of Renal OC Secretion ...... 16

1.4 Basolateral OC Transport – the OCTs ...... 18

1.4.1 Molecular Characteristics of OCT1 and OCT2 ...... 18

1.5 Apical OC Transport – the MATEs ...... 21

1.5.1 Molecular Characteristics of MATEs ...... 23

1.5.2 MATE Protein Structure & Function ...... 25

1.6 Kinetics & Selectivity of MATE transporters...... 27

1.7 Role of MATEs in Renal OC Secretion - Specific Aims ...... 30

1.8 Predictive Models: Use of Pharmacophores ...... 31

1.9 Competitive Exchange Diffusion ...... 35

ORGANIZATION OF CHAPTERS ...... 36

CHAPTER 2: MATERIALS AND METHODS ...... 37

2.1 Chemicals ...... 37

2.2 Cell Culture and Stable Expression of hMATE1 & hMATE2-K ...... 37

2.3 Transport by hMATE1- & hMATE2-K Expressing CHO cells ...... 38

2.4 Real-Time Quantitative-PCR ...... 38

2.5 Immunohistochemistry ...... 39 6

TABLE OF CONTENTS – Continued

2.6 Computational Modeling ...... 40

2.6.1 3D-QSAR Development Using the Hypogen Method in Discovery

Studio ...... 40

2.6.2 Quantitative Model Update With Variable Weights and Tolerances...... 41

2.7 Statistics ...... 41

CHAPTER 3: MOLECULAR CHARACTERIZATION & QUANTITATIVE STRUCTURE ACTIVITY RELATIONSHIPS FOR INHIBITION OF HUMAN MATE1 & MATE2-K ...... 42

3.1 Background ...... 42

3.2 The kinetics of MPP and H+ Interaction With hMATE1 & hMATE2-K ...... 45

3.3 Inhibitory Selectivity of hMATE1 and hMATE2-K: Test Set Selection ...... 49

3.4 Expression of MATE1 and MATE2/2-K in Human Kidney ...... 51

3.5 Modeling of MATE Selectivity ...... 52

3.5.1 Computational Analysis of hMATE1 Inhibition ...... 53

3.5.2 Initial Round: hMATE1 Common Features Pharmacophore...... 54

3.5.3 hMATE1 Common Features Pharmacophore Testing ...... 55

3.5.4 First Iteration: Quantitative Pharmacophore Development for

hMATE1 ...... 56

3.5.5 Second Iteration: Quantitative Pharmacophore Development for

hMATE1 ...... 56

3.5.6 Final Iteration: Quantitative Pharmacophore Development for

hMATE1 ...... 57

3.6 Discussion & Conclusions ...... 57

CHAPTER 4: ASSESSMENT OF THE COMPLEX BINDING SURFACE OF HUMAN MATE1 ...... 60

TABLE OF CONTENTS – Continued

4.1 Background ...... 60

4.2 hMATE1-Mediated Transport Kinetics for MPP & Metformin ...... 62

4.3 Kinetic Basis of Metformin’s Inhibition of hMATE1-Mediated [3H]MPP

Transport...…...... 62

4.4 Kinetic Basis of Creatinine Inhibition of hMATE1-Mediated [3H]MPP

Transport ...... 63

4.5 Kinetic Basis of PYR Inhibition of hMATE1-Mediated [3H]MPP Transport ...... 63

4.6 Using CED to Address Cis- and Trans- Stimulation of hMATE1-Mediated

[3H]MPP Transport ...... 65

4.7 Significance of PYR Structural Analogs: PYR-2 & PYR3…………………...... 68

4.8 Conclusions & Discussion……………………………………………………...... 70

CHAPTER 5: APPLICATION OF THE COMPETITIVE EXCHANGE DIFFUSION PROTOCOL TO DETERMINE SUBSTRATES AND INHIBITORS OF HUMAN MATE2-K ...... 76

5.1 Background ...... 76

5.2 Using CED to Assess Various Compounds as MATE2-K Substrates or

Inhibitors ...... 78

5.3 Known MATE Substrates ...... 79

5.3.1 N-Tetraalkylammonium Compounds as MATE Substrates……………79

5.3.2 Unknown MATE Substrates ...... 80

5.4 Discussion ...... 82

CHAPTER 6: CONCLUSIONS & FUTURE DIRECTIONS ...... 83

6.1 Conclusions ...... 83

6.2 Future Directions ...... 84

TABLE OF CONTENTS – Continued

APPENDIX A: FIGURES ...... 86

APPENDIX B: TABLES ...... 117

REFERENCES ...... 123

LIST OF FIGURES

Figure 1 – Cellular Model of Organic Cation Secretion ...... 86

Figure 2 – hMATE1 Secondary Structure ...... 87

Figure 3 – hMATE1 Homology Model ...... 88

Figure 4 – hMATE1 & hMATE2-K MPP Kinetics & H+ Inhibition Kinetic ...... 89

Figure 5 – Range of Inhibition of hMATE1 & hMATE2-K Mediated Transport ...... 90

Figure 6 – Inhibitory Profiles of Transport Activity of hMATE1 and hMATE2-K ...... 91

Figure 7 – Comparison of hMATE1 & hMATE2-K IC50 Values & Range of Affinities ...... 92

Figure 8 – mRNA Levels of Xenobiotic Transporters in Human Kidneys ...... 93

Figure 9 – Immunohistochemical Localization of hMATE1 & hMATE2-K in the Kidney ...... 94

Figure 10 – Correlation Between hMATE1 & hMATE2-K IC50 Values & LogP, TPSA & pKa

Values ...... 95

Figure 11 – Family of Kinetic Curves & hMATE1 Common Features Pharmacophore ...... 96

Figure 12 – hMATE1 Inhibition Kinetics of Cinchonidine and Ethohexadiol & Common

Features Pharamcophores ...... 97

Figure 13 – First Iteration of the Quantitative Pharmacophore for hMATE1 & Correlation

Graph ...... 98

Figure 14 – Second Iteration of the Quantitative Pharmacophore for hMATE1 & Correlation

Graph ...... 99

Figure 15 – Final Iteration of the Quantitative Pharmacophore for hMATE1 & Correlation

Graph ...... 100

Figure 16 – hMATE1 Quantitative Pharmacophore Generated With Values From

Kido et al (73) ...... 101

LIST OF FIGURES - Continued

Figure 17 – hMATE1 Kinetics of MPP & Metformin ...... 102

Figure 18 – hMATE1 Metformin Inhibition Kinetics of MPP & MPP Kinetics in the Absence

& Presence of 250 µM Metformin ...... 103

Figure 19 – hMATE1 Creatinine Inhibition Kinetics of MPP & MPP Kinetics in the Absence

& Presence of 1 mM Creatinine ...... 104

Figure 20 – PYR Inhibition Kinetics of MPP & MPP Kinetics in the Absence & Presence of 1

µM PYR ...... 105

Figure 21 – hMATE1 MPP Uptake Time Course, Efflux Time Course & MPP Concentration-

Dependent Trans-Stimulation (C)...... 106

Figure 22 – PYR Concentration-Dependent Trans-Stimulation ...... 107

Figure 23 – [3H]MPP Time Course in the Presence of MPP or PYR ...... 108

Figure 24 – PYR-2 & PYR-3 Inhibition Kinetics of MPP ...... 109

Figure 25 – Time Dependence of Inhibition & Time Dependence of Inhibition Normalized to

Total Tracer Uptake ...... 110

Figure 26 –MPP Kinetics in the Absence & Presence of 200 nM PYR-2 ...... 111

Figure 27 – Model for Competitive Inhibition & Linear Mixed-Type Inhibition ...... 112

Figure 28 – Competitive Exchange Diffusion With hMATE2-K Controls ...... 113

Figure 29 – Competitive Exchange Diffusion With Known MATE Substrates ...... 114

Figure 30 – Competitive Exchange Diffusion With N-Tetraalkylammonium Series ...... 115

Figure 31 – Competitive Exchange Diffusion With Unknown MATE Substrates ...... 116

LIST OF TABLES

Table 1A – hMATE1 Kinetic Profiles ...... 117

Table 1B – hMATE2-K Kinetic Profiles ...... 118

Table 1C – hMATE1 Inhibition Kinetics ...... 119

Table 1D – hMATE2-K Inhibition Kinetics ...... 120

Table 2 – IC50 Values Generated for hMATE1 & hMATE2-K ...... 121

Table 3 - Jmax & Kt values for Chapter 4 Kinetics ...... 122

ABSTRACT

Organic cations (OCs) make up ~40% of all prescribed drugs and renal secretion plays a major role in clearing these (and other OCs), from the plasma. The active and rate-limiting step of renal

OC secretion is mediated by luminal OC/H+ exchange, the molecular basis of which is suspected to involve two homologous transport proteins, Multidrug And Toxin Extruders 1 & 2-K (MATE1 and MATE2-K). This study has two aims to resolve outstanding issues dealing with the mechanism of MATE-mediated OC transport: (Aim 1) develop predictive models of ligand interaction with hMATE1; (Aim 2) establish the kinetic mechanism(s) of ligand interaction with

MATE transporters and the extent to which inhibitory ligands serve as transported substrates of

MATE transporters. Transport was measured using human MATE1 and MATE2-K stably expressed in Chinese Hamster Ovary cells. Both MATEs had similar affinities for the prototypic

OC substrate, 1-methyl-4-phenylpyridinium (MPP), and had overlapping selectivity for most of the test inhibitors. The IC50 values for 59 structurally diverse inhibitory ligands were used to generate a common features (HIPHOP) pharmacophore and three quantitative pharmacophores for hMATE1 (each displaying a significant correlation between predicted and measured IC50 values). The models identified (i) structural features that influence ligand interaction with hMATE1, including hydrophobic regions, H-bond donor and acceptor sites and an ionizable feature; and (ii) novel high affinity inhibitors of MATE-mediated transport from 13 new drug classes. Whereas metformin and creatinine were shown to be competitive inhibitors of MPP, the inhibition of MATE1-mediated MPP transport produced by pyrimethamine (PYR) and related analogs was not competitive but, instead, had a “linear, mixed-type” inhibitory profile suggestive of a MATE binding surface rather than a singular binding site. “Competitive exchange diffusion” showed that selected inhibitory ligands (including quinidine, caffeine, and the organic anion,

PAH) also serve as transported substrates for MATE1. In conclusion, these data are consistent

with the presence of a MATE binding surface with multiple, non-overlapping binding sites that can display different kinetic interactions with structurally distinct substrates. The creation of hMATE1 pharmacophores offers insight into development and interpretation of predictive models of drug-drug interaction in the kidney.

CHAPTER 1

INTRODUCTION

1.1 Background

Humans are exposed to a vast array of toxins on a daily basis. Nevertheless, we survive because the body is equipped with mechanisms to clear it of these dangerous threats. Exposure to toxins can occur in one of two ways: (1) endogenous molecules can be broken down to form toxic metabolites, or (2) humans can ingest toxins in their diets, or from pharmaceutical compounds, or other environmental exposures. In some cases toxins taken in from the outside world never cross the mucosa of the intestine and, therefore, are not a threat. However, toxins that are absorbed into the bloodstream must be cleared from the body in order to minimize any potential danger. The liver and kidney are highly specialized to protect humans from systemic perturbations resulting from the intake of toxins. The liver metabolizes and excretes compounds in the bile whereas the kidney filters, and/or secretes toxins into the urine. Transport proteins exist in the liver and kidney that move toxins through the epithelial cells of these organs, for subsequent secretion. These proteins have evolved strategies that focus on key chemical characteristics, the most general of which appears to be charge; there are ‘organic anion’ transporters and ‘organic cation’ transporters. This project focuses on organic cation (OC) transport (the interested reader is referred to several recent reviews on organic anion transport

(13; 49; 52; 132), with an emphasis on the cellular and molecular strategies found in the kidney; and a particular emphasis on a key family of transporters (Multidrug And Toxin Extruders;

‘MATE’) recently implicated in this process.

1.2 Organic Cations

OCs are endogenous or exogenous compounds that carry a net positive charge at physiological pH. Endogenous OCs include selected neurotransmitters and hormones such as

dopamine, epinephrine and norepinephrine, and metabolic precursor molecules such as choline

(used to make the neurotransmitter, acetylcholine) and phospholipids that comprise cellular membranes. Exogenous OCs comprise a much more complex group of molecules which include plant by-products and drugs; collectively, ‘xenobiotics.’ OCs have been categorized into two classes based on their structural and physicochemical properties: Type I and Type II (93). Type I

OCs are small (MW<400 Da), comparatively hydrophilic, and usually monovalent, such as tetraethylammonium (TEA) and 1-methyl-4-phenylpyridinium (MPP). Conversely, Type II OCs are large (MW >500 Da), frequently polyvalent and comparatively hydrophobic (e.g., vercuronium and hexafluorenium). The majority of prescribed medications that are cationic at physiological pH fall into the Type I category, and it is the renal handling of these compounds that is the focus of my work.

Approximately 40% of all prescribed medications are OCs (3). Therefore, determining how the kidney and liver handle and interact with OCs is crucial in learning how to protect against toxic threats and also to prevent unwanted drug-drug interactions (DDIs). DDIs occur when the availability, effectiveness or activity of a drug is altered by the presence of another drug and can lead to: (1) decreased excretion or secretion of a drug; (2) decreased metabolism of a drug; (3) high levels of a drug in the plasma; (4) decreased bioavailability of a drug; and (5) toxicity from high levels in the plasma or within cells. For example, cimetidine (drug used for treatment of peptic ulcers) reduces the renal clearance of procainamide (anti-arrhythmia drug), ranitidine (anti-peptic ulcer drug), triamterene (diuretic), metformin (anti-diabetic drug), flecainide (anti-arrhythmia drug) and the active metabolite N-acetylprocainamide. It also decreases the level of plasma concentrations of ketoconazole (antifungal), indomethacin (non- steroidal anti-inflammatory drug, NSAID) and chlorpromazine (antipsychotic) by reducing their absorption in the GI tract (128). These findings show that cimetidine not only has DDIs with

several drugs but also that it may compete with other drugs at the absorption site (GI tract) and/or excretion sites (kidney or liver). Another interesting DDI example involves the zwitterion, fexofenadine (FEX), which is taken into the RPT cells by Organic Anion Transporter 3 (OAT3) and secreted into the tubular lumen via MATE1 (91). Uptake of FEX is markedly inhibited on the basolateral membrane of the RPT by 10 µM probenecid, whereas on the apical membrane

FEX secretion is decreased by 60% with exposure to 3 µM cimetidine (91). Both OAT3 and

MATE1 have been implicated as “having a role in clinical DDIs” by the International Transport

Consortium (ITC) and, therefore, these findings are of great clinical relevance (42). Furthermore, this is an interesting example from a physiological perspective, in that drugs can have two potential sites for DDIs in the renal proximal tubule (RPT).

1.3 Cellular Physiology of Renal OC Secretion

As previously mentioned, clearance and/or secretion of OCs occurs in the liver or kidney.

This work focuses on OC secretion in the kidney and more specifically on the role of this process in RPT cells of the kidney. The liver expresses some of the same OC transporters as the kidney and for a review of the postulated role of these transport proteins in hepatic OC secretion the interested reader is referred to the following reviews (16; 37; 62; 102). Renal OC secretion was discovered in the late 1940s when Rennick et al. showed that subjects injected with 200 mg of the prototypic OC, TEA, rapidly cleared the majority of it via excretion in the urine in the first 15 minutes, with over 99% of it gone in 2 hours (118). Such high rates of excretion suggested that glomerular filtration was not the sole process responsible and, therefore, Rennick hypothesized that there must be a secretory process involved (118). Subsequently, studies with stop-flow and micropuncture techniques showed that the OCs choline (1; 2; 154), N1-methylnicotinamide

(NMN; (55)), mecamylamine, mepiperphenidol (110), TEA (117) and dihydromorphine (59) were secreted primarily by the mammalian RPT.

The cellular model of OC secretion in the RPT was developed by Holohan and Ross in the early 1980s. Using cortical membrane vesicles they showed that there were two separate steps required for OC secretion in the RPT: a basolateral (peritubular) entry step and an apical

(luminal) efflux step (56; 57). Figure 1 provides an updated cellular model of OC secretion in the

RPT (revised from the American Physiological Society’s Comprehensive Physiology: Renal

Handbook (121)). The first step in this process begins with OCs coming in from the blood across the basolateral membrane, via facilitated diffusion, driven by the inside negative membrane potential of the cell. OCs can also come into the cell in exchange for intracellular OCs. Once inside the cell OCs can exit the RPT cells across the luminal membrane into the tubular filtrate in exchange for a proton (56; 57). Protons are freely available in the filtrate. However, the presence of an apical Na/H Exchanger (NHE3; not shown in Figure 1) can affect OC/H+ exchange by maintaining the inwardly directed electrochemical potential for H+.

It is now widely agreed that the peritubular entry step in the human kidney, via electrogenic facilitated diffusion, involves Organic Cation Transporter 2 (OCT2; (100)). The molecular basis of the luminal step, in which OCs are exchanged for a proton and subsequently released into the lumen of the tubule, has been more elusive. But in recent years there has been a growing consensus that the luminal step (the active and rate-limiting step in OC secretion in the

RPT (123)) is mediated by two members of the Multidrug And Toxin Extrusion (MATE) family:

MATE1 (SLC47A1) and MATE2-K (SLC47A2). This study focuses on the selectivity and physiological mechanisms involved with the influence of MATE transporters on secretion of

Type I OCs in the RPT. But before detailing the current status of MATE transporters and their role in renal OC secretion, I will briefly describe the current understanding of the OCTs.

1.4 Basolateral OC Transport – the OCTs

In the RPT cells of the human kidney, OCT2 (SLC22A2) is expressed on the basolateral membrane. OCT1 was thought to be expressed in the basolateral membrane of the human RPT, until recent evidence showed that it is expressed, albeit at comparatively low levels, in the apical membrane (144). In the liver, OCT1 is the major organic cation transporter on the sinusoidal membranes of canilicular cells (95). The level of expression of each of these proteins in the kidney (and elsewhere) is very different. Based on protein and mRNA, OCT2 is the predominant

OCT protein expressed in the human kidney, with hOCT1 making up less than 1% of that of hOCT2 (100). Importantly, there are marked species differences in the expression of OCTs, and in rat, mouse and rabbit kidneys both Oct1 and Oct2 play a functional role in OC secretion (160).

Oct1 (-/-) knockout mice and Oct2 (-/-) knockout mice showed that OC secretion remains largely intact when one transporter is eliminated; it was not until a double knockout animal (Oct1 (-/-) and Oct2 (-/-)) was made that TEA secretion was abolished, thereby showing that (i) Oct homologs work in concert to secrete OCs in the mouse (ii) and these two processes (in the mouse) are completely responsible for basolateral TEA secretion (64).

1.4.1 Molecular Characteristics of OCT1 and OCT2

Rat Oct1 was cloned in 1994 by Gründemann, D et al and was the first of the OCT family, not only to be cloned, but also to be expressed in a heterologous expression system (46).

They named the 556 amino acid protein organic cation transport 1 (rOct1; Slc22a1) and showed that when expressed in Xenopus oocytes, it had kinetic profiles for prototypical OCs (e.g., NMN) that were similar to those observed in microperfused RPTs and membrane vesicles from rat kidney (46; 145). Subsequently, others cloned homologs and orthologs of rOct1 in the human, rabbit, and other species. In 1997, Koehler, MR et al (74) found the gene locations for hOCT1 and hOCT2 on chromosome 6 and then Gorboulev V et al (44) successfully cloned and

characterized their basic function in Xenopus oocytes. These findings led to significant advances in the field of basolateral OC uptake in the RPT cells. Below is a brief summary of key characteristics of the OCTs:

i. OCTs are part of the Major Facilitator Superfamily (MFS) of transport proteins. They

have 550-560 amino acid residues with 12 transmembrane spanning domains (TMDs).

ii. OCTs support electrogenic OC entry across the basolateral membrane or electroneutral

exchange of an OC for an OC (14). Changes in membrane potential affect OCT transport,

+ + influencing both Kt and Jmax (14); however, changes in Na and H gradients exert no

effect on OCT transport (14).

iii. The intracellular and extracellular binding regions of the OCTs can have markedly

different affinities for substrates and inhibitors. Experiments with rOct2 have shown that

the non-transported inhibitors, tetrabutylammonium (TBuA) and corticosterone, have a

differential affinity for either the extracellular or intracellular binding surface

(respectively; (43; 111; 155)).

iv. Human OCT1 and OCT2 (and OCT3) are regulated by the calcium-calmodulin (CaM)

pathway, such that inhibition of this pathway results in decreased affinity of TEA in

hOCT1 and hOCT2-expressing human embryonic kidney (HEK) cells (15; 21; 88).

v. The promoter for the gene encoding hOCT1 has two DNA response elements for the

hepatocyte nuclear factor, HNF-4α (122). Activation of hOCT1 transcription by HNF-4α

is inhibited by the bile acid, chenodeoxycholic acid (122).

vi. Renal cortical slices taken from male rats show higher rates of TEA uptake than those of

female rats (12; 151). The presence of testosterone increases expression of rOct2 in the

female rat kidney, whereas expression of rOct2 decreases in the male rat kidney in the

presence of estradiol (152). There are two androgen response elements in the promoter

region of the rOct2 gene (152). vii. Changes in the expression of the OCTs have been observed in pathological conditions.

In 5/6-nephrectomized rats (a model used to simulate renal failure) clearance of

cimetidine is decreased in conjunction with the expression of rOct2 with no changes in

the expression of rOct1, organic anion transporters (Oat) 1 and 3 or Na+/K+-ATPase (63).

Amongst other disease states, diabetes and hyperuricemia are also correlated with

decreases in the expression of the OCTs (47; 48; 139). viii. Single nucleotide polymorphisms (SNPs) of hOCT2 and hOCT1 not only influence

transport in heterologous expression systems have also have been linked to changes in

metformin pharmacokinetics in humans (18; 20; 72; 124; 156).

ix. Crystal structures for E. coli’s lactose permease (LacY) and glycerol-3-phosphate (GlpT)

have given significant insight into the tertiary structures of members of the MFS. The

tertiary structures of these proteins were used to create in silico homology models for

rOct1 and hOCT2 ((111; 179) amongst others) and they confirmed previous findings

about amino acid placement and accessibility to the binding regions of OCTs. Briefly,

seven amino acids implicated in being in the putative binding region of rOct1 were

corroborated by the homology model of rOct1; as well as the accessibility of selected

cysteine residues in the aqueous cleft region, which contains the putative translocation

pathways of OCT1 (130) and OCT2 (109).

x. In agreement with studies done using native renal preparations, researchers have shown

that OCTs are polyspecific, meaning they can interact with a structurally diverse array of

substrates and inhibitors. Studies that probed the multiselectivity of the OCTs showed

that inhibitors span multiple pharmacological classes. Consequently, rather than rely on

general, structural class criteria when categorizing these compounds, efforts to establish

the specific molecular determinants of substrate/inhibitor interaction with OCTs have

used so-called molecular ‘descriptors,’ such as lipophilicty, size, charge, flexibility etc.,

as a means to quantify the basis of transporter-ligand binding. Computational analysis of

the molecular determinants of substrate/inhibitor interaction with OCT1 (3; 10) and

OCT2 (131; 181) has been used to generate ‘pharmacophores’, which are in silico models

that describe the molecular features of a drug or ligand, required for recognition and/or

interaction with a biological macromolecule. The OCT pharmacophores have identified

several features necessary (or at least important) for the binding interaction between the

transporters and their ligands, including issues such as placement of hydrophobic mass

relative to a basic nitrogen-containing moiety, in conjunction with the unique shape and

size of the ligand.

In short, studies focusing on the secondary and tertiary structures of the OCTs, their

multispecificity, short and long-term regulation, genetic variations and the molecular

determinants necessary for a compound to interact with the OCTs, have contributed to a

growing body of information about OC secretion. Moreover, these studies have generated

ideas for studying the apical step in this process and how basolateral and apical transport

processes might work together to achieve transcellular secretion of OCs.

1.5 Apical OC Transport – the MATEs

Many studies using native tissues (i.e. isolated renal tubules, tissues, or membrane vesicles) were able to provide insight on the functional characteristics of the OC/H+ exchanger long before there was a clear idea of the identity of the protein. A brief outline of the physiological features of the OC/H+ exchanger developed from these studies is outlined below.

i. OC/H+ exchange is present in both the apical membrane RPT cells in the kidney and in

the canalicular membranes of heptocytes in the liver(107). ii. In renal brush border membrane vesicles (BBMV), extravesicular H+competively inhibits

OC uptake (cis-inhibition) and intravesicular H+ trans-stimulates OC uptake by

increasing the turnover rate (Jmax; (163)). Together these findings supported Holohan and

Ross’ initial theories about the OC/H+ Exchanger. iii. The OC/H+ exchanger is an obligatory exchanger with a 1:1 stoichiometry for its

substrates, which consequently makes each transport cycle electroneutral (162). iv. Whereas the basolateral rate of OC uptake along the length of an isolated rabbit RPT is

relatively constant, the apical efflux step is greatest in early proximal tubule and

decreases along its length (98; 123). Furthermore, the maximal rate of OC transport is

greater in BBMV taken from the outer renal cortex than in those taken from the inner

medulla (while the affinity of both populations of vesicles for OC is similar) (97). These

findings support the contention that the apical step in OC secretion in the RPT is the rate-

limiting as well as active step (159). v. The apical OC/H+ exchanger interacts with a multitude of OCs and its multispecificity is

comparable to that of OCT2 (146).

This‘physiological fingerprint’ of OC/H+ exchange, which includes (i) expression in the kidney (and liver); (ii) localization in the apical membrane of RPT cells (and hepatocytes); (iii) support of OC/H+ exchange; and (iv) broad selectivity for cationic substrates, established the conditions that had to be met by any transport process(es) suggested to mediate apical OC transport. Three proteins have been set forth because they fulfilled some of the ‘fingerprint’ requirements. The 'novel organic cation transporters' (OCTNs) 1 and 2 were the first proposed as

OC/H+ exchanger candidates, because of their ability to transport TEA (albeit, with low affinity)

and their expression in the apical membrane of RPTs (134; 168; 171). Furthermore, Xenopus oocytes injected with mRNA for OCTN1 showed a transient increase in TEA uptake in the presence of an initial outwardly directed proton gradient (168; 171). However, OCTN1, though found in many tissues, is not expressed in the liver (134), and more recent studies showed that

OCTN1 supports a preferential transport of ergothioneine that is stimulated by extracellular sodium, a functional characteristic completely absent in renal (and hepatic) OC/H+ exchange.

Thus, OCTN1 fails to meet two of the hallmark features of the OC/H+ exchanger. Likewise,

OCTN2 showed some characteristics of the OC/H+ exchanger (expression in the apical membrane of RPT cells; modest transport of TEA). However, upon further inspection it was discovered that it is expressed ubiquitously throughout the body and that it too is a Na+-dependent (40; 77; 133;

169) cotransporter, showing a marked selectivity for the zwitterion, carnitine. Consequently, both

OCTN1 and OCTN2 have been dismissed as candidates for the apical OC/H+ exchanger.

The other major candidate protein suggested as being the elusive OC/H+ Exchanger was

Multidrug Resistance Transporter 1 (MDR1) or P-glycoprotein (P-gP). However, its proposed role in renal OC secretion largely reflected its expression in the apical membrane of RPTs. MDR1 selectivity is strikingly different from that of OC/H+ exchange; it does not interact with most prototypical Type I OCs (e.g., TEA) and it is currently thought that its main role is in secreting

Type II OCs (Figure 1; (159)).

1.5.1 Molecular Characteristics of MATEs

In 2005, Otsuka et al. proposed that the MATE proteins could be the long sought OC/H+

Exchanger (107). MATEs belong to a family that includes over 800 xenobiotic/sodium or xenobiotic/proton exchangers expressed in Archaea, Eubacteria and Eukarya (principally plants and fungi), (106; 107). Otsuka et al., used Vibrio parahaemolyticus’ NorM, a Na+/multidrug antiporter (the first described member of the MATE family; (107)), as a reference sequence for

a BLAST search and found mRNA in human liver, kidney, skeletal muscle (and weakly in the brain and heart), which they named hMATE1 (SLC47A1; (107)). They tested the functional characteristics of this novel MATE protein and found that HEK cells transfected with hMATE1 supported H+-sensitive electroneutral OC exchange (107). In addition to the discovery of

MATE1, a homologous protein, MATE2 (SLC47A2), was found in the kidney (89). Following this discovery, Masuda et al, reported a kidney-specific splice variant of MATE2 they called

MATE2-K ('K' referring to the fact that it was only found in the kidney) and a brain-specific isoform, MATE2-B (‘B’ for brain; (89)). MATE2-K shares 94% sequence homology with hMATE2 (lacking 108 base pairs in exon 7, resulting in the elimination of 36 amino acids in the putative cytoplasmic loop between helices 2 and 3; (89)). Functional MATE2-K homologs have been found in the rabbit kidney (177), and sequence analysis showed the existence of MATE2-K splice variants in other species as well (81). Rats and mice lack a kidney specific MATE2, and mice express only Mate1 in the kidney. Mouse Mate2 is restricted to the testis and does not appear to be an ortholog of rbMate2 or hMATE1 (175). In fact there is phylogenetic evidence that suggests that “mMate2” should actually be referred to as mMate3 (175).

The MATE1/2-K orthologs that have been cloned in the human, rabbit, rat and mouse all display the ‘physiological fingerprint’ of the OC/H+ Exchanger. The following is a brief outline of the general characteristics of MATE1 (and MATE2-K).

i. Antibodies specific for hMATE1 showed that it is expressed in the kidney and liver

(107); particularly in the apical membranes of RPT cells and in the canilicular

membranes of hepatocytes. Immunohistochemical studies have also shown that

hMATE2 and hMATE2-K are localized in the RPT (75; 107).

ii. Terada et al, cloned rat MATE1 and found that it shares 79% homology with hMATE1

and that it shares the same functional characteristics as hMATE1 (138). Evidence in this

study suggested that TEA uptake was decreased as a result of intracellular alkalization,

whereas using nigericin to acidify the intracellular compartments of the cell stimulated

TEA uptake (138). Experiments involving rMATE1 and H+-gradients revealed that an

outwardly directed proton gradient ([H]inside > [H]outside) achieved through

intracellular acidification, stimulated extracellular TEA uptake (141). Lastly, membrane

vesicles from MATE1 expressing cells support concentrative TEA uptake in the presence

of an outwardly-directly proton-gradient, providing critical proof that MATEs support

OC/H+ exchange (141).

iii. Other studies have examined hMATE1’s and hMATE2-K's functional characteristics and

found that HEK cells expressing hMATE1 or hMATE2-K also exhibited the H+ gradient–

dependent antiport of TEA, MPP, cimetidine, metformin, creatinine, guanidine,

procainamide, and topotecan (89), supporting the ‘multiselectivity’ of MATE

transporters.

iv. A recent study from our lab of hMATE1-mediated MPP transport used an efflux protocol

to show that the kinetic interaction of H+ with the extracellular and cytoplasmic faces of

the exchanger is symmetrical (22).

v. A mMate1 knockout mouse created in 2010 showed that metformin clearance was

decreased by 85% in comparison to the wild-type mouse, thus solidifying the significance

of the MATEs in OC secretion in the RPT (140).

1.5.2 MATE Protein Structure & Function

The data accumulated since the 2005 discovery of the MATE proteins in mammals has not only shown that they match the ‘physiological fingerprint’ of OC/H+ exchange but also that without them secretion of OC is severely decreased. The field now widely accepts that, in the human kidney, some combination of MATE1 and MATE2/2-K plays a quantitatively significant

role in secretion of Type I OCs. However, despite this evidence comparatively little work has been directed toward understanding the molecular basis of MATE interaction with a broad suite of clinically important compounds. Such information requires knowledge of the structure of

MATE transporters and the kinetic basis of their interaction with substrates and inhibitors.

Following is a brief summary of the current limited understanding of MATE protein structure- function relationships.

i. Members of the MATE family typically have 12 TMDs; however mammalian MATEs

appear, by hydropathy analysis, to have 13 TMDs (Figure 2). This reflects the presence

at the C-terminal end of mammalian MATEs of ~20 hydrophobic amino acids not found

in typical MATE proteins. Zhang X et al provided experimental evidence, using epitope

tagging, site-directed mutagenesis and a thiol-reactive reagent, that rbMate1 has 13

TMDs (179). They confirmed that, as predicted, the N-terminus is in the cytoplasm, but

also showed that the C-terminus of the full length protein is exposed to the extracellular

face of the membrane. Moreover, truncation of the C-terminal hydrophobic sequence

results in a protein with a cytoplasmic C-terminus that retains transport function. Further

studies by Zhang, X et al have shown the same for human and mouse MATE1

(unpublished observation).

ii. An x-ray crystal structure of a bacterial MATE family member, NorM, was reported in

2010 (53). The NorM crystal structure showed 12 TMDs and was used as a template for

the development of a 3-dimensional homology model of the mammalian MATE1 (Figure

3; (121)).

iii. Thus far, two conserved cysteine residues (C63, C127) of MATE1 homologs have been

found (refer to Figure 2 orange residues; Figure 3 yellow residues) to influence transport.

C63 and C127 site-directed mutants (of human and rat MATE1) were created and shown

to support TEA transport, albeit at a markedly reduced rate (in comparison to the wild-

type (6)).

iv. Histidine residues were also examined and while there are 7 conserved across species,

only one appears to be essential, His-385 ((6); refer to Figure 2, blue residues; Figure 3,

purple residues). Experiments with the histidine reactive reagent, diethylpyrocarbonate

(DEPC), showed that TEA transport was inhibited in the wild-type hMATE1 (6).

DEPC’s inhibition of TEA transport could not be reversed in the presence of saturating

concentrations of TEA, which led to the hypothesis that histidines are not within the

substrate-binding site(s) but rather are proton detectors, therefore monitoring the proton-

driving force (and allowing for the exchange of OC; (6)).

v. In 2007, Matsumoto et al studied the role of glutamate residues in substrate recognition

by MATE1 and found that four glutamate residues played a critical role in TEA uptake,

E273, E389, E278 and E300 (Figure 2, green residues; Figure 3, red residues (90)). The

authors came to the latter conclusion by making glutamate-to-aspartic acid mutants (thus

retaining glutamate's negative charge but moving the location of the carboxyl group)

(90). Matsumoto el al, also made glutamate-to-alanine mutants and showed that E278A,

E300A and E389A, had no uptake, 48% and 16% of WT uptake, respectively (90). They

showed that glutamate mutants (both aspartic acid and alanine) had different pH profiles,

therefore suggesting that glutamate residues play a critical role in substrate and proton

binding (90).

1.6 Kinetics & Selectivity of MATE Transporters

The data outlined above show that MATEs play a crucial role in the secretion of OCs in the kidney. Understanding the apical OC exit-step is critical because OCs entering the cell via OCT2

may be trapped without an OC transporter on the apical membrane of RPT cells, resulting in cellular accumulation of a compound that could be nephrotoxic. An example of this was outlined in a recent study that showed that cisplatin, a platinum based cancer therapy drug, accumulates in the rat kidney causing nephrotoxicity (173), whereas, oxaliplatin, another related platinum based cancer therapy drug does not show any nephrotoxic effects (173). Interestingly, both platinum based drugs are transported by OCT2 but only oxaliplatin interacts with MATE1 and MATE2, suggesting differential selectivities based on substrate structure (173). This is a phenomenon that has many implications, not only for cisplatin, but for other DDIs and the impact of differing specificities between OCT2 and MATEs.

Few studies have examined either the kinetics or selectivity of substrate interaction with human MATE1, and even fewer have looked at hMATE2-K. Furthermore, of the studies that have focused on human MATE-mediated transport, few have generated kinetic parameter values, such as Jmax, Kt or IC50 (the concentration of inhibitor that blocks half of the mediated transport, which the field typically uses to determine a test ligand’s affinity for the processes with which it interacts). In Tables 1 A-B I compiled all kinetic data for transport, to date, for hMATE1- and hMATE2-K-mediated OC transport (respectively), listing the values for Jmax (pmol/min/mg protein) and Kt (µM). Tables 1C and 1D list the IC50 values that have been determined for inhibitor interaction with hMATE1 and hMATE2-K. These data prompt a discussion of issues relevant to the observations I report in this dissertation. i. First is simply how few kinetic measurements have been made. In fact, kinetic

parameters for transport (Tables 1A and 1B) have only been reported for 11 ‘organic

cations,’ and, with the exception of TEA and MPP, all those represent single reports.

Similarly, the kinetics of inhibition of human MATE transporters (Tables 1C and 1D)

have only been determined for a few compounds, and many of the studies that generated

these data focused on structurally restricted classes of compounds (e.g., the quinolone or

cephalosporin antibiotics). This paucity of data makes it difficult (as noted below) to

draw the broad comparisons required to interpret MATE transport activity within a

physiological, or mechanistic, context. ii. The second issue involves the several different conditions used in the experiments that

generated these data. For example, whereas some measurements were conducted at 37ºC,

others used room temperature, thereby making it difficult to compare rates of transport.

Of greater significance, however, is the range of pH conditions used in different studies.

Hydrogen ion is a substrate for the MATE transporters, and both the apparent affinity and

the rate of transport are sensitive to the external and internal H+ concentration and to the

direction of H+ gradients across the membrane (22; 164). Studies of MATE-mediated

transport (uptake) frequently maximize the rate of transport by acidifying the cytoplasm,

typically by using the standard ‘ammonia pulse’ method (e.g., (136)). Others maximize

transport by decreasing extracellular H+ concentration (thereby reducing the cis-inhibition

produced by competitive interaction of H+ with the transported test substrate; (22)). Both

approaches ‘suffer’ from the fact that neither reflects the normal physiological condition.

However, it must be noted that the normal physiological mode of MATE transport

activity is the mediation of OC efflux; all the kinetic values reported in Tables 1A and 1B

reflect the results of measurements of substrate uptake. The point to emphasize here is

that Kt values (and IC50 values reported in Tables 1C and 1D for hMATE1 and hMATE2-

K, respectively) are particularly sensitive to the extracellular pH; apparent affinities

measured at elevated pH will be higher (i.e., Kt/IC50 values will be lower), compared to

those measured at physiological pH (i.e., ~7.4). I will discuss this issue again when I

outline the experimental conditions I used, and interpretation of the results gained, in my

studies.

1.7 Role of MATEs in Renal OC Secretion - Specific Aims

There is, currently, a minimal understanding of the structure of mammalian MATE proteins and, therefore, no basis to predict the extent to which putative substrates or inhibitors will interact. Owing to the growing consensus of the critical role played by MATE proteins in mediating renal secretion of cationic drugs, there is a substantial need to understand the molecular basis of ligand interaction with MATE transporters. The goals of this project are to establish the selectivities of hMATE1 and hMATE2-K as critical elements for developing predictive models for ligand interactions. The removal of potentially toxic OCs is critical to maintaining overall health and therefore an understanding of the processes that allow the kidneys to do so will enhance the physiological understanding of DDIs and the development of nephrotoxicity. In this dissertation I address two specific aims regarding the cellular and molecular physiology of human

MATE transporters:

1. Develop predictive models of ligand interactions for hMATE1 and hMATE2-K,

to facilitate future modeling of the pharmacokinetics of drug secretion and

prediction of unwanted DDIs.

2. Establish the kinetic mechanism(s) of ligand interaction with MATE

transporters; and the extent to which inhibitory ligands also serve as

transported substrates of MATE transporters.

Pharamcophore development will be used to address Specific Aim 1. A brief history of pharmacophores and their applications is outlined in the following section (1.8). A modified

version of the competitive exchange diffusion (CED) method will be used to address aspects of

Specific Aim 2. An outline of the CED method and its applications is presented in section 1.9.

1.8 Predictive Models: Use of Pharmacophores

The development of a pharmacophore model of ligand interactions requires compiling a large data set of numbers that have a sufficient degree of ‘internal consistency’ to justify quantitative correlations/comparison to be made. Therefore, the values in the literature cannot be used; not only because they lack internal consistency but also because so few values exist. It is just such analyses that will be required to develop predictive models of ligand interaction of the type that can be derived from generation of pharmacophores, as described below. Researchers have long tried to find correlations between molecular descriptors such as charge, size, hydrophobicity etc., and biological activity with a given protein or enzyme. Some studies have been successful in finding such correlations. In fact in a study done with microperfused intact tubules Ulrich K et al, observed a strong correlation between ‘biological activity’ (Ki) and pKa values for a set of OCs (145). Unfortunately, this approach has not always been useful or reliable for complex experimental systems that express multiple transporters or for proteins that interact with structurally diverse group of compounds. In Ullrich’s work, for example, we now know that the results he reported involved interactions with at least two basolateral OC transporters (OCT1 and OCT2); and the marked correlation between Ki and pKa was probably influenced by the structurally restricted range of compounds included in the study. The renal secretory pathway for

OCs interacts with compounds that have a multitude of relevant molecular descriptors and therefore an approach that focuses on single descriptors is likely to be undependable. Recently, a method that utilizes the multitude of molecular descriptors available, in conjunction with empirically generated affinity values, has been created. Computer-driven computational methods seek correlations between a functional parameter (e.g., Km, Ki, or IC50 values) and a vast array of

'steric and electronic features' to generate an in silico predictive model, a ‘pharmacophore,’ for interaction of ligands with a given protein, receptor, enzyme or anything else with a binding site or surface. The International Union of Pure and Applied Chemistry (IUPAC) defines a pharmacophore as “an ensemble of steric and electronic features that is necessary to ensure the optimal supramolecular interactions with a specific biological target and to trigger (or block) its biological response” (158). Pharmacophores are now used to screen large drug databases and predict interactions and identify novel substrates or inhibitors. Below, I have highlighted some relevant pharmacophore findings for proteins and transcriptional factors involved in the handling of xenobiotics and drugs in the human, as an entree to the type of approach my own work has used. i. Cytochrome P-450 enzymes have been popular targets for pharmacophore analysis. One

of the first dates back to 1998, when Ekins S et al, used 3-Dimensional Quantitative

Structure Activity Relationships (3D-QSAR) to study the cytochrome P-459 2B6 enzyme

(CYP2B6; (29)). In this study they found for the first time robust and reliable predictions

about the molecular basis of ligand interaction with CYP2B6 (29). They determined that

the size, positive electrostatic potential, hydrogen bonding acceptor capacity and

hydrophobicity of putative ligands were the most relevant descriptors of effective

interaction with 2B6 (29). ii. The same group also studied cytochrome P-450 3A4 (CYP3A4) and the relationship

between a suite of molecular descriptors and experimentally observed Ki values for a

‘training set’ of test inhibitors. The resulting quantitative pharmacophore (28) produced

calculated Ki values with a correlation (r value) of 0.91, providing a powerful validation

of the method (28).

iii. Using the CYP3A4 pharmacophore Ekins S et al also identified, for the first time, novel

hydrophobic compounds that interacted with CYP3A4‘s active sites to stimulate its

metabolic sites, thus stimulating metabolism (30). iv. Pharmacophores have also been created for transcriptional factors such as Pregnane X

Receptor (PXR), which regulates expression of xenobiotic metabolism and transporter

genes. PXR was theorized to be stimulated by azoles binding to its activation site and

therefore a series of imadazoles, steroids and other experimental compounds were used to

generate a pharmacophore (31). The pharmacophore generated for PXR was used to

identify a novel steroidal agonist binding site and found several PXR agonist and

antagonist binding sites (31). Indeed, it is the ability of a pharmacophore to interrogate

external databases of structural information and identify novel ligands that has proven to

be one of the great advantages of this analytical approach to understanding ‘selectivity.’

v. Many projects have focused on predicting interaction with the Multidrug Resistance

transporter, MDR1 (or ‘P-gp’, for P glycoprotein), due to its promiscuous interactions

with many clinically relevant compounds. However, finding correlations or methods to

predict interactions was difficult. Using data available in the literature, in addition to new

in vitro data, computational methods predicted that verapamil, vinblastine and digoxin

have overlapping binding sites on P-gp (34). Verification of this prediction used four

different modeling approaches and three different cell lines (e.g. epithelial colorectal

adenocarcinoma (CaCo-2) cells and Lilly Laboratories Cell-Porcine Kidney (LLC-PK1)

to show that the original findings were valid and not an artifact of the specific cell lines

and/or methods used (33), an observation that supports the utility of heterologous

expression systems in the study of ligand-protein integration.

vi. Zolk and collegues, recently developed a pharmacophore for human OCT2 using 25 high

affinity inhibitors of [3H] MPP (181). They did show a significant correlation between a

single descriptor, topological polar surface area (TPSA), and IC50 values (r=0.71,

p<0.0001), suggesting a relationship between the hydrophobicity of a compound and its

affinity for hOCT2 (181). More importantly, the multiple descriptor-based hOCT2

pharmacophore they developed led to the discovery of two key structural features

(separated by 5 Å): (1) an ionizable feature and (2) a hydrophobic aromatic site (181).

In summary, pharmacophores have a host of different applications and are powerful tools

for identifying novel ligands and predicting interactions with binding sites/surfaces. More

relevant to this study, however, is determining which molecular descriptors are important for

interactions with transport proteins that mediate OC secretion.

1.9 Competitive Exchange Diffusion

To address the Second Aim of this dissertation (establish the kinetic mechanism(s) of

ligand interaction with MATE transporters; and the extent to which ligands serve as transported

substrates of MATE transporters), I elected to use the ‘competitive exchange diffusion’ (CED)

protocol outlined by Lacko and Burger in the 1960’s (78). CED can be used to indirectly

determine what compounds are substrates and which compounds are simply inhibitors of the

MATEs using the following steps:

i. Incubate cells with a substrate of choice (i.e. [3H]MPP; which can be called

“X*”) known for interaction with the membrane protein of interest, until steady-

state accumulation is achieved (i.e. intracellular capacity is at its maximum).

ii. Remove the labeled compound (X*) and add solution with an equal

concentration of compound X* and an unlabeled test compound (Y; i.e. another

OC, such as metformin) and allow the new buffer to incubate.

iii. Finally, measure how much compound X* is left inside the cells to determine if

compound Y trans-stimulates efflux of intracellular compound X*; a decrease in

X* content indicates that it was exchanged for compound Y, which consequently

shows that compound Y is a substrate. If, however the cellular content of

compound X remains the same as that of control (buffer with compound X*

only) then it can be assumed that compound Y either does not interact with the

membrane protein or that its interaction with the transporter is restricted to

inhibition.

Ideally, one determines which compounds are substrates versus inhibitors by using a radiolabeled form of a compound to directly measure transport, but unfortunately comparatively few OCs are available in their radiolabeled form and some of the ones that are, are extremely costly; thus the CED method allows for the measurement of trans-stimulation such that a compound can be confirmed as either ‘transported’ or ‘not transported’.

ORGANIZATION OF UPCOMING CHAPTERS

Chapter 2 describes materials and methods used to complete this dissertation. Chapters 3-5 are data chapters each containing background, results and discussion sections. Chapters 3 focuses on pharmacophore development and addresses Specific Aim 1. It also presents the work included in the manuscript Molecular Characterization and Quantitative Structure Activity Relationships for

Inhibition of human MATE1 and MATE-2K, by Bethzaida Astorga, Sean Ekins, Mark Morales and Stephen H Wright (in submission). Chapter 4 describes my assessment of the complex binding surface of hMATE1 and addresses portions of Specific Aim 2. Chapter 5 presents experiments done with hMATE2-K and the CED technique, describing a method to test if inhibitors of MATE can also serve as transported substrates. This chapter also addresses portions of Specific Aim 2. The last Chapter (6) provides an integrated discussion of conclusions and future directions given the findings presented in this dissertation.

CHAPTER 2

MATERIALS AND METHODS

2.1 Chemicals

Platinum® High Fidelity DNA polymerase, Zeocin, Hygromycin, Flprecombinase expression plasmid (pOG44), Chinese hamster ovary cells containing a single integrated Flp

Recombination Target (FRT) site (CHO Flp-In), and the mammalian expression vector pcDNA5/FRT/V5-His TOPO were obtained from Invitrogen Corporation (Carlsbad, CA). Ham’s

F12 Kaign’s modification cell culture medium was obtained from Sigma Chemical (St. Louis,

MO), as were the test inhibitors of MATE transport activity. [3H]1-Methyl-4-phenylpyridinium

([3H]MPP; 80 Ci/mmol) was synthesized by the Department of Chemistry and Biochemistry,

University of Arizona, [14C]Metformin was purchased from Moravak Biochemicals (Brea, CA) and [14C]Mannitol (58.8 mCi/mmol) was purchased from PerkinElmer. All drugs and other chemicals were purchased from Sigma (St. Louis, MI) and were of the highest quality available.

2.2 Cell Culture and Stable Expression of hMATE1 & hMATE2-K

Chinese Hamster Ovary cells (CHO) containing the Flp recombination target site were grown in Ham’s F12 Kaighn's modification medium supplemented with 10% fetal calf serum and

100 µg/ml Zeocin. Cultures were split every 3 days. 5 x 106 cells were transfected by electroporation (BTX ECM 630, San Diego; 260 volts and time constant of ~25 ms) with 10 μg of salmon sperm, 18 μg of pOG44, and 2 μg of pcDNA5/FRT/V5-His TOPO containing either a hMATE1 or hMATE-2K construct. Cells were seeded in a T-75 flask following transfection and maintained under selection pressure with hygromycin (100 µg/ml) for at least 3 weeks before use in transport studies.

2.3 Transport by hMATE1- & hMATE2-K Expressing CHO Cells

CHO cells expressing hMATE1 or hMATE2-K were grown to confluence in multi-well

(typically 24-well) plates. The general transport protocol was as follows (modifications of this procedure used in specific studies are described with the presentation of those results). Prior to transport experiments, the media was aspirated and the cells rinsed twice with room temperature

Waymouth buffer (WB) containing in mM: 135 NaCl, 28 D-glucose, 5 KCl, 1.2 MgCl2, 2.5

CaCl2, 0.8 MgSO4, and 13 HEPES-NaOH, pH 7.4. Transport was measured at room temperature

(~23-25oC) and was initiated by adding transport solution containing WB with (typically) 1

Ci/ml of radiolabeled compound (e.g., [3H]MPP). Since initial experiments showed that uptake of [3H]MPP was linear for ~15 min, 5 min uptakes were used to approximate the initial rate of transport of that substrate. After the transport period, the solution was aspirated and the wells were rinsed three times with 1 ml of ice-cold WB. The cells were solubilized in 0.2 ml of 0.5N

NaOH with 1% SDS (v/v), and the resulting lysate was neutralized with 0.1 ml of 1N HCl.

Accumulated radioactivity was determined by liquid scintillation spectrometry (Beckman model

LS3801). Individual transport observations were typically performed in duplicate for each experiment, and observations were usually confirmed, at least three times, in separate experiments using cells of a different passage.

2.4 Real-Time Quantitative-PCR

Total RNA was isolated from human cortical tissue (obtained from cadavers via the

International Institute For The Advancement of Medicine) using RNA later (Ambion) according to the manufacturer's protocol and reverse transcribed using the general procedure we employed in previous studies (QuantiTect Reverse Transcription, QIAGEN; (67; 85; 178)). The extracted

RNA was dissolved in diethylpyrocarbonate-treated water and then quantified by measuring the

absorbance at 260 nm. RNA (5 µg) was resolved on 1.5% formaldehyde-agarose gel to assess

RNA integrity. Real-time quantitative PCR on the resulting cDNA was carried out using an ABI

7300 Real-Time PCR System and TaqMan gene expression primers and probes (Applied

Biosystems). Standard curves were created using linearized purified plasmid DNA for the candidate transporters. DNA concentrations were measured using the BioPhotometer 6131, and dilutions of purified plasmid DNA from 10-2 to 10-8 were used in triplicate PCR reactions using the same mix and cycling procedures as with the RT product, producing a standard curve for each transporter. The Rotor Gene Analysis Software V6.0 was used to analyze the PCR results.

2.5 Immunohistochemistry

Surgically removed kidney specimens were immediately put on ice and fixed overnight by immersion in 2-4% paraformaldehyde at 4º C. The samples were rinsed in PBS for 2-3 minutes and dehydrated in ethanol (in series 50%, 70%, 80%, 95% and 100%) for 30 minutes.

The specimen blocks were then immersed in SPURR resin diluted in ethanol (1:1) for 24 hours at room temperature. The following day the samples were immersed in 100% SPURR for 48 hours at 4 C. Lastly, the kidney blocks were embedded in SPURR (100%) overnight in an oven at 60

C. Serial sections from plastic embedded blocks, cut by the University of Arizona's histology lab, were used to determine expression of hMATE1 and hMATE2-K. Specimens were washed with

PBS for 2 minutes at room temperature and then blocked with 10% goat serum for 30 minutes at room temperature in a humidified chamber. Primary antibodies for hMATE1 (diluted 1:200 in

10% goat serum; generously donated by Yosinori Moriyama Okayama University Dentistry &

Pharmaceutical Sciences Graduate School of Medicine Okayama-shi, Japan) or hMATE2-K

(diluted 1:100 in 10% goat serum; generously donated by Ken-ichi Inui Kyoto University

Hospital Department of Pharmacy, Kyoto, Japan) were applied and left overnight at 4°C, in a humidified chamber. The slides were gently shaken, ensuring that the reagents did not come off

the slides. The following day each slide was washed 3 times for 5 minutes at room temperature to remove any residual primary antibody. The fluorescently labeled secondary antibody (e.g.

Donkey anti-rabbit labeled with FITC) at a 1:500 dilution (diluted in 10% goat serum) were then applied for 1 hour at room temperature, while shaking (in the dark). The secondary antibody was removed and the slide was subsequently washed with PBS 3 times for 5 minutes each. The nuclear stain, TO-PRO3 was applied for 5 minutes and followed by 3 washes for 5 minutes each. The slides were then mounted with a cover slip with DAKO mounting medium.

2.6 Computational Modeling

A common features pharmacophore was developed using Accelrys Discovery Studio vers

2.5.5. (San Diego, CA) following the approach taken previously with other transporters (25; 26).

Template molecule structures were downloaded from ChemSpider (www.chemspider.com), and conformer generation was carried out using the CAESAR algorithm applied to the selected template molecules (maximum of 255 conformations per molecule and maximum energy of 20 kcal/mol).

2.6.1 3D-QSAR Development Using the Hypogen Method in Discovery Studio

hMATE1 IC50 values were used as the measure of biological activity. In the HypoGen approach (10; 34) ten hypotheses were generated using hydrophobic, hydrogen bond acceptor, hydrogen bond donor, and the positive and negative ionizable features, and the CAESAR conformer generation method was used. After assessing all generated hypotheses, the hypothesis with lowest energy cost was selected for further analysis, as this model possessed features representative of all the hypotheses and had the lowest total cost. The total energy cost of the generated pharmacophore was calculated from the deviation between the estimated activity and the observed activity, combined with the complexity of the hypothesis (i.e. the number of pharmacophore features). A null hypothesis, which presumes that there is no relationship

between chemical features and biological activities are normally distributed about their mean, was also calculated. Therefore, the greater the difference between the energy cost of the generated and null hypotheses, the less likely the generated hypothesis reflects a chance correlation. Also, the quality of the structure-activity correlation between the predicted and observed activity values was estimated via correlation coefficient.

2.6.2 Quantitative Model Update With Variable Weights and Tolerances

Hydrogen bond acceptor, hydrogen bond donor, hydrophobic, positive ionizable and negative ionizable features were selected for model building (again using CAESAR for conformation generation). Variable weights and tolerances were employed and a maximum of 10 pharmacophores were selected. The pharmacophore with the best correlation (lowest RMS error) was used for further analysis.

2.7 Statistics

Data were analyzed using Prism (GraphPad Software). Statistical tools such as paired t- test or unpaired t-test were applied according to the nature of each data set.

CHAPTER 3

MOLECULAR CHARACTERIZATION & QUANTITATIVE STRUCTURE ACTIVITY RELATIONSHIPS FOR INHIBITION OF HUMAN MATE1 & MATE-2Ka

3.1 Background

One of the main physiological functions of the kidneys is to clear the body of a multitude of structurally diverse organic compounds, the majority of which are xenobiotic, i.e., of exogenous origin. These include plant-derived compounds found in typical diets and, increasingly, synthetic pharmaceuticals of substantial clinical relevance. So-called ‘organic cations’ (OCs), i.e., molecules that carry a net positive charge at physiological pH, are a particularly significant subset of this group of compounds as they make up about 40% of all prescribed drugs (e.g., cimetidine, procainamide, pindolol and metformin) (103). Classic studies dating back to 1981 outlined what remains the basic cellular model of renal OC secretion: the sequential activity of (i) an electrogenic organic cation transporter on the basolateral membrane and (ii) an electroneutral OC/H+ exchanger on the luminal membrane of renal proximal tubule

(RPT) cells (57). Over the next 15-20 years, data gathered using in vivo stopped-flow tubular microperfusion, isolated single RPTs, cultured renal epithelial cells, and isolated membrane vesicles, characterized the activity of these two functionally distinct processes (112; 113).

Identification of the molecular basis of the basolateral “entry” step followed the cloning of organic cation transporter 1 (OCT1) in 1994 (46), and there is now a broad consensus that, in the

a Contents of this chapter were submitted for publication as ‘Bethzaida Astorga, Sean Ekins, Mark Morales and Stephen H Wright. Molecular Characterization and Quantitative Structure Activity Relationships for Inhibition of human MATE1 and MATE-2K. Am. J. Physiol. Renal Physiol., during preparation of this dissertation.

human kidney, basolateral OC transport is dominated by activity of organic cation transporter 2,

(OCT2; (100; 160)).

The molecular basis of the apical element in renal OC secretion, i.e., OC/H+ exchange, proved to be more elusive. Until recently, there were no viable candidates for the molecular identity of the apical OC/H+ exchanger, despite the fact that it is recognized as being both the active and rate-limiting step in renal OC secretion (98; 123; 161). Then, in 2005, Otsuka et al performed a BLAST search of the human and mouse genomes using the sequence of NorM (a

Na+/multidrug antiporter from Vibrio parahaemolyticus), a member of the newly described

Multidrug And Toxin Extrusion, or MATE, family of transport proteins (107). They identified two novel sequences, which they called MATE1 and MATE2. Functional characterization of hMATE1 led to the suggestion that it could be a component of the long sought OC/H+ exchange activity of the luminal membrane of RPT cells and hepatocytes (107). Masuda et al (89) subsequently found a kidney-specific splice variant of MATE2 they called MATE2-K that displayed the same “physiological fingerprint” of apical OC transport, as that of MATE1; both (i) were highly expressed in the kidney; (ii) were localized to the apical membrane of RPT cells; (iii) supported OC/H+ exchange; and (iv) mediated transport of several prototypic OC substrates.

Tanihara et al (136) extended the latter observation, showing that MATE1 and MATE2-K both support mediated uptake of at least 12 structurally diverse radiolabeled OCs. The quantitative link between MATE activity and renal OC secretion became evident in the observation that elimination of Mate1 significantly reduced renal clearance of metformin (143) and cephalexin

(157) in mice.

Despite increasing evidence that MATEs represent the rate-limiting step in OC secretion, virtually nothing is known about key parameters that may assist in the prediction of unwanted drug-drug interactions DDIs. Indeed, the quantitative profiles of selectivity for neither MATE

transporter have not been established, nor have the molecular determinants that define the basis of interaction of potential substrates and/or inhibitors of the MATEs been determined to date.

Nevertheless, there is interest in understanding the role of OC transporters, including MATEs, in

(DDIs; (36)). For example, therapeutic doses of cimetidine retard the renal elimination of procainamide (126; 127) and metformin (128). In addition fexofenadine (antihistamine drug), which is taken into the RPT cell via Organic Anion Transporter 3 (OAT3) and secreted into the tubular lumen by MATE1, is significantly inhibited by probenecid at the entry step (OAT3) and by cimetidine at the exit step (MATE1); showing that there are multiple potential sites for DDIs in the RPT (91). Moreover, the clinical cost of DDIs is substantial and responsible for about 1% of hospital admissions (almost 5% in elderly populations; (8)). Thus, understanding the transport mechanisms that underlie the processes of renal and hepatic OC clearance is particularly relevant to efforts to predict and pre-empt the unwanted outcomes of drug exposure.

Previously, combining in vitro data with computational modeling of transporters including OCT1 and OCT2 enabled the development of pharmacophore and quantitative structure activity relationships (QSARs) which facilitated an understanding of the likely structure activity relationships for each transporter (10; 131). More recently computational models of transporters have been used for searching a subset of FDA approved drugs to suggest additional compounds for testing to discover novel inhibitors for hPEPT1 (32; 79), P-gp (17), OCTN2 (25; 26) and

ASBT (180). These methods included pharmacophores and Bayesian approaches and generally had good levels of predictability. In the current study we have taken the approach of using multiple sequential rounds of pharmacophore development and searching of a more comprehensive set of FDA drugs in order to (i) characterize the relative selectivity of MATE1 and MATE2-K toward a set of clinically important OCs; (ii) develop initial predictive models of their selectivity; and (iii) identify novel inhibitors of these MATE1 and MATE-2K processes.

3.2 The Kinetics of MPP and H+ Interaction With hMATE1 & hMATE2-K

The kinetics of MATE1- and MATE2-K-mediated MPP transport (Fig. 4A, 4B) were described by the Michaelis-Menten equation for competitive interaction of labeled and unlabeled substrate (87).

J [S*] J*max D [S*] eq. 1 K [S*] [S] ns tapp where J* is the rate of transport of the radiolabeled substrate (in this case, [3H]MPP) from a concentration of the labeled substrate equal to [S*]; Jmax is the maximal rate of mediated substrate transport; Ktapp is the apparent Michaelis constant of the transported substrate; [S] is the concentration of unlabeled substrate; and Dns is a rate constant that describes the nonsaturable component of labeled substrate accumulation (reflecting the combined influence of diffusion, nonspecific binding, and incomplete rinsing of [3H]MPP from the cell culture well). In four separate experiments Kt for MATE1 and MATE2-K were 4.70 ± 0.55 µM and 3.30 ± 0.19 µM,

-2 respectively; with Jmax values of 2.14 ± 0.12 and 1.86 ± 0.65 pmol/(cm x min), respectively.

-1 -1 Expressed per mg of membrane proteins, these Jmax values become 21.4 and 18.6 pmol mg min for MATE1 and MATE2-K, respectively.

There are very few values for the Jmax and Kt of MPP transport in MATE1 and MATE2-

K-expressing cells in the literature. In comparison to data reported by others, our Kt values fall within range (refer to Table 1); however our Jmax differ by a factor of ~35. Taking a wider look at the literature reveals, some interesti,ng comparisons. Tanihara et al for instance, studied hMATE1

-1 -1 expressed in HEK 293 cells, and they reported a Jmax values of 740 pmol mg min (compared to

-1 -1 our value of approx. 20 pmol mg min ; (136)). Similarly, reported Kt values for hMATE2-K-

-1 -1 mediated MPP transport are between 94-110 µM, with a Jmax of 575 pmol mg min (89; 136), compared to our hMATE2-K values of ~4 μM and ~20 pmol mg-1 min-1. Interpreting the basis for the differences in kinetic values reported here and those reported in previous studies is complicated by the fact that our experiments and those of other studies typically used rather different experimental conditions, including a different cell expression system (HEK 293 versus

CHO), different incubation temperatures (37oC versus room temperature), and an external pH of

7.4 under ‘ammonia pulse’ conditions of intracellular acidification (versus our use of an external pH of 8.5). The 20- to 30-fold differences in Jmax most likely represents the combined influence of (i) different temperatures; (ii) the large intracellular acidification employed in the other studies noted above; and (iii) the different cell lines used to perform the measurements. By this, we do not mean CHO cells verses HEK 293 cells as the expression system; we have compared the kinetics of MPP transport and the inhibitory profiles of selected test agents in both CHO cells and

HEK 293 cells that express hMATE1 and found both Kt and Jmax values to be comparable

(unpublished observations). Rather, we are referring to our use of the Flp-recombinase system of stable transfection that results in insertion of a single copy of the MATE DNA sequence, into the cell genome. Studies by Otsuka and Tanihara (and others) typically have used the highest expressing cells arising from selection for antibiotic-resistance following traditional transient transfection, a method that can result in stable expression of multiple copies of the selected transporter. Thus the difference in Jmax between our studies and those in previous studies probably is due to differences in MATE expression level. Kt values, on the other hand, ought to be independent of expression level and, therefore, there is a reasonable expectation that Kt values measured by different groups should be reasonably comparable. But, as noted the sample size for comparison is small. In one study (136), the Kt for hMATE1-mediated MPP transport was 100

µM, much larger than ours, whereas in another study, the reported value was 16 µM (89), which

is quite comparable to ours (once difference in external pH is taken into account). We will have more opportunity during our discussion of the kinetics of MATE inhibition to draw comparisons with other reports in the literature.

Of course, it is also important to note that MATEs are sensitive to the intra-and extracellular concentration of protons and therefore we characterized the kinetics of H+ inhibition of [3H]MPP uptake. We previously showed that elevated concentrations of H+ in the extracellular solution inhibit transport mediated by hMATE1 (22). Figure 4C,D compares the pH sensitivity of

MPP transport mediated by hMATE1 and hMATE2-K. As anticipated, transport activity of both proteins was inhibited by increasing concentrations of H+ in the extracellular solution, and was described by the following relationship:

J [S*] J*app D [S*] eq. 2 IC [H ] ns 50 o

3 where Japp is the product of the maximum rate of S* (i.e., [ H]MPP) uptake (Jmax) and the ratio of

+ + the Ki of H and Kt for MPP transport; and IC50 is the concentration of [H ]o that reduced

3 + mediated (i.e., blockable) [ H]MPP transport by 50%. In 3 experiments, the IC50 for H -inhibited hMATE1-mediated MPP uptake was 19.6 ± 0.71 nM (pH 7.7, n = 3), similar to the value of 12.4 nM reported previously (22). MATE2-K proved to be substantially more sensitive to H+, displaying an IC50 value of 3.5 ± 0.55 nM (pH 8.5, n=3).

The physiological profile of luminal OC transport in native RPT cells H+ has long been known to include a functional interaction with H+. In renal brush border membrane vesicles, external (cis) H+ acts as a competitive inhibitor of OC interaction with the external face of the exchanger (163), consistent with the results in Figure 4C,D, and with the previous observation

(22) that inhibition of MATE1-mediated MPP transport produced by external H+ is largely competitive in nature. Similarly, the hallmark of luminal OC transport in RPT is the stimulatory

effect on OC transport of an oppositely-oriented H+ gradient (70), an effect shown in work with renal BBMV to reflect an increase in the turnover number of the OC/H+ exchanger produced by trans H+ gradients (165). Acknowledging that changes in both intra and extracellular pH exert direct and marked effects on the kinetics of MATE-mediated OC/H+ exchange introduces the issue of how best to probe the mechanism of ligand interaction with MATE transporters. Because rates of MATE-mediated uptake are comparatively low when the extracellular pH is 7.5 or less

(Figure 4B; (107; 142; 177)), many studies have employed the ‘ammonia-pulse’ method to acidify the cytoplasm, thereby creating an outwardly-directed pH gradient and a stimulation of

OC uptake (68; 89; 105; 107; 136; 172). However, the acidification of the cytoplasm following an ammonia pulse is generally short-lived and constantly changing (22) during the several minute time courses used to measure the rate of MATE-mediated transport, and these ill-defined conditions complicate the interpretation of kinetic measurements. Consequently, we elected to maximize control rates of MATE-mediate transport by running all transport experiments at an external pH of 8.5. Importantly, we previously showed that cytoplasmic pH is effectively constant (at a pH of 7.5-7.6) during exposure to an external pH of 8.5 (22), so transmembrane H+ gradients are both (i) outwardly-directed) and (ii) unchanging during transport measurements.

Furthermore, the interaction of H+ with MATE transport activity appears to be effectively restricted to that of a competitive ligand (22). We suggest that the rank-order of ligand selectivity at pH 8.5 and 7.4 is similar, if not identical, for the two transporters, as supported by the similar rank order of uptake ratios for transport of a structurally diverse set of organic cations into hMATE1 and hMATE2-K at these two pH values (136). However, given the apparent pKa values for the two MATE transporters the absolute IC50 values for inhibition of MATE1 and MATE2-K activity measured at pH 8.5 can be expected to underestimate the values anticipated at pH 7.4, by approximately 3- to 6-fold, respectively.

3.3 Inhibitory Selectivity of hMATE1 and hMATE2-K: Test Set Selection

The test compounds were selected to represent a structurally diverse collection of drug and drug-like compounds, the intention being to interrogate the complex chemical space expected to influence interaction with the binding regions of multiselective organic cation transporters, i.e.,

MATE1 and MATE2-K. Predominating the profile of test compounds were weak bases and cations; neutral compounds or those having a net negative charge at physiological pH were largely excluded. This ‘bias’ toward cations reflected the existing database from the early literature on transport in isolated renal membranes (56; 57) and in intact renal tubules (23; 92), and from more recent work with MATE transporters (136), showing that cationic charge is a key criterion of ligand interaction with these processes (a conclusion supported by the present study, as documented below). Indeed, it was a specific goal of this study to identify molecular determinants of interaction of organic cations with MATE1 and MATE2-K. The final battery of compounds included 23 from the list of compounds generated by Ahlin et al. (3), in their study of selectivity of OCT1; 13 compounds selected because of previous evidence of their interaction with OC/H+ exchange activity in either native renal membranes, intact tubules, BBMV or heterologous expression systems expressing MATE1 or MATE2-K (45; 61; 73; 99; 115; 116;

174); and 23 compounds selected from lists of target ‘hits’ from databases of compounds that were interrogated by the pharmacophore model(s) developed during the course of this study. The final collection of test inhibitors included 59 compounds (Table 2 for hMATE1 and hMAT2-K).

Figure 5 shows the range of inhibition of MATE-mediated transport activity produced by our battery of 59 test compounds. At 10 M, MPP transport was reduced by ≥50% by 20

(MATE1; 5A) or 14 (MATE2-K; 5B) of these compounds. Figure 6 shows inhibitory profiles against transport activity of MATE 1 and MATE2-K produced by four compounds (quinidine,

agmatine, nialamide and allopurinol) selected to emphasize the spectrum of inhibitory effectiveness of our battery of test agents, with IC50 values (determined using equation 2) that ranged from micromolar (e.g., quinidine), through micromolar (e.g., nialamide), to no effective interaction at all (e.g., allopurinol). Table 2 lists the IC50 values (as measured at pH 8.5, and as calculated for pH 7.4) for all 59 compounds used to inhibit transport activity of one or both

MATE transporters.

The current database on selectivity of human MATEs is sufficiently sparse that it is difficult to compare our observations with those in the literature. Nevertheless, a few comparisons are noteworthy for their rather close agreement with the observations reported here.

The antimalarial drug, pyrimethamine (PYR), the highest affinity inhibitor in the present study

(Table 2), was reported to have an IC50 of 93 nM against hMATE1-mediated metformin transport

pH7.4 (76), very similar to the IC50 value we predicted for inhibition at pH 7.4 (IC50 ) of MATE1- mediated MPP. Also, both hMATE1 and hMATE2-K support transport of the antidiabetic drug, metformin, with Kt values of 238 μM (94) and 1.1 mM (89), respectively, not unlike the

pH7.4 calculated IC50 we measured here (123 µM and 581 µM). But not all the comparisons of the present observations correspond so closely to those observed in previous studies. Tacrine for example has IC50s of 0.6 µM and 1.1 µM in hMATE1 and hMATE2-K expressing CHO cells

(respectively); however, literature values for tacrine inhibition of 4-4-dimethlamino-styryl)-N- methyl-pyridinium (ASP) are 1.1 µM and >100 µM (respectively; (73)). The experimental conditions used in the aforementioned study varied from those employed here (i.e. HEK-293 cells at pH 7.4) and could account for some of the difference. However, it is more plausible that these differences are due to the use of MPP versus ASP (73). Tacrine inhibition of MPP may utilize a different transport pathway from tacrine inhibition of ASP and, therefore, result in the observed

100-fold difference in affinity for hMATE2-K expressing cells. The MATEs may display

‘mixed-type inhibition’ (for either MPP or ASP), where tacrine inhibits MPP or ASP transport in a non-competitive manner, via binding of tacrine at a different site from MPP or ASP resulting in the marked difference. ASP or MPP could be binding at different sites and, therefore, tacrine may elicit a distinctive pathway of transport. Furthermore, these data suggest that there could be multiple binding sites or a ‘binding surface’ for the wide-array of chemical structures that interact with MATEs. Evidence and a discussion for this topic are presented in further detail in later sections of this manuscript.

As shown from the results presented in Figures 4 through 6, there was substantial overlap in the interaction of the test compounds with MATE1 and MATE2-K. The extent of this overlap is evident in the comparison of the MATE1 and MATE2-K IC50 values for the 59 test compounds used to probe both transporters (Figure7A); 69% of these agents had IC50 values for the two transporters that differed by less than a factor of 3 at pH 8.5 (78% of the compounds, based on the

pH 7.4 calculated IC50 ). The two proteins did markedly discriminate between a few of the compounds in this study. For example, as shown in Figure 7B, the apparent affinities of hMATE1 for atropine (IC50 of 5.90 µM) and amantadine (7.50 µM) were ~10-times greater than those displayed by MATE2-K (52.8 µM and 88.9 µM, for atropine and amantadine, respectively; whereas the apparent affinity of hMATE2-K for azidoprocainamide (APMI; IC50 of 0.50M) was

~10 times greater than that displayed by hMATE1 (6.2 M; Figure 7B). Despite these differences, the data support the view that MATE1 and MATE2-K show far more similarities in selectivity, than differences.

3.4 Expression of MATE1 and MATE2/2-K in Human Kidney

Unlike the mouse and rat kidneys, which do not express Mate2/2-K (54; 81), the human kidney expresses both homologs. In one previous study (89), comparable levels of mRNA for

MATE1 and MATE2/2-K were found to be expressed in a commercial sample of mRNA from

human kidney. Figure 8 shows our assessment of mRNA levels for several xenobiotic transporters in samples collected from outer cortex, inner cortex and medulla from two human kidneys. In both kidneys (i) mRNA for all the transport proteins was much higher in the outer and inner cortex (which generally showed similar levels of expression) than in the medulla; and

(ii) MATE2/2-K expression was 8-10 fold higher than that of MATE1. In the original report of the cloning of MATE2-K, the level of MATE2-K protein expression in Western blots of human renal BBMV appeared to be at least as large as that of MATE1. Figure 9A-B shows the immunohistochemical localization of MATE1 (Fig. 9A) and MATE2/2-K (Figure9B) in human renal cortex at 40X magnification (insert at 10X magnification). As reported previously (89;

107), both proteins were clearly localized to the brush border membrane of proximal tubules.

Although these data do not permit firm conclusions to be drawn concerning the possible co- expression of the two transporters within RPT cells, the extensive expression of each protein in outer cortex is consistent with the general expression of both transporters in RPTs. These data, along with the observations of marked overlap of selectivity of the two MATEs suggest that it is likely that each plays a quantitatively significant role in renal OC secretion (though each may play unique roles in the secretion of compounds for which they display uniquely high affinity).

3.5 Modeling of MATE Selectivity

Previous studies of inhibition of OC/H+ exchange activity in isolated renal brush border membrane vesicles (165-167) and intact microperfused renal proximal tubules (24; 129; 145; 146;

149) sought to correlate inhibitor effectiveness with selected, single physicochemical characteristics of the test agents included in these studies, such as logP (hydrophobicity) and pKa

(basicity). More recently, a marked correlation between total polar surface area (TPSA) and inhibition of OCT2 activity was reported (181). Using these molecular descriptors, we graphed

IC50s of hMATE1 as a function of logP, pKa and TPSA. There was a significant, albeit weak

correlation between hMATE1 IC50 values and logP (r value of 0.332, p < 0.05; Figure10A). This correlation was even more evident when the complicating influence of steric differences was minimized by limiting the comparison to a set of structural congeners; Figure10B shows the significant correlation between LogP values for a set of n-tetraalkylammonium compounds and their measured IC50 values for inhibition of MATE1-mediated MPP transport (r = 0.97 for TEA

+ through TPeA). Similalry, the IC50 for inhibition of OC/H exchange activity in rabbit renal

BBMV showed a marked correlation with LogP (145; 149), although the inhibitors studied were rather structurally constrained, and steric differences (and other physicochemical parameters) appear to mask this simplistic comparison. Indeed, there was no correlation between hMATE1 values and TPSA (r value of 0.045, p > 0.05; Figure10C). For pKa there was a weak, but significant, correlation with hMATE1 IC50 values (r value of 0.423, p < 0.01; Figure10C).

We were not surprised that single molecular discriptors failed to be particularly effective predictors of inhibitory interaction with the MATEs. Given the broad structural diversity of compounds that interact effectively with the MATEs, it is likely that binding is a more complex process requiring multiple molecular interactions and, therefore, that single physicochemical properties alone can provide only a partial description.

3.5.1 Computational Analysis of hMATE1 Inhibition

The value of the approach offered by computational assessment of structure/activity relationships (SARs) is that it may enable more insight with regard to the molecular basis of ligand interaction than can a focus on individual physicochemical parameters alone. The results presented here focus primarily on model development with hMATE1, in large part because initial modeling efforts with hMATE2-K did not produce stable and reliable models.

Following our previous in silico modeling efforts on drug transporters, (26; 39) we extended our in vitro-in silico (IVIS) strategy in this study by using multiple iterations of model

development. Normally researchers tend to produce a large amount of high throughput screening data against a target of interest and then generate a computational analysis (perhaps looking for correlations with simple physicochemical properties, from which some rules may be inferred).

The IVIS approach is different in that we started with a small amount of in vitro data and used it for initial models that were then used to screen a database of additional potential compounds for testing. The model retrieved compounds of interest from which we selected a new test set, and the data from the next round of testing then fed into a further round of model building and database searching, and so on. The advantage of such an approach is the model is co-developed with the data, and is validated and tuned with each additional set of test compounds. The approach does not require a large library of compounds to be tested and can save reagents and money (associated with testing many inactive compounds). This pharmacophore approach may suggest non-intuitive compounds as inhibitors because they include one or more of the initially mapped features but prove to be low affinity ligands (owing to the absence of what proves to be a missing critical feature). This in particular may be another potential valuable side effect of the approach, enabling us to find novel compounds that may have a similar shape to known inhibitors but different mapping of pharmacophore features.

3.5.2 Initial Round: hMATE1 Common Features Pharmacophore

There were 26 compounds in the initial round of inhibitors studied (Table 2, subset “a”), from which five were selected to generate a common features pharmacophore: two high affinity compounds, pyrimethamine (IC50 of 0.04 µM) and quinidine (IC50 of 1.57 µM) and three low affinity compounds, histamine (IC50 of 761 µM), caffeine (IC50 of 1096 µM) and chloramphenicol (IC50 of 1115 µM; Figure 11A). Common molecular and chemical features of the high affinity substrates were included in the pharmacophore, whereas the molecular features of the low affinity substrates were excluded from the pharmacophore. Figure 11B shows the

resulting ‘common features pharmacophore’, which included 2 hydrophobic regions (cyan), 1 H- bond donor (magenta), and 1 H-bond acceptor (green). The pharmacophore is depicted with

(Figure11B) the highest affinity substrate.

3.5.3 hMATE1 Common Features Pharmacophore Testing

The common features pharmacophore, including the van der Waals surface of pyrimethamine to provide a shape restriction, was used to search a 3D database of 2690 FDA approved compounds (www.collaborativedrug.com), and identified 126 molecules as potential inhibitors (Table 2). Of these, 15 commercially available compounds were selected and tested as inhibitors (Table 2, subset “b”) of MATE1 (and MATE2-K), and the resulting IC50 values are presented in Table 2. Nine compounds in this test set proved to be comparatively high affinity inhibitors of MATE1 (IC50 values of 0.9-25 µM), whereas four displayed modest affinity (IC50 values of 26-300 µM), two had weak interactions (<60% inhibition at 1000 M) and one exerted no inhibition (NI) of either transporter at a concentration of 1 mM.

The inhibitory profiles produced by two of the 15 compounds in this ‘test set,’ cinchonidine and ethohexadiol (Figure 12A, B), provided insight into the molecular determinants associated with ligand interaction with MATE1. Although both molecules were good fits for the common features pharmacophore (Figure12C,D; fit values of 3.1 and 2.7, respectively, data not shown), cinchonidine was an effective, high affinity inhibitor of hMATE1 (IC50 = 0.93 µM; 12A) whereas ethohexadiol (Figure 12B) had a weak interaction, with approximately 20% inhibition of transport at a concentration of 1 mM. Thus, the presence in ethohexadiol of the common structural features of the pharmacophore, i.e., the 2 hydrophobic regions, 1 H-bond donor site and

1 H-bond acceptor site), was not sufficient for an effective inhibitory interaction with MATE1. It is noteworthy that ethohexadiol is not an organic cation; its presence in the 'hit list' reflected the fact that the common features pharmacophore lacked a cationic (positive ionizable) feature,

because all three of the ‘weak inhibitors’ of MATE1 activity (Figure11A) included such a feature.

Consequently, whereas a cationic feature is not itself sufficient to insure a strong interaction with

MATE1, these data argue that charge exerts a strong influence on the binding interaction.

3.5.4 First Iteration: Quantitative Pharmacophore Development for hMATE1

We generated a quantitative pharmacophore model in parallel to the common features hMATE1 pharmacophore, taking advantage of the broad range of activities (IC50 values from 40 nM to >5 mM) displayed by the initial round of inhibitors. 24 of the initial 26 compounds (H+, due to its small chemical shape, and verapamil because it was a racemic mixture, were not included) were used in an analysis that resulted in a model containing two hydrophobic features

(cyan), one hydrogen-bond donor (magenta), and, unlike the common features pharmacophore, one positive ionizable feature (red: Figure 13A). The model had a small cost difference as total cost =125.97 and null cost = 137.57, suggesting a modest quality model. Nevertheless, unlike the modeling efforts based on single physical descriptors (Figure 10), the correlation between observed and predicted IC50 values resulted in r = 0.68 (p < 0.0001; Figure 13B).

3.5.5 Second Iteration: Quantitative Pharmacophore Development for hMATE1

Of the 43 compounds (the initial 24 plus the “test set” of 15 compounds and two other compounds, derived from the database searching, that probed the common features model) used to generate and validate the two pharmacophores, the most potent inhibitor of hMATE1 was pyrimethamine (PYR). Consequently, we chose to probe two structural analogs of PYR (5-(4- chlorophenyl)-6-ethyl-2,4-pyrimidinediamine):1-(2-chlorophenyl)-6,6-dimethyl-1,6-dihydro-

1,3,5-triazine-2,4-diamine (PYR-2); and 1-(3-chlorophenyl)-6,6-dimethyl-1,6-dihydro-1,3,5- triazine-2,4-diamine (PYR-3). The IC50 values of 0.04, 0.30, and 0.32 µM for PYR, PYR-2 and

PYR-3, respectively (Table 2), showed that the modest differences in structure between these three compounds had comparatively little impact on their inhibitory interactions with MATE1,

and suggested that the structural features of this suite of compounds provide insight into molecular characteristics that optimize ligand interactions with the protein’s binding site/surface transport. A second quantitative pharmacophore model for hMATE1 reflecting these data was generated (Figure 14A), and also included 2 hydrophobes (cyan), 2 hydrogen-bond acceptors

(green), and an ionizable feature (red). However, this model represented only a modest increase, over the first iteration QSAR, in its ability to effectively predict affinity. The observed and predicted data resulted in an r value of 0.71 (p < 0.0001; Figure 14B), therefore we tested more compounds in an attempt to increase the model’s predictive power.

3.5.6 Final Iteration: Quantitative Pharmacophore Development for hMATE1

We ultimately screened 59 compounds, adding several novel structural groups including the n-tetraalkylammonium series mentioned earlier. For the final quantitative pharmacophore we chose to remove all compounds that had a pKa value below 8.0, the rationale for doing so being to eliminate any interpretational issues associated with compounds that were not ionized at the experimental pH of 8.5. Of the 59 compounds screened we removed 13 to generate the pharmacophore depicted in Figure 15A (named the N46 model). The N46 model had several features in common with the previous iteration models, in that it included two hydrophobes

(cyan), a hydrogen bond-acceptor (magenta), and an ionizable feature (red), though the spatial distribution of these elements differed somewhat from the previous model. Figure 15B displays the relationship between measured and predicted IC50 values based on the N46 model (r value of

0.73, p < 0.0001).

3.6 Discussion & Conclusions

The present observations support several conclusions concerning the molecular basis of selectivity of the mammalian MATEs. First, it supports the hypothesis that no single physicochemical parameter of ligand structure is likely to provide an adequate predictor of

interaction with MATE1 or MATE2-K. Second, we identified several structural features that provide insight into the basis of ligand interaction with the MATEs, i.e., multiple hydrophobes

(logP); hydrogen donors; ionizable (i.e., cationic) feature. With respect to these two latter points, a recent study by Kido et al (73) that screened some 900+ compounds for inhibitory interaction with hOCT2 noted that inhibitory effectiveness was particularly influenced by (i) ligand liphophilicity and (ii) average charge, as well as (iii) molecular volume, (iv) TPSA, and (v) the number of hydrogen bond donors and acceptors. Third, the several iterations of pharmacophore development led to the identification of 11 new clinical classes of compounds as likely MATE ligands.

Finally, there is valuable insight concerning the molecular basis of ligand interaction with the MATEs in the relatively modest quantitative predictive power of the final pharmacophore

(Figure 15A). While it scored many unique ‘hits,’ it also scored some important ‘misses’

(including MPP and TEA). There is, we suggest, important information in this observation.

Pharmacophore analysis makes the tacit assumption that there is a ‘most effective structure’ for interaction with a binding ‘site,’ with the pharmacophore representing a statistical average of all the chemical features important for interaction with that site. The physiological role of MATE transporters, however, is to interact effectively with a multitude of very structurally diverse compounds. The structure of the P-glycoprotein (119; 120), the archetype multidrug transporter, provided insight into the molecular basis of interaction with a structurally diverse population of ligands by revealing a large binding surface with multiple binding sites. It is not difficult to imagine that ligand binding to the mammalian MATE transporters also involves interaction with a ‘surface’ containing multiple, overlapping sites, as mentioned previously. An interesting example to support this theory comes from the study by Kido et al (73) where they screened a series of OCs for inhibition of transport of the fluorescent OC, ASP, in HEK cells expressing

hMATE1 or hMATE2-K. We used the seven IC50 values from Kido et al to generate a quantitative pharmacophore, which resulted in a fundamentally different structure from that of the pharmacophores we generated with [3H]MPP inhibition. The Kido quantitative pharmacophore had 3 hydrophobes (cyan), 2 hydrogen bond acceptors (green) and 3 excluded volumes (grey;

Figure 16). Importantly, spatial configurations of these features differed substantially from those generated with our data. This is not only suggestive of the binding surface (vs. binding site) theory but also of the mixed-type inhibition theory. To that end, it is relevant to note that, whereas tacrine was a comparable inhibitor of MATE1-mediated transport of MPP (in our study;

IC50 of 0.6 µM) and ASP (Kido et al, IC50 of 1.1 µM), tacrine was an extremely weak inhibitor of

MATE2-K-mediated ASP transport (IC50>100 µM), while being a potent inhibitor of MATE2-K- mediated MPP transport (IC50 of 1.5 µM). This is similar to what has been seen with CYP3A4, where multiple substrate probes are needed to fully understand whether a compound is an inhibitor or not (39; 71). It is possible that the current probes are only revealing some of the potential inhibitory effects of compounds such that if a structurally different probe was used it might show a failed inhibitor of one probe was in fact an inhibitor of the second probe. This may be the case with many transporters in that compounds could show ‘probe-dependent inhibition.’

In conclusion, we generated the first database of considerable size for ligand interactions of hMATE1 and hMATE2-K. Applying these data over the course of several computational modeling iterations using the IVIS approach resulted in a series of pharmacophores for hMATE1.

The hMATE1 pharmacophores indentified key structural features strongly correlated with ligand binding to hMATE1. The observations also supported the view that inhibitory profiles derived from the use of structurally distinct transported substrates can result in distinct pharmacophores, consistent with the contention that hMATE1 may have a complex binding surface for ligand interaction, rather than a single binding site.

CHAPTER 4

ASSESSMENT OF THE COMPLEX BINDING SURFACE OF HUMAN MATE1

4.1 Background

Organic cations (OCs) are cleared from the body via the liver and/or the kidney.

However, the main site of OC secretion is the renal proximal tubule (RPT; (1; 2; 55; 59; 110;

117; 154)). At the cellular level OC secretion occurs via two main steps: uptake into the cell across the basolateral or peritubular membrane and efflux out of the cell across the luminal or apical membrane and into the tubule filtrate for subsequent excretion (56; 57). It is widely agreed that in the human kidney OCs enter the cell across the peritubular membrane, via facilitated diffusion using Organic Cation Transporter 2 (OCT2; (100)). For exit from the RPT cell, OCs are exchanged for a proton and are released into the lumen of the tubule (56; 57). In recent years there has been a growing consensus that the luminal step (the active and rate-limiting step in OC secretion in the RPT (123)) is mediated by two members of the Multidrug And Toxin Extrusion

(MATE) family: MATE1 (SLC47A1) and MATE2-K (SLC47A2). Humans consume OCs on a daily basis, as plant products (xenobiotics) and/or prescribed drugs. In fact, approximately 40% of all prescribed medications are OCs (103) such as the anti-diabetic drug metformin or cimetidine (histamine receptor antagonist used to treat peptic ulcers). Therefore, the pathway of

OC secretion in the RPT, and subsequent excretion, plays a significant role in homeostasis and human physiology.

In the preceding chapter I described my work toward development of a predictive model of ligand interaction with hMATE1 and hMATE2-K. Although that work succeeded in identifying novel drug classes (highlighted in bold in Table 2) of ligands that interact with hMATE1 and structural components of those ligands important for interaction (i.e. hydrophobic

regions, H-bond acceptor and donor sites etc.), the results could also be interpreted as indicating that ligand interaction with MATE proteins may not involve a ‘classical’ interaction with a single, common binding site. In this Chapter I examine in more detail the kinetic basis of ligand interaction with hMATE1. I found that, whereas some ligands do display a pure competitive profile of inhibitory interaction, others, including molecules sharing the structural features of the high affinity inhibitor of MATE activity, pyrimethamine (PYR; (76)), do not. The data are consistent with the view, introduced in the preceding chapter, that the binding of substrates and inhibitors of MATE transporters involves interaction with a large binding surface that contains multiple binding regions. Hence, future studies that attempt to define the molecular descriptors of ligand interaction with MATE transporters are advised to use multiple transported probes.

To further probe these ideas we elected to study five compounds in more depth. To keep the data consistent we continued our studies using MPP (unlabeled and radiolabeled) as the principal test substrate. The first compound that we studied was metformin (anti-diabetes drug), not only because of its pharmacological relevance but also because we knew that it is a transported substrate of MATE1 and MATE2-K (9; 19; 76; 89; 137; 140). We also chose creatinine because of its physiological importance and because like metformin it is a small, very basic OC. Lastly, we chose to probe PYR and two of its structural analogs, 1-(2-chlorophenyl)-

6,6-dimethyl-1,6-dihydro-1,3,5-triazine-2,4-diamine (PYR-2) and 1-(3-chlorophenyl)-6,6- dimethyl-1,6-dihydro-1,3,5-triazine-2,4-diamine (PYR-3). PYR, PYR-2 and PYR-3 were the highest affinity inhibitors (Table 2) of MATE1 -mediated [3H]MPP transport and therefore we wanted to elucidate if they interact with the same site(s) as MPP.

4.2 hMATE1-Mediated Transport Kinetics for MPP & Metformin

The kinetics of MATE1-mediated MPP and metformin transport were described by the

Michaelis-Menten equation for competitive interaction of labeled and unlabeled substrate (87). as described in Chapter 3. However, to assist the reader, it is presented again, here:

transport; Ktapp is the apparent Michaelis constant of the transported substrate; [S] is the concentration of unlabeled substrate; and Dns is a rate constant that describes the nonsaturable component of labeled substrate accumulation (reflecting the combined influence of diffusion, nonspecific binding, and incomplete rinsing of [3H]MPP from the cell culture well).

Figure 17A, B shows a Kt value of 4.9 µM for MPP (17A) and 64.3 µM for metformin

2 (17B) and Jmax values of 3.0 and 9.3 pmol/(cm x min) respectively (Table 3). Using the transport

-7 efficiency ratio (TE), which is calculated by dividing Jmax by Kt, MPP, with its TE of 6.12 x 10 cm-2 x min-1, is transported about 6 times more efficiently than metformin (1.45 x 10-7 cm-2 x min-1), suggesting that while a metformin-loaded transporter has a higher turnover number than when loaded with MPP, the higher affinity of MATE1 for MPP results in a more effective interaction with the transporter.

4.3 Kinetic Basis of Metformin’s Inhibition of hMATE1-Mediated [3H]MPP Transport

I noted in Chapter 3 that metformin was a moderately effective inhibitor of hMATE1 and hMATE2-K (Table 3). Figure 18A shows the inhibitory profile of metformin’s inhibition of

MPP, resulting in an IC50 of 47.0 µM. The similarity of this inhibitory constant to the Kt for metformin transport (64.3 µM) reported above, is consistent with the view that MPP and

metformin interact competitively at a common, mutually exclusive binding site. To further test this hypothesis, we measured the kinetics of MPP transport in the absence and presence of 250

µM metformin (Figure 18B) using Eadie-Hofstee plots of the carrier-mediated (i.e., saturable) components of these two data sets, which emphasizes the apparent competitive nature of the interaction of metformin and MPP (i.e. no change in Jmax with an increase in apparent Kt). The presence of metformin had no significant effect on the Jmax of MATE1-mediated MPP transport

(3.02 pmol/(cm2 x min) versus 3.93 pmol/(cm2 x min), in these experiments; Table 3), but increased the apparent Kt by 8-fold (3.28 µM to 35.5 µM).

4.4 Kinetic basis of Creatinine Inhibition of hMATE1-Mediated [3H]MPP Transport

Metformin is a small, extremely basic OC (MW of 129.1, pKa of 12.4). We elected to determine the kinetic basis of inhibition of MATE1-mediated MPP transport produced by another comparatively small, basic OC, creatinine (MW 113.1; pKa 9.4). Figure 19 shows that, like metformin, creatinine’s inhibition of MATE1-mediated MPP transport fits a classical competitive profile, elevating apparent Kt without influencing Jmax. Figure 19A shows the inhibitory profile of creatinine’s inhibition of hMATE1, resulting in an IC50 of 195 µM. As with metformin, the

3 presence of 1 mM creatinine had no significant effect on the Jmax of MATE1-mediated [ H]MPP

2 2 transport (3.02 pmol/(cm x min) versus 2.66 pmol/(cm x min)), but increased the apparent Kt by

~2-fold (4.89 µM to 10.4 µM), consistent with the view that creatinine is also a competitive inhibitor of MPP. Again an Eadie-Hofstee plot was used to highlight these differences (i.e. no change in Jmax with an increase in apparent Kt).

4.5 Kinetic Basis of PYR Inhibition of hMATE1-Mediated [3H]MPP Transport

In Chapter 3 we noted that PYR is the highest affinity inhibitor of all OCs included in our test set (Table 2). Figure 20A shows that PYR is a strong inhibitor of hMATE1 with an IC50 of

~40 nM. To assess the kinetic basis of this inhibition we determined the kinetics of MATE1-

mediated MPP transport in the absence and presence of 1 µM PYR (Figure 20B). In contrast to the inhibitory profiles produced by metformin and creatinine, PYR showed an inhibitory profile

2 that was distinctly not competitive in that it markedly decreased Jmax (from 3.02 pmol/(cm x min)

2 to 0.25 pmol/(cm x min), an 11-fold difference; Figure 20B) while increasing the Kt value (by

2.5-fold, from 4.89 µM to 11.1 µM).

The striking disparity in apparent inhibitory mechanism produced by PYR, compared to metformin and creatinine, led us to examine its basis in more detail. Of particular concern was the fact that PYR is a very weak base, with a pKa of ~7.3. Consequently, in our experiments, conducted at pH 8.5, only about 6% of the PYR in solution can be expected to be charged. Thus there was the possibility that the uncharged fraction of PYR in the external solution could diffuse across the membrane and interact with MATE1 at its cytoplasmic face. Arndt and colleagues made observations of this phenomenon with OCT2 and the weak base, quinine (5). Quinine’s pKa is 8.4 and at physiological pH about 10% of it is uncharged and therefore could move relatively freely across the plasma membrane. Arndt et al showed that by increasing the extracellular pH they also increased passive diffusion of quinine which lead to a decrease in IC50 values in Xenopus oocytes injected with OCT2 (5). They hypothesized that quinine is a non- transported inhibitor of OCT2 and that the uncharged species of quinine diffuses into the cell and inhibits uptake by binding at the intracellular face of OCT2, thereby producing a quinine/OCT2 complex that does not turnover. Moreover, the inhibitory interaction at the cytoplasmic face cannot be displaced by increasing extracellular concentrations of a competing ligand, thereby explaining the 'non-competitive' inhibitory profile of quinine's inhibition of OCT2-medited TEA uptake (5). This may also be the case for PYR at extracellular pH 8.5. Running the experiments at a pH of 7.4, however, does not eliminate this issue because even at this pH over half of the

PYR in solution is uncharged; running experiments at markedly lower pH values (which is one of

the approaches that Arndt used) was not an option for us given the inhibitory influence of extracellular H+ (as a competing co-substrate of MATE1; refer to Figure 4).

Our first approach was to test the extent to which diffusion of PYR could influence the transport characteristics of MATE1. We made use of a method first described over 50 years ago by Lacko and Burger in their studies of glucose transport in red blood cells (RBCs; (78)).

Dubbed “competitive exchange diffusion” (CED), the efflux of intracellular molecules (e.g. glucose for RBCs or MPP for CHO cells expressing MATEs) is measured by allowing cells to accumulate a labeled substrate to a steady-state level. The cells are then rinsed of the extracellular solution and exposed to an experimental solution containing either (i) the same concentration of the labeled substrate, or (ii) the same concentration of labeled substrate plus a test compound to determine if the presence of the test compound stimulates efflux of the intracellular compound.

4.6 Using CED to Address Cis- and Trans- Stimulation of hMATE1-Mediated [3H]MPP

Transport

We established the ‘proof-of-concept’ of the CED protocol for MATE-mediated transport by assessing the influence of extracellular unlabeled MPP on the efflux of preloaded [3H]MPP.

Figure 21 shows a series of experiments that established the time and concentration dependence of MPP transport using the CED protocol. Figure 21A shows that steady-state accumulation of

[3H]MPP (~13 nM) into hMATE1-expressing cells was achieved within approximately 30 minutes. In Figure 21B, cells were allowed to accumulate [3H]MPP to steady-state (30 min), at which point the extracellular medium was replaced with either (i) 13 nM [3H]MPP (identical to the loading solution); or (ii) one containing 13 nM [3H]MPP plus 100 µM unlabeled MPP. When the solution contained only [3H]MPP, cell content of labeled MPP remained unchanged, reflecting maintenance of the steady-state condition. However, the presence of 100 µM unlabeled

MPP in the extracellular solution resulted in a net loss of labeled MPP from the cells. This was expected because MATEs can exchange an OC for an OC in addition to exchange with H+ and therefore the unlabeled MPP interacted with the extracellular binding surface of hMATE1 to stimulate exchange for the intracellular [3H]MPP, thus displaying a trans- effect and re- establishing the fact that MPP is a substrate for hMATE1. Interestingly, the cell did not achieve

“0” [3H]MPP content because once the unlabeled MPP entered the cell it competed with

[3H]MPP, on the intracellular binding surface, creating a pseudo ‘efflux steady state’, which slowed down the measured rate of efflux. Consequently, the CED protocol cannot be used to study the kinetics of efflux due to the ambiguity created by the intracellular mixture of unlabeled and labeled MPP. The emphasis here, however, is that the CED protocol is a very powerful tool for determining if a compound is a transported substrate of membrane transport protein such as, hMATE1.

To further explore the practical applications of the CED protocol, we ran a concentration series with extracellular MPP (Figure 21C). At increasing concentrations of MPP the cell content of [3H]MPP significantly (p < 0.0001) decreased beginning at 10 µM. There is a slight decrease between 30 µM and 1000 µM; however, it is negligible, showing that at about 100 µM the transporter is unaffected by adding more MPP. In other words, increasing extracellular MPP beyond ~100 µM was incapable of stimulating more efflux, due to the maximal occupation of the binding surface of hMATE1 by MPP. Maximal turnover by MPP is achieved between 100-1000

µM.

The data in Figure 21 confirmed the effectiveness of the CED protocol to establish if an extracellular compound can effectively 'turnover' hMATE1, necessitating that the extracellular compound itself be transported. We then used the CED protocol to test whether PYR is a substrate of hMATE1 and if, like quinine, it exerts an inhibitory effect on MATE1 activity at the

cytoplasmic face of the transporter. Figure 22 shows the effect of increasing concentrations of extracellular PYR on the efflux of [3H]MPP from hMATE1-expressing cells. Similar to MPP, increasing concentration of PYR had a significant trans-stimulatory effect on efflux of intracellular [3H]MPP, thereby supporting the conclusion that PYR is a substrate for, as well as an inhibitor of, hMATE1. The concentration dependency of this effect was consistent with the cis- inhibitory profile, as well, with no effect seen at 1 nM, and a maximal effect seen between 100 nM and 1 µM. Of particular relevance to the issue at hand, however, was the observation that increasing the extracellular concentration of PYR to 10 µM which would have resulted in an 100- fold increase of the potentially diffusible, non-charged, species compared to the concentration at

100 nM, had no inhibitory effect on efflux of [3H]MPP. The latter would have been expected if intracellular PYR exerted a high-affinity, inhibitory effect reflecting formation of a non- transported complex after diffusing into the cell. These data suggest that, if PYR is diffusing into cells, it has little or no impact on the accessibility of intracellular MPP to the cytoplasmic face of

MATE1, inferring that the failure of PYR to exert a classical competitive inhibition of MATE1- mediated MPP transport did not reflect the type of non-ionic diffusion-driven effect observed by

Arndt and colleagues with quinine and OCT2. There is evidence here to suggest that there is a more complex mechanism occurring.

Figure 20A argues against the view that PYR exhibits a classical competitive profile of

MATE1 inhibition, and the evidence in Figure 22 suggests that this effect cannot be dismissed as merely being the result of non-ionic-diffusion and a subsequent inhibitory interaction of PYR at the cytoplasmic face of MATE1. Instead it argues that the effect reflects an inhibitory interaction of PYR at the external, or cis, face of the protein. We elected, therefore to test if the PYR-induced inhibition reflects a rapid pure cis interaction with MATE1. The [3H]MPP uptakes used to assess the kinetics of transport shown in Figure 20A were based on 5 minute incubations (which

provided linear rates of uptake and sufficient levels of substrate accumulation to facilitate accurate kinetic analysis). However, we wanted to elucidate what occurs at earlier time points. If

PYR inhibition of MATE1 activity reflected, and was restricted to, a rapid interaction with a site at the extracellular (cis) face of the transporter, then the fractional inhibition of MATE1 activity noted at five minutes might be expected to be evident at earlier time points. On the other hand, if the inhibition at five minutes was due, at least in part, to the influence of intracellular PYR, or to a slower interaction with one more (possibly allosteric) extracellular sites, then inhibition of MPP uptake at earlier time points (e.g., before intracellular PYR accumulation that involved nonionic diffusion), would be less than that at five minutes. Figure 23 shows the time course of 13nM

[3H]MPP uptake, from 15 seconds through 5 minutes, in the absence and presence of either 100 nM PYR or 1 mM MPP. As expected, 1 mM MPP exerted a large, robust inhibition at every time point. Unexpectedly, however, PYR exerted only a minimal amount of inhibition at early time points (15-60 seconds), whereas at 5 minutes we saw a large, robust inhibition of transport.

These data suggest that the inhibition of MATE1-mediated MPP uptake noted at extended time points is not restricted to a simple, rapid, interaction at a site shared with MPP. Of course, the inhibitory profile seen in Figure 20 implied that such an interaction does not occur. The exact mechanism for PYR interaction with MATE1 is unclear; however, the evidence presented here is consistent with a complex kinetic interaction with inhibitory ligands.

4.7 Significance of PYR Structural Analogs: PYR-2 & PYR-3

To further analyze PYR’s interaction with hMATE1 we chose to study two PYR structural analogs, PYR-2, PYR-3. Both molecules resemble PYR structurally, and are high

3 affinity inhibitors of hMATE1–mediated transport of [ H]MPP (Figure 24). PYR-2 has an IC50 of

140 nM, similar to PYR-3’s IC50 of 196 nM (Table 2). Not surprisingly, PYR-2 and PYR-3 are the second and third highest inhibitors of hMATE1 most likely due to the fact that they display

several key structural features of PYR. However, of relevance to the present issue, both of these compounds are much more basic than PYR with pKa values of 9.2 (PYR-2) and 9.1 (PYR-3).

The analogs thus provide a tool for focusing on steric effects of PYR without the weak base effect that complicates interpretation of the PYR data. Therefore, we decided to probe PYR-2 and

PYR-3’s cis interactions with the protein to determine if their more basic properties display a different effect on hMATE1 than PYR.

Figure 25A shows uptake of [3H]MPP in the presence and absence of two concentrations of MPP: 1 mM (to block all MATE1-mediated transport) and 20 µM (a concentration about five- times over the Kt for MATE1-mediated MPP uptake. Also shown is the effect at both time points of 250 µM metformin (about five-times over the Kt for metformin transport), 200 nM PYR (about four-times over the PYR IC50), and 400 nM PYR-2 and PYR-3 (~three and two-times their respective IC50 values). As expected, MPP and metformin both displayed very similar inhibitory profiles at both time points. Interestingly, PYR-2 and PYR-3 also displayed essentially the same patterns of cis inhibition as MPP and metformin at both time points, therefore showing that they are effectively restricted to the interaction at the cytoplasmic face of hMATE1. Conversely, PYR showed very little inhibition at 30 seconds, while again showing effects at 5 minutes. To highlight the inhibitory interaction of each compound, we normalized each one to the percent inhibition of 13 nM [3H]MPP uptake, at 30 seconds versus 5 minutes (Figure 25B). Whereas the degree of inhibition produced by MPP, metformin, PYR-2 and PYR-3 at the two time points did not differ (P > 0.05), PYR inhibition at 30 sec was significantly less (P < 0.05) than at 5 min.

Lastly, we chose to further interrogate the kinetic basis of the PYR analogs’ interaction with hMATE1. For the purpose of this chapter, PYR-2 and PYR-3 were essentially the same so we chose one of the analogs (PYR-2) as a representative for further experiments. The aim was to determine if PYR-2 is a competitive inhibitor, like metformin and creatinine or if it displays a

more complex interaction, like PYR. Figure 26 shows MPP kinetics in the absence and presence of 400 nM PYR-2. Like PYR, the inhibitory profile produced by PYR-2 was not competitive

(Figure 26); in addition to producing a significant (P < 0.05) increase in Kt (from 5.0 to 9.2 µM), the presence of PYR-2 resulted in a significant (P < 0.005) decrease in Jmax (from 1.44 to 0.72 pmol/(cm2 x min)).

4.8 Conclusions & Discussion

The principal purpose of this Chapter was to further study the complex binding surface of hMATE1. We chose five compounds and studied their kinetic profiles of inhibition and used the

CED protocol to establish the influence of trans-effects on the interaction of transported substrates with the cis- and trans- aspects of hMATE1. The data show that metformin and creatinine are classical competitive inhibitors of hMATE1-mediated transport of MPP, while

PYR, PYR-2 and (by extension) PYR-3 appear not to be. Furthermore, all of the compounds tested appear to be transported substrates of MATE1. Our metformin data confirm previous reports that it is a transported substrate of hMATE1, but again there are a limited number of sources that report kinetic values for MATE-mediated transport. Previously, Kt values of 238 µM and 253 µM have been reported for hMATE1-mediated metformin transport, which are higher than our 64.3 µM (Figure 17B; (76; 94)). However, as noted before experimental conditions and cell lines may play a role in these differences. Moreover, if our experiments were run at physiological pH the predicted Kt value would be approximately 224 µM, strikingly similar to those reported in the literature. The IC50 reported here for the MPP inhibition by metformin (47.0

µM; Figure 18A) was very similar to the Kt value for metformin kinetics, as predicted if MPP and metformin interact competitively at a common, mutually exclusive binding site. These data were further supported in Figure 18B where MPP kinetics were run in the absence and presence of 250

µM Metformin and showed that presence of metformin had no effect on Jmax while producing an

increase in Kt. Taken together, these data support the conclusion that metformin is a competitive inhibitor of MPP interaction on MATE1’s binding surface.

The other small basic OC tested in this chapter was creatinine. Creatinine is not a high affinity inhibitor of MATE1 (IC50 = 195 µM at pH 8.5), but data presented here shows that it, too, displays a competitive interaction on hMATE1 mediated transport of MPP (no change in Jmax with a change in Kt; Figure 19). This is of clinical significance because creatinine is used as a measure of glomerular filtration rate (GFR), and these findings confirm that doing so may be an overestimation due to it being secreted in the RPT by the MATEs. Furthermore, high affinity

MATE substrates could block creatinine secretion by the MATEs and speciously reduce measured GFR.

PYR stood out as the first non-competitive inhibitor of the group as well as the highest inhibitory substrate of hMATE1. Figure 20 shows that PYR has an IC50 of 40 nM and that the presence of 1 µM PYR, markedly decreased Jmax while having little effect on Kt (from 4.89 µM to

11.1 µM, respectively), which is not consistent with competitive inhibition. Others have reported

Ki values of 77 and 93 nM for PYR inhibition of TEA and metformin, respectively (61; 76), which are very similar to our 40 nM IC50 value. Kusuhara et al reported finding that patients taking a microdose (100 µg) or a therapeutic dose (250 mg) of metformin in conjunction with 50 mg PYR saw a 23 and 35% decrease in metformin’s renal clearance, implying that PYR has an effective inhibitory interaction with the OC secretory pathway (76). Ito et al, showed that in mice, PYR inhibits the efflux step of TEA and metformin secretion in the kidney (i.e. mMate1;

(61)). They further tested their findings by using HEK cell transfected with mMate1 or hMATE1 and found that PYR is a potent inhibitor of TEA uptake (Ki = 145 nM and 77 nM, respectively;

(61)). To further validate their findings they showed that metformin uptake was greatly inhibited by PYR (Ki = 31 nM) in human BBMVs (61). Taken together, these data and those presented in

this chapter show that PYR is a potent inhibitor of MPP, TEA and metformin uptake by MATE1- expressing cells and that its effects can be seen at a systemic level in humans and mice. PYR’s mode of action however remains unclear and therefore we probed its interactions with more experiments.

As noted above, we found PYR did not behave as a competitive inhibitor of hMATE1- mediated MPP transport. Interestingly, however, Kusuhara et al, showed that PYR is a competitive inhibitor of metformin transport in hMATE1-expressing cells, which appears to be in contrast to our findings. Again, it is worth first comparing the experimental conditions used in these two studies. We ran our experiments at RTº with extracellular pH 8.5, whereas Kusuhara et al, performed their studies at 37º C with an extracellular pH of 7.4. As commented on in Chapter

3, we think it unlikely that the difference lies in the choice of CHO vs. HEK-293 cells; we have run side-by-side studies on the kinetics of MPP transport in these two cells lines using our experimental conditions and found them to be virtually identical (unpublished observations). The use of different incubation temperatures may well have influenced the rate of transport

(Kusuhara’s rates of transport were substantially higher than those in our studies), but it is unlikely to have influenced that mechanism of ligand binding. The different pH values are, however, particularly relevant. As noted earlier, the pKa for pyrimethamine is about 7.3 and, so, the concentration of unprotonated PYR in our test solutions would have been twice that in

Kusuhara’s studies, which is the reasonable basis for predicting that passive diffusion of PYR into the cells could have set up a situation in our studies that permitted interaction of PYR at the cytoplasmic face of hMATE1. Results obtained using CED method (Figure 22) confirmed that

PYR, like MPP, is a transported substrate of hMATE1, which was expected due to the findings of

Kusuhara et al and Ito et al (61; 76) in in vivo systems, where it was present in the urine. But the concentration dependency of PYR’s induced CED was consistent with its cis-inhibitory profile

(no effect seen at 1 nM, and a maximal effect seen between 100 nM and 1 µM), and argued against the influence of nonionic diffusion of cytoplasmic PYR as the basis for its inhibition of

MATE-mediated transport. Nevertheless, the failure of cis PYR to inhibit MPP uptake at 30 sec to the same (or similar) extent as that noted at 5 min (Figure 23) indicated that PYR does not exert a ‘classical’ competitive interaction with MATE1-mediated MPP transport.

PYR-2 and PYR-3, with their pKa values of >9, offered the opportunity to assess the influence of PYR structure on the kinetic interaction with hMATE1 without the interpretive complications introduced from PYR’s low pKa. As expected, both PYR analogs were high affinity inhibitors of hMATE1-mediated MPP transport (Figure 24), but unlike PYR, that strong inhibitory interaction was effectively as strong at 30 sec as it was at 5 min (Figure 25).

Nevertheless, when MPP kinetics were determined in the absence and presence of 400 nM PYR-2 the profile was not competitive in nature, inducing both a decrease in Jmax, as well as an increase in apparent Kt. These data suggest that, whereas some inhibitory interactions of ligands on the binding surface of hMATE1, including those associated with the inhibition of MPP transport produced by metformin and creatinine, reflect competition between ligands for a common site (or a set of mutually exclusive sites), other interactions, including those with PYR-like structures, involve more kinetically complex mechanisms.

The data in hand are consistent with the presence in hMATE1 of a binding surface that includes multiple binding ‘sites.’ Some substrates interact preferentially with sites sufficiently close together within this surface that the presence of one (e.g., metformin) precludes effective interaction with another (e.g., MPP). These ligands would, consequently, display a competitive profile of inhibition (Figure 27A) in which sufficiently large concentrations of either would eliminate interaction with the other, the result being the classical ‘increase in apparent Kt with no change in Jmax.’ But in this model, other substrates (e.g., PYR and its homologs) interact at a site

that is spatially distinct from other sites (e.g., that for MPP). The result is an inhibitory profile described as ‘linear mixed-type inhibition” (Figure 27B), in which multiple substrates can interact, but when two are bound simultaneously, no transport occurs. Because neither substrate can prevent binding of the other, the kinetic consequence is a decrease in Jmax; apparent Kt can also change, if the presence of one substrate exerts an allosteric influence on binding of the other.

There are several attractive features of this model. First it aids in the understanding of the data presented here, where Jmax changes with little change in Kt. Furthermore, it provides a potential explanation for the Kusuhara et al findings, that whereas MPP and PYR occupy separate sites, it is possible that PYR and metformin occupy mutually exclusive sites. With regards to

MATE pharmacophore development, this model clarifies why an ‘all inclusive’ model is not possible and it is likely because substrate binding can involve multiple regions and interaction with a single transported probe (e.g. [3H]MPP) may not be indicative of the full range of ligand interaction with the transporter. Fourth, it can be argued that it is reasonable and expected that a xenobiotic transporter be capable of interacting effectively (i.e., with high affinity) with a structurally diverse suite of substrates, which would require more than just a binding ‘site’ but a binding ‘region’ or ‘surface’. This is the case for transport proteins such as P-glycoprotein (P- gP), which interacts with an extraordinarily diverse group of ligands, including cations and anions. In fact, recently P-gP’s crystal structure was resolved and showed that there are a multitude of different residues facing the putative drug-binding pocket, which is suggestive of a polyspecific drug recognition region (4). In other words, P-gP requires a polyspecific binding surface in order to interact with so many diverse ligands. To a lesser degree this may also be the case in MATE proteins.

In conclusion, we found that, whereas, metformin and creatinine are competitive inhibitors of MPP on hMATE1’s binding surface, PYR and its structural analogs, PYR-2 and

PYR-3, though appearing to exert their effects on MATE1-mediated transport from the external, cis, aspect of the protein, do not inhibit via simple competition for a common binding site. We suggest that these data are consistent with the presence in hMATE1 of a binding surface with multiple, non-overlapping binding sites that can display different kinetic interactions with structurally distinct substrates.

CHAPTER 5

APPLICATION OF THE COMPETITIVE EXHANGE DIFFUSION PROTOCOL TO DETERMINE SUBSTRATES AND INHIBITORS OF HUMAN MATE2-K

5.1 Background

In the kinetic analysis of MATE transport (and other transport proteins), if a compound

(e.g. [3H]MPP) is taken into a cell expressing MATE (and not into control cells that do not express the protein), then it is considered to be a MATE substrate. It is also frequently assumed

(tacitly, at least) that if uptake of a compound is inhibited by the presence of a second compound, then that second compound is likely to be a MATE substrate; however this is not always an accurate assumption. The danger in this approach is that uptake of an OC can be inhibited by another substrate or by a non-transported (e.g., allosteric) inhibitor of MATE. Researchers make this assumption (that inhibitors are also substrates) because there are many practical limitations associated with determining which compounds are substrates versus inhibitors. Ideally, one would have their compound of interest in radiolabeled (or fluorescently labeled) form to measure uptake directly. However few OCs are available commercially in radiolabeled form (or fluorescently labeled form) and those that are available can be expensive. Commercial synthesis of a radiolabeled substrate is generally prohibitively expensive. Of course other methods exist to measure uptake, such as HPLC or mass spectroscopy (MS), but they, too, can be expensive and time consuming, particularly if one is interested in multiple compounds. As noted in the previous chapter, however, determining which compounds are MATE substrates can have a substantial impact on pharmacophore development and in uncovering the kinetic mechanisms by which

MATE transport works (i.e. competitive inhibition versus non-competitive and/or mixed-type

inhibition). Therefore, finding a method to address this issue is central to advancing our knowledge of the physiological mechanism by which MATE proteins work.

50 years ago, Lacko and Burger, in their studies of glucose transport in red blood cells

(RBCs; (78)) created a technique called “competitive exchange diffusion” (CED), wherein the efflux of intracellular molecules (in their case, glucose for RBCs; in our case, for example, MPP for CHO cells expressing MATEs) is used as a surrogate for transport of a second, test, compound. In this method, cells are allowed to accumulate a labeled substrate to steady-state level (i.e. maximum intracellular concentration). The cells are then rinsed of the extracellular solution and exposed to an experimental solution containing either (i) the same concentration of the labeled substrate, or (ii) the same concentration of labeled substrate plus a test compound to determine if the presence of the test compound stimulates efflux of the intracellular compound.

The CED protocol, outlined in the previous chapter, can be used to address the issue of whether an inhibitor of transport can also serve as a substrate for the transporter. CED takes advantage of the fact that the MATEs are obligatory exchangers and that trans-stimulation of intracellular OCs, such as [3H]MPP, can be used to determine which compounds are transported (i.e. substrates) and which ones simply inhibit MATEs from turning over and/or binding another OC (i.e. non- transported inhibitors). It is important, however, to make the distinction between measuring efflux using the CED protocol and erroneously using the CED method to measure efflux kinetic parameters (such as Kt or Jmax). CED cannot be used to define kinetic parameters of efflux because of the uncertainty created by the mixture of OCs coming in from the extracellular space with the intracellular labeled OC, which may compete for the intracellular binding surface of

MATE. Nevertheless, as shown here, CED provides a valuable means to identify substrates of

MATE transporters.

This Chapter focuses on using the CED technique to study 14 compounds in more depth, using hMATE2-K expressing CHO cells. MATE2-K was chosen, because it displayed a higher signal-to-noise ratio for trans-stimulation than MATE1 (data not shown). In Chapter 3 I noted that MATE1 and MATE2-K have overlapping selectivites for a multitude of compounds and therefore, it can be inferred that their kinetic mechanisms are similar, hence substrates of

MATE2-K are likely MATE1 substrates.

5.2 Using CED to Assess Various Compounds as MATE2-K Substrates or Inhibitors

Of the 59 compounds screened in Chapter 3, I chose 14 compounds of various affinities, shapes and molecular descriptors. As part of the group of compounds tested I chose to use a negative control, mannitol, in order to ensure the validity of using the CED model to test for substrates versus inhibitors. Figure 28, shows the cellular content of [3H]MPP at time “0” (e.g. after [3H]MPP accumulation has reached steady-state, and before adding an extracellular compound), control (buffer at pH 8.5), control-mannitol (control buffer with 20 mM mannitol) and MPP. As expected mannitol did not trans-stimulate efflux of the intracellular [3H]MPP; only

MPP was able to stimulate efflux significantly (p < 0.05). Furthermore, this experiment showed that the presence of extracellular compounds up to an external concentration of 20 mM did not produce an osmotic effect (e.g. through generation of an outwardly-directed gradient of [3H]MPP following cell shrinkage). In the following experiment near-saturating concentrations of each ligand were used (calculated at five-times their IC50 values for MATE2-K) to ensure equivalent

'occupancy' of the transporter with potential substrate. In order to normalize the osmotic effect, mannitol was added to all of the solutions to give osmolar concentration of 20 mOsM over

Weymouth buffer (mannitol plus the compound of interest; control included 20 mM mannitol only).

5.3 Known MATE Substrates

Figure 29 shows that metformin, paraquat, creatinine and PYR are transported substrates of MATEs, confirming direct demonstrations (20; 69; 76; 89; 105; 136; 150), including some of our own measurements of transport (Figure 17). Again, the goal here was to display the ability of the CED method to show that a compound known to inhibit MATE-mediated transport of

[3H]MPP can also support turnover of the transporter. Interestingly, although PYR and creatinine stimulated a significant (p < 0.0001) efflux of [3H]MPP, they did not produce the same degree of efflux that MPP, metformin and paraquat do, suggesting that while PYR and creatinine are transported substrates of MATE1, the turnover number of MATE2-K when occupied by PYR or creatinine is lower than the turnover number of the transporter when occupied with MPP, metformin or paraquat. Of course this could be a result of linear mixed-type inhibition, as was chronicled in the previous chapter, where the simultaneous presence of two substrates at different locations of the binding decreases transport turnover.

5.3.1 N-Tetraalkylammonium Compounds as MATE Substrates

In the following experiment (Figure 30) I chose to test some of the ‘n-tetraalkyls’ probed in Chapter 3. Many have shown that tetraethylammonium (TEA) is a transported substrate of both MATEs (69; 89; 105; 107; 136); therefore, it was used to confirm those findings and to validate the use of the CED protocol. Tetramethylammonium (TMA) and tetrapentylammonium

(TPeA) have been less extensively probed. However, a study using BBMV from rabbits showed that TMA and TPeA can trans-stimulate TEA transport via the OC/H+ exchanger of native renal membranes (165), thus suggesting that they are substrates for MATE transporters. Interestingly, while they were both able to stimulate efflux of [3H]TEA from BBMV, TMA did so at a much higher rate than TPeA (165). This study was done before the molecular cloning of the MATEs in

2005. Therefore, I wanted to confirm these findings and solidify the fact that the MATEs are the

long sought after OC/H+ exchanger. As expected, all three compounds stimulated efflux significantly and showed cellular content levels similar to those seen with MPP (Figure 30; p <

0.0001), thus confirming that TEA, TMA and TPeA are MATE substrates.

There are no reported IC50 values for hMATE2-K with TMA or TPeA. However, what we know about them with hMATE1 (Table 2) suggests that, whereas TMA will have a weak interaction with MATE2-K, TPeA will be a comparatively strong inhibitor of hMATE2-K (22;

167). Experiments to elucidate more about their kinetic parameters with hMATE2-K would be helpful in analyzing their interactions with MATE2-K.

5.3.2 Unknown MATE Substrates

Lastly, I chose to use CED to test the extent to which a few novel MATE inhibitors are capable of supporting transporter turnover. These test compounds included two that were very strong inhibitors of MATE2-K (azido-procainamide (APMI), and quinidine); two that were quite weak inhibitors (ethohexadiol and caffeine); and the prototypical substrate of the renal organic anion secretory pathway, p-aminohippurate (PAH). Interestingly, as shown in Figure 31, all of these compounds proved to be transported substrates of hMATE2-K, capable of trans-stimulating significant levels of intracellular [3H]MPP (Figure 31; p <0.0001). APMI is one of the highest affinity substrates of MATE2-K (IC50 = 0.501 µM; Table 2) and therefore I expected that it would interact effectively with MATE2-K to stimulate turnover. Quinidine, was somewhat of an unexpected substrate, because it is an enantiomer of quinine, which has previously been shown to diffuse across membranes to cause inhibition at the intracellular face of OCT1, via non-ionic diffusion (5). That does not appear to be the case with MATE2-K; these data suggest that it supports robust turnover of MATE2-K. The lowest amount of trans-stimulation was seen with caffeine, PAH and ethohexadiol. Caffeine and ethohexadiol clearly interact weakly with MATE transporters, with IC50 values of 451 µM and >1000 µM (respectively, for MATE2-K). However,

so too does TMA (IC50 of 5 mM for MATE1), yet it supported turnover of the transporter at least as well as did MPP; Figure 30). The data suggest that the turnover number of the transporter fully occupied with effectively saturating concentration of either caffeine or ethohexadiol was much lower than when occupied with MPP, metformin, or other ‘good’ substrates (including

TMA). It is interesting to note that ethohexadiol is not a cation and caffeine has a pKa of 0.10 so, consequently, at pH 8.5 virtually none of it is charged. Moreover, PAH is negatively charged at pH 8.5. As noted in Chapter 3, the pharamcophore developed for ligand binding to MATE1 included a cationic feature. Thus it is predictable that all three of these compounds would interact weakly, if at all with MATE1. The data suggest that at sufficiently high concentrations, the transporter can be ‘occupied’ and can mediate flux of compounds not viewed as traditional OC substrates. However, the striking difference in the ‘turnover’ of the transporter occupied by a true cation, like TMA, compared to caffeine or ethohexadiol, underscores the importance of charge in effecting interaction of ligand with MATE transporters.

It is also interesting to consider the work of Ullrich and colleagues, who suggested some

20 years ago that the renal OC and Organic Anion secretory pathways are each capable of interacting functionally with (at least) selected substrates of the other (the so-called ‘bi-substrates’

(147; 148)). To that end, two MATE genes expressed in barley and sorghum, HvAACT1 and

AltSB, respectively, have been shown to be involved in aluminum tolerance (38; 86), and human

MATE1 and MATE2-K have also been implicated in transporting anions such as estrone sulfate, acyclovir and ganciclovir (137). Therefore it is not unreasonable that PAH could also have interaction, albeit weak, with the MATEs.

PAH is a compound that could benefit from further tests. PAH has been shown to be a substrate of organic anion transporters (OATs, Multidrug Resistant Proteins (MRPs) and

Inorganic Phosphate Transporters (Npts)) due in part to its pronounced negative charge at

physiological pH (7; 58; 60; 65; 80; 83; 84; 114; 125; 153). However, it has also been shown to be a substrate of OCT2 (5); there are no reports of it interacting with MATEs. Members of the

MATEs do appear to be capable of interacting with some anions (38; 86; 137) and zwitterions

(11; 27; 41; 50; 51; 66; 82; 96; 104; 108; 170), so PAH may be a viable substrate for MATE2-K.

However, more experiments should be done.

5.4. Discussion

In conclusion, 14 substrates (seven novel: APMI, quinidine, caffeine, PAH, ethohexadiol, TMA and TPeA), of hMATE2-K were identified using the CED method. The CED protocol offers a way to design high throughput studies that combine assessment of inhibition and assessment of whether test agents serve as substrates. This approach offers pharmaceutical companies an opportunity to identify pathways for potential DDIs, as well as pathways for renal secretion of organic cations and other drugs. CED also offers the opportunity to test the effect of chemical structure on actual transport, rather than just ‘ligand binding’, which can have significant effects on pharmacophore development and kinetic measurements.

CHAPTER 6

CONCLUSIONS & FUTURE DIRECTIONS

6.1 Conclusions

MATEs are the major candidates for apical OC/H+ exchange in the RPT, which also makes them the active and rate limiting step for renal OC secretion. Since the discovery of

MATEs in 2005, MATE data has steadily increased. The emphasis of studies on MATE transporters has been largely directed toward (i) securing their place as key players in mediating renal (and hepatic) OC secretion; and (ii) establishing their role in the transport of selected classes of drugs. Lacking in this approach has been work to establish the selectivity and kinetic characteristics of transport that may lead to development of predictive models of drug interaction with these proteins. In this study, I generated the first database of considerable size for ligand interactions of both MATEs. These data led to several iterations of computational modeling for hMATE1 pharmacophores. Not only did the models identify key structural features strongly correlated with ligand binding to hMATE1, but I was also able to identify 11 new structural drug classes that are likely to interact with MATE transporters. Moreover, the observations associated with the generation of distinct pharamcophores is consistent with the contention that hMATE1 may have a complex binding surface for ligand interaction, rather than a single binding site.

After I probed the kinetic parameters of MATE interaction with a host of OCs I used the

CED method developed by Lacko & Burger (78), to establish that, whereas, metformin and creatinine are competitive inhibitors, PYR and its structural analogs, PYR-2 and PYR-3, do not inhibit via simple competition for a common binding site on hMATE1. Again, these data were consistent with the idea that hMATE1 has a binding surface with multiple, non-overlapping binding sites that can display different kinetic interactions with structurally distinct substrates.

In Chapter 5, I used the CED model to test 14 OCs and determine if they were substrates of hMATE2-K. I determined that five inhibitory ligands are also novel substrates of hMATE2-K:

APMI, quinidine, caffeine, PAH, ethohexadiol, TMA and TPeA. Moreover, I validated the use of the CED protocol as a high throughput assessment of substrate versus inhibitors of MATEs.

6.2 Future Directions

Most studies on MATEs have focused on single classes of drugs or OCs, providing only limited information about the molecular characteristics of binding to MATEs. Quantitative kinetic data are needed in the field in order to accurately assess the physiological functions of the

MATEs. Furthermore, the limited kinetic data available for the MATEs is invariably directed at measurement of uptake rather than the more physiologically relevant efflux of OCs. Obvious limitations exist in measuring the kinetic parameters of efflux. However, this is of critical importance to the advancement of the field. The results of my work suggest at least three new sets of studies that could substantially advance the understanding of the role of MATEs in renal physiology:

i. More structurally diverse groups of OCs should be tested to address the

hypothesis of ‘linear-mixed type inhibition’ and the implication that it might

have on DDIs and renal secretion of OCs. Expanding the list of test compounds

to include new drug classes and novel molecules will provide more data that can

be used to further refine future pharmacophores.

ii. Studies done with different probes, i.e. ASP or PYR instead of MPP, could

provide more information about different binding sites on the MATE binding

surface, as well as knowledge about potential DDIs and ligand interactions.

iii. Also, further analysis of which OCs are substrates (versus inhibitors) will give

insight into the idea of multiple binding sites, by elucidating which compounds

are competitive inhibitors of each other and which appear to be exhibiting linear

mixed type inhibition. Pharamcophore development of a MATE1 or MATE2-K

model will benefit from understanding MATE interaction with different

substrates and also from clarifying which OCs are substrates versus inhibitors.

There are multiple facets to be explored in light of this study and while I believe it contributes to the advancement of the field I also believe that this study created various interesting questions and ideas that need further testing and analysis.

APPENDIX A: FIGURES

Figure 1 Cellular Model of Organic Cation Secretion: OC secretion at the cellular level, as discussed in text (revised from the American Physiological Society’s Comprehensive Physiology:

Renal Handbook (121)). Abbreviations: MATE, Multdrug And Extrusion; NKA, sodium- potassium ATPase; OC+, Organic Cation; OCT, Organic Cation Transporter; OCTN, Organic

Cation/Carnitine Transporter; P-gP, P-Glycoprotein.

Figure 2 hMATE1 Secondary Structure: The depicted membrane topology of hMATE1 generated from http://www.sacs.ucsf.edu/TOPO2/. hMATE1 has 13 predicted transmembrane domains (TMDs), with a long intracellular loop between TMD 12 and 13. Highlighted residues are C63, C127 (orange), G273, G278, G300, G300 (green) and H 385 (blue).

Figure 3 hMATE1 Homology Model: The side view of the model with the extracellular face at the top and the cytoplasm at the bottom. This model used NorM (a member of the MATE family) as a template. Figure adapted from the American Physiological Society’s Comprehensive

Physiology: Renal Handbook (121).

A. B. 10 10 ) -2 -1 J = 1.86 fmol cm -2 min-1 Jmax = 2.14 pmol cm min max -1 8 K = 3.72 M 8 Kt = 4.37 M t

min 6

-2 6

4 hMATE1 4 hMATE2-K

H]MPP Uptake 2

3 2

[ (fmol cm

0 0 0 1 10 100 1000 0 1 10 100 1000 [MPP] M [MPP] M

10 D.

6 C. )

-1 8 IC50 = 3.5 nM IC50 = 19.6 nM (pH 8.5) (pH 7.7)

4 6

x min -2 4

2 1 mM MPP hMATE2-K 1 mM MPP

H]MPP Uptake hMATE1 2

[ (fmol x cm (fmol 0 0 0.1 1 10 100 1000 0.1 1 10 100 1000 [H+] nM [H+] nM

Figure 4 hMATE1 & hMATE2-K MPP Kinetics & H+ Inhibition Kinetics: MPP Kinetics in hMATE1 (A) and hMATE2-K expressing CHO cells. [3H]MPP uptake was measured in the presence of increasing concentrations of unlabeled MPP. Inhibition of [3H]MPP kinetics with increasing concentrations of H+ are also displayed in Figures C (hMATE1) and D (hMATE2-K).

Each point is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº). Jmax, Kt and IC50 values were calculated as an average of 3 experiments.

A. B.

100 hMATE1 100 hMATE2-K

75 75

50 50

25 25 % Inhibition %

H]MPPUptake 0 0

3 [

-25 -25 Compounds Compounds

Figure 5 Range of Inhibition of hMATE1 & hMATE2-K Mediated Transport: The percent inhibition of [3H]MPP at 10 µM of all 59 compounds screened. The profiles of hMATE1 (A) versus hMATE2-K (B) are similar with 20 (MATE1, A) or 14 (MATE2-K, B) compounds reducing [3H]MPP transport by ≥50%.

A. B.

hMATE1 IC = 53.9 M hMATE1 IC50 = 1.57 M 50 120 120 hMATE2-K IC = 60.8 M hMATE2-K IC50 = 1.48 M 50 100 100

80 80

60 OH 60 N

H-MPP Uptake H2N 3 40 N 40 NH2 HN NH 20 O 20

0 0 Percent 0 0.01 0.1 1 10 100 1000 0 1 10 100 1000 [Quinidine] M [Agmatine] M

C. D.

hMATE1 IC50 = 212 M 120 hMATE2-K IC50 = 236 M 120

100 100

80 N 80 H N N O hMATE1: NI 60 HN 60 N H-MPP Uptake NH N hMATE2-K: NI 3 40 40 O NH 20 20 OH

0 0 Percent 0 1 10 100 1000 0 10 31.6 100 316.2 1000 [Nialamide] M [Allopurinol] M

Figure 6 Inhibitory Profiles of Transport Activity of hMATE1 and hMATE2-K: The inhibitory profiles against transport activity of hMATE1 (closed circles) and hMATE2-K (open circles) produced by four compounds: quinidine (A), agmatine (B), nialamide (C) and allopurinol

(D). These compounds were selected to emphasize the spectrum of inhibitory effectiveness of our battery of test agents, with IC50 values that ranged from micromolar (e.g., quinidine), through micromolar (e.g., nialamide), to no effective interaction (e.g., allopurinol). Each point is the percent of control (e.g. [0] of compound) measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº). IC50 values were calculated as an average of 3 experiments. NI= No interaction.

A. pH 8.5 pH 7.4 10000

10000



( 50 100 100

1 1 hMATE2-K IC

1 100 10000 1 100 10000 hMATE1 IC50 (M) hMATE1 IC50 (M)

B. hMATE1 IC50 = 7.50 M 125 hMATE2-K IC50 = 88.9 M 125 125 hMATE1 IC50 = 6.21 M hMATE1 IC 50 = 5.90 M hMATE2-K IC 50 = 0.50 M hMATE2-K IC 50 = 52.8 M 100 100 100 O N 75 75 NH 75 H N 2 N

NH2 H]MPP Uptake H]MPP 50 3 50 50 N

25 O 25 NH2 25 O HO

Percent [ Percent 0 0 0 0 1 10 100 1000 0 1 10 100 1000 0 1 10 100 1000 [Atropine] M [Amantidine] M [APMI] M

Figure 7 Comparison of hMATE1 & hMATE2-K IC50 Values & Range of Affinities: hMATE1 IC50 values were graphed as a function of hMATE2-K IC50 values for the 59 test compounds (A) at pH 8.5 (left) and calculated 7.4 (right). There is marked overlap between the two proteins with the exception of a few of the compounds some of which are highlighted in B.

The apparent affinities of hMATE1 (closed circles) and hMATE2-K (open circles) for atropine

(left), amantadine (center) and azidoprocainamide (APMI; right) are displayed in panel B. Each point is the percent of control (e.g. [0] of compound) measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº). IC50 values were calculated as an average of 3 experiments.

OAT1 OAT3 OCTN2 OCT2 Outer Cortex Inner Cortex MATE1 Medulla MATE2/2-K 0 20 40 60 80 100 Relative mRNA level (GAPDH= 100)

Figure 8 mRNA Levels of Xenobiotic Transporters in Human Kidneys: Assessment of mRNA levels for Organic Anion Transporter (OAT) 1 and 3, Organic Cation/Carnitine

Transporter (OCTN) 2, Organic Cation Transporter (OCT2) and Multidrug And Toxin Extrusion

(MATE) 1 and 2-K transporters in samples collected from outer cortex (white), inner cortex

(gray) and medulla (black) from two human kidneys. Each point is the mean (±SEM, n of 3) and levels of each are normalized to the level of GAPDH expression in each sample.

A. hMATE1 (40X) B. hMATE2-K (40X) 94

10X 10X

Figure 9 Immunohistochemical Localization of hMATE1 & hMATE2-K in the Kidney:

Immunohistochemical localization of MATE1 (A) and MATE2/2-K (B) in human renal cortex at

40X magnification (insert at 10X magnification). Both proteins were localized to the brush border membrane of proximal tubules (labeled in green; nuclear stain in red).

B. A. 4 4 TMA 3 2 TEA

M) 2

TEMA



( 50

TPrA 0 TBA log log IC 1 TPeA r = 0.332 r = 0.972 p < 0.05 p < 0.001 -2 -3 0 3 6 9 12 -2 0 2 4 6 8 logP logP

C. 4 4 D.

M) 2

 2

( 50

0 log log IC

0 r= 0.021 r = 0.423 p > 0.05 p < 0.01 -2 0 50 100 150 200 -3 0 3 6 9 12 2 TPSA (Å ) pKa

Figure 10 Correlation Between hMATE1 & hMATE2-K IC50 Values & LogP, TPSA & pKa: Relationship between hMATE1 IC50 values with logP (A-B), TPSA (C) and pKa (D). None of these molecular descriptors correlated strongly to MATE1 interaction, except for the relationship between a group of n-tetraalkylammonium compounds and logP.

A. 125

100

H]MPP Uptake 3

[ 25 % 0 0 1 10 100 1000 10000 [X] M

B. B.

Figure 11 Family of Kinetic Curves hMATE1 & Common Features Pharmacophore:

Kinetics for the five compounds used to generate the hMATE1 common features pharmacophore

(A): PYR (open triangles; IC50 = 0.04 µM), quinidine (closed triangles; IC50 = 1.57 µM), histamine (closed diamonds; IC50 = 761 µM), caffeine (closed squares; IC50 = 1096 µM) and chloramphenicol (closed circles; IC50 = 1115 µM). The resulting common features pharmacophore (B) had 2 hydrophobic regions (cyan), 1 H-bond donor (magenta), and 1 H-bond acceptor (green). Each point is the percent of control (e.g. [0] of compound) measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº).

A. B.

) 3 10 -1 IC50= 0.93 M 8

x min 2

6 Weak Intearction -2 N 4 OH 1 HO

OH H]MPP Uptake

3 2

[ (fmol x cm (fmol 0 0 0 1 10 100 1000 0 1 10 100 1000 10000 [Cinchonidine] M [Ethohexadiol] M

C.A. B.D.

Figure 12 hMATE1 Inhibition Kinetics of Cinchonidine and Ethohexadiol & Common

Features Pharamcophores: Cinchonidine inhibition kinetics of [3H]MPP (A) and ethohexadiol

(B). Both molecules (cinchonidine, C and ethohexadiol, D) were good fits for the common features pharmacophore (fit values of 3.1 and 2.7, respectively, data not shown), cinchonidine was an effective, high affinity inhibitor of hMATE1. Each point in A and B is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº). IC50 values were calculated as an average of 3 experiments.

4.00 50

M) 2.00

 (

Log Predicted IC 0.00 r= 0.68 p < 0.0001 0.00 2.00 4.00 Log Experimental IC50 (M)

Figure 13 First Iteration of the Quantitative Pharmacophore for hMATE1 & Correlation

Graph: 24 of the initial 26 compounds were used in an analysis that resulted in a model containing two hydrophobic features (cyan), one hydrogen-bond donor (magenta), and one positive ionizable feature (red) seen with cinchonidine’s structure (A). The graph on the right (B) displays the relationship between measured and predicted IC50 values based on the model, which has a correlation r value of 0.68 (p < 0.0001).

A. B.

4.00 50

M) 2.00

 (

Log Predicted IC 0.00 r= 0.71 p < 0.0001 0.00 2.00 4.00 Log Experimental IC50 (M)

Figure 14 Second Iteration of the Quantitative Pharmacophore for hMATE1 & Correlation

Graph: Using 43 compounds (the initial 24 plus the “test set” of 15 compounds and two other compounds, derived from the database searching, that probed the common features model) this second quantitative model was generated and included 2 hydrophobes (cyan), 2 hydrogen-bond acceptors (green), and an ionizable feature (red) seen with cinchonidine’s structure. The graph on the right (B) displays the relationship between measured and predicted IC50 values based on the model, which has a correlation r value of 0.71 (p < 0.0001).

100

A. B.

4.00 50

M) 2.00

 (

Log Predicted IC 0.00 r= 0.73 p < 0.0001 0.00 2.00 4.00 Log Experimental IC50 (M)

Figure 15 Final Iteration of the Quantitative Pharmacophore for hMATE1 & Correlation

Graph: The final model included two hydrophobes (cyan), a hydrogen bond-acceptor (magenta), and an ionizable feature (red). The graph on the right (B) displays the relationship between measured and predicted IC50 values based on the N46 model, which has a correlation r value of

0.73 (p < 0.0001).

101

Figure 16 hMATE1 Quantitative Pharmacophore Generated With Values From Kido et al

(73): The Kido quantitative pharmacophore had 3 hydrophobes (cyan), 2 hydrogen bond acceptors (green) and 3 excluded volumes (grey) seen here with orphenadrine. Kido et al (73) used ASP as their inhibitory compound whereas we used MPP the result is fundamentally different hMATE1 quantitative pharamcophores.

102

A. B.

25 2 1200 2 Jmax= 3.02 pmol/cm x min Jmax= 9.31 pmol/cm x min Kt= 4.88 M Kt= 64.3 M 20

900

min

min 2 15 2 600 10

300

H]MPP Uptake pmol/cm

5 pmol/cm

[

C]Metformin Uptake 14

0 [ 0 0 1 100 10000 0 1 100 10000 [MPP] M [Metformin] M

Figure 17 hMATE1 Kinetics of MPP & Metformin: hMATE1 MPP (A) and metformin (B) kinetics reveal that while metformin has a higher turnover number than MPP, the higher affinity of MATE1 for MPP results in a more effective interaction with the transporter. Each point is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº).

Jmax, and Kt values were calculated as an average of 3 experiments.

103

B. A. 8 6

IC = 47.0 M 50 6

4 x min

x min

2 2 NH2 4 HN NH 2 NH

N CH 2 H]MPP Uptake

3 H]MPP Uptake

fmol/cm

pmol/cm [ H3C [ 0 0 0 10 100 1000 10000 0.0 0.2 0.4 0.6 0.8 [Metformin] M [3H]MPP/[MPP] pmol/cm2 x min x M

Figure 18 hMATE1 Metformin Inhibition Kinetics of MPP, MPP Kinetics in the Absence &

Presence of 250 µM Metformin: Inhibitory profile of metformin’s inhibition of MPP, resulting in an IC50 of 47.0 µM (A). B shows Eadie-Hofstee plots of the carrier-mediated (i.e., saturable) components of these two data sets, which emphasizes the apparent competitive nature of the interaction of metformin and MPP. Each point is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº) in five separate experiments. Jmax, Kt and

IC50 values were calculated as an average of 3-5 experiments.

104

A. 8 B. 4 IC50 = 195 M 6

x min x min

O 2 2 4 N CH3 2 N

NH2 2 H]MPP Uptake

H]MPP Uptake 1

fmol/cm

3 3

pmol/cm

[ [ 0 0 0 10 100 1000 0.0 0.2 0.4 0.6 0.8 [Creatinine] uM [3H]MPP/[MPP] fmol/cm2 x min x M

Figure 19 hMATE1 Creatinine Inhibition Kinetics of MPP & MPP Kinetics in the Absence

& Presence of 1 mM Creatinine: Inhibitory profile of creatinine’s inhibition of hMATE1, resulting in an IC50 of 195 µM (A). The Eadie-Hofstee plot in B emphasizes the apparent competitive nature of the interaction of creatinine and MPP. Each point is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº) in five separate experiments. Jmax, Kt and IC50 values were calculated as an average of 3-5 experiments.

105

0.8 A. B. 8 8 -1

0.6 min

-2 0.4 6 6

0.2

H]MPP Uptake 3 IC = 0.042 M fmol cm

50 [ x min

2 4 0.0 4 NH2 0.00 0.01 0.02 0.03 N [3H]MPP/[MPP] H2N N fmol cm -2 min-1 M-1 2

H]MPP Uptake

fmol/cm

3 [ Cl 0 0 0 0.01 0.10 1.00 10.00 0.0 0.2 0.4 0.6 0.8 [PYR] M [3H]MPP/[MPP] fmol cm-2 min-1 M-1

Figure 20 hMATE1 PYR Inhibition Kinetics of MPP & MPP Kinetics in the Absence &

Presence of 1 µM PYR: PYR inhibition of MPP has an IC50 of ~40nM (A). B show Eadie-

Hofstee plots of MPP kinetics and MPP kinetics in the presence of 1 µM PYR (inset is MPP kinetics in the presence of 1 µM PYR), where Jmax is largely decreased with the presence of PYR and Kt values are only slightly altered. Each point is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº) in five separate experiments. Jmax, Kt and IC50 values were calculated as an average of 3-5 experiments.

106

extracellular medium replaced B. A. 60 C.

20 tracer 100

)

2 ) 15 2

40 2 75 *** ***

10 *** 50 *** ***

(fmol/cm 20

(fmol/cm H]MPP Uptake 5 (fmol/cm

3 tracer + 100 M MPP

[ 25

+1 mM MPP H]MPP Content Cell

H]MPP Content Cell

3 [

0 3 0 20 40 60 0 [ 0 0 100 200 300 0 1 3 10 30 100 300 1000 Time (Min) Time (Sec) trans [MPP] M

Figure 21 hMATE1 MPP Uptake Time Course, Efflux Time Course & MPP Concentration-

Dependent Trans-Stimulation: Steady-state accumulation of [3H]MPP (~13 nM) into hMATE1-expressing cells was achieved within approximately 30 minutes (A). Then the extracellular medium was replaced with either (i) 13 nM [3H]MPP (identical to the loading solution); or (ii) one containing the 13 nM [3H]MPP plus 100 µM unlabeled MPP (B). C shows that at increasing concentrations of extracellular MPP the cell content of [3H]MPP significantly decreased beginning at 10 µM. Each point is the mean (±SEM) of uptake or efflux measured in three wells of a 24-well plate, at pH 8.5 (RTº). Significance was calculated using a 1-way

ANOVA where p < 0.0001.

107

125

100 ***

H]MPP

2 3 75 *** *** ***

50 fmol/cm 25

Cell ContentCell [ 0 0 1 10 100 1000 10000 trans [PYR] nM

Figure 22 PYR Concentration-Dependent Trans-Stimulation: Increasing concentrations of extracellular PYR trans-stimulate [3H]MPP efflux (e.g. cell content) significantly beginning at 10 nM. Each point is the mean (±SEM) of uptake or efflux measured in three wells of a 24-well plate, at pH 8.5 (RTº). Significance was calculated using a 1-way ANOVA where p < 0.0001.

108

-1 40

min 30 Tracer only -2

20 + 100 nM PYR

H]MPP Uptake 10

fmol cm [ + 1mM MPP 0 0 100 200 300 Time (Sec)

Figure 23 [3H]MPP Time Course in the Presence of MPP or PYR: Time course of 13 nM

[3H]MPP uptake, from 15 seconds through 5 minutes, in the absence (closed circles) and presence of either 100 nM PYR (open circles) or 1 mM MPP (closed squares). Each point is the mean

(±SEM) of uptake measured in three wells of a 24-well plate, at pH 8.5 (RTº).

109

A. B.

8 6

IC50= 0.14 M 6 IC50= 0. 196 M x min

2 4 H2N 4 N H2N N NH2

N NH2 Cl N N 2 N H C 2 H-MPP Uptake 3 H3C CH3 fmol/cm Cl

3 CH3

0 0 0 0.1 1 10 0 0.1 1 10 [PYR-2] M [PYR-3] M

Figure 24 PYR-2 & PYR-3 Inhibition Kinetics of MPP: PYR structural analogs, PYR-2 (A) and PYR-3 (B) are high affinity inhibitors of hMATE1–mediated transport of [3H]MPP. Each point is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH

8.5 (RTº). IC50 values were calculated as an average of 3-4 experiments.

110

A. Tracer 1 mM MPP Compound X 100 100 100

75 75 75

50 50 50

%Tracer Uptake %Tracer 25 25 25

0 0 0 30s 5 min 30s 5 min 30s 5 min MPP Metformin PYR

A. Cont. 100 B. 100 MPP 75 Metformin 75 PYR 30s PYR-2 * 300s PYR-3 50 50 MPP Metformin PYR %Tracer Uptake %Tracer 25 25 PYR-2 PYR-3 0 25 50 75 100 0 0 % Inhibition of [3H]MPP Uptake 30s 5 min 30s 5 min PYR-2 PYR-3

Figure 25 Time Dependence of Inhibition & Time Dependence of Inhibition Normalized to

Total Tracer Uptake: Uptake of [3H]MPP in the presence and absence of MPP (1 mM (gray) and 20 µM (striped); top left), 250 µM metformin (striped; middle top) and 200 nM of PYR

(striped; top right), PYR-2 (striped; bottom left), and PYR-3 (striped; bottom right) at 30 seconds and 5 minutes (A). Normalized data (to percent inhibition of 13 nM [3H]MPP uptake, at 30 seconds versus 5 minutes) shown in B. Creatinine, metformin and both PYR analogs displayed marked inhibition; however, PYR, showed a significantly lower degree of inhibition (P < 0.05) than all of the other compounds at 30 seconds. Each point is the mean (±SEM) of uptake or efflux measured in three wells of a 24-well plate, at pH 8.5 (RTº). Significance was calculated using a 1-way ANOVA where p < 0.05.

111

4 MPP MPP + 200 nM PYR-2

x min 3 2

H-MPP Uptake 1

fmol/cm 3

0 0.0 0.2 0.4 0.6 0.8 1.0 [3H] MPP/[MPP] fmol cm-2 min-1 M-1

Figure 26 MPP Kinetics in the Absence & Presence of 200 nM PYR-2: Example (n =1) of an

Eadie-Hofstee plot of MPP kinetics (closed circles) and MPP kinetics in the presence of 200 nM

PYR-2 (open circles) where the decrease in Jmax and Kt values is highlighted. Each point is the mean (±SEM) of uptakes measured in three wells of a 24-well plate, after 5 min at pH 8.5 (RTº).

112

Figure 27 Model for Competitive Inhibition & Linear Mixed-Type Inhibition: A model for

Competitive Inhibition can be seen in panel A, where T is the transporter T, S 1 and 2 are the substrates, Ks is the dissociation constant for the appropriate substrate and subscripts o and i represent the transporter facing the outside and inside of the cell, respectively. In competitive inhibition the transporter may bind to an inhibitor, substrate, or neither, but it cannot bind both at the same time. Panel B shows Linear Mixed-Type Inhibition, where the presence of two substrates, binding to two different binding sites, either decreases or completely inhibits the turnover of the transporter.

113

H]MPP

2 3 40

** fmol/cm 20

Cell ContentCell [ 0

MPP Control Time "0"

Control-Mannitol

Figure 28 Competitive Exchange Diffusion With hMATE2-K Controls: Cellular content of

[3H]MPP at time “0” (e.g. after [3H]MPP accumulation has reached steady-state, and before adding an extracellular compound), control (buffer at pH 8.5), control-mannitol (control buffer with 20 mM mannitol) and MPP are shown in this figure. Mannitol did not trans-stimulate efflux of the intracellular [3H]MPP; only MPP was able to stimulate efflux significantly (p < 0.05).

Each point is the mean (±SEM) of uptake or efflux measured in four wells of a 48-well plate, at pH 8.5 (RTº). Significance was calculated using a 1-way ANOVA where p < 0.05.

114

H]MPP

2 3 40 *** *** *** *** ***

fmol/cm 20

Cell ContentCell [ 0

MPP PYR Control MetforminParaquat Creatinine

Figure 29 Competitive Exchange Diffusion With Known MATE Substrates: This figure confirms previous studies that showed that MPP, metformin, paraquat, PYR and creatinine are

MATE substrates. All of the trans-stimulated [3H]MPP efflux significantly (p < 0.0001). Each point is the mean (±SEM) of uptake or efflux measured in four wells of a 48-well plate, at pH 8.5

(RTº). Significance was calculated using a 1-way ANOVA where p < 0.05.

115

H]MPP

2 3 40 *** ***

fmol/cm *** *** 20

Cell ContentCell [ 0

TEA MPP TmA TPeA Control

Figure 30 Competitive Exchange Diffusion With N-Tetraalkylammonium Series: All members of this series stimulated significant (p< 0.0001) efflux of [3H]MPP, comparable to that or unlabeled MPP. Each point is the mean (±SEM) of uptake or efflux measured in four wells of a 48-well plate, at pH 8.5 (RTº). Significance was calculated using a 1-way ANOVA where p <

0.05.

116

H]MPP 2

3 *** *** *** *** 40 ***

fmol/cm *** 20

Cell ContentCell [ 0

MPP APMI PAH Control Caffeine Quinidine Ethohexadiol

Figure 31 Competitive Exchange Diffusion With Unknown MATE Substrates: These test compounds included two very potent inhibitors of MATE2-K, APMI and quinidine; two weak inhibitors, ethohexadiol and caffeine; and the prototypical substrate of the renal organic anion secretory pathway, PAH. All of these compounds proved to be transported substrates of hMATE2-K, capable of trans-stimulating significant levels of intracellular [3H]MPP (p <0.0001).

Each point is the mean (±SEM) of uptake or efflux measured in four wells of a 48-well plate, at pH 8.5 (RTº). Significance was calculated using a 1-way ANOVA where p < 0.05.

117

APPENDIX B: TABLES

Table 1A: hMATE1 Kinetic Profiles Substrate Kt Jmax TE Cell Condition Ref. (μM) nmol mg-1 min-1 μl mg-1 min-1 acyclovir 2640 0.62 0.2 HEK NH4; pH (136) 7.4; 37ºC cimetidine 8.0 0.17 21.3 HEK NH4; pH (73) 7.4; 37ºC DAPI2 1.1 *** NA MDCK NH4; pH (172) 7.4; 37ºC estrone sulfate 470 0.27 0.6 HEK NH4; pH (136) 7.4; 37ºC ganciclovir > 5000 1.06 0.2 HEK NH4; pH (136) 7.4; 37ºC guanidine 2100 0.89 0.4 HEK NH4; pH (136) 7.4; 37ºC metformin 238 1.81 7.6 HeLa pH 8; RTº (94) 253 10.8 42.7 HEK pH 7.4; 37ºC (76) MPP 100 7.35 73.5 HEK NH4; pH (107) 7.4; 37ºC MPP1 4.0 *** NA CHO pH 8.5; RTº (22) paraquat 212 0.29 1.4 HEK pH 8; RTº (20) procainamide 1230 3.78 3.1 HEK NH4; pH (136) 7.4; 37ºC TEA 366 7.54 20.6 HEK NH4; pH (73) 7.4; 37ºC 220 2.26 10.3 HEK pH 8; 37ºC (107) 380 1.19 3.1 HEK NH4; pH (136) 7.4; 37ºC 490 1.94 4.0 HEK NH4; pH (69) 7.4; 37ºC topotecan 70 0.21 3.0 HEK NH4; pH (136) 7.4; 37ºC

118

Table 1B: hMATE2-K Kinetic Profiles Substrate Kt Jmax TE Cell Condition Ref. (μM) nmol mg-1 min-1 μl mg-1 min-1 cimetidine 18.2 0.22 12.1 HEK NH4; pH (73) 7.4; 37ºC guanidine 4200 0.52 0.1 HEK NH4; pH (136) 7.4; 37ºC metformin 1050 NR 11.9 HEK pH 8.5; (89) 37ºC 362 2.4 6.6 HEK pH 7.4; (76) 37ºC MPP 94 NR 3.1 HEK NH4; pH (89) 7.4; 37ºC procainamide 4100 NR 1.4 HEK NH4; pH (89) 7.4; 37ºC TEA 375 1.2 3.1 HEK NH4; pH (73) 7.4; 37ºC 760 0.80 1.1 HEK NH4; pH (136) 7.4; 37ºC 139 2.2 14.3 HEK NH4; pH (69) 7.4; 37ºC 830 NR 7.6 HEK pH 8.5; (89) 37ºC 110 5.8 52.7 HEK NH4; pH (136) 7.4; 37ºC topotecan 60 1.3 2.2 HEK NH4; pH (136) 7.4; 37ºC

119

Table 1C: hMATE1 Inhibition Kinetics Inhibitor IC50 Substrate Cell Condition Ref. (μM) cefazolin NI [14C]TEA HEK NH4; pH 7.4; (136) 37ºC cephalexin > 5000 [14C]TEA HEK NH4; pH 7.4; (136) 37ºC cephradine 4040 [14C]TEA HEK NH4; pH 7.4; (136) 37ºC chloroquine 2.5 [14C]metformin HEK pH 7.8; 37ºC (101) cimetidine 5.7 ASP HEK pH 7.4; RTº (73) ciprofloxacin 231 [14C]TEA HEK NH4; pH 7.4; (136) 37ºC DAPI 6.9 [3H]cimetidine HEK NH4; pH 7.4; (73) 37ºC dipryridamole 26 ASP HEK pH 7.4; RTº (73) disopyramide 66 ASP HEK pH 7.4; RTº (73) EB 5.7 [3H]cimetidine HEK NH4; pH 7.4; (105) 37ºC imatinib 1.0 [14C]creatinine HEK NH4; pH 7.4; (135) 37ºC imipramine 10 ASP HEK pH 7.4; RTº (73) levofloxacin 38.2 [14C]TEA HEK NH4; pH 7.4; (136) 37ºC ondansetron 0.15 ASP HEK pH 7.4; RTº (73) orphenadrine 65 ASP HEK pH 7.4; RTº (73) propidium iodine 15.3 [3H]cimetidine HEK NH4; pH 7.4; (73) 37ºC PYR 0.08 [14C]TEA HEK pH 7.4; 37ºC (61) 0.06 [14C]metformin HEK pH 7.4; 37ºC (76) tacrine 1.1 ASP HEK pH 7.4; RTº (73) TBA 18.3 [14C]paraquat HEK pH 8; RTº (20) TEA 121 [14C]paraquat HEK pH 8; RTº (20) TMA > 5000 [14C]paraquat HEK pH 8; RTº (20) TPrA 63 [14C]paraquat HEK pH 8; RTº (20) trimethoprim 6.2 [14C]metformin HEK pH 7.8; 37ºC (101)

120

Table 1D: hMATE2-K Inhibition Kinetics Inhibitor IC50 Substrate Cell Condition Ref. (μM) cefazolin NI [14C]TEA HEK NH4; pH (136) 7.4; 37ºC cephalexin NI [14C]TEA HEK NH4; pH (136) 7.4; 37ºC cephradine 1040 [14C]TEA HEK NH4; pH (136) 7.4; 37ºC cimetidine 39 ASP HEK pH 7.4; RTº (73) ciprofloxacin 99 [14C]TEA HEK NH4; pH (136) 7.4; 37ºC DAPI 15.7 [3H]cimetidine HEK NH4; pH (73) 7.4; 37ºC dipryridamole 74 ASP HEK pH 7.4; RTº (73) disopyramide >100 ASP HEK NH4; 37ºC (73) EB 3.0 [3H]cimetidine HEK NH4; pH (73) 7.4; 37ºC imatinib 4.3 [14C]creatinine HEK NH4; pH (135) 7.4; 37ºC imipramine >100 ASP HEK pH 7.4; RTº (73) levofloxacin 82 [14C]TEA HEK NH4; pH (136) 7.4; 37ºC ondansetron 6.9 ASP HEK pH 7.4; RTº (73) orphenadrine >100 ASP HEK pH 7.4; RTº (73) propidium iodine 1.3 [3H]cimetidine HEK NH4; pH (73) 7.4; 37ºC PYR 0.01 [14C]metformin HEK pH 7.4; (76) 37ºC 0.05 [14C]TEA HEK pH 7.4; (61) 37ºC tacrine >100 ASP HEK pH 7.4; RTº (73) Abbreviations: ASP, 4-(4-dimethlamino)styryl)-N-methyl-pyridinium; CHO, Chinese Hamster Ovary cell; DAPI, 4',6-diamidino-2-phenylindole; EB, ethidium bromide; HEK, Human Embryonic Kidney cell; HeLa, immortal cervical cancer cells; MDCK II, Madin Darby Canine Kidney cells; MPP, l-methyl-4- phenylpyridinium; NH4, Intracellular acidification using the ammonia pulse technique; NI, No Interaction; NR, Data Not Reported; PYR, pyrimethamine; TBA, tetrabutylammonium; TEA, tetraethylammonium; TMA, tetramethylammonium; and TPrA, tetraproprylammonium. 1Jmax reported as, pmol min-1 cm-2 ; 2Jmax reported as, F1/10 min/mg protein

121

Table 2: IC50 Values Generated for hMATE1 & hMATE2-K Compound Drug Class (1) IC50 (1) Cal. IC50 (2-K) IC50 pH 8.5 (2-K) Cal. IC50 pH 8.5 pH 7.4 pH 7.4 Agmatine a NT 53.9 ± 1.40 141 60.8 ± 9.12 395 Allopurinol a XAI NI N/A NI N/A Amantidine a AV 7.53 ± 1.49 19.7 88.9 ± 9.0 578 Amiloride a DI 2.41 ± 0.183 6.3 3.06 ± 0.615 19.9 APMI a PAL 6.21 ± 0.255 16.2 0.501 ± 0.224 3.30 Atropine a AC 5.9 ± 1.31 15.3 52.8 ± 13.7 343 Baclofen AS, MR NI N/A NI N/A Caffeine a S 1096 ± 42.5 2863 451 ± 120 2932 Chloramphenicol a AB 1114 ± 156 2909 1951 ± 45 12683 Cinchonidine b SC 0.927 ± 0.154 2.38 4.38 ± 1.16 27.1 Cinchonine b SC 1.86 ± 0.375 4.78 0.939 ± 0.043 5.81 Cisplatin CT NI N/A NI N/A Clonidine a A2A 8.09 ± 0.786 21.1 54.0 ± 1.30 351 Creatinine a MBP 195 ± 39.2 510 150 ± 25.8 977 Ethohexadiol b IR >2000 N/A >1000 N/A Famotidine b H2RA 2.16 ± 0.388 5.56 6.28 ± 0.45 38.8 Guanfacine A2AN 3.51 ± 616 9.18 218 1346 Guanidine MBP >2100 5405 >4000 N/A H+ EI 0.0196 ± 0.001 N/A 0.0035 ± 0.001 N/A Histamine IRR, NT 761 ± 196 1986 775 ± 148 5036 Imiquimod b IMR 13.9 ± 5.92 35.9 19.1 ± 5.66 118 Ketoconazole AF 1.33 ± 0.165 3.43 9.33 ± 0.811 57.7 Metformin a AD 47.0 ± 2.24 123 89.3 ± 17.7 581 Midodrine b VP 109 ± 18.4 281 87.1 ± 38.4 539 MPP NTX 4.70 ± 0.548 12 3.3 ± 0.192 21.0 Nalxone OAN 24.1 ± 3.72 62.1 43.2 ± 12.9 280 Nialamide MAOI 212 ± 4.67 546 236 ± 30.6 1460 Nicotine S 167 ± 36.7 437 134 ± 27.9 869 NMN a POC 147 ± 80.7 385 46.3 ± 5.30 501 Paraquat a H 50.5 ± 7.67 132 15.5 ± 1.3 101 Phenformin a AD 6.10 ± 0.186 16 11.2 ± 2.81 73.0 Phentolamine b AAA 4.64 ± 0.368 11.9 5.3 ± 0.527 32.8 Procainamide a AA 24.0 ± 1.27 63 19.1 ± 3.36 124 Proguanil b AM 4.35 ± 1.70 11.4 1.39 ± 0.512 9.02 Propranolol b BB 7.81 ± 0375 20.1 7.71 ± 0.104 47.7 PYR a AM 0.042 ± 0.01 0.547 0.415 ± 0.144 2.70 PYR-2 SC 0.135 ± 0.01 0.789 4.71 ± 0.278 30.6 PYR-3 SC 0.196 ± 0.01 0.833 6.32 ± 0.447 41.1 Quinidine a AA 1.57 ± 0.318 4.1 1.47 ± 0.273 9.60 Quinine b AA 1.90 ± 0.410 5 6.4 ± 1.66 42.0 Ranitidine a H2RA 5.40 ± 1.02 14 10 ± 1.88 65.0 Scopolamine b AC, AMS 47.6 ± 4.83 123 272 ± 24.8 1683 Serotonin MA, NT 28.8 ± 5.30 75 18.3 ± 1.40 119 Sulfadimethoxine AB >>1000 N/A >>1000 N/A Sulfamerazine AB >>1000 N/A >>1000 N/A Tacrine ACS 0.51 ± 0.10 1.31 1.10 ± 0.21 7.17 TBA SC 11.7 30.4 N/A N/A TEA a POC 40.6 ± 4.57 106 14.4 ± 8.75 94.0 TEMA a SC 51.6 ± 10.2 135 28.4 ± 0.721 184 Ticlopidine AP >1000 N/A >1000 N/A TMA SC >5000 N/A N/A N/A Topiramate b ACT ~1000 N/A NI N/A TPeA SC 6.34 ± 0.23 16.4 N/A N/A TPrA SC 14.8 ± 4.03 38.5 N/A N/A Tramadol OA 17.8 ± 1.60 46.3 74.6 ± 17.6 485 Trichlormethiazide b DI 249 ± 9.8 640 679 4198 Trimethoprim a AB 9.70 ± 1.22 225 2.61 ± 1.12 17.0 Tryptophan AAM NI N/A NI N/A Tyramine a MA 86.6 ± 8.17 226 138 ± 10.8 899 Verapamil LTCB 41.9 ± 6.75 109 37.9 ± 9.74 247 Abbreviations: A2A, Alpha-2 Adrenergic Agonist; A2AN, Alpha-2A Norepinephrin Receptor Agonist; AA, Antiarrhythmic; AAA, Alpha Adrenergic Antagonist; AAC, Amino Acid; AB, Antibiotic; AC, Anticholinergic; ACS, Anticholinesterase; ACT, Anticonvulsant; AD, Antidiabetic; AF, Antifungal; AM, Antimalarial; AMS, Antimuscurinic; AP, Antiplatelet; AS, Antispastic; AV, Antiviral; BB, Beta Blocker; CT, Chemotherapy Drug; DI, Diuretic; EI, Endogenous Ion; H, Herbacide; H2RA, Histamine H2- Receptor Antagonist; IMR, Immune Response Modifier; IR, Insect Repellent; IRR, Immune Response Regulator; LTCB, L- Type Calcium Blocker; MA, Momoamine; MAOI, Monoamine Oxidase Inhibitor; MBP, Metabolism By-Product; MS, Muscle Relaxer; NT, Neurotransmitter; NTX, Neurotoxin; OA, Opiate Agonist; OAN, Opiate Antagonist; PAL, Photoaffinity Label; POC, Prototypical Organic Cation; S, Stimulant; SC, Synthetic Chemical; VP, Vasopressor; XOI, Xanthine Oxidase Inhibitors; and NI, No interaction. Each value is the mean (±SEM) of 3-4 experiments, a represents the initial 24 compounds used for the first iteration of pharamacophores, b represents the 15 compounds use to test the common features pharmacophore and the bolded compounds are those identified by the pharamacophores.

122

Table 3: hMATE1 Jmax & Kt values for MPP Kinetics in Absence and Presence of Five Test Compounds and Metformin Kinetics

Compound Jmax Kt pmol/cm2 x min µM MPP 3.02 ± 0.43 4.88 ± 0.43 Metformin 9.31 ± 1.86 64.3 ± 6.33 MPP + 250 µM Metformin 3.93 ± 0.76 35.5 ± 5.28 MPP + 1000 µM Creatinine 2.66 ± 0.76 10.4 ± 1.50 MPP + 1 µM PYR 0.25 ± 0.07 11.1 ± 3.06 MPP + 0.4 µM PYR-2 0.72 ± 0.03 9.21 ± 2.35 Data are means (±SEM) of 3-5 separate experiments. MPP, l-methyl-4- phenylpyridinium; PYR, pyrimethamine; PYR-2, 1-(2-chlorophenyl)-6,6-dimethyl-1,6- dihydro-1,3,5-triazine-2,4-diamine.

123

REFERENCES

1. Acara M and Rennick B. Regulation of plasma choline by the renal tubule: bidirectional transport of choline. Am J Physiol 225: 1123-1128, 1973.

2. Acara M, Roch-Ramel F and Rennick B. Bidirectional renal tubular transport of free choline: a micropuncture study. Am J Physiol 236: F112-F118, 1979.

3. Ahlin G, Karlsson J, Pedersen JM, Gustavsson L, Larsson R, Matsson P, Norinder U, Bergstrom CA and Artursson P. Structural requirements for drug inhibition of the liver specific human Organic Cation Transport protein. J Med Chem 2008.

4. Aller SG, Yu J, Ward A, Weng Y, Chittaboina S, Zhuo R, Harrell PM, Trinh YT, Zhang Q, Urbatsch IL and Chang G. Structure of P-glycoprotein reveals a molecular basis for poly-specific drug binding. Science 323: 1718-1722, 2009.

5. Arndt P, Volk C, Gorboulev V, Budiman T, Popp C, Ulzheimer-Teuber I, Akhoundova A, Koppatz S, Bamberg E, Nagel G and Koepsell H. Interaction of cations, anions, and weak base quinine with rat renal cation transporter rOCT2 compared with rOCT1. Am J Physiol Renal Physiol 281: F454-F468, 2001.

6. Asaka J, Terada T, Tsuda M, Katsura T and Inui K. Identification of essential histidine and cysteine residues of the H++/organic cation antiporter Multidrug and Toxin Extrusion (MATE). Mol Pharmacol 71: 1487-1493, 2007.

7. Bahn A, Hagos Y, Reuter S, Balen D, Brzica H, Krick W, Burckhardt BC, Sabolic I and Burckhardt G. Identification of a new urate and high affinity nicotinate transporter, hOAT10 (SLC22A13). J Biol Chem 283: 16332-16341, 2008.

8. Becker ML, Kallewaard M, Caspers PW, Visser LE, Leufkens HG and Stricker BH. Hospitalisations and emergency department visits due to drug-drug interactions: a literature review. Pharmacoepidemiol Drug Saf 16: 641-651, 2007.

9. Becker ML, Visser LE, van Schaik RH, Hofman A, Uitterlinden AG and Stricker BH. Interaction between polymorphisms in the OCT1 and MATE1 transporter and metformin response. Pharmacogenet Genomics 20: 38-44, 2010.

10. Bednarczyk D, Ekins S, Wikel JH and Wright SH. Influence of molecular structure on substrate binding to the human organic cation transporter, hOCT1. Mol Pharm 63: 489- 498, 2003.

11. Begum A, Rahman MM, Ogawa W, Mizushima T, Kuroda T and Tsuchiya T. Gene cloning and characterization of four MATE family multidrug efflux pumps from Vibrio cholerae non-O1. Microbiol Immunol 49: 949-957, 2005.

12. Bowman HM and Hook JB. Sex differences in organic ion transport by rat kidney. Proc Soc Exp Biol Med 141: 258-262, 1972.

124

13. Burckhardt G and Burckhardt BC. In vitro and in vivo evidence of the importance of organic anion transporters (OATs) in drug therapy. Handb Exp Pharmacol 29-104, 2011.

14. Busch AE, Quester S, Ulzheimer JC, Waldegger S, Gorboulev V, Arndt P, Lang F and Koepsell H. Electrogenic properties and substrate specificity of the polyspecific rat cation transporter rOCT1. J Biol Chem 271: 32599-32604, 1996.

15. Cetinkaya I, Ciarimboli G, Yalcinkaya G, Mehrens T, Velic A, Hirsch JR, Gorboulev V, Koepsell H and Schlatter E. Regulation of human organic cation transporter hOCT2 by PKA, PI3K, and calmodulin-dependent kinases. Am J Physiol Renal Physiol 284: F293-F302, 2003.

16. Chandra P and Brouwer KL. The complexities of hepatic drug transport: current knowledge and emerging concepts. Pharm Res 21: 719-735, 2004.

17. Chang C, Bahadduri PM, Polli JE, Swaan PW and Ekins S. Rapid identification of P- glycoprotein substrates and inhibitors. Drug Metab Dispos 34: 1976-1984, 2006.

18. Chen Y, Li S, Brown C, Cheatham S, Castro RA, Leabman MK, Urban TJ, Chen L, Yee SW, Choi JH, Huang Y, Brett CM, Burchard EG and Giacomini KM. Effect of genetic variation in the organic cation transporter 2 on the renal elimination of metformin. Pharmacogenet Genomics 19: 497-504, 2009.

19. Chen Y, Teranishi K, Li S, Yee SW, Hesselson S, Stryke D, Johns SJ, Ferrin TE, Kwok P and Giacomini KM. Genetic variants in multidrug and toxic compound extrusion-1, hMATE1, alter transport function. Pharmacogenomics J 9: 127-136, 2009.

20. Chen Y, Zhang S, Sorani M and Giacomini KM. Transport of paraquat by human organic cation transporters and multidrug and toxic compound extrusion family. J Pharmacol Exp Ther 322: 695-700, 2007.

21. Ciarimboli G, Struwe K, Arndt P, Gorboulev V, Koepsell H, Schlatter E and Hirsch JR. Regulation of the human organic cation transporter hOCT1. J Cell Physiol 201: 420- 428, 2004.

22. Dangprapai Y and Wright SH. Interaction of H+ with the extracellular and intracellular aspects of hMATE1. Am J Physiol Renal Physiol 2011.

23. Dantzler WH, Brokl O and Wright SH. Brush-border TEA transport in intact proximal tubules and isolated membrane vesicles. Am J Physiol 256: F290-F297, 1989.

24. David C, Rumrich G and Ullrich KJ. Luminal transport system for H+/organic cations in the rat proximal tubule. Kinetics, dependence on pH; specificity as compared with the contraluminal organic cation-transport system. Pflugers Arch 430: 477-492, 1995.

25. Diao L, Ekins S and Polli JE. Novel inhibitors of human organic cation/carnitine transporter (hOCTN2) via computational modeling and in vitro testing. Pharm Res 26: 1890-1900, 2009.

26. Diao L, Ekins S and Polli JE. Quantitative structure activity relationship for inhibition of human organic cation/carnitine transporter. Mol Pharm 7: 2120-2131, 2010.

125

27. Dridi L, Tankovic J and Petit JC. CdeA of Clostridium difficile, a new multidrug efflux transporter of the MATE family. Microb Drug Resist 10: 191-196, 2004.

28. Ekins S, Bravi G, Binkley S, Gillespie JS, Ring BJ, Wikel JH and Wrighton SA. Three- and four-dimensional quantitative structure activity relationship analyses of cytochrome P-450 3A4 inhibitors. J Pharmacol Exp Ther 290: 429-438, 1999.

29. Ekins S, Bravi G, Ring BJ, Gillespie TA, Gillespie JS, Vandenbranden M, Wrighton SA and Wikel JH. Three-dimensional quantitative structure activity relationship analyses of substrates for CYP2B6. J Pharmacol Exp Ther 288: 21-29, 1999.

30. Ekins S, Bravi G, Wikel JH and Wrighton SA. Three-dimensional-quantitative structure activity relationship analysis of cytochrome P-450 3A4 substrates. J Pharmacol Exp Ther 291: 424-433, 1999.

31. Ekins S, Chang C, Mani S, Krasowski MD, Reschly EJ, Iyer M, Kholodovych V, Ai N, Welsh WJ, Sinz M, Swaan PW, Patel R and Bachmann K. Human pregnane X receptor antagonists and agonists define molecular requirements for different binding sites. Mol Pharmacol 72: 592-603, 2007.

32. Ekins S, Johnston JS, Bahadduri P, D'Souza VM, Ray A, Chang C and Swaan PW. In vitro and pharmacophore-based discovery of novel hPEPT1 inhibitors. Pharm Res 22: 512-517, 2005.

33. Ekins S, Kim RB, Leake BF, Dantzig AH, Schuetz EG, Lan LB, Yasuda K, Shepard RL, Winter MA, Schuetz JD, Wikel JH and Wrighton SA. Application of three- dimensional quantitative structure-activity relationships of P-glycoprotein inhibitors and substrates. Mol Pharmacol 61: 974-981, 2002.

34. Ekins S, Kim RB, Leake BF, Dantzig AH, Schuetz EG, Lan LB, Yasuda K, Shepard RL, Winter MA, Schuetz JD, Wikel JH and Wrighton SA. Three-dimensional quantitative structure-activity relationships of inhibitors of P-glycoprotein. Mol Pharmacol 61: 964-973, 2002.

35. Elferink RP, Ottenhoff R, Liefting W, de Haan J and Jansen PL. Hepatobiliary transport of glutathione and glutathione conjugate in rats with hereditary hyperbilirubinemia. J Clin Invest 84: 476-483, 1989.

36. Endres CJ, Hsiao P, Chung FS and Unadkat JD. The role of transporters in drug interactions. Eur J Pharm Sci 27: 501-517, 2006.

37. Erlinger S. Mechanisms of hepatic transport and bile secretion. Acta Gastroenterol Belg 59: 159-162, 1996.

38. Furukawa J, Yamaji N, Wang H, Mitani N, Murata Y, Sato K, Katsuhara M, Takeda K and Ma JF. An aluminum-activated citrate transporter in barley. Plant Cell Physiol 48: 1081-1091, 2007.

126

39. Galetin A, Ito K, Hallifax D and Houston JB. CYP3A4 substrate selection and substitution in the prediction of potential drug-drug interactions. J Pharmacol Exp Ther 314: 180-190, 2005.

40. Gao M, Loe DW, Grant CE, Cole SPC and Deeley RG. Reconstitution of ATP- dependent leukotriene C4 transport by Co- expression of both half-molecules of human multidrug resistance protein in insect cells. J Biol Chem 271: 27782-27787, 1996.

41. German N, Wei P, Kaatz GW and Kerns RJ. Synthesis and evaluation of fluoroquinolone derivatives as substrate-based inhibitors of bacterial efflux pumps. Eur J Med Chem 43: 2453-2463, 2008.

42. Giacomini KM, Huang SM, Tweedie DJ, Benet LZ, Brouwer KL, Chu X, Dahlin A, Evers R, Fischer V, Hillgren KM, Hoffmaster KA, Ishikawa T, Keppler D, Kim RB, Lee CA, Niemi M, Polli JW, Sugiyama Y, Swaan PW, Ware JA, Wright SH, Wah YS, Zamek-Gliszczynski MJ and Zhang L. Membrane transporters in drug development. Nat Rev Drug Discov 9: 215-236, 2010.

43. Gorboulev V, Shatskaya N, Volk C and Koepsell H. Subtype-specific affinity for corticosterone of rat organic cation transporters rOCT1 and rOCT2 depends on three amino acids within the substrate binding region. Mol Pharmacol 67: 1612-1619, 2005.

44. Gorboulev V, Ulzheimer JC, Akhoundova A, Ulzheimer-Teuber I, Karbach U, Quester S, Baumann C, Lang F, Busch AE and Koepsell H. Cloning and characterization of two human polyspecific organic cation transporters. DNA Cell Biol 16: 871-881, 1997.

45. Groves CE, Morales MN, Gandolfi AJ, Dantzler WH and Wright SH. Peritubular paraquat transport in isolated renal proximal tubules. J Pharmacol Exp Ther 275: 926- 932, 1995.

46. Gründemann D, Gorboulev V, Gambaryan S, Veyhl M and Koepsell H. Drug excretion mediated by a new prototype of polyspecific transporter. Nature 372: 549-552, 1994.

47. Habu Y, Yano I, Okuda M, Fukatsu A and Inui K. Restored expression and activity of organic ion transporters rOAT1, rOAT3 and rOCT2 after hyperuricemia in the rat kidney. Biochem Pharmacol 69: 993-999, 2005.

48. Habu Y, Yano I, Takeuchi A, Saito H, Okuda M, Fukatsu A and Inui K. Decreased activity of basolateral organic ion transports in hyperuricemic rat kidney: roles of organic ion transporters, rOAT1, rOAT3 and rOCT2. Biochem Pharmacol 66: 1107-1114, 2003.

49. Hagenbuch B. Drug uptake systems in liver and kidney: a historic perspective. Clin Pharmacol Ther 87: 39-47, 2010.

50. Hasdemir U. [The role of cell wall organization and active efflux pump systems in multidrug resistance of bacteria]. Mikrobiyol Bul 41: 309-327, 2007.

127

51. He GX, Kuroda T, Mima T, Morita Y, Mizushima T and Tsuchiya T. An H(+)- coupled multidrug efflux pump, PmpM, a member of the MATE family of transporters, from Pseudomonas aeruginosa. J Bacteriol 186: 262-265, 2004.

52. He SM, Li R, Kanwar JR and Zhou SF. Structural and functional properties of human multidrug resistance protein 1 (MRP1/ABCC1). Curr Med Chem 18: 439-481, 2011.

53. He X, Szewczyk P, Karyakin A, Evin M, Hong WX, Zhang Q and Chang G. Structure of a cation-bound multidrug and toxic compound extrusion transporter. Nature 467: 991-994, 2010.

54. Hiasa M, Matsumoto T, Komatsu T, Omote H and Moriyama Y. Functional characterization of testis-specific rodent multidrug and toxic compound extrusion 2, a class III MATE-type polyspecific H+/organic cation exporter. Am J Physiol Cell Physiol 293: C1437-C1444, 2007.

55. Holohan PD, Pessah NI and Ross CR. Binding of N1-methylnicotinamide and p- aminohippuric acid to a particulate fraction from dog kidney. J Pharmacol Exp Ther 195: 22-33, 1975.

56. Holohan PD and Ross CR. Mechanisms of organic cation transport in kidney plasma membrane vesicles: 1. Countertransport studies. J Pharmacol Exp Ther 215: 191-197, 1980.

57. Holohan PD and Ross CR. Mechanisms of organic cation transport in kidney plasma membrane vesicles: 2. DpH studies. J Pharmacol Exp Ther 216: 294-298, 1981.

58. Hosoyamada M, Sekine T, Kanai Y and Endou H. Molecular cloning and functional expression of a multispecific organic anion transporter from human kidney. Am J Physiol 276: F122-F128, 1999.

59. Hug CC, Jr., Mellett LB and Cafruny EJ. Stop-flow analysis of the renal excretion of tritium-labeled dihydromorphine. J Pharmacol Exp Ther 150: 259-269, 1965.

60. Iharada M, Miyaji T, Fujimoto T, Hiasa M, Anzai N, Omote H and Moriyama Y. Type 1 sodium-dependent phosphate transporter (SLC17A1 Protein) is a Cl(-)-dependent urate exporter. J Biol Chem 285: 26107-26113, 2010.

61. Ito S, Kusuhara H, Kuroiwa Y, Wu C, Moriyama Y, Inoue K, Kondo T, Yuasa H, Nakayama H, Horita S and Sugiyama Y. Potent and specific inhibition of mMate1- mediated efflux of type I organic cations in the liver and kidney by pyrimethamine. J Pharmacol Exp Ther 333: 341-350, 2010.

62. Jansen PL. Foreword: from classic bile physiology to cloned transporters. Semin Liver Dis 20: 245-250, 2000.

63. Ji L, Masuda S, Saito H and Inui K. Down-regulation of rat organic cation transporter rOCT2 by 5/6 nephrectomy. Kidney Int 62: 514-524, 2002.

128

64. Jonker JW, Wagenaar E, Van Eijl S and Schinkel AH. Deficiency in the organic cation transporters 1 and 2 (Oct1/Oct2 [Slc22a1/Slc22a2]) in mice abolishes renal secretion of organic cations. Mol Cell Biol 23: 7902-7908, 2003.

65. Jutabha P, Anzai N, Kitamura K, Taniguchi A, Kaneko S, Yan K, Yamada H, Shimada H, Kimura T, Katada T, Fukutomi T, Tomita K, Urano W, Yamanaka H, Seki G, Fujita T, Moriyama Y, Yamada A, Uchida S, Wempe MF, Endou H and Sakurai H. Human sodium phosphate transporter 4 (hNPT4/SLC17A3) as a common renal secretory pathway for drugs and urate. J Biol Chem 285: 35123-35132, 2010.

66. Kaatz GW, McAleese F and Seo SM. Multidrug resistance in Staphylococcus aureus due to overexpression of a novel multidrug and toxin extrusion (MATE) transport protein. Antimicrob Agents Chemother 49: 1857-1864, 2005.

67. Kaewmokul S, Chatsudthipong V, Evans KK, Dantzler WH and Wright SH. Functional mapping of rbOCT1 and rbOCT2 activity in the S2 segment of rabbit proximal tubule. Am J Physiol Renal Physiol 285: F1149-F1159, 2003.

68. Kajiwara M, Terada T, Asaka J, Ogasawara K, Katsura T, Ogawa O, Fukatsu A, Doi T and Inui K. Critical roles of Sp1 in gene expression of human and rat H+/organic cation antiporter MATE1. Am J Physiol Renal Physiol 293: F1564-F1570, 2007.

69. Kajiwara M, Terada T, Ogasawara K, Iwano J, Katsura T, Fukatsu A, Doi T and Inui KI. Identification of multidrug and toxin extrusion (MATE1 and MATE2-K) variants with complete loss of transport activity. J Hum Genet 2009.

70. Katsura T, Maegawa H, Tomita Y, Takano M, Inui K-I and Hori R. Trans- stimulation effect on H+-organic cation antiport-system in rat renal brush-border membranes. Am J Physiol 261: F774-F778, 1991.

71. Kenworthy KE, Bloomer JC, Clarke SE and Houston JB. CYP3A4 drug interactions: correlation of 10 in vitro probe substrates. Br J clin Pharmacol 48: 716-727, 1999.

72. Kerb R, Brinkmann U, Chatskaia N, Gorbunov D, Gorboulev V, Mornhinweg E, Keil A, Eichelbaum M and Koepsell H. Identification of genetic variations of the human organic cation transporter hOCT1 and their functional consequences. Pharmacogenetics 12: 591-595, 2002.

73. Kido Y, Matsson P and Giacomini KM. Profiling of a prescription drug library for potential renal drug-drug interactions mediated by the organic cation transporter 2. J Med Chem 54: 4548-4558, 2011.

74. Koehler MR, Wissinger B, Gorboulev V, Koepsell H and Schmid M. The two human organic cation transporter genes SLC22A1 and SLC22A2 are located on chromosome 6q26. Cytogenet Cell Genet 79: 198-200, 1997.

75. Komatsu T, Hiasa M, Miyaji T, Kanamoto T, Matsumoto T, Otsuka M, Moriyama Y and Omote H. Characterization of the human MATE2 proton-coupled polyspecific organic cation exporter. Int J Biochem Cell Biol 43: 913-918, 2011.

129

76. Kusuhara H, Ito S, Kumagai Y, Jiang M, Shiroshita T, Moriyama Y, Inoue K, Yuasa H and Sugiyama Y. Effects of a MATE protein inhibitor, pyrimethamine, on the renal elimination of metformin at oral microdose and at therapeutic dose in healthy subjects. Clin Pharmacol Ther 89: 837-844, 2011.

77. Kyte J and Doolittle RF. A simple method for displaying the hydropathic character of a protein. J Mol Biol 157: 105-132, 1982.

78. Lacko L and Burger M. Common carrier system for sugar transport in human red cells. Nature 191: 881-882, 1961.

79. Larsen SB, Omkvist DH, Brodin B, Nielsen CU, Steffansen B, Olsen L and Jorgensen FS. Discovery of ligands for the human intestinal Di-/Tripeptide transporter (hPEPT1) using a QSAR-assisted virtual screening strategy. ChemMedChem 4: 1439- 1445, 2009.

80. Leier I, Hummel-Eisenbeiss J, Cui Y and Keppler D. ATP-dependent para- aminohippurate transport by apical multidrug resistance protein MRP2. Kidney Int 57: 1636-1642, 2000.

81. Lickteig AJ, Cheng X, Augustine LM, Klaassen CD and Cherrington NJ. Tissue distribution, ontogeny and induction of the transporters Multidrug and toxin extrusion (MATE) 1 and MATE2 mRNA expression levels in mice. Life Sci 83: 59-64, 2008.

82. Long F, Rouquette-Loughlin C, Shafer WM and Yu EW. Functional cloning and characterization of the multidrug efflux pumps NorM from Neisseria gonorrhoeae and YdhE from Escherichia coli. Antimicrob Agents Chemother 52: 3052-3060, 2008.

83. Lopez-Nieto CE, You G, Bush KT, Barros EJ, Beier DR and Nigam SK. Molecular cloning and characterization of NKT, a gene product related to the organic cation transporter family that is almost exclusively expressed in the kidney. J Biol Chem 272: 6471-6478, 1997.

84. Lu R, Chan BS and Schuster VL. Cloning of the human kidney PAH transporter: narrow substrate specificity and regulation by protein kinase C. Am J Physiol 276: F295- F303, 1999.

85. Lungkaphin A, Chatsudthipong V, Evans KK, Groves CE, Wright SH and Dantzler WH. Interaction of the metal chelator, DMPS, with OAT1 and OAT3 in intact isolated rabbit renal proximal tubules. Am J Physiol Renal Physiol 286: F68-F76, 2004.

86. Magalhaes JV, Liu J, Guimaraes CT, Lana UG, Alves VM, Wang YH, Schaffert RE, Hoekenga OA, Pineros MA, Shaff JE, Klein PE, Carneiro NP, Coelho CM, Trick HN and Kochian LV. A gene in the multidrug and toxic compound extrusion (MATE) family confers aluminum tolerance in sorghum. Nat Genet 39: 1156-1161, 2007.

87. Malo C and Berteloot A. Analysis of kinetic data in transport studies: new insights from kinetic studies of Na+-D-glucose cotransport in human intestinal brush-border membrane vesicles using a fast sampling, rapid filtration apparatus. J Membrane Biol 122: 127-141, 1991.

130

88. Martel F, Keating E, Calhau C, Gründemann D, Schomig E and Azevedo I. Regulation of human extraneuronal monoamine transporter (hEMT) expressed in HEK293 cells by intracellular second messenger systems. Naunyn Schmiedebergs Arch Pharmacol 364: 487-495, 2001.

89. Masuda S, Terada T, Yonezawa A, Tanihara Y, Kishimoto K, Katsura T, Ogawa O and Inui KI. Identification and functional characterization of a new human kidney- specific H+/organic cation antiporter, Kidney-Specific Multidrug and Toxin Extrusion 2. J Am Soc Nephrol 2127-2135, 2006.

90. Matsumoto T, Kanamoto T, Otsuka M, Omote H and Moriyama Y. Role of glutamate residues in substrate recognition by human MATE1 polyspecific H+/organic cation exporter. Am J Physiol Cell Physiol 294: C1074-C1078, 2008.

91. Matsushima S, Maeda K, Inoue K, Ohta KY, Yuasa H, Kondo T, Nakayama H, Horita S, Kusuhara H and Sugiyama Y. The inhibition of human multidrug and toxin extrusion 1 (hMATE1) is involved in the drug-drug interaction caused by cimetidine. Drug Metab Dispos 2008.

92. McKinney TD. Procainamide uptake by rabbit proximal tubules. J Pharmacol Exp Ther 224: 302-306, 1983.

93. Meijer DKF, Mol WEM, Müller M and Kurz G. Carrier-mediated transport in the hepatic distribution and elimination of drugs, with special reference to the category of organic cations. J Pharmacokin Biopharmaceut 18: 35-70, 1990.

94. Meyer zu Schwabedissen HE, Verstuyft C, Kroemer HK, Becquemont L and Kim RB. Human multidrug and toxin extrusion 1 (MATE1/SLC47A1) transporter: functional characterization, interaction with OCT2 (SLC22A2), and single nucleotide polymorphisms. Am J Physiol Renal Physiol 298: F997-F1005, 2010.

95. Meyer-Wentrup F, Karbach U, Gorboulev V, Arndt P and Koepsell H. Membrane localization of the electrogenic cation transporter rOCT1 in rat liver. Biochem Biophys Res Commun 248: 673-678, 1998.

96. Miyamae S, Ueda O, Yoshimura F, Hwang J, Tanaka Y and Nikaido H. A MATE family multidrug efflux transporter pumps out fluoroquinolones in Bacteroides thetaiotaomicron. Antimicrob Agents Chemother 45: 3341-3346, 2001.

97. Montrose-Rafizadeh C, Mingard F, Murer H and Roch-Ramel F. Carrier-mediated transport of tetraethylammonium across rabbit renal basolateral membrane. Am J Physiol 257: F243-F251, 1989.

98. Montrose-Rafizadeh C, Roch-Ramel F and Schäli C. Axial heterogeneity of organic cation transport along the rabbit renal proximal tubule: studies with brush-border membrane vesicles. Biochim Biophys Acta 904: 175-177, 1987.

99. Moseley RH, Morrissette J and Johnson TR. Transport of N1-methylnicotinamide by organic cation-proton exchange in rat liver membrane vesicles. Am J Physiol 259: G973- G982, 1990.

131

100. Motohashi h, Sakurai Y, Saito H, Masuda S, Urakami Y, Goto M, Fukatsu A, Ogawa O and Inui K. Gene expression levels and immunolocalization of organic ion transporters in the human kidney. J Am Soc Nephrol 13: 866-874, 2002.

101. Muller F, Konig J, Glaeser H, Schmidt I, Zolk O, Fromm MF and Maas R. Molecular mechanism of renal tubular secretion of the antimalarial drug chloroquine. Antimicrob Agents Chemother 55: 3091-3098, 2011.

102. Muller M and Jansen PL. Molecular aspects of hepatobiliary transport. Am J Physiol 272: G1285-G1303, 1997.

103. Neuhoff S, Ungell AL, Zamora I and Artursson P. pH-dependent bidirectional transport of weakly basic drugs across Caco-2 monolayers: implications for drug-drug interactions. Pharm Res 20: 1141-1148, 2003.

104. Ohta KY, Imamura Y, Okudaira N, Atsumi R, Inoue K and Yuasa H. Functional characterization of multidrug and toxin extrusion protein 1 as a facilitative transporter for fluoroquinolones. J Pharmacol Exp Ther 328: 628-634, 2009.

105. Ohta KY, Inoue K, Yasujima T, Ishimaru M and Yuasa H. Functional characteristics of two human MATE transporters: kinetics of cimetidine transport and profiles of inhibition by various compounds. J Pharm Pharm Sci 12: 388-396, 2009.

106. Omote H, Hiasa M, Matsumoto T, Otsuka M and Moriyama Y. The MATE proteins as fundamental transporters of metabolic and xenobiotic organic cations. Trends Pharmacol Sci 27: 587-593, 2006.

107. Otsuka M, Matsumoto T, Morimoto R, Arioka S, Omote H and Moriyama Y. A human transporter protein that mediates the final excretion step for toxic organic cations. Proc Natl Acad Sci U S A 102: 17923-17928, 2005.

108. Otsuka M, Yasuda M, Morita Y, Otsuka C, Tsuchiya T, Omote H and Moriyama Y. Identification of essential amino acid residues of the NorM Na+/multidrug antiporter in Vibrio parahaemolyticus. J Bacteriol 187: 1552-1558, 2005.

109. Pelis RM, Dangprapai Y, Wunz TM and Wright SH. Inorganic mercury interacts with cysteine residues (C451 and C474) of hOCT2 to reduce its transport activity. Am J Physiol Renal Physiol 292: F1583-F1591, 2007.

110. PILKINGTON LA and KEYL MJ. STOP-FLOW ANALYSES OF MEPIPERPHENIDOL AND MECAMYLAMINE IN THE DOG. Am J Physiol 205: 471-476, 1963.

111. Popp C, Gorboulev V, Muller TD, Gorbunov D, Shatskaya N and Koepsell H. Amino acids critical for substrate affinity of rat organic cation transporter 1 line the substrate binding region in a model derived from the tertiary structure of lactose permease. Mol Pharmacol 67: 1600-1611, 2005.

112. Pritchard JB and Miller DS. Mechanisms mediating renal secretion of organic anions and cations. Physiol Rev 73: 765-796, 1993.

132

113. Pritchard JB and Miller DS. Intracellular compartmentation of organic anions and cations during renal secretion. Cell Physiol Biochem 6: 50-59, 1996.

114. Race JE, Grassl SM, Williams WJ and Holtzman EJ. Molecular cloning and characterization of two novel human renal organic anion transporters (hOAT1 and hOAT3). Biochem Biophys Res Commun 255: 508-514, 1999.

115. Rafizadeh C, Manganel M, Roch-Ramel F and Schäli C. Transport of organic cations in brush border membrane vesicles from rabbit kidney cortex. Pflugers Arch 407: 404- 408, 1986.

116. Rafizadeh C, Roch-Ramel F and Schäli C. Tetraethylammonium transport in renal brush border membrane vesicles of the rabbit. J Pharmacol Exp Ther 240: 308-313, 1987.

117. Rennick BR and Moe GK. Stop-flow localization of renal tubular excretion of tetraethylammonium. Am J Physiol 198: 1267-1270, 1960.

118. Rennick BR, Moe GK and . Absorption and renal excretion of the tetraethylammonium ion. J Pharmacol Exp Ther 91: 210-217, 1947.

119. Rosenberg MF, Callaghan R, Ford RC and Higgins CF. Structure of the multidrug resistance P-glycoprotein to 2.5 nm resolution determined by electron microscopy and image analysis. J Biol Chem 272: 10685-10694, 1997.

120. Rosenberg MF, Callaghan R, Modok S, Higgins CF and Ford RC. Three-dimensional structure of P-glycoprotein: the transmembrane regions adopt an asymmetric configuration in the nucleotide-bound state. J Biol Chem 280: 2857-2862, 2005.

121. Ryan M.Pelis SHW. Renal Transport of Organic Anions and Cations. 2011, 1: . 2011.

122. Saborowski M, Kullak-Ublick GA and Eloranta JJ. The human organic cation transporter-1 gene is transactivated by hepatocyte nuclear factor-4alpha. J Pharmacol Exp Ther 317: 778-785, 2006.

123. Schali C, Schild L, Overney J and Roch-Ramel F. Secretion of tetraethylammonium by proximal tubules of rabbit kidneys. Am J Physiol 245: F238-F246, 1983.

124. Shu Y, Leabman MK, Feng B, Mangravite LM, Huang CC, Stryke D, Kawamoto M, Johns SJ, DeYoung J, Carlson E, Ferrin TE, Herskowitz I, Giacomini KM and Pharmacogenetics Of Membrane Transporters Investigators. Evolutionary conservation predicts function of variants of the human organic cation transporter, OCT1. Proc Natl Acad Sci U S A 100: 5902-5907, 2003.

125. Smeets PH, van Aubel RA, Wouterse AC, van den Heuvel JJ and Russel FG. Contribution of multidrug resistance protein 2 (MRP2/ABCC2) to the renal excretion of p-aminohippurate (PAH) and identification of MRP4 (ABCC4) as a novel PAH transporter. J Am Soc Nephrol 15: 2828-2835, 2004.

126. Somogyi A and Heinzow B. Cimetidine reduces procainamide elimination. New Eng J Med 307: 1080, 1982.

133

127. Somogyi A, McLean A and Heinzow B. Cimetidine-procainamide pharmacokinetic interaction on man: evidence of competition for tubular secretion of basic drugs. Eur J Clin Pharmacol 25: 339-345, 1983.

128. Somogyi A, Stockley C, Keal J, Rolan P and Bochner F. Reduction of metformin renal tubular secretion by cimetidine in man. Br J clin Pharmacol 23: 545-551, 1987.

129. Somogyi AA, Rumrich G, Fritzsch G and Ullrich KJ. Stereospecificity in contraluminal and luminal transporters of organic cations in the rat renal proximal tubule. J Pharmacol Exp Ther 278: 31-36, 1996.

130. Sturm A, Gorboulev V, Gorbunov D, Keller T, Volk C, Schmitt BM, Schlachtbauer P, Ciarimboli G and Koepsell H. Identification of cysteines in rat organic cation transporters rOCT1 (C322, C451) and rOCT2 (C451) critical for transport activity and substrate affinity. Am J Physiol Renal Physiol 293: F767-F779, 2007.

131. Suhre WM, Ekins S, Chang C, Swaan PW and Wright SH. Molecular determinants of substrate/inhibitor binding to the human and rabbit renal organic cation transporters, hOCT2 and rbOCT2. Mol Pharmacol 67: 1067-1077, 2005.

132. Svoboda M, Riha J, Wlcek K, Jaeger W and Thalhammer T. Organic anion transporting polypeptides (OATPs): regulation of expression and function. Curr Drug Metab 12: 139-153, 2011.

133. Tamai I, Ohashi R, Nezu J, Yabuuchi H, Oku A, Shimane M, Sai Y and Tsuji A. Molecular and functional identification of sodium ion-dependent, high affinity human carnitine transporter OCTN2. J Biol Chem 273: 20378-20382, 1998.

134. Tamai I, Yabuuchi H, Nezu J-I, Sai Y, Oku A, Shimane M and Tsuji A. Cloning and characterization of a novel human pH-dependent organic cation transporter, OCTN1. FEBS Letters 419: 107-111, 1997.

135. Tanihara Y, Masuda S, Katsura T and Inui K. Protective effect of concomitant administration of imatinib on cisplatin-induced nephrotoxicity focusing on renal organic cation transporter OCT2. Biochem Pharmacol 78: 1263-1271, 2009.

136. Tanihara Y, Masuda S, Sato T, Katsura T, Ogawa O and Inui KI. Substrate specificity of MATE1 and MATE2-K, human multidrug and toxin extrusions/H(+)- organic cation antiporters. Biochem Pharmacol 2007.

137. Tanihara Y, Masuda S, Sato T, Katsura T, Ogawa O and Inui KI. Substrate specificity of MATE1 and MATE2-K, human multidrug and toxin extrusions/H(+)- organic cation antiporters. Biochem Pharmacol 2007.

138. Terada T, Masuda S, Asaka JI, Tsuda M, Katsura T and Inui KI. Molecular Cloning, Functional Characterization and Tissue Distribution of Rat H(+)/Organic Cation Antiporter MATE1. Pharm Res 23: 1696-1701, 2006.

134

139. Thomas MC, Tikellis C, Kantharidis P, Burns WC, Cooper ME and Forbes JM. The role of advanced glycation in reduced organic cation transport associated with experimental diabetes. J Pharmacol Exp Ther 311: 456-466, 2004.

140. Toyama K, Yonezawa A, Tsuda M, Masuda S, Yano I, Terada T, Osawa R, Katsura T, Hosokawa M, Fujimoto S, Inagaki N and Inui K. Heterozygous variants of multidrug and toxin extrusions (MATE1 and MATE2-K) have little influence on the disposition of metformin in diabetic patients. Pharmacogenet Genomics 20: 135-138, 2010.

141. Tsuda M, Terada T, Asaka J, Ueba M, Katsura T and Inui K. Oppositely directed H+ gradient functions as a driving force of rat H+/organic cation antiporter MATE1. Am J Physiol Renal Physiol 292: F593-F598, 2007.

142. Tsuda M, Terada T, Asaka J, Ueba M, Katsura T and Inui K. Oppositely directed H+ gradient functions as a driving force of rat H+/organic cation antiporter MATE1. Am J Physiol Renal Physiol 292: F593-F598, 2007.

143. Tsuda M, Terada T, Mizuno T, Katsura T, Shimakura J and Inui KI. Targeted disruption of the multidrug and toxin extrusion 1 (Mate1) gene in mice reduces renal secretion of metformin. Mol Pharmacol 2009.

144. Tzvetkov MV, Vormfelde SV, Balen D, Meineke I, Schmidt T, Sehrt D, Sabolic I, Koepsell H and Brockmoller J. The effects of genetic polymorphisms in the organic cation transporters OCT1, OCT2, and OCT3 on the renal clearance of metformin. Clin Pharmacol Ther 86: 299-306, 2009.

145. Ullrich KJ, Papavassiliou F, David C, Rumrich G and Fritzsch G. Contraluminal transport of organic cations in the proximal tubule of the rat kidney. I. Kinetics of N1- + methylnicotinamide and tetraethylammonium, influence of K , HCO3, pH; inhibition by aliphatic primary, secondary and tertiary amines, and mono- and bisquaternary compounds. Pflugers Arch 419: 84-92, 1991.

146. Ullrich KJ and Rumrich G. Luminal transport system for choline+ in relation to the other organic cation transport systems in the rat proximal tubule. Kinetics, specificity: alkyl/arylamines, alkylamines with OH, O, SH, NH2, ROCO, RSCO and H2PO4-groups, methylaminostyryl, rhodamine, acridine, phenanthrene and cyanine compounds. Pflügers Arch 432: 471-485, 1996.

147. Ullrich KJ, Rumrich G, David C and Fritzsch G. Bisubstrates: substances that interact with both, renal contraluminal organic anion and organic cation transport systems. II. Zwitterionic substrates: dipeptides, cephalosporins, quinolone-carboxylate gyrase inhibitors and phosphoamide thiazine carboxylates; nonionizable substrates: steroid hormones and cyclophosphamides. Pflugers Arch 425: 300-312, 1993.

148. Ullrich KJ, Rumrich G, David C and Fritzsch G. Bisubstrates: substances that interact with renal organic anion and organic cation transport systems. I. Amines, piperidines, piperazines, azepines, pyridines, quinolines, imadazoles, thiazoles, guanidines, and hydrazines. Pflugers Arch 425: 280-299, 1993.

135

149. Ullrich KJ, Rumrich G and Fritzsch G. Contraluminal transport of organic cations in the proximal tubule of the rat kidney. II. Specificity: anilines, phenylalkylamines (catecholamines), heterocyclic compounds (pyridines, quinolines, acridines). Pflugers Arch 420: 29-38, 1992.

150. Urakami Y, Kimura N, Okuda M, Masuda S, Katsura T and Inui K. Transcellular transport of creatinine in renal tubular epithelial cell line LLC-PK1. Drug Metab Pharmacokinet 20: 200-205, 2005.

151. Urakami Y, Nakamura N, Takahashi K, Okuda M, Saito H, Hashimoto Y and Inui K. Gender differences in expression of organic cation transporter OCT2 in rat kidney. FEBS Lett 461: 339-342, 1999.

152. Urakami Y, Okuda M, Saito H and Inui K. Hormonal regulation of organic cation transporter OCT2 expression in rat kidney. FEBS Lett 473: 173-176, 2000.

153. van Aubel RA, Peters JG, Masereeuw R, van Os CH and Russel FG. Multidrug resistance protein mrp2 mediates ATP-dependent transport of classic renal organic anion p-aminohippurate. Am J Physiol Renal Physiol 279: F713-F717, 2000.

154. Vander AJ. Renal excretion of choline in the dog. Am J Physiol 202: 319-324, 1962.

155. Volk C, Gorboulev V, Budiman T, Nagel G and Koepsell H. Different affinities of inhibitors to the outwardly and inwardly directed substrate binding site of organic cation transporter 2. Mol Pharmacol 64: 1037-1047, 2003.

156. Volk C, Gorboulev V, Kotzsch A, Mueller TD and Koepsell H. Five Amino Acids in the Innermost Cavity of the Substrate Binding Cleft of Organic Cation Transporter 1 Interact with Extracellular and Intracellular Corticosterone. Mol Pharmacol 2009.

157. Watanabe S, Tsuda M, Terada T, Katsura T and Inui KI. Reduced renal clearance of a zwitterionic substrate cephalexin in Mate1-deficient mice. J Pharmacol Exp Ther 2010.

158. Wermuth C, anellin CR, indberg P and itscher LA. Glossary of terms used in medicinal chemistry (IUPAC Recommendations 1998). Pure and Applied Chemistry 70: 1129-1143, 1998.

159. Wright SH. Role of organic cation transporters in the renal handling of therapeutic agents and xenobiotics. Toxicol Appl Pharmacol 204: 309-319, 2005.

160. Wright SH and Dantzler WH. Molecular and cellular physiology of renal organic cation and anion transport. Physiol Rev 84: 987-1049, 2004.

161. Wright SH, Evans KK, Zhang X, Cherrington NJ, Sitar DS and Dantzler WH. Functional map of TEA transport activity in isolated rabbit renal proximal tubules. Am J Physiol Renal Physiol 287: F442-F451, 2004.

162. Wright SH and Wunz TM. Transport of tetraethylammonium by rabbit renal brush- border and basolateral membrane vesicles. Am J Physiol 253: F1040-F1050, 1987.

136

163. Wright SH and Wunz TM. Mechanism of cis- and trans-substrate interactions at the tetraethylammonium/H+ exchanger of rabbit renal brush-border membrane vesicles. J Biol Chem 263: 19494-19497, 1988.

+ 164. Wright SH and Wunz TM. Correlation between Jmax and Kt of organic cation/H exchange in renal brush border membranes. FASEB J 4: A728, 1990.

165. Wright SH and Wunz TM. Influence of substrate structure on turnover of the renal organic cation/H+ exchanger of the renal luminal membrane. Pflügers Arch 436: 469-477, 1998.

166. Wright SH and Wunz TM. Influence of substrate structure on substrate binding to the renal organic cation/H+ exchanger. Pflügers Arch 437: 603-610, 1999.

167. Wright SH, Wunz TM and Wunz TP. Structure and interaction of inhibitors with the TEA/H+ exchanger of rabbit renal brush border membranes. Pflugers Arch 429: 313-324, 1995.

168. Wu X, Huang W, Prasad PD, Seth P, Rajan DP, Leibach FH, Chen J, Conway SJ and Ganapathy V. Functional characteristics and tissue distribution pattern of organic cation transporter 2 (OCTN2), an organic Cation/Carnitine transporter. J Pharmacol Exp Ther 290: 1482-1492, 1999.

169. Wu X, Prasad PD, Leibach FH and Ganapathy V. cDNA sequence, transport function, and genomic organization of human OCTN2, a new member of the organic cation transporter family. Biochem Biophys Res Commun 246: 589-595, 1998.

170. Xu XJ, Su XZ, Morita Y, Kuroda T, Mizushima T and Tsuchiya T. Molecular cloning and characterization of the HmrM multidrug efflux pump from Haemophilus influenzae Rd. Microbiol Immunol 47: 937-943, 2003.

171. Yabuuchi H, Tamai I, Nezu JI, Sakamoto K, Oku A, Shimane M, Sai Y and Tsuji A. Novel membrane transporter OCTN1 mediates multispecific, bidirectional, and pH- dependent transport of organic cations. J Pharmacol Exp Ther 289: 768-773, 1999.

172. Yasujima T, Ohta KY, Inoue K, Ishimaru M and Yuasa H. Evaluation of 4',6- diamidino-2-phenylindole as a fluorescent probe substrate for rapid assays of the functionality of human multidrug and toxin extrusion proteins. Drug Metab Dispos 38: 715-721, 2010.

173. Yokoo S, Yonezawa A, Masuda S, Fukatsu A, Katsura T and Inui K. Differential contribution of organic cation transporters, OCT2 and MATE1, in platinum agent- induced nephrotoxicity. Biochem Pharmacol 74: 477-487, 2007.

174. Yokoo S, Yonezawa A, Masuda S, Fukatsu A, Katsura T and Inui K. Differential contribution of organic cation transporters, OCT2 and MATE1, in platinum agent- induced nephrotoxicity. Biochem Pharmacol 74: 477-487, 2007.

137

175. Yonezawa A and Inui KI. Importance of the Multidrug and Toxin Extrusion MATE/SLC47A Family to Pharmacokinetics, Pharmacodynamics/Toxicodynamics and Pharmacogenomics. Br J Pharmacol 2011.

176. Zakharyan RA, Wildfang E and Aposhian HV. Enzymatic methylation of arsenic compounds. III. The marmoset and tamarin, but not the rhesus, monkeys are deficient in methyltransferases that methylate inorganic arsenic. Toxicol Appl Pharmacol 140: 77-84, 1996.

177. Zhang X, Cherrington NJ and Wright SH. Molecular Identification and Functional Characterization of Rabbit MATE1 and MATE2-K. Am J Physiol Renal Physiol 2007.

178. Zhang X, Evans KK and Wright SH. Molecular cloning of rabbit organic cation transporter rbOCT2 and functional comparisons with rbOCT1. Am J Physiol Renal Physiol 283: F124-F133, 2003.

179. Zhang X and Wright SH. MATE1 has an external COOH terminus, consistent with a 13-helix topology. Am J Physiol Renal Physiol 297: F263-F271, 2009.

180. Zheng X, Ekins S, Raufman JP and Polli JE. Computational models for drug inhibition of the human apical sodium-dependent bile acid transporter. Mol Pharm 6: 1591-1603, 2009.

181. Zolk O, Solbach TF, Konig J and Fromm MF. Structural determinants of inhibitor interaction with the human organic cation transporter OCT2 (SLC22A2). Naunyn Schmiedebergs Arch Pharmacol 379: 337-348, 2008.