Altered Metabolism of Daunorubicin and Doxorubicin by Genetic Variants of Human Aldo-Keto and Carbonyl Reductases

ALTERED METABOLISM OF DAUNORUBICIN AND DOXORUBICIN BY GENETIC VARIANTS OF HUMAN ALDO-KETO AND CARBONYL REDUCTASES

Onkar Singh Bains

M.E.T., Simon Fraser University, 2004 B.Sc. (Biology), Simon Fraser University, 2001

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

The Faculty of Graduate Studies

(Pharmaceutical Sciences)

THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)

October 2010

ABSTRACT

The anthracyclines, doxorubicin (DOX) and daunorubicin (DAUN) are commonly used to treat a variety of cancers. Their use is associated with life-threatening adverse events, especially chronic cardiotoxicity, in some patients. There may be a genetic basis for this variation, arising from altered metabolism by non-synonymous single nucleotide polymorphisms (ns-SNPs) in genes encoding for the aldo-keto reductase (AKR) and carbonyl reductase (CBR) enzymes, which are responsible for the biotransformation of these two drugs.

The first two studies (Chapters 2 and 3) of this thesis examined the effect of ns-

SNPs in 8 AKR and 3 CBR genes on the in vitro metabolism of both anthracyclines to their major metabolites, doxorubicinol (DOXol) and daunorubicinol (DAUNol) using purified, human wild-type and genetic variant enzymes. Michaelis-Menten kinetic curves were plotted and the metabolic capacities of the wild-type and variant enzymes were compared using catalytic efficiency ( kcat /Km). In the presence of DAUN, 7 AKR and 5

CBR variants exhibited significantly reduced metabolic activity while 3 AKR and 5 CBR

variants demonstrated significantly reduced activity with DOX as substrate. These

findings suggest that genetic variants of human AKRs and CBRs are capable of

decreasing the in vitro metabolism of DOX and DAUN.

There is considerable controversy in the literature on how DAUN and DOX

contribute to the variable adverse events seen in patients treated with these drugs. Some

studies suggest that the toxic species are the major metabolites, DAUNol and DOXol,

while others suggest the parent drug is more toxic. To study this, I examined whether a

strong and consistent association exists between metabolic activity and drug toxicity

among nine cell lines from different tissues (Chapter 4). My findings indicated that there

ii is a strong, and inversely proportional, association between cytotoxicity and DAUN or

DOX metabolism. Furthermore, the cell lines that were resistant to the toxic effects of these drugs had significantly greater expression of the AKRs and CBRs.

Overall, these data provide a foundation of biochemical evidence to design in vivo studies that will elucidate the role of altered metabolism by genetic variants of human

AKRs and CBRs in the development of anthracycline-induced cardiotoxicity.

iii TABLE OF CONTENTS

ABSTRACT...... ii TABLE OF CONTENTS...... iv LIST OF TABLES...... vii LIST OF FIGURES ...... ix LIST OF ABBREVIATIONS...... xiii ACKNOWLEDGEMENTS...... xvi DEDICATION...... xviii PREFACE...... xix Co-authorship...... xix Publications arising from work presented in this thesis...... xx

CHAPTER 1: Overview ...... 1 1.1 Anthracyclines—chemical structure and mechanism of action...... 2 1.2 DAUN and DOX—their use in cancer treatment and cardiac side effects...... 4 1.3 Proposed mechanisms of DAUN/DOX-induced cardiotoxicity...... 9 1.4 The phases of biotransformation and DOX/DAUN metabolism...... 13 1.5 Aldo-keto reductases and carbonyl reductases ...... 20 1.6 Genetic polymorphisms in genes producing drug metabolizing enzymes...... 24 1.7 Rationale...... 34 1.8 Research hypotheses and predictions...... 35

CHAPTER 2: Altered metabolism of daunorubicin and doxorubicin by allelic variants of human aldo-keto reductases...... 38 2.1 Preface...... 38 2.2 Materials and Methods...... 42 2.2.1 Chemicals and enzymes...... 42 2.2.2 Molecular cloning of human AKR genes and creation of allelic variants 43 2.2.3 Expression and purification of recombinant human AKR wild-type and variant enzymes ...... 47 2.2.4 Rate of enzymatic activity of AKRs in presence of test substrates ...... 50 2.2.5 Kinetic activity of AKR enzymes in the presence of anthracyclines...... 51 2.2.6 Statistical analysis...... 53 2.3 Results...... 53 2.3.1 Expression and purification of the AKR enzymes...... 53 2.3.2 Enzyme activities of AKR wild-type and variant enzymes using test substrates...... 55 2.3.3 Kinetic characterization of wild-type and variant enzymatic activities with DOX and DAUN as substrates ...... 59 2.4 Discussion...... 68

CHAPTER 3: Allelic variants of human carbonyl reductases demonstrate reduced in vitro metabolism of daunorubicin and doxorubicin ...... 76 3.1 Preface...... 76 3.2 Methods...... 79 3.2.1 Chemicals and enzymes...... 79

iv 3.2.2 Molecular cloning of human CBR genes and creation of the genetic variants...... 80 3.2.3 Expression and purification of recombinant human CBR wild-type and variant enzymes ...... 82 3.2.4 Kinetic analysis of CBR wild-type and variant enzymes ...... 84 3.2.5 Statistical analysis...... 85 3.3 Results...... 86 3.3.1 Expression and purification of the human CBRs...... 86 3.3.2 Kinetic characterization of CBR wild-type and variant enzymatic activities with menadione...... 87 3.3.3 Kinetic characterization of wild-type and variant enzymatic activities with DOX and DAUN...... 91 3.4 Discussion...... 96

CHAPTER 4: Ex vivo and in vitro studies suggest a negative association between cytotoxicity and metabolism of the anthracyclines daunorubicin and doxorubicin by aldo- keto and carbonyl reductases ...... 103 4.1 Preface...... 103 4.2 Methods...... 106 4.2.1 Chemicals and enzymes...... 106 4.2.2 Cell culture...... 106 4.2.3 MTT cell viability assay to measure cytotoxicity...... 107 4.2.4 DAUN and DOX m etabolic assays with cytosolic fractions derived from cell lines ...... 108 4.2.5 Instrumentation and experimental conditions for UPLC-MS/MS...... 109 4.2.6 Western blotting of cytosols to detect AKRs and CBRs ...... 111 4.2.7 Statistical analysis...... 112 4.3 Results...... 113 4.3.1 Cytotoxicity of cells following exposure to anthracyclines...... 113 4.3.2 Metabolism of the anthracyclines using cell line cytosols...... 116 4.3.3 Expression of cytosolic AKRs and CBRs in cell lines ...... 118 4.3.4 Increase in DAUN/DOX metabolism and induction of AKR and CBR gene expression are responses in cell lines following exposure to anthracycline drugs ...... 121 4.3.5 Association between cytotoxicity and DOX/DAUN metabolism in cell lines exposed to these anthracyclines ...... 127 4.3.6 UPLC-MS/MS method validation ...... 127 4.4 Discussion...... 131

CHAPTER 5: Overall summary and conclusions...... 137 5.1 Novel findings...... 137 5.2 Implications of findings...... 139 5.3 Potential limitations of research...... 141 5.3.1 The use of a bacterial expression system for the in vitro studies...... 141 5.3.2 Focusing solely on the polymorphisms in the coding regions of the AKR and CBR genes...... 145

v 5.3.3 Allele frequencies for two variants are not reported in the NCBI database ...... 146 5.4 Suggested future research directions ...... 146 5.4.1 Biochemical investigations...... 147 5.4.2 Cancer patient correlation studies to determine if AKR and CBR variant alleles are genetic biomarkers of cardiotoxicity arising from DAUN/DOX therapy...... 153

REFERENCES ...... 157 APPENDICES ...... 180

vi LIST OF TABLES

Table 1—The human AKR proteins within the AKR1, AKR6, and AKR7 families identified in the AKR Superfamily website along with the location of the gene in the chromosome...... 21

Table 2—The human CBR proteins along with the location of the gene in the chromosome...... 23

Table 3—A summary of studies that looked at the effect of ns-SNPs in genes for different drug metabolizing enzymes...... 31

Table 4—Allele frequencies of the non-synonymous single nucleotide polymorphic variants of human AKR enzymes from different ethnic groups...... 40

Table 5—Enzymatic rates for reported test substrates by recombinant 6x-His tagged AKR wild-type and variant allele enzymes...... 56

Table 6—Kinetic constants for DAUN metabolism by recombinant 6x-His tagged AKR wild-type and variant enzymes...... 64

Table 7—Kinetic constants for DOX metabolism by recombinant 6x-His tagged AKR wild-type and variant enzymes...... 67

Table 8—Allele frequencies of the non-synonymous single nucleotide polymorphic variants of human CBR enzymes from different ethnic groups...... 78

Table 9—Michaelis-Menten kinetic parameters for test substrate menadione by recombinant 6x-His tagged CBR wild-type and variant allele enzymes...... 89

Table 10—Kinetic constants for DAUN metabolism by recombinant 6x-His tagged CBR wild-type and variant enzymes...... 94

Table 11—Kinetic constants for DOX metabolism by recombinant 6x-His tagged CBR wild-type and variant enzymes...... 95

Table 12—Mean LC 50 values of DAUN and DAUNol for cell lines at specified time intervals following exposure to the drug...... 115

Table 13—Mean LC 50 values of DOX and DOXol for cell lines at specified time intervals following exposure to the drug...... 116

Table 14—Results for accuracy and precision (intra-day and inter-day) for the UPLC- MS/MS determinations of DOXol and DAUNol in pooled matrix of nine cell lines (days 1 to 3)...... 130

vii Table 15—Enzymatic rates for reported test substrates by recombinant 6x-His tagged AKR wild-type along with kinetic parameters for CBR1 wild-type in the presence of menadione...... 145

Table 16—Allele frequencies of the non-synonymous single nucleotide polymorphic variants of human AKR enzymes from different ethnic groups...... 147

viii LIST OF FIGURES

Figure 1—Chemical structure of an anthracycline illustrating the fused tetracycline ring (denoted by A, B, C, and D) as as well the daunosamine sugar...... 2

Figure 2—Chemical structures of (A) DOX and (B) DAUN...... 5

Figure 3—Chemical structures of anthracycline analogs epirubicin, idarubicin, and aclarubicin...... 8

Figure 4—DOX-related redox cycling generation of ROS and reactive nitrogen species (RNS) by cellular enzymatic mechanisms...... 12

Figure 5—A simplified overview illustrating Phase I, II, and III metabolism of aspirin occurring in a cell...... 14

Figure 6—Schematic representative of DOX metabolism...... 17

Figure 7—The location of AKRs and CBRs within all existing categories of metabolic enzymes...... 19

Figure 8—Schematic representations of the different human genetic polymorphisms: missense, nonsense, insertion, deletion, duplication and frameshift...... 26

Figure 9—A schematic representation of a gene comprised of exons and introns, as well as other regulatory regions of DNA that control gene expression (promoter, silencer and enhancer)...... 28

Figure 10—(Top ) Michaelis-Menten kinetic curve plotting metabolic reaction activity rate against substrate concentration. ( Bottom ) Kinetic parameter values and how they are affected with low and high metabolizing enzymes...... 30

Figure 12—Interaction between neighboring histidine residues in the 6x-His tag and the nickel-nitrilotriacetic acid (Ni-NTA) matrix ...... 49

Figure 13—Purification of human recombinant 6x-His-tagged wild-type enzymes: (A) AKRA1, (B) AKR1B1, (C) AKR1B10, (D) AKR1C1, (E) AKR1C2, (F) AKR1C3, (G) AKR1C4, and (H) AKR7A2...... 54

Figure 14—Representative Western blot detection of purified 6x-His tagged AKR proteins [(A) AKR1C3, (B) AKR1C4, and (C) AKR7A2 wild-type enzymes; lane 2; 3 µg] and native AKR proteins...... 58

ix Figure 15—Generation of DAUNol and DOXol in vitro by purified AKR incubated with DOX (top panel) and DAUN (bottom panel)...... 61

Figure 18—Three-dimensional molecular structures of human (A) AKR1A1, (B) AKR1C3, (C) AKR1C4, and (D) AKR7A2 wild-type enzymes...... 71

Figure 19—Proposed catalytic mechanism for AKR-mediated reduction reaction [modified based on the scheme provided by Barski et al. (2008)] ...... 72

Figure 21—Purification of human recombinant 6x-His-tagged wild-type enzymes: (A) CBR1, (B) CBR3, and (C) CBR4...... 87

Figure 22—Representative Western blot detection of purified 6x-His tagged CBR proteins [(A) CBR1, (B) CBR4, and (C) CBR4 wild-type enzymes] and native AKR proteins...... 91

Figure 23—Generation of DOXol and DAUNol in vitro by purified CBR3 incubated with DOX (top panel), and purified CBR1 incubated with DAUN (bottom panel)...... 92

Figure 24 —In vitro enzymatic activities for the purified 6x-His tagged (A) CBR1, (B) CBR3, and (C) CBR4 wild-type and variant enzymes with daunorubicin...... 93

Figure 25—In vitro enzymatic activities for the purified 6x-His tagged (A) CBR1, (B) CBR3, and (C) CBR4 wild-type and variant enzymes with with doxorubicin...... 95

Figure 27—A proposed catalytic mechanism for CBR-mediated reduction reaction (see text for details) ...... 100

Figure 28—Representative UPLC-MS/MS chromatogram of doxorubicin, doxorubicinol, daunorubicin, daunorubicinol, and idarubicin...... 111

x Figure 29—Sample dose-response curves of (A) H9c2 rat heart cell line and (B) HepG2 liver cell line in the presence of varying concentrations of daunorubicin (DAUN), doxorubicin (DOX), daunorubicinol (DAUNol), and doxorubicinol (DOXol) following a 48 hr incubation...... 114

Figure 30—Enzymatic activities for rat and human cell line cytosols incubated with 10 µM of daunorubicin (A) and doxorubicin (B)...... 117

Figure 31—Western blot detection of the sensitive cell line cytosols (lane 1: heart, lane 2: prostate, lane 3: ovary, lane 4: pancreas, and lane 5: breast) as well as the resistant cell line cytosols (lane 6: liver, lane 7: colon, lane 8: lung, lane 9: kidney) confirms expression of the desired AKR and CBR proteins...... 119

Figure 32—Relative expression ratio levels of AKR and CBR proteins in the sensitive (heart, prostate, ovary, pancreas, and breast) as well as the resistant cell line cytosols (liver,colon, lung, and kidney)...... 120

Figure 36—Bar graphs illustrating the relative abundance of the cytosolic AKRs and CBRs in human heart lysate (H) and pooled human liver lysate (L). The relative abundance values of the individual reductases are ordered from highest to lowest kcat /Km values towards DAUN (A) and DOX (B), which were determined from the in vitro studies with the recombinant enzymes. The reductase enzymes in boxes represent those with variants that significantly reduce DAUN or DOX metabolism compared to their respective wild-type enzymes. ( Below bar graphs ) Lysates (20 g total protein) were run on 18% SDS-PAGE gels and subjected to Western blotting for detection of AKR1A1, AKR1B1, AKR1B10, AKR1C1, AKR1C2, AKR1C3, AKR1C4, AKR7A2, CBR1, and CBR3 proteins...... 133

xi Figure 37—Purification of human recombinant 6x-His-tagged wild-type enzymes for Sf9 insect cells: (1) AKRA1, (2) AKR1B1, (3) AKR1B10, (4) AKR1C2, (5) AKR1C4, (6) AKR7A2, and (7) CBR1...... 144

Figure 38—A spin trapping chemical reaction in which 5-Diethoxyphosphoryl-5-methyl- 1-pyrroline N-oxide (DEPMPO), the spin trapping agent, reacts with a free radical to form a radical adduct that is more stable than the initial radical (Swartz et al., 2007)... 150

xii LIST OF ABBREVIATIONS

6x-His Six histidine oC Degree Celsius xg Times gravity (centrifugal force) α Alpha β Beta µg Microgram µl Microlitre µM Micromolar A Adenine ABC ATP-binding cassette ACN Acetonitrile AKR Aldo-keto reductase ANOVA Analysis of variance C Cytosine CBR Carbonyl reductase CE Collision energy C.I. Confidence interval C.V. Coefficient of variation CYP Cytochrome-P450 DAUN Daunorubicin DAUNol Daunorubicinol dB Decibel DEPMPO 5-Diethoxyphosphoryl-5-methyl-1-pyrroline N-oxide DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid DOX Doxorubicin DOXol Doxorubicinol E. coli Escherichia coli EPR Electrón paramagnetic resonance g Gram G Guanine GHz Gigahertz H2O2 Hydrogen peroxide HPLC High performance liquid chromatography hr Hour Hz Hertz i.d. Internal diameter IPTG Isopropyl β-D-thiogalactopyranoside k Kilo kDa Kilodalton kcat Substrate turnover rate kcat /Km Catalytic efficiency Km Substrate affinity – the substrate concentration at which the reaction rate reaches half of its maximum value kHz Kilohertz

xiii kV Kilovolt KH 2PO 4 Potassium dihydrogen phosphate LC 50 Lethal Concentration 50—the concentration which causes 50% death LOQ Limit of quantitation m Metre M Molar (mole/litre) ml Millilitre mM Millimolar min Minute MRM Multiple reaction monitoring MS/MS Tandem mass spectrometry msec Millisecond MTT 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide m/z Mass to charge ratio n Number of subjects or samples N/A Not available NADP + β-Nicotinamide adenine dinucleotide phosphate NADPH β-Nicotinamide adenine dinucleotide phosphate, reduced form NaH 2PO 4 Sodium dihydrogen phosphate NCBI National Centre for Biotechnology Information nM Nanomolar nmol Nanomole ns-SNP Non-synonymous single nucleotide polymorphism - ·O 2 Superoxide anion ·OH Hydroxyl radical p Statistical probability (of obtaining a result at least as extreme as the one that was observed) PBS Phosphate-buffered saline PCR Polymerase chain reaction pH Negative logarithm of hydrogen ion concentration pmol Picomole POR Cytochrome P450 reductase QC Quality control r Corelation coefficient r2 Coefficient of determination ROS Reactive oxygen species s second S.D. Standard deviation SDR Short-chain dehydrogenases/reductase SDS-PAGE Sodium dodecyl sulphate polyacrylamide gel electrophoresis sec Second SNP Single nucleotide polymorphism T Thymine TIC Total ion current U Uracil UPLC Ultra-high performance liquid chromatography V Volt

xiv v/v Volume to volume ratio Vmax Maximal rate of velocity W Watt

Amino acid one letter abbreviations

A Alanine (Ala) C Cysteine (Cys) D Aspartic acid (Asp) E Glutamic acid (Glu) F Phenylalanine (Phe) G Glycine (Gly) H Histidine (His) I Isoleucine (Ile) K Lysine (Lys) L Leucine (Leu) M Methionine (Met) N Asparagine (Asn) P Proline (Pro) Q Glutamine (Gln) R Arginine (Arg) S Serine (Ser) T Threonine (Thr) V Valine (Val) W Tryptophan (Trp) Y Tyrosine (Tyr)

Variant nomenclature (using two variants of AKR1A1 as an example)

N52S A change from asparagine to serine at amino acid position 52 E55D A change from glutamic acid to aspartic acid at amino acid position 55

NCBI population groups

CEU Utah residents with ancestry from northern and western Europe, collected by Centre d'Etude du Polymorphisme Humain (CEPH) CHB Han Chinese in Beijing, China JPT Japanese in Tokyo, Japan YRI Yoruba in Ibadan, Nigeria HSP Hispanic PRH Pacific Rim Heritage UFV Combination of residents from Utah, as well as French and Venezuelan X Combined population from Michigan along with UFV MITO MITOGPOP6: combination of Caucasian, Japanese, Chinese, Amerindian, Pygmy, Russian, etc.

xv ACKNOWLEDGEMENTS

First of all, I owe my deepest gratitude to my principal supervisor, Dr. K. Wayne

Riggs, for accepting me as his graduate student in this project, which focused on an area that was vastly different from my background in biology and environmental toxicology. I will never forget the encouragement and support he provided during my tenure as a Ph.D. student. Thank you to my co-supervisor, Dr. Ronald E. Reid, and Dr. Thomas A.

Grigliatti for their constructive feedback during lab meetings, professional attitude, and guiding me to becoming a better critical thinker in science. In addition, this thesis would not have been possible without the expert advice and amazing support by my supervisory committee members, Dr. Judy Wong, Dr. Pierre Kennepohl, and Dr. Stelvio Bandiera.

I am forever grateful to Dr. Tom Pfeifer, Dr. Ryan H. Takahashi, Dr. Joanna M.

Lubieniecka, and Dr. Randy Mottus for their valuable advice on ways to strengthen my research project and teaching me scientific techniques in molecular biology.

Furthermore, it is an honor for me to have worked with Dr. Andras Szeitz for his technical assistance with HPLC and UPLC/MS/MS, as well as our entertaining conversations we had about our favorite NHL hockey team, the Vancouver Canucks. I would like to show my gratitude to the past members of the Grigliatti lab for their technical assistance and advice on the project, as well as their support during the lab meetings: Omid Toub, Dr. Pamela Kalas, Dr. Greg Doheny, and Madalene Earp. In addition, I thank Dr. Vlad-Martin Diaconescu for assisting with the electron paramagnetic resonance (EPR) part of my studies.

I express many thanks to my colleagues and friends in the Faculty of

Pharmaceutical Sciences for making my five years unforgettable. Also, thank you to the incredible and hardworking staff that made my journey in completing this thesis much

xvi easier: Suzana Topic, Barb Conway, Jamal Kurtu, Joan Cosar, Wes Wong, Violet Yuen,

Marie Langton, Josie Lim, Arti Maharaj, Sandra Stallberg, and June Chow.

I am deeply grateful for having family and friends that supported me in completing this project. Thank you to my parents (Mr. Gian Singh Bains and Mrs.

Joginder Kaur Bains), brothers (Sukhjit and Sunny), sister-in-law (Mandeep), grandparents, and all of my uncles and aunts that encouraged me to work hard and focus on finishing my Ph.D. degree. Also, thank you to my friends for your heartfelt encouragement and positivity, especially when times were tough.

Personal financial support was gratefully accepted from the Canadian Institutes of

Health Research (CIHR) and the Faculty of Graduate Studies at the University of British

Columbia. Project funding for the project was received from CIHR.

xvii DEDICATION

To my parents

xviii PREFACE

Co-authorship

This thesis includes five manuscripts (four published, one submitted) for which I am co-first author for one and first author for the rest. In relation to the findings with

AKR1A1 (Chapter 2), equal work was performed by Dr. Ryan Takahashi and myself towards the research and preparation of the manuscript. Dr. Takahashi had established the HPLC-fluorescence detection assay method of detecting the doxorubicinol (DOXol) and daunorubicinol (DAUNol) metabolites before I started my Ph.D. research project.

For the studies involving the carbonyl reductases (CBRs) in Chapter 3, I was assisted by a summer research student, Miss. Morgan J. Karkling. I had created the prokaryotic expression constructs of the wild-type and variant CBR1 and CBR3 enzymes, and we were both responsible for purification of the CBR enzymes from bacteria and enzyme kinetic analyses using daunorubicin (DAUN) and doxorubicin (DOX). I analyzed the data arising from the CBR1 and CBR3 studies and prepared the manuscripts, which have been published.

In the last study in Chapter 4, Dr. Joanna M. Lubieniecka taught me proper techniques for using mammalian cell lines and we both participated in the experimental design. Also, Dr. András Szeitz and I developed the UPLC-MS/MS analytical method to measure the concentrations of DAUNol and DOXol in the metabolic assays.

In recognition of the significant contributions of the aforementioned individuals in my studies, they will be co-authors in future manuscripts, which are ready for submission. Other than the exceptions provided above, the contributions of the co- authors in all the manuscripts were through their involvement in intellectual discussion regarding research design and manuscript preparation.

xix

Publications arising from work presented in this thesis

1. Bains OS, Takahashi RH, Pfeifer TA, Grigliatti TA, Reid RE, and Riggs KW (2008) Two allelic variants of aldo-keto reductase 1A1 exhibit reduced in vitro metabolism of daunorubicin. Drug Metab Dispos 36(5): 904-910.

2. Bains OS, Karkling MJ, Grigliatti TA, Reid RE, and Riggs KW (2009) Two non- synonymous single nucleotide polymorphisms of carbonyl reductase 1 demonstrate reduced in vitro metabolism of daunorubicin and doxorubicin. Drug Metab Dispos 37: 1107-1114.

3. Bains OS, Karkling MJ, Grigliatti TA, Reid RE, and Riggs KW (2010) Naturally occurring variants of human CBR3 alter anthracycline in vitro metabolism. J Pharmacol Exp Ther 332(3): 755-763.

4. Bains OS, Grigliatti TA, Reid RE, and Riggs KW (2010) Naturally occurring variants of human aldo-keto reductases with reduced in vitro metabolism of daunorubicin and doxorubicin. J Pharmacol Exp Ther . (accepted Sept 13 th , 2010, doi:10.1124/jpet.110.173179)

The first and fourth publications are used in Chapter 2 while the second and third

publications are mentioned in Chapter 3, along with the findings for CBR4 in the fourth

publication. The manuscript with the findings for the cell line study in Chapter 4 looking

at the association between cytotoxicity and DOX/DAUN metabolism has been submitted

for publication.

xx CHAPTER 1: OVERVIEW

Cancer is one of the most common causes of death, accounting for nearly 1 of every 4 deaths in the US and Canada (Marrett et al., 2008; American Cancer Society,

2010). One common method of treating cancer is through the use of chemotherapy, which involves the administration of anti-cancer drugs to slow or stop cancer cells from growing, multiplying or spreading to other parts of the body. In the past several decades, these drugs have prolonged the lives of countless cancer patients; however, the effectiveness of many anti-cancer drugs is limited by their toxicity to normal rapidly dividing cells. This is the case with one group of anti-cancer drugs known as the anthracyclines, in particular, daunorubicin (DAUN) and doxorubicin (DOX). There is considerable inter-patient variability in the onset of serious life-threatening adverse effects associated with anthracyclines, such as chronic cardiotoxicity. A possible reason for this variability is altered metabolism due to genetic polymorphisms in the genes encoding the metabolic enzymes for these drugs. The enzymes of interest are the aldo- keto reductases (AKRs) and carbonyl reductases (CBRs) since previous studies have demonstrated that they are capable of metabolizing DAUN and/or DOX. This thesis describes the detailed studies conducted to examine the role of genetic variants in the

AKRs and CBRs on the metabolism and cytotoxicity of DOX and DAUN:

• Chapter 1 provides the general background in support of the studies;

• Chapters 2 and 3 examine the role of genetic variation in AKRs and CBRs using

bacterially-expressed human recombinant enzymes and comparing DAUN and

DOX metabolism between the wild-type and variants;

1 • Chapter 4 assesses the association between cytotoxicity and altered metabolism of

DAUN and DOX mediated by AKRs/CBRs using a variety of cell lines; and

• Chapter 5 provides a global summary of the studies with some suggestions for

future directions of this work.

1.1 Anthracyclines—chemical structure and mechanism of action

Anthracyclines are a class of anti-cancer drugs used in clinical practice for the treatment of carcinomas (invasive malignant tumors derived from epithelial tissue), leukemias (cancer of the bone marrow), lymphomas (cancer of the lymph nodes or lymphoid tissue), and sarcomas (cancer of the body’s connective tissues such as bone, muscle, and cartilage) (Wojtacki et al., 2000; Danesi et al., 2002). The chemical structure

of anthracyclines consist of a fused tetracycline ring containing (i) hydroquinone and

quinone structures found adjacent to each other in rings B and C, respectively, which

gives the ring system both electron-donating and -accepting properties, (ii) a methoxy

group at the carbon-4 position on ring D, and (iii) a short carbonyl-containing side chain

at carbon-9 and glycosidic linkage to a daunosamine sugar at carbon-7 (Figure 1).

Figure 1—Chemical structure of an anthracycline illustrating the fused tetracycline ring (denoted by A, B, C, and D) as as well the daunosamine sugar.

The introduction of anthracyclines to cancer chemotherapy has been one of the major successes of oncology. For example, in pediatric oncology, the 5-year survival rate for childhood cancer has increased from approximately 30% in the 1960s to 70–80% today (Gatta et al., 2002; Jemal et al., 2006) with more than 50% of cancer survivors receiving anthracycline treatment (Krischer et al., 1997).

Notwithstanding the widespread use of anthracyclines in patient care and experimental studies, the exact mechanisms by which these drugs exert their cytotoxic actions are not certain (Gewirtz, 1999; Minotti et al., 2004). The most commonly described mechanisms of action are DNA intercalation (Sinha et al., 1984; Gerwitz,

1999) and inhibition of topoisomerase II (Mordente et al., 2001; Binaschi et al., 2001).

The intercalation of the anthracyclines into the DNA structure inhibits replication of cancer cells since enzymes required for DNA replication (i.e., DNA polymerases, helicases, and primases) are unable to interact with the anthracycline-DNA complex.

Likewise, DNA repair enzymes, such as the DNA glycosylases (enzymes involved in base excision repair), would not be able to bind to the anthracycline-DNA complex.

Without proper repair of DNA, cancer cells will eventually die through the activation of apoptotic response systems (Goto et al., 2001).

Inhibition of topoisomerase II is another mechanism that is believed to play a significant role in anthracycline-induced cytotoxicity (Minotti et al., 2004). Its role in

DNA replication and cell division is vital, since it catalyzes several reactions, including catenation–decatenation of sister chromosomes, DNA knotting–unknotting, and removal of supercoils, by making transient double stranded nicks in the DNA strand, all of which reduce the topological complexity of DNA (Tewey et al., 1984; Wilstermann and

3 Osheroff, 2003; Roca, 2009). Anthracyclines act as “topoisomerase poisons” by stabilizing the DNA-topoisomerase II covalent adduct, thus preventing re-ligation of the

DNA strands (Guano et al., 1999; Maxwell et al., 2005). DNA damage due to topoisomerase poisoning by anthracyclines can overwhelm the cell leading to growth arrest in the G1 and G2 phases of the cell division cycle, ultimately triggering cancer cell apoptosis via the p53-dependent pathway (Ruiz-Ruiz et al., 2003; Minotti et al., 2004;

Simunek et al., 2009). In addition to these proposed mechanisms, some studies have

suggested that the production of reactive oxygen species (ROS), direct damage to cell

membranes (Vichi and Tritton 1992; Binaschi et al., 2001), and lipid peroxidation

participate in the anti-cancer effects of anthracyclines (Muindi et al., 1984). However,

these suggested mechansisms are ascertained from in vitro studies; hence, it remains

unclear to what extent they contribute to the cytotoxicity of cancer cells, since the drug

concentrations or experimental conditions for these studies often do not reflect the in vivo

situation (Gewirtz, 1999).

1.2 DAUN and DOX—their use in cancer treatment and cardiac side effects

Many of the anthracyclines continue to undergo clinical trials for cancer therapy.

Two of these anthracyclines, DAUN and DOX, are commonly used in cancer therapy.

The chemical structures of these two drugs differ solely by the presence of a hydroxyl group at the carbon-14 position (Figure 2). DAUN (otherwise known as daunomycin and rubidomycin) and DOX (also referred to as adriamycin), were isolated in the 1960s from

Streptomyces peucetius , a species of actinobacteria (Tan et al., 1967; Gewirtz, 1999;

Simunek et al., 2009). DOX is used in the treatment of a broad spectrum of cancers, including breast cancer, non-Hodgkin’s disease, childhood solid tumors, and soft tissue

4 carcinomas, while DAUN is employed in the treatment of acute myeloid and acute lymphoblastic leukemias (Hunault-Berger et al., 2001; Danesi et al., 2002; Fassas and

Anagnostopoulos, 2005; Cortes-Funes and Coronado, 2007).

Figure 2—Chemical structures of (A) DOX and (B) DAUN.

Even though DAUN- and DOX-based chemotherapies have contributed to

improved life expectancy in cancer patients, there are common adverse effects, but

fortunately they are largely reversible and clinically manageable: myelosuppression (a

drop in white blood cell count), stomatitis (inflammation of the mucous lining of any of

the structures in the mouth), hair loss, mucositis (inflammation and ulceration of the

mucous membranes lining the digestive tract), and gastrointestinal disturbances (nausea

and vomiting) (Gianni et al., 1995; Rubin and Hait, 2003; Lang-Bicudo et al., 2008;

Simunek et al. 2009; Majem et al., 2009). On the other hand, there are some adverse

effects that require specific attention due to their irreversible, life-threatening nature. In

5 particular, a major concern is the considerable inter-patient variability seen in the development of chronic cardiotoxicity (Shan et al., 1996; Wojtacki et al., 2000; Mordente et al., 2001; Minotti et al., 2001; Lipshultz et al., 2008; Simunek et al., 2009). This type of toxicity is severe because it can ultimately lead to decreased left ventricular ejection fractions, cardiomegaly (enlargement of the heart), and eventual congestive heart failure, which accounts for at least 20% of mortalities in patients treated with DOX (Praga et al.,

1979; Von Hoff et al., 1979; Shan et al., 1996). Chronic cardiotoxicity arises through a progressive decrease in ventricular contractile function, normally within the first year after completion of treatment with anthracyclines (Singal and Iliskovic, 1998); however, it also occurs as a delayed form, presenting itself at 4 to 20 years after conclusion of cancer therapy (Steinherz et al., 1991).

The incidence of chronic cardiotoxicity and congestive heart failure is dose- dependent, escalating in occurrence with increasing total lifetime dosing of anthracyclines. In a retrospective study of 399 patient records, congestive heart failure was detected in 4% of cancer patients who received 500 to 550 mg DOX/m 2 body

surface, 18% in patients given 551 to 600 mg/m 2, and 36% in patients given ≥ 601 mg/m 2

(Lefrak et al., 1973). Another retrospective study verified this trend of dose-dependent

increase in congestive heart failure with DOX after analysis of 4018 patient records: 3%

at a cumulative dose of 400 mg DOX/m2; 7% at 550 mg/m2, and 18% at 700 mg/m2 (Von

Hoff et al., 1979). Dose-dependent increases in congestive heart failure have also been

reported for DAUN following the review of 5613 patient records with the incidence

rising from 1.5% at a cumulative dose of 600 mg/m2 to 12% at 1000 mg/m2 (Von Hoff et al., 1977). Despite variations among patients in their tolerance of DOX and DAUN, an empirical cumulative lifetime dose limit of 500 mg/m 2 is suggested as a strategy to

6 minimize the risk of chronic cardiotoxicity and congestive heart failure (Gilladoga et al.,

1976; Wallace, 2003). While 500 mg/m 2 is the recommended dose, there is considerable

inter-patient variation observed with some cancer patients developing congestive heart

failure at a cumulative dose of 300 mg DOX/m 2; however, other patients exhibited no

signs of cardiomyopathy with cumulative doses greater than 1000 mg DOX/m 2 (Deng

and Wojnowski, 2007). At this time, there are no biomarkers to determine the risk a

cancer patient has for developing chronic cardiotoxicity or other forms of cardiac damage

prior to starting anthracycline therapy. Therefore, cancer patients who can endure doses

of anthracyclines higher than the recommended limit will not receive optimal therapy,

whereas those patients who are highly susceptible to congestive heart failure are put in a

dangerous situation since the recommended dose may prove to be lethal or permanently

detrimental.

The clinical use of DAUN and DOX spans more than fifty years and an analog

that is a potent cancer cell killer with diminished cardiotoxic properties has yet to be

made. Over 2000 synthetic analogs of DAUN and DOX (i.e., aclarubicin, epirubicin, and

idarubicin are just to name a few) have been developed with modifications made to the

chemical structure in attempts to retain antitumor efficacy while eliminating the

incidence of cardiotoxicity (Figure 3). These modifications involve adding or removing

substituents from the tetracycline rings and carbon-9 side chain, as well as adding a

longer chain of daunosamine sugars. Only a few analogs have reached the stage of

clinical development and approval; among them, epirubicin and idarubicin enjoy

popularity as useful alternatives to DOX or DAUN, respectively. Unfortunately, the

major side effect of epirubicin and idarubicin, as well as the other clinically used analogs

in cancer therapy, is the development of cardiotoxicity (Dabich et al., 1986; Natale et al.,

7 1993; Anderlini et al., 1995; Dhingra et al., 1995; Ryberg et al., 1998; Minotti et al.,

2004).

Figure 3—Chemical structures of anthracycline analogs epirubicin, idarubicin, and aclarubicin.

In addition, the concomitant use of anthracyclines with certain other anti-cancer drugs has resulted in an increased risk of anthracycline-induced chronic cardiotoxicity

(Wojtacki et al., 2000; Mordente et al., 2001; Danesi et al., 2002; Floyd et al., 2005;

Gianni et al., 2007). For example, taxanes such as paclitaxel are microtubule inhibitors that induce apoptosis in breast cancer cells and inhibit tumor angiogenesis (Grant et. al.,

2003; Minotti et al., 2004). Therefore, the combination of anthracyclines with taxanes came as a natural step toward an improved treatment of metastatic breast cancer.

However, trials incorporating bolus DOX followed by a 3 hr infusion of paclitaxel have

8 resulted in 18% of patients developing congestive heart failure at cumulative doses of

480 mg/m 2 (Gianni et al., 1995). This finding was supported by a previous study, which demonstrated that 50% of breast cancer patients undergoing DOX-paclitaxel treatment had significant reductions in left ventricular ejection fractions, and 20% of these patients developed congestive heart failure (Gehl et al., 1996). Another anti-cancer drug,

Herceptin ® (also known as trastuzumab), is a humanized monoclonal antibody that selectively inhibits the growth and proliferation of breast cancer cells overexpressing tyrosine kinase human epidermal growth factor receptor 2 (HER-2) (Feldman et al., 2000;

 Tan and Swain 2003; Jones and Smith, 2004). Although Herceptin has been successful in breast cancer chemotherapy, its use with DOX has resulted in cardiotoxic effects. For example, a study by Seidman et al., (2002) reported that the incidence of cardiac

 dysfunction in patients receiving Herceptin alone was approximately 3-7%, yet this incidence increased substantially to 27% for patients undergoing treatment with a

 combination involving DOX and Herceptin .

1.3 Proposed mechanisms of DAUN/DOX-induced cardiotoxicity

There is evidence that the accumulation of DOX in the heart is dangerous due to

damage of the cardiac tissue (Menna et al., 2007). A slow infusion of DOX in patients

has led to reductions in cardiotoxicity compared to the administration methods involving

bolus dosing and drug encapsulation, both of which prolong drug circulation (Berry et al.,

1998; Speyer and Wasserheit, 1998; Minotti et al., 2004; Cortes-Funes and Coronado,

2007). However, the exact mechanism of anthracycline-induced cardiotoxicity is not

certain. Previous studies have suggested that there are several pathogenic mechanisms

involved: (i) cellular toxicity from metabolites (Boucek et al., 1987; Minotti et al., 1995);

9 (ii) selective inhibition of gene expression for proteins associated with contraction of the myocardium (i.e., cardiac troponins, actin thin filaments, myosin light chains, and creatine kinase) (Ito et al., 1990); (iii) excessive calcium overload inside cardiomyocytes due to altered activity of calcium pumps (Mitrius and Vogel, 1990); (iv) intensive release of vasoactive amines in the myocardium and peripheral tissue (Rossi et al., 1994); and (v) inhibition of topoisomerase II activity (Cummings et al., 1991).

Although the cause of anthracycline-induced cardiotoxicity is multifactorial, a

large body of evidence points to the involvement of reactive oxygen species (ROS) in

cardiomyocyte damage (Myers et al., 1977; Doroshow, 1983; Jackson et al., 1984;

Rajagopalan et al., 1988; Minotti et al., 2004; Gilleron et al., 2009; Zhang et al., 2009;

Menna et al., 2010). ROS are reactive molecules that include hydrogen peroxide (H 2O2)

and highly reactive free radicals such as the hydroxyl radical (·OH) and superoxide anion

- (·O 2 ), all of which have been documented to be the major perpetrators responsible for

DOX-induced chronic cardiotoxicity in cancer patients following anthracycline treatment

(Rajagopalan et al., 1988; Minotti et al., 1998; Xu et al., 2001; Tokarska-Schlattner et al.,

2006). ROS are naturally produced in biological systems and are capable of damaging

cells through a number of different mechanisms, in particular, oxidation of cell

membrane lipids, DNA disintegration, and dysfunction of enzymes containing sulfhydryl

groups (Wojtacki et al., 2000). Another mechanism involves anthracyclines inhibiting

the mitochondrial respiratory chain by interacting with mitochondrial DNA or binding to

cardiolipin, a polyunsaturated, fatty acid-rich phospholipid with a high affinity for

anthracyclines found in elevated concentrations in the inner mitochondrial membrane

(Giantris et al., 1998; Herman et al., 1999; Wouters et al., 2005; Simbre et al., 2005;

Lipshultz et al., 2008). On the other hand, it is also important to mention that ROS have

10 beneficial roles such as activating phagocytic cells responsible for killing harmful bacteria and regulating smooth muscle tone in vascular tissue (Horenstein et al., 2000).

ROS production in the presence of DOX has been speculated to occur through one or a combination of non-enzymatic and enzymatic processes. The non-enzymatic process tends to be iron-mediated. Cells normally contain a small amount of total cellular cytosolic iron, less than 5%, that is either free or loosely bound (Esposito et al.,

2002; Xu et al., 2005). Anthracyclines like DOX can directly bind to Fe 3+ in the presence

of oxygen, forming a DOX-Fe 3+ complex. This complex can be reduced to DOX-Fe 2+ in

the presence of glutathione or by intramolecular autooxidation of the hydroquinone

moiety in the tetracycline ring. The DOX-Fe 2+ complex can go through either of two

- reaction pathways: (i) it can react with molecular oxygen (O 2) to form ·O 2 as one of the

products, which may then dismutate to form H 2O2; or (ii) the complex can react with

3+ H2O2 to generate the extremely reactive ·OH radical. In both cases, DOX-Fe is

regenerated. This non-enzymatic process is comparable to that of DOX redox cycling

with no metabolites being produced and the reaction proceeding indefinitely (Jung and

Reszka et al., 2001; Xu et al., 2005). Studies have demonstrated that the use of

dexrazoxane, a prodrug analog of the iron chelator EDTA, has had some success in

preventing the cardiotoxicity of anthracycline therapy, therefore providing good clinical

evidence for involvement of the iron-mediated non-enzymatic process in ROS formation

following anthracycline treatment (Hasinoff et al., 2003; Schroeder and Hasinoff, 2005;

Hasinoff and Herman, 2007; Hasinoff, 2008; Sanchez-Medina et al., 2010).

The enzymatic process of anthracycline-mediated ROS generation (Figure 4) involves a one electron reduction of the quinone moiety in the tetracycline ring by cellular enzymes (i.e., cytochrome-P450 or cytochrome b5 reductases, mitochondrial

11 NADH dehydrogenases, xanthine dehydrogenases, and endothelial nitric oxide synthases) which results in the formation of a semiquinone radical (Vasquez-Vivar et al., 1997;

Minotti et al., 1999; Minotti et al., 2004). The DOX semiquinone can subsequently

- transfer an electron to O 2 to form ·O 2 (Olson and Mushlin, 1990; Singal et al., 1997;

Minotti et al., 1998; Wallace and Starkov, 2000; Tokarska-Schlattner et al., 2006), initiating a reaction cascade leading to the formation of other ROS and reactive nitrogen

- - species: superoxide dismutases can convert ·O 2 into H 2O2; ·O 2 and H 2O2 can interact

- with iron or other transition metal ions to create ·OH; ·O 2 can also initiate lipid

peroxidation forming lipid peroxides and derived alkoxyl and peroxyl radicals (ROOH,

RO·, ROO·) or react with nitric oxide (NO·) to form peroxynitrite (ONOO -) (Tokarska-

Schlattner et al., 2006).

Figure 4—DOX-related redox cycling generation of ROS and reactive nitrogen species (RNS) by cellular enzymatic mechanisms (explained in the text). The redox cycling involves a one electron (e-) reduction of the quinone moiety to a semiquinone radical in the tetracycline ring of DOX by cellular enzymes. The - DOX semiquinone transfers an electron to O 2 to form ·O 2 , which subsequently leads to the formation of other ROS and RNS. The quinone moeity and semiquinone radical within the DOX chemical structure are shown in dashed boxes.

However, as oxidation is the primary means by which human get energy and accompanies many metabolic processes, the human body has evolved a natural counterbalance to oxidative cell damage from ROS with the widespread use of antioxidant molecules and the expression of antioxidant enzymes such as superoxide

- dismutases (responsible for catalyzing the conversion of ·O 2 into H 2O2 and O 2), as well as catalases and glutathione peroxidases (both enzymes convert H 2O2 to H 2O and O 2).

Nonetheless, under conditions where ROS production is significantly increased, the antioxidant protective system may be overwhelmed. This is especially the case with the heart since cardiomyocytes are characterized by a low content of antioxidant enzymes, in particular, glutathione peroxidases and superoxide dismutases (Cummings et al., 1991;

Horenstein et al., 2000). Also, it has been demonstrated that during DOX treatment, concentrations of glutathione peroxidase are decreased, which further enhances sensitivity of the heart to the destructive effects of ROS.

1.4 The phases of biotransformation and DOX/DAUN metabolism

Biotransformation is the process whereby a substance is converted from one chemical to another by an enzyme-catalyzed chemical reaction in the body. The biotransformation reactions catalyzed by metabolic enzymes are generally divided into,

Phase I and Phase II processes. This may be followed by a Phase III process, which involves the removal of the metabolites from the cell with the aid of efflux transporters

(Figure 5).

Figure 5—A simplified overview illustrating Phase I, II, and III metabolism of aspirin occurring in a cell. Phase I and II reactions require enzymes (in green; either free in the cytosol or membrane bound) while Phase III focuses on transporters (in red) embedded in the plasma membrane that assist in the removal of Phase I and Phase II metabolites.

Phase I biotransformation normally involves oxidation, reduction, or hydrolysis reactions that lead to the addition or exposure of polar functional groups on the substrate molecule. This increases the hydrophilicity of the molecule, albeit a small increment which facilitates elimination from the body (Klaasen, 1995; Ioannides and Lewis, 2004).

Examples of functional entities added or unmasked are the hydroxyl (-OH), carboxyl (-

COOH), sulfhydryl (-SH) and amino (-NH 2) groups. Cytochrome P450s (CYPs) are the

major enzymes in Phase I metabolism of exogenous and endogenous compounds, which

carry out a number of oxidation and reduction reactions, including aliphatic and aromatic

hydroxylation, epoxidation, sulfoxidation, and oxidative dehalogenation. Other Phase I

14 enzymes consist of flavin-dependent monooxygenases (catalyze N- and S-oxidation), peroxidases (catalyze reduction of hydroperoxides to alcohols), carboxylesterases and peptidases (both catalyze hydrolysis of ester or peptide bonds, respectively), and alcohol dehydrogenases (catalyze oxidation of alcohols to aldehydes (Yin 1994; Krueger et al.,

2006; Imai, 2006)

Phase II biotransformation generally follows Phase I, and generally introduces a hydrophilic group to the compound through a conjugation reaction with the previously formed functional groups (Klaasen, 1995; Knights, 1998). This leads to a substantial increase in water solubility of the compound so that it can be easily excreted from the body. Some of the prominent examples of enzymes (and conjugates) participating in

Phase II reactions are: UDP-glucuronosyltransferases (catalyzes the transfer of glucuronic acid to the substrate); sulfotransferases (sulfate group); glutathione S- transferases (glutathione); and amino acid N-acyl transferases (amino acid, such as glycine). If a compound already has an exposed functional group, it can bypass Phase I and go directly to Phase II.

Phase III encompasses the export of conjugated metabolites from Phase II reactions, as well as unconjugated metabolites from Phase I metabolism, across the plasma membrane into the extracellular space via membrane-bound transporter pumps

(Homolya et al., 2003; Xu et al., 2005). Phase III does not chemically change the molecule as seen in Phase I and II; however, it is included in the metabolic process since it prevents the accumulation of conjugated and unconjugated metabolites in the cell, and therefore, is important in the detoxification process. There are a number of examples of

Phase III transporters that exist in humans: P-glycoproteins [also known as multi-drug resistance 1 (MDR1) protein]; multi-drug resistance associated proteins (MRPs), and

15 breast cancer resistance proteins (BCRPs) (Homolya et al., 2003; Haimeur et al., 2004;

Okamura et al., 2004). These three transporters fall under a superfamily of transmembrane proteins referred to as ATP-binding cassette (ABC) transporters, which efflux a broad spectrum of endogenous and exogenous compounds out of the cell, such as chemotherapeutic drugs, peptides, glucocorticoids, steroids, as well as Phase I

(unconjugated) and Phase II (conjugated) biotransformation metabolites. There are also

ATP-independent Phase III transporters, such as lung resistance-related proteins (LRPs), which are involved in intracellular redistribution of drugs, thereby, reducing exposure of nuclear targets from cytotoxic agents (Zurita et al., 2003; Perez-Tomas, 2006).

DOX is extensively metabolized by Phase I enzymes, such as oxidoreductases

(enzymes that catalyze the transfer of electrons from one molecule, the oxidant, to another molecule, the reductant) to yield the alcohol metabolite, doxorubicinol (DOXol), and by cytochrome P450 reductases (POR) that cleave the glycosidic bond and release the aglycones, 7-desoxydoxorubicinon and 7-desoxydoxorubicinol (Figure 6) (Felsted et al., 1977; Oki et al., 1977; Pan et al., 1981; Lown JW, 1993; Robert and Gianni, 1993; Le

Guellec et al., 1993; Riddick et al., 2005). Phase II processes do not appear to be important metabolic routes for DOX; however, 4-demethyl-7-deoxydoxorubicinol aglycone-4-O-β-glucuronide (dA3-4-O-glucuronide) has been identified in urine following DOX therapy (Takanashi and Bachur, 1976) as has the sulfate conjugate, 4- demethyl-7-deoxydoxorubicinol aglycone-4-O-sulfate (dA3-4-O-sulfate) (Takanashi and

Bachur, 1976; Andrews et al., 1980) (Figure 5). Renal clearance of DOX is low with approximately 12% of the total dose recovered in urine during 6 days of treatment, while hepatic clearance is high with more than half of the total dose excreted in bile within 7 days of treatment (Danesi et al., 2002). Of the fraction eliminated in bile, approximately

16 50% is the parent drug, 23% is DOXol (the major metabolite), and the remainder consists of other metabolites like the aglycones (Lipp and Bokemeyer, 1999). DAUN follows a similar metabolic pathway in humans with daunorubicinol (DAUNol) being the major metabolite (Bachur, 1971; Andrews et al., 1980; Akopfure et al., 1982; Pea et al., 2000).

Figure 6—Schematic representative of DOX metabolism. The parent drug, DOX, is shown in the solid box while the major alcohol metabolite, doxorubicinol (DOXol) is indicated in the dashed box. AKR, CBR, and POR correspondingly refer to aldo-keto reductase, carbonyl reductase, and cytochrome P-450 reductase. NADPH (nicotinamide adenine dinucleotide phosphate, reduced form) is the enzyme cofactor involved in the chemical process.

17 The high variability in the response of patients receiving anthracycline treatment is reflected in the various pharmacokinetic parameters for the parent drug as well as the alcohol and aglycone metabolites, which have been extensively reviewed and reported in the literature (Andrews et al., 1980; Balis et al., 1983; Eksborg et al., 1986; Camaggi et al., 1988a; Camaggi et al., 1988b; Mross et al., 1988; Speth et al., 1988; Jacquet et al.,

1990; Launchbury and Habboubi, 1993; Piscitelli et al., 1993; Robert and Gianni, 1993).

There is a large variation in the values of the pharmacokinetic parameters compiled for

DOX from these references, which include: elimination half life (t 1/2 ), 22 to 74 hrs;

volume of distribution (V d), 21 to 40 L; total body clearance (CL), 24 to 120 ml/min; and

area-under-the-curve (AUC), 1000 to 3700 ng•h/ml. A wide range in t 1/2 values (12 to 52 hrs) for DOXol has also been reported (Jacquet et al., 1990). The considerable inter- patient variation observed in DOX pharmacokinetics, which in turn could explain the variation in chronic cardiotoxicity development in cancer patients, is thought to be related in part to the individual differences in anthracycline metabolism (Cummings et al., 1986).

Two groups of oxidoreductase enzymes have been shown to metabolize DAUN and DOX: the aldo-keto reductases (AKRs) and carbonyl reductases (CBRs) (Cummings et al., 1991; Ohara et al., 1995; O’Connor et al., 1999; Forrest and Gonzalez et al., 2000;

Gonzalez-Covarrubias et al., 2007; Kassner et al., 2008). Within the realm of metabolic

enzymes, AKRs and CBRs are classified as oxidoreductases [a class of enzymes that

catalyze the reversible transfer of electrons from one substance to another (oxidation–

reduction, or redox reaction)] (Figure 7). AKRs are one superfamily of oxidoreductases

while CBRs are under the umbrella of another protein superfamily known as the short-

chain dehydrogenases/reductases (SDRs). There are 15 families of AKRs with human

AKR enzymes categorized into 3 of these families: AKR1, AKR6, and AKR7. For the

18 CBRs, there are 3 human families identified: CBR1, CBR3, and CBR4. A detailed description of the AKRs and CBRs are addressed in the next section.

Figure 7—The location of AKRs and CBRs within all existing categories of metabolic enzymes. The main AKR and CBR enzymes of interest are the ones found in humans (highlighted in green): AKR1, AKR6, AKR7, CBR1, CBR3, and CBR4. The categories for metabolic enzymes were derived by the International Union of Biochemistry and Molecular Biology following their numerical classification scheme [the Enzyme Commission (EC) number] of the enzymes based on the chemical reactions they catalyze: oxidoreductases (EC1), transferases (EC2), hydrolases (EC3), lyases (EC4), isomerases (EC5), and ligases (EC6).

1.5 Aldo-keto reductases and carbonyl reductases

Aldo-keto reductases (AKRs) are members of a highly divergent superfamily of

NAD(P)H-dependent oxidoreductase enzymes with more than 140 members divided into

15 families (AKR1 to AKR15 as seen in Figure 7) (Barski et al., 2008; Mindnich and

Penning, 2009). They are generally described as monomeric proteins with an approximate length of 320 amino acids and are found in all phyla ranging from prokaryotes, protozoans, and yeasts to plants, amphibians, and mammals. The crystal structures determined for a number of AKRs reveal a general protein structure described as an ( α/β)8 barrel motif. In this motif the α-helix and β-strand alternate eight times, and

the β-strands coalesce in the core of the structure to form the staves of a barrel (Penning,

2003; Barski et al., 2008). A nomenclature system for these enzymes was developed based on amino acid sequence similarity that distinguishes the unique protein isoforms

(Jez et al., 1997). Using AKR1A1 as an example, the designation “AKR” identifies the protein as a member of the superfamily; the first numeric figure “1” designates family, defined by 40% shared amino acid sequence identity; the letter “A” represents the subfamily, defined by 60% shared amino acid sequence identity; and the last numeric figure “1” representing the unique protein sequence. Proteins that share 97% or greater amino acid sequence identity must be demonstrated to have distinct functions (i.e., different enzymatic activities or distinct 3' untranslated regions) to be considered unique proteins, rather than variant alleles of a single protein isoform.

A webpage that is publicly available with current information about AKR proteins is maintained at the University of Pennsylvania by Dr. T. Penning

(http://www.med.upenn.edu/akr/). From this website, human AKR proteins are

20 categorized into 3 of the 15 families: AKR1, AKR6, and AKR7. There are 10 human

AKR proteins identified in the AKR1 family, 3 for the AKR6 family, and 2 for the AKR7 family (Table 1). The AKR1 proteins are the largest of all the families and include aldose reductases, aldehyde reductases, hydroxysteroid reductases, and steroid 5 β- reductases (Jin and Penning, 2007; Penning and Drury, 2007). The AKR6 and AKR7 proteins are human homologs of potassium channel subunits and aflatoxin aldehyde reductase, respectively. These AKR enzymes are widely distributed in human tissues, with high levels of expression being notable in the liver, heart, small intestine, kidney, and brain (O’Connor et al., 1999; Jin and Penning, 2007).

Table 1—The human AKR proteins within the AKR1, AKR6, and AKR7 families identified in the AKR Superfamily website along with the location of the gene in the chromosome (http://www.med.upenn.edu/akr/).

CHROMOSOME GENES ALTERNATIVE NAMES LOCATIONS AKR1A1 Aldehyde reductase 1p33-p32

AKR1B1 Aldose reductase 7q35 Aldose reductase-related protein; Small intestine AKR1B10 7q33 reductase AKR1B15 Aldose reductase related protein 7q33 Trans-1,2-dihydrobenzene-1,2-diol dehydrogenase; 20α- AKR1C1 Hydroxysteriod dehydrogense; Dihydrodiol 10p15-p14 dehydrogenase 1; Indanol dehydrogenase Trans-1,2-dihydrobenzene-1,2-diol dehydrogenase; 3 α- AKR1C2 Hydroxysteriod dehydrogense type 3; Dihydrodiol 10p15-p14 dehydrogenase 2; Bile acid binding protein

Trans-1,2-dihydrobenzene-1,2-diol dehydrogenase; 3α- Hydroxysteriod dehydrogense type 2; Dihydrodiol AKR1C3 dehydrogenase x; 17 β-Hydroxysteriod dehydrogenase 10p15-p14 type 5; Testosterone 17 β-dehydrogenase 5; Indanol dehydrogenase

3α-Hydroxysteriod dehydrogense type 1; Dihydrodiol AKR1C4 10p15-p14 dehydrogenase 4; Chlordecone reductase

AKR1D1 Steroid 5 β-reductase; 3-oxo-5β-steroid 4-dehydrogenase 7q32-q33 AKR1E2 Testis-specific protein 10p15.1

21 CHROMOSOME GENES ALTERNATIVE NAMES LOCATIONS Potassium voltage-gated channel, shaker-related AKR6A3 3q26.1 subfamily, β member 1 Potassium voltage-gated channel, shaker-related AKR6A5 1p36.3 subfamily, β member 2 Potassium voltage-gated channel, shaker-related AKR6A9 17p13.1 subfamily, β member 3 AKR7A2 Aflatoxin aldehyde reductase 1p36.13 AKR7A3 Aflatoxin aldehyde reductase 1p36.13

Carbonyl reductases (CBRs) are members of the short chain dehydrogenase/ reductases (SDRs), a highly divergent superfamily of NADPH-dependent oxidoreductase enzymes that has grown by several orders of magnitude, from 20 in 1991 to over 47,000 today (Persson et al., 1991; Persson et al., 2009; Kallberg et al., 2010). In phylogenetic comparisons, members of the SDRs show early divergence where the majority have only low pairwise amino acid sequence identity (typically 20–30%), but have several properties in common (Persson et al., 1991; Kavanagh et al., 2008; Persson et al., 2009).

These common properties include the presence of a single-domain, cofactor binding

Rossmann fold consisting of a β-sheet sandwiched between three α-helices on each side, as well as an active site containing a catalytic tetrad of highly conserved tyrosine, lysine, serine and asparagine residues (Filling et al., 2002; Kavanagh et al., 2008). SDRs average between 250–350 amino acids in length and exhibit no structural similarities to aldo-keto reductases (AKRs), in spite of the overlapping substrate specificities (Flynn and Green, 1993; Maser, 1995; Hoffmann and Maser, 2007).

Currently, there are three isoforms of CBRs found in humans: CBR1, CBR3, and

CBR4 (Table 2) (Matsunaga et al., 2006; Hoffman and Maser, 2007; Oppermann, 2007;

Endo et al., 2008). CBRs are ubiquitously expressed in different human tissues, which

22 include the liver, heart, lung, epidermis, kidney, small intestine, and brain (Wirth and

Wermuth, 1992; Forrest and Gonzalez, 2000). CBR1 and CBR3 are considered to be of clinical importance since they play a major role in metabolism of the anthracycline drugs, daunorubicin (DAUN) and doxorubicin (DOX) (Licata et al., 2000; Lakhman et al., 2005;

Gonzalez-Covarrubias et al., 2007; Blanco et al., 2008). There are no published studies on the ability of CBR4 to metabolize these drugs; however, this enzyme was included in our studies along with CBR1 and CBR3 for the sake of completeness.

Table 2—The human CBR proteins along with the location of the gene in the chromosome (Matsunaga et al., 2006; Hoffman and Maser, 2007; Oppermann, 2007; Endo et al., 2008).

CHROMOSOME GENES ALTERNATIVE NAMES LOCATIONS Prostaglandin-E(2) 9-reductase; Prostaglandin 9- CBR1 21q22.12 ketoreductase; 15-hydroxyprostaglandin dehydrogenase

Short chain dehydrogenase/reductase family 21C, CBR3 21q22.2 member 2

Short chain dehydrogenase/reductase family 45C, CBR4 member 1; Carbonic reductase 4; 3-oxoacyl-[acyl-carrier- 4q32.3 protein] reductase

AKRs and CBRs are classified as Phase I biotransformation enzymes, which play an essential role in converting a broad spectrum of endogenous and exogenous compounds containing reactive aldehyde and ketone functional groups to their corresponding hydroxy metabolites (Matsunaga et al., 2006; Jin and Penning, 2007;

Hoffman and Maser, 2007; Oppermann, 2007). This metabolic conversion increases the water solubility of the metabolites and facilitates their elimination from the body directly or via Phase II conjugation reactions. Some examples of compounds that are metabolized

23 by AKRs are steroid hormones, aflatoxins, sugar/lipid aldehydes, polycyclic aromatic hydrocarbons and prostaglandins (Matsunaga et al., 2006; Jin and Penning, 2007;

Penning and Drury, 2007). The detoxification and activation of many of these substrates are necessary for complex processes such as drug metabolism, toxicant elimination, or carcinogenesis (Jin and Penning, 2007).

As stated previously (on page 18 of section 1.4), AKRs and CBRs are of

particular interest clinically since they play a major role in the metabolism of

anthracycline drugs DAUN and DOX, which rank among the most effective

antineoplastic agents ever developed in cancer therapy (Licata et al., 2000; Jin and

Penning, 2007). It is possible that altered metabolic activity of AKRs and CBRs resulting

from genetic polymorphisms in the genes producing the AKR and CBR enzymes may be

one factor contributing to the considerable inter-patient variability in DAUN and DOX

pharmacokinetics and occurrence of chronic cardiotoxicity.

1.6 Genetic polymorphisms in genes producing drug metabolizing enzymes

Genetic and non-genetic factors are determinants of the inter-individual and inter-

ethnic variability in drug disposition (absorption, distribution, metabolism, and excretion

of an administered drug). Non-genetic risk factors include age, sex, diet, concomitant

therapy, drug interactions, organ function, and disease states (i.e., chronic liver disease or

human immunodeficiency virus infection), all of which can influence the therapeutic

effects of drugs, as well as the onset of adverse drug reactions (Jackson, 2004; Shah,

2005). For this thesis, the focus is on genetic factors, more specifically, the genetic

polymorphisms associated with the drug metabolizing enzymes, that could alter the

metabolism of the anthracyclines DAUN and DOX, and therefore play a role in the inter-

24 patient variability associated with the development of serious adverse effects such as chronic cardiotoxicity.

Genetic polymorphisms refer to variations in the DNA sequence among individuals; they may, or may not, affect the physical or biochemical phenotype

(collective characteristics) of an organism. Some examples of genetic polymorphisms are illustrated in Figure 8, and include:

• silent mutations–a change in one DNA base pair that does not the change the amino

acid in the protein made by the gene;

• missense mutations–a change in one DNA base pair that results in a change in amino

acid in the protein made by the gene;

• nonsense mutations–similar to a missense mutation, except that the DNA base pair

change gives rise to one of three stop codons, UAA, UAG, or UGA, which

prematurely signals the cell to stop building the protein and generally leads to

production of a truncated protein;

• insertions–changes the number of DNA base pairs in a gene by adding a piece of

DNA;

• deletions–changes the number of DNA base pairs by removing a piece of DNA; small

deletions involve the removal of one of a few base pairs of DNA; and larger deletions

remove an entire gene or segments of a chromosome;

• duplications–a segment of DNA, usually comprising one or more genes, is present in

one or more extra copies one or more times; and

• frameshift mutations–the addition or loss of a DNA base pair that shifts the reading

frame of the gene; the consequence of altering the reading frame is that the sequence

of amino acids added from the point of the mutation onward is altered.

Figure 8—Schematic representations of the different human genetic polymorphisms: missense, nonsense, insertion, deletion, duplication and frameshift. The silent mutation is not shown, but it involves a single base pair change leading to the same encoded amino acid. Images were obtained from “Genetics Home Reference” website (http://ghr.nlm.nih.gov/)

26 All of these polymorphisms can ultimately affect the levels of protein expression for a particular enzyme as well as its functionality. For example, a missense mutation that alters the three-dimensional structure of either the substrate or cofactor binding sites may produce an enzyme with either a lowered or even non-functional metabolic capability.

Single nucleotide polymorphisms (SNPs) involve a solitary base pair change in the DNA sequence and are the most common genetic variations in humans (Shastry,

2002). Within the human genome, which is comprised of 3 billion base pairs of DNA,

SNPs occur on the average of one in every 300 to 400 base pairs (Kruglyak and

Nickerson, 2001), and can occur in the protein coding regions of genes known as exons

(Figure 9). Within exons, SNPs that are silent mutations are termed synonymous since they do not lead to a replacement of one amino acid with another, that is, they do not alter the amino acid sequence of the protein. Missense or nonsense mutations alter the amino acid sequence of the protein; hence, these changes are referred to as non-synonymous

mutations . SNPs that are not in protein-coding regions may still have consequences on protein expression and function (Figure 9). For example, SNPs in the promoter site (a region of DNA that facilitates the transcription of a particular gene), may alter the binding of the transcription initiation complex which is comprised of many proteins and recruits RNA polymerase II, an enzyme required for transcription of DNA to messenger

RNA. Also, SNPs at the junction between exons and introns (non-coding region of

DNA) may affect splicing during pre-messenger RNA processing. Splicing is important in removing introns (non-coding regions regions in the gene and primary transcript) and joining exons (coding regions) before producing the correct protein through translation.

Furthermore, SNPs in introns can potentially affect gene expression since some of these

27 non-protein coding regions are known to increase expression of genes that they are located in by a process identified as intron-mediated enhancement (Buchman and Berg,

1988; Mascarenhas et al., 1990). The mechanism by which this occurs is still unknown.

Figure 9—A schematic representation of a gene comprised of exons and introns, as well as other regulatory regions of DNA that control gene expression (promoter, silencer and enhancer). Exons are protein coding regions of the gene while introns are the non-coding regions. Promoters are segments of DNA usually occurring upstream from the introns and act as a controlling element in the expression of that gene. They are an important site where RNA polymerase binds and signals where RNA synthesis (transcription) should begin. Enhancers are control regions of DNA that can turn on promoters with the help of activator proteins (in blue), thereby increasing the rate of transcription of a gene. Silencers are control regions of DNA that, upon binding with repressor proteins, can halt the expression of the gene. SNPs that occur at any of the regions (designated by yellow lightning bolt icons) can potentially alter enzyme expression and activity. Other important proteins of note are: (i) the TATA binding protein (in red), a transcription factor that binds specifically to a DNA sequence within the promoter called the TATA box, and initiates the recruitment of other proteins that are required for RNA polymerase to begin transcription, such as coactivators and basal transcription factors; (ii) coactivators (in purple), which are “adapter” molecules that integrate signals from activators; and (iii) basal transcription factors (in green), which position RNA polymerase at the start of transcription and initiate the transcription process. This figure is adapted from Griffiths et al., (2000).

There are bioinformatics databases for SNPs such as the “Database for single nucleotide polymorphisms” (dbSNP) from the National Center for Biotechnology

Information (NCBI) (http://www.ncbi.nlm.nih.gov/snp). This database is an annotated collection of all publicly available nucleotide sequences and their protein translations, which was produced by NCBI as part of an international collaboration with the European

Molecular Biology Laboratory (EMBL) Data Library from the European Bioinformatics

Institute (EBI) and the DNA Data Bank of Japan (DDBJ). This is the database used in

28 identifying the polymorphisms that occur naturally within the human population, for the various AKR and CBR genes and enzymes that were studied. The focus of this research is on the naturally-occurring ns-SNPs in the coding sequence of the human AKR and

CBR genes. This type of polymorphism, as previously described, has the potential to alter the function of the enzymes. However, the extent of altered function depends on the type of amino acid replacement that occurs and can vary from reducing activity by a few percent, perhaps a physiologically insignificant amount, to a complete null allele (no detectable enzyme activity). The activity of the non-synonymous polymorphic enzymes

(i.e., allelic variants) can be ascertained by simply creating the appropriate base pair change, via site-directed mutagensis, in the DNA sequence coding for the AKR/CBR wild-type enzymes (wild-type refers to the ‘normal’ and usually most predominant sequence in the human population based on the NCBI database), transforming the modified genes into a host cell for expression, purifying the expressed enzyme from other endogenous proteins of the host cell, and performing in vitro assays with purified enzyme and a substrate of interest. These methods are addressed in greater detail in Chapters 2 and 3 of my thesis in relation to the AKR and CBR genes and their enzymes.

In vitro assays are carried out to compare metabolic rates of wild-type with that of variant enzymes at specific substrate concentrations, or through Michaelis-Menten studies to determine rates for a range of substrate concentrations. These rates are plotted against the substrate concentrations to compare the values of the enzyme kinetic parameters of the wild-type and allelic variants (Figure 10). The parameters looked at in this study are: the maximum rate at which the enzyme can perform ( Vmax ), the affinity or strength of binding between the enzyme and its substrate ( Km), the rate at which the

substrate is converted into the product ( kcat ), and catalytic efficiency ( kcat /Km). In order

29 for a variant enzyme to exhibit reduced activity, there will be an increase in Km and decreases in Vmax , kcat , and kcat /Km, compared to the wild-type enzyme, while the reverse

is true for increased activity. These kinetic parameter values will assist in understanding

the results from the upcoming examples of studies provided in this Chapter that looked at

the metabolic effect of variant enzymes arising from ns-SNPs, and they are defined and

described more extensively below.

30 Previous studies have demonstrated that ns-SNPs are capable of modifying the activity of an enzyme (Table 3). This is especially the case with certain variants of the human AKRs and CBRs. For example, a recent study by Takahashi et al. (2009) revealed that each of the two naturally occurring allelic variants of human AKR1C2, F46Y and

L172Q, have significantly reduced ability to metabolize the androgen 5 α-

dihydrotestosterone (DHT), which is the most potent natural ligand for the human

androgen receptor to its metabolite, 5 α-androstane-3α,17 β-diol (3 α-diol), a less potent

receptor ligand. This finding suggests that these variants may have a pivotal role in

modulating DHT levels and, consequently, in regulating receptor signalling in the

androgen-dependent tissues such as the prostate (Ji et al., 2003; Ji et al., 2007; Penning et

al., 2008). In addition, in vitro studies by Gonzalez-Covarrubias et al. (2000) and Blanco

et al. (2008) have demonstrated that ns-SNPs in genes encoding human CBRs (V88I and

V244M) reduce metabolism of the anthracycline based drugs, DAUN and DOX,

repectively. The CBR1 V88I variant had a 47% lower production of DAUNol, while the

CBR3 V244M variant exhibited a 2.6-fold decrease in DOXol production, compared to

their respective wild-type CBR enzymes.

Table 3—A summary of studies that looked at the effect of ns-SNPs in genes for different drug metabolizing enzymes.

VARIANT METABOLIC EFFECT STUDIES ENZYMES PATHWAY OR ROLE VARIANT(S) COMPARED TO OF ENZYME WILD-TYPE Lower activity AKR1C2 (aldo- Reduction of Takahashi et (decrease of ~40% in keto reductase dihydrotestosterone to F46Y, L172Q al., 2009 V for both 1C2) 3α-diol max variants) 2.6-fold decrease in Converting carbonyl group DOXol production in Blanco et al., CBR3 (carbonyl of DOX to a hydroxyl V244M relation to the wild- 2008 reductase 3) (DOXol) at the carbon-13 type human CBR3 position enzyme

31 VARIANT METABOLIC EFFECT STUDIES ENZYMES PATHWAY OR ROLE VARIANT(S) COMPARED TO OF ENZYME WILD-TYPE Converting carbonyl group 47% less DAUNol Gonzalez- CBR1 (carbonyl of DAUN to a hydroxyl production compared Covarrubius V88I reductase 1) (DAUNol) at the carbon-13 to wild-type human et al., 2007 position CBR1 enzyme

Both lower activity (S)-warfarin 7'- (V of R144C hydroxylation, phenytoin 4'- max CYP2C9 lowered by 47 to Lee et al., hydroxylation, tolbutamide (cytochrome R144C, I359L 86% for all 2002 methylhydroxylation, P450 2C9) substrates; K of diclofenac 4'-hydroxylation, m I359L increased by fluriprofen 4'-hydroxylation 3.2-11.5 fold) CYP2C8 Lowers activity ( K Hanioka et 6α-hydroxylation of m (cytochrome A238P value 2.9-fold higher al., 2010 paclitaxel P450 2C8) than wild-type) Sutton et al., 2005; 30–40% less activity; MnSOD Catalyzes the conversion of Shimoda- affects localization (manganese superoxide radicals to Matsubayashi A-9V and transport of the superoxide hydrogen peroxide et al., 1996; enzyme into dismutase) Rosenblum et mitochondria al., 1996 Decreased enzyme NAT1 ( N- R33stop, Zhu and Hein, N-OH-PhIP) O- expression and acetyltransferase R187stop, 2008 acetyltransferase activity activity for all 1) R64W, D251V variants

In addition to the AKR and CBR enzymes, there are extensive studies on the effect of ns-NSPs in genes encoding for cytochrome P450 enzymes (CYPs), see Table 3 for some examples. CYPs represent the most important Phase I drug-metabolizing enzymes that oxidize a broad spectrum of endogenous substances and xenobiotics, including more than 90% of all drugs, into more hydrophilic compounds (Nebert and

Russell, 2002). Kinetic studies using a number of different pharmaceutical compounds e.g., warfarin (an anticoagulant), phenytoin (an antieplipetic), tolbutamide (potassium- channel blocker), diclofenac, and fluriprofen (both non-steroidal anti-inflammatory drugs) have demonstrated that mutations in the CYP2C9 gene, and hence its protein

32 product, significantly reduces the rate of hydroxylation of each of these drugs compared to the wild-type enzyme. For example, the R144C variant has shown to lower Vmax of the enzyme product by 47 to 86% of that of the wild-type (Lee et al., 2002) . Also, the catalytic activity of the I359L variant enzyme is significantly reduced for most CYP2C9

substrates due to an increase in Km (3.2 to 11.5-fold) and a reduction in Vmax (50-92%)

(Lee et al., 2002). One variant of the CYP2C8 gene, specifically the A238P substitution, has been associated with reduced 6α-hydroxylation activity in the presence of the anti-

cancer drug, paclitaxel (a 2.9-fold decrease in Km compared to wild-type) (Hanioka et al.,

2010). These findings with the different variants suggest the amino acid substitutions may influence the clinical response to drugs that are metabolized mainly by CYP2C8 and

CYP2C9 enzymes.

Studies looking at the effect of ns-SNPs on enzyme function have also encompassed other metabolic enzymes that are not involved in Phase I reactions. For instance, an alanine to valine substitution at position -9 (A-9V) in the mitochondrial targeting signal peptide of human manganese superoxide dismutase (Mn-SOD) has been shown to change the structural conformation of this targeting sequence, which affects transport and localization of this enzyme into the mitochondria (Shimoda-Matsubayashi et al., 1996; Rosenblum et al., 1996; Sutton et al., 2003). Furthermore, the alanine substitution by valine has been demonstrated to reduce activity of MnSOD by 30–40% compared to the alanine-containing wild-type counterpart (Rosenblum et al., 1996).

MnSOD is vital for catalyzing the conversion of superoxide radicals (liberated as a by- product of the mitochondrial electron transport chain) to hydrogen peroxide.

Also, four ns-SNPs in the gene coding for human arylamine N-acetyltransferase 1

(NAT1), a Phase II enzyme responsible for N-acetylation of arylamines and hydrazine,

33 have demonstrated effects on both protein expression and function. The N-acetylation mechanism requires cysteine at amino acid at position 68 (C68), in the active site, to accept an acetyl group from acetyl-CoA, and then subsequently transfer this group to the substrate. Two of the ns-SNPs in the NAT1 gene are nonsense mutations that lead to

truncated enzymes (R33Stop and R187Stop) while two others are missense mutations,

R64W and D251V. When the nonsense allele was expressed in transfected monkey

kidney COS-1 cells, no protein was detected. When the R64W and D251V missense

alleles were expressed in the same in vitro system, the amount of protein they produced

was 50- and 40-fold, respectively, lower than the amount produced by the wild-type

NAT1 allele (Zhu and Hein, 2008). Furthermore, the activity of all these variants, as

determined by the reduction of N-hydroxy-2-amino-1-methyl-6-phenylimidazo[4,5-

b]pyridine (N-OH-PhIP), were below the limit of detection (Zhu and Hein, 2008).

1.7 Rationale

It is well documented that some patients develop significant cardiac adverse effects following treatment with the anthracycline-based drugs DAUN or DOX, and the likelihood of developing these adverse effects is directly correlated with the life-time cumulative dose of the drug treatment. The inter-patient variability in relation to adverse effects associated with DAUN and DOX treatment cannot be explained by known clinical parameters. It is likely that a significant proportion of this variability may be due to the differences between patients in the activity among enzymes that metabolize these anthracyclines, in particular, the AKRs and CBRs. The AKR and CBR families of interest in this thesis are the ones found in humans: AKR1, AKR6, AKR7, CBR1, CBR3 and CBR4. These families will be further addressed in Chapters 2 and 3.

34 It has been documented in previous studies that human AKR and CBR enzymes play a vital role in the bioactivation and detoxification of a broad spectrum of anthropogenic and natural chemical species, which also includes the anthracycline anticancer drugs, DAUN and DOX. The lack of understanding of factors that modulate the function of these oxidoreductase enzymes makes it important to characterize these factors for the purposes of relating their contributions to the inter-patient variability seen in anthracycline metabolism and, potentially, as determinants of the development of chronic cardiotoxicity.

1.8 Research hypotheses and predictions

HYPOTHESIS #1: The cause of the inter-patient variation in DAUN- or DOX- induced

adverse effects is unknown; however, allelic variation in the genes expressing enzymes

that metabolize these anthracyclines may be one factor contributing to this inter-patient

variable toxicity.

PREDICTION: Naturally occurring genetic polymorphisms in one or more of these genes will significantly reduce the activity of its protein product.

As previously mentioned, the enzymes that metabolize DAUN and DOX belong to the aldo-keto-reductase (AKR) and carbonyl reductase (CBR) families of genes/proteins. Therefore, to test the validity of our overall hypothesis, we examined whether naturally occurring ns-SNPs in the coding regions of the human AKR and CBR genes produce enzymes with altered capacity for the in vitro metabolism of DAUN

and/or DOX. This study compares the wild-type and allelic variants of the selected

human AKRs and CBRs with the in vitro metabolism of DAUN and DOX to the

35 corresponding carbon-13 alcohol metabolites, DAUNol and DOXol. Using purified, bacterially-expressed, human AKR enzymes, we demonstrate that, of the 29 allelic variants that were examined in this thesis, 7 resulted in significantly reduced catalytic efficiencies ( kcat /Km) of DAUN compared to the wild-type enzyme, while 3 variants reduced the kcat /Km of DOX. The analyses for the AKR genes and their enzymes are described in Chapter 2. A similar set of analyses were completed for the CBR genes, and are presented in Chapter 3. Of the 10 CBR allelic variants, 5 were found to significantly reduce the kcat /Km of DAUN and DOX.

HYPOTHESIS #2: There is a strong and consistent association between DAUN/DOX metabolic activity and drug toxicity.

PREDICTION: Alterations in DAUN and DOX metabolism are associated with the development of cytotoxic effects induced by these drugs.

Using nine cell lines from different tissues, cytotoxicity studies were conducted by incubating cell lines with different concentrations of DAUN or DOX (0 to 150 µM) for specified time periods (0, 6, 24, or 48 hr) and determining LC 50 values (the concentration attributing to 50% cell death) using cell viability assays. Metabolic studies were measured by treating cell lines with 100 nM DAUN or DOX for the same time periods, extracting the cytosols, introducing 10 µM of DAUN or DOX to the each cytosolic extracts, and measuring production of the carbon-13 metabolites. The findings from this study indicate that there is a negative relationship between cytotoxicity and

DAUN or DOX metabolism. In addition, we looked at the expression of AKRs and

CBRs in the cell line cytosols, and discovered that cells that are susceptible to toxic

36 effects by DAUN or DOX are the ones with the lowest expression of these enzymes, while the reverse is true for the cells that are resistant to DAUN/DOX toxicity. These analyses are addressed in greater detail in Chapter 4.

37 CHAPTER 2: ALTERED METABOLISM OF DAUNORUBICIN AND DOXORUBICIN BY ALLELIC VARIANTS OF HUMAN ALDO-KETO REDUCTASES 1,2

2.1 Preface

Aldo-keto reductases (AKRs) are members of a highly divergent superfamily of

NAD(P)H-dependent oxidoreductase enzymes that have been shown to metabolize a broad range of endogenous and exogenous carbonyl-containing compounds, including steroid hormones, aflatoxins, sugar/lipid aldehydes, and prostaglandins (Matsunaga et al.,

2006; Hoffmann and Maser, 2007; Jin and Penning, 2007; Penning and Drury, 2007).

AKRs and CBRs are categorized as Phase I metabolizing enzymes responsible for converting compounds containing reactive aldehyde and ketone functional groups to their corresponding hydroxy metabolites. This metabolic conversion increases the water solubility of the metabolites, and facilitates their elimination from the body directly or via

Phase II conjugation reactions.

In addition to the aforementioned compounds, AKRs and CBRs have also been implicated in the metabolism of the anthracycline antibiotics, daunorubicin (DAUN) and

doxorubicin (DOX) (Cummings et al., 1991; Jin and Penning, 2007). As stated in

Chapter 1, DAUN is extensively used in the treatment of acute myeloid and

1 A version of this chapter has been published:

Bains OS, Takahashi RH, Pfeifer TA, Grigliatti TA, Reid RE, and Riggs KW (2008) Two allelic variants of aldo-keto reductase 1A1 exhibit reduced in vitro metabolism of daunorubicin. Drug Metab Dispos 36: 904- 910.

Bains OS, Grigliatti TA, Reid RE, and Riggs KW (2010) Naturally occurring variants of human aldo-keto reductases with reduced in vitro metabolism of daunorubicin and doxorubicin. J Pharmacol Exp Ther . (accepted Sept 13 th , 2010, doi:10.1124/jpet.110.173179

2 See Co-authorship statement for details of contribution of this manuscript .

38 lymphoblastic leukemias while DOX is employed against numerous cancers, such as breast cancer, childhood solid tumors, non-Hodgkin’s lymphoma, and soft tissue carcinomas (Hunault-Berger et al., 2001; Fassas and Anagnostopoulos, 2005; Cortes-

Funes et al., 2007; Deng and Wojnowski et al., 2007). Even though cancer treatments involving these drugs have resulted in greater life expectancy, both anthracyclines have been associated with variation in the onset of life-threatening adverse events, such as chronic cardiotoxicity (Minotti et al., 1995; Barry et al., 2007; Deng and Wojnowski,

2007; Menna et al., 2007). Both the frequency of chronic cardiotoxicity and mortality rate are correlated with the lifetime cumulative dose of either drug. Moreover, the risk of chronic cardiotoxicity increases when DOX and DAUN are used in combination with other anticancer drugs, such as herceptin, paclitaxel, docetaxel, vincristine, and cyclophosphamide (Mordente et al., 2001; Danesi et al., 2002; Floyd et al., 2005; Gianni et al., 2007).

The cause of the inter-patient variation in DAUN- or DOX- induced adverse effects is unknown; however, allelic variation in the genes expressing enzymes that metabolize these anthracyclines may be one factor contributing to this interpatient variable toxicity. To test this hypothesis, we examined whether naturally occurring non- synonymous single nucleotide polymorphisms (ns-SNPs) in the coding regions of human

AKR genes produce enzymes with altered capacity for in vitro metabolism of DAUN

and/or DOX. Ns-SNPs are single base pair changes that may either produce a missense

or nonsense mutation: a missense mutation results in a different amino acid, while a

nonsense mutation results in a premature stop codon. This thesis will focus on the ns-

SNPs of a select group of human AKRs, for which the wild-type enzymes have been

demonstrated to metabolize DAUN and/or DOX: AKR1A1, AKR1B1, AKR1B10,

39 AKR1C1, AKR1C2, AKR1C3, AKR1C4, and AKR7A2 (Ohara et al., 1995; O’Connor et al., 1999; Martin et al., 2006; Kassner et al., 2008). There are 29 documented ns-SNPs in these human genes listed in the National Centre for Biotechnology Information Database with allele frequencies ranging from 0.8 to 94.3% (Table 4). Presently, there is no information on the catalytic properties of the remaining enzyme variants in the presence of either anthracycline. The other remaining AKRs from Table 1 (page 21 of thesis) were excluded from this study since they either (i) have no documented ns-SNPs (i.e.,

AKR1D1), or (ii) have not been reported in literature to be capable of metabolizing

DAUN and/or DOX.

Table 4—Allele frequencies of the non-synonymous single nucleotide polymorphic variants of human AKR enzymes from different ethnic groups.

NCBI rs- ENZYMES a VARIANTS ALLELE FREQUENCIES b number CEU=2.5% ( n=120); HCB=5.6% ( n=90); AKR1A1 N52S rs2229540 JPT=3.3% ( n=90); YRI=1.7% ( n=120)

(EC 1.1.1.2) E55D rs6690497 Not available

I15F rs5054 UFV=1.0% ( n=184); X=10.0% ( n=68)

YRI=5.0% ( n=80); UFV=1.0% ( n=184); H42L rs5056 JPT=1.1% ( n=88) AKR1B1 L73V rs5057 X=5.0% ( n=68) (EC 1.1.1.21) K90E rs2229542 Multiple (CEU, HCB, JPT, etc)=4.7% (n=64) G204S rs5061 X=5.0% ( n=68) T288I rs5062 YRI=5.0% ( n=80)

CEU=1.7% ( n=118); HCB=24.4% ( n=90); P87S rs2303312 AKR1B10 JPT=23.3% ( n=90)

(EC 1.1.1.-) M286T rs3735042 HCB=6.0% ( n=84); JPT=4.7% ( n=86) N313D rs4728329 CEU=9.2% ( n=120) AKR1C1 R170H rs17295755 UFV=52.0% ( n=184) (EC 1.1.1.-, EC 1.1.1.149, EC 1.3.1.20, Q172L rs1735444 UFV=32.0% ( n=184) EC 1.1.1.112)

40 NCBI rs- ENZYMES a VARIANTS ALLELE FREQUENCIES b number

AKR1C2

(EC 1.-.-.-, F46Y rs2854482 CEU=5.8% ( n=120); YRI=14.2% ( n=120) EC 1.3.1.20, EC 1.1.1.213)

CEU=1.3% ( n=76); HSP=1.4% ( n=76); R66Q rs35961894 JPT=3.9% ( n=76); YRI=6.6% ( n=76) AKR1C3 R170C rs35575889 CEU=1.4% ( n=76) (EC 1.-.-.-, EC 1.3.1.20, P180S rs34186955 CEU=8.1% ( n=76) EC 1.1.1.213, A106T rs4987102 PRH=2.1% ( n=76) EC 1.1.1.63, EC 1.1.1.64, E36Term rs1804062 UFV=14.0% ( n=48) EC 1.1.1.188, CEU=40.0% ( n=120); HCB=82.2% ( n=90); EC 1.1.1.112) H5Q rs12529 JPT=94.3% ( n=88); YRI=48.3% ( n=76)

G135E rs11253043 YRI=2.1% ( n=48); HSP=2.2% ( n=46) AKR1C4 CEU=38.3% ( n=120); HCB=34.4% ( n=90); S145C rs3829125 (EC 1.1.1.-, JPT=41.1% ( n=90); YRI=50.8% ( n=120) EC 1.1.1.225, EC 1.1.1.50) CEU=15.8% ( n=120); HCB=15.6% ( n=90); L311V rs17134592 JPT=8.9% ( n=90) V135M rs6670759 YRI=0.9% ( n=116) S255N rs2231203 YRI= 2.1% ( n=116) G198S rs2231200 CEU=0.9% ( n=116) AKR7A2 CEU=5.0% ( n=120); HCB=1.1% ( n=90); A142T rs1043657 YRI=0.8% ( n=120) (EC 1.1.1.n11) E180K rs859210 UFV=5.0% ( n=184)

CEU=0.8% ( n=120); HCB=13.6% ( n=88); Q157H rs859208 JPT=3.4% ( n=88); YRI=28.0% ( n=118) a Below the name of the enzyme are the Enzyme Commission (EC) numerical classification numbers, which were obtained from the Universal Protein (UniProt) database (http://www.uniprot.org/). Based on the chemical reactions they catalyze, some of the enzymes have more than one EC number. b Allele frequencies were obtained from The National Center for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov/). The ethnic groups are designated as follows: YRI (African), CEU (European), HCB (Chinese), HSP (Hispanic), JPT (Japanese), PRH (Pacific Rim Heritage), and UFV (Utah, French and Venezuelan). X refers to a combined population from Michigan along with UFV. The chromosome sample counts ( n) for each variant are also given. Note that the amino acid location numbers for all the AKR7A2 variants in the NCBI SNP database are 29 residues greater than their location within the protein coding sequence from the NCBI Protein database. An explanation for this discrepancy can be found in Kelly et al., (2002) with numbering of amino acid residues on the amino terminus starting from a methionine located 29 residues ahead of the methionine used as the start site in AKR7A2 coding sequence. b

The purpose of this study is to improve our understanding of the effect of genetic

variation in the human AKR genes on the in vitro metabolism of DAUN and DOX to the

corresponding carbon-13 alcohol metabolites, daunorubicinol (DAUNol) and

doxorubicinol (DOXol). Using purified, bacterially-expressed, human AKR enzymes,

we demonstrate that 7 AKR variants resulted in significantly reduced catalytic

efficiencies ( kcat /Km) of DAUN compared to the wild-type enzyme, while 3 variants reduced kcat /Km of DOX. Furthermore, we observed that DAUN is generally a better substrate than DOX for the AKR wild-type and variant isoforms, as shown by the significant increases in kcat /Km.

2.2 Materials and Methods

2.2.1 Chemicals and enzymes

Agarose, ampicillin (sodium salt), chloramphenicol, daunorubicin, DL- glyceraldehyde, DNAseI, doxorubicin, idarubicin, kanamycin, ( S) -1-indanol, lysozyme, methanol, potassium phosphate (KH 2PO 4), N,N,N’,N’ tetramethylethylenediamine,

+ NADP , NADPH, RNAseI, 1-acenaphthenol, and 9,10-phenanthrenequinone were supplied by Sigma-Aldrich (St. Louis, MO). High performance liquid chromatography

(HPLC)-grade acetonitrile, agar, ammonium persulfate, formic acid, ethanol, glycine, glycerol, glacial acetic acid, imidazole, and Tris were purchased from Thermo Fisher

Scientific (Waltham, MA). NaCl and yeast extract were ordered from EMD Chemicals

Inc. (Darmstadt, Germany). Bacto tryptone and pooled human liver cytosol were obtained from BD Biosciences (Franklin Lakes, NJ) while isopropyl β-D-1-

thiogalactopyranoside (IPTG) was supplied by Fermentas (Hanover, MD), respectively.

Tween 20 was purchased from EMD Biosciences (San Diego, CA). Klenow fragment,

42 T4 ligase, Factor Xa, and restriction enzymes were purchased from New England Biolabs

(Ipswich, MA). Doxorubicinol was obtained from Qventas Inc. (Branford, CT).

2.2.2 Molecular cloning of human AKR genes and creation of allelic variants

The AKR1A1 and AKR1C2 genes were excised from insect cell expression vector

constructs (p2Z0p2N-AKR1A1 and p2Z0p2N-AKR1C2) available in Dr. T. Grigliatti’s

lab oratory (Department of Zoology, University of British Columbia) using DraI and

NotI , and subcloned into the NheI (blunt end with Klenow fragment)-NotI sites of the prokaryotic expression vector, pET28a (Novagen, San Diego, CA) with T4 DNA ligase.

These constructs encoded the AKR with a six histidine purification affinity tag (6x-His) separated by a 17-amino acid residue linker and a Factor Xa (FXa) recognition site

[isoleucine (I), glutamic acid (E), glycine (G), and arginine (R)] on the amino terminus

(Figure 11). The FXa cleavage site is located after the FXa recognition site and before the AKR gene.

Figure 11—A translated product of the pET28a-AKR1A1 construct with modifications to the amino terminus of AKR1A1: a 6x-His-tag followed by an amino acid linker and a Factor Xa (FXa) recognition site. The translated product of the pET28a-AKR1C2 (not shown) has the same amino terminus modifications.

The purpose of the 6x-His affinity tag is to allow for efficient purification of the recombinant protein from the endogenous bacterial proteins using nickel-nitrilotriacetic

43 acid (Ni-NTA) affinity chromatography, which will be discussed in the purification protocol on pages 12 and 13 (Gaberc-Porekar and Menart, 2001; Waugh 2005; Arnau et al., 2006). The amino acid linker allowed the 6x-His tag to be exposed for Ni-NTA affinity chromatography so that more purified protein could be obtained for the enzymatic activity assays. In addition, the FXa cleavage site permitted removal of the tag and linker from the enzyme in order to see if these amino terminus modifications affected enzyme activity. This is addressed further in the Results section of this chapter on pages

20 to 21. The translated products for the remaining AKR constructs were set up in a similar manner.

The human AKR1C3 wild-type coding region was excised from a pOTB7

recombinant plasmid (Invitrogen, Carlsbad, CA) using XmnI and DdeI (blunt end with

Klenow fragment) and subcloned into HindIII (blunt end)-XhoI (blunt end) sites of the pET28a prokaryotic expression vector (EMD, Novagen, San Diego, CA) with T4 ligase.

This construct gave rise to a human AKR1C3 enzyme with an amino terminal 6x-His tag separated from the enzyme by a 32-amino acid residue linker. A FXa cleavage site and methionine start site were inserted at the amino terminus between the linker and AKR1C3 gene using the QuikChange ® Site-Directed Mutaganesis Kit (Stratagene, La Jolla, CA)

with the 5' –

CTCCGTCGACAAGCTATAGAAGGAAGAATG GATTCCAAACACCAG – 3'

(forward) and 5' –

CTGGTGTTTGGAATCCATTCTTCCTTCTAT AGCTTGTCGACGGAG – 3' (reverse) primers (the FXa and methionine sites are underlined in the primer sequences). The

QuikChange ® polymerase chain reaction (PCR) amplification protocol for site directed

mutagenesis was modified as follows: two separate 50 µl reactions (one for each primer)

44 were subjected to 10 cycles of denaturation at 95 oC for 1 min, annealing at 60 oC for 1.5

min, and elongation at 68 oC for 6.5 min. A 25 µl aliquot of each PCR amplification reaction was combined with 0.75 µl PfuTurbo ® DNA polymerase (Stratagene). This reaction was subjected to 18 cycles of the same PCR protocol above.

The human AKR1B10 wild-type coding region was PCR amplified from a pOTB7

(Invitrogen) recombinant plasmid using the following primers, which contained an EcoRI

adapter in the forward primers and a XhoI adapter in the reverse primers (adapters are

underlined): 5' – GTACCGCTCGAATTC ATGGCCACGTTTGTGG – 3' (forward) and

5' – GTCTGCTACCTCGAG TCAATATTCTGCATCG – 3' (reverse). The PCR was performed in a 50 µl reaction buffer containing 100 ng of template, 200 ng of each primer, 0.2 mM dNTPs, and 1.25 units of PfuTurbo ® DNA polymerase. The amplifying

conditions for PCR involved an initial denaturation step at 95 oC for 1 min, followed by

35 cycles of denaturation at 95 oC for 30 sec, annealing at 55 oC for 1 min, and extension at 68 oC for 10 min. Subsequently, a final extension reaction was performed at 68 oC for

10 min. The AKR1B10 PCR product was cut with EcoRI and XhoI , then subcloned into

pET28a. This construct gave rise to a human AKR1B10 enzyme with an amino terminal

6x-His tag as well as a 26-amino acid residue linker between the tag and enzyme. A FXa

cleavage site was inserted at the amino terminus between the linker and gene via site

directed mutagenesis with the subsequent primers: 5' –

GGATCCGAATTCATAGAAGGAAGA ATGGCCACGTTTGTGG – 3' (forward) and 5'

– CCACAAACGTGGCCATTCTTCCTTCTAT GAATTCGGATCC– 3' (reverse) (the

FXa sites are underlined in the primer sequences).

The pET15b constructs containing the human AKR1B1, AKR1C1, AKR1C4 , and

AKR7A2 genes were generously provided by Dr. C.R. Wolf (Biomedical Research

45 Centre, Ninewells Hospital and Medical School, University of Dundee, Scotland, U.K.).

These constructs gave rise to their respective AKR genes with a 10-amino acid residue linker between the 6x-His tag and gene. A FXa cleavage site was inserted at the amino terminus between the linker and gene using site directed mutagenesis with the following primers (the FXa site is underlined in the primer sequences) [AKR1B1: 5' –

CGCGGCAGCCATATAGAAGGAAGA ATGGCAAGCCGTCTCC – 3' (forward) and

5' – GGAGACGGCTTGCCATTCTTCCTTCTAT ATGGCTGCCGCG – 3' (reverse);

AKR1C1: 5' – CGCGGCAGCCATATAGAAGGAAGAA TGGATTCGAAATATC– 3'

(forward) and 5' –GATATTTCGAATCCATTCTTCCTTCTAT ATGGCTGCCGCG– 3'

(reverse); AKR1C4: 5' –

GCGGCAGCCATATAGAAGGAAGA ATGGATCCCAAATATC – 3' (forward) and 5'

– GATATTTGGGATCCATTCTTCCTTCTAT ATGGCTGCCGC – 3' (reverse);

AKR7A2: 5' – GCGGCAGCCATATAGAAGGAAGA ATGGATCCCAAATATC – 3'

(forward) and 5' – GATATTTGGGATCCATTCTTCCTTCTAT ATGGCTGCCGC – 3'

(reverse)] . In the case of the construct with the AKR1C4 gene, a ns-SNP resulting in a change from cysteine to tyrosine at position 170 was detected after sequencing.

Therefore, we performed site directed mutagenesis as described above with the following primers (ns-SNP is underlined in the primer sequences) to generate the wild-type

AKR1C4 gene: 5' – GTCAACTTCAACTGCAGGCAGCTGGAG – 3' (forward) and 5' –

CTCCAGCTGCCTGCAGTTGAAGTTTGAC – 3' (reverse). The construct with the

AKR7A2 gene included a ns-SNP at position 142 resulting in a change from alanine to threonine, which turned out to be one of the variants that we used in this study.

Therefore, we performed site directed mutagenesis as described above with the following primers to generate the wild-type gene: 5' – CTTCTACCTACACGCACCTGACCACG –

46 3' (forward) and 5' – CGTGGTCAGGTGCGTGTAGGTAGAAG – 3' (reverse). This construct gave rise to a wild-type AKR7A2 enzyme with a 10-amino acid residue linker between the 6x-His tag and gene.

The pET28a- and pET15b-variant constructs were created by site directed mutagenesis using the QIAGEN (Mississauga, Ontario) protocol with primers listed in

Supplemental Table 1 (see Appendices, pages 180 to 181). However, p2Z0pie2N laboratory vectors containing the variant genes for AKR1A1 had been created previously and did not require site-directed mutagenesis. As a result, these genes were sub-cloned into the pET28a in the same manner as the wild-type AKR1A1 gene, which was described earlier on page 7. All constructs were verified by dideoxy sequencing at the

University of British Columbia Nucleic Acid Protein Service unit. The truncated variant for AKR1C3 (E36Term) was not created since the translated protein is devoid of the important amino acid residues required for catalysis (aspartic acid-50, tyrosine-55, lysine-

84, and histidine-117) (Schlegel et al., 1998; Di Luccio et al., 2006; Jin and Penning,

2007) . Therefore, it was assumed that the protein would be inactive.

2.2.3 Expression and purification of recombinant human AKR wild-type and variant enzymes

The pET constructs of the AKR wild-type and variants were heat-shock transformed into Escherichia coli BL21 (DE3) pLysS competent cells and expressed under the control of an IPTG-inducible T7 polymerase. Cells were plated on Luria-

Bertani broth agar (1% bacto-tryptone, 0.5% yeast extract, 0.5% NaCl) supplemented with antibiotics (25 µg/ml chloramphenicol and 50 µg/ml kanamycin sulphate for the pET28a constructs or 100 µg/ml ampicillin for the pET15b constructs). Colonies were

47 randomly picked and cultured overnight at 37 oC in 3 ml of Luria-Bertani broth with

kanamycin and chloramphenicol at the same concentrations stated previously. Cultures

were expanded to 800 ml and grown at 37°C until an OD 600 of 0.5 was reached. IPTG

was added to a final concentration of 1 mM and cells were allowed to grow for an

additional 5 hrs, after which the cultures were harvested by centrifugation (4000x g for 20

min at 4 oC).

The supernatant was removed and the cells containing the recombinant 6x-His

tagged AKR enzymes resuspended at 5 ml per gram wet weight with Buffer A (300 mM

NaCl, 50 mM NaH 2PO 4, pH 8.0). The cell suspensions were then lysed with lysozyme

(final concentration of 1 mg/ml, incubated on ice for 30 min) and the cells disrupted using six 10-sec burst (with a 10 sec cooling period between each burst) from a sonic dismembranator with a microtip set at 200-300 W. This was followed by incubation with DNaseI and RNAseI (5 and 10 µg/ml, respectively) for 15 min on ice and then centrifugation (10,000x g, 20 min, 4 oC). The cell lysate was subjected to Ni-NTA affinity

chromatography, with the recombinant protein being isolated according to the

manufacturer’s instructions (QIAGEN, Mississauga, Ontario). Briefly, the supernatant

was incubated with Ni-NTA agarose beads for 1.5 hrs at 4 oC to allow for the 6x-His tagged AKR proteins to interact with the Ni-NTA. Ni-NTA occupies four of the six ligand binding sites in the coordination sphere of the nickel ion, leaving two sites free to interact with the 6x-His tag (Figure 12). The supernatant mixture was then transferred to a gravity-flow polypropylene column, after which the 6x-His-tagged AKRs bound to the agarose beads were washed with Buffer A containing 20 mM imidazole to remove non- specific endogenous bacterial proteins bound to the beads. AKRs were further eluted

48 with multiple fractions of Buffer A with increasing concentrations of imidazole (30, 50,

100 and 250 mM).

Figure 12—Interaction between neighboring histidine residues in the 6x-His tag and the nickel- nitrilotriacetic acid (Ni-NTA) matrix. This figure is modified from The QIAexpressionistTM: A handbook for high-level expression and purification of 6x-His tagged proteins (QIAGEN; June 2003, 5 th edition).

Protein purity was assessed by running elution fractions on 18% SDS- polyacrylamide gels, which were stained with SYPRO ® Ruby (Invitrogen Canada, Inc.,

Burlington, Ontario) overnight (16 hrs). After staining, the protein was detected using a

Storm 840 Molecular Dynamics Imager (GE Healthcare, Piscataway, NJ) at excitation

and emission wavelengths of 450 and 520 nm, respectively.

Western blot analyses of the purified fractions were conducted according to the

procedure described by Odyssey TM (LI-COR Biosciences, Lincoln, NE). Following 18%

SDS-polyacrylamide gel electrophoresis, proteins were transferred at 20 V in Towbin's buffer (25 mM Tris, 192 mM glycine, and 20% v/v methanol) overnight (at 4 oC) to a

Hybond-C Extra nitrocellulose membrane (GE Healthcare, Piscataway, NJ). The

membranes were blocked in Odyssey blocking buffer, and the enzyme detected using

49 either a MaxPab polyclonal mouse anti-human AKR1A1, AKR1B1, AKR1B10,

AKR1C1, AKR1C2, AKR1C3, AKR1C4, or AKR7A2 antibody (Abnova ® Corporation,

Taipei City, Taiwan) (diluted 1:2500) as the primary antibodies and IRDye 800CW goat anti-mouse IgG as the secondary antibody (diluted 1:5000) (LI-COR). Both primary and secondary antibodies were in blocking buffer containing 0.1% Tween 20. The bound secondary antibody was detected using the Odyssey TM Infrared Imaging system (LI-

COR).

2.2.4 Rate of enzymatic activity of AKRs in presence of test substrates

The rate of enzyme activities of the purified 6x-His tagged AKR wild-type and variant enzymes were measured using a Fluoroskan Ascent FL (Thermo Fisher Scientific,

Waltham, MA) by following the initial rate of either NADPH oxidation or NADP +

reduction (depending on the test substrate) at excitation and emission wavelengths of 355

and 460 nm, respectively. The assays were conducted using 1-acenaphthenol (for

AKR1C isoforms), DL-glyceraldehyde (for AKR1A1 and AKR1B1), ( S)-1-indanol

(AKR1B10), and 9,10-phenanthrenequinone (for AKR7A2) as test substrates

(Burczynski et al., 1998; O’Connor et al., 1999; Martin et al., 2006; Takahashi et al.,

2008; Byrns et al., 2008). NADPH oxidation was measured with DL-glyceraldehyde,

9,10-phenanthrenequinone, and ( S)-1-indanol, while NADP + reduction was measured with 1-acenapthenol. In brief, 3 µg purified protein was incubated with 2.3 mM NADP +

or 0.2 mM NADPH and test substrate (1 mM acenapthenol and DL-glyceraldehyde, 0.5

mM ( S)-1-indanol, and 0.05 mM 9,10-phenanthrenquinone) in a reaction mixture of 150

µl of 100 mM KH 2PO 4, pH 7.4. Assays involving AKR1A1, AKR1B1, AKR1C1,

AKR1C2, AKR1C3, and AKR7A2 were performed at 25 oC while assays with AKR1B10

50 and AKR1C4 were conducted at 37 oC (these were the temperatures in which the assays were performed in the published literature). In these assays, the concentration of DMSO- methanol (4:1 mixture) required to dissolve the substrates (except for DL-glyceraldehyde for which water was used) was below 4% (v/v) in the final reaction mixture. Readings were collected at 20-sec intervals for 1.5 hrs with shaking between each reading.

Maximal rates were calculated from the Ascent software version 2.6 computer program

(Thermo Fisher Scientific) using a 5-min interval (15 total readings) with the steepest slope. The rate of enzymatic activity was calculated from the maximal rates using a standard curve constructed from the fluorescence measurements of solutions of known

NADPH concentrations. Enzyme activity rates in the presence of 1-acenapthenol and

(S)-1-indanol were measured as nanomoles of NADP + reduced per minute per milligram of purified protein while activity rates for the other substrates was measured as nanomoles NADPH oxidized per minute per milligram of purified protein. Activity rates were ascertained and compared with published rates to determine if the purified enzymes were functional.

2.2.5 Kinetic activity of AKR enzymes in the presence of anthracyclines

Activity measurements for the reduction of the anthracyclines were performed by incubating either DOX or DAUN (10-700 µM; in some cases up to 1000 µM) with 3 µg purified AKR enzyme in a total volume of 150 µl containing 100 mM potassium phosphate, pH 7.4, and 1 mM NADPH at 37°C. Protein concentrations were based on the Bradford protein assay using bovine serum albumin as a standard. The reaction was stopped by adding 300 µl of ice-cold acetonitrile, which contained idarubicin as an internal standard, followed by vortex mixing and centrifugation at 10,000 g for 10 min at

51 4°C to pellet the protein. The supernatant was removed for high performance liquid chromatography (HPLC) analysis. HPLC separation was performed using a Waters

Alliance 2695 system (Waters Corporation, Milford, MA) with a C18, 75 x 4.6 mm i.d., 3

µm analytical column (Waters Symmetry) and a C18, 40 x 4.6 mm i.d. guard column

(Phenomenex SecurityGuard; Phenomenex, Torrance, CA). HPLC conditions were as follows: mobile phase A, 0.1% formic acid and B, acetonitrile; gradient elution, 0 to 1 min, 15% B; 1 to 8 min, 15% B to 35% B; 8 to 10 min, 35% B; 10 to 10.1 min, 35% B to

15% B; column re-equilibration for 2 min at a constant flow rate of 1 ml/min. The column was heated to 30°C and the autosampler temperature was set to 10°C with 5 µl of sample injected onto the column. Fluorescence detection of DOX and DAUN and their carbon-13 hydroxy major metabolites, DOXol and DAUNol, was performed with excitation and emission wavelengths of 460 and 550 nm, respectively (Waters 2475 Multi

Wavelength Detector).

Quantitation of DOXol and DAUNol was performed using a weighted (1/x 2)

linear regression determined from solutions of known concentrations of an authentic

chemical standard for DOXol. Since a chemical standard for DAUNol could not be

obtained, its concentrations were calculated as DOXol equivalents using a response ratio

of 1.0. All HPLC data processes, including chromatogram integration, calibration, and

quantitative calculations, were performed with Waters Empower software (version 2.0).

Maximal reaction velocity (Vmax ) and substrate affinity ( Km) were determined by fitting the rate measurement data using nonlinear least-squares fitting to a Michaelis-

Menten hyperbola (GraphPad Prism version 4.0; GraphPad Software Inc., San Diego,

CA). Values for substrate turnover rate ( kcat ) were calculated from Vmax values using the apparent molecular weight for the 6x-His tagged AKRs. The kcat /Km values for the wild-

52 type and variant enzymes were also calculated. Following Michaelis-Menten data analysis, Eadie-Hofstee plots were constructed (a graphical representation of enzyme kinetics in which reaction velocity is plotted as a function of the velocity vs. substrate concentration ratio) in order to check for deviation from linearity with changing substrate concentrations.

2.2.6 Statistical analysis

Statistical analyses were performed by GraphPad Instat ® (version 3.6; GraphPad

Software Inc., San Diego, CA). Results were expressed as means ± S.D. Enzyme

activities were compared using a one-way analysis of variance followed by Tukey-

Kramer multiple comparisons tests. Differences were considered significant at p<0.05.

Sample sizes ( n) for the following experiments with the AKR wild-type isoforms and variants were as follows: enzymatic activity rates using appropriate test substrates ( n=9),

determination of Michaelis-Menten kinetic parameters using DAUN as substrate ( n=9), and determination of kinetic parameters using DOX as substrate ( n=9).

2.3 Results

2.3.1 Expression and purification of the AKR enzymes

The expression of the 6x-His tagged recombinant human AKR wild-type enzyme

was confirmed by Western blot analysis, which showed a band with mobility

corresponding to the calculated molecular mass of the tagged AKR or CBR (41 kDa for

AKR1A1, 38.5 kDa for AKR1B1, 40.3 kDa for AKR1B10, 39.4 kDa for AKR1C1, 40.2

kDa for AKR1C2, 41.9 kDa for AKR1C3, 39.7 kDa for AKR1C4, and 39.3 kDa for

AKR7A2) (Figure 13). Total protein staining of the SDS-PAGE gel demonstrated that

53 the wild-type fraction was purified from its transformed bacterial lysate, since no other proteins were detected (Figure 13). For the AKR wild-type and variant enzymes, the majority of the pure enzyme was recovered in the 250 mM imidazole elution fractions; no further protein was eluted with higher imidazole concentrations. The results for the expression and purification of the variant forms of each enzyme paralleled that of the corresponding wild-type enzyme (Supplemental Figures 1 to 4; see Appendices, pages

182 to 185).

Figure 13—Purification of human recombinant 6x-His-tagged wild-type enzymes: (A) AKRA1, (B) AKR1B1, (C) AKR1B10, (D) AKR1C1, (E) AKR1C2, (F) AKR1C3, (G) AKR1C4, and (H) AKR7A2. (Left ) Gel stained with SYPRO ® Ruby following SDS-PAGE showing purified protein fraction (lane 3; 2 µg), free of contaminating proteins from the bacterial lysate (lane 1; 10 µg total protein). Removal of contaminating proteins is observed in flow through fraction from Qiagen purification procedures [Ni-NTA column flow through (lane 2; 10 µg total protein)]. ( Right ) Western blot detection of transformed lysate (lane 6) and purified protein fractions (lane 8), confirms expression of the desired AKR protein. Little immunoreactivity was detected in the flow through fraction (lane 7) suggesting that majority of the enzyme was bound to the Ni-NTA resin prior to its elution. GST-tagged full recombinant proteins (Abnova ® Corporation, Taipei City, Taiwan; lane 5; 2 µg total protein) were used as positive controls for antibody immunoreactivity for all enzymes. No antibody immunoreactivity was observed for untransformed bacterial lysate (lane 4; 10 g total protein). M refers to the molecular weight markers.

2.3.2 Enzyme activities of AKR wild-type and variant enzymes using test substrates

First, we examined whether the purified human 6x-His tagged AKR proteins were active and how the mutations affected enzyme activity compared to the wild-type using published test substrates (Table 5). As mentioned previously, enzymatic rates for the

AKR1C isoforms were determined using 1-acenaphthenol while the rates for AKR1A1 and AKR1B1 were ascertained using DL-glyceraldehyde. The test substrates, 9,10- phenanthrenequinone and ( S)-1-indanol, were introduced in metabolic assays to analyze

AKR7A2 and AKR1B10 activity, respectively. In the presence of the test substrates, the

6x-His tagged wild-type AKR isoforms were found to have rates in accordance with the reported values in previous literature (these values are given in Table 5 along with the references from which they were obtained). The reported rate values for AKR1B10,

AKR1C1, AKR1C2, and AKR1C3 were from assays using native (untagged) proteins, while the remaining four AKR isoforms were 6x-His tagged.

A total of eight 6x-His tagged variants exhibited significantly reduced activity rates compared to their corresponding AKR wild-type enzymes: a 29 to 42% decrease in activity for three variants of AKR1C3 (A106T, R170C and P180S; n=9), one variant of

AKR1C2 (F46Y; n=9), one variant of AKR1C4 (L311V; n=9), two variants of AKR1A1

(E55D and N52S; n=9), and one variant of AKR7A2 (A142T; n=9). None of the allelic variants of AKR1B1, AKR1B10, and AKR1C1 were shown to have significantly altered activity compared to their respective wild-type enzymes.

55 Table 5—Enzymatic rates for reported test substrates by recombinant 6x-His tagged AKR wild-type and variant allele enzymes.

REPORTED RATES (nmol/min ● ENZYMES VARIANTS SUBSTRATES RATES (nmol/min ● mg protein) mg protein) a AKR1A1 DL -glyceraldehyde 1240±170 1263±56 [1] N52S 850±170 *

E55D 520±170 * AKR1B1 DL -glyceraldehyde 354±48 441±14 [1] I15F 287±36 H42L 347±42 L73V 342±25

K90E 336±51 G204S 327±45 T288I 313±45 AKR1B10 (S) -1-indanol 980±80 1000 [2] P87S 870±90 M286T 820±90 D313N 900±50 AKR1C1 1-acenaphthenol 2050±200 1800 [3]; 2100 [4] R170H 1891±164

Q172L 1870±149 2400 [3]; 2500 [4]; AKR1C2 1-acenaphthenol 2980±151 2230±80 [5] F46Y 2075±170 * AKR1C3 R66Q 1-acenaphthenol 3370±160 2800 [4] H5Q 3380±53 R66Q 3078±276 A106T 2060±81 * R170C 2298±115 * P180S 2410±165 * AKR1C4 1-acenaphthenol 1660±190 1951±49 [1] G135E 1484±181 S145C 1820±121 L311V 958±132 * 9,10- AKR7A2 2553±229 3100±107 [1] phenanthrenequinone

V135M 2385±295

56 REPORTED RATES (nmol/min ● ENZYMES VARIANTS SUBSTRATES RATES (nmol/min ● mg protein) mg protein) a A142T 1660±130 * Q157H 2425±237 E180K 2303±247 G198S 2483±271 S255N 2450±230

Values correspond to the mean ± S.D obtained from three experiments performed with three independent enzyme preparations ( n=9) for each isoform. Reported rates for the wild-type enzymes are also provided for comparison purposes. * Variants significantly different from their respective wild-type enzymes ( p<0.05) a [1]=O’Connor et al., 1999; [2]=Martin et al., 2006; [3]=Byrns et al., 2008 ; [4]=Burczynski et al., 1998; [5] Takahashi et al., 2008

Following the studies with the test substrates, we wanted to see if there was a significant difference in activity between the tagged and untagged enzymes. The untagged enzymes were generated by incubating the tagged wild-type AKR proteins with

Factor Xa (FXa) for 6 hrs at 23 oC. FXa cleaves the 6x His tag linker sequence after the arginine residue in the isoleucine-glutamic acid-glycine-arginine recognition sequence, thereby creating the native AKR enzyme. Subsequently, the reaction was mixed with Ni-

NTA resin and run through a column to separate the 6x-His tag and amino acid linker from the native enzyme. Western blot analysis demonstrated this treatment completely removed the tag and linker from the wild-type enzymes (Figure 14).

Figure 14—Representative Western blot detection of purified 6x-His tagged AKR proteins [(A) AKR1C3, (B) AKR1C4, and (C) AKR7A2 wild-type enzymes; lane 2; 3 µg] and native AKR proteins (lane 3; 3 µg). The 6x-His tagged proteins were incubated with Factor Xa for 6 hrs at 23 oC to allow for removal of the tag and linker from the amino terminus. GST-tagged AKR recombinant proteins (Abnova Corporation) were used as a positive controls for antibody immunoreactivity (lane 1; 1 g). M refers to the molecular weight markers.

In the case of the 6x-His tagged human AKR1A1 enzyme, there was unsuccessful cleavage of the tag and linker from the native enzyme, which was likely due to the FXa recognition site being hidden within the protein structure, thereby making it inaccessible to the enzyme. The following attempts were undertaken in hopes of removing the tag and linker by FXa cleavage: (i) altering concentrations of the reaction buffer components,

Tris-HCl, CaCl 2 and NaCl; (ii) using different ranges of pH (6 to 9) and temperature (4 to

37 oC) to increase FXa specificity; (iii) increasing FXa enzyme concentration; (iv)

incubating the 6x-His tagged protein with thrombin to cleave the tag and a portion of the

linker at the thrombin recognition site (leucine-valine-proline-arginine-glycine-serine)

which may alter the folding of AKR1A1 and expose the primary cleavage site for FXa,

and (v) performing FXa incubation while the enzyme was bound to the Ni-NTA resin to

aid in exposing the recognition site. Although these attempts were unsuccessful, the

58 activities of the 6x-His tagged AKR1A1 enzyme were in accordance to activity values of native human AKR1A1 as reported in literature using comparable assays (Table 4). As a result, we deemed it unnecessary to cleave the tag and linker from AKR1A1 as they did not appear to alter enzyme activity.

For the other AKR enzymes, the enzymatic rates of tagged and native (untagged) enzymes were compared. Using 1-acenaphthenol, there were no significant differences in enzymatic activity between the tagged and native AKR1C isoforms [rates of the native enzymes were 1800±130, 2710±140, 3060±210, and 1500±210 nmol/min ●mg for

AKR1C1, AKR1C2, AKR1C3, and AKR1C4, respectively; n=9], suggesting that the

amino acid linker and 6x-His tag engineered on the amino terminus of the wild-type gene products has no effect on enzyme activity. Likewise, we found similar results between the tagged and native AKR1B isoforms and AKR7A2 [rates of the native enzyme were

300±50 (AKR1B1), 910±40 (AKR1B10), and 2550±250 (AKR7A2) nmol/min ●mg;

n=9]. Although we did not cleave off the tag and linker for the variants, we assumed that

there would be no significant difference between the rates of the tagged and untagged

variant enzymes. Keeping these results and assumptions in mind, the uncleaved wild-

type and variant enzymes were used for subsequent activity assays using the

anthracyclines as substrates.

2.3.3 Kinetic characterization of wild-type and variant enzymatic activities with DOX and DAUN as substrates

To evaluate the impact of the single amino acid substitutions in the human AKR enzymes on the reduction of anthracycline drugs, we measured the in vitro formation of the major alcohol metabolites. A 50 min incubation period was found to be in the linear range of enzymatic activity for each of the concentrations of DOX and DAUN used to

59 conduct the enzymatic assays (3 µg purified protein). In addition, the concentration of cofactor (1 mM NADPH) was sufficient for maximal enzymatic activity during this 50 min incubation period. Higher concentrations of cofactor were also examined (1.5 mM and 2 mM); however, the enzymatic rates associated with these concentrations did not differ from that of 1 mM.

Full chromatographic resolution of DAUNol, DOXol, DAUN, DOX, and idarubicin (internal standard) was achieved for all chemical standards and in vitro samples

(Bains et al., 2008). DOXol, DOX, DAUNol, DAUN, and idarubicin were observed to

elute at 4.6, 5.5, 6.1, 6.9, and 7.4 min, respectively (Figure 15). Incubation of the 6x-His

tagged AKR wild-type and variant enzymes with DOX generated a single new

chromatographic peak that was identified as DOXol. Similarly, incubation with DAUN generated a single new chromatographic peak that was identified as DAUNol. The identification of the metabolite peaks was confirmed by incubation of DOX and DAUN with human liver cytosol and the generation of compounds that had identical

chromatographic behaviors. There were no detectable peaks at the DAUNol or DOXol

retention time in the absence of the AKR proteins. In addition to the enzymatic assays

performed in this study, we incubated enzyme only as well as enzyme and substrate

without the addition of cofactor, as controls. The production of DOXol and DAUNol

from these controls was below the limit of detection (25 nM) using HPLC-fluorescence.

Therefore, we conclude the production of the metabolites is due to enzymatic processes.

Figure 15 —Generation of DAUNol and DOXol in vitro by purified AKR incubated with DOX (top panel) and DAUN (bottom panel). Measurement of DOXol and DAUNol was performed using HPLC- fluorescence. Representative chromatograms show clear resolution of DOXol and DAUNol from DOX, DAUN, and idarubicin (IDA, internal standard). Retention times observed for DOXol, DOX, DAUNol, DAUN, and IDA are 4.6, 5.5, 6.1, 6.9, and 7.4 minutes, respectively.

Michaelis-Menten kinetic curves were constructed for the AKR wild-type (Figure

16) and each of the variant enzymes in the presence of differing concentrations of

DAUN, and the kinetic parameter values determined (Table 6). A total of 7 allelic variants of AKR1A1, AKR1C3, AKR1C4, and AKR7A2 exhibited significant reductions in enzymatic activity with DAUN compared to their respective wild-type enzymes: a 29 and 46% reduction in kcat /Km for both the E55D and N52S variants of AKR1A1, respectively [this reduction is due to corresponding increases in K m (30 and 79%); n=9];

a 28 to 40% reduction in kcat /Km for the A106T, R170C, and P180S variants of AKR1C3

[this reduction is due to corresponding decreases in Vmax (23-47%) and kcat (22-47%); n=9]; a 45% reduction in kcat /Km for the L311V variant of AKR1C4 [this reduction is due

to corresponding decreases in Vmax (47% lower) as well as kcat and kcat /Km (43% lower); n=9]; and a 85% reduction in kcat /Km for the A142T variant of AKR7A2 [this reduction is

due to corresponding decreases in Vmax and kcat (both 61%) and an increase in Km (156%); n=9].

With DOX as a substrate (Figure 17; Table 7), 3 allelic variants of AKR1C3 and one for AKR7A2 demonstrated significant reductions in enzymatic activity with respect to the wild-type enzymes: a 52 and 69% reduction in kcat /Km for both the R170C and

P180S variants of AKR1C3, respectively [this reduction is due to corresponding decreases in Vmax (41 and 44%) and k cat (39 and 45% decrease); n=9]; and a 60%

reduction in kcat /Km for the A142T variant of AR7A2 [this reduction is due to corresponding decreases in Vmax (41%) and k cat (44%) and an increase in Km (47%); n=9].

Figure 16 —In vitro enzymatic activities for the purified 6x-His tagged (A) AKR1A1, (B) AKR1B1, (C) AKR1B10, (D) AKR1C1, (E) AKR1C2, (F) AKR1C3, (G) AKR1C4, and (H) AKR7A2 wild-type and variant enzymes with daunorubicin. Activities were measured by following the rate of daunorubicinol production. Three independent batches of each enzyme were purified and assays were performed in triplicate with each batch. Enzymatic activities are reported as mean ± S.D. ( n=9). Dotted lines refer to variants that have significant differences in kcat /Km parameter values compared to the wild-type.

63 Table 6—Kinetic constants for DAUN metabolism by recombinant 6x-His tagged AKR wild-type and variant enzymes.

V (nmol/min ● k / K ENZYMES VARIANTS max K (µM) k (s -1) a cat m mg protein) m cat (s -1 M -1) AKR1A1 4042±233 737±39 2.76±0.10 3750±210 N52S 3714±156 957±59* 2.54±0.11 2650±183*

E55D 3910±260 1321±135* 2.67±0.16 2020±194* AKR1B1 427±16 897±51 0.27±0.01 329±14 I15F 405±12 939±50 0.25±0.01 277±15 H42L 414±56 781±91 0.27±0.04 343±60 L73V 382±38 775±85 0.25±0.02 320±48

K90E 389±61 725±65 0.25±0.04 348±66 G204S 456±51 790±78 0.29±0.03 375±68 T288I 440±37 839±88 0.29±0.02 339±47 AKR1B10 1829±77 447±37 1.23±0.06 2752±382 P87S 1700±91 404±30 1.14±0.07 2822±360 M286T 1654±98 452±28 1.11±0.07 2453±287 D313N 1790±84 410±32 1.20±0.06 2927±391 AKR1C1 87±9 551±41 0.057±0.006 103±19 R170H 77±7 585±57 0.051±0.006 87±16

Q172L 72±10 534±39 0.047±0.006 88±17 AKR1C2 114±17 1236±168 0.076±0.01 62±7 F46Y 120±19 1322±152 0.080±0.02 61±10 AKR1C3 1570±37 122±16 1.09±0.03 8937±701 H5Q 1640±76 114±9 1.15±0.04 10066±770 R66Q 1534±28 122±14 1.07±0.02 8854±844 A106T 1082±103* 118±19 0.76±0.07* 6440±740* R170C 837±53* 108±21 0.58±0.05* 5348±676* P180S 1211±46* 134±21 0.85±0.03* 6432±584* AKR1C4 377±28 1096±78 0.23±0.02 228±21 G135E 340±25 1156±64 0.22±0.02 194±25 S145C 356±22 1122±89 0.23±0.02 210±19 L311V 199±24* 1060±91 0.13±0.02* 125±18* AKR7A2 3659±294 329±33 2.39±0.19 7369±858 V135M 3883±240 411±62 2.54±0.16 6322±661 A142T 1418±83* 842±77* 0.93±0.05* 1109±117* Q157H 3363±163 379±26 2.24±0.15 5836±510

64 V (nmol/min ● k / K ENZYMES VARIANTS max K (µM) k (s -1) a cat m mg protein) m cat (s -1 M -1) E180K 3204±159 366±20 2.10±0.10 5750±412 G198S 3557±174 353±37 2.33±0.11 6652±638 S255N 3274±294 339±21 2.14±0.19 6354±718

Values correspond to the mean ± S.D obtained from three experiments performed with three independent enzyme preparations ( n=9) for each isoform. * Variants significantly different from their respective wild-type enzymes ( p<0.05) a kcat calculated from Mr 41000 (AKR1A1), 38500 (AKR1B1), 40300 (AKR1B10), 39400 (AKR1C1), 40200 (AKR1C2), 41900 (AKR1C3), 39700 (AKR1C4), and 39300 (AKR7A2)

Figure 17—In vitro enzymatic activities for the purified 6x-His tagged (A) AKR1A1, (B) AKR1B1, (C) AKR1B10, (D) AKR1C1, (E) AKR1C2, (F) AKR1C3, (G) AKR1C4, and (H) AKR7A2 wild-type and variant enzymes with doxorubicin. Activities were measured by following the rate of doxorubicinol production. Three independent batches of each enzyme were purified and assays were performed in triplicate with each batch. Enzymatic activities are reported as mean ± S.D. ( n=9). Dotted lines refer to variants that have significant differences in kcat /Km parameter values compared to the wild-type.

66 Table 7—Kinetic constants for DOX metabolism by recombinant 6x-His tagged AKR wild-type and variant enzymes.

Vmax (nmol/min ● -1 a kcat / Km ENZYMES VARIANTS K (µM) k (s ) -1 -1 mg protein) m cat (s M ) AKR1A1 75±10 540±124 0.051±0.006 95±11 N52S 72±10 605±150 0.049±0.007 82±19

E55D 54±10 479±102 0.037±0.004 77±12 AKR1B1 193±12 813±17 0.12±0.01 153±10 I15F 191±23 851±16 0.12±0.01 144±18 H42L 167±8 840±11 0.11±0.01 127±10 L73V 167±19 829±19 0.11±0.01 129±14

K90E 168±19 854±23 0.11±0.01 126±12 G204S 160±15 844±17 0.10±0.01 123±21 T288I 160±17 805±26 0.10±0.01 128±16 AKR1B10 11±1 508±64 0.007±0.001 14±3 P87S 9±1 530±48 0.006±0.001 11±3 M286T 15±2 521±34 0.010±0.002 19±4 D313N 12±2 489±40 0.008±0.001 16±3 AKR1C1 17±2 643±52 0.011±0.002 17±4 R170H 14±1 670±48 0.009±0.001 13±3

Q172L 19±1 651±34 0.012±0.001 18±3 AKR1C2 56±11 780±67 0.038±0.007 49±12 F46Y 63±16 708±104 0.043±0.011 62±21 AKR1C3 546±27 142±29 0.38±0.02 2514±268 H5Q 570±18 161±12 0.40±0.01 2485±241 R66Q 509±14 158±14 0.36±0.01 2266±194 A106T 527±16 138±13 0.37±0.01 2692±297 R170C 323±23 * 188±17 0.23±0.02 * 1208±115 * P180S 306±27 * 273±20 * 0.21±0.02 * 786±95 * AKR1C4 73±13 962±64 0.049±0.008 51±9 G135E 70±12 974±37 0.047±0.008 48±8 S145C 71±11 992±63 0.047±0.007 48±9 L311V 69±12 935±41 0.046±0.008 49±8 AKR7A2 237±15 340±25 0.16±0.01 457±17 V135M 215±24 361±20 0.14±0.02 393±56

67 Vmax (nmol/min ● -1 a kcat / Km ENZYMES VARIANTS K (µM) k (s ) -1 -1 mg protein) m cat (s M ) A142T 141±19 * 501±23 * 0.09±0.01 * 184±24 * Q157H 214±16 362±20 0.14±0.01 385±26 E180K 226±28 367±19 0.15±0.02 404±52 G198S 222±19 374±17 0.15±0.01 389±33 S255N 220±13 377±18 0.14±0.01 383±38

Eadie-Hofstee plots for the AKR wild-type and variant enzymes verified linearity at the same concentrations of DAUN and DOX used to conduct the assays ( r2 > 0.84 for

all plots). Furthermore, by comparing kcat /Km values, we observed that DAUN is a better

substrate for the AKR wild-type and variant enzymes compared to DOX (26.2 to 39.5-

fold higher for AKR1A1, 1.9 to 3-fold higher for AKR1B1, 129.1 to 256.5-fold higher

for AKR1B10, 4.9 to 6.7-fold higher for AKR1C1, 2.4-8.2 fold higher for AKR1C3, 2.5-

4.5 fold higher for AKR1C4, and 6.0-17.1 fold higher for AKR7A2). Only the AKR1C2

enzyme (wild-type and variant) exhibited no significant difference in kcat /Km values

between DAUN and DOX; thereby, suggesting that both drugs are equally good

substrates for this enzyme.

2.4 Discussion

The central focus of this in vitro study was to examine the effect of ns-SNPs on

the metabolic activity of human AKRs using test substrates, as well as DAUN and DOX

as substrates. The wild-type and naturally occurring variant alleles of the human AKR

genes were cloned into an E. coli expression vector with a 6x-His tag added to the amino

68 terminus of the expressed protein. The respective wild-type and variant proteins for each gene were purified and purity assessed by electrophoretic and Western blot analyses. To determine whether the 6x-His tag influenced enzyme activity, the tag was removed by incubation with FXa and the enzyme repurified. Using the test substrates, the activity of the tagged enzyme did not significantly differ from its native counterpart. As a result, we were able to assess the impact of the single amino acid substitutions on the activity of each enzyme, without cleaving the tag in the subsequent assays involving DAUN and

DOX.

In these assays, the carbon-13 alcohol metabolites of DOX and DAUN, DOXol and DAUNol, respectively, were quantified since previous published studies demonstrated that they are the major metabolites in cancer patients receiving treatment with either of these anthracycline anti-cancer drugs (Lipp and Bokemeyer, 1999; Plebuch et al., 2007). Our assays show that both DOX and DAUN are able to be converted to their respective major metabolites by the wild-type and variants enzymes for the AKRs examined in this study.

Out of the 28 allelic variants that were studied (excluding the E36Term truncated variant), 7 were found to significantly decrease metabolism of DAUN compared to their respective wild-type enzymes: the E55D and N52S variants forms of AKR1A1; the

A106T, R170C, and P180S variant forms of AKR1C3; the L311V variant of AKR1C4; and the A142T variant of AKR7A2. In the presence of DOX as a substrate, 3 of the 7 aforementioned variants significantly reduced activity compared to their respective wild- type enzyme: R170C, P180S and A142T. To our knowledge, this study is the first demonstration of the effect of these variant enzymes on DAUN and DOX metabolism.

69 We used three-dimensional models of human AKR1A1, AKR1C3, AKR1C4 and

AKR7A2 enzymes (Figure 18) to examine the location of the 7 mutations relative to the active and cofactor binding sites. Reduction of the carbonyl group by AKR enzymes is thought to involve the co-operation of four amino acids (tyrosine, lysine, aspartic acid and histidine) forming a catalytic tetrad, which are positionally conserved in the individual reductases within this superfamily (Jez et al., 1997; Penning 2003; Barski et al., 2008). The locations of these conserved residues for the 4 aforementioned AKR isoforms are as follows: AKR1A1: aspartic acid-45, tyrosine-50, lysine-80, and histidine-

113; AKR1C3 and 1C4: aspartic acid-50, tyrosine-55, lysine-84, and histidine-117;

AKR7A2: aspartic acid-72, tyrosine-77, lysine-105, and histidine-141. The roles that these residues play in the reduction of carbonyl-containing substrates to their alcohol products are explained in Figure 19 (Schlegel et al. 1998; Di Luccio et al., 2006; Jin and

Penning, 2007; Barski et al., 2008; Di Costanzo et al., 2009). In the case of AKR1A1, the locations of the N52S and E55D mutations occur close to the catalytic tyrosine-50 residue, which can likely hinder the role of tyrosine-50 in the metabolism of the anthracyclines.

The R170C and P180S mutations in AKR1C3 are located close to amino acid residues in the cofactor binding site that participate in hydrogen bonding to the nicotinamide ring of NADPH (serine-166 in β-sheet 5, asparagine-167 in loop 5, and

glutamine-190 in β-sheet 6), and therefore help anchor the cofactor to the AKR1C3 enzyme (Komoto et al., 2004). It is possible that these mutations affect the binding of the cofactor, which ultimately reduces the enzymatic conversion of DAUN and DOX to

DAUNol and DOXol, respectively. Currently, there is no data to indicate why or how the

A106T polymorphism affects cofactor or substrate binding. It is likely chemical

Figure 18—Three-dimensional molecular structures of human (A) AKR1A1, (B) AKR1C3, (C) AKR1C4, and (D) AKR7A2 wild-type enzymes. In these structures, the latter three are shown to be complexed with the cofactor, NADP + (green) [Protein Data Bank ID: 2ALR (AKR1A1), 2FGB (AKR1C3), 2FVL (AKR1C4), and 2BP1 (AKR7A2)]. The mutations that significantly altered enzyme activity are shown (purple) as well as those that did not alter enzyme activity (red). In addition, the residues comprising the catalytic tetrad are illustrated (blue) [AKR1A1: (1) histidine-113, (2) lysine-80, (3) tyrosine-50, and (4) aspartic acid-45; AKR1C3 and 1C4: (1) histidine-117, (2) lysine-84, (3) tyrosine-55, and (4) aspartic acid- 50; AKR7A2: (1) histidine-141, (2) lysine-105, (3) tyrosine-77, and (4) aspartic acid-72]. Also, provided are three residues of AKR1C3 that participate in cofactor binding: serine-166, asparagine-167, and glutamine-190 (pink). The molecular graphic images of the AKRs were produced using the UCSF Chimera program (Resource for Biocomputing, Visualization, and Informatics, University of California, San Francisco, CA).

Figure 19—Proposed catalytic mechanism for AKR-mediated reduction reaction [modified based on the scheme provided by Barski et al. (2008)]. The numbering of the four catalytic residues is illustrated for AKR1A1. Top panel: In the reduction of the carbonyl-containing substrate, the tyrosine-50 residue is shown to be a proton donor for the carbonyl substrate. It also forms a hydrogen bond with the substrate, resulting in carbonyl polarization, which accelerates the hydride transfer of the pro-R hydrogen from NADPH to the carbonyl carbon. The hydrogen-bonding network provided by lysine-80 and aspartic acid- 45 serves to lower the pK a of tyrosine, making the proton transfer from tyrosine to the substrate easier. Bottom panel: The reduced carbonyl (i.e., the alcohol-containing substrate) then dissociates from the active site, and a net charge on the tyrosinate anion is stabilized by the hydrogen bonding network. Histidine-113 also participates in the catalytic mechanism by donating protons to the tyrosinate anion to restore the conserved tyrosine residue.

72 structural differences between the side chain groups of these amino acids [alanine: -CH 3; threonine: -CH(OH)(CH 3)] may be enough to alter the conformation of the AKR1C3 protein, thereby changing the structure of the active and/or cofactor binding sites. Since a significant difference in metabolic capability was detected between this variant and the wild-type in the presence of DAUN, and not DOX, it is possible that the active site is affected.

Previous studies have shown the importance of leucine-311 in substrate binding to human AKR1C4. For example, a study by Matsuura et al. (1998), which used chimeric enzymes produced by switching the C-terminal loop in AKR1C4 with that of AKR1C1, demonstrated that the binding of substrates, inhibitors and activators of AKR1C4 requires that the amino acid residues are located in the C-terminal loop of the wild-type AKR1C4 enzyme. Furthermore, the authors found that a leucine to valine mutation in amino acid

311 (L311V) decreased the catalytic activity of AKR1C4 for its substrates, but it did not affect the enzyme’s sensitivity to an inhibitor or affect the enzyme’s response to an activator. Hence, this study suggested that leucine-311 is important in substrate binding to the active site of the AKR1C4 enzyme. This finding was further substantiated in a study by Kume at al. (1999), which used purified recombinant human AKR1C4 enzymes to show that the L311V polymorphism resulted in a 3- to 5-fold decrease in enzyme activity compared to the wild-type for a variety of xenobiotic and steroidal substrates

[e.g., ( S)-1-indanol, cholic acid, andosterone, naloxone, naltrexone]. In addition to the effect of the L311V mutation, this study also revealed that the S145C ns-SNP (serine to cysteine mutation at amino acid position 145) had no significant effect on enzyme activity compared to the wild-type for a variety of xenobiotic and steroidal substrates

[e.g., ( S)-1-indanol, cholic acid, andosterone, naloxone, naltrexone]. As well as the effect

73 of the L311V mutation, this study also revealed that the S145C ns-SNP (serine to cysteine mutation at amino acid position 145) had no significant effect on enzyme activity compared to the wild-type. Our study demonstrates that the L311V mutation dramatically alters the ability of the enzyme to metabolize DAUN and DOX and, like the previous studies on xenobiotic and steroidal substrates, indicates that the S145C mutation has little affect on the enzymes ability to use DAUN and DOX as substrates.

The alanine to threonine mutation at amino acid position 142 in human AKR7A2

(the A142T ns-SNP) is adjacent to the histidine-141 residue, which is one of the four amino acids that forms a catalytic tetrad for reduction of the substrate (others being tyrosine-77, lysine-105, and aspartic acid-72). Due to this proximity, the A142T polymorphism may hinder the ability of histidine-141 to participate in catalysis of DOX and DAUN, thus, leading to reduced enzymatic activity compared to the wild-type

AKR7A2 enzyme.

In addition to looking at the differences in enzyme activity between the wild-type and variant enzymes with both anthracyclines, our studies demonstrated that DAUN was generally a better substrate for all wild-type and variant forms of the human AKRs, as shown by 1.9- to 257-fold increases in kcat /Km over DOX. However, this was not the case

for AKR1C2 as both DAUN and DOX are equally good substrates. Also, we found that

the human AKR1C3 wild-type enzyme is the most efficient at metabolizing both DAUN

and DOX as demonstrated by having the highest kcat /Km values compared to the

remaining 7 wild-type AKR enzymes. The same could be said for AKR7A2 in the

presence of DAUN.

Overall, our study demonstrates that AKR enzymes are able to metabolize both

DAUN and DOX, with the AKRs having higher specificity for the former anthracycline

74 drug. Significant reductions in enzyme activity were discovered in 7 of 28 allelic variants of AKRs that we examined. All of these variants may contribute to the interpatient variability seen with the development of serious cardiac side effects in

DAUN- and DOX-treated patients. If AKRs play a major role in anthracycline-induced cardiotoxicity, we propose that this condition is largely the result of accumulation of the parent drug. The data collected begs the question of whether individuals with one or more of the aforementioned AKR polymorphisms are at higher risk for developing cardiac side effects after treatment with either DAUN or DOX. This issue of correlating naturally-occurring allelic variation with cardiotoxicity can be addressed with clinical association studies, which are currently underway in our laboratory.

75 CHAPTER 3: ALLELIC VARIANTS OF HUMAN CARBONYL REDUCTASES DEMONSTRATE REDUCED IN VITRO METABOLISM OF DAUNORUBICIN AND DOXORUBICIN 1,2

3.1 Preface

Aldehydes and ketones are reactive carbonyl-containing compounds present in a broad range of natural and synthetic compounds to which living organisms are exposed: aflatoxins, food ingredients, pharmaceutical drugs, and environmental pollutants

(Matsunaga et al., 2006; Hoffman and Maser, 2007). The reactive nature of aldehydes and ketones can be detrimental to living organisms. For example, they can (i) interact directly with proteins and membranes, thereby resulting in loss of function to enzymes, membrane transporters, transcription factors, signaling components, microtubules, and other proteins (Karlhuber et al., 1997; Picklo et al., 2002); (ii) trigger apoptotic pathways ultimately leading to cell death via mechanisms involving the activation of caspases

(Kruman et al., 1997; Ji et al., 2001; Li et al., 2006); and (iii) form adducts with DNA

(Ellis, 2007). All of these molecular perturbations can ultimately lead to cell death (Li et al., 1996; Li et al., 2006). Luckily, organisms have evolved several enzyme systems for detoxifying aldehydes and ketones in order to minimize cellular damage. Such well established pathways include the oxidation of aldehydes to their corresponding carboxylic acids by aldehyde dehydrogenases and aldehyde oxidases, as well as the

1 A version of this chapter has been published :

Bains OS, Karkling MJ, Grigliatti TA, Reid RE, and Riggs KW. Two non-synonymous single nucleotide polymorphisms of human carbonyl reductase 1 demonstrate reduced in vitro metabolism of daunorubicin and doxorubicin. Drug Metab Dispos 2009; 37: 1107-1114.

Bains OS, Karkling MJ, Lubieniecka JM, Grigliatti TA, Reid RE, Riggs KW. Naturally occurring variants of human CBR3 alter anthracycline in vitro metabolism. J Pharmacol Exp Ther , 2010; 332(3): 755-763.

2 See Co-authorship statement for details of contribution of this manuscript .

76 reduction of aldehydes and ketones into the corresponding alcohols by NADPH- dependent reductases (Matsunaga et al., 2006).

One group of enzymes capable of metabolizing carbonyl-containing compounds is the carbonyl reductases (CBRs). CBRs are ubiquitously expressed in different human tissues and are considered to be of clinical importance since they play a major role in metabolism of the highly effective anthracycline drugs, daunorubicin (DAUN) and doxorubicin (DOX) (Wirth and Wermuth, 1992; Forrest and Gonzalez, 2000; Licata et al., 2000; Lakhman et al., 2005; Gonzalez-Covarrubias et al., 2007; Blanco et al., 2008).

Even though DAUN and DOX are effective cancer therapies, their use is limited due to the onset of adverse side effects which are highly variable from patient to patient. The most concerning is the development of life-threatening cardiotoxicity since it can result in irreversible complications such as reduced left ventricular ejection fraction or congestive heart failure (Barry et al., 2007; Deng and Wojnowski, 2007; Menna et al.,

2007). The cause of the interpatient variation in DAUN- or DOX-associated adverse effects is unknown; nonetheless, one contributing factor may be the altered metabolism of these anthracyclines by allelic variants of CBR enzymes. The variants are those arising from non-synonymous single nucleotide polymorphisms (ns-SNPs) in CBR genes.

Therefore, the central focus of this study is to improve our understanding, and

compare the function, of the wild-type and variant enzymes on the in vitro metabolism of

DOX and DAUN. There are a total of 10 documented ns-SNPs in the human CBR gene

listed in the National Centre for Biotechnology Information Database with allele

frequencies ranging from 1.7 to 62.5% in specific ethnic populations (Table 8). While

there are documented studies demonstrating the catalytic properties of the V88I variant

enzyme with DAUN and V244M variant with DOX (Lakhman et al., 2005; Gonzalez-

77 Covarrubias et al., 2007; Blanco et al., 2008), none are reported for these variants in the presence of DOX and DAUN, respectively. Furthermore, there is no information on the catalytic properties of the 8 remaining enzyme variants with either anthracycline as substrates. Here we investigate the wild-type and the 10 known variants of human CBR3 for the in vitro metabolism of DAUN and DOX to the corresponding carbon-13 alcohol

metabolites, daunorubicinol (DAUNol) and doxorubicinol (DOXol). Using purified,

bacterially-expressed, human histidine-tagged enzymes, we demonstrate that 5 variants

(CBR1: V88I and P131S; CBR3: C4Y, V244M, and V93I) had significantly reduced

catalytic efficiencies ( kcat /Km) compared to their corresponding wild-type enzymes. No

significant differences in kcat /Km were detected for the lone L70M variant of CBR4 with either DAUN or DOX. In addition, we observed that DAUN is a better substrate than

DOX for the wild-type and variant isoforms of CBR1 as seen with a 35 to 46-fold increase in kcat /Km values. On the other hand, DOX was found to be a better substrate than DAUN for both CBR3 (2.3 to 4.7-fold increase in kcat /Km) and CBR4 (32 to 40-fold increase in kcat /Km).

Table 8—Allele frequencies of the non-synonymous single nucleotide polymorphic variants of human CBR enzymes from different ethnic groups .

NCBI rs- a VARIANTS b ENZYMES number ALLELE FREQUENCIES CBR1 V88I rs1143663 YRI=2.1% (n=48)

(EC 1.1.1.184, EC 1.1.1.189, P131S rs41557318 CEU=2.2% (n=46) EC 1.1.1.197)

CBR3 P131S rs16993929 YRI=15.8% (n=120)

(EC 1.1.1.1849) CEU=30.0% (n=120); HCB=33.3% (n=90); V244M rs1056892 JPT=30.0% (n=90); YRI=52.5% (n=120)

CEU=48.3% (n=120); HCB=53.3% (n=90); C4Y rs8133052 JPT=47.7% (n=88); YRI=23.7% (n=118)

M235L rs4987121 YRI=2.5% (n=120)

78 L84V rs9282628 CEU=2.2% (n=46); YRI=11.8% (n=34)

V93I rs2835285 CEU=1.7% (n=120)

D239V rs11701643 Not available

CBR4 CEU=24.2% (n=120); HCB=2.2% (n=90); L70M rs2877380 (EC 1.-.-.-, JPT=3.3% (n=90); YRI=62.5% (n=120) EC 1.1.1.) a Below the name of the enzyme are the Enzyme Commission (EC) numerical classification numbers, which were obtained from the Universal Protein (UniProt) database (http://www.uniprot.org/). Based on the chemical reactions they catalyze, some of the enzymes have more than one EC number. b Allele frequencies were obtained from The National Center for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov/). The ethnic groups are designated as follows: YRI (African), CEU (European), HCB (Chinese), and JPT (Japanese). The chromosome sample counts ( n) for each variant are also given.

3.2 Methods

3.2.1 Chemicals and enzymes

Agarose, chloramphenicol, daunorubicin, doxorubicin, idarubicin, kanamycin sulfate, lysozyme, menadione, methanol, potassium phosphate (KH 2PO 4), RNaseI, sodium phosphate (NaH 2PO 4), N,N,N’,N’ tetramethylethylenediamine and NADPH were supplied by Sigma-Aldrich (St. Louis, MO). HPLC-grade acetonitrile, agar, ammonium persulfate, formic acid, ethanol, glycine, glycerol, glacial acetic acid, imidazole, and Tris were purchased from Fisher Scientific Co. (Fair Lawn, NJ). NaCl and yeast extract were ordered from EMD Chemicals Inc. (Darmstadt, Germany). Bacto tryptone and isopropyl

β-D-1-thiogalactopyranoside (IPTG) were obtained from BD Biosciences (Franklin

Lakes, NJ) and Fermentas Inc. (Hanover, MD), respectively. Tween 20 was purchased from EMD Biosciences (La Jolla, CA), and DNaseI was provided by Boehringer

Manheim GmbH (Manheim, Germany). Klenow fragment, T4 ligase, Factor Xa, and

79 restriction enzymes were purchased from New England Biolabs (Ipswich, MA).

Doxorubicinol was obtained from Qventas Inc. (Branford, CT). Pooled human liver cytosol was purchased from BD Biosciences (San Jose, CA).

3.2.2 Molecular cloning of human CBR genes and creation of the genetic variants

The human CBR1 wild-type coding region, which was excised from a pOTB7

recombinant plasmid (MGC-1920, American Type Culture Collection (ATCC), Manassas,

VA) using BssHII (blunt end with Klenow fragment) and XhoI , was subcloned into NheI

(blunt end)-XhoI sites of the pET28a prokaryotic expression vector (EMD, Novagen, San

Diego, CA) with T4 ligase. This construct gave rise to a human CBR1 enzyme with a

6x-His tag separated by an 18-amino acid residue linker on the amino terminus (Figure

20). A Factor Xa (FXa) site was inserted at the amino terminus between the linker and

CBR1 gene using the QuikChange ® Site-Directed Mutaganesis Kit (Stratagene, La Jolla,

CA) with the 5' - CCCCGTTCAGCCATAGAAGGAAGA ATGTCGTCCGGC - 3'

(forward) and 5' - GCCGGACGACATTCTTCCTTCTAT GGCTGAACGGGG - 3'

(reverse) primers (the FXa site is underlined in the primer sequences). The QuikChange ® polymerase chain reaction (PCR) amplification protocol for site directed mutagenesis was modified as follows: two separate 50 µl reactions (one for each primer) were subjected to

10 cycles of denaturation at 95 oC for 1 min, annealing at 60 oC for 1.5 min, and elongation at 68 oC for 6.5 min. A 25 µl aliquot of each PCR amplification reaction was combined with 0.75 µl PfuTurbo ® DNA polymerase (Stratagene). This reaction was subjected to 18 cycles of the same PCR protocol above.

Figure 20—A translated product of the pET28a-CBR1 construct with modifications to the amino terminus of AKR1A1: a 6x-His-tag followed by an amino acid linker and a Factor Xa (FXa) recognition site. The translated product of the other pET28a-CBRs (not shown) has similar amino terminus modifications.

The human CBR3 wild-type coding region was excised from a pOTB7

recombinant plasmid (MGC-3489, American Type Culture Collection (ATCC), Manassas,

VA) using BlpI (blunt end with Klenow fragment) and XhoI and subcloned into NheI

(blunt end)-XhoI sites of the pET28a prokaryotic expression vector (EMD, Novagen, San

Diego, CA) with T4 ligase. This construct gave rise to a human CBR3 enzyme with an

amino terminal 6x-His tag separated from the enzyme by a 15-amino acid residue linker.

A Factor Xa (FXa) cleavage site was inserted at the amino terminus between the linker

and CBR3 gene using the same site directed mutagenesis conditions for CBR1 with the 5'

– GCTAGCTCAGGCATAGAAGGAAGA ATGTCGTCCTGC – 3' (forward) and 5' –

CGAGGACGACATTCTTCCTTCTAT GGCTGAGCTAGC – 3' (reverse) primers (the

FXa site is underlined in the primer sequences).

The human CBR4 wild-type coding regions were PCR amplified from a pCMV-

SPORT6 recombinant plasmid (Invitrogen) using the following primers, which contained

an EcoRI adapter in the forward primers and a XhoI adapter in the reverse primers

(adapters are underlined): 5' – GTACCGCTCGAATTC ATGGACAAAGTGTGTGCTG

– 3' (forward) and 5' – GTCTGCTAACTCGAG GTGCCCTTGATGCTAATC – 3'

(reverse) for CBR4. PCR was performed in a 50 µl reaction buffer containing 100 ng of

81 template, 200 ng of each primer, 0.2 mM dNTPs, and 1.25 units of PfuTurbo ® DNA

polymerase. The amplifying conditions for PCR involved an initial denaturation step at

95 oC for 1 min, followed by 35 cycles of denaturation at 95 oC for 30 sec, annealing at

55 oC for 1 min, and extension at 68 oC for 10 min. Subsequently, a final extension

reaction was performed at 68 oC for 10 min. The CBR4 PCR product was cut with EcoRI

and XhoI , and then subcloned into pET28a. This gave rise to a human CBR4 enzyme

with an amino terminal 6x-His tag as well as a 26-amino acid residue linker between the

tag and enzyme. A FXa cleavage site was inserted at the amino terminus between the

linker and gene via site directed mutagenesis with the subsequent primers: 5' –

CGGATCCGAATTCATAGAAGGAAGA ATGGACAAAGTGTGTGC– 3' (forward)

and 5' –GCACACACTTTGTCCATTCTTCCTTCTAT GAATTCGGATCCG– 3'

(reverse) for CBR4 (the FXa sites are underlined in the primer sequences).

The pET28a-variant constructs were created by site directed mutagenesis using

the QIAGEN (Mississauga, Ontario) protocol with primers listed in Supplemental Table

2 (see Appendices, page 186). All constructs were verified by dideoxy sequencing at the

University of British Columbia Nucleic Acid Protein Service unit

(http://naps.msl.ubc.ca/).

3.2.3 Expression and purification of recombinant human CBR wild-type and variant enzymes

The pET constructs of the AKR wild-type and variants were heat-shock

transformed into Escherichia coli BL21 (DE3) pLysS competent cells and expressed

under the control of an IPTG-inducible T7 polymerase. Cells were plated on Luria-

Bertani broth agar (1% bacto-tryptone, 0.5% yeast extract, 0.5% NaCl) supplemented

82 with antibiotics (25 µg/ml chloramphenicol and 50 µg/ml kanamycin sulphate for the pET28a constructs or 100 µg/ml ampicillin for the pET15b constructs). Colonies were randomly picked and cultured overnight at 37 oC in 3 ml of Luria-Bertani broth with

kanamycin and chloramphenicol at the same concentrations stated previously. Cultures

were expanded to 800 ml and grown at 37°C until an OD 600 of 0.5 was reached. IPTG

was added to a final concentration of 1 mM and cells were allowed to grow for an

additional 5 hrs, after which the cultures were harvested by centrifugation (4000x g for 20

min at 4 oC).

Expression and purification of the 6x-His tagged CBR1 and CBR3 enzymes (both wild-type and variants) were performed using the same protocols as described in Chapter

2 with the AKRs. In the case of CBR4 wild-type and the L70M variant, the expression protocol was similar to that of CBR1 and CBR3 (starting with the heat shock transformation of the pET construct up to the bacterial culture harvestation by centrifugation); however, the purification protocol was different. The harvested cells were resuspended in Buffer B (100 mM NaH 2PO 4, 10 mM Tris-Cl, pH 8.0) with 8 M urea for lysis. The extracted CBR4 and L70M proteins were subjected to purification by

Ni-NTA chromatography under denaturing conditions according to the QIAGEN protocol. In the end, CBR4 and L70M were eluted using multiple fractions of Buffer B with decreasing pH levels (6.3, 5.9 and 4.5). The eluted fractions were then dialyzed at

4oC to gradually remove the urea (Buffer B with 6 M, 4 M, 2 M and 1 M urea for 2 hrs

each followed by 0 M urea overnight).

The protocols for protein purity assessment and Western blotting were similar to that of the AKR study (section 2.2.3, pages 47-50 in thesis). However, for Western blotting, different primary monoclonal antibodies were used: a monoclonal mouse anti-

83 human CBR1 antibody (Abnova ® Corporation, Taipei City, Taiwan) (diluted 1:3000), as

well as MaxPab polyclonal mouse anti-human CBR3 and CBR 4 antibodies (Abnova ®)

(diluted 1:2500).

3.2.4 Kinetic analysis of CBR wild-type and variant enzymes

The enzyme activities of the purified 6x-His-tagged CBR wild-type and variant enzymes were measured at 37°C using a Fluoroskan Ascent ® FL fluorometer (Thermo

Fisher Scientific, Waltham, MA) by following the initial rate of NADPH oxidation at

excitation and emission wavelengths of 355 and 460 nm, respectively. The assays were

conducted as described previously using menadione as the test substrate (Lakhman et al.,

2005; Gonzalez-Covarrubias et al., 2007; Endo et al., 2008). In brief, purified protein (3

µg) was incubated with NADPH (1 mM for CBR1 and CBR3; 0.2 mM for CBR4) and

menadione (CBR1: ranging from 20 to 150 µM; CBR3: 20-100 µM; CBR4: 5-500 µM)

in a reaction mixture of 150 µl of 100 mM potassium phosphate, pH 7.4. Protein amount

and incubation times were selected for each enzyme and substrate concentration to ensure

that measured rates were in the linear range of the enzyme kinetic curve. In these assays,

the concentration of 95% ethanol, which was required to dissolve the substrate, was

below 4% (v/v) in the final reaction mixture. Readings were collected at 20-sec intervals

for 1.5 hrs with shaking between each reading. Maximal rates were calculated from the

Ascent program (version 2.6) using a 5-min interval (15 total readings) with the steepest

slope. Enzymatic activity (nanomoles of NADPH consumed per minute per milligram of

purified protein) was calculated from these rates using a standard curve constructed from

the fluorescence measurements of solutions of known NADPH concentrations.

84 Activity measurements for the reduction of the anthracyclines were performed by incubating either DOX or DAUN (CBR1: 10 to 400 µM; CBR3 and CBR4: 10 to 700

µM) with purified enzyme (3 µg) for 30 min in a total volume of 150 µl buffer containing

100 mM potassium phosphate, pH 7.4, and 1 mM NADPH at 37°C. The samples were mixed at 250 rpm. Protein concentrations were based on the Bradford protein assay using bovine serum albumin as a standard. The reaction was stopped by adding 300 µl of ice- cold acetonitrile, which contained idarubicin as an internal standard, followed by vortex mixing and centrifugation at 10,000xg for 10 min at 4°C to remove protein. The supernatant was removed for high performance liquid chromatography (HPLC) analysis.

The procedures for HPLC separation and fluorescence detection of DOX and DAUN and their carbon-13 hydroxy metabolites, DOXol and DAUNol, were carried out as previously described (Bains et al., 2008).

The kinetic constants of Vmax and Km were determined by fitting the rate measurement data using nonlinear least-squares fitting to a Michaelis-Menten hyperbola

(GraphPad Prism version 4.0; GraphPad Software Inc., San Diego, CA). The turnover values ( kcat ) were calculated from Vmax values using the apparent molecular weight for the

6x-His-tagged CBR and variant proteins of 34 kDa (CBR1 and CBR3) or 29.6 kDa

(CBR4). The values for kcat /Km were also calculated. Following Michaelis-Menten data analysis, Eadie-Hofstee plots were generated to check for deviation from linearity with changing substrate concentrations.

3.2.5 Statistical analysis

Statistical analyses were performed using the same protocol as described in the

AKR study (section 2.2.6, page 53 of thesis) with results expressed as means ± S.D and

85 enzyme activities compared using a one-way analysis of variance followed by a Tukey-

Kramer multiple comparisons test. Differences were considered significant at p<0.05.

Sample sizes ( n) for the following experiments with the CBR wild-type isoforms and variants were as follows: determination of Michaelis-Menten kinetic parameters using menadione as substrate (n=9), determination of kinetic parameters using DAUN as substrate ( n=9), and determination of kinetic parameters using DOX as substrate ( n=9).

3.3 Results

3.3.1 Expression and purification of the human CBRs

The expression of the 6x-His-tagged recombinant human CBR wild-type enzyme was confirmed by Western blot analysis, which showed a band with mobility corresponding to the calculated molecular mass of the tagged CBR (34 kDa for CBR1, 34 kDa for CBR3, and 29.6 kDa for CBR4) (Figure 21). Total protein staining of the SDS-

PAGE gel demonstrated that the wild-type fraction was purified from its transformed bacterial lysate as no other proteins were detected (Figure 21). For the CBR1, and CBR3 wild-type and variant enzymes, a majority of the pure enzyme was recovered in the 250 mM imidazole elution fractions; no further protein was eluted with higher imidazole concentrations. For the CBR4 wild-type and variant enzymes, the majority of the pure enzyme was recovered in pH 4.5 elution fractions following protein purification under denaturing conditions with 8 M urea. The results for the expression and purification of the variant forms of each enzyme paralleled that of the corresponding wild-type enzyme

(Supplemental Figure 5; see Appendices, page 187).

Figure 21—Purification of human recombinant 6x-His-tagged wild-type enzymes: (A) CBR1, (B) CBR3, and (C) CBR4. ( Left ) Gel stained with SYPRO ® Ruby following SDS-PAGE showing purified protein fraction (lane 3; 2 µg), free of contaminating proteins from the bacterial lysate (lane 1; 10 µg total protein). Removal of contaminating proteins is observed in flow through fraction from Qiagen purification procedures [Ni-NTA column flow through (lane 2; 10 µg total protein)]. ( Right ) Western blot detection of transformed lysate (lane 6) and purified protein fractions (lane 8), confirms expression of the desired CBR protein. Little immunoreactivity was detected in the flow through fraction (lane 7) suggesting that majority of the enzyme was bound to the Ni-NTA resin prior to their elution. GST-tagged human CBR4 recombinant protein (Abnova ® Corporation, Taipei City, Taiwan; lane 5; 2 µg total protein) and human liver cytosol (lane 5; 20 µg total protein; for CBR1 and CBR3) was used as positive controls for antibody immunoreactivity for these enzymes. No antibody immunoreactivity is observed for untransformed bacterial lysate (lane 4; 10 g total protein). M refers to the molecular weight markers.

3.3.2 Kinetic characterization of CBR wild-type and variant enzymatic activities with menadione

Before running the kinetic assays with DAUN and DOX, the human CBR wild- type and variant enzymes were subjected to assays with test substrates in order to confirm that they were active following purification by comparing enzymatic activity values with published studies. Another reason to perform these assays was to determine which variant enzymes exhibit significantly altered activity compared to the wild-type. This

87 information is important since we would expect that these variants are capable of altering metabolism of DAUN and DOX.

For the CBR wild-type and variant enzymes, menadione was used as the test substrate, and Michaelis-Menten kinetic parameter values [maximal reaction velocity

(Vmax ), substrate affinity ( Km), turnover rate ( kcat ), and kcat /Km] were determined and compared with published values (Table 9). The Vmax for the CBR1 wild-type enzyme in

this study is approximately two-fold higher than the value in the literature (Gonzalez-

Covarrubias et al., 2007) for bacterially expressed recombinant human 6x-His-tagged

CBR1. Reasons for this discrepancy between these two studies include: (i) greater CBR1

protein activity following purification in our study, as well as (ii) the use of a more

sensitive method for measuring the rate of NADPH oxidation (fluorescence) in our study

versus UV-visible spectrophotometry, which was used in the Gonzalez-Covarrubias

study. In relation to the reported studies for the CBR3 wild-type enzyme, there were

discrepancies reported for kinetic parameter values ( Vmax , Km, kcat , and kcat /Km) between two studies (Table 9) (Lakhman et al. 2005; Miura et al., 2008). When comparing the kinetic parameter values in these two studies, menadione is metabolized more readily by

CBR3 in the Lakhman study. The parameter values obtained in our study were consistent with the Miura study, even though both studies used the same assays conditions provided by the Lakhman group. To test if CBR3 activity would increase, the menadione assays in our study were performed with higher concentrations of cofactor (1.5 mM and 2 mM); however, the kinetic parameter values did not differ from those reported in the study with

1 mM NADPH. Taken together, menadione is able to be metabolized by CBR3; however, not as efficiently as claimed by the Lakhman group. With the CBR4 wild-type enzyme, the kinetic parameter values that were obtained in this study for menadione were

88 consistent with those reported in Endo et al. (2008). Overall, the CBR1 enzyme exhibited the greatest capability of metabolizing menadione ( kcat /Km: 17400±2800), followed by CBR4 ( kcat /Km: 5670±650), and CBR3 ( kcat /Km: 1320±290).

Table 9—Michaelis-Menten kinetic parameters for test substrate menadione by recombinant 6x-His tagged CBR wild-type and variant allele enzymes.

KINETIC PARAMETERS V (nmol/min ● mg ENZYMES VARIANTS max K (µM) k (s -1) a k / K (s -1 M -1) protein) m cat cat m CBR1 582±43 20±6 0.33±0.02 17400±2800 REPORTED: 220±27 [1] c N/A N/A N/A V88I 300±9* 21±2 0.17±0.02* 8090±710*

P131S 561±43 40±5* 0.32±0.06 7950±900* CBR3 170±29 76±23 0.096±0.014 1320±290 REPORTED: 19600±2000 [2] c 25±3 [2] c 9.9 [2] c 402000 [2] c REPORTED: 126 b [3] c 54±10 [3] c 0.071 [3] c 1300 [3] c P131S 145±22 64±10 0.082±0.013 1280±248 V244M 90±10* 134±6* 0.051±0.007* 381±42* C4Y 85±5* 140±7* 0.048±0.004* 343±48* M235L 158±31 65±13 0.090±0.018 1380±371 L84V 164±26 71±14 0.093±0.015 1310±345 V93I 96±9* 142±8* 0.054±0.006* 380±51* D239V 171±29 80±11 0.097±0.016 1210±253 CBR4 543±41 47±9 0.27±0.02 5670±650 REPORTED: 508 b [4] c 36 [4] c 0.21 [4] c 5833 [4] c L70M 525±37 49±10 0.26±0.02 5330±543

Values correspond to the mean ± S.D obtained from three experiments performed with three independent enzyme preparations ( n=9) for each isoform. Reported parameter values for the wild-type enzymes are also given for comparison purposes . * Variants significantly different from their respective wild-type enzymes ( p<0.05) a k cat calculated for 6x-His tagged CBRs from Mr 34000 (CBR1), 34000 (CBR3), and 29600 (CBR4) b Calculated Vmax based on other parameter values given in study c [1]=Gonzalez-Covarrubias et al., 2007; [2]=Lakhman et al., 2005; [3]=Miura et al., 2008; [4]=Endo et al., 2008 N/A=No value available from the study

89 Some of the CBR variants were found to decrease activity significantly compared to their respective wild-type enzymes ( n=9 ): a 54% reduction in kcat /Km for both the V88I and P131S variants (CBR1) as well as a 71-74% decrease for the V244M, C4Y, and V93I variants (CBR3). There was no significant difference in activity between the CBR4 wild- type and the L70M variant ( n=9). The kcat /Km value was primarily used for comparison

between the variants and wild-type enzymes since it incorporates Vmax , K m, and kcat ; thus,

making it a better measure of enzyme activity.

The 6x-His tag and amino acid linker were removed with Factor Xa (FXa), as

described in Chapter 2, in order to see if this modification on the amino terminus of the

enzyme disrupted activity in our preparation (Figure 22). This was not the case since the

kinetic parameters did not differ significantly from the tagged enzyme [untagged CBR1:

-1 Vmax (489±38 nmol/min•mg), Km (22±5 µM), kcat (0.28±0.02 s ), and kcat /Km

-1 -1 (12600±2100 s M ); untagged CBR3: Vmax (149±31 nmol/min•mg), Km (60±15 µM), kcat

-1 -1 -1 (0.084±0.016 s ), and kcat /Km (1410±400 s M ); untagged CBR4: Vmax (483±39

-1 -1 -1 nmol/min•mg), Km (58±7 µM), kcat (0.24±0.02 s ), and kcat /Km (4140±780 s M )] (all n=9) . Therefore, we deemed it not necessary to cleave off the tag and linker from the wild-type and variant enzymes for the subsequent assays involving DAUN and DOX.

Figure 22—Representative Western blot detection of purified 6x-His tagged CBR proteins [(A) CBR1, (B) CBR4, and (C) CBR4 wild-type enzymes; lane 2; 3 µg] and native AKR proteins (lane 3; 3 µg). The 6x- His tagged proteins were incubated with Factor Xa for 6 hrs at 23 oC to allow for removal of the tag and linker from the amino terminus. GST-tagged human CBR4 recombinant protein (Abnova ® Corporation, Taipei City, Taiwan; lane 1; 2 µg total protein) and human liver cytosol (lane 1; 20 µg total protein; for CBR1 and CBR3) were used as positive controls for antibody immunoreactivity for these enzymes. M refers to the molecular weight markers.

3.3.3 Kinetic characterization of wild-type and variant enzymatic activities with DOX and DAUN

To evaluate the impact of the single amino acid substitutions in the human CBR enzymes on the reduction of the anthracycline drugs, we measured the formation of the

major alcohol metabolites in vitro , as performed in the AKR study (Chapter 2). Full

chromatographic resolution of DAUNol and DOXol from DAUN and DOX, respectively,

and idarubicin (internal standard) was achieved for all chemical standards and in vitro samples. DOXol, DOX, DAUNol, DAUN, and idarubicin were observed to elute at 4.6,

5.5, 6.1, 6.9, and 7.4 min, respectively (Figure 23). There were no detectable peaks at the

DAUNol or DOXol retention time in the absence of the CBR proteins.

Figure 23—Generation of DOXol and DAUNol in vitro by purified CBR3 incubated with DOX (top panel), and purified CBR1 incubated with DAUN (bottom panel). Measurement of DOXol and DAUNol was performed using HPLC-fluorescence. Representative chromatograms show clear resolution of DOXol and DAUNol from DOX, DAUN, and idarubicin (IDA, internal standard). Retention times observed for DOXol, DOX, DAUNol, DAUN, and IDA are 4.6, 5.5, 6.1, 6.9, and 7.4 minutes, respectively.

92 Michaelis-Menten kinetic curves were generated for the CBR wild-types (Figure

24) and each of the variant enzymes in the presence of differing concentrations of

DAUN, ultimately leading to the determination of kinetic parameter values (Table 10). A total of 5 allelic variants of CBR1 and CBR3 exhibited significant reductions in enzymatic activity with DAUN compared to their respective wild-type enzymes: a 50 and

58% reduction in kcat /Km for the V88I and P131S variants of CBR1, respectively [this reduction is due to a decrease in Vmax and kcat (36% for both parameters) for V88I and an increase in Km (93%) for P131S; n=9]; and a 29 to 43% reduction in kcat /Km for the C4Y,

V93I, and V244M variants of CBR3 [this reduction due to corresponding decreases in

Vmax (22-43%) and kcat (21-43%) for all variants; n=9].

Figure 24 —In vitro enzymatic activities for the purified 6x-His tagged (A) CBR1, (B) CBR3, and (C) CBR4 wild-type and variant enzymes with daunorubicin. Activities were measured by following the rate of daunorubicinol production. Three independent batches of each enzyme were purified and assays were performed in triplicate with each batch. Enzymatic activities are reported as mean ± S.D. ( n=9). Dotted lines refer to variants that have significant differences in kinetic parameter values compared to the wild- type.

93 Table 10—Kinetic constants for DAUN metabolism by recombinant 6x-His tagged CBR wild-type and variant enzymes.

V (nmol/min ● k / K ENZYMES VARIANTS max K (µM) k (s -1) a cat m mg protein) m cat (s -1 M -1) CBR1 2990±245 57±15 1.69±0.14 31500±3580 V88I 1900±100* 68±8 1.08±0.06* 15600±2800*

P131S 2540±150 110±10* 1.44±0.09 13090±3460* CBR3 166±5 461±25 0.094±0.003 204±13 P131S 157±15 503±88 0.088±0.009 180±24 V244M 130±9* 587±75 0.074±0.005* 128±22* C4Y 95±7* 471±63 0.054±0.004* 116±19* M235L 149±14 524±91 0.084±0.008 165±33 L84V 161±8 499±50 0.091±0.005 185±28 V93I 120±7* 472±47 0.068±0.004* 145±16* D239V 156±7 454±38 0.088±0.004 197±19 CBR4 9±1 673±94 0.0044±0.0007 7±2 L70M 10±1 664±86 0.0049±0.0007 7±2

With DOX as a substrate (Figure 25; Table 11), the same aforementioned variants of CBR1 and CBR3 demonstrated significant reductions in enzymatic activity with respect to the wild-type enzymes: a 43 and 45% reduction in kcat /Km for both the V88I and P131S variants of CBR1, respectively [this reduction is due to a corresponding decrease in Vmax (32%) for V88I and an increase in Km (54%) for P131S; n=9]; and a 41 to 66% reduction in kcat /Km for the C4Y, V93I, and V244M variants of CBR3 [this reduction is due to corresponding decreases in Vmax and kcat (25-34% for both parameters) for all variants along with an increase in Km (120%) for V93I; n=9]. When compared against the CBR4 wild-type enzyme, there were no significant differences detected in any of the kinetic parameters for the L170M variant with either anthracycline.

Figure 25—In vitro enzymatic activities for the purified 6x-His tagged (A) CBR1, (B) CBR3, and (C) CBR4 wild-type and variant enzymes with with doxorubicin. Activities were measured by following the rate of doxorubicinol production. Three independent batches of each enzyme were purified and assays were performed in triplicate with each batch. Enzymatic activities are reported as mean ± S.D. (n=9). Dotted lines refer to variants that have significant differences in kinetic parameter values compared to the wild-type.

Table 11—Kinetic constants for DOX metabolism by recombinant 6x-His tagged CBR wild-type and variant enzymes.

V (nmol/min ● k / K ENZYMES VARIANTS max K (µM) k (s -1) a cat m mg protein) m cat (s -1 M -1) CBR1 319±29 273±47 0.18±0.02 682±77 V88I 218±19* 310±15 0.12±0.02 390±60*

P131S 280±28 421±28* 0.16±0.02 377±71* CBR3 466±22 278±30 0.264±0.013 965±151 P131S 439±24 317±38 0.249±0.013 797±121 V244M 334±16* 311±33 0.189±0.009* 567±47* C4Y 308±10* 382±25 0.175±0.006* 458±27* M235L 345±23* 270±42 0.196±0.014* 745±151 L84V 452±24 310±36 0.256±0.013 842±130 V93I 352±16* 613±48* 0.199±0.009* 327±24*

95 V (nmol/min ● k / K ENZYMES VARIANTS max K (µM) k (s -1) a cat m mg protein) m cat (s -1 M -1) D239V 432±35 334±46 0.245±0.020 743±116 CBR4 63±8 139±16 0.031±0.005 223±55 L70M 71±9 125±19 0.035±0.004 280±61

3.4 Discussion

The main goal of this in vitro study was to examine the effect of ns-SNPs on the metabolic activities of human CBR1, CBR3, and CBR4 enzymes using menadione as a test substrate, as well as for the anthracyclines DAUN and DOX. The wild-type and naturally occurring variant alleles of the human CBR genes were expressed in bacteria, and then purified by Ni-NTA affinity chromatography. As seen in the AKR study, the removal of the 6x-His tag and amino acid linker from the CBR wild-type enzymes did not influence enzyme activity in the presence of menadione. Therefore, in the ensuing assays with the anthracyclines, the 6x-His tagged enzymes (both wild-type and variants) were used to assess the impact of the single amino acid substitutions on the ability of the enzyme to metabolize DOX and DAUN to their corresponding alcohols DOXol and

DAUNol, respectively. Our assays demonstrated that both DOX and DAUN are readily converted to their respective alcohol metabolites by the wild-type and variants enzyme for the CBRs examined in this study.

Out of the 10 allelic variants that were examined in this study, 5 were found to significantly decrease metabolism of both DOX and DAUN compared to their respective wild-type enzymes: the V88I and P131S variant forms of CBR1, and the C4Y, V93I, and

V244M variant forms of CBR3. To our knowledge, this study is the first to (i) look at the

96 effect of the CBR1, CBR3, and CBR4 variant enzymes on DAUN and DOX metabolism

[except for V244M since and earlier study by Blanco et al. (2008) demonstrated a significant reduction in DAUNol production compared to the wild-type], and (ii) demonstrate that human CBR3 metabolizes DAUN and CBR4 metabolizes both anthracyclines, in vitro.

The finding in this study that DOX is metabolized by the CBR3 wild-type was corroborated by Blanco et al. (2008); however, the metabolic activity of our human

CBR3 was considerably higher. Both studies were in disagreement with an in vitro study by Kassner et al. (2008), which suggested that their histidine-tagged human CBR3 does not metabolize DOX. The reason for the discrepancy between these results is not known.

In relation to the CBR variant, V88I, our enzymatic studies with this enzyme in the presence of DAUN is consistent with the study by Gonzalez-Covarrubias et al.

(2007), which used cofactor (NADPH) consumption to demonstrate that the Vmax for

DAUN metabolism was significantly decreased in the V88I variant compared to the wild- type enzyme. While co-factor usage is an indirect test of substrate conversion, they further confirmed this kinetic difference by directly measuring DAUNol levels using

HPLC-fluorescence at a single DAUN concentration (500 µM), which showed the variant enzyme produced 47% less metabolite than that of the wild-type. These authors suggested that the V88I mutation has an impact on NADPH binding affinity due to its close proximity to the cofactor site (Figure 26a). Their finding demonstrates that the valine to isoleucine substitution leads to a lower affinity for the cofactor. Since NADPH is a vital cofactor in the metabolic conversion of DOX and DAUN to their major alcohol metabolites, this might be the basis for the significant decrease in the levels of DOXol and DAUNol produced with the V88I variant enzyme.

97 In addition, the P131S mutation (Figure 26a) may alter the structural conformation of the cofactor and/or active site of CBR1, which would explain the significant differences seen in the metabolism of both anthracyclines compared to the wild-type enzyme. The distinctive cyclic structure of proline's side chain is an important structural property, which contributes to the precise three-dimensional shapes of proteins.

Proline's linkage to other amino acids through the amino group contributes to various bends and kinks in the shape of the protein, without which the protein could not function properly (Duclohier, 2004; Fu et al., 2009; Moroder and Budisa, 2010). Therefore, the substitution of proline with a serine may disrupt the folding of the CBR1 enzyme, ultimately affecting the active and cofactor binding sites.

With human CBR3, 3 of the 7 allelic variants (C4Y, V93I, and V244M) were found to have significantly altered DOX and DAUN metabolism in our studies. The locations of these alleles are shown in a three-dimensional model of the CBR3 enzyme

(Figure 26b), along with the residues involved in catalysis: tyrosine-194, serine-140, lysine-198, and asparagine-114 (Jornvall et al., 1995; Filling et al., 2002). These are the same residues for catalysis with human CBR1. In the reduction of the carbonyl-

containing substrate, the tyrosine-194 residue has two functions: (i) it forms a hydrogen

bond with the substrate carbonyl, resulting in carbonyl polarization, which accelerates the

hydride transfer of the pro-R hydrogen from the nicotinamide ring of NADPH cofactor to

the carbonyl carbon of the substrate, and (ii) it acts as a proton donor for the substrate

(Figure 27) (Ghosh et al., 1994; Hoffmann and Maser, 2007). Serine-140 serves to

stabilize the substrate via hydrogen bonding, which facilitates proton and hydride transfer from the conserved tyrosine and cofactor, respectively (Grimm et al., 2000). In addition, the hydrogen-bonding network provided by lysine-198 and asparagine-114 serves to

98 lower the pK a of tyrosine, making the proton transfer easier. (Ghosh et al., 1994;

Auerbach et al., 1997; Benach et al., 1999; Hoffmann and Maser, 2007). After the proton transfer from tyrosine-194 and hydride transfer from NADPH, the reduced carbonyl substrate then dissociates, and a net charge on the tyrosinate anion is stabilized by the hydrogen bonding network of lysine-198 and asparagine-114. The conserved tyrosine can be regenerated from the tyrosinate anion through the donation of a hydrogen atom by the surrounding solvent or by serine-140 (Ghosh et al., 1994; Duax et al., 2000; Hoffman and Maser, 2007).

Figure 26—Three-dimensional molecular structure of human (A) CBR1 and (B) CBR3 wild-type enzymes complexed with the cofactor, NADP + (green) [Protein Data Bank ID: 3BHI (CBR1) and 2HRB (CBR3)]. The mutations that significantly altered enzyme activity are shown (purple) as well as those that did not alter enzyme activity (red). For CBR3, since the amino acid sequence begins at position 5 with serine, there is no C4Y ns-SNP illustrated. In addition, the residues comprising the catalytic tetrad are illustrated (blue) [CBR1 and CBR3: (1) tyrosine-194, (2) lysine-198, (3) asparagine-114, and (4) serine-140]. The molecular graphic images of the CBRs were produced using the UCSF Chimera program (Resource for Biocomputing, Visualization, and Informatics, University of California, San Francisco, CA).

Figure 27—A proposed catalytic mechanism for CBR-mediated reduction reaction (see text on pages 98 to 99 for details). This figure is modified based on the scheme provided by Hoffmann and Maser (2007). The numbering of the catalytic residues (tyrosine, lysine, asparagine, and serine) follows that of the human CBR1 and CBR3 wild-type enzymes.

100 Of the mutations significantly affecting DAUN and DOX metabolism, only V93I

is physically located in close proximity to the catalytic and cofactor binding sites. The

change from valine to isoleucine at this position has been proposed to disrupt side-chain

interactions with asparagine-114 (Pilka et al., 2006). This disruption to the active site

may lead to decreased metabolism of DOX and DAUN. Further studies are needed to

provide a more definitive explanation on how the other variants alter CBR3 activity.

Besides looking at the differences in enzyme activity between the wild-type and

variant enzymes with either DAUN or DOX, our studies also demonstrated that DOX is a

better substrate than DAUN for both the wild-type and variant isoforms of human CBR3

and CBR4, as shown by increases in kcat /Km of 2.3 to 4.7-fold and 32 to 40-fold, respectively. The reverse is true for human CBR1 with DAUN being the better substrate than DOX (35 to 46-fold increase in kcat /Km for the wild-type and variants).

In conclusion, this study demonstrates that the 6x-His tagged CBR enzymes are able to metabolize DAUN and DOX. Significant reductions in enzyme activity were discovered in five variant isoforms, which may contribute to the inter-patient variability seen with the development of serious cardiac side effects in patients treated with either

DOX or DAUN. If these reductases play a major role in anthracycline-induced cardiotoxicity, we propose that this condition is largely the result of accumulation of the parent drug. The data collected in this study begs the question of whether individuals bearing one or more of the CBR polymorphisms are at higher risk for developing cardiac side effects after treatment with either of these anthracyclines. This issue can be addressed with clinical association studies, which are currently underway in our laboratory to determine whether there is a correlation with chronic cardiotoxicity. We realize that there are single nucleotide polymorphisms (SNPs) in the non-coding regions

101 of the human CBR genes, including untranslated regions, introns, and promoter regions, which may influence the amount of protein produced. Even though a study by Zhang and

Blanco (2009) found that two common SNPs in the CBR3 promoter [-725 T>C

(rs2239566, q=12.7-22.5%) and -326 T>A (rs8132243, q=2.4-17.5%)] did not modify gene promoter activity or significantly alter hepatic CBR3 mRNA levels compared to the wild-type, there are a number of common SNPs in the non-coding regions that remain to be investigated. The clinical association studies will examine these SNPs as well as the ns-SNPs in the human CBR enzymes to see if there is a correlation with chronic cardiotoxicity.

102 CHAPTER 4: EX VIVO AND IN VITRO STUDIES SUGGEST A NEGATIVE ASSOCIATION BETWEEN CYTOTOXICITY AND METABOLISM OF THE ANTHRACYLINES DAUNORUBICIN AND DOXORUBICIN BY ALDO-KETO AND CARBONYL REDUCTASES

4.1 Preface

Daunorubicin (DAUN) and doxorubicin (DOX) are two of the most effective antineoplastic agents ever developed (Licata et al., 2000; Lakhman et al., 2005; Blanco et al., 2008). Conversely, their usage is often associated with considerable inter-patient variability in the onset of life-threatening adverse cardiotoxic events (Wojtacki et al.,

2000; Mordente et al., 2001; Danesi et al., 2002; Barry et al., 2007; Deng and Wojnowski,

2007; Menna et al., 2007). One possible cause of this variation could be disparity in the rate at which these drugs are metabolized, among patients subjected to them. Two groups of enzymes that have been implicated in the metabolism of DAUN and DOX are the aldo-keto reductases (AKRs) and carbonyl reductases (CBRs) (Cummings et al., 1991;

Jin and Penning, 2007).

Currently, there is disagreement on the role that drug metabolism plays in generating the DAUN- and DOX-associated toxicity. Some studies suggest that the major metabolites, daunorubicinol (DAUNol) and doxorubicinol (DOXol), are responsible for the adverse effects. For example, Olson et al. (2003) observed that inhibition of CBR1 blocks DOX-induced cardiotoxicity in mice. Similarly, a study by

Forrest et al. (2000) demonstrated that transgenic mice, in which human CBR1 had been

overexpressed in the heart, exhibited high levels of intracardiac DOXol and increased

signs of myocardial damage after DOX administration. Furthermore, Behnia and

Boroujerdi (1999) reported that when DOX was administered to rats, those pre-treated

103 with an AKR inhibitor showed no signs of cardiotoxicity, while the ones that received no pre-treatement exhibited some forms of cardiotoxicity.

On the other hand, there are studies suggesting the parent drug is more toxic. For example, Veitch et al. (2009) showed that a DOX-resistant MCF-7 breast cancer cell line had a 2.17-fold higher DOXol production than wild-type cells. Furthermore, the introduction of an AKR1C inhibitor dramatically altered the DOX-associated toxicity in this resistant cell line, reducing it to a level that was comparable to that of the wild- type. A study by Ax et al. (2000) revealed an 8-fold increase in LC 50 values of DAUN in

DAUN-resistant human stomach carcinoma cells compared to wild-type cells. DAUNol production in the resistant cell line was 6-fold higher and, based on finding significantly elevated levels of CBR1, AKR1B1, and AKR1C2 mRNAs in the resistant cell line, the authors suggested that resistance may be associated with increased DAUN metabolism.

Furthermore, in human erythroleukemia CBR1 transfected cells that were subsequently treated with DAUN, CBR1 activity (conversion of DAUN to DAUNol) increased 83-fold and was associated with a 2- to 3-fold reduction in cytotoxicity compared to non- transfected cells (Gonzalez et al., 1995).

This controversy prompted us to determine whether there is a consistent association between DAUN/DOX metabolism, by AKRs and CBRs, and cellular toxicity. Eight human carcinoma cell lines and one rat embryonic cardiac cell line were chosen for this study. Ex vivo cytotoxicity studies demonstrated that all cell lines were

100% viable to begin with (0 hr incubation time period). In the presence of DAUN or

DOX for 6, 24, and 48 hrs, four of the cell lines [HepG2 (liver), NCI-H460 (lung), HCT-

15 (colon) and A-498 (kidney)] were more tolerant of the toxic effects of these drugs, based on their significantly higher LC 50 values, than the five remaining cell lines [H9c2

104 (heart), PC-3 (prostate), PANC-1 (pancreas), OVCAR-4 (ovary), and MCF-7 (breast)], which were categorized as the sensitive cell lines. These studies were repeated using

DAUNol and DOXol at the 48 hr incubation time period, and LC 50 values for these compounds were found to be significantly higher than those for the parent drugs, thus suggesting that the metabolites are less toxic.

In vitro metabolic studies were performed with extracted cytosolic fractions of the

cell lines following pre-treatment with 100 nM DAUN and DOX separately for 0, 6, 24,

and 48 hrs. These studies revealed the following: (i) at 0 hrs, the tolerant cell lines

metabolized DAUN and DOX to their major metabolites more efficiently compared to

the sensitive cell lines, and (ii) there were similar incremental increases in enzyme

activity for all the cell lines following 6, 24, and 48 hrs incubation with either

anthracycline. One likely hypothesis for these findings is that the tolerant cell lines have

a greater abundance of one or more of the cytosolic AKR and CBR enzymes. Previous in

vitro studies suggest that the following enzymes are able to metabolize DAUN and/or

DOX: AKR1A1, AKR1B1, AKR1B10, AKR1C1, AKR1C2, AKR1C3, AKR1C4,

AKR7A2, CBR1, and CBR3 (Ohara et al., 1995; O’Connor et al., 1999; Martin et al.,

2006; Kassner et al., 2008; Bains et al., 2008; Bains et al., 2009; Bains et al., 2010a;

Bains et al., 2010b). To begin, at the 0 hr incubation time period, the cytosolic

expression levels of these enzymes in the tolerant cell lines were significantly greater in

abundance compared to the sensitive group. Next, DAUN and DOX were found to

induce the expression of the enzymes (focus on AKR1C3 and CBR1) by similar

increments in all the cell lines treated for 6, 24, and 48 hrs. The induction of the

reductases in response to DAUN/DOX exposure at 6, 24, and 48 hrs correlated with the

increases in enzyme activity seen for all the cell lines. Overall, there was a negative

105 relationship between DAUN/DOX-induced cytotoxicity and AKR/CBR-mediated metabolism of DAUN/DOX to DAUNol/DOXol.

4.2 Methods

4.2.1 Chemicals and enzymes

Daunorubicin, doxorubicin, dimethyl sulfoxide (DMSO), 100X antibiotic/antimycotic, potassium phosphate (KH 2PO 4), 3-(4,5-dimethythiazol-2-yl)-2,5-

diphenyl tetrazolium bromide (MTT), N,N,N',N'-tetramethylethylenediamine, sodium

® chloride (NaCl), sodium monobasic phosphate (NaH 2PO 4), Tween 20, and β-

Nicotinamide adenine dinucleotide 2 ′-phosphate reduced tetrasodium salt (NADPH) were provided by Sigma-Aldrich (St. Louis, MO). Doxorubicinol and daunorubicinol were obtained from SynFine Research (Richmond Hill, Ontario). High performance liquid chromatography (HPLC)-grade acetonitrile, ammonium persulfate, tetramethylethylenediamine, formic acid, glycine, glycerol, and Tris were purchased from

Thermo Fisher Scientific (Waltham, MA). Phosphate-buffered saline (PBS, pH 7.4) was ordered from Invitrogen TM Corporation (Carlsbad, CA). Fetal bovine serum albumin was supplied by the American Type Culture Collection (Manassas, VA).

4.2.2 Cell culture

OVCAR-4 (ovarian), NCI-H460 (lung), and PANC-1 (pancreas) human carcinoma cell lines were kindly provided Dr. Thomas Hamilton (Fox Chase Cancer

Center, Jenkintown, PA), Dr. Urs Häfeli’s laboratory (Faculty of Pharmaceutical

Sciences, University of British Columbia, Vancouver, British Columbia), and Dr. Sylvia

Ng’s laboratory (Faculty of Pharmaceutical Sciences, University of British Columbia),

106 respectively. The H9c2 rat heart embryonic cell line, as well as HepG2 (liver), A-498

(kidney), MCF-7 (breast), PC-3 (prostate), and HCT-15 (colon) human carcinoma cell lines were obtained from the American Type Culture Collection. HepG2, A-498 and

MCF-7 cells were cultured in Eagle's minimum essential medium. H9c2 and PANC-1 cells were maintained in Dulbecco’s modified Eagle’s medium, while the PC-3 cells were cultured with F-12K medium. OVCAR-4, NCI-H460, and HCT-15 cells were maintained in RPMI-1640 medium. All cells were supplemented with 10% fetal bovine serum and 1X antibiotic/antimycotic. Cells were grown on BD Falcon™ cell culture dishes (100 mm x 20 mm; BD Biosciences, Mississauga, Ontario) and maintained in a humidified incubator with 5% CO 2 at 37°C.

4.2.3 MTT cell viability assay to measure cytotoxicity

Cells were seeded at densities of approximately 50-70% in Nunclon TM 48-well cell

culture plates (Thermo Fisher Scientfic) and were allowed to attach overnight in a

humidified incubator with 5% CO 2 at 37°C. The next day, cells were treated with

varying concentrations of DAUN or DOX (0, 0.01, 0.025, 0.05, 0.1, 0.25, 0.5, 1, 2, 5, 10,

25, 50, 100, and 150 µM) for specified time periods (6, 24, and 48 hrs post-incubation).

After treatment for each of the time periods, MTT solution [3-(4,5-dimethythiazol-2-yl)-

2,5-diphenyl tetrazolium bromide; 5 mg/mL in PBS, pH 7.4] was added (62.5 µl/well),

and the plates were incubated for another 4 hrs at 37 oC. This assay measures the ability of live cells to reduce MTT to an insoluble, colored (dark purple) formazan product by mitochondrial succinate dehydrogenase. Following the 4 hr incubation period, cells were solubilized by DMSO and absorbance values of the formazan product were measured at

TM 570 nm (and 37°C) using a monochromator-based Synergy M x multi-mode microplate

107 reader with Ultra Fine-Tuned TM performance. Absorbance values were recorded using the Gen5 TM version 1.08.4 software program (BioTek Instruments, Inc., Winooski, VT).

Cells without drug exposure were used as controls. For each cell line exposed to DAUN or DOX at each time interval, MTT assays were performed in three independent experiments and repeated in triplicate. The percentage of live cells was calculated relative to the control wells using the following equation: [(Absorbance value of experimental well)/(Absorbance value of control well)] x 100. These percentage values were used to construct lethal concentration (LC 50 ) dose-response curves with the GraphPad Prism version 4.0 program (GraphPad Software Inc., San Diego, CA). LC 50 values for each cell line were determined following DAUN and DOX exposure at the aforementioned time intervals. MTT assays were repeated again to determine LC 50 values in all cell lines for

the 48 hr exposure period with the major metabolites, DAUNol and DOXol (0, 0.5, 1, 5,

10, 25, 50, 100, 250, 500, 1000, 1500, 2000, and 3000 µM) .

4.2.4 DAUN and DOX m etabolic assays with cytosolic fractions derived from cell lines

Cytosolic fractions derived from the rat heart and human carcinoma cell lines were extracted according to the protocol provided for mammalian cells by QIAGEN Inc.

(Mississauga, Ontario). Cells were washed in PBS (pH 7.4), collected for 5 min at

1000x g, resuspended in lysis buffer (50 mM NaH 2PO 4, 300 mM NaCl, pH 8.0) with

0.05% Tween ® 20, and lysed by sonication. The lysate was centrifuged at 10,000x g for

10 min at 4 oC to pellet cellular debris and DNA. The supernatant was stored in 20% glycerol at -20 oC and saved for metabolic assays.

108 The enzymatic conversion of DAUN and DOX to their respective major metabolites, DAUNol and DOXol, were measured in cell cytosols. Assays were performed similar to the Kassner et al. (2008) study: 20 µg total cytosolic protein, 10 µM

o DAUN or DOX substrate, 2 mM NADPH, 100 mM KH 2PO 4, pH 7.4 at 37 C, 30 min.

Cytosolic protein concentrations were determined using the Bradford protein assay with

bovine serum albumin as a standard. The enzymatic reaction was stopped by adding 300

µl of ice-cold acetonitrile, which contained idarubicin as an internal standard, followed

by vortex mixing and centrifuging at 10,000xg for 10 min at 4°C to remove protein. The

supernatant was removed for detection and quantification of the DAUNol and DOXol

metabolites by an ultra performance liquid chromatography-tandem mass spectrometry

(UPLC-MS/MS) method.

4.2.5 Instrumentation and experimental conditions for UPLC-MS/MS

The UPLC/MS/MS system consisted of a Waters Acquity ultra performance

liquid chromatograph connected to a Waters Quattro Premier XE triple quadrupole mass

spectrometer equipped with an electrospray ionization source (Waters ® Corporation,

Milford, MA). Chromatographic separation of DOX, DOXol, DAUN, DAUNol and idarubicin (internal standard) was achieved using a Waters Acquity UPLC TM bridged

ethyl hydrid (BEH) C 18 column (100 mm x 2.1 mm i.d., 1.7 µm particle size) with the

following mobile phase composition and gradient programming: water with 0.1% formic

acid (A), acetonitrile with 0.1% formic acid (B); 0 min, 5% B; 1 min, 5% B; 2 min, 98%

B; 4 min 98% B; 4.1 min, 5% B; and 6 min, 5% B. The flow rate was maintained at 0.2

ml/min with a total run time of 6 mins. The mobile phase flow was diverted to the waste

before 2.5 mins and after 3.5 mins during the chromatographic run. The mass

109 spectrometer was operated in positive electrospray ionization mode with a capillary voltage of 3.5 kV, cone voltage of 40 V, and desolvation temperature of 300 ºC. The analytes were detected in multiple reaction monitoring (MRM) mode using the total ion current (TIC) of the following ion transitions ( m/z ), and collision energy (CE, eV) values: for DOX—m/z 544.1 361.2 (CE 20), m/z 544.1 379.0 (CE 20), and m/z 544.1

397.2 (CE 10); for DOXol—m/z 546.0 345.0 (CE 35), m/z 546.0 363.1 (CE 20), and m/z 546.0 398.9 (CE 10); for DAUN—m/z 527.9 321.0 (CE 20), m/z 527.9

363.2 (CE 20), and m/z 527.9 381.0 (CE 10), for DAUNol—m/z 530.1 306.2 (CE

35), m/z 530.1 321.2 (CE 30), and m/z 530.1 383.1 (CE 10); and for IDA—m/z

497.9 291.2 (CE 25), m/z 497.9 333.0 (CE 25). Data were acquired and processed

using MassLynx TM version 4.1 software (Waters ® Corporation). A representative UPLC-

MS/MS chromatogram of DOX, DOXol, DAUN, DAUNol, and IDA is presented in

Figure 28.

110

Figure 28—Representative UPLC-MS/MS chromatogram of doxorubicin, doxorubicinol, daunorubicin, daunorubicinol, and idarubicin. These analytes were monitored using the total ion current (TIC) of three multiple reaction monitoring (MRM) signals for doxorubicin, doxorubicinol, daunorubicin, daunoribicinol, and two MRM signals for idarubicin. The retention time (in minutes) for each of these analytes are labelled on top of the peaks. The peaks in the chromatogram appear broad since this is a close up a 1 min time interval (from 2.5 to 3.5 mins).

4.2.6 Western blotting of cytosols to detect AKRs and CBRs

Western blot analyses of the cytosols extracted from each cell line were conducted according to the procedure described by Odyssey TM (LI-COR Biosciences,

Lincoln, NE). Following 18% SDS-polyacrylamide gel electrophoresis, cytosolic

proteins were transferred at 20 V in Towbin's buffer (25 mM Tris, 192 mM glycine, and

20% v/v methanol) overnight (at 4 oC) to a Hybond-C Extra nitrocellulose membrane (GE

Healthcare, Piscataway, NJ). The membranes were blocked in Odyssey blocking buffer,

and the enzyme was detected using either a MaxPab polyclonal mouse anti-human

111 AKR1A1, AKR1B1, AKR1B10, AKR1C1, AKR1C2, AKR1C3, AKR1C4, AKR7A2,

CBR1 or CBR3 antibody (Abnova ® Corporation, Taipei City, Taiwan) (diluted 1:2500)

as the primary antibodies and IRDye 800CW goat anti-mouse IgG as the secondary

antibody (diluted 1:5000) (LI-COR). Both primary and secondary antibodies were in

blocking buffer containing 0.1% Tween 20. The bound secondary antibody was detected

using the Odyssey TM Infrared Imaging system (LI-COR). β-tubulin served as a loading

control and was detected using the same aforementioned Western protocol, except that a

mouse monoclonal anti- β-tubulin primary antibody (Abcam ® Inc., Cambridge, MA) was used (diluted 1:1000). Band intensities for AKR or CBR proteins were determined using

Odyssey TM densitometry software, and were normalized to band intensities of β-tubulin in order to quantify the relative expression of the individual reductases in each cytosolic extract.

4.2.7 Statistical analysis

Statistical analyses were performed by GraphPad Instat ® (version 3.6; GraphPad

Software Inc., San Diego, CA). Percent viabilities to create dose-response curves from

the cytotoxicity assays were expressed as means ± S.D while LC 50 values were provided

along with the 95% confidence intervals. Results from the expression and metabolic

assays were reported as means ± S.D. and compared using one-way analysis of variance

followed by Tukey-Kramer multiple comparisons tests. Differences were considered

significant at p<0.05. Sample sizes ( n) for the experiments performed in this study are as

follows: determination of LC 50 values with the 9 cell lines in presence of either DAUN,

DOX, DAUNol or DOXol ( n=3 at each exposure time period), determination of

DAUNol/DOXol levels after performing metabolic assays with the cell line cytosolic

112 extracts ( n=12), and determination of the relative expression ratios of the reductase

enzymes in the individual cell lines after Western blotting and densitometry analyses

(n=3).

4.3 Results

4.3.1 Cytotoxicity of cells following exposure to anthracyclines

The cytotoxicities of the anticancer drugs DAUN, DOX, and their metabolites

(DAUNol and DOXol) were determined in the cell lines using the MTT assay. Cell viability was measured after incubating each cell line with various concentrations of the anthracycline parent drug or its major alcohol metabolite for 0 (100% viable), 6, 24, or 48 hrs. Dose-response curves were plotted for all 9 cell lines; two representative plots are presented in Figure 29 ( n=3). The LC 50 values were calculated from these curves, and

the LC 50 values for each of the cell lines were compared to see which compounds were more toxic. Cell lines with lower LC 50 values are more sensitive to the drug, while cell

lines with higher LC 50 values are less sensitive (more resistant). The LC 50 values for all

nine cell lines are provided in Tables 12 (for DAUN and DAUNol) and 13 (for DOX and

DOXol). Four observations are worth mentioning from these data sets. First, four cell

lines, HepG2 (liver), HCT-15 (colon), NCI-H460 (lung), and A-498 (kidney), were found

to be significantly more resistant to DAUN and DOX compared to the five remaining cell

lines , H9c2 (heart), PC-3 (prostate), OVCAR-4 (ovary), PANC-1 (pancreas), and MCF-7

(breast), which I have called the sensitive group of cell lines. This finding was consistent

for all time points examined. Second, DOX was more toxic than DAUN and, again, this

was consistent for all of the cell lines. Third, LC 50 values for DAUNol and DOXol were

significantly higher (i.e., less toxic) than the LC50 values of their respective parent drug

113 and this was consistently observed for all cell lines tested. Indeed, the DAUNol LC 50 values were 30- to 108-fold greater than those for DAUN; the DOXol LC 50 values were

59- to 162-fold greater than those for DOX. Taken together, these findings strongly suggest that the parent drugs, DAUN or DOX, are far more toxic to these cells than are their metabolites, DAUNol and DOXol, respectively.

Figure 29—Sample dose-response curves of (A) H9c2 rat heart cell line and (B) HepG2 liver cell line in the presence of varying concentrations of daunorubicin (DAUN), doxorubicin (DOX), daunorubicinol (DAUNol), and doxorubicinol (DOXol) following a 48 hr incubation. Three independent batches of cells for each cell line were grown and subjected to MTT viability assays. The % viability values are reported as the mean ± S.D. ( n=3).

114 Table 12—Mean LC 50 values of DAUN and DAUNol for cell lines at specified time intervals following exposure to the drug.

MEAN LC (µM) a 50 (95% C.I.) DAUN DAUN DAUN DAUNol CELL LINES 6 hrs 24 hrs 48 hrs 48 hrs 12.6 3.2 0.4 42.7 H9c2 (heart) (11.2-14.1) (3.0-3.5) (0.3-0.5) (37.6-48.5) 10.2 3.2 0.4 43.2 PC-3 (prostate) (9.2-11.4) (2.9-3.6) (0.3-0.5) (38.1-49.1) 11.1 3.1 0.7 41.6 OVCAR-4 (ovary) (9.8-12.5) (2.8-3.4) (0.6-0.8) (35.5-48.9) PANC-1 11.4 3.6 0.6 45.0 (pancreas) (10.1-13.0) (3.2-4.0) (0.5-0.7) (39.5-51.4) 14.1 4.2 0.6 48.7 MCF-7 (breast) (11.9-16.6) (3.8-4.6) (0.5-0.7) (42.6-55.7) 21.5 9.5 1.5 54.5 Hep G2 (liver) (18.4-25.1) (8.3-10.9) (1.3-1.9) (47.8-62.0) 19.7 9.9 1.4 54.0 HCT-15 (colon) (17.0-22.7) (8.7-11.3) (1.2-1.7) (47.7-61.1) 21.0 10.3 1.7 55.0 NCI-H460 (lung) (18.1-24.3) (9.3-11.5) (1.4-2.0) (48.8-62.0) 22.9 13.7 1.7 51.7 A-498 (kidney) (20.5-25.6) (12.4-15.0) (1.5-1.9) (45.3-59.2) a LC50 values were calculated from dose-response curves generated from MTT viability experiments. 95% confidence intervals (C.I.s) are provided in brackets below the mean LC50 values. The dark grey shaded section of the table refers to the sensitive cell lines while the unshaded section refers to the resistant cell lines.

115 Table 13—Mean LC 50 values of DOX and DOXol for cell lines at specified time intervals following exposure to the drug.

MEAN LC (µM) a 50 (95% C.I.) DOX DOX DOX DOXol CELL LINES 6 hrs 24 hrs 48 hrs 48 hrs 5.1 1.1 0.2 25.8 H9c2 (heart) (4.4-5.8) (1.0-1.3) (0.1-0.3) (22.5-29.5) 5.1 0.9 0.2 27.0 PC-3 (prostate) (4.6-5.7) (0.8-1.1) (0.1-0.3) (23.2-31.5) 6.6 1.1 0.2 27.2 OVCAR-4 (ovary) (5.9-7.5) (0.9-1.2) (0.1-0.3) (24.0-30.7) 7.4 1.3 0.2 29.6 PANC-1 (pancreas) (6.3-8.6) (1.1-1.5) (0.1-0.3) (25.6-34.1) 7.3 1.3 0.2 32.4 MCF-7 (breast) (6.4-8.3) (1.1-1.5) (0.1-0.3) (28.2-37.1) 11.9 3.4 0.5 37.2 Hep G2 (liver) (10.1-13.8) (3.0-3.9) (0.4-0.6) (30.2-45.6) 10.1 2.4 0.5 38.8 HCT-15 (colon) (9.1-11.1) (2.2-2.6) (0.4-0.6) (34.8-43.3) 10.9 3.0 0.6 35.5 NCI-H460 (lung) (9.8-12.0) (2.6-3.4) (0.5-0.7) (30.7-41.0) 12.2 4.1 0.5 38.9 A-498 (kidney) (10.6-14.0) (3.7-4.6) (0.4-0.6) (34.4-44.0) a LC50 values were calculated from dose-response curves generated from MTT viability experiments. 95% confidence intervals (C.I.s) are provided in brackets below the mean LC50 values. The dark grey shaded section of the table refers to the sensitive cell lines while the unshaded section refers to the resistant cell lines.

4.3.2 Metabolism of the anthracyclines using cell line cytosols

Since the liver, colon, lung, and kidney derived cell lines seemed to be considerably less sensitive to DAUN or DOX than the remaining five cell lines, we wished to ascertain the basis for this difference. More specifically, we wanted to determine whether this difference was associated with differences among the cell types in the rate at which they were able to metabolize the drugs or were based on some other physiological property, perhaps transport into or out of the cell, as an example. Hence,

116 we measured the rate of enzymatic conversion of DAUN and DOX to their respective major metabolites, DAUNol and DOXol, in cytosolic extracts from each of the nine cell lines. Among the five cell lines that were more sensitive to the drugs (heart, prostate, ovary, pancreas, and breast), the rates of conversion of DAUN to DAUNol ranged from

17.6±1.0 to 27.3±1.9 pmol DAUNol/min•mg cytosolic protein and the rate of conversion of DOX to DOXol ranged from 3.6±1.0 to 5.0±0.8 pmol DOXol/min•mg cytosolic protein (Figure 30) (all n=12). The four tolerant (less sensitive) cell lines (liver, colon,

lung, and kidney) had significantly higher metabolic rates ranging from 50.0±3.2 to

80.7±9.7 pmol DAUNol/min•mg and 9.3±0.8 to 10.5±1.1 pmol DOXol/min•mg (Figure

30) (all n=12). As a result, the five sensitive cell lines were classified as the low

metabolizers of these anthracycline drugs and the four tolerant cell lines as the high

metabolizers. Furthermore, all cell lines were shown to metabolize DAUN more rapidly

(a 3.3- to 7.8-fold higher rate) than DOX.

Figure 30—Enzymatic activities for rat and human cell line cytosols incubated with 10 µM of daunorubicin (A) and doxorubicin (B). Activities were measured by following the rate of daunorubicinol and doxorubicinol production. Three independent batches of cells for each cell line were grown and cytosols extracted. Assays were performed in quadruplicate with each batch. Enzymatic activities are reported as mean ± S.D. ( n=12). Cell lines to the left of the vertical dashed line are sensitive to DAUN and DOX while the cell lines to the right are the resistant ones.

117

4.3.3 Expression of cytosolic AKRs and CBRs in cell lines

Next, we wanted to determine whether the difference in metabolic activity between the tolerant and sensitive cell lines resulted from a difference in the relative abundance of specific AKR and/or CBR enzymes, or was due to other physiological factors that might differentiate the cell types. Western blot analysis for each of eight

AKR (AKR1A1, 1B1, 1B10, 1C1, 1C2, 1C3, 1C4, and 7A2) and two CBR enzymes

(CBR1 and CBR3) was performed on cytosolic extracts from each of the nine cell lines.

Distinct bands with mobilities corresponding to the calculated molecular mass for each of the enzymes, 35-38 kDa for the AKRs and 30 kDa for the CBRs, were detected in each extract (Figure 31). Expression levels of the reductases for each cell line were ascertained by densitometry analysis and normalized against the expression levels of β-

tubulin (as an internal standard) in order to obtain the relative abundance of each enzyme

in each cell type. Comparing these values between cell lines, each of the cell types in the

tolerant group (liver, colon, lung, and kidney) express AKRs and CBRs at significantly

higher levels than any of the five cell types that comprise the sensitive group (Figure 32)

(all n=3). Therefore, it appears that the higher expression of the reductase genes

contributes to the ability of the tolerant cell lines to metabolize DAUN and DOX more

efficiently than the sensitive cell lines.

118

Figure 31—Western blot detection of the sensitive cell line cytosols (lane 1: heart, lane 2: prostate, lane 3: ovary, lane 4: pancreas, and lane 5: breast; all at 15 µg total protein) as well as the resistant cell line cytosols (lane 6: liver, lane 7: colon, lane 8: lung, lane 9: kidney; all at 15 µg total protein) confirms expression of the desired AKR and CBR proteins. β-tubulin (~55 kDa) was used as an internal loading control. 6x-His tagged AKR and CBR recombinant proteins were used as positive controls for antibody immunoreactivity (~1 g). Cell lines to the left of the vertical dashed line are sensitive to DAUN and DOX while the cell lines to the right are the resistant ones.

119

Figure 32—Relative expression ratio levels of AKR and CBR proteins in the sensitive (heart, prostate, ovary, pancreas, and breast; all at 15 µg total protein) as well as the resistant cell line cytosols (liver,colon, lung, and kidney; all at 15 µg total protein). Densitometry was used to determine expression values of reductases in each cell line following Western blotting. These individual reductase expression values were then divided by the expression values of β-tubulin with the purpose of calculating relative expression ratios. Three independent batches of cells for each cell line were grown and subjected to Western blotting and densitometry analyses. Relative expression values are reported as mean ± S.D. ( n=3). Cell lines to the left of the vertical dashed line are sensitive to DAUN and DOX while the cell lines to the right are the resistant ones.

120

4.3.4 Increase in DAUN/DOX metabolism and induction of AKR and CBR gene expression are responses in cell lines following exposure to anthracycline drugs

If there is a connection between the AKR and/or CBR enzymes and detoxification or protection against these drugs, then either the expression of the AKR and/or CBR may be induced in response to exposure to either drug. Therefore, it was necessary to ask whether exposure to either DAUN or DOX elicited any response in one or more of the nine cell lines. If a response was elicited by the cell lines following anthracyline exposure, then the next step was to determine if the resistant cell lines (liver, colon, lung, and kidney) respond differently than the sensitive cell lines (heart, prostate, ovary, pancreas, and breast). Finally, it was important to see if any increase from the metabolism of the drugs was correlated with gene expression rather than some post- translational alteration in enzyme activity or some other physiological event. Hence, the relative abundance of each enzyme in the cell lines was measured following exposure to the drug at each of the aforementioned time periods.

To begin, all the cell lines were exposed to 100 nM of either DAUN or DOX for

6, 24, or 48 hrs. This drug concentration was selected since cell viability was 75% or greater based on the dose-response curves generated from the cytotoxicity study; therefore, the cells would have a chance to respond to the drug, if they could. At each time interval, the cells were washed three times with PBS, immediately thereafter cytosolic extracts were made from each cell line, and their ability to metabolize DAUN or DOX was tested. The metabolic assays were conducted for 30 min as described above; 10 µM of DAUN was used as the substrate for the DAUN-treated cell lines and 10

µM DOX was used as the substrate for the DOX-treated cell lines. Metabolic activity

121 rates of all the treated cell lines were measured by the production of the major metabolites DAUNol and DOXol, and compared to rates of their respective non-treated

(0 hr) cell lines. The results are shown in Figure 33 (all n=12); three observations are

noteworthy. First, while there was no significant difference in metabolic activity between

the non-treated and 6 hr DAUN- or DOX-treated cell lines, there was a clear upward

trend in the measured activity of nearly all cell lines. The single exception, among the 18

experiments, is the lung cells exposed to DAUN where there was no detectable change

after 6 hrs. Hence, while the increase in metabolic activity after a 6 hr exposure is not

statistically significant, the trend is clear. Second, after a 24 hr exposure, all 9 cell lines

showed a statistically significant increase in their ability to metabolize DAUN or DOX,

and after exposure to the drug for 48 hrs, all 9 cell lines showed an even greater increase

in their ability to metabolize the drug. In the case of those cells exposed to DAUN for 24

and 48 hrs, there was a 1.4- to 2.5-fold respective increase in DAUNol production versus

the untreated cells. Likewise, the metabolic activity of cells exposed to DOX for 24 and

48 hrs was 1.6- to 2.6-fold higher than their untreated counterparts. Lastly, the four cell

lines defined previously as high metabolizers (liver, colon, lung, and kidney) did not

differ from the low metabolizing group (heart, prostate, ovary, pancreas, and breast) in

their ability to respond to exposure to either drug, that is, the induction was similar in all

cell lines.

122

Figure 33—The metabolic activity rates (measured as pmol DAUNol or DOXol/minute•mg total protein) of cytosols that were extracted from cell lines treated with 100 nM DAUN (A) and DOX (B) after 0, 6, 24, and 48 hr exposure time periods. Significant increases (*, p<0.05) in activity rates are seen for the cytosols treated for 24 and 48 hrs. Metabolic activity rates on the left side of the dashed vertical line represent the sensitive cell lines while activities to the right are the resistant cell lines. Three independent batches of cells for each cell line were grown and cytosols extracted. Assays were performed in quadruplicate with each batch. Enzymatic activities are reported as mean ± S.D. ( n=12).

123 Although observable at 6 hrs, the increase in enzyme activity of all nine cell lines was not statistically significant compared to the untreated control until the 24 hr exposure period; this data suggests that the induction by either DAUN or DOX was not immediate, but rather it took some time. This delay may be consistent with the time needed for induction of transcription and subsequent translation processes. Hence, it is reasonable to predict that the increase in metabolic activity in response to exposure to DAUN or DOX was linked to an increase in the amount of the various AKR and/or CBR enzymes produced in each cell line. To test this, Western blotting was used to examine the relative abundance (i.e., expression) of two enzymes (AKR1C3 and CBR1) in each of the 9 cell lines exposed to DAUN and, separately, DOX for either 0, 6, 24, or 48 hrs. Rather than perform Western blot analyses for the cytosolic expression of all ten of the aforementioned enzymes in each of the 9 different cell lines, the focus turned towards the

AKR1C3 and CBR1 enzymes because the in vitro enzymatic assays from Chapters 2 and

3 have shown that these two enzymes have the highest catalytic efficiency ( kcat /Km) towards DOX and DAUN, respectively, compared to the eight remaining AKR and CBR enzymes. Western blots of AKR1C3 and CBR1 are shown in Figure 34A. The quantitative analyses of these Western blots are illustrated for AKR1C3 and CBR1 enzymes from DAUN- or DOX-treated cells in Figures 34B to 34E (all n=3). These

analyses demonstrated that the relative abundance of each of these AKR and CBR

enzymes increased over time in all 9 cell lines exposed to either these anthracyclines.

The increase was detectable after a 6 hr exposure, although not statistically significant

compared to the non-treated cells (0 hr exposure), while the expression levels continued

to increase with statistically significant differences detected after 24 and 48 hrs of

exposure. For AKR1C3, significant increases in expression were detected in both the 24

124 and 48 hr DAUN-treated (1.2 to 2.3-fold increase compared to non-treated cell lines) and

DOX-treated (1.2 to 2.7-fold increase) cell lines. A similar trend was found for CBR1 expression for the cells exposed for 24 and 48 hr to either DAUN (1.3 to 2.8-fold increase) or DOX (1.5 to 2.4-fold increase). Therefore, the induction in AKR and CBR gene expression, which leads to an increase in abundance of these enzymes, is a contributing factor to the increase in DAUN and DOX metabolism seen with the treated cell lines over their respective untreated counterparts. Previous studies have demonstrated signficiant induction of AKRs and CBRs in the presence of DOX. For example, exposure of MCF-7 breast tumor cells to low (nanomolar) concentrations of

DOX resulted in a significant increase in CBR1 protein levels (Gavelova et al., 2008), as well as AKR1C2 and AKR1C3 (Veitch et al., 2009). In relation to DAUN, the current study was the first to demonstrate that exposure to this anthracycline can lead to significant induction of CBR1 and AKR1C3 enzymes.

125

Figure 34—Sample Western blot analyses (A) of cytosols from H9c2 rat heart and HCT-15 human colon carcinoma cell lines for purposes of assessing induction of AKR1C3 and CBR1 after 0, 6, 24, and 48 hr exposure to either 100 nM DAUN or DOX. Densitometry was performed on the immunoreactive bands and normalized against β-tubulin (loading control) in order to calculate relative expression ratios. With the treated cell lines, the relative expression ratios for AKR1C3 (B: DAUN-treated; C: DOX-treated) and CBR1 (D: DAUN-treated; E: DOX-treated) indicate significant induction (*, p<0.05) in cytosols treated for 24 and 48 hrs. Expression levels on the left side of the dashed vertical line represent the sensitive cell lines while expression levels to the right are the resistant cell lines. Three independent batches of cells for each cell line were grown and subjected to Western blotting and densitometry analyses. Relative expression values are reported as mean ± S.D. ( n=3).

126

4.3.5 Association between cytotoxicity and DOX/DAUN metabolism in cell lines exposed to these anthracyclines

For each of the incubation time periods, the LC 50 values (DAUN and DOX) from the cytotoxicity studies as well as metabolic activity rates from the DAUN and DOX- treated cell lines were plotted against each other to see if an association exists between these two variables (Figure 35). The high values obtained for correlation coefficients ( r) and coefficients of determination ( r2) suggest that cytotoxicity and metabolism are

strongly related to each other with these two variables being negatively correlated.

Figure 35—Scatterplot of cytotoxicity (LC 50 values) and metabolic activity following DAUN (A) and DOX (B) treatment in the nine cell lines for the purposes of seeing if there is an association between these two variables for 6, 24, and 48 hr incubations. Low LC 50 values are synonymous with higher cytotoxicity while the reverse is true for high LC 50 values. The high correlation coefficient ( r) and coefficient of determination ( r2) values provided for each of the exposure time periods suggest a strong correlation between cytotoxicity and metabolism.

4.3.6 UPLC-MS/MS method validation

The UPLC/MS/MS method was validated for accuracy, precision, linearity, range, limit of quantitation (LOQ), and selectivity. The validation was performed in spiked samples of a pooled blank cytosol matrix, which was prepared from nine separate cancer cell lines. Accuracy was expressed as the percentage deviation (% Deviation) between

127 the measured and nominal concentrations of six replicate samples separately spiked at three concentration levels (QC-Low, QC- Mid, and QC-High). Precision was expressed as the coefficient of variation (% CV) of the measured concentrations of six replicate samples obtained from a single spiked sample of the QC-Low, QC- Mid, and QC-High samples. Six replicate samples were analyzed at the three QC levels. For DOXol and

DAUNol, the results for accuracy and precision are presented in Table 14. For DOXol, intra-day and inter-day accuracy ranged from -4.90% to 10.2% and 3.21% to 9.53%, respectively, while intra-day and inter-day precision ranged from 1.77% to 10.5% and

3.43% to 10.4%, respectively. For DAUNol, intra-day and inter-day accuracy ranged from -6.84% to 11.4% and 3.67% to 8.90%, respectively, while intra-day and inter-day precision ranged from 1.33% to 10.9% and 3.43% to 9.31%, repectively. Accuracy and precision met the suitability criteria of less than 15% CV for calculated concentrations and less than 15% bias from expected concentration with six replicate samples prepared at three QC levels within the range of the method. With respect to linearity, the calibration curves for the metabolites were weighted (1/X 2, n=3) and were linear with coefficient of determination ( r2) values for DOXol ≥ 0.994 and for DAUNol ≥ 0.988.

The range of the method was 2.50 to 500 ng/ml for DOXol and DAUNol requiring 150 µl of sample. The LOQ for DOXol was 2.50 ng/ml (i.e., 4.58 nM) with an accuracy of -

0.552%±3.63% (mean±SD, n=6), and precision of 3.65% ( n=6). The LOQ for DAUNol was 2.50 ng/ml (4.42 nM) with the accuracy of 0.885%±12.3% (mean±SD, n=6), and

precision 12.2% ( n=6). The method was selective with no interference at the retention

times of DOXol and DAUNol when the blank matrix pooled from the nine cell line

cytosols were spiked at LOQ level ( n=6). The criteria used to assess the suitability of the

128 method were based on the best practices developed for bioanalytical method validation that fulfill the current regulatory guidance (Shah et al., 2000; Nowatzke and Woolf, 2007).

129 Table 14—Results for accuracy and precision (intra-day and inter-day) for the UPLC-MS/MS determinations of DOXol and DAUNol in pooled matrix of nine cell lines (days 1 to 3). Acceptance criteria for the method was set at less than 15% CV (precision) and less than 15% bias at all QC levels. Each determination was based on six replicate samples.

DOXol DAUNol Intra-day Inter-Day Intra-day Inter-Day Accuracy Day 1 Day 2 Day 3 Days 1-3 Accuracy Day 1 Day 2 Day 3 Days 1-3 (n=6) (n=6) (n=6) (n=18) (n=6) (n=6) (n=6) (n=18) QC-Low (7.5 ng/ml) QC-Low (7.5 ng/ml) Mean (ng/ml) 7.13 7.17 7.58 7.29 Mean (ng/ml) 6.99 7.54 7.71 7.4 SD (ng/ml) 0.992 0.343 0.627 0.695 SD (ng/ml) 0.90 0.50 0.27 0.66 % Deviation -4.90 -4.40 1.09 9.53 % Deviation -6.84 0.55 2.78 8.90 QC-Mid (80 ng/ml) QC-Mid (80 ng/ml) Mean (ng/ml) 81.2 77.9 84.8 81.3 Mean (ng/ml) 80.1 83.7 87.1 83.7 SD (ng/ml) 3.25 4.89 1.50 4.39 SD (ng/ml) 4.36 2.31 1.66 4.08 % Deviation 1.44 -2.67 5.95 5.40 % Deviation 0.149 4.65 8.94 4.88 QC-High (400 ng/ml) QC-High (400 ng/ml) Mean (ng/ml) 439 427 441 436 Mean (ng/ml) 412 432 446 430 SD (ng/ml) 10.3 14.2 15.1 14.0 SD (ng/ml) 8.86 5.76 7.34 15.8 % Deviation 9.73 6.80 10.2 3.21 % Deviation 3.04 8.08 11.4 3.67

Intra-day Inter-Day Intra-day Inter-Day Precision Day 1 Day 2 Day 3 Days 1-3 Precision Day 1 Day 2 Day 3 Days 1-3 (n=6) (n=6) (n=6) (n=18) (n=6) (n=6) (n=6) (n=18) QC-Low (7.5 ng/ml) QC-Low (7.5 ng/ml) Mean (ng/ml) 6.4 7.2 7.6 7.1 Mean (ng/ml) 6.68 7.54 7.71 7.31 SD (ng/ml) 0.67 0.34 0.63 0.73 SD (ng/ml) 0.726 0.495 0.273 0.680 % CV 10.5 4.79 8.27 10.4 % CV 10.9 6.57 3.54 9.31 QC-Mid (80 ng/ml) QC-Mid (80 ng/ml) Mean (ng/ml) 81.0 77.9 84.8 81.2 Mean (ng/ml) 79.2 83.7 87.1 83.4 SD (ng/ml) 2.97 4.89 1.50 4.33 SD (ng/ml) 2.64 2.31 1.66 3.94 % CV 3.67 6.28 1.77 5.33 % CV 3.33 2.76 1.91 4.72 QC-High (400 ng/ml) QC-High (400 ng/ml) Mean (ng/ml) 432 427 441 433 Mean (ng/ml) 415 432 446 431 SD (ng/ml) 14.4 14 15.1 15 SD (ng/ml) 9.50 5.76 7.34 14.8 % CV 3.34 3.32 3.43 3.43 % CV 2.29 1.33 1.65 3.43

130

4.4 Discussion

DAUN and DOX are among the most commonly used and most successful drugs to treat a variety of leukemias, carninomas, and lymphomas. However, a considerable number of patients treated with either of these drugs develop serious chronic cardiomyopathies. While the likelihood of developing some form of cardiotoxicity is directly linked to the cumulative dose of the drug that each patient receives, the underlying cause of the cellular toxicity remains largely unknown. One hypothesis for the inter-patient variability of these toxic effects is that the patients differ in their metabolic efficiency of removing either the parent drug or its down stream metabolites.

However, there is no conclusive evidence in the literature that links altered DAUN or

DOX metabolism to toxicity. The attractiveness of this hypothesis is weakened further by conflicting assertions of whether the cytotoxicity is caused by the parent drug (DAUN or DOX) or its primary metabolite (DAUNol or DOXol, respectively). To address these issues, we performed ex vivo and in vitro studies on nine different cell lines to determine:

(i) whether the parent drugs, DAUN or DOX, or their metabolites, DAUNol and DOXol respectively, were more toxic, (ii) the metabolic efficiency of each cell type, that is, the efficiency with which each cell type converted the parent drug to its primary metabolite,

(iii) whether the metabolic efficiency of each cell line correlated with the relative abundance of one or more of the selected AKRs and CBRs, (iv) whether one or more cell types responded to the presence of DAUN or DOX by increasing their capacity to metabolize the drug, and if so, (v) if the increase in the ability to metabolize the drug correlated with an increase in the relative abundance of the DAUN or DOX metabolizing enzymes or resulted from some other physiological event, that is yet to be determined.

131 The cytotoxicity studies revealed that the LC 50 values for the metabolites

DAUNol and DOXol were far greater (30- to 108-fold and 59- to 162-fold, respectively) than those for the parent drugs, DAUN and DOX respectively, and thus clearly demonstrated that the parent drugs were far more toxic to the cells than were their primary metabolites. This was true for all nine cell-types that were examined. Hence, it appears that the parent drugs are physiologically more detrimental to cells, regardless of their tissue of origin. While all nine cell-types responded similarly, they did not respond equally. The nine different cell lines could be divided into two different groups based on their sensitivity (LC 50 values) to either the parent drugs or their metabolites. Four cell types (liver, colon, lung, and heart) were less sensitive to DAUN and DOX; while the remaining five cell types (heart, prostate, ovary, pancreas, and breast) were more sensitive, and the differences were statistically significant. Although we could also divide the nine cell types into two groups based on their response to either DAUNol or

DOXol, these differences were not statistically significant.

Since previous studies, performed on cancer patients receiving treatment with either DAUN or DOX, demonstrated that DAUNol and DOXol are the major metabolites of these anti-cancer drugs (Lipp and Bokemeyer, 1999; Plebuch et al., 2007), we assessed the efficiency with which the various cell lines metabolized DAUN and DOX by quantifying the rate at which the carbon-13 alcohol metabolites were produced. Once again, all nine cell lines were able to metabolize either DAUN or DOX, but they differed in the rate at which they converted the parent drug to its metabolite. Four cell lines (liver, colon, lung, and heart) metabolized DAUN or DOX faster than the remaining five cell lines (heart, prostate, ovary, pancreas, and breast). The assemblage of the two groups of

132 cell-types based on metabolic efficiency parallels that found for the cytotoxicity studies.

Thus, there is an association between these two physiological phenomena, with high metabolizing cell lines exhibiting elevated LC 50 values (greater resistance) and conversely, the low metabolizing cell lines having decreased LC 50 values (greater sensitivity) for the parent drugs or their metabolites. Lastly, the relative abundance of the eight AKR and two CBR enzymes in each cell type correlated with the relative efficiency with which the cell types metabolized DAUN or DOX. The high metabolizing cell lines had considerably more of each AKR and CBR enzyme in their cytosol (often 2 to 3-fold greater) than the amount present in the cell lines comprising the low metabolizing group.

Therefore, there is a striking parallel between the efficiency with which each cell type metabolizes DAUN or DOX and the relative abundance (expression level) of the anthracycline metabolizing enzymes. Our results indicate a negative relationship between toxicity following DOX or DAUN exposure and both metabolism by, and relative abundance of, the eight AKR and two CBR enzymes. The high metabolizing cell lines are less sensitive to DAUN or DOX and produce more of each type of enzyme, while the low metabolizing enzymes are more sensitive to these drugs and produce less of each enzyme.

Some caveats should, however, be stated. First, the cell lines used are permanent cell lines derived from human tissues, and one from rat, which were established some time ago. As immortalized cells, their growth properties have, no doubt, been altered; however, we believe that these cell lines do retain much of the biochemical distinctiveness of their parent cell types. Second, although there is a striking negative relationship between the relative abundance of all eight AKR and both CBR enzymes and

133 the sensitivity of each of the cell-lines to DAUN/DOX-induced cytotoxicity, the findings do not establish a direct involvement of any of these reductases with drug sensitivity.

Nevertheless, we believe the results provide a compelling reason for additional studies to determine the basis for the correlation between AKR and CBR abundance and cytotoxicity following exposure to these anthracycline drugs. These future studies should include (i) inhibiting/reducing the metabolic activity of the individual AKRs/CBRs (i.e., via knockdown/knockout procedures or the use of specific inhibitors) in cell lines with high levels of reductase expression and DAUN/DOX metabolic activity to see if cell toxicity is increased, and (ii) over-expressing individual AKRs and CBRs (i.e., via transfection with vectors containing the reductase gene of interest) in cell lines with low levels of reductase expression/metabolic activity to see if cell toxicity is decreased. A detail description of these proposed studies is addressed in the “Suggested future research directions” section of Chapter 5.

Overall, this study demonstrated a negative association between metabolic activity by AKRs/CBRs and cytotoxicity using nine cell lines. If cytotoxicity observed in the cell lines, which appears to be correlated with the level of DAUN/DOX metabolism, is mechanistically similar to the toxicity observed following anthracycline therapy in cancer patients, then one would expect that the relative toxicities observed among the nine cell lines would parallel the relative sensitivity of normal cells from which these cell lines were derived. This may be true in relation to cardiotoxicity, since the AKRs and

CBRs are found in human heart tissue, but are in lower abundance than in the liver

(Figure 36) (all n=3). A study by O’Connor et al. (1999) also demonstrated that

AKR7A2 and AKR1B1 are expressed in the human heart using Western blot analyses,

134 however, the expression of these two enzymes were not quantified. Given the correlation between the relative abundance of these enzymes, the relative efficiency with which each cell type metabolizes DAUN or DOX, and their relative sensitivity to the cytotoxic effects of these drugs, one can predict that cancer patients with mutant alleles of the AKR and CBR genes that encode these enzymes are more likely to develop adverse effects

(i.e., cardiotoxicity) following DAUN or DOX therapy. Indeed, if metabolism of DAUN and DOX are associated with sensitivity to the cytotoxic effects of these drugs then, those individuals with the more severe mutant alleles should be at the greatest risk, while those with mutations that have rather minor affects on enzyme function should be a far lower risk. The next step is to perform clinical studies to see if there are strong associations between frequency of one or more of these reductase variants and the susceptibility of

DAUN/DOX-treated cancer patients to the development of serious life-threatening adverse effects such as cardiotoxicity. This would help determine if these variants are clinically relevant genetic biomarkers for guiding anthracycline therapy. We have initiated these clinical studies.

135

Figure 36—Bar graph illustrating the relative expression ratios of the cytosolic AKRs and CBRs in human heart lysate (H; single donor, a 38 year old female; ProSci Incorporated, Poway, CA) and pooled human liver lysate (L; 50 donors; BD Biosciences, San Jose, CA). Densitometry was used to determine expression values of reductases in each cell line following Western blotting. These individual reductase expression values were then divided by the expression values of β-tubulin with the purpose of calculating relative expression ratios. Western blotting was repeated three times to determine relative expression values of the reductases, which are reported as mean ± S.D. ( n=3). (Below the bar graphs ) Human heart (H) and liver (L) lysates (~20 g total protein) were run on 18% SDS-PAGE gels and subjected to Western blotting for detection of AKR1A1, AKR1B1, AKR1B10, AKR1C1, AKR1C2, AKR1C3, AKR1C4, AKR7A2, CBR1, and CBR3 proteins (ranging from 30 to 38 kDa). In addition to the reductase enzymes, β-tubulin band was also detected using Western blotting (~55 kDa). The 6x-His tagged purified proteins were used as positive controls (PCs) for antibody immunoreactivity for these enzymes. M refers to the molecular weight markers.

136 CHAPTER 5: OVERALL SUMMARY AND CONCLUSIONS

5.1 Novel findings

There are two novel findings that I observed after analyzing the results from the in vitro and ex vivo experiments performed in this thesis. The first of these findings is that:

1. A total of 12 variant alleles (7 for the AKRs and 5 for the CBRs) were associated

with significantly reduced in vitro metabolic activities of DAUN and/or DOX

compared to their respective wild-type counterparts (Chapters 2 and 3).

To date, this thesis offers the most comprehensive set of data on the impact that naturally-occurring allelic variants of human AKRs and CBRs, in the form of ns-SNPs, have on the functional activities of these reductase enzymes. Out of the 38 variants that were examined in the in vitro studies, 12 resulted in significantly decreased enzymatic activities in the metabolic conversion of these drugs to their corresponding major alcohol metabolites, DAUNol and DOXol. This number can be increased to 13 if one includes the truncated enzyme (E36Term) of AKR1C3. As stated in Chapter 2, this variant cannot metabolize either DAUN or DOX, since it is missing critical residues that are involved in catalysis and cofactor binding. As stated previously, the in vitro studies involved a series of experiments comprised of cloning the wild-type AKR and CBR genes into pET vectors, creating the variants by site-directed mutagenesis, transforming the wild-type and variant constructs into E.coli for expression, purifying the wild-type and variant enzymes using Ni-NTA affinity chromatography, and finally performing in vitro enzymatic assays with DAUN and DOX for the purposes of comparing kinetic parameter values between the wild-type and variant enzymes.

137

The second novel finding, which was ascertained from the cell line study, is as follows:

2. A reduction in AKR and CBR-mediated metabolism of DAUN and DOX in cell lines

is associated with increased cytotoxicity following exposure with either of these

anthracyclines (Chapter 4).

The rationale for these experiments was based on previous literature suggesting that metabolism plays a major role in anthracycline-induced cardiotoxicity; however, there is disagreement on whether the alcohol metabolites, DAUNol and DOXol, or the parent drugs are the toxic species (Behenia and Boroujerdi, 1999; Ax et al., 2000; Olson et al. 2003; Veitch et al., 2009). To recap, nine cell lines, representing tissues from different organs (heart, prostate, ovary, pancreas, breast, liver, colon, lung, and kidney), were subjected to varying concentrations of either DAUN or DOX at specified time periods ranging from 6 to 48 hrs and cell viability (cytotoxicity) was measured. Cell viability, as a consequence of exposure to various concentrations of either the parent drug or its metabolite was determined using MTT assays, from which LC 50 values were determined. Four of the cell lines (liver, colon, lung, and kidney) were found to be more resistant to DAUN and DOX while the five remaining cell lines (heart, prostate, ovary, pancreas, and breast) were more sensitive. For the metabolic studies, the resistant and sensitive cells lines were treated with a specific concentration of either DAUN or DOX for the same time periods as in the cytotoxicity studies. The cytosolic extract derived from each cell line were prepared and subjected to in vitro assays to measure the rate of

138 conversion of DAUN/DOX to DAUNol/DOXol. The resistant cell lines had significantly higher rates of conversion of the parent drug to its metabolite than did the sensitive cells lines. In addition to the in vitro assays, the cytosols from the DAUN- and DOX-treated

cell lines were subjected to Western blot analyses, and the relative abundance of each of

the AKRs and CBRs, in each cell line, were measured. The levels of the AKR and CBR

enzymes were significantly higher in the resistant cell lines than those of the sensitive cell

lines. Finally, the association between cytotoxicity and AKR-/CBR-mediated

metabolism of DAUN and DOX at each exposure time period was assessed using

correlation analyses.

5.2 Implications of findings

If a reduction in DAUN/DOX metabolism, perhaps caused by variants of AKRs

and CBRs, plays a major role in the onset and progression of cardiotoxicity and other

toxic adverse effects, then one strategy to reduce these effects is to lower the dose of the

drug to be administered. Ultimately, the dosage should be lowered so that it is sufficient

to kill cancer cells while, at the same time prevent damage to the normal cells. Indeed,

this strategy has also been shown to result in a lower incidence of cardiotoxicity

compared with that seen in standard dosing controls (Safra 2003). A recent study has

demonstrated the lowering dosage of anthracyclines prevented cardiac side effects, yet

the dosage was still high enough to successfully treat the cancer. The study by Brouwer

et al. (2007) found no clinically relevant deterioration of cardiac function in survivors of

acute lymphoblastic leukemia treated with a lower dosage of DAUN (cumulative dose of

100 mg/m 2), even after a follow-up of 20 years post-treatment.

139 If the ideal therapeutic response is not attainable following low dosage infusion, another alternative is to use liposome-encapsulated formulations of anthracyclines

(Vergely et al., 2007). Liposomes easily accumulate at tumor sites by exiting the bloodstream at sites of leaky vasculature; however, they do not readily exit the circulation in healthy tissues, such as the heart, due to ‘tight’ endothelial capillary junctions

(Outomuro et al., 2007; Leonard et al., 2009). Thus, they may represent a more targeted delivery system, and this may lead to a reduction in toxicity compared with the conventional delivery (i.e., non-encapsulated infusion) these anthracyclines. In addition, liposomes should allow a slow release anthracyclines - over several days, thereby lowering the concentration of free drug in the bloodstream and limiting exposure to the heart and other non-targeted tissues (Vaage et al., 1994; Hobbs et al., 1998; Allen and

Martin 2004; Vail et al., 2004). This enhanced cardio-sparing effect has been reported in two studies, which required endomyocardial biopsy in order to detect anthracycline- induced cardiac damage (Berry et al., 1998; Safra et al., 2000). Currently, three formulations of liposomal anthracyclines are approved for clinical use: non-pegylated liposomal DOX (Myocet TM ), pegylated liposomal DOX (Caelyx ®, Doxil ®), and

liposomal DAUN (DaunoXome ®; mainly used for treatment of advanced AIDS-related

Kaposi’s sarcoma).

In addition to the aforementioned strategies to reduce potential deleterious effects

associated with these anthraycline drugs, other considerations might include the

implementation of antioxidants and antioxidant enzymes in DAUN/DOX-based

chemotherapies, which have been shown to protect against cardiotoxicity in several

experimental models. For example, the use of flavonoids has been linked to cardio-

140 protection due to their iron chelating and antioxidant properties (Bast et al., 2007;

Mojzisova et al., 2009; Li et al., 2009). Furthermore, the administration of coenzyme

Q10 , an essential component of the electron transport system and a potent intracellular

antioxidant, appears to prevent damage to the mitochondria of the heart, thus preventing

the development of anthraycline-induced cardiomyopathy (Conklin, 2005). Finally,

previous studies have demonstrated that the use of antioxidant vitamins (i.e., vitamins C

and E) has led to a significant reduction of anthracycline-induced cardiotoxicity in rat

models and isolated ventricular myocytes (Kalender et al., 2002; Wold et al., 2005), and

therefore should be taken into account as part of treatment in cancer patients undergoing

DAUN/DOX therapy.

5.3 Potential limitations of this research

5.3.1 The use of a bacterial expression system for the in vitro studies

A bacterial expression system was used to express the human genes and produce

the human AKR and CBR enzymes for the enzymology studies. In the last two decades

or so, E. coli based expression systems have been exploited for the production of a

variety of therapeutic proteins (Sahdev et al., 2008). This expression system provides

many benefits for in vitro applications, for example, E. coli are (i) inexpensive to culture,

(ii) can be grown quickly to high cell densities, and (iii) can be scaled up for high yield

production of the recombinant protein of interest (Crespi and Miller, 1999; Terpe, 2006;

Zerbs et al., 2009). However, bacterial hosts generally lack the enzymes and cellular

machinery that is required to generate the post-translational modifications often

associated with the final, functional form of mammalian proteins. These modifications

141 play a significant role in many proteins, including proper folding into three-dimensional structures, localization to the correct intracellular compartments, enzyme activation/deactivation, and enhanced stability (Seo and Lee, 2004; Diella et al., 2004;

Lee at al., 2006). Some examples of post-translational modifications consist of:

• Glycosylation: the addition of oligosaccharide (complex sugar) chains to either

asparagine, hydroxyllysine, serine, or threonine;

• Acetylation: the addition of an acetyl group, usually at the N-terminus of the protein;

• Methylation: a type of alkylation that involves the addition of a methyl group, usually

at lysine or arginine residues;

• Phosphorylation: the addition of a phosphate group, usually to serine, tyrosine,

threonine or histidine; and

• Sulfation: the addition of a sulfate group to a tyrosine.

Most of the post-translational modifications documented for AKRs and CBRs

involve N-acetylalanines, N-acetyllysines, phosphoserines, and phosphotyrosines

[Universal Protein database (http://www.uniprot.org/)]. These modifications are

associated with AKR1A1, AKR1B1, AKR1B10, AKR1C2, AKR1C4, AKR7A2, and

CBR1. To determine whether post-translational modifications affected enzymatic

activity of these reductases, I excised the histidine-tagged wild-type coding regions of

the AKRs and CBR1 from the pET vectors and cloned the coding regions of each gene

into an insect cell expression vector, p2ZOpie2F (Hegedus et al., 1998). These constructs

were then transfected into Sf9 insect cells (cell lines isolated from ovarian tissue of the

fall army worm, Spodoptera frugiperda ) using the cationic-lipid formulation, Cellfectin ®

(Invitrogen), and expression of the reductases were verified by Western blotting with

142 specific antibodies (those used in Chapters 2 and 3). Insect expression systems are known to provide virtually all of the post-translation modifications that occur in mammalian cells, with the exception of sylation (Altmann et al., 1999; Hollister et al.,

2002; Harrison and Jarvis 2006). These InsectSelect™ produced “human” reductases were purified using Ni-NTA affinity chromatography (Figure 37) and the purified enzymes were subjected to metabolic assays using appropriate test substrates. The results from these assays are provided in Table 15. No significant differences in activity were detected between the enzymes expressed and purified from the insect-cell expression system and those obtained from the bacterial expression system (all n=9). Therefore, I proceeded to use the bacterial system for my detailed in vitro enzymology studies with the test substrates as well as DAUN and DOX. No post-translational modifications were noted in the database for the remaining enzymes: AKR1C1, AKR1C3, CBR3, and CBR4.

143

Figure 37—Purification of human recombinant 6x-His-tagged wild-type enzymes for Sf9 insect cells: (1) AKRA1, (2) AKR1B1, (3) AKR1B10, (4) AKR1C2, (5) AKR1C4, (6) AKR7A2, and (7) CBR1. ( Top ) Gel stained with SYPRO ® Ruby following SDS-PAGE showing purified protein fractions (1 µg). ( Bottom ) Western blot detection of purified protein fractions confirms expression of the desired AKR and CBR proteins. GST-tagged full recombinant proteins (Abnova ® Corporation, Taipei City, Taiwan; 2 µg total protein) and human liver cytosol (20 µg total protein) were used as positive controls (P) for antibody immunoreactivity for the AKRs and CBR1, respectively. M refers to the molecular weight markers.

144 Table 15 —Enzymatic rates for reported test substrates by recombinant 6x-His tagged AKR wild-type along with kinetic parameters for CBR1 wild-type in the presence of menadione.

RATES (nmol/min ● mg RATES (nmol/min ● mg ENZYMES SUBSTRATES protein) for insect cell protein) for bacterial expressed recombinant expressed recombinant AKR1A1 DL -glyceraldehyde 1430±210 1240±170 AKR1B1 DL-glyceraldehyde 410±37 354±48 AKR1B10 (S)-1-indanol 1070±90 980±80 AKR1C2 1-acenaphthenol 3120±190 2980±151 AKR1C4 1-acenaphthenol 1450±135 1660±190

9,10- AKR7A2 2800±260 2553±229 phenanthrenequinone

CBR 1 KINETIC PARAMETERS using menadione V (nmol/min ● max K (µM) k (s -1)a k / K (s -1 M-1) mg protein) m cat cat m Insect cell expressed 654±31 19±3 0.37±0.02 19500±2100 recombinant Bacterial expressed 582±43 20±6 0.33±0.02 17400±2800 recombinant

The rates for both the insect cell and bacterial expressed recombinants are provided for comparison purposes. Values correspond to mean ± S.D obtained from three experiments performed with three independent enzyme preparations ( n=9) for each isoform. Reported rates for the wild-type enzymes are also provided for comparison purposes a kcat calculated for 6x-His tagged CBR1 from Mr 34000.

5.3.2 Focusing solely on the polymorphisms in the coding regions of the AKR and CBR genes

The polymorphisms selected for this thesis were the ns-SNPs in the AKR and

CBR genes. Synonymous SNPs were excluded since these base pair changes do not result in a change from one amino acid to another in the polypeptide sequence, which means that the activity should be identical to that of the wild-type enzymes. Also, the polymorphisms associated with the non-coding regions (i.e., introns and promoters) were not included since there was already a tremendous amount of work required to determine the level to which each of the 38 ns-SNPs (excluding the E36Term variant of AKR1C3)

145 affected enzyme activity in comparison to their respect wild-type counterparts. More importantly, there is a lack of information on the impact that genetic polymorphisms in non-coding regions of human AKRs and CBRs may have on enzyme amounts or function. In addition, the regulatory regions of these genes, such as enhancers, silencers, and response elements, are not well defined. Clearly, more studies need to be performed on these genes to define the important regulatory regions.

5.3.3 Allele frequencies for two variants are not reported in the NCBI database

For two of the variants (E55D: significantly altered metabolic activity compared to the AKR1A1 wild-type enzyme; and D239V: did not significantly alter activity of the

CBR3 wild-type enzyme), there is no frequency data reported in the NCBI database.

Therefore, studies on the clinical relevance of these allelic variants may be somewhat premature. Clearly, if the E55D and D239V alleles do not occur with any significant frequency in the human population, they will have little or no physiological meaning in determining patient variability when it comes to metabolism of DAUN and DOX.

Nevertheless, the data for these two variants do provide new protein structure-function information that identifies important residues for function of their respective enzyme

(E55D), or residues that are not critical for function (D239V).

5.4 Suggested future research directions

Recommendations for future research directions focus on two areas: (i) biochemical investigations and (ii) cancer patient correlation studies.

146 5.4.1 Biochemical investigations

The studies in this thesis focused on AKR enzymes that have been reported to be involved in the metabolism of DOX and DAUN. A new direction would be to repeat these in vitro studies with other AKRs for which ns-SNP frequencies have been reported, namely: AKR1E2, AKR6A5, and AKR7A3 (Table 16). This would increase the breadth of knowledge pertaining to anthracycline metabolism since these enzymes may potentially be added to the list of those that have been demonstrated to metabolize DAUN and/or DOX to their corresponding alcohols. Such data may also suggest that these variants could also be involved in the dose-dependent cardiotoxicity and/or other side effects seen in cancer patients during and following DAUN/DOX therapy.

Table 16 —Allele frequencies of the non-synonymous single nucleotide polymorphic variants of human AKR enzymes from different ethnic groups.

NCBI rs- ENZYMES VARIANTS ALLELE FREQUENCIES a number

D35N rs61745201 CEU=2.8% ( n=72) C52G rs35429729 YRI=2.5% ( n=120)

CEU=5.8% ( n=120); CHB=12.2% ( n=90); AKR1E2 K86R rs17133693 JPT=7.8% ( n=90)

L186W rs76596595 CEU=1.4% ( n=72) R301Term rs12240276 CEU=10.8% ( n=120); YRI=12.5% ( n=120) AKR6A5 E74K rs2229003 MITO=47.7% ( n=38) CEU=83.3% ( n=120); CHB=80.0% ( n=90); T323A rs1738025 JPT=79.5% ( n=88) AKR7A3 CEU=84.2% ( n=120); CHB=80.0% ( n=90); N215D rs1738023 JPT=78.9% ( n=90); YRI=65.8% ( n=120) a Allele frequencies are obtained from The National Center for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov/). The ethnic groups are designated as follows: YRI (African), CHB (Chinese), CEU (European), JPT (Japanese), and MITO (MITOGPOP6: combination of Caucasian, Japanese, Chinese, Amerindian, Pygmy, Russian, etc.). The chromosome sample counts ( n) for each variant are also given.

147

Future investigations can also be undertaken in relation to the findings in Chapter

4. Even though an association between AKR/CBR metabolism and DAUN/DOX- induced cytotoxicity was established in this study, these findings do not confirm a direct relationship between the reductases and cytotoxicity. Nevertheless, the results provide good reason to look into performing future experiments in order to determine the degree to which AKRs and CBRs are linked to toxicity following exposure to the anthracyclines.

These experiments would include:

• Taking cell lines with high expression of AKRs and CBRs (i.e., liver, colon, lung, and

kidney) and inhibiting/knocking out or knocking down (reducing) the metabolic

ability of the individual reductase enzymes and determining whether sensitivity to

DAUN and/or DOX is increased. Expression of AKRs and CBRs can be decreased

or eliminated using gene-silencing systems such as RNA interference (RNAi). The

basic mechanism of RNAi-mediated gene silencing involves delivering small

interfering RNAs (siRNAs), which are double-stranded and about 21 to 28

nucleotides in length, into a cell where they associate with specific proteins to form a

ribonucleoprotein complex (Bonetta, 2004). This complex then scans the mRNA to

degrade the corresponding mRNA target in a highly specific manner; the net effect is

that the protein in question is not made. However, RNAi is not always specific and

can affect the expression of a number of off-target products (Jackson and Linsley,

2004). Alternately metabolic inhibition experiments can be performed through the

use of specific chemical agents or antibodies specific to the AKR and CBR isoforms

to reduce or eliminate activity of these enzymes (Coleman, 2005; Gelboin and Krausz

148 2006). Like RNAi, there are challenges with using inhibitors and antibodies, namely,

that they may cross-react with other enzymes, thereby, confounding the findings

regarding the contribution of the enzyme of interest to the observed toxicity.

• Taking cell lines with low expression of AKRs and CBRs (i.e., heart, prostate, ovary,

pancreas, and breast) and increasing the expression of the individual AKR and CBR

enzymes to determine whether the increase in enzymatic activity correlates with a

decrease in sensitivity to the drugs. The increase in reductase expression can be

attained by transfecting these cell lines with expression vectors containing the AKR

or CBR gene of interest. For each cell line, enhanced expression of the individual

reductases in the transfected cells would be assessed by Western blotting and

measured against untransfected cells. Once this is established, the cytotoxicity

experiments of the transfected and untransfected cells could be performed in the

presence of DAUN and DOX to evaluate and compare LC50 values.

Another set of future experiments would involve an extension of the cell toxicity

study with the focus on measuring reactive oxygen species (ROS) levels in the resistant

and sensitive cell lines during DAUN and DOX exposure. It would be intriguing to see if

an association exists between the amount of ROS generated and the sensitivity of the cell

line to each drug. If significant increases in ROS levels are detected with the sensitive

cell lines, then these reactive molecules likely contribute to DAUN/DOX-induced

cytotoxicity. Also, since the sensitive cell lines exhibited lower metabolic activity

towards these anthracyclines, then it is possible decreased metabolism is linked to an

increase in ROS production. One technique that allows for identification and quantitation

of different ROS species is electron paramagnetic resonance (EPR) spin trapping, which

149 is of particular interest since it allows direct detection of free radicals in intact biological

- systems (Berliner and Fujii et al., 2004). The detection of ROS, such as ·O 2 and ·OH, is

difficult to attain since these free radicals are short lived species; however, the

introduction of spin trapping agents have alleviated this problem since they are able to

react with free radicals to form another free radical product that is considerably more

stable (Figure 38) (Khan and Swartz, 2002; Swartz et al., 2007).

Figure 38—A spin trapping chemical reaction in which 5-Diethoxyphosphoryl-5-methyl-1-pyrroline N- oxide (DEPMPO), the spin trapping agent, reacts with a free radical to form a radical adduct that is more stable than the initial radical (Swartz et al., 2007).

I have performed some preliminary studies using EPR to see if: (i) ROS are generated in the in vitro metabolic assays with the AKRs and CBRs in the presence of

either DAUN or DOX, and (ii) ROS levels are altered in metabolic assays with the

variants exhibiting significantly reduced activity towards DAUN or DOX compared to

their respective wild-type enzyme. As a starting point, I compared the following wild-

type and variant enzymes for the EPR experiments: (i) CBR1 and V88I; (ii) CBR3 and

C4Y; (iii) AKR7A2 and A142T; as well as (iv) AKR1C3 and R170C. The metabolic

assays for the EPR experiments were performed using identical conditions to the assays

used in determining DAUNol/DOXol levels (Chapters 2 and 3), except for the reaction

time being reduced to 30 min, as well as the addition of the DEPMPO spin-trapping agent

150 (20 mM) and the chelating agent, diethylenetriaminepentaacetic acid (DTPA) (1 mM).

DTPA sequesters any transition metal ions in the reaction medium since they can induce

ROS by Fenton reactions, an oxidation process in which reaction of H 2O2 with iron

- generates ·O 2 and ·OH (Anipsitakis and Dionysiou, 2004; Valko et al., 2005; Valko et

al., 2006). This alleviates any potential background ROS signal contribution from the

medium. An in vitro positive control [xanthine (0.25 mM), xanthine oxidase (0.08 U/ml),

o DEPMPO (20 mM) and DTPA (1 mM) in PBS, pH 7.4 at 37 C for 30 min (Dambrova et al., 2000)] as well as a negative control (without the addition of DEPMPO) were performed concurrently. Following completion of the assays, aliquots of the reaction medium were dispensed into quartz EPR tubes, frozen in liquid nitrogen, and subjected to

EPR analysis. An example of the EPR spectra generated from the positive and negative control assays is illustrated in Figure 39.

Figure 39—EPR spectra of the positive and negative controls assays using DEPMPO. For the positive control assay (xanthine/xanthine oxidase), the spectra representing the DEPMPO-superoxide spin adduct is given in red while the negative control (no DEPMPO) is represented in blue. Recordings shown were performed on samples withdrawn 15 min after the start of the incubation. The EPR spectrometer settings were as follows: microwave frequency, 9.41 GHz; microwave power, 0.6339 mW; modulation frequency, 100 kHz; modulation amplitude, 1 Gauss; receiver gain, 60 dB; and time constant, 1.28 msec.

151

For the in vitro assays involving the AKRs and CBRs, no ROS levels were

detected. Based on the expert advice given by Dr. Pierre Kennepohl (Department of

Chemistry, UBC) as well as Mr. Vlad Martin-Diaconescu (a recent Ph.D. alumnus from

Dr. Kennepohl’s laboratory), modifications were made to the assay conditions for the

purposes of increasing ROS detection. These modifications included (i) increasing the

concentration of DEPMPO in the assays to 100 mM, (ii) dispensing the assay media into

EPR tubes that hold larger volumes, and (iii) repeating the assays using another spin

trapping agent, N-tert-butyl-α-phenylnitrone (PBN) at 20 and 100 mM. Again, no ROS levels were detected using any of the aforementioned assay modifications. Since the techniques employed in these EPR studies with DEPMPO are sensitive enough to detect

ROS at the levels typically found in cells (Liu et al., 1999; Valgimigli et al., 2001;

Mojovic et al., 2005; Hardy et al., 2007; Bacic et al., 2008), I conclude that no significant amount of ROS are produced by these enzymatic reactions. Due to these findings, I did not proceed further with looking at ROS in my research project.

Even though the in vitro assays did not generate detectable levels of ROS, these preliminary findings do not rule out the contribution of ROS to anthracycline-induced cardiotoxicity. Future studies could be performed ex vivo by comparing ROS levels

following either DAUN or DOX exposure between non-transfected H9c2 heart cell lines

and those that are transfected with the wild-type AKR or CBR based on other studies.

L'Ecuyer et al. (2001) demonstrated that H9c2 cells can be used to study free radical

production, and can be engineered to express foreign genes at controllable levels, making

them a suitable system to study molecular responses to oxidative damage. Previous

152 works in this area include studies of oxidative cell damage caused by DOX (L’Ecuyer et al., 2001; Filigheddu et al., 2001; Spallarossa et al., 2004). Successful transfection can be verified by Western blotting with AKR or CBR specific antibodies, and these cell lines should have a greater ability to metabolize DAUN and DOX. If the findings from this set of future studies demonstrate that non-transfected H9c2 cell lines generate ROS at significantly greater levels than the transfected H9c2 cell lines, then it is likely that decreased metabolism (i.e., more of the parent drug present), which is inversely related to cytotoxicity from my findings in Chapter 4, is also inversely related to ROS generation.

5.4.2 Cancer patient correlation studies to determine if AKR and CBR variant alleles may serve as genetic biomarkers of cardiotoxicity arising from DAUN/DOX therapy

Since the studies performed in this thesis reveal that the parent drug is more cytotoxic than its primary metabolite, it is possible that those individuals who have an ns-

SNP in the appropriate gene or combination of genes (which includes one or more of the

AKRs and/or CBRs) could be at higher risk for developing cardiotoxicity following exposure to DAUN or DOX. This leads into the following hypothesis and prediction:

HYPOTHESIS: Individuals with ns-SNPs in genes encoding AKR and CBR enzymes are at greater risk for the development of cardiotoxicity if they receive DAUN or DOX treatment.

PREDICTION: If ns-SNPs in one or more of the AKR or CBR genes contribute to, and thus are mechanistically associated with, the likelihood of developing cardiotoxicity

153 following drug treatment, then this hypothesis makes the prediction that those ns-SNPs that have the strongest effect on enzyme activity are more likely to be found among patients who develop cardiotoxicity. At the same time, those variants that have a lesser affect on enzyme activity would be found at slightly lower levels among cardiotoxic patients, or at approximately the same levels as found in the normal population.

Genotyping analyses are now quite easy to do, although they are not routinely done. The testing of the hypothesis requires genotyping a relatively large number

(several hundred) cardiotoxic patients along with a control population, and performing association studies between each genotype (the individual ns-SNPs and haplotype combinations) and (i) cardiotoxicity, as well as (ii) no affect on the heart (control). The measurement of a cardiotoxicity is a reduction in left ventricular ejection fraction (LVEF; the fraction of blood pumped out of the left ventricle with each heart beat) drop by more than 10% after the start of treatment, or an LVEF of less than 50% (Schwartz et al., 1987;

Nousiainen et al., 2002; Daugaard et al., 2005; Belham et al., 2007).

Whether a correlation exists between the individual ns-SNPs or haplotype combinations and the development of cardiotoxicity, the next question becomes whether the mutation is a simple biomarker or does it play a mechanistic role in the likelihood of developing cardiotoxicity following drug treatment. If the mutation has a mechanistic role, rather than being a simple biomarker, then there should be a strong correlation between strong mutant alleles and cardiotoxicity and a lesser correlation with the weaker alleles.

The establishment of AKR and CBR genetic biomarkers would allow clinicians to

154 identify individuals at high risk for developing cardiotoxicity by simply collecting blood samples from patients prior to DAUN or DOX therapy and extracting DNA for purposes of genotyping for the presence of the high risk variant allele(s). This would allow patients and clinicians to choose alternate therapies that will minimize their risk. For patients that are at low risk of developing cardiotoxic effects, dosage regimens of DAUN and DOX could be increased to treat cancers effectively.

If AKRs and CBRs are solely responsible for DAUN/DOX metabolism, then one would expect that cancer patients carrying alleles leading to the expression of AKR/CBR enzyme variants with significantly reduced metabolic activity will be likely to have lower rates for metabolizing these anthracyclines. However, AKR/CBR-mediated metabolism is likely a single aspect of what is highly complex patient phenotype when it comes to cardiotoxicity. The likelihood that a single variant allele in one enzyme will be the lone determinant of an individual’s susceptibility to this adverse effect is low. Therefore, genotyping studies should include looking at other enzymes capable of metabolizing

DAUN and DOX, such as the cytochrome P450 enzymes (CYPs), for which there are variant alleles in the human population (Goeptar et al., 1993; Zhang et al., 2009).

However, CYPs (i.e., CYP2D6) are apparently involved to a minor extent in the metabolism of anthracyclines (Le Guellec et al., 1993). Like the AKRs and CBRs, it is possible that some of the CYP variants may significantly reduce metabolism of these anthracyclines compared to their wild-type counterparts. In addition to looking at drug metabolizing enzymes, genotyping studies look at the frequency of genes representing efflux transporters since they may alter DOX/DAUN pharmacokinetics to the point of developing anthracycline-induced cardiotoxicity. For example, one study suggested that

155 cardiotoxicity was associated with the G671V variant of the DOX efflux transporter multidrug resistance associated protein 1 (MRP1) and with the V1188E-C1515Y haplotype (i.e., set of alleles of closely linked loci on a chromosome that tend to be inherited together) of the functionally similar efflux transporter multidrug resistance associated protein 2 (MRP2) (Deng and Wojnowski et al., 2007). Overall, genetic biomarkers in the form of SNPs in genes encoding drug metabolizing enzymes and transporters could be successfully used to identify patients at high risk of cardiotoxicity prior to therapy. This would facilitate early diagnosis of susceptibility to cardiotoxicity and improve DAUN and DOX therapy. On the other hand, if the metabolite, or a down stream product of the drug metabolism, is the primary cause of the damage to the heart, then those individuals that have an ns-SNP in the appropriate AKR or CBR gene, or combination of genes, would be more resistant to the development of cardiotoxicity. In other words, the ns-SNPs would prove to be cardio-protective, and they might be safely given a higher dose of the drug, which correlates with improved chance of remission from cancer.

156 REFERENCES

Akpofure C, Riley CA, Sinkule JA, Evans WE (1982) Quantitation of daunorubicin and its metabolites by high-performance liquid chromatography with electrochemical detection. J Chromatogr 232(2): 377-383.

Allen TM, and Martin FJ. (2004) Advantages of liposomal delivery systems for anthracyclines. Semin Oncol 31: 5–15.

Altmann F, Staudacher E, Wilson IB, and Marz L (1999) Insect cells as hosts for the expression of recombinant glycoproteins. Glycoconj J 16(2): 109-123.

American Cancer Society (2010) Cancer facts & figures 2010. http://www.cancer.org/downloads/STT/Cancer_Facts_and_Figures_2010.pdf Anderlini P, Benjamin RS, Wong FC, Kantarjian HM, Andreeff M, Kornblau SM, O'Brien S, Mackay B, Ewer MS, and Pierce SA (1995) Idarubicin cardiotoxicity: a retrospective study in acute myeloid leukemia and myelodysplasia. J Clin Oncol 13 : 2827 -2834.

Andrews PA, Brenner DE, Chou FT, Kubo H, and Bachur NR (1980) Facile and definitive determination of human adriamycin and daunoribicin metabolites by high-pressure liquid chromatography. Drug Metab Dispos 8(3): 152-156.

Anipsitakis GP, and Dionysiou DD (2004) Radical generation by the interaction of transition metals with common oxidants. Environ Sci Technol 38(13): 3702-3712.

Arnau J, Lauritzen C, Petersen GE, and Pedersen (2006) Current strategies for the use of affinity tags and tag removal for the purification of recombinant proteins. Protein Expr Purif 48: 1-13.

Auerbach G, Herrmann A, Gutlich M, Fischer M, Jacob U, Bacher A, and Huber R (1997) The 1.25 A crystal structure of sepiapterin reductase reveals its binding mode to pterins and brain neurotransmitters. Embo J 16: 7219–7230.

Ax W, Soldan M, and Maser E (2000) Development of daunorubicin resistance in tumour cells by induction of carbonyl reduction. Biochem Pharmacol 59(3): 293-300.

Bachur NR (1971) Daunorubicinol, a major metabolite of daunorubicin: isolation from human urine and enzymatic reactions. J Pharmacol Exp Ther 177(3): 573-578.

Bacic G, Spasojevic I, Secerov B, and Mojovic M (2008) Spin-trapping of oxygen free radicals in chemical and biological systems: new traps, radicals and possibilities. Spectrochim Acta A Mol Biomol Spectrosc 69(5): 1354-1366.

157 Bains OS, Takahashi RH, Pfeifer TA, Grigliatti TA, Reid RE, and Riggs KW (2008) Two allelic variants of aldo-ketoreductase 1A1 exhibit reduced in vitro metabolism of daunorubicin. Drug Metab Dispos 36: 904-910.

Bains OS, Karkling MJ, Grigliatti TA, Reid RE, and Riggs KW (2009) Two non- synonymous single nucleotide polymorphisms of carbonyl reductase 1 demonstrate reduced in vitro metabolism of daunorubicin and doxorubicin. Drug Metab Dispos 37: 1107-1114.

Bains OS, Karkling MJ, Grigliatti TA, Reid RE, and Riggs KW (2010a) Naturally occurring variants of human CBR3 alter anthracycline in vitro metabolism. J Pharmacol Exp Ther 332(3): 755-763.

Bains OS, Grigliatti TA, Reid RE, and Riggs KW (2010b) Naturally occurring variants of human aldo-keto reductases with reduced in vitro metabolism of daunorubicin and doxorubicin. J Pharmacol Exp Ther doi:10.1124/jpet.110.173179.

Balis FM, Holcenberg JS, and Bleyer WA (1983) Clinical pharmacokinetics of commonly used anticancer drugs. Clin Pharmacokinet 8(3): 202-232.

Barry E, Alvarez JA, Scully RE, Miller TL, and Lipshultz SE (2007) Anthracycline- induced cardiotoxicity: course, pathophysiology, prevention and management. Expert Opin Pharmacother 8: 1039-1058.

Barski OA, Tipparaj SM, and Bhatnagar A (2008) The aldo-keto reductase superfamily and its role in drug metabolism and detoxification. Drug Metab Rev 40(4): 553- 624.

Bast A, Haenen GR, Bruynzeel AM, and Van der Vijgh WJ (2007) Protection by flavonoids against anthracycline cardiotoxicity: from chemistry to clinical trials. Cardiovasc Toxicol 7(2): 154-159.

Behnia K, and Boroujerdi M (1999) Inhibition of aldo-keto reductases by phenobarbital alters metabolism, pharmacokinetics and toxicity of doxorubicin in rats. J Pharm Pharmacol 51(11): 1275-1282.

Belham M, Kruger A, Mepham S, Faganello G, and Pritchard C (2007) Monitoring left ventricular function in adults receiving anthracycline-containing chemotherapy. Eur J Heart Fail 9(4): 409-414.

Benach J, Atrian S, Gonzalez-Duarte R, and Ladenstein R (1999). The catalytic reaction and inhibition mechanism of Drosophila alcohol dehydrogenase: Observation of an enzyme-bound NAD-ketone adduct at 1.4 A resolution by X-ray crystallography. J Mol Biol 289: 335–355.

158 Berliner LJ, and Fujii H (2004) In vivo spin trapping of nitric oxide. Antioxid Redox Signal 6(3) : 649-656.

Berry G, Billingham M, Alderman E, Richardson P, Torti F, Lum B, Patek A, and Martin FJ (1998) The use of cardiac biopsy to demonstrate reduced cardiotoxicity in AIDS Kaposi’s sarcoma patients treated with pegylated liposomal doxorubicin. Ann Oncol 9(7): 711-716.

Binaschi M, Bigioni M, Cipollone A, Rossi C, Goso C, Maggi CA, Capranico G, and Animati F (2001) Anthracyclines: selected new developments. Curr Med Chem Anticancer Agents 1: 113–130.

Blanco JG, Lesisenring WM, Gonzalez-Covarrubias VM, Kawashima TI, Davies SM, Relling MV, Robison LL, Sklar CA, Stovall M, and Bhatia S (2008) Genetic polymorphisms in carbonyl reductase 3 gene CBR3 and the NAD(P)H:quinone oxidoreductase 1 gene NQO1 in patients who developed anthracycline-related congestive heart failure after childhood cancer. Cancer 112: 2789-2795.

Bonetta, L (2004) RNAi: silencing never sounded better. Nat Methods 1(1): 79-86.

Boucek RJ Jr, Olson RD, Brenner DE, Ogunbunmi EM, Inui M, and Fleischer S (1987) The major metabolite of doxorubicin is a potent inhibitor of membrane-associated ion pumps: a correlative study of cardiac muscle with isolated membrane fractions. J Biol Chem 262: 15851-15856.

Brouwer CA, Gietema JA, van den Berg MP, Bink-Boelkens MT, Elzenga NJ, Haaksma J, Kamps WA, Vonk JM, and Postma A (2007) Low-dose anthracyclines in childhood acute lymphoblastic leukemia (ALL): no cardiac deterioration more than 20 years post-treatment. J Cancer Surviv 1(4): 255-260.

Buchman AR, and Berg P (1988) Comparison of intron-dependent and intron- independent gene expression. Mol Cell Biol 8(10): 4395-4405.

Burczynski ME, Harvey RG and Penning TM (1998) Expression and characterization of four recombinant human dihydrodiol dehydrogenase isoforms: oxidation of trans - 7, 8-dihydroxy-7,8-dihydrobenzo[ a]pyrene to the activated o-quinone metabolite benzo[ a]pyrene-7,8-dione. Biochemistry 37(19): 6781–6790.

Byrns MC, Steckelbroeck S and Penning TM (2008) An indomethacin analogue, N-(4- chlorobenzoyl)-melatonin, is a selective inhibitor of aldo-keto reductase 1C3 (type 2 3 α-HSD, type 5 17 β-HSD, and prostaglandin F synthase), a potential target for the treatment of hormone dependent and hormone independent malignancies. Biochem Pharmacol 75(2): 484-493.

159 Camaggi CM, Comparsi R, Strocchi E, Testoni F, Angelelli B, and Pannuti F (1988a) Epirubicin and doxorubicin comparative metabolism and pharmacokinetics. A cross-over study. Cancer Chemother Pharmacol 21(3): 221-228.

Camaggi CM, Comparsi R, Strocchi E, Testoni F, and Pannuti F (1988b) HPLC analysis of doxorubicin, epirubicin and fluorescent metabolites in biological fluids. Cancer Chemother Pharmacol 21(3): 216-220.

Coleman MD (2005) Human drug metabolism: an introduction. John Wiley & Sons: West Sussex, UK.

Conklin KA (2005) Coenzyme Q 10 for prevention of anthracycline-induced cardiotoxicity. Integr Cancer Ther 4(2): 110-130.

Cortes-Funes H, and Coronado C (2007) Role of anthracyclines in the era of targeted therapy. Cardiovasc Toxicol 7(2): 56-60.

Crespi CL, and Miller VP (1999) The use of heterologously expressed drug metabolizing enzymes—state of the art and prospects for the future. Pharmacol Ther 84: 121- 131.

Cummings J, Milstead R, Cunningham D, and Kaye S (1986) Marked inter-patient variation in adriamycin biotransformation to 7-deoxyaglycones: evidence from metabolites identified in serum. Eur J Cancer Clin Oncol 22(8): 991-1001.

Cummings J, Anderson L, Willmott N and Smyth JF (1991) The molecular pharmacology of doxorubicin in vivo . Eur J Cancer 27: 532-535.

Dabich L, Bull FE, Beltran G, Athens JW, Coltman CA Jr, Weick JK, Van Slyck EJ, and Amare M (1986) Phase II evaluation of aclarubicin in refractory adult acute leukemia: a Southwest Oncology Group Study. Cancer Treat Rep 70 : 967-969.

Dambrova M, BAumane L, Kalvinsh I, and Wikberg JE (2000) Improved method of EPR detection of DEPMPO-superoxide radicals by liquid nitrogen freezing. Biochem Biophys Res Commun 275(3): 895-898.

Danesi R, Fogli S, Gennari A, Conte P, and Del Tacca M (2002) Pharmacokinetic pharmacodynamic relationships of the anthracycline anticancer drugs. Clin Pharmacokinet 41: 431-444.

Daugarrd G, Lassen U, Bie P, Pedersen EB, Jensen KT, Abildgaard U, Hesse B, and Kjaer A (2005) Natriuretic peptides in the monitoring of anthracycline induced reduction in left ventricular ejection fraction. Eur J Heart Fail 7(1): 87-93.

Deng S, and Wojnowski L (2007) Genotyping the risk of anthracycline-induced cardiotoxicity. Cardiovasc Toxicol 7(2): 129-134.

160

Dhingra K, Frye D, Newman RA, Walters R, Theriault R, Fraschini G, Smith T, Buzdar A, and Hortobagyi GN (1995) Phase II clinical and pharmacological study of pirarubicin in combination with 5-fluorouracil and cyclophosphamide in metastatic breast cancer. Clin Cancer Res 1: 691 -697.

Di Costanzo L, Penning TM and Christianson DW (2009) Aldo-keto reductases in which the conserved catalytic histidine is substituted. Chem Biol Interact 178(1-3): 127- 133.

Diella F, Cameron S, Gemund C, Linding R, Via A, Kuster B, Sicheritz-Ponten T, Blom N, and Gibson TJ (2004) Phospho.ELM: a database of experimentally verified phosphorylation sites in eukaryotic proteins. BMC Bioinformatics 5: 79.

Di Luccio E, Elling RA and Wilson DK (2006) Identification of a novel NADH-specific aldo-keto reductase during sequence and structural homologies. Biochem J 400(1): 105-114.

Doroshow JH (1983) Anthracycline antibiotic-stimulated superoxide, hydrogen peroxide, and hydroxyl radical production by NADH dehydrogenase. Cancer Res 43(10): 4543-4551.

Duax WL, Ghosh D, Pletnev V (2000) Steroid dehydrogenase structures, mechanism of action, and disease. Vitam Horm 58: 121–148.

Duclohier H (2004) Helical kink and channel behaviour: a comparative study with peptaibols alamethicin, trichotoxin, and antiamoebin. Eur Biophys J 33(3): 169- 174.

Eksborg S, Stendahl U, and Lonroth U (1986) Comparative pharmacokinetic study of adriamycin and 4'epi-adriamycin after their simultaneous intravenous administration. Eur J Clin Pharmacol 30(5): 629-631.

Ellis EM (2007) Reactive carbonyl and oxidative stress: potential for therapeutic intervention. Pharmacol Ther 115(1): 13-24.

Endo S, Matsunaga T, Kitade Y, Ohno S, Tajima K, El-Kabbani O, and Hara A (2008) Human carbonyl reductase 4 is a mitochondrial NADPH-dependent quinone reductase. Biochem Biophys Res Commun 377: 1326-1330.

Esposito BP, Epsztejn S, Breuer W, Cabantchik ZI (2002) A review of fluorescence methods for assessing labile iron in cell and biological fluids. Anal Biochem 304(1): 1-18.

Fassas A, and Anagnostopoulos A (2005) The use of liposomal daunorubicin (DaunoXome) in acute myeloid leukemia. Leuk Lymphoma 46(6): 795-802.

161

Feldman AM, Lorell BH, and Reis SE (2000) Trastuzumab in the treatment of metastatic breast cancer: anticancer therapy versus cardiotoxicity. Circulation 102(3): 272- 274.

Felsted RL, Richter DR, and Bachur NR (1977) Rat liver aldehyde reductase. Biochem Pharmacol 26(12): 1117-1124.

Filigheddu N, Fubini A, Baldanzi G, Cutrupi S, Ghe C, Catapano F, Broglio F, Bosia A, Papotti M, Muccioli G, Ghigo E, Deghenghi R, Graziani A(2001) Hexarelin protects H9c2 cardiomyocytes from doxorubicin-induced cell death. Endocrine 14: 113-119.

Filling C, Berndt KD, Benach J, Knapp S, Prozorovski T, Nordling E, Ladenstien R, Jornvall H, and Oppermann U (2002) Critical residues for structure and catalysis in short-chain dehydrogenases/reductases. J Biol Chem 277: 25677- 25684.

Floyd JD, Nguyen DT, Lobins RL, Bashir Q, Doll DC, and Perry MC (2005) Cardiotoxicity of cancer therapy. J Clin Oncol 23(30): 7685-7696.

Flynn TG, and Green NC (1993) The aldo-keto reductases: an overview. Adv Exp Med Biol 328:251–257.

Forrest GL and Gonzalez B (2000) Carbonyl reductase. Chem Biol Interact 129(1-2): 21- 40.

Fu H, Grimsley GR, Razvi A, Scholtz JM, and Pace CN (2009) Increasing protein stability by improving β-turns. Proteins 77(3): 491-498.

Gaberc-Porekar V, and Menart V (2001) Perspectives of immobilized-metal affinity chromatography. J Biochem Biophys Methods 49: 335-360.

Gatta G, Capocaccia R, Coleman MP, Ries LA, and Berrino F (2002) Childhood cancer survival in Europe and the United States. Cancer 95: 1767–1772.

Gavelova M, Hladikova J, Vildova L, Novotna R, Vondracek J, Krcmar P, Machala M, and Skalova L (2008) Reduction of doxorubicin and oracin and induction of carbonyl reductase in human breast carcinoma MCF-7 cells. Chem Biol Interact 176(1): 9-18.

Gehl J, Boesgaard M, Paaske T, Vittrup Jensen B, and Dombernowsky P (1996) Combined doxorubicin and paclitaxel in advanced breast cancer: effective and cardiotoxic. Ann Oncol 7(7): 687-693.

162 Gelboin HV, and Krausz K (2006) Monoclonal antibodies and multifunctional cytochrome P450: drug metabolism as paradigm. J Clin Pharmacol 46(3): 353- 372.

Gewirtz DA (1999) A critical evaluation of the mechanisms of action proposed for the antitumor effects of the anthracycline antibiotics adriamycin and daunorubicin. Biochem Pharmacol 57: 727-741.

Ghosh D, Wawrzak Z, Weeks CM, Duax WL, Erman M (1994) The refined three- dimensional structure of 3 α,20 β-hydroxysteroid dehydrogenase and possible roles of the residues conserved in short-chain dehydrogenases. Structure 2: 629– 640.

Gianni L, Munzone E, Capri G, Fulfaro F, Tarenzi E, Villani F, Spreafico C, Laffranchi A, Caraceni A, Martíni C, Stefanelli M, Balagussa P, and Bonadonna G (1995) Paclitaxel by 3-hour infusión in combination with bolus doxorubicin in women with untreated metastatic breast cancer: high antitumor efficacy and cardiac effects in a dose-finding and sequence-finding study. J Clin Oncol 13(11): 2688- 2699.

Gianni L, Salvatorelli E, and Minotti G (2007) Anthracycline cardiotoxicity in breast cancer patients: synergism with trastuzumab and taxanes. Cardiovasc Toxicol 7(2): 67-71.

Giantris A, Abdurrahman L, Hinkle A, Asselin B, and Lipshultz SE (1998) Anthracycline-induced cardiotoxicity in children and young adults. Crit Rev Oncol Hematol 27: 53–68.

Gilladoga AC, Manuel C, Tan CT, Wollnet N, Sternberg SS, and Murphy ML (1976) The cardiotoxicity of adriamycin and daunomycin in children. Cancer 37(2 Suppl): 1070-1078.

Gilleron M, Marechal X, Montaigne D, Franczak J, Neviere R, and Lancel S (2009) NADPH oxidases participate to doxorubicin-induced cardiac myocyte apoptosis. Biochem Biophys Res Commun 388(4): 727-731.

Goeptar AR, Te Koppele JM, Lamme EK, Pique JM, and Vermeulen NP (1993) Cytochrome P450 2B1-mediated one-electron reduction of adriamycin: a study with rat liver microsomes and purified enzymes. Mol Pharmacol 44(6): 1267- 1277.

Gonzalez B, Akman S, Doroshow J, Rivera H, Kaplan WD, and Forrest GL (1995) Protection against daunorubicin cytotoxicity by expression of a cloned human carbonyl reductase cDNA in K562 leukemia cells. Cancer Res 55(2): 4646-4650.

Gonzalez-Covarrubias V, Ghosh D, Lakhman SS, Pendyala L, and Blanco JG (2007) A

163 functional genetic polymorphism on human carbonyl reductase 1 (CBR1 V88I) impacts on catalytic activity and NADPH binding affinity. Drug Metab Dispos 35(6): 973-980.

Goto S, Ihara Y, Urata Y, Izumi S, Abe K, Koji T, and Kondo T (2001) Doxorubicin- induced DNA intercalation and scavenging by nuclear glutathione S-transferase. FASEB J 15: 2702-2714.

Griffiths AJF, Miller JH, Suzuki DT, Lewontin RC, and Gelbart WM (2000) An introduction to genetic analysis, 7 th ed. WH Freeman and Company, New York.

Grimm C, Maser E, Mobus E, Klebe G, Reuter K, and Ficner R (2000) The crystal structure of 3 α-hydroxysteroid dehydrogenase/carbonyl reductase from Comamonas testosterone shows a novel oligomerization pattern within the short chain dehydrogenase/reductase family. J Biol Chem 275: 41333–41339.

Guano F, Pourquier P, Tinelli S, Binaschi M, Bigioni M (1999) Topoisomerase poisoning activity of novel disaccharide anthracyclines. Mol Pharmacol 56: 77-84.

Haimeur A, Conseil G, Deeley RG, and Cole SP (2004) The MRP-related and BCRP/ABCG2 multi-drug resistance proteins: biology, substrate specificity, and regulation. Curr Drug Metab 5(1): 21-53.

Hanioka N, Matsumoto K, Saito Y, and Narimatsu S (2010) Functional characterization of CYP2C8.13 and CYP2C8.14: catalytic activities towards paclitaxel. Basic Clin Pharmcol Toxicol [Epub ahead of print]

Hardy M, Rockenbauer A, Vasquez-Vivar J, Felix C, Lopez M, Srinivasan S, Avadhani N, Tordo P, and Kalyanaraman B (2007) Detection, characterization, and decay kinetics of ROS and thiyl adducts of mito-DEPMPO spin trap. Chem Res Toxicol 20(7): 1053-1060.

Harrison RL, and Jarvis DL (2006) Protein N-glycosylation in the baculovirus-insect cell expression system and engineering of insect cells to produce “mammalianized” recombinant glycoproteins. Adv Virus Res 68: 159-191.

Hasinoff BB, Schnabl KL, Marusak RA, Patel D, and Huebner E (2003) Dexrazoxane (ICRF-187) protects cardiac myocytes against doxorubicin by preventing damage to mitochondria. Cardiovasc Toxicol 3(2): 89-99.

Hasinoff BB and Herman EH (2007) Dexrazoxane: how it works in cardiac and tumor cells: is it a prodrug or is it a drug? Cardiovasc Toxicol 7: 140-144.

Hasinoff BB (2008) The use of dexrazoxane for the prevention of anthracycline extravasation injury. Expert Opin Investig Drugs 17(2): 217-223.

164 Hegedus DD, Pfeifer TA, Theilmann DA, Kennard ML, Gabathuler R, Jefferies WA, and Grigliatti TA (1999) Differences in the expression and localization of human melanotransferrin in lepidopteran and dipteran insect cell lines. Protein Expr Purif 15: 296-307.

Herman EH, Zhang J, Lipshultz SE, Rifai N, Chadwick D, Takeda K, Yu Z, and Ferrans VJ (1999) Correlation between serum levels of cardiac troponin-T and the severity of the chronic cardiomyopathy induced by doxorubicin. J Clin Oncol 17: 2237–2243.

Hobbs SK, Monsky WL, Yuan F, Roberts WG, Griffith L, Torchilin VP, and Jain RK (1998) Regulation of transport pathways in tumor vessels: role of tumor type and microenvironment. Proc Natl Acad Sci USA 95: 4607–4612.

Hoffmann F, and Maser E (2007) Carbonyl reductases and pluripotent hydroxysteroid dehydrogenases of the short-chain dehydrogenase/reductase superfamily. Drug Metab 39(1): 87-144.

Hollister J, Grabenhorst E, Nimitz M, Conradt H, and Jarvis DL (2002) Engineering the protein N-glycosylation pathway in insect cells for production of biantennary, complex N-glycans. Biochemistry 41(50): 15093-15104.

Homolya L, Varadi A, and Sarkadi B (2003) Multidrug resistance-associated proteins: export pumps for conjugates with glutathione, glucuronate or sulfate. Biofactors 17(1-4): 103-114.

Horenstein MS, Vander Heide RS, L’Ecuyer TJ (2000) Molecular basis of anthracycline- induced cardiotoxicity and its prevention. Mol Genet Metab 71(1-2): 436-444.

Hunault-Berger M, Milpied N, Bernard M, Jouet JP, Delain M, Desablens B, Sadoun A, Guilhot F, Casassus P, and Ifrah N (2001) Daunorubicin continuous infusion induces more toxicity than bolus infusion in acute lymphoblastic leukemia induction regimen: a randomized study. Leukemia 15: 898-902.

Imai T (2006) Human carboxylesterase isozymes: catalytic properties and rational drug design. Drug Metab Pharmacokinet 21(3): 173-185.

Ioannides C, and Lewis DFV (2004) Cytochromes P450 in the bioactivation of chemicals. Curr Top Med Chem 4: 1767-1788.

Ito H, Miller SC, Billingham ME, Akimoto H, Torti SV, Wade R, Gahlmann R, Lyons G, Kedes L, Torti FM (1990) Doxorubicin selectivity inhibits muscle gene expression in cardiac muscle cells in vivo and in vitro . Proc Natl Acad Sci USA 87(11) : 4275-4279.

Jackson AL, and Linsley PS (2004) Noise amidst the silence: off-target effects of siRNAs? Trends Genet 20: 521-524.

165

Jackson FL (2004) Human genetic variation and health: new assessment approaches based on ethnogenetic layering. Br Med Bull 69: 215-235.

Jackson JA, Reeves JP, Muntz KH, Kruk D, Prough RA, Willerson JT, and Buja LM (1984) Evaluation of free radical effects and catecholamine alterations in adriamycin cardiotoxicity. Am J Pathol 117(1) : 140-153.

Jacquet JM, Bressolle F, Galtier M, Bourrier M, Donadio D, Jourdan J, and Rossi JF (1990) Doxorubicin and doxorubicinol: intra- and inter-individual variations of pharmacokinetic parameters. Cancer Chemother Pharmacol 27: 219-225.

Jemal A, Siegel R, Ward E, Murray T, Xu J, Smigal C, and Thun MJ (2006) Cancer statistics, 2006. CA Cancer J Clin 56: 106–130.

Jez JM, Flynn TG, Penning TM (1997) A new nomenclature for the aldo-keto reductase superfamily. Biochem Pharmacol 54(6): 639-647.

Ji C, Amarnath V, Pietenpol JA, and Marnett LJ (2001) 4-hydroxynonenal induces apoptosis via caspase-3 activation and cytochrome c release. Chem Res Toxicol 14(8): 1090–1096.

Ji Q, Chang L, VanDenBerg D, Stanczyk FZ, and Stolz A (2003) Selective reduction of AKR1C2 in prostate cancer and its role in DHT metabolism. Prostate 54: 275– 289.

Ji Q, Chang L, Stanczyk FZ, Ookhtens M, Sherrod A, and Stolz A (2007) Impaired dihydrotestosterone catabolism in human prostate cancer: critical role of AKR1C2 as a pre-receptor regulator of androgen receptor signaling. Cancer Res 67: 1361– 1369.

Jin Y and Penning TM (2007) Aldo-keto reductases and bioactivation/detoxification. Annu Rev Pharmacol Toxicol 47: 263-292.

Jones RL and Smith IE (2004) Efficacy and safety of trastuzumab. Expert Opin Drug Saf 3(4): 317-327.

Jornvall H, Persson B, Krook M, Atrian S, Gonzalez-Duarte R, Jeffery J, and Ghosh D (1995) Short-chain dehydrogenases/reductases (SDR). Biochemistry 34: 6003- 6013.

Jung K and Reszka R (2001) Mitochondria as subcellular targets for clinically useful anthracyclines. Adv Drug Deliv Rev 49(1-2): 87-105.

166 Kalender S, Kalender Y, Ates A, Yel M, Olcay E, and Candan S (2002) Protective role of antioxidant vitamin E and catechin on idarubicin-induced cardiotoxicity in rats. Braz J Med Biol Res 35(11): 1379-1387.

Kallberg Y, Oppermann U, and Persson B (2010) Classification of the short-chain dehydrogenase/reductase superfamily using hidden Markov models. FEBS J 277(10): 2375-2386.

Karlhuber GM, H.C. Bauer HC, and Eckl PM (1997) Cytotoxic and genotoxic effects of 4-hydroxynonenal in cerebral endothelial cells. Mutat Res 381 : 209–216.

Kassner N, Huse K, Martin HJ, Godtel-Armbrust U, Metzger A, Meineke I, Brockmoller J, Klein K, Zanger UM, Maser E, and Wojnowski L (2008) Carbonyl reductase 1 is a predominant doxorubicin reductase in the human liver. Drug Metab Dispos 36(10): 2113-2120.

Kavanagh KL, Jornvall H, Persson B, and Oppermann U (2008) Functional and structural diversity within the short-chain dehydrogenase/reductase (SDR) superfamily. Cell Mol Life Sci 65: 3895–3906.

Kelly VP, Sherratt PJ, Crouch DH, and Hayes JD (2002) Novel homodimeric and heterodimeric rat gamma-hydroxybutyrate synthases that associate with the Golgi apparatus define a distinct subclass of aldo-keto reductase 7 family proteins. Biochem J 366(Pt 3): 847-861.

Khan N, and Swartz H (2002) Measurements in vivo of parameters pertinent to ROS/RNS using EPR spectroscopy. Mol Cell Biochem 234/235: 341-357.

Klaasen C (1995) Casarett & Doull’s Toxicology. The basic science of poisons, 5 th ed. McGraw-Hill Companies Inc., New York.

Knights KM (1998) Role of hepatic fatty acid:coenzyme A ligases in the metabolism of xenobiotic carboxylic acids. Clin Exp Pharmacol Physiol 25(10): 776-782.

Komoto J, Yamada T, Watanabe K and Takusagawa F (2004) Crystal structure of human prostaglandin F synthase (AKR1C3). Biochemistry 43(8): 2188-2198.

Krischer JP, Epstein S, Cuthbertson DD, Goorin AM, Epstein ML, and Lipshultz SE (1997) Clinical cardiotoxicity following anthracycline treatment for childhood cancer: the Pediatric Oncology Group experience. J Clin Oncol 15: 1544–1552.

Krueger SK, Vandyke JE, Williams DE, and Hines RN (2006) The role of flavin- containing monooxygenase (FMO) in the metabolism of tamoxifen and other tertiary amines. Drug Metab Rev 38(1-2): 139-147.

167 Kruglyak L, and Nickerson DA (2001) Variation is the spice of life. Nat Genet 27: 234– 236.

Kruman I, Bruce-Keller AJ, Bredesen D, Waeg G, and Mattson MP (1997) Evidence that 4-hydroxynonenal mediates oxidative stress-induced neuronal apoptosis. J Neurosci 17: 5089–5100.

Kume T, Iwasa H, Shiraishi H, Yokoi T, Nagashima K, Otsuka M, Terada T, Takagi T, Hara A and Kamataki T (1999) Characterization of a novel variant (S145C/L311V) of 3 α-hydroxysteroild/dihydrodiol dehydrogenase in human liver. Pharmacogenetics 9(6): 763-771.

Lakhman SS, Ghosh D, and Blanco JG (2005) Functional significance of natural allelic variant of human carbonyl reductase 3 (CBR3). Drug Metab Dispos 33: 254- 257.

Lang-Bicudo L, Eduardo Fde P, Eduardo Cde P, and Zezell DM (2008) LED phototherapy to prevent mucositis: a case report. Photomed Laser Surg 26(6): 609-613.

Launchbury AP and Habboubi N (1993) Epirubicin and doxorubicin: a comparison of their characteristics, therapeutic activity and toxicity . Cancer Treat Rev 19(3): 197-228. L'Ecuyer T, Horenstein MS, Thomas R, Vander Heide R (2001) Anthracycline-induced cardiac injury using a cardiac cell line: potential for gene therapy studies . Mol Genet Metab 74: 370-379.

Le Guellec C, Lacarelle B, Catalin J, and Durand A (1993) Inhibitory effects of anticancer drugs on dextromethorphan-O-demethylase activity in human liver microsomes. Cancer Chemother Pharmacol 32(6): 491-495.

Lee CR, Goldstein JA, and Pieper JA (2002) Cytochrome P450 2C9 polymorphisms: a comprehensive review of the in-vitro and human data. Pharmacogenetics 12: 251- 263.

Lee TY, Huang HD, Hung JH, Huang HY, Yang YS, and Wang TH (2006) dbPTM: an information repository of protein post-translational modification. Nucleic Acids Res 34(Database issue): D622-D627.

Lefrak EA, Pitha J, Rosenheim S, and Gottlieb JA (1973) A clinicopathologic analysis of adriamycin cardiotoxicity. Cancer 32 : 302-314.

Leonard RC, Williams S, Tulpule A, Levine AM, and Oliveros S (2009) Improving the therapeutic index of anthracycline chemotherapy: focus on liposomal doxorubicin (Myocet). Breast 18(4): 218-224.

168 Li D, Hinshelwood A, Gardner R, McGarvie G, and Ellis EM (2006) Mouse aldo-keto reductase AKR7A5 protects V79 cells against 4-hydroxynonenal-induced apoptosis. Toxicology 226(2-3): 172–180.

Li L, Hamilton RF, Kirichenko A, and Holian A (1996) 4-Hydroxynonenal-induced cell death in murine alveolar macrophages. Toxicol Appl Pharmacol 139(1): 135– 143.

Li W, Xu B, Xu J, and Wu XL (2009) Procyanidins produce significant attenuation of doxorubicin-induced cardiotoxicity via suppression of oxidative stress. Basic Clin Pharmacol Toxicol 104(3): 192-197.

Licata S, Saponiero A, Mordente A and Minotti G (2000) Doxorubicin metabolism and toxicity in human myocardium: role of cytoplasmic deglycosidation and carbonyl reduction. Chem Res Toxicol 13: 414-420.

Lipp HP, and Bokemeyer C (1999) Anthracyclines and other intercalating agents, in Anticancer drug toxicity: prevention, management and clinical pharmacokinetics (Lipp HP ed.) pp 81-113, Marcel Dekker, New York.

Lipschultz SE, Alvarez JA, and Scully RE (2008) Anthracycline associated cardiotoxicity in survivors of childhood cancer. Heart 94: 525-533.

Liu KJ, Miyake M, Panz T, and Swartz H (1999) Evaluation of DEPMPO as a spin trapping agent in biological systems. Free Radic Biol Med 26(5-6): 714-721.

Lown JW (1993) Anthracycline and anthraquinone anticancer agents: current status and recent developments. Pharmacol Ther 60(2): 185-214.

Majem M, Salazar R, Garcia M, Clopes A, Munoz C, Pardo B, Martinez M, Cuadra C, Montes A, Mis R, Savulsky C, Ramon Germa-Lluch J (2009) A phase I clinical and pharmacokinetic study (LIPOTEC-GP PHARM/DOXO 01) of a new liposomal doxorubicin given as 3-week schedule in patients with solid tumors. J Liposome Res 19(4): 261-266.

Marrett LD, De P, Airia P, and Dryer D (2008) Cancer in Canada 2008. CMAJ 179(11): 1163-1170.

Martin HJ, Breyer-Pfaff U, Wsol V, Venz S, Block S, and Maser E (2006) Purification and characterization of AKR1B10 from human liver: role in carbonyl reduction of xenobiotics 34(3): 464-470.

Mascarenhas D, Mettler IJ, Pierce DA, Lowe HW (1990) Intron-mediated enhancement of heterologous gene expression in maize. Plant Mol Biol 15(6): 913-920.

169 Maser E (1995) Xenobiotic carbonyl reduction and physiological steroid oxidoreduction: the pluripotency of several hydroxysteroid dehydrogenases. Biochem Pharmacol 49: 421-440.

Matsunaga T, Shintani S, and Hara A (2006) Multiplicity of mammalian reductases for xenobiotic carbonyl compounds. Drug Metab Pharmacokinet 21: 1–18.

Matsuura K, Hara A, Deyashiki Y, Iwasa H, Kume T, Ishikura S, Shiraishi H and Katagiri Y (1998) Roles of the C-terminal domains of human dihydrodiol dehydrogenase isoforms in the binding of substrates and modulator: probing with chimaeric enzymes. Biochem J 336(2): 429-436.

Maxwell A, Costenaro L, Mitelheiser S, and Bates AD (2005) Coupling ATP hydrolysis to DNA strand passage in type IIA DNA topoisomerases. Biochem Soc Trans 33: 1460-1464.

Menna P, Recalcati S, Cairo G, and Minotti G (2007) An introduction to the metabolic determinants of anthracycline cardiotoxicity. Cardiovasc Toxicol 7(2): 80-85.

Menna P, Salvatorelli E, and Minotti G (2010) Anthracycline degradation in cardiomyocytes: a journey to oxidative survival. Chem Res Toxicol 23(1): 6-10.

Mindich RD and Penning TM (2009) Aldo-keto reductase superfamily: genomics and annotation. Hum Genomics 3(4): 362-370.

Miners JO, and Birkett DJ (1998) Cytochrome P4502C9: an enzyme of major importance in human drug metabolism. Br J Clin Pharmacol 45:525–538.

Minotti G, Cavaliere AF, Mordente A, Rossi M, Schiavello R, Zamparelli R, and Possati G (1995) Secondary alcohol meabolites mediate iron delocalization in cytosolic fractions of myocardial biopsies exposed to anticancer anthracyclines: novel linkage between anthracycline metabolism and iron-induced cardiotoxicity. J Clin Invest 95(4): 1595-1605.

Minotti G, Recalcati S, Mordente A, Liberi G, Calafiore AM, Mancuso C, Preziosi P, and Cairo G (1998) The secondary alcohol metabolite of doxorubicin irreversibly inactivates aconitase/iron regulatory protein-1 in cytosolic fractions from human myocardium. FASEB J 12(7): 541-542.

Minotti G, Cairo G, and Monti E (1999) Role of iron in anthracycline cardiotoxicity: new tunes for an old song? FASEB J 13(2): 199-212.

Minotti G, Ronchi R, Salvatorelli E, Menna P, and Cairo G (2001) Doxorubicin irreversibly inactivates iron regulatory proteins 1 and 2 in cardiomyocytes: evidence for distinct metabolic pathways and implications for iron-mediated cardiotoxicity of antitumor therapy. Cancer Res 61(23): 8422–8428.

170

Minotti G, Menna P, Salvatorelli E, Cairo G, and Gianni L (2004) Anthracyclines: molecular advances and pharmacologic developments in antitumor activity and cardiotoxicity. Pharmacol Rev 56(2): 185-229.

Mitrius JC and Vogel SM (1990) Doxorubicin-induced automaticity in cultured chick heart cell aggregates. Cancer Res 50(14): 4209-4215.

Miura T, Nishinaka T, and Terada T (2008) Different functions between human monomeric carbonyl reductase 3 and carbonyl reductase 1. Mol Cell Biochem 315: 113-121.

Mojovic M, Vuletic M, and Bacic GG (2005) Detection of oxygen-centered radicals using EPR spin-trap DEPMPO: the effect of oxygen. Ann N Y Acad Sci 1048: 471-475.

Mojzisova G, Sarissky M, Mirossay L, Martinka P, Mojzis J (2009) Effect of flavonoids on daunorubicin-induced toxicity in H9c2 cardiomyoblasts. Phytother Res 23(1): 136-139.

Mordente A, Meucci E, Martorana GE, Giardina B, and Minotti G (2001) Human heart cytosolic reductases and anthracycline cardiotoxicity. IUBMB Life 52: 83-88.

Moroder L, and Budisa N (2010) Synthetic biology of protein folding. Chemphyschem 11(6): 1181-1187.

Mross K, Maessen P, van der Vijgh WJ, Gall H, Boven E, and Pinedo HM (1988) Pharmacokinetics and metabolism of epidoxorubicin and doxorubicin in humans. J Clin Oncol 6(3): 517-526.

Muindi JR, Sinha BK, Gianni L, and Myers CE (1984) Hydroxyl radical production and DNA damage induced by anthracycline-iron complex. FEBS Lett 172: 226–230.

Myers CE, McGuire WP, Liss RH, Ifrim I, Grotzinger K, and Young RC (1977) Adriamycin: the role of lipid peroxidation in cardiac toxicity and tumor response. Science 197(4299): 165-167.

Natale RB, Cody RL, Simon MS, and Wheeler RH (1993) An in vivo and in vitro trial of aclarubicin in metastatic breast cancer: a novel approach to the study of analogs. Cancer Chemother Pharmacol 31 : 485-488.

Nebert DW, and Russell DW (2002) Clinical importance of the cytochromes P450. Lancet 360: 1155–1162.

Nousiainen T, Jantunen E, Vanninen E, and Hartikainen J (2002) Early decline in left ventricular ejection fraction predicts doxorubicin cardiotoxicity in lymphoma

171 patients. Br J Cancer 86(11): 1697-1700.

Nowatzke W, and Woolf E (2007) Best practices during bioanalytical method validation for the characterization of assay reagents and the evaluation of analyte stability in assay standards, quality controls, and study samples. Aaps J 9: E117-E122.

O’Connor T, Ireland LS, Harrison DJ, and Hayes JD (1999) Major differences exist in the function and tissue-specific expression of human aflatoxin B1 aldehyde reductase and the principal human aldo-keto reductase AKR1 family members. Biochem J 343(2): 487-504.

Ohara H, Miyabe Y, Deyashiki Y, Matsuura K and Hara A (1995) Reduction of drug ketones by dihydrodiol dehydrogenases, carbonyl reductase and aldehyde reductase of human liver. Biochem Pharmacol 50: 221–227.

Okamura N, Sakaeda T, and Okumura K (2004) Pharmacogenomics of MDR and MRP subfamilies. Personalized Med 1(1): 85-104.

Oki T, Komiyama T, Tone H, Inui T, Takeuchi T, and Umezawa H (1977) Reductive cleavage of anthracycline glycosides by microsomal NADPH-cytochrome C reductase. J Antibiot (Tokyo) 30(7): 613-615.

Olson LE, Bedja D, Alvery SJ, Cardounel AJ, Gabrielson KL, and Reeves RH (2003) Protection from doxorubicin-induced cardiac toxicity in mice with a null allele of carbonyl reductase 1. Cancer Res 63(20): 6602-6606.

Olson RD and Mushlin PS (1990) Doxorubicin cardiotoxicity: analysis of prevailing hypotheses. Faseb J 4: 3076-3086.

Oppermann UC (2007) Carbonyl reductases: the complex relationships of mammalian carbonyl- and quinone-reducing enzymes and their role in physiology. Annu Rev Pharmacol Toxicol 47: 293-322.

Outomuro D, Grana DR, Azzato F, and Milei J (2007) Adriamycin-induced myocardial toxicity: new solutions for an old problem? Int J Cardiol 117(1): 6-15.

Pan SS, Pedersen L, and Bachur NR (1981) Comparative flavoprotein catalysis of anthracycline antibiotic: reductive cleavage and oxygen consumption. Mol Pharmacol 19(1): 184-186.

Pea F, Russo D, Michieli M, Baraldo M, Ermacora A, Damiani D, Baccarani M, and Furlanut M (2000) Liposomal daunorubicin plasmatic and renal disposition in patients with acute leukemia. Cancer Chemother Pharmacol 46(4): 279-286.

172 Penning TM (2003) Introduction and overview of the aldo-keto reductase superfamily, in Aldo-keto reductases and toxicant metabolism (Penning TM, Petrash JM eds.) pp 3-22, American Chemical Society, Washington, DC.

Penning TM and Drury JE (2007) Human aldo-keto reductases: function, gene regulation, and single nucleotide polymorphisms. Arch Biochem Biophys 464(2): 241-250.

Penning TM, Jin Y, Rizner TL, and Bauman DR (2008) Pre-receptor regulation of the androgen receptor. Mol Cell Endocrinol 281: 1–8.

Perez-Tomas R (2006) Multidrug resistance: retrospect and prospects in anti-cancer drug treatment. Curr Med Chem 13(16): 1859-1876.

Persson B, Krook M, and Jornvall (1991) Characteristics of short-chain alcohol dehydrogenases and related enzymes. Eur J Biochem 200: 537-543.

Persson B, Kallberg Y, Bray JE, Bruford E, Dellaporta SL, Favia AD, Duarte RG, Jörnvall H, Kavanagh KL, Kedishvili N, Kisiela M, Maser E, Mindnich R, Orchard S, Penning TM, Thornton JM, Adamski J, and Oppermann U (2009) The SDR (short-chain dehydrogenase/reductase and related enzymes) nomenclature initiative. Chem Biol Interact 178(1-3): 94-98.

Picklo MJ, Montine T,Amarnath V, and Neely M (2002) Carbonyl toxicology and Alzheimer's disease. Toxicol Appl Pharmacol 184: 187–197.

Pilka ES, Rojkova A, Kavanagh KL, Niesen F, Sundstrom M, Arrowsmith C, Weigelt J, Edwards A, Opperman U (2006) Crystal structure of human carbonyl reductase 3 complexed with NADP +, RSCB Protein Data Bank ID: 2HRB, http://www.sgc.ox.ac.uk/

Piscitelli SC, Rodvold KA, Rushing DA, Tewksbury DA (1993) Pharmacokinetics and pharmacodynamics of doxorubicin in patients with small cell lung cancer. Clin Pharmacol Ther 53(5): 555-561.

Plebuch M, Soldan M, Hungerer C, Koch L and Maser E (2007) Increased resistance of tumor cells to daunorubicin after transfection of cDNAs coding for anthracycline inactivating enzymes. Cancer Lett 255(1): 49-56.

Praga C, Beretta G, Vigo PL, Lenaz GR, Pollini C, Bonadonna G, Canetta R, Castellani R, Villa E, Gallagher CG, von Melchner H, Hayat M, Ribaud P, De Wasch G, Mattsson W, Heinz R, Waldner R, Kolaric K, Buehner R, Ten Bokkel-Huyninck W, Perevodchikova NI, Manziuk LA, Senn HJ, and Mayr AC (1979) Adriamycin cardiotoxicity: a survey of 1273 patients. Cancer Treat Rep 63(5): 827-834.

Qiagen (2003) The QIAexpressionist: a handbook for high-level expression and purification of 6x-His tagged proteins, 5 th ed. (http://www.qiagen.com/literature/handbooks/PDF/Protein/Expression/QXP_QIA

173 expressionist/1024473_QXPHB_0603.pdf).

Rajagopalan S, Politi PM, Sinha BK, and Myers CE (1988) Adriamycin-induced free radical formation in the perfused rat heart: implications for cardiotoxicity. Cancer Res 48(17): 4766-4769.

Riddick DS, Lee C, Ramji S, Chinje EC, Cowen RL, Williams KJ, Patterson AV, Stratford IJ, Morrow CS, Townsend AJ, Jounaidi Y, Chen CS, Su T, Lu H, Schwartz PS, and Waxman DJ (2005) Cancer chemotherapy and drug metabolism. Drug Metab Dispos 33(8): 1083-1096.

Robert J and Gianni L (1993) Pharmacokinetics and metabolism of anthracyclines. Cancer Surv 17: 219-252.

Roca J (2009) Topoisomerase II: a fitted mechanism for the chromatin landscape. Nucleic Acids Res 37(3): 721-730.

Rosenblum JS, Gilula NB, and Lerner RA (1996) On signal sequence polymorphisms and diseases of distribution. Proc Natl Acad Sci USA 93: 4471–4473.

Rossi F, Filippelli W, Russo S, Filippelli A, Berrino L (1994) Cardiotoxicity of doxorubicin: effects of drugs inhibiting the release of vasoactive substances. Pharmacol Toxicol 75(2): 99-107.

Rubin EH and Hait WN (2003) Anthracyclines and DNA Intercalators/ Epipodophyllotoxins/DNA Topoisomerases , in Cancer Medicine, 6 th ed. (Bast RC Jr, Kufe DW, Pollock RE, Weichselbaum RR, Holland JF, Frei E, and Gansler TS eds.) Chapter 49, BC Decker, Hamilton, Ontario.

Ruiz-Ruiz C, Robledo G, Cano E, Redondo JM, and Lopez-Rivas A (2003) Characterization of p53-mediated up-regulation of CD95 gene expression upon genotoxic treatment in human breast tumor cells. J Biol Chem 278: 31667–31675.

Ryberg M, Nielsen D, Skovsgaard T, Hansen J, Jensen BV, and Dombernowsky P (1998) Epirubicin cardiotoxicity: an analysis of 469 patients with metastatic breast cancer. J Clin Oncol 16 : 3502 -3508.

Safra T, Muggia F, Jeffers S, Tsao-Wei DD, Groshen S, Lyass O, Henderson R, Berry G, and Gabizon A (2000) Pegylated liposomal doxorubicin (doxil): reduced clinical cardiotoxicity in patients reaching or exceeding cumulative doses of 500 mg/m 2. Ann Oncol 11(8): 1029-1033.

Safra T (2003) Cardiac safety of liposomal anthracyclines. Oncologist 8 Suppl 2: 17-24.

174 Sahdev S, Khattar SK, and Saini KS (2008) Production of active eukaryotic proteins through bacterial expression systems: a review of the existing biotechnology strategies. Mol Cell Biochem 307(1-2): 249-264.

Sanchez Medina J, Gonzalez-Ramella O, Gallegos-Castorena S (2010) The effect of dexrazoxane for clinical and subclinical cardiotoxicity in children with acute myeloid leukemia. J Pediatr Hematol Oncol 32(4): 294-297.

Schlegel BP, Ratnam K and Penning TM (1998) Retention of NADPH-linked quinine reductase activity in an aldo-keto reductase following mutation of the catalytic tyrosine. Biochemistry 37(31): 11003-11011.

Schroeder PE and Hasinoff BB (2005) Metabolism of the one-ring open metabolites of the cardioprotective drug dexrazoxane to its active metal-chelating form in the rat. Drug Metab Dispos 33(9): 1367-1372.

Schwartz RG, McKenzie WB, Alexander J, Sager P, D'Souza A, Manatunga A, Schwartz PE, Berger HJ, Setaro J, Surkin L, Wackers F JTh, and Zaret BL (1987) Congestive heart failure and left ventricular dysfunction complicating doxorubicin therapy: seven-year experience using serial radionuclide angiocardiography. Am J Med 82(6): 1109-1118.

Seidman A, Hudis C, Pierri MK, Shak S, Paton V, Ashby M, Murphy M, Stewart SJ, and Keefe D (2002) Cardiac dysfunction in the trastuzumab clinical trials experience . J Clin Oncol 20(5): 1215-1221.

Seo J, and Lee KJ (2004) Post-translational modifications and their biological functions: proteomic analysis and systematic approaches. J Biochem Mol Biol 37(1): 35-44.

Shah RR (2005) Pharmacogenetics in drug regulation: promise, potential and pitfalls. Phil Trans R Soc 360(1460): 1617-1638.

Shah VP, Midha KK, Findlay JW, Hill HM, Hulse JD, McGilveray IJ, McKay G, Miller KJ, Patnaik RN, Powell ML, Tonelli A, Viswanathan CT, and Yacobi A (2000) Bioanalytical method validation--a revisit with a decade of progress. Pharm Res 17: 1551-1557.

Shan K, Lincoff AM, and Young JB (1996) Anthracycline-induced cardiotoxicity. Ann Intern Med 125(1): 47-58.

Shastry BS (2002) SNP alleles in human disease and evolution. J Hum Genet 47(11): 561-566.

Shimoda-Matsubayashi S, Matsumine H, Kobayashi T, Nakagawa-Hattori Y, Shimizu Y, and Mizuno Y (1996) Structural dimorphism in the mitochondrial targeting sequence in the human manganese superoxide dismutase gene. A predictive

175 evidence for conformational change to influence mitochondrial transport and a study of allelic association in Parkinson's disease. Biochem Biophys Res Commun 226: 561–565.

Simbre VC, Duffy SA, Dadlani GH, Miller TL, and Lipshultz SE (2005) Cardiotoxicity of cancer chemotherapy: implications for children. Pediatr Drugs 7: 187–202.

Simunek T, Sterba M, Popelova O, Adamcova M, Hrdina R, and Gersl V (2009) Anthracycline-induced cardiotoxicity: overview of studies examining the roles of oxidative stress and free cellular iron. Pharmacol Rep 61(1): 154-171.

Singal PK, Iliskovic N, Li T, and Kumar D (1997) Adriamycin cardiomyopathy: pathophysiology and prevention. FASEB J 11(12): 931-936.

Singal PK, and Iliskovic N (1998) Doxorubicin-induced cardiomyopathy. N Engl J Med 339: 900–905.

Sinha BK, Trush MA, Kennedy KA, and Mimnaugh EG (1984) Enzymatic activation and binding of adriamycin to nuclear DNA. Cancer Res 44: 2892–2896.

Spallarossa P, Garibaldi S, Altieri P, Fabbi P, Manca V, Nasti S, Rossettin P, Ghigliotti G, Ballestrero A, Patrone F, Barsotti A, Brunelli C (2004) Carvedilol prevents doxorubicin-induced free radical release and apoptosis in cardiomyocytes in vitro . J Mol Cell Cardiol 37: 837-846.

Speth PA, van Hoesel QG, and Haanen C (1988) Clinical pharmacokinetics of doxorubicin. Clin Pharmacokinet 15(1): 15-31.

Speyer J and Wasserheit C (1998) Strategies for reduction of anthracycline cardiac toxicity. Semin Oncol 25: 525-537.

Steinherz LJ, Steinherz PG, Tan CT, Heller G, and Murphy ML (1991) Cardiac toxicity 4 to 20 years after completing anthracycline therapy. JAMA 266(12): 1672-1677.

Sutton A, Khoury H, Prip-Buus C, Cepanec C, Pessayre D, and Degoul F (2003) The Ala16Val genetic dimorphism modulates the import of human manganese superoxide dismutase into rat liver mitochondria. Pharmacogenetics 13: 145–157.

Swartz HM, Khan N, and Khramtsov VV (2007) Use of electron paramagnetic resonance spectroscopy to evaluate the redox state in vivo . Antioxid Redox Signal 9(10): 1757-1771.

Takahashi RH, Bains OS, Pfeifer TA, Grigliatti TA, Reid RE and Riggs KW (2008) Aldo-keto reductase 1C2 fails to metabolize doxorubicin and daunorubicin in vitro. Drug Metab Dispos 36(6): 991-994.

Takahashi RH, Grigliatti TA, Reid RE, and Riggs KW (2009) The effect of allelic

176 variation of aldo-keto reductase 1C2 on the in vitro metabolism of dihydrotestosterone. J Pharamcol Exp Ther 329(3): 1032-1039.

Takanashi S and Bachur NR (1976) Adriamycin metabolism in man. Evidence from urinary metabolites. Drug Metab Dispos 4(1): 79-87.

Tan AR and Swain SM (2003) Adjuvant chemotherapy for breast cancer: an update. Semin Oncol 30: 54-64.

Tan C, Tasaka H, Yu KP, Murphy ML, and Karnofsky DA (1967) Daunomycin, an antitumor antibiotic, in the treatment of neoplastic disease. Clinical evaluation with special reference to childhood leukemia. Cancer 20: 333–353.

Terpe K (2006) Overview of bacterial expression systems for heterologous protein production: from molecular and biochemical fundamentals to commercial systems. Appl Microbiol Biotechnol 72(2): 211-222.

Tewey KM, Rowe TC, Yang L, Halligan BD, and Liu LF. (1984) Adriamycin induced DNA damage mediated by mammalian DNA topoisomerase II. Science 226: 466- 468.

Tokarska-Schlattner M, Zaugg M, Zuppinger C, Wallimann T, Schlattner U (2006) New insights into doxorubicin-induced cardiotoxicity: the critical role of cellular energetics. J Mol Cell Cardiol 41(3): 389-405.

Vaage J, Barbera-Guillem E, Abra R, Huang A, and Working P (1994) Tissue distribution andtherapeutic effect of intravenous free or encapsulated liposomal doxorubicin on human prostate carcinoma xenografts. Cancer 73: 1478–1484.

Vail DM, Amantea MA, Colbern GT, Martin FJ, Hilger RA, and Working PK (2004) Pegylated liposomal doxorubicin: proof of principle using preclinical animal models and pharmacokinetic studies. Semin Oncol 31:16–35.

Valgimigli L, Pedulli GF, and Paolini M (2001) Measurement of oxidative stress by EPR radical-probe technique. Free Radic Biol Med 31(6): 708-716.

Valko M, Morris H, and Cronin MT (2005) Metals, toxicity and oxidative stress. Curr Med Chem 12(10): 1161-1208.

Valko M, Rhodes CJ, Moncol J, Izakovic M, and Mazur M (2006) Free radicals, metals and antioxidants in oxidative stress-induced cancer. Chem Biol Interact 160(1): 1- 40.

Vasquez-Vivar J, Martasek P, Hogg N, Masters BS, Pritchard KA Jr, and Kalyanaraman B (1997) Endothelial nitric oxide synthase-dependent superoxide generation from adriamycin. Biochemistry 36: 11293-11297.

177

Veitch ZW, Guo B, Hembruff SL, Bewick AJ, Heibein AD, Eng J, Cull S, Maclean DA, and Parissenti AM (2009) Induction of 1C aldoketoreductases and other drug dose-dependent genes upon acquisition of anthracycline resistance. Pharmacogenet Genomics 19(6): 477-488.

Vergely C, Delemasure S, Cottin Y, and Rochette L (2007) Preventing the cardiotoxic effects of anthracyclines: from basic concepts to clinical data. Heart Metab 35: 1- 7.

Vichi P and Tritton TR (1992) Adriamycin: protection from cell death by removal of extracellular drug. Cancer Res 52(15): 4135-4138.

Von Hoff DD, Rozencweig M, Layard M, Slavik M, and Muggia FM (1977) Daunomycin-induced cardiotoxicity in children and adults: a review of 110 cases. Am J Med 62(2): 200-208.

Von Hoff DD, Layard MW, Basa P, Davis HL Jr, Von Hoff AL, Rozencweig M, Muggia FM (1979) Risk factors for doxorubicin-induced congestive heart failure. Ann Intern Med 91(5): 710-717.

Wallace KB. Doxorubicin-induced cardiac mitochondrionopathy . Pharmacol Toxicol 2003; 93(3): 105-115.

Wallace KB and Starkov AA (2000) Mitochondrial targets of drug toxicity. Annu Rev Pharmcol Toxicol 40: 353-388.

Waugh DS (2005) Making the most of affinity tags. Trends Biotechnol 23(6): 316-320.

Wilstermann AM and Osheroff N (2003) Stabilization of eukaryotic topoisomerase II- DNA cleavage complexes. Curr Top Med Chem 3: 321-338.

Wirth H, and Wermuth B (1992) Immunohistochemical localization of carbonyl reductase in human tissues. J Histochem Cytochem 40: 1857–1863.

Wold LE, Aberle NS, and Ren J (2005) Doxorubicin induces cardiomyocyte dysfunction via a p38 MAP kinase-dependent oxidative stress mechanism. Cancer Detect Prev 29(3): 294-299.

Wojtacki J, Lewicka-Nowak E, and Lesniewski-Kmak K (2000) Anthracycline-induced cardiotoxicity: clinical course, risk factors, pathogenesis, detection and prevention--review of the literature. Med Sci Monit 6(2): 411-420.

Wouters KA, Kremer LC, Miller TL, Herman EH, and Lipshultz SE (2005) Protecting against anthracycline-induced myocardial damage: a review of the most promising strategies. Br J Haematol 131: 561–578.

178

Xu X, Persson HL, and Richardson DR (2005) Molecular pharmacology of the interaction of anthracyclines with iron. Mol Pharmacol 68(2): 261-271.

Xu MF, Tang PL, Qian ZM, and Ashraf M (2001) Effects by doxorubicin on the myocardium are mediated by oxygen free radicals. Life Sci 68(8): 889-901.

Xu C, Li CY, and Kong AN (2005) Induction of phase I, II, III drug metabolism/transport by xenobiotics. Arch Pharmacal Research 28(3): 249-268.

Yin SJ (1994) Alcohol dehydrogenase: enzymology and metabolism. Alcohol Alcohol Suppl 2: 113-119.

Zerbs S, Frank AM, and Collart FR (2009) Bacterial systems for production of heterologous proteins. Methods Enzymol 463: 149-168.

Zhang J, and Blanco JG (2009) Identification of the promoter of human carbonyl reductase 3 (CBR3) and impact of common promoter polymorphisms on hepatic CBR3 mRNA expression. Pharm Res 26: 2209-2215.

Zhang YW, Shi J, Li YJ, and Wei L (2009) Cardiomyocyte death in doxorubicin-induced cardiotoxicity. Arch Immunol Ther Exp (Warsz) 57(6): 435-445.

Zhou S, and Chan E (2005) Drug bioactivation, covalent binding to target proteins and toxicity relevance. Drug Metab Rev 1: 41-213.

Zhu Y, and Hein DW (2008) Functional effects of single nucleotide polymorphisms in the coding region of human N-acetyltransferase 1. Pharmacogenomics J 8(5): 339-348.

Zurita AJ, Diestra JE, Condom E, Garcia del Muro X, Scheffer GL Scheper RJ, Perez J, Germa-Lluch JR, and Izquierdo MA (2003) Lung resistance-related protein as a predictor of clinical outcome in advanced testicular germ-cell tumors. Br J Cancer 88: 879-886.

179 APPENDICES

Supplemental Table 1 —Primers for creating non-synonymous single nucleotide polymorphic variants of human AKRs using site-directed mutagenesis.

GENES VARIANTS a FORWARD PRIMER (5'3') REVERSE PRIMER (5'3') I15F CAAGATGCCCTTCCTGGGGTTGG CCAACCCCAGGAAGGGCATCTTG [A T] H42L CGGGTACCGCCTCATCGACTGTGC GGGCACAGTCGATGAGGCGGTA [A T] CC CCCG L73V GCGTGAGGAGGTCTTCATCGTCAG GCTGACGATGAAGACCTCCTCAC [C G] C GC AKR1B1 K90E GGGCCTGGTGGAAGGAGCCTGCC GGCAGGCTCCTTCCACCAGGCCC [A G] G204S GCCAGTCCAAAAGCATCGTGGTGA GTCACCACGATGCTTTTGGACTG [G A] C GC T288I CCAGGATATGACCATCTTACTCAGC GTAGCTGAGTAAGATGGTCATAT [C T] TAC CCTGG P87S CACTTTCTTTGAGAGATCCCTTGTG CTTTCCTCACAAGGGATCTCTCA [C T] AGGAAAG AAGAAAGTG M286T GTGATGAGGAGACGGCAACCATAC GAGTATGGTTGCCGTCTCCTCAT AKR1B10 [T C] TC CAC N313D GACTATCCCTTCGATGCAGAATATT CAATATTCTGCATCGAAGGGATA [A G] G GTC R170H CCAACTTCAACCACAGGCAGCTGG CTCCAGCTGCCTGTGGTTGAAGT [G A] AG TGG AKR1C1 Q172L CTTCAACCGCAGGCTGCTGGAGAT GGATCATCTCCAGCAGCCTGCGG [A T] GATCC TTGAAG F46Y CAATAGAAGCCGGGTACCACCATA GAATCAATATGGTGGTACCCGGC AKR1C2 [T A] TTTGATTC TTCTATTG H5Q GGAAGAATGGATTCCAAACAGCAG CATTTAGCTTTACACACTGCTGTT [C G] TGTGTAAAGCTAAATG TGGAATCCATTCTTCC R66Q GGTTGGACTGGCCATCCAAAGCAA CATCTGCAATCTTGCTTTGGATG [G A] GATTGCAGATG GCCAGTCCAACC A106T GGAAAACTCACTGAAAAAAACTCA GTCAACATAGTCCAATTGAGTTT AKR1C3 [G A] ATTGGACTATGTTGAC TTTTCAGTGAGTTTTCC R170C GGGTGTCAAACTTCAACTGCAGGC CATCTCCAGCTGCCTGCAGTTGA [C T] AGCTGGAGATG AGTTTGACACCC P180S GAGATGATCCTCAACAAGTCAGGA GCTTGTACTTAAGTCCTGACTTGT [C T] CTTAAGTACAAGC TGAGGATCATCTC G135E GCCACTACCAAAAGATGAAAATGA CCACTGTGTCGAATATTACTTTTT AKR1C4 [G A] AAAAGTAATATTCGACACAGTGG CATTTTCATCTTTTGGTAGTGGC S145C CAGTGGATCTCTGTGCCACATGGG CTCCCATGTGGCACAGAGATCCA [CG] AG CTG

180 GENES VARIANTS a FORWARD PRIMER (5'3') REVERSE PRIMER (5'3') L311V GATATGTTGTCATGGATTTTGTTAT GATAATCAGGATGGTCCATAACA [C G] GGACCATCCTGATTATC AAATCCATGACAACATATC V135M CTGCAGTGTCCCCAAATGGACCTCT GGTAGAAGAGGTCCATTTGGGGA [G A] TCTACC CACTGCAG Q157H GCATGCCTGCCACCGGCTGCACCA CTGGTGCAGCCGGTGGCAGGCAT [GC] G GC E180K GCTGGGAAGTGGCCAAGATCTGTA GAGGGTACAGATCTTGGCCACTT AKR7A2 [G A] CCCTC CCCAGC G198S CCACTGTGTACCAGAGCATGTACA GGCGTTGTACATGCTCTGGTACA [G A] ACGCC CAGTGG S255N CGCTTCTTTGGGAATAACTGGGCTG GTAGGTCTCAGCCCAGTTATTCC [G A] AGACCTAC CAAAGAAGCG a The base pair mutation is given beneath each variant [in square brackets]. Also, the mutated base pair is underlined in the forward and reverse primer sequences.

181

Supplemental Figure 1—Purification of human recombinant 6x-His-tagged AKR1A1 variant (E55D and N52S) and AKR1B1 variant enzymes (I15F, H42L, L73V, K90E, G204S, and T288I). (Left ) Gel stained with SYPRO ® Ruby following SDS-PAGE showing purified protein fraction (lane 3; 2 µg), free of contaminating proteins from the bacterial lysate (lane 1; 10 µg total protein). Removal of contaminating proteins is observed in flow through fraction from Qiagen purification procedures [Ni-NTA column flow through (lane 2; 10 µg total protein)]. ( Right ) Western blot detection of transformed lysate (lane 6) and purified protein fractions (lane 8), confirms expression of the desired AKR protein. Little immunoreactivity was detected in the flow through fraction (lane 7) suggesting that majority of the enzyme was bound to the Ni-NTA resin prior to its elution. GST-tagged full recombinant proteins (Abnova ® Corporation, Taipei City, Taiwan; lane 5; 2 µg total protein) were used as positive controls for antibody immunoreactivity for all enzymes. No antibody immunoreactivity was observed for untransformed bacterial lysate (lane 4; 10 g total protein). M refers to the molecular weight markers.

182

Supplemental Figure 2—Purification of human recombinant 6x-His-tagged AKR1C4 variant (G135E, S145C, and L311V), AKR1C2 variant (F46Y), and AKR1C1 variant enzymes (R170C and Q172L). (Left ) Gel stained with SYPRO ® Ruby following SDS-PAGE showing purified protein fraction (lane 3; 2 µg), free of contaminating proteins from the bacterial lysate (lane 1; 10 µg total protein). Removal of contaminating proteins is observed in flow through fraction from Qiagen purification procedures [Ni-NTA column flow through (lane 2; 10 µg total protein)]. ( Right ) Western blot detection of transformed lysate (lane 6) and purified protein fractions (lane 8), confirms expression of the desired AKR protein. Little immunoreactivity was detected in the flow through fraction (lane 7) suggesting that majority of the enzyme was bound to the Ni-NTA resin prior to its elution. GST-tagged full recombinant proteins (Abnova ® Corporation, Taipei City, Taiwan; lane 5; 2 µg total protein) were used as positive controls for antibody immunoreactivity for all enzymes. No antibody immunoreactivity was observed for untransformed bacterial lysate (lane 4; 10 g total protein). M refers to the molecular weight markers.

183

Supplemental Figure 3—Purification of human recombinant 6x-His-tagged AKR1C3 variant (H5Q, R66Q, A106T, R170C, and P180S) and AKR1B10 variant enzymes (P87S, M286T, and N313D). (Left ) Gel stained with SYPRO ® Ruby following SDS-PAGE showing purified protein fraction (lane 3; 2 µg), free of contaminating proteins from the bacterial lysate (lane 1; 10 µg total protein). Removal of contaminating proteins is observed in flow through fraction from Qiagen purification procedures [Ni-NTA column flow through (lane 2; 10 µg total protein)]. ( Right ) Western blot detection of transformed lysate (lane 6) and purified protein fractions (lane 8), confirms expression of the desired AKR protein. Little immunoreactivity was detected in the flow through fraction (lane 7) suggesting that majority of the enzyme was bound to the Ni-NTA resin prior to its elution. GST-tagged full recombinant proteins (Abnova ® Corporation, Taipei City, Taiwan; lane 5; 2 µg total protein) were used as positive controls for antibody immunoreactivity for all enzymes. No antibody immunoreactivity was observed for untransformed bacterial lysate (lane 4; 10 g total protein). M refers to the molecular weight markers.

184

Supplemental Figure 4—Purification of human recombinant 6x-His-tagged AKR7A2 variant enzymes (V135M, A142T, Q157H, E180K, G198S, S255N). (Left ) Gel stained with SYPRO ® Ruby following SDS- PAGE showing purified protein fraction (lane 3; 2 µg), free of contaminating proteins from the bacterial lysate (lane 1; 10 µg total protein). Removal of contaminating proteins is observed in flow through fraction from Qiagen purification procedures [Ni-NTA column flow through (lane 2; 10 µg total protein)]. ( Right ) Western blot detection of transformed lysate (lane 6) and purified protein fractions (lane 8), confirms expression of the desired AKR protein. Little immunoreactivity was detected in the flow through fraction (lane 7) suggesting that majority of the enzyme was bound to the Ni-NTA resin prior to its elution. GST- tagged full recombinant proteins (Abnova ® Corporation, Taipei City, Taiwan; lane 5; 2 µg total protein) were used as positive controls for antibody immunoreactivity for all enzymes. No antibody immunoreactivity was observed for untransformed bacterial lysate (lane 4; 10 g total protein). M refers to the molecular weight markers.

185 Supplemental Table 2—Primers for creating non-synonymous single nucleotide polymorphic variants of human CBRs using site-directed mutagenesis.

GENES VARIANTS a FORWARD PRIMER (5' 3') REVERSE PRIMER (5' 3') V88I CTGGACGTGCTGATCAACAACGC CCGCGTTGTTGATCAGCACGTCC [G A] GG AG CBR1 P131S CTCCCTCTAATAAAATCCCAAGG CCACTCTCCCTTGGGATTTTATTA [C T] GAGAGTGG GAGGGAG P131S CTGCCGATAATGAAATCTCATGG CCACTCTCCCATGAGATTTCATTA [C T] GAGAGTGG TCGGCAG V244M GAAAGACAGCATCAGGACTATG GTCTCAGCCCCCTCCTCCATAGTC [G A] GAGGAGGGGGCTGAGAC CTGATGCTGTCTTTC C4Y GAATGTCGTCCTACAGCCGCGTG CGCCACGCGGCTGTAGGACGACA [G A] GCG TTC M235L GACCAGTGAAGACAGACTTGGAT CTGTCTTTCCCATCCAAGTCTGTC CBR3 [A T] GGGAAAGACAG TTCACTGGTC L84V CAAGGAGTACGGGGGGGTCAAT GTTGACCAGTACATTGACCCCCC [C G] GTACTGGTCAAC CGTACTCCTTG V93I CAACAACGCGGCCATCGCCTTCA CTCTTGAAGGCGATGGCCGCGTT [G A] AGAG GTTG D239V GACATGGATGGGAAAGTCAGCAT CACAGTCCTGATGCTGACTTTCC [A T] CAGGACTGTG CATCCATGTC L70M CATTTGAAGAGATGGAGAAACAT CTCGACCTAAATGTTTCTCCATCT CBR4 [C A] TTAGGTCGAG CTTCAAATG a The base pair mutation is given beneath each variant [in square brackets]. Also, the mutated base pair is underlined in the forward and reverse primer sequences.

186

Supplemental Figure 5—Purification of human recombinant 6x-His-tagged CBR variant enzymes. The top two panels represent variants of CBR1 (V88I and P131S) followed by the CBR4 variant (L70M) and the CBR3 variants (C4Y, L84V, V93I, P131S, M235L, D239V, and V244M). ( Left ) Gel stained with SYPRO ® Ruby following SDS-PAGE showing purified protein fraction (lane 3; 2 µg), free of contaminating proteins from the bacterial lysate (lane 1; 10 µg total protein). Removal of contaminating proteins is observed in flow through fraction from Qiagen purification procedures [Ni-NTA column flow through (lane 2; 10 µg total protein)]. ( Right ) Western blot detection of transformed lysate (lane 6) and purified protein fractions (lane 8), confirms expression of the desired CBR protein. Little immunoreactivity was detected in the flow through fraction (lane 7) suggesting that majority of the enzyme was bound to the Ni-NTA resin prior to their elution. GST-tagged human CBR4 recombinant protein (Abnova ® Corporation, Taipei City, Taiwan; lane 5; 2 µg total protein) and human liver cytosol (lane 5; 20 µg total protein; for CBR1 and CBR3) was used as positive controls for antibody immunoreactivity for these enzymes. No antibody immunoreactivity is observed for untransformed bacterial lysate (lane 4; 10 g total protein). M refers to the molecular weight markers.

187