The Role of a Liver-Testis Axis in the Development of Leydig Cell Hyperplasia and Tumours

THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

Michelle Coulson

August 2002

Supervisors: Professor G.Gordon Gibson, Dr. Mark Graham and Dr. Nick Plant

Studentship funded by AstraZeneca

Molecular Toxicology Group University of Surrey Guildford GU2 7XH ProQuest Number: 27557925

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a com plete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest

ProQuest 27557925

Published by ProQuest LLO (2019). Copyright of the Dissertation is held by the Author.

All rights reserved. This work is protected against unauthorized copying under Title 17, United States C ode Microform Edition © ProQuest LLO.

ProQuest LLO. 789 East Eisenhower Parkway P.Q. Box 1346 Ann Arbor, Ml 48106- 1346 Abstract

Leydig cell hyperplasia and adenomas are frequently observed during chronic carcinogenicity studies with new therapeutic agents, however, the significance of this effect to humans is unknown. Many compounds that produce Leydig cell tumours also induce hepatic cytochromes P450 (CYP) but it is unknown whether these two phenomena are causally related. The aim of this project was to investigate the existence of a liver-testis axis wherein microsomal enzyme inducers enhance testosterone metabolic clearance, resulting in a drop in circulating hormone levels and a consequent hypertrophic response from the hypothalamic-pituitary axis.

Lansoprazole was selected as the model compound as it induces hepatic CYPs and produces LCTs in rats (Atkinson et al., 1990; Masubuchi et al., 1997a). Effects of lansoprazole treatment (150 mg/kg/day for 14 days) on the liver, testis and endocrine control of the testis in male rats were investigated. The effects of lansoprazole were compared to model CYP inducers (P-NF, PB, PCN and ciprofibrate), the latter not being associated with LCTs.

Lansoprazole produced effects on the liver consistent with an increased metabolic capacity, including an increase in total microsomal CYP content, CYP induction (CYPlAl, CYP1A2, CYP2B1, CYP3A and CYP4A1) and enhanced CYP-

dependent testosterone metabolism in vitro (i.e. 6 ^-, 7a-, 16a-, 16p-, 2a- and 2p- OHT and androstenedione formation). Furthermore, lansoprazole-treated rats exhibited a significantly smaller AUCiast and significantly higher plasma clearance and volume of distribution (Vgg) following intravenous administration of ^"^C- testosterone. Reductions in plasma testosterone levels were observed in lansoprazole-treated animals but no significant changes in plasma LH or FSH levels were detected. Lansoprazole-treated animals exhibited lower plasma prolactin and intratesticular testosterone concentrations, which might play a role in LCT development. No marked effects on testicular CYP-dependent testosterone metabolism were observed.

In conclusion, these findings are consistent with the hypothesis that hepatic CYP induction contributes to the reduction in circulating testosterone levels in lansoprazole-treated animals, which might play a role in LCT development. Acknowledgements

I would like to thank my supervisors Professor Gordon Gibson, Dr. Mark Graham and Dr. Nick Plant for their continued guidance, enthusiasm and support throughout this project. I would also like to thank AstraZeneca, in particular Tim Hammond, for providing the funding for this research.

I would like to thank the various members of the Safety Assessment group for their assistance during the time I spent at AstraZeneca, in particular Paula Barkby, Julie Brett, Hamish Wilson, Isobel Clamp, James Finney and Tony Goodall. I would especially like to thank Dr Alex Bell for his input over the course of this project, particularly with the model inducers study and TaqMan work. I would also like to thank Louise Reid and Ruth Storer for their technical expertise, hard work and enthusiasm in conducting the ^"^C-testosterone clearance study. Finally, I would like to acknowledge Dr.Craig Lambert for his time and expertise in helping me to perform the pharmacokinetic analysis for the clearance study.

I would like to thank the various members of the School of Biological and Life Sciences for their assistance during this project. I would especially like to thank Peter Kentish for his time, technical expertise and encouragement throughout this project (and for lots of coffees!). I would also like to thank Ben Buckle for his help setting up the HPLC method. Thankyou to everyone from the Molecular Toxicology group past and present. In particular, I would like to thank Sarah Crunkhom and Karen Swales for all of their help and friendship over the last three years.

To my Mum and Jean-Philippe, thanks for your neverending support and patience. Table of Contents

Abstract...... i Acknowledgements ...... ii Table of contents ...... iii List of figures ...... x List of tables ...... xiv List of abbreviations ...... xvi

CHAPTER 1: INTRODUCTION 1.1 Preface...... 2 1.1.1 The Liver-Testis Axis and Leydig Cell Tumours ...... 2 1.1.2 The Liver-Thyroid Axis ...... 4 1.2 Testicular Physiology and Function ...... 7 1.2.1 Anatomy ...... 7

1.2.2 Testicular Cell Types ...... 8 1.2.3 Regulation of Testicular Function ...... 10 1.2.3.1 Endocrine Regulation ...... 10 1.2.3.2 Paracrine Regulation ...... 14 1.3 Xenobiotic-Induced Leydig Cell Tumours ...... 17 1.3.1 Introduction ...... 17 1.3.2 Pathology of Leydig Cell Tumours ...... 18 1.3.3 Role of LH in Leydig Cell Tumorigenesis ...... 18 1.3.4 Role of Paracrine Factors in Leydig Cell Tumorigenesis ...... 19 1.3.5 Mechanisms of Induction of Leydig Cell Tumours ...... 20

1.3.6 Human Relevance ...... 25 1.4 Cytochrome P450 and Testosterone Metabolism ...... 28 1.4.1 CytochromeP450 ...... 28 1.4.2 Testosterone Biosynthesis ...... 30 1.4.3 Testosterone Catabolism ...... 33 1.4.3.1 Introduction ...... 33 1.4.3.2 Regulation of Steroid Hormone Hydroxylation ...... 37 1.4.3.3 Testosterone Metabolism by Rat Liver Microsomes ...... 39 1.4.4 Testosterone Metabolism by Human Liver Microsomes ...... 44 1.5 Testicular Cytochrome P450 Enzymes ...... 46 1.5.1 Introduction ...... 46 1.5.2 Testicular Metabolism of Exogenous Compounds ...... 46

111 1.5.3 Testicular Metabolism of Endogenous Compounds ...... 48 1.5.4 Other Rat Testicular Cytochrome P450 ...... 50 1.6 Aims and Experimental Approach ...... 52 1.6.1 Aims...... 52 1.6.2 Experimental Approach ...... 52

CHAPTER 2: MATERIALS AND METHODS 2.1 Materials ...... 57 2.1.1 Chemicals ...... 57 2.1.2 HPLC Solvents ...... 59 2.1.3 Western Blotting Solutions ...... 59 2.1.4 ELISA Solutions ...... 61 2.1.5 Agarose Gel Electrophoresis Solutions ...... 61 2.2 Methods ...... 62 2.2.1 Animal Studies...... 62 2.2.1.1 Animal Housing ...... 62 2.2.1.2 Study 1 ...... 62 2.2.1.3 Study2 ...... L ...... 63 2.2.1.4 Study 3 ...... 64 2.2.1.5 Study 4 ...... 65

2.2.1.6 Study 5 ...... 65

2.2.2 Preparation of Microsomes ...... 6 6 2.2.3 Preparation of Testicular Supernatant ...... 67

2.2.4 Hormone Assays ...... 6 8

2.2.5 Protein Determination ...... , 6 8 2.2.6 Microsomal P450 Content Determination ...... 69 2.2.7 High Pressure Liquid Chromatography (HPLC) ...... 69 2.2.7.1 HPLC Apparatus ...... 69 2.2.7.2 Preparation of HPLC Standards ...... 70 2.2.7.S Chromatographic Conditions ...... 70 2.2.8 Testosterone Hydroxylase Assay; Microsomal Incubations ...... 71 2.2.8.1 Preparation of the Substrate ...... 71 2.2.8.2 Incubation Conditions ...... 71 2.2.8.3 Extraction of Testosterone and Metabolites ...... 72 2.2.8.4 Calculation of the Extraction Efficiency ...... 73 2.2.8.5 Calculation of Enzyme Activity ...... 73

IV 2.2.9 Optimisation of the Testosterone Hydroxylase Assay ...... 74 2.2.9.1 Incubation Time ...... 74 2.2.9.2 Microsomal Protein Concentration .-...... 74 2.2.9.3 Stability of Incubation Extracts ...... 75 2.2.9.4 Inclusion of 4-MA in Incubation Mixtures ...... 75 2.2.10 Extraction and Quantification of ‘'^C-Testosterone in ...... 76 Plasma Samples 2.2.10.1 Extraction of ’“^C-Testosterone from Plasma ...... 76 2.2.10.2 HPLC Analysis of Plasma Sample Extracts ...... — , 76 2.2.10.3 Quantification of ''‘C-Testosterone in Plasma ...... 77 Sample Extracts 2.2.10.4 Calculation of Kinetic Parameters ...... 77 2.2.11 Western Blotting ...... 79 2.2.11.1 Preparation of Samples ...... 79 2.2.11.2 SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE) 79 2.2.11.3 Transfer of Proteins to Nitrocellulose Membrane ...... 80 2.2.11.4 Immunodetection ...... 80 2.2.11.5 Staining of Membranes with Ponceau S ...... 81 2.2.12ELISA ...... 81 2.2.13 Optimisation of the ELISA Assay ...... 83 2.2.13.1 Specificity of the Anti-CYP Primary Antibodies ...... 83 2.2.13.2 Linear Range of Microsomal Protein Concentrations 83 2.2.14 Preparation of Samples for Real-time PCR (TaqMan™) ...... 84 2.2.14.1 Total RNA Extraction ...... 84 2.2.14.2 Purification of RNA ...... 85

2.2.14.3 Determination of Total RNA Yield and Quality ...... 8 6 2.2.14.4 First-Strand cDNA Synthesis ...... 87

2.2.15 Real-time PCR (TaqMan™) Analysis of Candidate Genes ...... 8 8

2.2.15.1 Background ...... :...... 8 8 2.2.15.2 Primer and Probe Design ...... 91 2.2.15.3 TaqMan Reaction ...... 93 2.2.15.4 Data Analysis ...... 94 2.2.16 Statistical Analysis ...... 94 2.3 Experimental Strategy ...... 95 CHAPTERS; METHOD OPTIMISATION 3.1 Introduction ...... 102 3.2 HPLC-Based Testosterone Hydroxylase Assay ...... 102 3.2.1 Introduction ...... 102 ,3.2.2 High Pressure Liquid Chromatography (HPLC) ...... 103 3.2.2.1 Separation of Authentic Testosterone Metabolites ...... 103 3.2.2.2 Linear Range of Radiochemical Detection ...... 105 3.2.3 Microsomal Incubations using Liver Microsomes ...... 106 3.2.3.1 Timecourse ...... 106 3.232 Protein Study ...... 108 3.2.3.3 Effect of 4-MA bn Microsomal Testosterone Metabolism ..... 109 3.2.3.4 Stability of Incubation Extracts ...... I ll 3.2.4 Microsomal Incubations using Testis Microsomes ...... I ll 3.2.5 Discussion ...... 113 3.3 ELISA ...... ' ...... 114 3.3.1 Introduction ...... 114 3.3.2 Specificity of the Anti-CYP Primary Antibodies ...... 115 3.3.3 Linear Range of Microsomal Protein Concentrations ..... 115 3.3.4 Discussion ...... 117

CHAPTER 4: EFFECTS OF MODEL INDUCERS ON CYPs, TESTOSTERONE METABOLIsk AND PLASMA HORMONE LEVELS 4.1 Introduction ...... 119 4.2 Study Design ...... 119 4.3 Results...... 120 4.3.1 Body, Liver and Testes Weights ...... 120 4.3.2 Plasma Hormone Levels ...... 120 4.3.3 Protein Content of Liver and Testis Microsomes ...... 123 4.3.4 Total CYP Content of Liver Microsomes ...... 124 4.3.5 Microsomal Testosterone Metabolism ...... 124 4.3.5.1 Testosterone Metabolism by Liver Microsomes ...... '...... 124 4.3.5.2 Testosterone Metabolism by Testis Microsomes ...... 134 4.3.6 Western Blotting ...... 137 4.4 Discussion ...... 139 4.4.1 Effects on Hepatic CYPs and Testosterone Metabolism ...... 139 4.4.1.1 Testosterone Metabolism by Control Liver Microsomes 139 4.4.1.2 Effects of Model Inducers ...... 140

VI 4.4.2 Effects on Testicular CYPs and Testosterone Metabolism ...... 145 4.4.2.1 Organ Weights ...... 145 4.4.2.2 Testosterone Metabolism ...... 145 4.4.3 Effects on Plasma Hormone Levels ...... 148 4.4.4 Conclusions ...... 150

CHAPTER 5: EFFECTS OF LANSOPRAZOLE ON CYPs, TESTOSTERONE METABOLISM AND PLASMA HORMONE LEVELS 5.1 Introduction ...... 153 5.2 Study Design...... 153 5.3 Results ...... 154 5.3.1 Body, Liver and Testes Weights ...... ,...... 154 5.3.2 Plasma Hormone Levels ...... 154 5.3.3 Protein Content of Liver and T estis Microsomes ...... 156 5.3.4 Total CYP Content of Liver Microsomes ...... 156 5.3.5 Microsomal Testosterone Metabolism ...... 157 5.3.5.1 Testosterone Metabolism by Liver Microsomes ...... 157 5.3.5.2 Testosterone Metabolism by Testis Microsomes ...... 159 5.3.6 ELISA...... 161 5.3.7 Western Blotting ...... 161 5.3.7.1 Liver Microsomes ...... 161 5.5.7.2 Testis Microsomes ...... ; ...... 167 5.4Discussion ...... 168 5.4.1 Final Body and Organ Weights ...... 168 5.4.2 Effect of Lansoprazole on Hepatic CYP Proteins ...... 169 5.4.3 Effect of Lansoprazole on Hepatic Testosterone Metabolism 171 5.4.4 Effect of Lansoprazole on Testicular CYP Proteins ...... 172 5.4.5 Effect of Lansoprazole on Testicular Testosterone Metabolism 172 5.4.6 Plasma Hormone Levels ...... 173 5.4.7 Conclusions ...... 175

CHAPTER 6: EFFECT OF LANSOPRAZOLE ON GENE EXPRESSION IN THE LIVER AND TESTIS 6.1 Introduction ...... 177 6.2 Study Design...... 177 6.3 Results ...... 178 6.3.1 Body, Liver and Testis Weights ...... 178

V l l 6.3.2 Total CYP Content of Liver Microsomes ...... 178 6.3.3 ELISA...... 179 6.3.4 Plasma Hormone Levels ...... ;...... 179 6.3.5 Intratesticular Testosterone Levels ...... 181 6.3.6 TaqMan Analysis ...... 183

6.3.6 .1 Preparation of cDNA ...... 183 6.3.6.2 Real-time PCR (TaqMan) Quantitation ...... 183 6.4 Discussion ...... : . 187 6.4.1 Final Body and Organ Weights ...... 187 6.4.2 Effect of Lansoprazole on Hepatic CYPs ! ...... 187 6.4.3 Plasma Hormone Levels ...... 188 6.4.4 Intratesticular Hormone Levels ...... 189 6.4.5 Effect of Lansoprazole Treatment on Gene Expression in the Liver.. 191 and Testis 6.4.6 Conclusions ...... 196

CHAPTER 7: EFFECT OF LANSOPRAZOLE ON GENE EXPRESSION IN THE PITUITARY GLAND 7.1 Introduction ...... 198 7.2 Study Design ...... 198 7.3 Results ...... 199 7.3.1 Body, Liver and Testis Weights ...... 199 7.3.2 Plasma Hormone Levels ...... !...... 199 7.3.3 TaqMan Analysis ...... 201 7,4Discussion ...... 202 7.4.1 Final Body and Organ Weights ...... 202 7.4.2 Plasma Hormone Levels ...... 202 7.4.3 Effect of Lansoprazole Treatment on Gene Expression in th e 203 Pituitary Gland 7.4.4 Conclusions ...... 204

CHAPTER 8 : EFFECT OF LANSOPRAZOLE TREATMENT ON THE PLASMA CLEARANCE OF ^^C-TESTOSTERONE 8.1 Introduction ...... 206 8.2 Study Design ...... 206 8.3 Results ...... 207 8.3.1 Body and Liver Weights ...... 207

V l l l 8.3.2 Analysis of Plasma Samples ...... 207 8.3.3 Quantification of ^'^C-Testosterone in Plasma Samples ...... 209 8.3.4 Calculation of Kinetic Parameters ...... 212 8.4Discussion ...... 213 8.4.1 Analysis of Plasma Samples ...... 213 8.4.2 Extractable Radioactivity ...... 215 8.4.2.1 ^'^C-Testosterone Plasma Concentration-Time Profiles ...... 215

8 .4.2.2 Kinetic Analysis ...... 216 8.4.3 Unextractable Radioactivity ...... 219 8.4.4 Conclusions ...... 221

CHAPTER 9: FINAL DISCUSSION 9.1 Introduction ...... 223 9.2 Effects of Lansoprazole on the L iver ...... 223 9.3 Effects of Lansoprazole on the Plasma Clearance of ^‘‘C-Testosterone 225 9.4 Effects of Lansoprazole bn Plasma Hormone Levels ...... 226 9.4.1 Testosterone ...... 226 9.4.2 Luteinising Hormone (LH) ...... 227 9.4.3 Other Hormones ...... 229 9.5 Effects of Lansoprazole on the Testis ...... 230 9.6 Conclusions ...... 233 9.7 Future Work ...... : ...... 234

References...... 237

IX List of Figures

1.1 Schematic representation of the liver-testis axis 3 1.2 Regulation of the thyroid gland by the hypothalamic-anterior pituitary- 5 thyroid axis 1.3 The structure of the mammalian testis 7 1.4 Regulation of the testis by the hypothalamic-anterior pituitary - testis 11 axis 1.5 Five mechanisms through which xenobiotics can disrupt the HPT axis - 21 to induce Leydig cell hyperplasia and tumours. 1.6 A schematic representation of the testicular steroidogenic pathway and 31 the cytochromes P450 involved 1.7 The major cytochromes P450 catalysing testosterone hydroxylation 35 reactions in rat liver microsomes 1.8 The chemical structure of lansoprazole 53 2.1 Diagrammatic representation of the steps in a typical TaqMan reaction 89 3.1 A representative chromatogram showing the separation of a mixture of 103 authentic testosterone metabolites by HPLC 3.2 The linear range of detection of the HPLC radioactivity detector 105 3.3 The formation of three testosterone metabolites as a function of the 107 incubation time by rat liver microsomes 3.4 The formation of three testosterone metabolites as a function of the 107 microsomal protein concentration 3.5 Effect of the 5a-reductase inhibitor, 4-MA, on testosterone metabolism 110 by liver microsomes 3.6 The stability of testosterone and hydroxytestosterone metabolites in a 112 reconstituted incubation extract. 3'7 Western blot of rat liver microsomes developed with anti-CYP 1A 116

3 .8 CYP 1A ELISA standard curves constructed using liver microsomes 116 from control and inducer-treated animals 4.1 Plasma testosterone levels in control, p-NF, PB, PCN and ciprofibrate- 121 treated animals 4.2 Plasma luteinising hormone levels in control, p-NF, PB, PCN and 121 ciprofibrate-treated animals 4.3 Plasma follicle-stimulating hormone levels in control, p-NF, PB, PCN 122 and ciprofibrate-treated animals 4.4 Plasma prolactin levels in control, p-NF, PB, PCN and ciprofibrate- 122 treated animals 4.5 Representative chromatographic separation of testosterone metabolites ' 125 formed following incubation of ^"^C-testosterone with liver microsomes from a control animal. 4.6 Representative chromatographic separation of testosterone metabolites 126 detected following incubation of ^"^C-testosterone with liver microsomes in the absence of NADPH 4.7 The effects of omission of NADPH from incubation mixtures on 127 testosterone metabolism by rat liver microsomes from a control animal. 4.8 Comparison of the levels of metabolites detected in blank incubations 127 with standard or washed liver microsomes 4.9 Testosterone metabolism catalysed by liver microsomes from control 130 and p-naphthoflavone-treated animals 4.10 Testosterone metabolism catalysed by liver microsomes from control 131 and PB-treafed animals 4.11 Testosterone metabolism catalysed by liver microsomes from control 132 and PCN-treated animals 4.12 Testosterone metabolism catalysed by liver microsomes from control 133 and ciprofibrate-treated animals 4.13 Chromatographic separation of testosterone metabolites formed 129 following incubation of ^"^C-testosterone with liver microsomes from a PB-treated animal. 4.14 Chromatographic separation of testosterone metabolites formed 134 following incubation of ^"^C-testosterone with pooled testis microsomes from control animals 4.15 Testosterone metabolism catalysed by pooled testis microsomes from 136 control and xenobiotic-treated animals 4.16 Western blot of rat liver microsomes developed with anti-CYP 1A2 137 primary antibody 5.1 Plasma testosterone levels in control and lansoprazole-treated rats 155

XI 5.2 Plasma prolactin, follicle-stimulating hormone and luteinising hormone 155 levels in control and lansoprazole-treated rats 5.3 The effect of treatment of rats with lansoprazole on testosterone 158 metabolism by liver microsomes 5.4 The effect of treatment of rats with lansoprazole on testosterone 160 metabolism by testis microsomes 5.5 Western blot of rat liver microsomes developed with anti-rat CYP1A2 162 primary antibody 5.6 Western blot of rat liver microsomes developed with anti-rat CYP2B 162 primary antibody 5.7 Western blot of rat liver microsomes developed with anti-rat CYP3A 164 primary antibody 5.8 Western blot of rat liver microsomes developed with anti-rat CYP3A 164 primary antibody 5.9 Western blot of rat liver microsomes developed with anti-rat CYP3A2 166 primary antibody from Gentest 5.10 Western blot of rat testis microsomes developed with anti-rat CYP4A1 166 primary antibody 6.1 Plasma testosterone levels in control and lansoprazole-treated rats 180 6.2 Plasma prolactin, follicle-stimulating hormone and luteinising hormone 180 levels in control and lansoprazole-treated rats 6.3 Intratesticular testosterone levels in control and lansoprazole-treated 182 animals - 6.4 Plasma testosterone levels in the subgroup of control and lansoprazole- 182 treated animals used to measure intratesticular hormone levels 6.5 Total RNA extracted from rat testis 183

6 . 6 Effect of lansoprazole treatment on liver 7 mRNA levels as 184 determined by TaqMan 6.7 Fold changes in liver mRNA levels of selected genes following treatment 185 of rats with lansoprazole as determined by TaqMan

6 . 8 Fold changes in testis mRNA levels of selected genes following treatment 186 of rats with lansoprazole as determined by TaqMan 7.1 Plasma testosterone levels in control and lansoprazole-treated rats 200

XU 7.2 Plasma prolactin, follicle-stimulating hormone and luteinising hormone 200 levels in control and lansoprazole-treated rats 7.3 Fold changes in pituitary gland mRNA levels for LH and prolactin 201 following treatment of rats with lansoprazole as determined by TaqMan 8.1 Representative HPLC profiles of serial plasma samples collected 208 following an intravenous injection of ^'^C-testosterone in a control rat

8 .2 ^"^C-testosterone plasma concentration decay curves in control and 210 lansoprazole-treated rats 8.3 Unextractable radioactivity present in serial plasma samples collected 211 following an intravenous injection of ^"^C-testosterone in a control rat

Xlll List of Tables

1.1 Examples of testicular regulatory factors and their effects on rat Leydig 15 cell numbers and function 1.2 Incidence of spontaneous Leydig cell tumours in rats 18 1.3 Examples of nongenotoxic pharmaceutical compounds that produce 22 Leydig cell hyperplasia and adenomas in rats 1.4 General functions of the cytochrome P450 gene families 29 1.5 Properties of the major rat hepatic cytochromes P450 that catalyse 36 testosterone metabolism. 2.1 Dosing schedules for compounds administered to rats for the model 63 inducers study. 2.2 The solvent gradient used to separate testosterone and metabolites 70 2.3 Components of incubation mixtures for the testosterone hydroxylase 72 assay 2.4 Guidelines for the design of TaqMan primers and probes 91 2.5 ^ Nucleotide sequences of primers and probes used in TaqMan analysis 92 of selected genes 2.6 The universal thermal cycling conditions used for TaqMan PCR 94 reactions 3.1 Chromatographic retention times for testosterone and metabolites 104 3.2 Linear range of metabolite formation by rat liver microsomes with 106 respect to incubation time 3.3 Results of linear regression analysis of data from the protein study 108 4.1 Effect of xenobiotic pretreatment on relative liver and testes weights 120 4.2 Effect of xenobiotic pretreatment on the protein yields of liver 123 and testis microsomes 4.3 Effect of xenobiotic pretreatment on the total cytochrome P450 content 124 of liver microsomes 4.4 Effect of xenobiotic pretreatment on CYP proteins levels in rat liver 138 microsomes 5.1 Effect of lansoprazole treatment on relative liver and testes weights 154 5.2 Effect of lansoprazole treatment on the protein yields of liver and testis 156 microsomes

XIV 5.3 Effect of lansoprazole treatment on the total cytochrome P450 content 156 of liver microsomes 5.4 ELISA quantitation of CYP proteins in liver microsomes from control 161 and lansoprazole-treated animals 6.1 Efrect of lansoprazole treatment on relative liver and testes weights 178 6.2 Effect of lansoprazole treatment on the total cytochrome P450 content 178 of liver microsomes 6.3 ELISA quantitation of CYP proteins in liver microsomes from control 179 and lansoprazole-treated animals 7.1 Effect of lansoprazole treatment on relative liver and testes weights 199 8.1 Effect of lansoprazole treatment on relative liver weights 207 8.2 ^"^C-testosterone plasma elimination kinetics in control and lansoprazole- 212 treated rats

XV Abbreviations

ABP Androgen-binding protein AhR Aryl hydrocarbon receptor AIS Androgen insensitivity syndrome ARNT Aryl hydrocarbon receptor nuclear translocator AUC Area under the plasma concentration-time curve bFGF Basic fibroblast growth factor BrdU Bromodeoxyuridine BROD Benzoxyresorufîn-0-dealkylase

C8 Ammonium perfluorooctanoate cAMP Cyclic adenosine monophosphate CMC Carboxymethylcellulose CYP Cytochrome P450 DES Diethylstilbestrol DHT Dihydrotestosterone DMBA 7,12-Dimethyl benz(a)anthracene DPM Disintegrations per minute EH Epoxide hydrolase ELISA Enzyme-linked immunosorbant assay EROD Ethoxyresorufin 0-dealkylase FSH Follicle-stimulating hormone GH Growth hormone GnRH Gonadotropin-releasing hormone hCG Human chorionic gonadotropin HPT Hypothalamic-pituitary-testis axis HRP Horseradish peroxidase 3P-orl7p-HSD 3p-or 17p-hydroxysteroid dehydrogenase HPLC High pressure liquid chromatography HPT Hypothalamic-pituitary-testis axis I3C Indole-3-carbinol IGF-1 Insulin-like growth factor-1 IHC hnmunohistochemistry i.p. Intraperitoneal

XVI l.V . Intravenous LCT Leydig cell tumour LSC Liquid scintillation counting LH Luteinising hormone 4-MA 17p -A, A-Diethylcarbamoyl-4-methyl-4-aza-5a-androstan-3 -one 3-MC 3-Methylcholanthrene MROD Methoxyresorufîn-0-dealkylase MTBE Methyl tert-h\xXy\ ether MTMP (Methylthio)methyl pyrazine NADPH Nicotinamide adenine dinucleotide phosphate NDMA N-Nitrosodiethylamine P-NF p-Naphthoflavone PB Phénobarbital PCN Pregnenolone-16a-carbonitrile PCR Polymerase chain reaction PDGF Platelet-derived growth factor PDR Peripheral benzodiazepine receptor PModS Peritubular factor that Modulates Sertoli cell function p.o. Per os (by mouth) PROD Pentoxyresorufin 0-dealkylase PTI 1-Methyl-3-propylimidazole-2-thione 5a-R Steroid 5a-reductase RIA Radioimmunassay mRNA Messenger ribonucleic acid rRNA Ribosomal ribonucleic acid SAP Steroidogenesis activating polypeptide SCP-2 Sterol carrier protein-2 SHBG Sex hormone binding globulin ST Sulfotransferase StAR Steroidogenic acute regulatory protein TAG Triacetyloleandomycin TCDD 2,3,7,8-Tetrachloro-dibenzo-/?-dioxin Ts Triiodothyronine

X V ll T4 Thyroxine TGF-a or -j3 Transforming growth factor-a or p TLC Thin layer chromatography TMB 3,3’,5,5’-tetramethylbenzidine

TNFa Tumour necrosis factor-a TRH Thyrotropin-releasing hormone TSH Thyroid-stimulating hormone UDP-GT Uridine diphosphoglucuronosyltransferase Vss Volume of distribution at steady state ZES Zollinger-Ellison syndrome

X V lll Chapter 1

Introduction 1.1 Preface

1.1.1 The Liver-Testis Axis and Leydig Cell Tumours

Leydig cell hyperplasia and adenomas are frequently observed during chronic toxicity and carcinogenicity studies with a diverse range of compounds. One common property of such chemicals is that they induce hepatic cytochromes P450 (CYPs)

(e.g. oxazepam, felbamate, ammonium perfluorooctanoate (C 8 ), lansoprazole and methyl tert-buXyl ether (MTBE)) (Diwan et al., 1986; reviewed by Glue et al., 1997; Biegel et al., 1995a; Masubuchi et al., 1997a; Williams & Borghoff, 2000). It is currently unclear whether these two phenomena are causally related.

CYPs play a role in steroid hormone homeostasis through their ability to metabolise steroids, such as testosterone. Exposure of rodents to microsomal enzyme inducers alters the composition of the hepatic CYP complement, resulting in concomitant changes in metabolism. It has been hypothesised that hepatic microsomal enzyme inducers produce Leydig cell tumours (LCTs) in rodents through their ability to increase the metabolic clearance of testosterone (Figure 1.1). The resultant decrease in circulating hormone levels would stimulate a compensatory increase in luteinising hormone (LH) secretion from the pituitary gland. It is widely accepted that chronic elevation of circulating LH levels is associated with induction of Leydig cell hyperplasia and tumours in rodents (Christensen & Peacock, 1980; Chatini et al., 1990; reviewed by Cook et al., 1999).

At present, very little research has been conducted to determine whether such a liver- testis axis exists. A limited number of studies have been performed to investigate the impact of microsomal enzyme inducers on androgen homeostasis in rodents. Wilson & Le Blanc (1998) examined the effects of the organochloride pesticide, endosulfan, on the rate of steroid hormone metabolism in female CD-I mice. Endosulfan-induced increases in the rate of hepatic testosterone metabolism were associated with enhanced urinary elimination of ^^"^C-testosterone (-3.6 fold). This increase in androgen clearance was associated with a small non-significant decrease in serum testosterone levels, suggesting that homeostatic feedback mechanisms were able to compensate for the effects of this compound in female mice. MTBE is an oxygenated fuel additive, which produces a dose-dependent increase in the incidence of LCTs in rats (Bird et al.. Figure 1.1 : Schematic representation of the liver-testis axis, a potential mechanism through which hepatic microsomal enzyme inducers might produce Leydig cell tumours.

Hepatic microsomal T clearance of enzyme inducer testosterone

circulating testosterone levels

T LH secretion

Trophic stimulation of the Leydig cells

Leydig cell hyperplasia and hypertrophy

Leydig cell tumours 1997; Belpoggi et al., 1995). Studies designed to investigate the mechanism of LCT induction revealed that MTBE induces CYPs involved in testosterone metabolism (e.g. CYP2B1/2, CYP3A1/2 and CYP2A1) and produces a significant reduction in serum testosterone levels (Williams & Borghoff, 2000; Williams et al., 2000). It is currently unknown whether the MTBE-induced decrease in circulating androgen levels occurs due to enhanced clearance of testosterone.

Further research is required to investigate the impact of microsomal enzyme inducers on androgen homeostasis and the potential of these compounds to produce LCTs secondary to increased metabolic clearance of steroid hormones.

1.1.2 The Liver-Thyroid Axis

The concept of the liver-testis axis is analogous to the liver-thyroid axis, which has an established role in the induction of thyroid follicular cell hyperplasia and tumours by microsomal enzyme inducers in rats (reviewed by McClain, 1992; Hard, 1998).

The thyroid gland is composed of two distinct cell types: the follicular cells and parafollicular cells (also called C-cells). The follicular cells synthesise and secrete the thyroid hormones T 4 (thyroxine) and T 3 (triiodothyronine) under the influence of the pituitary factor, thyroid-stimulating hormone (TSH). This forms part of the hypothalamic-pituitary-thyroid axis, which maintains the appropriate levels of circulating thyroid hormones (Figure 1.2). TSH is the primary regulator of thyroid growth and differentiated function but a variety of local regulatory factors have also been identified (e.g. insulin-like growth factor-1 (IGF-1), transforming growth factor beta (TGF-P) and basic fibroblast growth factor (bFGF)) (reviewed by Hard, 1998).

Three major pathways are involved in the inactivation and excretion of thyroid hormones (Eelkman Rooda et al., 1989; De Herder et al., 1988). The first pathway involves sequential deiodination catalysed by microsomal monodeiodinase enzymes, which are expressed in tissues such as the liver, kidney and brain. The second and third pathways involve conjugation with glucuronic acid or sulfate, catalysed by uridine diphosphoglucuronosyltransferase (UDP-GT) and sulfotransferase (ST) enzymes respectively. These reactions occur predominantly in the liver and facilitate Figure 1.2: Regulation of the thyroid gland by the hypothalamic - anterior pituitary - thyroid axis

Higher Brain Centres

Hypothalamus

TRH

Anterior Pituitary

Hepatic Metabol TSH -ism

Thyroid Gland Biliary Excretion

Thyroid Hormones (Ts, T4)

T3 , triiodothyronine; T 4 , thyroxine; TRH, thyrotropin- releasing hormone; TSH, thyroid-stimulating hormone; + or -, stimulation or inhibition of target cell population biliary and urinary excretion of the conjugated hormone. Other pathways, such as deamination and decarboxylation are quantitatively less important.

Long term exposure to a diverse range of compounds induces thyroid follicular cell hyperplasia and tumours in rodents. Many of these chemicals are hepatic microsomal enzyme inducers, which produce histological changes in the thyroid gland secondary to increased metabolic clearance of thyroid hormones (e.g. l-methyl-3- propylimidazole-2-thione (PTI), phénobarbital (PB) and ciprofibrate) (McClain et al., 1989; Visser et al., 1991; Biegel et al., 1995b; Schuur et al., 1997). The consequent reduction in circulating thyroid hormone levels stimulates a compensatory increase in TSH secretion. The role of sustained elevation of serum TSH levels in thyroid gland neoplasia is a well-established phenomenon (reviewed by McClain, 1992).

PB is an example of a well-characterised hepatic microsomal enzyme inducer, which acts as a tumour promoter in the rodent thyroid gland (McClain et al., 1988). Treatment of rats with PB increases the plasma clearance of thyroxine through marked induction of hepatic UDP-GT activity and increased biliary flow (McClain et al.,

1989). PB-treated rats show characteristic reductions in plasma T 4 and T3 levels and elevated TSH levels (McClain et al., 1989). These endocrine changes are accompanied by an increase in thyroid weight and follicular cell hyperplasia and tumours (McClain et al., 1989). McClain et al. (1988) demonstrated that administration of thyroxine to PB-treated rats completely blocked the tumour promoting effect of this compound. In humans, several studies have shown that PB induces hepatic microsomal enzymes and increases thyroid hormone turnover, but circulating thyroid hormone and TSH levels are generally unaffected (reviewed by Curran & Degroot, 1991). Consistent with these findings, there is no epidemiological data linking PB with an increased incidence of thyroid cancer in humans (reviewed by Curran & DeGroot, 1991).

For the current project, the concept of the liver-thyroid axis was used as a model to investigate the existence of a liver-testis axis in rats. 1.2 Testicular Physiology and Function

1.2.1 Anatomy

The male reproductive system is composed of the testes, a system of ducts, the accessory sex organs (prostate, seminal vesicles and bulbourethral glands) and several supporting structures. The testes have both endocrine and reproductive functions as they produce male sex hormones (mainly testosterone) and mature sperm cells. The system of ducts store and transport the sperm, while secretions from the accessory sex organs provide optimal conditions for sperm motihty, survival and transport.

Figure 1.3: The structure of the mammahan testis a) A sagittal section of the testis and associated epididymis b) A transverse section through a seminiferous tubule

■ Sperm atic co rd

Blood vessels a n d n erv es

p-Seminiferous tubule

Head of epididymis

Efferent ductule

Ductus (vas) Ad acent saminile/cus tubules d eferen s L obule interstitial cells Rete testis S eptum Tubulus rectus

Tunica albuginea Body of epidiidymis Tunica vaginalis

'— Cavity of Tail of epididym is tunica vaginalis

Spe^fnaiozoon

perm atiQ

Blood vessel

Nucleus of SofioH cell/ Spermaiogoniom

Primary spermaiocyic

Secondary soorrnatocyie

(Adapted from Marieb, 1995) The testes are paired oval glands enclosed in a scrotal sac that lies outside of the abdominal cavity (Figure 1.3). A thick connective tissue capsule made up of three tunica layers surrounds each testis: tunica vasculosa (innermost), tunica albuginea and tunica vaginalis (outermost). The tunica albuginea extends inwards, dividing each testis in to a number of compartments or lobules. Each lobule contains one to three highly coiled seminiferous tubules, which are lined with Sertoli cells and spermatogenic cells at various stages of development (tubular compartment). The seminiferous tubules are surrounded by interstitial tissue composed of loose connective tissue containing mast cells, fibroblasts, macrophages, nerve fibres, lymph vessels, blood vessels and clusters of Leydig cells (interstitial compartment). A layer of peritubular tissue composed of extracellular matrix, fibroblasts and myoid cells forms the border of the interstitial space. -

1.2.2 Testicular Cell Types

The endocrine and reproductive functions of the testis involve three major cell types: Leydig cells that secrete testosterone, Sertoli cells that support spermatogenesis and the male germ cells. The roles of each cell type are discussed in the following section.

Leydig Cell

Leydig cells appear as clusters of plump cells that reside in the interstitial space between adjacent seminiferous tubules. Closer examination reveals numerous mitochondria, abundant smooth endoplasmic reticulum, a well-developed Golgi complex, variable numbers of lipid droplets and a network of peroxisomes. Leydig cells are often found intimately associated with blood and lymph vessels, which facilitates the rapid exchange of materials with the systemic circulation (Fawcett et al., 1973).

' ■ ■ ■ Leydig cells are the principal site of androgen biosynthesis (mainly testosterone), which underlies their central role in the paracrine control of spermatogenesis and endocrine control of structures such as the accessory sex organs, bone and skeletal muscle (Christensen & Mason, 1965; Hall et al., 1969). The androgèn-dependence of these structures is illustrated following androgen withdrawal, which leads to rapid germ cell depletion and atrophy of the accessory sex organs in rats (Bartlett et al.. 1986). In the adult testis, the Leydig cell is also the major site for aromatisation of testosterone to form oestradiol (Tsai-Morris et al., 1985).

In the mammalian testis, at least two distinct generations of Leydig cells exist: fetal and adult Leydig cells (reviewed by Habert et al., 2001). The first generation develops during fetal life and these cells are responsible for masculinisation of the male reproductive system. The second population of Leydig cells predominate in the adult testis and produce the testosterone required for the onset of spermatogenesis and maintainance of reproductive function. Both generations of Leydig cells appear to be formed by differentiation from mesenchymal-like stem cells (Hardy et al., 1989). In the adult rat, Leydig cells comprise approximately 2.7% of the testicular volume (Mori & Christensen, 1980). Under normal conditions Leydig cells rarely divide and this cell population remains relatively stable throughout adult life (Teerds et al., 1989a).

Sertoli Cell

The Sertoli cells are found embedded between the germ cells and play a crucial role in the regulation of spermatogenesis. They provide nutritional, metabolic and structural support to the developing germ cells. The Sertoli cells secrete tubular fluid, which transports newly formed sperm cells to the epididymis and contains a number of important proteins such as transferrin, plasminogen activator and androgen binding protein (ABP). ABP sequesters testosterone and maintains the high intratesticular hormone levels required for spermatogenesis (Zirkin et al., 1989). The unique tight junctions between adjacent Sertoli cells form the blood-testis barrier, which separates the interstitial and tubular compartments. This barrier aids regulation of the chemical environment of the tubule and affords some protection of the developing germ cells from potentially mutagenic substances. The Sertoli cell also appears to play an important role in intratesticular communication through secretion of a variety of proteins and peptides (e.g. TGF-P and IGF-1) (reviewed by Gnessi et al., 1997; Saez, 1994) (se el.2.3.2). Germ Cell

Spermatogenesis is the highly synchronised and complex process through which primitive germ cells divide and mature to form specialised spermatozoa. There are essentially three different germ cell types: spermatogonia, spermatocytes and spermatids. The spermatogonia are the primitive germ cells, which divide and differentiate to form mature sperm cells and continuously replenish the stem cell population (Huckins, 1971). Spermatogenesis occurs within the seminiferous tubules, where the spermatogonia are closely associated with the basement membrane and the developing germ cells progressively migrate towards the lumen of the tubule (reviewed by Foote & Bemdtson, 1992). This process occurs throughout adult life and takes approximately 56 days in the rat (Leblond & Clermont, 1952).

1.2.3 Regulation of Testicular Function

1.2.3.1 Endocrine Regulation

The functions of the testis are primarily regulated by the hypothalamic-pituitary-testis axis (HPT axis) which maintains the dynamic equilibrium of circulating testosterone levels using a closed negative feedback loop (Figure 1.4). Overall control is at the level of the hypothalamus which releases pulses of gonadotropin-releasing hormone (GnRH) at a critical frequency and amplitude. GnRH stimulates the anterior pituitary gland to release luteinising hormone (LH) and follicle-stimulating hormone (FSH), which act on specific target cells within the testis.

LH

LH acts exclusively on the Leydig cells and is the primary regulator of testosterone biosynthesis and secretion (Menon et al., 1967; Purvis & Hansson, 1978). Binding of LH to a specific plasma membrane receptor activates a cAMP-dependent protein kinase pathway leading to an acute stimulation of testosterone production (Cooke et al., 1976). LH also exerts a long-term trophic effect on the Leydig cell and is the major factor required to maintain the fully differentiated structure and function of this cell population (reviewed by Saez, 1994). This long-term effect of LH is dependent upon RNA and protein synthesis (reviewed by Saez, 1994). Hypophysectomy or suppression of LH is associated with Leydig cell atrophy, diminished volume of the

10 Figure 1.4: Regulation of the testis by the hypothalamic - anterior pituitary - testis axis

Higher Brain Centres

Dopamine, opiates, neuropeptides

Hypothalamus

GnRH

Anterior Pituitary

Testosterone Inhibin LH FSH

Testosterone Leydig Cell Sertoli Cell

GnRH, gonadotropin-releasing hormone; LH, luteinising hormone; FSH, follicle- stimulating hormone; TGF-p, transforming growth factor beta; IGF-1, insulin-like

growth factor - 1 ; + or -, stimulation or inhibition of target cell population.

11 smooth endoplasmic reticulum, reduced steroid secretory capacity and a decrease in LH receptor numbers (Russell et al, 1992; Keeney et al, 1988).

Testosterone completes a negative feedback loop to regulate further LH secretion from the pituitary gland. Testosterone acts mainly at the level of the hypothalamus, but also has a weak inhibitory effect on the pituitary gland (Roselli & Resko, 1990). Oestradiol, produced by the aromatization of testosterone in tissues such as the brain and testis, also participates in this negative feedback loop acting mainly at the level of the pituitary gland (Nishihara & Takahashi, 1983; Roselli & Resko, 1990). These regulatory pathways allow the Leydig cells to adapt to the fluctuating demands of the body for steroid hormones. In addition, stimulation of the Leydig cell by LH has been reported to evoke adaptive changes (e.g. downregulation of LH receptor numbers and uncoupling of the receptor from adenylate cyclase), which tend to reduce cell responsiveness to further stimulation (Wang et al., 1991; Saez et al., 1978a,b). However, steroidogenic desensitisation appears to represent a pharmacological response, which is probably not involved in the regulation of Leydig cell steroidogenesis in vivo. Studies have demonstrated that treatment of rats with LH/hCG causes downregulation of testicular LH receptors, but steroidogenic desensitisation is only observed following a single large, non-physiological dose of LH/hCG (Saez et al., 1978a,b; Zipf et al., 1978b). Indeed, repeated treatment of rats with low levels of LH enhances LH-stimulated steroidogenesis in vivo, despite downregulation of LH receptor numbers (Zipf et al., 1978b).

FSH

FSH acts primarily on the Sertoli cells, but also has a mitogenic effect on the spermatogonia (reviewed by Bardin et al., 1994; Griswold, 1993). Binding of FSH to a plasma membrane receptor activates a cAMP-dependent protein kinase pathway leading to changes in protein synthesis and cellular function. Little is known about the precise hormonal roles of FSH in adult males, but it appears to be essential for the initiation and maintenance of normal spermatogenesis (Matsumoto et al., 1986). A close interrelationship exists between FSH, testosterone and the Sertoli cell. In the immature testis, FSH triggers seminiferous tubule growth, initiates spermatogenesis and stimulates the secretion of Sertoli cell products such as tubular fluid and ABP (reviewed by Sharpe, 1994). In the adult testis, most of these functions appear to be

12 more responsive to testosterone (Dym et al, 1979; Ahmad et al., 1975). FSH may also act, via the Sertoli cell population, to regulate Leydig cell structure and function through the production of various paracrine factors (see 1.2.3.2).

FSH stimulates the Sertoli cell to release of the non-steroidal factor inhibin, which completes a negative feedback loop to regulate further FSH secretion from the pituitary gland (Bicsak et al, 1987; Franchimont et al., 1975). Inhibin has no effect on pituitary LH secretion (Franchimont et al., 1975). Gonadal sex steroids have minimal effects on FSH secretion, although oestrogens have an inhibitory influence (Gay & Dever, 1971). Severe damage to the seminiferous tubules or germ cell population is associated with elevated circulating FSH levels, probably due to reduced secretion of inhibin from the Sertoli cells (Zylber-Haran et al., 1982).

Prolactin

The anterior pituitary gland also secretes prolactin, which appears to play an important role in the regulation of gonadal function. Secretion of prolactin is influenced by multiple factors including dopamine (inhibits release) and thyrotropin-releasing hormone (TRH) (stimulates release) (Hill-Samli & MacLeod, 1974; MacLeod & Lehmeyer, 1974).

Prolactin has no well-defined physiological role in males but adequate levels appear to be essential for normal testicular function (reviewed by Bartke, 1980). Prolactin itself has little influence on the male reproductive tract, but it potentiates the effects of LH (Welsh et al., 1986; Bartke et al., 1978; Hafiez et al., 1972). Prolactin appears to be required for the maintenance of LH receptor numbers, which increases the sensitivity of the Leydig cells to gonadotropin stimulation (Welsh et al., 1986; Bartke et al., 1978; Zipf et al., 1978a). This hormone also appears to influence testosterone biosynthesis through effects on the availability of androgen precursors (Welsh et al., 1986). In addition to the direct effects on the testis, prolactin appears to exert indirect effects on testicular function through modulation of gonadotropin secretion from the pituitary gland (McNeilly et al., 1978). Prolactin suppresses LH and FSH secretion, which may be due to decreased pituitary responsiveness to GnRH or reduced secretion of GnRH from the hypothalamus (Winters & Loriaux, 1978; Smith & Bartke, 1987; Cheung, 1983). Prolactin appears to be essential for normal testicular function but

13 pathological increases in circulating hormone levels can have a deleterious effect. In humans and rodents, hyperprolactinaemia is associated with a hypogonadism characterised by testicular atrophy, decreased testosterone production and infertility (Carter et al., 1978; Sharpe & McNeilly, 1979; Bartke et al., 1977; Fang et al., 1974).

1.2.3.2 Paracrine Regulation

The HPT axis provides the overall control of testis function but a number of intratesticular regulatory factors have also been identified. These may constitute a more sensitive local control system to enable rapid “fine-tuning” of testicular cellular function with changing requirements. Many of these factors are expressed and secreted in a highly regulated manner, suggesting potential roles in the regulation of testicular development and function. Locally secreted factors may mediate regulatory pathways between the interstitial and tubular compartments. Numerous potential paracrine factors have been detected in the mammalian testis but in general their precise roles remain poorly understood (reviewed by Habert et al., 2001; Gnessi et al., 1997; Saez, 1994). For the majority, effects on testicular cells have been observed in vitro and their physiological significance is unknown. Paracrine regulation of the Leydig and Sertoli cell populations is discussed in the following sections.

Leydig Cell

Leydig cell structure and function is primarily regulated by pituitary LH, but may also be influenced by paracrine factors. Leydig cells appear to regulate their own activity through gap junctions linking adjacent cells and secretion of autocrine factors such as testosterone and oestradiol, which have inhibitory effects on steroidogenesis (Risley et al., 1992; Damey et al., 1996). Sertoli cells, peritubular cells, germ cells and. macrophages secrete numerous regulatory factors, which appear to act upon the Leydig cell (reviewed by Habert et al., 2001; Gnessi et al., 1997; Saez, 1994). Table 1.1 summarises the effects of various paracrine factors on Leydig cell number and function and two well-documented examples are discussed in more detail below.

IGF-1, derived from the circulation or produced locally, may influence Leydig cell proliferation, differentiation and function (Gelber et al., 1992; Khan et al., 1992). During puberty, circulating IGF-1 levels rise concomitantly with the period of rapid Leydig cell proliferation in rats (reviewed by Daughaday & Rotwein, 1989). IGF-1

14 e I U I I y Cû. I o i I ü î S3 O î u I g î II I 00 I 00 ’.d •2 00 O

I 0 .S2 u I •S .Sf S rO e 0 1 "d y & 1 î î I r I 1 1 d o % f I r 0 S3 I I 00 00 00 ÛÛ O 'a :§ .sI u 00 I 1 M s § k 2 I & + + + + + + + + I u I O-S3 U O Ü *-Cu u S' 3 o u ci J- d 2 P4 y CJ U o O î P4 S y y O co co H-l O y a y I 2 00 ci y % y co K

Platelet-derived growth factor (PDGF), which is produced by the Sertoli cells, enhances the LH-stimulated steroidogenic response in adult Leydig cells (Risbridger, 1993). PDGF gene knockout mice develop marked testicular abnormalities. These animals have normal testis at birth, but subsequently show progressive reductions in testicular size and a marked absence of adult Leydig cells (Gnessi et al., 2000). These studies suggest that PDGF may play a crucial role in the commitment of Leydig cell precursors to proliferate and differentiate to form the adult Leydig cell population.

Sertoli Cell

Sertoli cells are a major site for the testicular synthesis and secretion of paracrine factors but are also a target for the effects of certain regulatory molecules. The Leydig cells produce factors such as testosterone and |3-endorphin, which influence the secretion of several of Sertoli cell products (Fabbri et al., 1985; Skinner & Fritz, 1985). p-Endorphin has inhibitory effects on the Sertoli cell, including inhibition of ABP production and cell proliferation, and may have a local role in antagonism of the stimulatory effects of FSH on this cell population (Fabbri et al., 1985; Orth, 1986). The peritubular myoid cells produce a nonmitogenic paracrine factor termed PModS, which stimulates Sertoli cell functions such as ABP and transferrin secretion (Norton & Skinner, 1989; Skinner & Fritz, 1985). There is evidence that androgens may regulate the secretion of PModS from the peritubular cells (Skinner & Fritz, 1985). Sertoli cell morphology and secretory activity also appears to vary with the stage of the seminiferous epithelium, suggesting that the germ cells may release factors to modulate Sertoli cell function (reviewed by Sharpe, 1994). Elimination of specific germ cell types from the testis is associated with stage-dependent changes in inhibin secretion from this cell population (reviewed by Weinbauer & Wessels, 1999).

16 1.3 Xenobiotic-Induced Leydig Cell Tumours

1.3.1 Introduction

Mammalian reproduction is a complex and highly regulated process, which provides many potential sites for toxicological disruption. There are a vast number of publications describing adverse effects of chemicals on the male reproductive system, including drugs, pesticides, industrial chemicals and environmental contaminants (reviewed by Creasey, 1998). For many of these agents, their primary target site and biochemical mechanism of toxicity are unknown.

The testes are the most significant site of xenobiotic-induced damage that affects fertility. Testicular toxicants can be divided in to two broad classes based on their mode of action. The first class of compounds directly interfere with the cells involved in spermatogenesis (i.e. germ and Sertoli cells), often due to some inherent chemical reactivity. This class includes agents that damage cellular macromolecules (DNA, RNA or proteins), which may have cytotoxic or genotoxic consequences for the cell (e.g. cyclophosphamide, daunorubicin). The second class of compounds indirectly affect spermatogenesis through disturbance of endocrine or paracrine regulatory pathways (e.g. ethanol, ethane dimethane sulphbnate). v Morphological responses observed in testicular cells following a toxic insult include degeneration, necrosis, inflammation, hyperplasia and neoplasia.

In general, testicular tumours are rarely observed in laboratory animals with the exception of Leydig cell tumours (LCTs), which frequently occur in rats. Rat LCTs occur spontaneously and are readily induced by a diverse range of xenobiotics. The spontaneous incidence of LCTs is age-related and varies considerably depending on the strain, most notably the incidence in two-year-old Fischer 344 rats is approximately 76% (Mitsumori & Elwell, 1988) (Table 1.2). The underlying causes of the high incidence in this strain are currently unknown, but it might be related to a progressive hormonal imbalance (Amador et al., 1985). The spontaneous incidence of LCTs is considerably lower in all strains of mice (Mitsumori & Elwell, 1988; Lang, 1995).

17 Table 1.2: Incidence of spontaneous Leydig cell tumours in strains of rats commonly used in safety assessment studies at 24 months old.

Strain Incidence %

Wistar 87/ 1249 7

Sprague-Dawley 16/349 4.6 F344 39,253/51,230 76.6 Crl:CD®BR 37/721 5.1

■ F344/DuCij 537/569 94.4

(Data derived from Cook et al., 1999)

1.3.2 Pathology of Leydig Cell Tumours

The morphologic appearance of the proliferative changes in the Leydig cells that result in the generation of LCTs appears to be the same in both spontaneous and chemically- induced tumours (McConnell et al., 1992). This usually begins with focal hyperplasia, which may progress to form benign tumours (adenomas). In rare cases, adenomas undergo malignant transformation to form carcinomas. No consistent criteria have been identified to differentiate between Leydig cell hyperplasia and adenomas therefore this distinction is often based primarily on the size of the lesion. Guidelines released by The Society of Toxicologic Pathologists propose that an adenoma be defined as a lesion that is larger than the diameter of three seminiferous tubules (McConnell et al, 1992). In rodents, it may take up to two years for xenobiotic- induced LCTs to develop and generally there are no distinctive early (i.e. following one to two weeks of compound exposure) indications of Leydig cell proliferation (reviewed by Clegg et al., 1997).

1.3.3 Role of LH in Leydig Cell Tumorigenesis

It is well established that LH has a major role in the control of Leydig cell function but this hormone may also regulate Leydig cell proliferation. There is considerable evidence to support a causal relationship between elevated circulating LH levels and the development of LCTs. Leydig cells rarely divide but prolonged administration of

18 LH to rats, mice and monkeys can induce Leydig cell proliferation (Christensen & Peacock, 1980; Pfeiffer & Hooker, 1943; Simpson & Van Wagenen, 1954; Teerds et al., 1989b). Chatini et al (1990) reported that spontaneously occurring LCTs in Fischer 344 rats are associated with high circulating LH levels and testosterone supplementation prevented LCT development. Elevation of circulating LH levels has been reported in animals treated with a diverse range of Leydig cell tumourigens, suggesting a common mechanism of tumour induction (reviewed by Cook et al., 1999). In rats, prolonged elevation of serum gonadotropins is correlated to the increased incidence of LCTs (Brown et al., 1979; Chatini et al., 1990; Yamada et al., 1994b). The evidence described herein suggests that any condition capable of producing a chronic elevation of circulating LH levels may have the potential to induce Leydig cell hyperplasia and tumours. It is currently unclear whether chronic elevation of LH levels alone is a sufficient stimulus to cause Leydig cells to divide or whether other factors are involved.

Very little is known about the mechanism through which LH may induce LCTs, but it might involve the activation of proto-oncogenes (reviewed by Saez, 1994). Studies using cultured pig Leydig cells and MA-10 tumour cells have confirmed that LH/human chorionic gonadotropin (hCG) exposure induces a rapid and transient increase in c-fos and c-jun mRNA levels (Hall et al., 1991; Czerwiec et al., 1989). The role these transcription factors in the long-term trophic effects of LH are currently unknown.

1.3.4 Role of Paracrine Factors in Leydig Cell Tumorigenesis

There is substantial evidence to support a role for LH in Leydig cell tumorigenesis, but other findings indicate that testicular paracrine factors might also be involved. Cook et al (1999) postulated that paracrine factors are likely to play a role in the aetiology of LCTs in rats, either as a primary factor or as a consequence of other changes (e.g. gonadotropin levels or spermatogenesis). Studies have demonstrated that local regulatory factors can influence Leydig cell proliferation, therefore changes in the balanced secretion of inhibitory and stimulatory factors might predispose the testis to uncontrolled cell division. The paracrine pathways that may be involved are currently unknown.

19 1.3.5 Mechanisms of Induction of Leydig Cell Tumours

An extensive and diverse range of pharmaceutical preparations have been reported to increase the incidence of Leydig cell hyperplasia and tumours during rodent toxicity and carcinogenicity studies (reviewed by Cook et al, 1999) (Table 1.3). Figure 1.5 illustrates five mechanisms through which chemicals can disrupt the HPT axis and produce LCTs. The common point is that all of these compounds increase circulating LH levels. The following section uses a mechanistic approach to classify some of the nongenotoxic compounds that induce Leydig cell hyperplasia and adenomas.

Testosterone Biosynthesis Inhibitors. Testosterone biosynthesis inhibitors reduce serum testosterone levels and increase serum LH levels in rats (Fort et al., 1995; Hamada & Futamura, 1998). Most testosterone biosynthesis inhibitors belong to the imidazole, benzimidazole, dicarboximide or dimethylpyridine classes. Imidazole and benzimidazoles inhibit cytochrome P450-catalysed steps in the testosterone biosynthetic pathway by binding to the protoporphyrin iron of the enzyme (Vanden Bossche, 1985; Feldman, 1986).

Lacidipine is a calcium channel antagonist that inhibits testosterone biosynthesis, increases plasma LH levels and produces LCTs in rats but not mice (Hamada & Futamura, 1998). This compound reduces basal and LH-stimulated testosterone production in isolated Leydig cells from rats, mice and monkeys (Hamada & Futamura, 1998). Lacidipine increased 5-bromodeoxyuridine (BrdU) incorporation in LH-stimulated rat Leydig cells, but not in Leydig cells from mice or monkeys (Hamada & Futamura, 1998). These findings are consistent with data from carcinogenicity studies in rats and mice and suggest that elevated LH levels may be responsible for LCT induction in rats. Other examples of testosterone biosynthesis inhibitors known to produce LCTs in rats include felodipine (Oradell, 1995a); cimetidine (Leslie et al., 1981) and metronidazole (Rustia & Shubik, 1979).

Androgen receptor antagonists. These compounds bind to the androgen receptor to reduce the negative feedback signal of gonadal steroids at the hypothalamus and pituitary gland (Simard et al., 1986). This leads to hypersecretion of LH and a consequent increase in circulating testosterone levels (Cook et al., 1993; Viguier-

20 Figure 1.5: Five mechanisms through which xenobiotics can disrupt the HPT axis to induce Leydig cell hyperplasia and tumours.

Higher Brain Centres 1. Dopamine agonists

Dopamine, opiates, neuropeptides

Hypothalamus

Antiandrogens GnRH © ! ■3. 5a-reductase

> Anterior Pituitary

LH

Leydig Cell 4. Testosterone TESTIS W Biosynthesis ^ Inhibitors Testosterone ------

5. Aromatase CD Inhibitors

Oestradiol

(Adapted from Cook et al., 1999)

21 Table 1.3: Examples of nongenotoxic pharmaceutical compounds that produce Leydig cell hyperplasia and adenomas in rats

Compound Clinical Indication Leydig Cell Response

Buserelin Prostate and breast carcinoma Hyperplasia Carbamazepine Anticonvulsant/ analgesic Adenoma Cimetidine Reduction of gastric acid secretion Adenoma Finasteride Benign prostatic hyperplasia Hyperplasia Flutamide Prostate carcinoma Adenoma Gemfibrozil Hyperlipidaemia Adenoma Isradipine -Antihypertensive Adenoma Leuprolide Endometriosis Adenoma Mesulergine Hyperprolactinaemia Adenoma Norprolac Hyperprolactinaemia Adenoma Metronidazole Antibacterial Adenoma Indomethacin Anti-inflammatory Adenoma Bicalutamide Prostate carcinoma Adenoma Lansoprazole Reduction of gastric acid secretion Adenoma Guanadrel Antihypertensive Adenoma Lacidipine Antihypertensive Adenoma Vinclozolin Fungicide Adenoma Procymidone Fungicide Adenoma Oxlolinic acid Antimicrobial Adenoma Felbamate Anticonvulsant Adenoma Oxazepam Anxiolytic Adenoma

(Derived from Cook et al., 1999)

Martinez et al., 1983a,b). Flut^ide is a potent non-steroidal antiandrogen that induces Leydig cell adenomas in rats following one year of treatment (Oradell, 1995b). This is associated with elevated circulating LH and testosterone levels and reductions in prostate and seminal vesicle weights (Viguier-Martinez et al, 1983a,b). Other examples of compounds with anti-androgenic activity that produce LCTs in rats

22 include cimetidine (Brimblecombe & Leslie, 1984), linuron (Cook et al., 1993) and bicalutamide (reviewed by Iswaran et al, 1997).

5a~reductase inhibitors. 5a-reductase inhibitors, such as finasteride, block the conversion of testosterone to the more potent androgen, dihydrotestosterone (DHT) (Prahalada et al, 1994; Oradell, 1995c). This reduces the net androgenic feedback signal to the hypothalamus and pituitary gland leading to a compensatory increase in LH secretion (Prahalada et al., 1994). Chronic treatment (83 weeks) of mice with finasteride has been reported to increase the incidence of Leydig cell hyperplasia and adenomas (Prahalada et al., 1994). In short term studies (5 to 14 weeks), a positive correlation was observed between the increased incidence of Leydig cell hyperplasia and the increase in serum LH levels in finasteride-treated mice (Prahalada et al., 1994).

Aromatase inhibitors. Aromatase inhibitors, such as formestane and letrozole, block the conversion of testosterone to oestradiol resulting in a decrease in serum oestradiol levels and a consequent increase in LH secretion (Junker-Walker & Nogues, 1994). In subchronic studies (3 months), formestane and letrozole produced Leydig cell hyperplasia in beagle dogs but not rats (Junker-Walker & Nogues, 1994). It was postulated that these species differences might have been due to the more prominent role of oestradiol in the feedback regulation of LH secretion in dogs compared to rodents (Junker-Walker & Nogues, 1994).

Dopamine agonists. Dopamine agonists or agents that enhance dopamine levels reduce serum prolactin levels. This may lead to downregulation of Leydig cell LH receptors resulting in a partial inhibition of testosterone biosynthesis (Prentice at al, 1992). This would provoke a compensatory increase in LH secretion. Alternatively, dopamine agonists may increase serum LH levels through direct effects on the hypothalamus or pituitary gland (Yamada et al., 1994b).

Oxolininc acid is an antimicrobial agent that increases the incidence of Leydig cell hyperplasia and tumours in rats (Yamada et al., 1994a). Dietary exposure to oxolinic acid is associated with reduced serum prolactin levels and a significant increase in LH

23 secretion (Yamada et al., 1994b). Co-treatment of rats with the dopamine receptor

(D2) antagonist, haloperidol, completely blocked the oxolinic acid-induced increase in LH levels (Yamada et al., 1994b). These findings suggest that oxolinic acid may induce LCTs through facilitation of the hypothalamic dopaminergic system leading to an increase in LH secretion. Other examples of dopamine agonists known to produce LCTs in rats include mesulergine (Dirami et al., 1996) and norprolac (Roberts et al., 1993).

GnRH agonists. GnRH agonists, such as buserelin, histrelin and leuprolide, induce LCTs in rats by binding to Leydig cell GnRH receptors (Donaubauer et al., 1987; Oradell, 1995d; Oradell, 1995e; Hunter et al., 1982). Low doses of GnRH agonists have similar effects to LH and may stimulate the Leydig cell directly or by enhancing LH secretion from the pituitary gland (Hunter et al., 1982). This mode of tumour induction appears to be a specific to rats because mice, monkey and human Leydig cells do not express GnRH receptors (Wang et al., 1983; Mann et al., 1989; Clayton & Huhtaniemi, 1982).

Oestrogen agonists. Oestrogen agonists, such as diethylstilbestrol (DES) and ethinyl estradiol, appear to induce LCTs in certain strains of mice through their ability to elevate circulating LH levels (Baroni et al., 1966; Huseby, 1976; Yasuda et al., 1988). In mice, oestrogens increase pituitary LH secretion and also have a direct carcinogenic effect on the Leydig cells (Huseby, 1980). Oestrogen agonists may also induce LCTs by increasing prolactin levels, which would stimulate LH secretion in mice (Klemcke & Bartke., 1981; Chandrashekar et al., 1991). Species differences in susceptibility to LCT induction by oestrogen agonists have been illustrated in studies using DES, which is a model compound used for the induction of LCTs in mice (Baroni et al., 1966). In contrast, the hyperprolactinaemia resulting from DES treatment is associated with a reduction in the incidence of LCTs in Fischer 344 rats (Bartke et al., 1985).

The role of oestrogens in the development of LCTs in rats is currently a matter of debate. Oestrogen agoinists do not induce LCTs in rats at dosages that cause testicular atrophy, but this condition may interfere with the detection of the Leydig cell lesions (Gibson et al., 1967; reviewed by Biegel et al., 2001). There is some evidence to

24 suggest that oestradiol may enhance Leydig cell tumorigenesis in rats. The high spontaneous incidence of LCTs in Fischer 344 rats is associated with an age-related increase in serum oestradiol levels (Turek & Desjardins, 1979). In contrast, serum oestradiol levels show an age-related decrease in CD rats, which have a low spontaneous incidence of LCTs (Biegel et al., 2001). Further research is required to clarify the role of oestradiol in LCT induction in rats.

Peroxisome proliferators. Peroxisome proliferator is the term used to describe a diverse range of compounds that cause peroxisome proliferation, induction of peroxisomal enzymes and hepatocellular carcinomas in rats (reviewed by Gibson & Lake, 1993). A number of these compounds also induce LCTs, including ammonium perfluorooctanoate (C 8 ) (Biegel et al., 2001; Sibinski, 1987), clofibrate (Oradell, 19951) and Wyeth-14,643 (Biegel et al., 2001). These compounds do not appear to induce peroxisomes in the Leydig cell, suggesting a distinct mechanism of tumour induction in the testis (Biegel et al., 1992; Cook et al., 1994; Mori et al., 1980).

The mechanism of LCT induction has been investigated for two peroxisome proliferators, C 8 and Wyeth 14,643. These compounds increase hepatic aromatase activity and produce a sustained increase in circulating oestradiol levels, which is correlated with the potency of LCT induction (Biegel et al., 2001; Liu et al., 1996; Biegel et al., 1995a). Elevated serum oestradiol levels are associated with increased testicular hormone levels, which might modulate testicular growth factor expression to produce LCTs (Biegel et al., 1995a; Cook et al., 1992). Consistent with this hypothesis, C 8 treatment was associated with an increase in testicular transforming growth factor alpha (TGFa) levels in rats (Biegel et al., 1995a). In addition, peroxisome proliferators directly modify Leydig cell fimction in vitro, reducing basal and LH-stimulated testosterone production (Liu et al., 1996).

1.3.6 Human Relevance

The common occurrence of Leydig cell hyperplasia and tumours during rodent toxicity studies with new therapeutic agents demands careful assessment of the risk presented to man. The significance of this effect to humans is currently unknown but there is evidence to suggest that men might be less susceptible than rodents to this

25 drug-induced lesion. First, there are marked species differences in the spontaneous incidence of LCTs. In contrast to the common occurrence of LCTs in rodents, the age-adjusted human incidence is 0.4 per million and the tumours are usually benign (Gilliland & Key, 1995). The magnitude of the difference in tumour incidence may be partially exaggerated by the different methods of diagnosis, usually by palpation in humans compared to detailed histological examination of the testis in laboratory animals. This may tend to underestimate the true incidence in the human population because small, impalpable LCTs would not be detected.

Second, physiological differences between rats and humans appear to confer a differential sensitivity of the Leydig cells to elevated circulating LH levels. The following factors might help to explain the apparent higher susceptibility of the rat Leydig cell to tumour formation secondary to hormonal imbalance. First, rats express a greater number of LH receptors per Leydig cell than humans, which may confer a greater sensitivity to circulating LH (Huhtaniemi, 1983). This is consistent with the finding that rat and human Leydig cells respond differently to exogenous hCG, showing a hyperplastic or hypertrophic response respectively (Christensen & Peacock, 1980; Heller & Leach, 1971). Second, rat Leydig cells express GnRH receptors, which have not been detected on mouse, monkey or human Leydig cells (Hunter et al., 1982; Wang et al., 1983; Mann et al., 1989; Clayton & Huhtaniemi, 1982). Low levels of GnRH may stimulate the Leydig cell directly or indirectly by enhancing LH secretion from the pituitary gland (Hunter et al., 1982). Third, prolactin is necessary for the maintenance of Leydig cell LH receptor numbers in rats but not humans (Zipf et al., 1978a; Walhstrom et al., 1983). Finally, humans express a high affinity sex hormone binding globulin (SHBG), which binds circulating testosterone retarding its metabolism and clearance (reviewed by Moore & Bulbrook, 1988; Dunn et al., 1981). In contrast, rats lack a high affinity circulating SHBG and consequently circulating testosterone levels may be perturbed more readily (Tenniswood et al., 1982).

The apparent species differences in Leydig cell responsiveness to LH are supported by data from certain human endocrine diseases. One example is androgen insensitivity syndrome (AIS) in which the body is unable to recognise androgens due to defective androgen receptor function (reviewed by Quigley et al., 1995). Affected individuals have elevated oestradiol and LH levels, normal or elevated testosterone and FSH

26 levels, and an increased incidence of testicular tumours (germ, Sertoli and Leydig cell origin). One group reported the frequent occurrence of Leydig cell hyperplasia and a tumour incidence of 2.3% in men with AIS (Rutgers & Scully, 1991). Cook et al (1999) contrasted this to the high incidence of LCTs in rats treated with the androgen receptor antagonist, flutamide (approaches 1 0 0 %), and concluded that this provided further evidence for the lower sensitivity of humans to induction of LCTs.

Data from epidemiology studies have reported no increase in the incidence of LCTs in humans despite widespread exposure to compounds such as 1,3-butadiene, clofibrate, lactose arid nicotine, which all induce LCTs in rats (reviewed by Cook et al., 1999). However, Cook et al. (1999) emphasised that relatively insensitive methods were used to detect tumours (e.g. palpation, abnormal endocrine profiles) and human exposure is generally much lower than the dosages used in rodent bioassays. Similarly, post marketing surveillance for therapeutic agents that induce LCTs in rodents (e.g. flutamide) have reported no increased incidence of these tumours in man.

In conclusion, most of the available evidence suggests that human Leydig cells are less sensitive than rat cells in their proliferative response to LH, but the possibility of LCT induction by therapeutic agents cannot be ruled out. Clegg et al. (1997) concluded that the occurrence of LCTs in rodent bioassays with new therapeutic agents is of concern unless there is sufficient mechanistic data regarding the mode of induction to conclude that it is not relevant to humans. By this principle, induction of LCTs by GnRH agonists and dopamine agonists are considered to present no risk to men because humans Leydig cells do not express GnRH receptors and prolactin is not permissive for LH receptor function in humans (Clegg et al., 1997; Clayton & Huhtaniemi, 1982). For the remaining modes of induction there is currently insufficient information available to rule out their relevance to humans. Further research is therefore required to improve the detection of LCTs in humans, to further investigate modes of induction and to clarify the mechanistic and sensitivity differences between species (Clegg et al., 1997).

27 1.4 Cytochrome P450 and Testosterone Metabolism

1.4.1 Cytochrome P450

Cytochrome P450 (CYP) enzymes constitute a superfamily of haemoproteins that catalyse the biotransformation of a diverse range of endogenous and exogenous compounds which include drugs, pesticides, carcinogens, food additives, environmental contaminants, steroid hormones, vitamins and fatty acids (reviewed by Lewis, 2001).. CYPs function as monooxygenases, which catalyse the general reaction shown below:

, Cytochrome P450 , . NADPHH"* + O 2 + RH ______► NADP^ + H 2 O + ROH where RH is the substrate and ROH is the hydroxylated metabolite. During this reaction NADPH H^ supplies two reducing equivalents and molecular oxygen is cleaved such that one oxygen atom is inserted in to the substrate while the other is reduced to water.

CYPs are found in every biological kingdom, including animals, plants and bacteria (Nebert et al., 1989). In mammalian species, CYPs are most abundantly expressed in the liver, but have been detected in virtually all tissues studied including the brain, lung, kidney, intestine, spleen, pancreas and gonads (reviewed by Lewis, 2001). These enzymes are membrane bound and are mainly localised in the endoplasmic reticulum, although certain forms are found in the mitochondria.

The CYP superfamily is composed of multiple forms, each showing distinct and sometimes overlapping substrate specificities, which are classified in to a number of gene families (members share over 40% sequence similarity) (e.g. CYPl to 3) and subfamilies (members share over 70% sequence similarity) (e.g. CYP2A to Q). Individual members of a subfamily are assigned a number (e.g. CYP2D6). The general functions of the different CYP families are shown in Table 1.4.

28 Table 1.4: General functions of the cytochrome P450 gene families

CYP Family Function

C Y P1-C Y P3 Drug, xenobiotic and steroid metabolism CYP4 Fatty acid and prostaglandin metabolism CYP5 Thromboxane synthesis CYP7 Cholesterol 7a-hydroxylation CY Pll Cholesterol side chain cleavage CYP17 Steroid 17 a-hydroxylation CYP19 Aromatisation of steroids CYP21 Steroid 21-hydroxylase CYP24 Vitamin D hydroxylation CYP27 Cholesterol 27-hydroxylase

(Adapted from Gibson & Skett, 2001)

Many CYPs are expressed constitutively (i.e. in the absence of an external stimulus) but certain forms can be reversibly induced following exposure to xenobiotics. Induction of the synthesis of a pre-existing or previously unexpressed enzyme may result in qualitative and quantitative changes in hepatic metabolism. Mechanisms involved in enzyme induction include transcriptional activation (e.g. CYPl Al, 2B1/2, 3A), post-transcriptional effects (e.g. stabilisation of CYP1A2 mRNA) and post- translational effects (e.g. reduced degradation of the CYP2E1 protein) (reviewed by Gibson & Skett, 2001). Conversely, CYP enzyme expression may be suppressed by certain xenobiotics (Yeowell et al, 1987; Waxman et al., 1984). Expression of CYP enzymes is also influenced by endogenous factors such as species, age, sex, and hormones (reviewed by Gibson & Skett, 2001).

CYPs catalyse key reactions in the biosynthesis and catabolism of steroid hormones, and consequently play an important role in steroid hormone homeostasis in mammalian systems. The CYP enzymes involved in testosterone biosynthesis are markedly different to those associated with catabolic reactions in that they tend to possess specific substrate specificities and usually produce a single steroid metabolite (reviewed by Zimniak & Waxman, 1993).

29 1.4.2 Testosterone Biosynthesis

The majority of testosterone biosynthesis occurs within the Leydig cells of the testis, which have been estimated to secrete over 95% of the circulating hormone after puberty (Christensen & Mason, 1965; Hall et ah, 1969). Testosterone biosynthesis involves a sequence of reactions, the majority being catalysed by CYP enzymes (Figure 1.6).

Testosterone is synthesised from the common steroid precursor, cholesterol, which may be derived from three sources: intracellular lipid droplets, de novo synthesis from acetate or circulating lipoproteins (Hall, 1994). Transfer of cholesterol to the inner mitochondrial membrane forms the rate-limiting step in steroid biosynthesis and is primarily regulated by LH (Crivello & Jefcoate, 1980; Privalle et al., 1983; Hall et al., 1979). Several proteins have been postulated to play a role in cholesterol transport, including sterol carrier protein 2 (SCP-2), steroidogenesis activating polypeptide (SAP) and the peripheral benzodiazepine receptor (PDR) (Yamamoto et al., 1991; Pedersen & Brownie, 1987; Papadopoulos, 1998). More recently a protein named steroidogenic acute regulatory protein (StAR), which transports cholesterol between the outer and inner mitochondrial membranes, has been purified and cloned (Clark et al., 1994). StAR is expressed almost exclusively in steroidogenic tissues and its presence is correlated with steroid hormone production (Sugawara et al., 1995; Clark et al., 1995; Luo et al., 2001). LH regulates transcription of the StAR gene, leading to increases in mRNA and protein levels (Clark et al., 1995; Caron et al., 1997; Luo et al., 1998; Clark et al., 1994). StAR gene expression is also influenced by aiitocrine and paracrine factors such as IGF-1, TGFP and tumour necrosis factor alpha (TNFa) (reviewed by Christensen & Strauss, 2000).

Conversion of cholesterol to pregnenolone is catalysed by the mitochondrial side- chain cleavage complex, CYP450scc (product of the C Y P llA l gene), which forms the rate-limiting enzymatic step during steroid biosynthesis (reviewed by Simpson, 1979). CYPscc expression is confined to steroidogenic tissues and in the testis is localised exclusively to the Leydig cells (O’Shaugnessy & Murphy, 1991). Once formed, pregnenolone rapidly enters the smooth endoplasmic reticulum where subsequent reactions occur. Two parallel pathways produce testosterone from pregnenolone, the

30 as

I 00

% 1 I I i I I % I0) I B§ f I I I 1

CO.I IS. t CQ.i 8 1 I T3t

Î03

\o Î CQ.I I CO I !O CO. I fCO relative importance of each being species-dependent. In humans the major pathway of testosterone biosynthesis is the A5-pathway, whereas the A4-pathway predominates in the rat (Weusten et al, 1987). Four principal steroidogenic enzymes are localised in the endoplasmic reticulum: 3 |3-hydroxysteroid dehydrogenase (3p-HSD), CYP17A1,

17p-hydroxysteroid dehydrogenase (17p-HSD) and CYP19A1 (Reviewed by Saez,

1994). CYP17A1 is a bifunctional protein supporting both 17a-hydroxylase and C l7-20 lyase activity, which converts progesterone to androstenedione (Nakajin et al., 1981). In the testis, CYP17A1 expression is confined to the Leydig cells and is regulated by LH (O’Shaugnessy & Murphy, 1991).

17p-HSD catalyses the reversible interconversion of androstenedione and testosterone

(reviewed by Andersson & Moghrabi, 1997). Several distinct 17P-HSD enzymes have been identified, which differ in their substrate specificity, cofactor requirements, preference for oxidative or reductive reactions and subcellular and tissue distribution (reviewed by Peltoketo et al., 1999). This enzyme plays a key role in the formation of active androgen and is present in gonadal and extragonadal tissues (e.g. liver, prostate, lung, testis and brain) (Martel et al., 1992).

CYP19A1 catalyses the conversion of androgens to oestrogens, under the influence of LH (Valladares & Payne, 1981). CYP19A1 is primarily located in steroidogenic tissues, but activity is also present in adipose tissue, liver and skeletal muscle (Lephart & Simpson, 1991; Longcope et al., 1978). CYP19A1 exhibits an age-dependent cellular distribution in the testis, with activity predominantly localised to the Sertoli cells in immature rats whereas the Leydig cells are the major site of oestradiol production in mature animals (Tsai-Morris et al., 1985; Rommerts et al., 1982). Oestrogens have an inhibitory effect on 17a-hydroxylase activity in adult rat Leydig cells, which may be involved in the autocrine regulation of testosterone biosynthesis (Onoda & Hall, 1981; Kalla et al., 1980).

Steroid biosynthesis requires the controlled expression and activity of the enzymes and other proteins involved in this pathway. Testosterone biosynthesis is dependent upon LH stimulation, which exerts both acute and chronic effects on the Leydig cell (reviewed by Saez, 1994). The acute steroidogenic response occurs within minutes of

32 LH exposure and involves increased translocation of cholesterol to the inner mitochondrial membrane (Hall et al., 1979). Chronic trophic stimulation of the Leydig cell is required for optimal expression of steroidogenic enzymes (Wing et al., 1984; O’Shaughnessy & Payne, 1982; Purvis et al., 1973). LH increases the mRNA, protein and activity of StAR, 3p-HSD, CYPscc and CYP17A1 (Lejeune et al., 1998; Keeney & Mason, 1992; Anakwe & Payne, 1987; Payne et al., 1992). Cultured Leydig cells show high basal expression of CYPscc, although maximal levels of expression require LH stimulation (Anakwe & Payne, 1987; Payne et al., 1992, Anderson & Mendleson, 1985). In contrast, CYP17A1 expression is absolutely dependent upon LH stimulation (Anakwe & Payne, 1987; Payne et al., 1992). Intratesticular autocrine and paracrine factors may also modulate the expression and activity of steroidogenic enzymes (Saez, 1994).

1.4.3 Testosterone Catabolism

1.4.3.1 Introduction

In humans, testosterone becomes loosely bound with plasma albumin or more tightly bound with sex hormone binding globulin (SHBG), such that a relatively small percentage of the circulating hormone exists in an unbound form (Dunn et al., 1981). In contrast, male rats lack a high affinity circulating SHBG and consequently virtually the entire circulating hormone is loosely associated with plasma albumin (Tenniswood et al., 1982). This might represent an important physiological difference between the species as plasma protein binding may modulate the biological availability of testosterone, thus influencing its effects on target tissues and the rate of metabolism and clearance (Hobbs et al., 1992).

The 5a-reductase enzyme (5a-R) converts a small proportion of the circulating testosterone to the more potent androgen, DHT (Moore & Wilson, 1972; Zhou et al., 1995). This occurs primarily in peripheral tissues, but enzyme activity is also present in the Leydig cells of the testis (Normington & Russell, 1992; 0 ’Shaugnessy & Murphy, 1991). In rats, two genes encoding 5a-R have been identified: 5a-Rl (major form found in the liver) and 5a-R2 (major form expressed in reproductive tissues, such as the prostate) (Normington & Russell, 1992). Hepatic 5a-R activity exhibits sex-specific developmental regulation with activity increasing at puberty in females

33 only, such that liver microsomes from adult male rats possess low 5a-R activity (Sonderfan & Parkinson, 1988). Testicular 5a-R activity shows a similar pattern of expression with high activity in immature animals, which progressively declines to reach the low levels observed in during adulthood (Lacroix et al., 1975).

A number of pathways contribute to the hepatic biotransformation of circulating testosterone to form inactive or less active metabolites, which are excreted in the urine or bile. First, testosterone is hydroxylated at a number of different positions in reactions catalysed by CYP enzymes (reviewed by Waxman, 1988). Second, testosterone is conjugated with glucuronic acid or sulfate, either directly or following hydroxylation, in reactions catalysed by UDP-GT and ST enzymes respectively (Matsui et al., 1974). These reactions generally result in the formation of more polar products, facilitating excretion in the urine or bile. Finally, testosterone may be dehydrogenated to form androstenedione through pathways catalysed by CYPs and 17p-HSD (reviewed by Waxman, 1988; Andersson & Moghrabi, 1997). Testosterone is metabolised by intact rat parenchymal cells to form the following major metabolites: androstenedione (46% total metabolites), 16a-0HT (21.5%), 2a-0HT

(18.5%) and 6p-0HT (9.5%) (Utesch et al, 1992). Waxman (1992) suggested that the physiological importance of steroid hydroxylation reactions might not only involve hormone deactivation. Testosterone metabolites may also possess unique biological or endocrine activities, although the physiological functions of the individual hydroxylated metabolites are largely unknown. There is however evidence to support a role for 7a-hydroxylated androgens in the regulation of testicular testosterone biosynthesis (Inano et al., 1973; Rosness et al., 1977; Mittler, 1985) (see 1.5).

Various CYPs hydroxylate testosterone with a high degree of regio- and stereospecificity (Figure 1.7 & Table 1.5). Certain steroid hydroxylation reactions are predominantly catalysed by a single CYP enzyme (e.g. CYP2C11 at the 2a-position) whilst others involve multiple forms (e.g. 16a-hydroxylation catalysed by CYP2B1, 2B2 and 2C11) (Waxman, 1988). The hydroxysteroid metabolite profile can be used as a diagnostic marker to monitor the relative concentrations of individual CYP enzymes in hepatic microsomes (Waxman, 1988). For example, testosterone 7a-,

34 CO. VO VO CO. I/) Ü Ü

VO

I u G s '

CO I CO

Ï VO (S CO I oo U U 5 O 5 I I I o e Î 'O I i

g CO CO

I0 ! 1 § 1) g ê % 5 u I 01 W Table 1.5: Properties of the major rat hepatic cytochromes P450 that catalyse testosterone metabolism.

CYP Site(s) of Constitutive/ Inducers Gender % Hepatic Testosterone Inducible Selectivity P450 t Metabolism

lA l 6p I 3-MC, None 1 PNF

1A2 6p C 3-MC, Female <5 pNF dominant

2A1 7a*, 6a C PB, 3-MC Female 12 dominant

2A2 15a*, 6p, 7a, 16a c - Male 3.5 specific

2B1 16a, 16P, 17 I PB Male <1 dominant

2B2 16p, 16a c PB Male <1 dominant

2C6 2a, 16a, 17 c PB None 2 0 -3 0

2C7 16a c Ethanol Female 15 dominant

2C11 16a*, 2a*, 6p, 17 c - Male 40 specific

2C12 15P* c - Female 27 specific

2C13 6P, 15a, 16a c - Male 23 specific

3A1 6P*, 2P, 15P I PB, PCN Male <20 dominant

3A2 6P,2P c PB, PCN Male Unknown specific

*, Almost all (> 85%) of the indicated metabolite is' produced by this form; 3-MC, 3- methylcholanthrene; PNF, P-naphthoflavone; PB, phénobarbital; PCN, pregnenolone-16a-carbonitrile; t, based on immunological quantification of control rat liver. (Adapted from Lewis, 2001 & Lewis, 1996).

36 16p- and 2a-hydroxylase activities reflect the levels of CYP2A1, 2B1/2 and 2C11 respectively (Sonderfan et al, 1987).

1.4.3.2 Regulation of Steroid Hormone Hydroxylation

In rodents, numerous studies have demonstrated that the relative rate of steroid hormone hydroxylation at each position is influenced by the age, sex and monooxygenase induction status of the animal (Waxnian et al, 1988a; Waxman et al., 1985). This relationship exists because the levels of expression of individual CYPs can be altered dramatically during development, by hormonal status and following exposure to xenobiotics (Waxman et al, 1988a; Waxman et al., 1985).

Age- and Sex-Dependent Regulation

Age- and sex-dependent regulation of CYP expression, and the concomitant changes in steroid hormone metabolism, maintains the appropriate levels of circulating hormones in male and female animals at various stages of development (reviewed by Kedderis & Mugford, 1998). In general, male rats display higher rates of metabolism of xenobiotic and steroid substrates compared to females (reviewed by Kedderis & Mugford, 1998).

Several CYPs that catalyse testosterone metabolism show sex-specific developmental regulation in rats (Table 1.5). CYP2A2 and 2C11 are male specific forms due to induction of gene expression in male rats only at puberty (Imaoka et al., 1991; Waxman et al., 1988a; Waxman et al., 1985). This underlies the male specific post- pubertal rise in testosterone 2a- and 16a-hydroxylase activity (Waxman et al., 1985). Similarly, the female specific form, CYP2C12, is induced at puberty in females only (Imaoka et al., 1991). CYP2A1 expression is activated in male and female rats at birth, but is subsequently suppressed to low levels in males at the onset of puberty (Imaoka et al., 1991; Waxman et al., 1985). CYP3A2 is expressed in both sexes at low levels after birth but expression is suppressed in females at puberty, in parallel to the loss of steroid 6|3-hydroxylase activity (Imaoka et al., 1988; Waxman et al., 1985).

37 Sex-dependent expression of CYPs is complex and is regulated, at least in part, by androgens. In rats, exposure to critical levels of androgen during the neonatal period “imprints” the expression of specific CYP fonris in the adult animal (e.g. CYP2C11 and 3A2) (Gustafsson et al., 1983; Waxman et al., 1985). During adulthood, the presence of circulating androgens may be required to maintain full expression of certain CYP forms (e.g. CYP2C11) (Waxman et al., 1985). The ability of androgens to regulate CYP expression does not involve a direct effect on the liver, but is related to the ability of androgens to modulate growth hormone (GH) secretion from the pituitary gland (reviewed by Kedderis & Mugford, 1998). In many species, including rats and humans, GH secretory patterns are sexually differentiated during adulthood (Winer et al., 1990; reviewed by Jansson et al., 1985b). Secretion of GH is pulsatile in male rats (peaks every 3 to 4 hours with low levels of GH between) whereas females show a more continuous secretion (irregular peaks with high inter peak GH levels) (reviewed by Jansson et al., 1985b). These sex-dependent GH secretory patterns appear to regulate CYP expression. For example, pulsatile secretion of GH stimulates the expression of the male specific form, CYP2C11 (Waxman et al., 1991a).

Effects o f Xenobiotic Exposure

The levels of individual CYPs show marked changes following exposure to compounds that induce or suppress their expression, with concomitant effects on hepatic steroid metabolism. Monooxygenase enzyme inducers may influence steroid metabolism in several ways including: induction of a new hydroxylation pathway (e.g.~25 fold induction of CYP2B1-mediated 16|3-hydroxylation by PB), enhancement of an existing pathway (e.g. 3-4 fold increase in CYP3A2-mediated 6 p-hydroxylation by pregnenolone-16a-carbonitrile (PCN)) or induction of pathways that are usually subject to endocrine control (e.g. ~15 fold induction of CYP3A- mediated 6 p~ hydroxylation in female rats upon exposure to PCN) (Waxman el al., 1988a; Waxman et al., 1985). Inducers such as PB and dexamethasone increase the total rate of hydroxylated metabolite formation by liver microsomes (Purdon & Lehman- McKeeman, 1997; Parkinson et al., 1992; Sonderfan et al., 1987). Administration of N-phenylbarbital to humans has no effect on the plasma clearance rate of testosterone but induced a change in the pattern of metabolites excreted in the urine (Southren et al., 1969).

38 Suppression (up to 70 to 80%) of constitutive hepatic CYPs has been reported following treatment of rats with several classical monooxygenase enzyme inducers (Waxman, 1984). Selective repression of hepatic CYP2C11 protein and catalytic activity (e.g. 16a- and 2 a-hydroxylase activity) has been demonstrated following treatment of male rats with compounds such as 3-methylcholanthrene (3-MC), PB, PCN, dexamethasone and p-naphthoflavone (p-NF) (Waxman, 1984; Yeowell et al, 1987). 3-MC treatment also produces an initial suppression of CYP3A2 catalytic activity (testosterone 6 p-hydroxylation) and protein levels, which is followed by an increase in expression above control levels (Jones & Riddick, 1996).

The quantitative and qualitative effects of many compounds on the hepatic CYP complement have been well characterised. In contrast, the physiological consequences of the concomitant changes in steroid hormone metabolism are poorly understood. These secondary changes may have an important impact on steroid hormone homeostasis within the animal, with potential implications for steroid hormone regulated processes, such as spermatogenesis.

1.4.3.3 Testosterone Metabolism by Rat Liver Microsomes

The following section will briefly describe the major CYP forms known to catalyse hepatic testosterone hydroxylation reactions in rats (Table 1.5 and Figure 1.7).

CYPIA

CYPlAl and 1A2 are expressed at low levels in the livers of male and female rats, but are extensively induced following exposure to compounds such as 3-MC, P-NF, benzo(a)pyrene, 2,3,7,8-tetrachloro-dibenzo-p-dioxin (TCDD) and other polycyclic aromatic hydrocarbons (Thomas et al., 1983).

Purified CYPlAl and 1A2 both show testosterone 6 p-hydroxylase activity in reconstituted systems but appear to make a relatively modest contribution to the overall rate of 6 p-hydroxylation catalysed by rat liver microsomes (Sonderfan et al., 1987; Wood et al., 1983). Treatment of immature Long Evans rats with 3-MC produces a dramatic increase in hepatic CYPlAl content (>70% total CYP content), however this is associated with a 60% reduction in the rate of 6 P-hydroxylation

39 (Wood et al, 1983). Studies employing anti-CYPlA antibodies indicate that CYPIA catalyses less than 1 0 % of the microsomal testosterone 6 p-hydroxylase activity of intact liver microsomes from untreated and inducer-treated rats (PB, 3-MC and Aroclor 1254) (Wood et al, 1983). Purified CYPlAl also supports testosterone 2p- hydroxylation, albeit at a very modest rate (Sonderfan et al., 1987).

CYP2A

CYP2A1 and 2A2 share approximately 8 8 % amino acid sequence homology but possess markedly different regioselectivities for testosterone hydroxylation and their expression is differentially regulated (Matsunaga et al., 1988). CYP2A1 is weakly induced by compounds such as PB, p-NF and 3-MC whereas CYP2A2 appears to be relatively refractory to induction (Waxman et al., 1985; Nagata et al., 1987; Thomas et al., 1981; Matsunaga et al., 1988). In fact, CYP2A2 expression is suppressed (50 to 80%) following exposure to PB, p-NF, 3-MC and isosafrole (Waxman et al., 1988a; Thummel et al., 1988).

CYP2A1 shows narrow regioselectivity and metabolises testosterone to one major product, 7a-hydroxytestosterone (7a-0H T) (Waxman et al., 1990; Wood et al, 1983; Sonderfan et al, 1987). Incubation of liver microsomes from.untreated and inducer- treated rats (e.g. PB, 3-MC and dexamethasone) with anti-CYP2A antibodies completely abolishes microsomal testosterone 7a-hydroxylase activity (Levin et al.,

1987; Wood et al., 1983; Waxman et al., 1987). The formation of 7a-0H T can therefore be used as a diagnostic marker for the expression of CYP2A1 in liver microsomes. Purified CYP2A1 also shows some degree of testosterone 6 a- hydroxylase activity (~5% the rate of 7a-hydroxylation), which is abolished following incubation of liver microsomes with anti-CYP2Al antibody (Sonderfan et al., 1987; Wood et al., 1983).

In contrast, CYP2A2 hydroxylates testosterone with little regio- or stereoselectivity to produce a number of metabolites (e.g. 6 P-, 7a-, 16a-, 15a- and 15p-0HT) (Jansson et al., 1985a; Thummel et al., 1988; Arlotto et al., 1989). The major metabolite formed by purified CYP2A2 is 15a-0HT, which accounts for between 40 and 50% of the

40 total testosterone metabolism catalysed by this form in reconstituted systems (Waxman et ah, 1990; Jansson et al, 1985a). Polyclonal anti-CYP2A antibodies (cross-react with CYP2A1 and 2A2) inhibit approximately 75% of the microsomal 15a-hydroxylase activity whereas a monoclonal anti-CYP2Al (selective for CYP2A1) had little effect on this activity, indicating that CYP2A2 is the major testosterone 15a- hydroxylase present in liver microsomes (Waxman et al., 1988a). Formation of 15a- OHT is therefore a useful marker for the levels of CYP2A2 in liver microsomes.

CYP2B

CYP2B1 and 2B2 enzymes share 97% amino acid sequence homology, but show differential expression and regulation (Yuan et al., 1983). CYP2B1 and 2B2 are expressed at low levels in untreated male rats but are extensively induced following exposure to the prototype inducer, PB (Waxman & Azaroff, 1992). PB treatment causes marked induction of CYP2B1 (up to 40 fold) and a smaller increase GYP2B2 levels (Omiecinski, 1986; Waxman & Azaroff, 1992). Other CYP2B inducers include DDT, chlordane, dexamethasone and isosafrole (Lubet et al., 1992; Thomas et al., 1981; Wortelboer et al., 1991; Waxman et al., 1985).

CYP2B1 and 2B2 appear to lack specific stereoselectivity for testosterone hydroxylation but oxidation sites are limited to the 16- and 17- positions (Wood et al, 1983). CYP2B1 catalyses testosterone 16a- and 16p-hydroxylation in addition to oxidation at the 17-position to form androstenedione (Wood et al., 1983; Sonderfan et al, 1987). These oxidation products are produced in approximately equal amounts (Wood et al, 1983; Sonderfan et al., 1987). CYP2B2 metabolises testosterone to form the same major metabolites as CYP2B1 but at a much slower rate (-10% of the turnover rate of CYP2B1) (Wood et al., 1983). Studies using anti-CYP2B antibodies revealed that CYP2B enzymes catalyse the majority (85 to 95%) of the steroid 16p- hydroxylase activity present in liver microsomes from PB-treated rats (Wood et al., 1983). Similar levels of inhibition were observed for microsomal testosterone 16a- hydroxylase activity and androstenedione formation (Wood et al., 1983). However, all of these reactions were inhibited to a markedly reduced extent (< 2 0 % total activity) when anti-CYP2B antibodies were incubated with control liver microsomes.

41 suggesting that other CYP forms catalyse these reactions in untreated animals (Wood et a l, 1983; Thomas et ah, 1981; Waxman et ah, 1987).

CYP2C

CYP2C is a large subfamily of enzymes that support a range of testosterone hydroxylase activities (Schenkman et ah, 1989). In general, CYP2C forms are constitutively expressed and show sex-specific developmental regulation in the rat (Schenkman et ah, 1989; reviewed by Waxman, 1991).

CYP2C11 is abundantly expressed in the liver and catalyses testosterone 16a- and 2a- hydroxylation in addition to oxidation at the 17-position to form androstenedione (Sonderfan et ah, 1987). Studies employing anti-CYP2Cl 1 antibodies have demonstrated that this enzyme supports the majority (>90%) of the microsomal testosterone 2a- and 16a-hydroxylase activity in untreated male rats (Morgan et ah, 1985; Waxman, 1984; Waxman et ah, 1987). However, treatment of adult rats with PB is associated with a reduction in CYP2C11-catalysed 16a-hydroxylation to approximately 30% of this pathway, with the remaining activity being catalysed by CYP2B1 (Waxman et al., 1987). The formation of 2a-0H T is used as a diagnostic marker for the expression of CYP2C11 in liver microsomes (Sonderfan et ah, 1987).

Purified CYP2C11 also hydroxylates testosterone at the 6 p-position, albeit at a very modest rate (Sonderfan et ah, 1987). Significant suppression of hepatic CYP2C11 expression (up to 70 to 80%) has been reported following treatment of rats with several prototype CYP inducers (e.g. PB, P-NF, PCN, 3-MC) and is associated with a reduction of microsomal testosterone 2a-hydroxylase activity (Waxman, 1984; Yeowell et al, 1987).

Other members of the CYP2C subfamily hydroxylate testosterone at several positions but with slower turnover rates than CYP2C11 (Table 1.5) (Schenkman et ah, 1989). Purified CYP2C6 metabolises testosterone to form the same major products as CYP2C11 (androstenedione, 16a- and 2a-0HT) and is induced following exposure to PB (~2 to 3 fold) (Schenkman et ah, 1989; Guengerich et ah, 1982; Omiecinski et ah, 1990b).

42 CYP2C7 is a female predominant form, which shows low testosterone 16a- hydroxylase activity (Schenkman et al., 1989). CYP2C12 is a female-specific testosterone 15P-hydroxylase, which appears refractory to induction by several prototype inducers (e.g. PB, PCN, isosafirole) and is suppressed following exposure to dexamethasone in rats (Waxman et al., 1985; Wright & Morgan, 1991).

CYP2C13 is a male-specific form, which appears refractor to induction by several classical inducers (Bandiera et al., 1986). Purified CYP2C13 primarily metabolises

testosterone at the 6 P-position, but also produces small amounts of 15a- and

16a-0HT along with traces of 7a- and 2a-0HT (McClellan-Green et al., 1987). Antibody inhibition studies indicate that CYP2C13 makes a relatively minor

contribution to the overall rate of testosterone 6 p-hydroxylation catalysed by rat liver microsomes (McClellan-Green et al., 1987).

CYP3A

A major pathway of steroid metabolism in both humans and rats involves

hydroxylation at the 6 p-position, which is primarily catalysed by CYP3A enzymes (Waxman, 1988; Zimniak & Waxman, 1993; Maenpaa et al., 1993). In the rat, five CYP3A genes have been identified: CYP3A1, 3A2, 3A9, 3A18 and 3A23 (Nelson et al., 1996). CYP3A1 and 3A2 proteins share 89% amino acid sequence similarity and

both support testosterone 6 p-hydroxylase activity (Gonzalez et al., 1986). Antibody inhibition studies have demonstrated that CYP3A2 catalyses over 85% of the total

testosterone 6 p-hydroxylation in liver microsomes from untreated rats (Waxman et al.,

1985). In contrast, CYP3A1 catalyses over 85% of the microsomal 6 P-hydroxylation in dexamethasone-treated rats (Gonzalez et al., 1986). Other CYP forms support

testosterone 6 p-hydroxylase activity in purified, reconstituted systems (e.g.CYPlAl,

2 A2 , 2C13) but appear to make a relatively modest contribution to the total rate of 6 p- hydroxylation catalysed by intact liver microsomes (Waxman et al., 1987; Wood et al.,

1983; McClellan-Green et al., 1987). In addition to testosterone 6 p-hydroxylation, purified CYP3A1 catalyses 2p- and 15p-hydroxylation and purified CYP3A2

catalyses 2P-hydroxylation (Sonderfan et al., 1987; Schenkman, 1992). The

43 combined production of 2p~, 6 p- and ISp-OHT reflects the combined levels of CYP3A1 and 3A2 in rat liver microsomes (Sonderfan et al., 1987).

CYP3A1 is expressed at low levels in liver microsomes from male rats, but is extensively induced following exposure to compounds such as dexamethasone, PCN, triacetyloleandomycin (TAO), rifampicin and isosafrole (Cooper et al., 1993; Debri et al., 1995; Waxman et al., 1985; Gonzalez et al., 1986). CYP3A2 is a constitutive enzyme that is induced by many of the compounds known to induce CYP3A1 (e.g. PB), albeit to a smaller extent (Cooper et al., 1993).

1.4.4 Testosterone Metabolism by Human Liver Microsomes

Testosterone metabolism catalysed by rat liver microsomes is characterised by hydroxylation at multiple sites and involves a number of CYPs belonging to the CYPl, 2 and 3 families (Sonderfan et al., 1987; Schenkman, 1992). In contrast, human liver microsomes metabolise testosterone to form one major product, 6p-0HT

(>7 5 % hydroxylated metabolites formed), and smaller amounts of androstenedione,

2p-, 15p, 2a- and 15a-0H T (Waxman et al., 1988b; Maenpaa et al, 1993; Draper et al., 1998). Hydroxylation at the 6 p-position is predominantly catalysed by CYP3A enzymes, which are abundantly expressed in human liver (-30% total CYP content) (Wang & Lu, 1997; Waxman et al., 1988b; Shimada et al., 1994). Hydroxylation of testosterone at the 2 P- and 15P- positions is correlated with 6 P-hydroxylase activity and specifically inhibited by anti-CYP3A antibodies, suggesting that these reactions are also primarily catalysed by CYP3A enzymes in human liver (Waxman et al., 1988b).

CYP3A4 is the major CYP3A form expressed in human adult liver and is induced following exposure to compounds such as rifampicin, phenytoin and dexamethasone (Shimada et al., 1994; Pichard et al., 1990). CYP3A5 is structurally and functionally related to CYP3A4 and is expressed in 10 to 20% of human livers (Aoyama et al., 1989). Certain human livers also express CYP3A7, which is otherwise considered to be a fetal form (Lacroix et al., 1997).

44 Purified CYP3A4 and 3A5 hydroxylate testosterone at the 6 p~position in reconstituted systems, but CYP3A4 appears to be the major testosterone 6 p-hydroxylase in human liver microsomes (Waxman et al., 1991b; Maenpaa et al, 1993; Wang & Lu., 1997). Hydroxylation reactions at the 2p-, 2a- and ISP- positions are also predominantly catalysed by CYP3A4 (Maenpaa et al, 1993). In contrast, androstenedione formation was not significantly inhibited by midazolam (CYP3A4 substrate) or anti-3 A4 antibodies, indicating that this reaction is predominantly catalysed by other CYP enzymes in human liver (Maenpaa et al, 1993).

CYP2C enzymes are abundantly expressed in human liver (-20% total hepatic CYP content) and certain forms are induced following exposure to compounds such as rifampicin (Shimada et al., 1994; Kay et al., 1985). Yamazaki and Shimada (1997) used recombinant CYP proteins to demonstrate that CYP2C19 metabolises testosterone to the major product androstenedione and the minor metabolites 6 P-, 16p- and 2p-0HT. Recombinant CYP2C9 also metabolises testosterone to form the same products but with a lower turnover rate (Yamazaki & Shimada, 1997). Incubation of human liver microsomes with anti-CYP2C9 (inhibits both CYP2C9 and CYP2C19) produced a marked inhibition of androstenedione formation (Yamazaki & Shimada, 1997). These findings suggest that CYP2C19, and to a lesser extent CYP2C9, are primarily responsible for metabolism of testosterone to form androstenedione in human liver microsomes.

45 1.5 Testicular Cytochrome P450 Enzymes

1.5.1 Introduction

CYPs are most abundantly expressed in the liver but are also found in a number of extrahepatic tissues, including the testis (Menard & Purvis, 1973). In addition to CYP enzymes involved in testosterone biosynthesis, CYPl, 2 and 4 family members have been detected in rat testis using a variety of techniques (e.g. enzyme assays, nucleic acid hybridisation, polymerase chain reaction (PCR), immunological detection) (e.g. Walker et al., 1995; Omiecinski, 1986; Seng et al, 1991). Expression of CYP enzymes along with epoxide hydrolases (EH) and various transferases (e.g. ST and UDP-GTs) indicates that the testis may mediate the local biotransformation of endogenous and exogenous compounds (Lee et al, 1980; Magnanti et al., 2000). The overall metabolic capacity will be dependent upon the qualitative and quantitative composition of the testicular enzyme complement. Expression of testicular CYP enzymes is regulated by anterior pituitary hormones (e.g. LH) and gonadal steroids (e.g. testosterone) (Waterman & Simpson, 1985; Otto et al., 1992; Lee et al, 1980). In general, testicular CYP enzymes appear to be less sensitive to exogenous chemical inducers than their hepatic counterparts (Goldstein & Linko, 1984; Omiecinski, 1986).

1.5.2 Testicular Metabolism of Exogenous Compounds

The mammalian testis appears to' be capable of metabolising a range of exogenous compounds (Thomas et al., 1987; Georgellis et al., 1987; Flowers et al., 1989). The Leydig cells, where a substantial proportion of the enzyme activity is localised, are in direct contact with substances entering the testis in the blood or lymph, whereas the germ cells are somewhat protected by the blood-testis barrier (Rommerts et al., 1982; Mukhtar et al., 1978). CYP enzymes play a crucial role in the detoxification of xenobiotics, but also “bioactivate” certain chemicals to form reactive metabolites. The contribution of the testis to the overall rate of xenobiotic metabolism is probably relatively small, but local formation of reactive metabolites may be an important determinant of the testicular toxicity of certain compounds.

7,12-dimethyl benz(a)anthracene (DMBA) is a potent carcinogen whose toxicity is correlated to the extent of biotransformation it undergoes within the target tissues

46 (Otto et al, 1992; Miyata et al., 1999). This compound exerts toxic effects at a number of sites including the testis, where it causes germ cell damage and sterility (Ford & Huggins, 1963). DMBA is metabolised by CYPs and microsomal epoxide hydrolases to form 3,4-dihydrodiol, which is a precursor for the ultimate carcinogenic and toxic metabolite, 3,4-dihydrodiol-1,2 epoxide (Georgellis & Rydstrom, 1989; Georgellis et al., 1987; Miyata et al., 1999). In the rat testis, DMBA bioactivation is predominantly localised to the Leydig cells and is stimulated by LH (Georgellis & Rydstrom, 1989; Lee et al., 1980). The CYP enzyme primarily responsible for the metabolic activation of DMBA in the rodent testis is CYPIBI (Otto et al., 1992; Buters et al., 1999). CYP IB 1 mRNA is constitutively expressed in the adrenal ^ d testis of untreated rats, where it may function as a physiological steroid hydroxylase (Walker et al., 1995). The crucial requirement for CYP IB 1 in DMBA carcinogenesis has been confirmed by the generation of CYP IB 1-knock out mice (Buters et al., 1999). CYP IB 1 null mice failed to show significant metabolism of DMBA and were resistant to the toxic effects of this compound (Buters et al., 1999).

The balanced expression of enzymes that catalyse the bioactivation and detoxification of DMBA may be an important factor influencing the ultimate toxicity of this compound. Furthermore, certain xenobiotic metabolising enzymes show differential expression in the various testicular cell types. For example, CYPs are predominantly found in the Leydig cells whereas EH and glutathione transferase activities are mainly localised to the seminiferous tubules (Mukhtar et al., 1978). This may influence the susceptibility of specific cell populations to the adverse effects of certain toxicants.

CYP2E1 is responsible for the hepatic bioactivation of numerous environmental toxicants and carcinogens, including halogenated hydrocarbons and nitroso compounds (Raucy et al., 1993; Kokkinakis et al., 1985). Low levels of CYP2E1 protein have been detected in testis microsomes from untreated rats by Western blotting (Jiang et al., 1998). Futhermore, testicular CYP2E1 protein levels are induced following treatment of rats with pyridine (> 2 fold) or isoniazid (- 3 fold) (Jiang et al., 1998; Thomas et al., 1987). N-Nitrosodiethylamine (NDMA) is a procarcinogen, which requires metabolic activation catalysed by a CYP-dependent NDMA demethylase to exert its toxic and carcinogenic effects (Fahmy & Fahmy, 1976).

47 Antibody inhibition studies have confinned that déméthylation of NDMA is predominantly catalysed by CYP2E1 in liver and testis microsomes (Thomas et al., 1987). NDMA demethylase activity is constitutively expressed at low levels in the testis, but is induced (-3.6 fold) following treatment of rats with isoniazid (Thomas et al., 1987). The physiological significance of testicular CYP2E1 is currently unknown but it may play a role in the local bioactivation of certain toxicants, which may be exacerbated following exposure to chemical inducers.

1.5.3 Testicular Metabolism of Endogenous Compounds

The testis may also play a role in the metabolism of endogenous substances, such as testosterone. The liver is the major site of androgen inactivation, but testicular microsomes have been shown to catalyse several pathways of testosterone metabolism (Sonderfan et al, 1989; Lacroix et al., 1975). Local formation of androgen metabolites, which might have altered biological activities, may underlie important physiological roles within the testis. Several pathways of testicular testosterone metabolism appear to show age-related changes in activity (Lacroix et al., 1975; Sonderfan et al., 1989). For example, immature testes possess high 5a-R activity, which progressively declines to reach the low levels observed during adulthood (Lacroix et al., 1975).

The major pathway of testosterone metabolism catalysed by testicular microsomes from adult rats involves oxidation at the 17-position to form androstenedione

(Sonderfan et al, 1989). Approximately equal amounts of 6 a- and 7a-0HT are also formed, along with traces of 6 p- and 16a-0HT (Sonderfan et al, 1989). In contrast to 7a-0HT, formation of androstenedione and 6a-0HT are not dependent upon NADPH or inhibited by carbon monoxide (Sonderfan et al., 1989). 17p-HSD catalyses the formation of androstenedione, but the identity of the testicular 6 a-hydroxylase is currently unknown (Sonderfan et al., 1989). Antibody inhibition studies revealed that testosterone 7a-hydroxylation is catalysed by CYP2A1 in testicular microsomes (Sonderfan et al, 1989).

CYP2A1 is expressed in the testis (- 2 - 5% hepatic levels) where it is predominantly localised to the Leydig cells (Sonderfan et al, 1989; Seng et al, 1991). Steroid 7a-

48 hydroxylase activity shows age-dependent expression with low or undetectable activity reported in microsomes from immature rats, which increases to form the higher levels observed in the adult testis (Sonderfan et a l, 1989; Inano et ah, 1973; Seng et ah, 1996; Rosness et ah, 1977). This occurs in parallel to the age-dependent increase in testicular CYP2A1 expression (Sonderfan et ah, 1989). Interestingly, testicular 7a-hydroxylase activity continues to increase with age until the rate of formation of 7a-0HT approaches the rate of testosterone biosynthesis (Eechaute et al, 1974).

Testicular microsomes also show 7a-hydroxylase activity towards androstenedione and DHT (Sunde et ah, 1982; Inano et ah, 1970). 7a-hydroxylated androgens lack androgenic activity but inhibit the activity of several enzymes involved in testosterone biosynthesis, including 3p-HSD, 3a-HSD, CYPllAl and 5a-R (Inano et ah, 1973; Inano & Tamaoki, 1971; Sunde et ah, 1982; Rosness et ah, 1977; Mittler, 1985). 7a- hydroxylated metabolites may therefore play a role in the regulation of testosterone biosynthesis and factors capable of modulating CYP2A1 expression could perturb androgen homeostasis. Hepatic CYP2A1 expression is weakly induced by compounds such as PB and p-NF, but it is currently unclear whether the testicular form is sensitive to induction by xenobiotics (Waxman et ah, 1985; Thomas et ah, 1981).

Testicular CYP2A1 expression may be one factor involved in the aetiology of spontaneous Leydig cell lesions in rodents. Seng et ah (1996) reported a correlation between high testicular CYP2A1 expression and severe Leydig cell hyperplasia in Fischer 344 rats. In contrast, Sprague Dawley rats, which exhibit a lower incidence of Leydig cell hyperplasia, possess lower CYP2A1 protein and testicular 7a-hydroxylase activity (Seng et al, 1996). Sustained caloric restriction produces a marked reduction in the incidence of LCTs in Fischer 344 rats and is associated with reduced expression of CYP2A1 and testicular 7a-hydroxylase activity (Seng et ah, 1996; Thurman et ah, 1994). These findings suggest that Leydig cell hyperplasia may, at least in part, be a consequence of prolonged excessive expression of CYP2A1 or alternatively that factors which produce such lesions also induce CYP2A1 (Seng et al, 1996).

49 1.5.4 Other Rat Testicular Cytochrome P450

CYPIA

CYPIA enzymes are responsible for hepatic bioactivation of numerous carcinogenic and mutagenic substances, including polycyclic aromatic hydrocarbons and aromatic amines (Guengerich, 1988). Extrahepatic expression of CYPIA enzymes may therefore be an important determinant of the organ specific toxicity of certain xenobiotics. Dibasio et al. (1991) reported low levels of CYP 1 A-associated ethoxyresorufin 0-dealkylase (EROD) activity in rat testicular microsomes. Goldstein and Linko (1984) reported low levels of CYPlAl and 1A2 proteins in testis microsomes from untreated Sprague Dawley rats using a radioimmunassay (RIA) method. 2,3,7,8,-tetrachloro-dibenzo-p-dioxin (TCDD) treatment is associated with elevated testicular aryl hydrocarbon hydroxylase activity ( 2 . 1 fold) and induction of CYPlAl but not CYP1A2 protein levels (quantified by RIA) (Goldstein & Linko, 1984). These results could not be confirmed by Western blotting, as CYPlAl and

1 A2 levels were below the limits of detection (Goldstein & Linko, 1984).

Subsequent studies have yielded conflicting information concerning the testicular expression of CYPIA, which may be related to strain-related differences in levels of expression and the sensitivity of the techniques used. Omiecinski et al (1990a) reproducibly failed to detect CLPiv47mRNA in testis firom Sprague Dawley rats using a PCR-based assay. Foster et al. (1986) detected no CYPlAl immunoreactive protein in testis firom untreated and inducer-treated (P-NF, PB or clofibrate) Alp/Apk rats (Wistar-derived strain) by immunohistochemistry (IHC). In contrast, Kobayashi et al. (1991) localised CYPIA protein to the Leydig cells in Wistar rats using IHC. More recently, Roman et al. (1998) reported that testicular CYPlAl expression is localised to spermatid tails in untreated Holtzman rats and TCDD treatment was shown to have no effect on the expression of this protein (by IHC). Consistent with previous studies, no testicular CYPlAl protein was detected by Western blotting (Roman et al., 1998). Interestingly, Western blotting confirmed that the aryl hydrocarbon receptor (AhR) and aryl hydrocarbon receptor nuclear translocator (ARNT) proteins are expressed in the testis (Roman et al., 1998). Using IHC, ARNT expression was localised to certain germ cell stages (e.g. mature spermatocytes and some elongating spermatids), Leydig cells and endothelial cells (Roman et al., 1998). The physiological significance of the

50 testicular expression of these proteins is currently unknown.

CYP2B

ÇYP2B1 mRNA is expressed constitutively at low levels (~ 6 % levels in PB-induced liver) in the testis of untreated Sprague Dawley rats (Omiecinski, 1986; Omiecinski et al., 1990b). Treatment of rats with 3-MC or PB had negligible effects on testicular CYP2B1 mRNA levels (Omiecinski, 1986). In contrast, CYP2B2 mRNA was not detected in testis from control, 3-MC or PB-treated rats (Omiecinski, 1986). CYP2B- specifrc pentoxyresorufin 0-dealkylase (PROD) activity and immunoreactive protein (Western blotting) were not detected in testicular microsomes from Fischer 344 and Sprague Dawley rats respectively (Dibiasio et al, 1991; Bengtsson et al., 1990). Kobayashi et al. (1991) used IHC to show that CYP2B protein is exclusively localised to the Leydig cells in Wistar rats. This group also reported a significant temporal variation in the number of Leydig cells showing CYP2B-immunoreactivity, which might have important physiological and toxicological implications.

CYP2C

CYP2C11 immunoreactive protein has been detected in testis microsomes from Long Evans and Harlan Sprague Dawley rats (Ryan et al., 1993; Seng et al., 1996). In contrast, CYP2C11-dependent testosterone 2a-hydroxylase activity and CYP2C immunoreactive protein were not detected in testicular microsomes from Fischer 344 or Sprague Dawley rats (National Centre for Toxicological Research (NTCR) colony), suggesting that this form shows strain-specific regulation (Seng et al, 1996).

CYP3A

CYP3A immunoreactive protein was not detected in testicular microsomes from Sprague Dawley rats by Western blotting (Bengtsson et al., 1990).

CYP4A

CYP4A1 RNA (RNase protection assay) and immunoreactive protein (Western blotting) were not detected in testis from untreated and methylclofenapate-treated

Alderley Park rats (Bell et al., 1992). Low levels of acyl-coA oxidase RNA (~ 6 % of hepatic levels) were detected in testis from untreated animals and were significantly induced (1.7 fold) following treatment with methylclofenapate (Bell et al., 1992).

51 1.6 Aims and Experimental Approach

1.6.1 Aims

The aim of the current project was to investigate the existence of a liver-testis axis wherein microsomal enzyme inducers enhance the metabolic clearance of testosterone causing a compensatory increase in LH secretion. Elevated circulating LH levels may be responsible for the induction of LCTs in rats. The ultimate aim of this work was to provide mechanistic biochemical and molecular data to help explain the aetiology of Leydig cell hyperplasia and tumour formation to aid human risk assessment.

1.6.2 Experimental Approach

Lansoprazole

Lansoprazole was selected as the model compound for studies designed to investigate the existence of a liver-testis axis as it induces hepatic CYPs and chronic exposure produces LCTs in rats (Masubuchi et al., 1997a; Atkinson et al., 1990). The following section will summarise the relevant background information regarding this compound.

Lansoprazole is a gastric proton pump inhibitor that is frequently prescribed for the treatment of conditions such as peptic ulcers, reflux oesophagitis and pathological hypersecretory states such as Zollinger-Ellison syndrome (ZES). The chemical structure of lansoprazole contains an asymmetric sulphur atom and this compound is administered clinically as a racemic mixture of the (+) and (-) enantiomers (Figure 1.8). This compound is extensively metabolised in the liver by CYP-dependent pathways to form the major plasma metabolites, 5-hydroxylansoprazole (primarily catalysed by CYP2C19) and lansoprazole sulphone (primarily catalysed by CYP3A4/5) (Pearce et al., 1996). The half-life of lansoprazole in the rat has been estimated to be 1 . 8 8 and 2.62 hours for the (+) and (-) enantiomers respectively following oral administration (Arimori et al., 1998).

52 Figure 1.8: The chemical structure of lansoprazole

C H g - S ^

2-[[[3-methyl-4-(2,2,2-trifluoroethoxy)-2-pyridyl]-methyl]-sulfinyl]benzimidazole * indicates the asymmetric sulphur atom.

Treatment of female Sprague Dawley rats with lansoprazole (300mg/kg/day for 7 days) is associated with a significant increase in total hepatic CYP content and induction of CYPlAl, 1A2, 2B1, 2B2, 3A2, 2C6 and 4A1 proteins (Masubuchi et al., 1997a). In cultured human hepatocytes lansoprazole is a “mixed inducer” of CYPIA and CYP3A, increasing both mRNA and protein levels (Curi-Pedrosa et al., 1994; Masubuchi et al., 1998). Lansoprazole induces CYPIA in humans but it is currently unclear whether this compound induces CYP3A in vivo (Kokufu et al., 1995; Fuchs et al., 1994). The mechanism through which lansoprazole induces CYPIA is unclear as high affinity binding of this compound to the Ah receptor has not been demonstrated (Curi-Pedrosa et al., 1994).

Lansoprazole has been subject to an extensive series of toxicity studies, many of which indicate adverse effects on the male reproductive system. In a one-year rat study, lansoprazole treatment increased the incidence of Leydig cell hyperplasia and one benign Leydig cell tumour was observed (50 mg/kg/day) (Atkinson et al., 1990). In a two-year rat study, an increased incidence of seminiferous tubule atrophy was recorded at 50 mg/kg/day along with an increased incidence of Leydig cell hyperplasia and tumours at dosages of 15 mg/kg/day or higher (Unpublished report by Takeda Chemical Industries Ltd., A-29-681). In contrast, lansoprazole treatment was not associated with Leydig cell hyperplasia or tumours in mice or dogs (Unpublished reports by Takeda Chemical Industries Ltd., A-29-680; A-29-439). In subchronic mouse studies (3 months) an increased incidence of seminiferous tubule atrophy and epididymal hypospermia were reported at high dosages (1200 and 2400 mg/kg/day) (Unpublished report by Takeda Chemical Industries Ltd., A-126-049).

53 Lansoprazole was found to be negative in an extensive array of in vitro and in vivo genotoxicity studies (e.g. bacterial mutagenicity, Ames test, rat bone marrow chromosome aberration test) (reviewed by Fort et al., 1995). Fort et al. (1995) performed mechanistic studies to investigate the mode of LCT induction following lansoprazole treatment in Sprague Dawley rats. In a four-week study, lansoprazole treatment (150 mg/kg/day) was associated with the following effects:

• Significant reduction in serum testosterone levels (maximal decrease of 50%)

• Significant increase in serum LH levels ( 6 8 % above control levels)

• Significant decrease in intratesticular testosterone levels ( 6 6 % of control levels) • Significant increase in intrapituitary LH levels (61% above control levels)

Providing testosterone supplementation to lansoprazole-treated Fischer 344 rats lowered serum LH levels and completely suppressed the induction of LCTs (Fort et al., 1995). These findings indicate that reduced negative feedback of testosterone at the hypothalamus and pituitary gland leading to an increase in LH secretion may be involved in the induction of LCTs by lansoprazole. Lansoprazole inhibits several steps in the testosterone biosynthetic pathway in cultured rat Leydig cells, the most sensitive site being transport of cholesterol to the CYPscc enzyme (Fort et al., 1995). In addition, increased metabolic clearance of testosterone secondary to induction of hepatic CYPs might contribute to the lansoprazole-induced reduction in serum testosterone levels. Interestingly, lansoprazole induces hepatic UDP-GT activity, increases biliary excretion of thyroid hormone and reduces plasma T 4 levels in female rats (Masubuchi et al., 1997b). However, in a one-year carcinogenicity study no adverse effects on thyroid gland histology were reported in lansoprazole-treated rats (> 50mg/kg/day) (Atkinson et al., 1990).

The ability of lansoprazole to induce LCTs in rats is currently considered to represent a species-specific effect. Studies indicate that this compound has no clinically significant effects on male hormone levels or gonadal function in humans (Meikle et al., 1994; Gaetani et al., 1995).

54 General Experimental Approach

To investigate the existence of a liver-testis axis, the effects of lansoprazole treatment on the following major areas have been studied:

• endocrine control of the testis

• expression of hepatic and testicular CYPs

• hepatic and testicular testosterone metabolism

In addition, the effects of dosing rats with model inducers of the major CYP families (P-NF (CYPIA), PB (CYP2B), PCN (CYP3A) and ciprofibrate (CYP4A)) were characterised. There is no evidence that these compounds produce LCTs in rats therefore it was interesting to compare and contrast their effects to those of lansoprazole.

Detailed information concerning the design of these experiments can be found in the experimental strategy section (2.3).

55 Chapter 2

Materials and Methods 2.1 Materials

2.1.1 Chemicals

The suppliers of all reagents used in this thesis are listed below. All chemicals and solvents were of the highest grade available and were purchased from Sigma or Fisher Chemicals unless otherwise stated.

Chemical Supplier

Acetonitrile (HPLC-grade) Fisher Acrylamide (30%, w / v) Ultrapure Agarose Flowgen Ammonium acetate BDH Ammonium persulphate Sigma Androstenedione Sigma Bisacrylamide (2%, w / v) Ultrapure P-Naphthoflavone Sigma Bovine serum albumin (Fraction V) Sigma Carbonate/ bicarbonate capsules Sigma Carboxymethylcellulose Sigma Chloroform (molecular biology grade) Sigma Cholic acid Sigma Ciprofibrate Sanofr dNTP mixture (lOmM) Gibco Dried milk powder (fat free) Tesco ECL™ western blotting kits (Rat C Y PlA l, 1A2, 2B, Amersham 3 A and 4A1) ECL Plus western blotting detection reagents Amersham Ethyl acetate (HPLC-grade) Fisher Folin-Ciocalteu’s phenol reagent Sigma Gel loading solution Sigma Genomic DNA (rat) Clontech

57 HPLC standards (6 a-, 6 P-, 15a-, 16a-, llp -, 2a- and Sigma

2P-hydroxytestosterone)

HPLC standards (7a- and 16p-hydroxytestosterone) Ultrafrne Isopropanol (molecular biology grade) Sigma Lansoprazole Sigma Methanol (HPLC-grade) Fisher 4-MA ( 17P-M A'-Diethylcarbamoyl-4-methyl-4-aza-5a- AstraZeneca androstan-3-one) NADPH (nicotinamide adenine dinucleotide phosphate. Melford reduced form tetrasodium salt) Nitrocellulose membrane (Hybond C) Amersham Nuclease-free water Ambion Phénobarbital Sigma Phosphate buffered saline tablets (minus Ca^^ and Mg^^ Fisher Ponceau S concentrate Sigma Pregnenolone-16a-carbonitrile Sigma Pyronin Y Sigma qPCR^'^ Mastermix Eurogentec Random primers Gibco RNAqueous™-4PCR kit Ambion Scintillation fluid (Optiphase ‘Safe’ scintillation cocktail) Perkin Elmer Sodium dithionite Sigma Supelcosil LC-18 HPLC Column (3pm particle size) Supelco

Supelguard LC-18 Guard column (5pm particle size) Supelco

Superase-In (20U/pl) Ambion Superblock buffer in PBS Fisher Superscript II RNase H- Reverse Transcriptase Kit Gibco Testosterone Sigma 4-^"^C-Testosterone (50pCi/ml in toluene, 50-60mCi/mmol) Amersham N,N,N’ ,N ’ -T etramethyethylenediamine (TEMED) Sigma Tris-Borate-EDTA (TBE) Buffer (10 X) Sigma Triton-X 100 Sigma Trizol® Reagent Gibco

58 Turbo-TMB Substrate Fisher Tween 20 and 80 Sigma Water (HPLC-grade) Sigma

2.1.2 HPLC Solvents

Mobile Phase A: 90: 10 (v / vl Methanol:Acetonitrile 900ml HPLC-grade Methanol 100ml HPLC-grade Acetonitrile Mixed by shaking in a closed container and 0.5% (v / v) glacial acetic acid was added. Stored at room temperature for up to 1 month. lOOmM Ammonium Acetate. pH 4.5 7.708g Ammonium Acetate Made up to IL with MilliQ water. Adjusted to pH 4.5 with glacial acetic acid. Stored at room temperature for up to 3 months.

Mobile Phase B: 30:70 (v / v) Methanol:lOmM Ammonium Acetate. pH 4.5 300ml HPLC-grade Methanol 700ml 1 OmM Ammonium Acetate (1:10 (v / v) 1 OOmM ammonium acetate in HPLC-grade water) Stored at room temperature for up to 1 month with continuous stirring.

2.1.3 Western Blotting Solutions

All solutions were stored at 2 to 5°C for up to 3 months unless otherwise stated.

Loading Buffer • 4.8ml MilliQ water

1.2ml 0.5M Tris-HCL (pH 6 .8 ) 0.96ml Glycerol 1.92ml 10% (w/v) Sodium Dodecyl Sulphate (SDS) 0.48ml P-Mercaptoethanol 0.6ml 0.05% (w / v) Pyronin Y Stored protected from light at room temperature for up to 2 weeks.

59 2 X Resolving Gel Buffer, pH 8.8 181.6g Tris 4g SDS

Made up to IL with MilliQ water. Adjusted to pH 8 . 8 with lOM sodium hydroxide.

2 X Stacking Gel Buffer, pH 6.8 60.6g Tris 4g SDS

Made up to IL with MilliQ water. Adjusted to pH 6 . 8 with lOM sodium hydroxide.

5 X Running Buffer. pH 8.3 I5.15g Tris 72g Glycine 5g SDS Made up to IL with MilliQ water.

5 X Running buffer was diluted 1:5 (v/v) in MilliQ water prior to use.

5 X Transfer Buffer. pH 8.3 11.9g Tris 56.25g Glycine Made up to IL with MilliQ water.

5 X Transfer buffer was diluted as follows prior to use: 160ml 5 x Transfer buffer, 640ml MilliQ water and 200ml methanol.

Tris-Buffered Saline with 0.1% (y / y) Tween (TBS-TweenL pH 7.6 2.42g Tris

8 g Sodium Chloride Made up to IL with MilliQ water, adjusted to pH 7.6 (with lOM hydrochloric acid) and 1ml Tween 20 added.

60 2.1.4 ELISA Solutions

All solutions were stored at 2 to 5°C for up to 1 month unless otherwise stated.

Carbonate/ bicarbonate Buffer One carbonate/bicarbonate capsule dissolved in 100ml MilliQ water.

Solubilisation Buffer lOg Cholic acid lOg Sodium Chloride 10ml Triton-X 100 . Made up to IL with MilliQ water.

Wash Buffer 10 PBS tablets (minus Ca^"^ and Mg^"^ 5ml 10% (v/v) Tween 20 Made up to IL with MilliQ water

2.1.5 Agarose Gel Electrophoresis Solutions

10 X Tris-acetate-EDTA (TAEl buffer, pH 7.5 484g Tris base 200ml 0.5M EDTA (pH 8) ^ 114.2ml Glacial acetic acid 500ml MilliQ water Made up to lOL with MilliQ water. Adjusted to pH 8.3 with lOM sodium hydroxide.

10 X TAE buffer was diluted 1:10 (v7 v) in MilliQ water prior to use.

Tris-Borate-EDTA (TBE) Buffer (0.5 X) 10 X TBE buffer (0.89M Tris borate, 0.02M EDTA (pH 8.3)) was purchased from Sigma and diluted 1:20 (v / v) in MilliQ water prior to use.

61 2.2 Methods

2.2.1 Animal Studies

Five animal studies formed the basis of the experimental approach for this project. The aims and experimental design of each study are summarised in the Experimental Strategy section (see 2.3).

2.2.1.1 Animal Housing

Male Sprague Dawley rats were purchased from Charles River Laboratories (study 1, 4 and 5) or Bantin & Kingman Universal (study 2 and 3) and were allowed to acclimatise for one week prior to initiation of treatment. Animals had free access to food (SDS R & M No.l (Special Diet Services Ltd., UK)) (study 1, 4 and 5) or Rat and Mouse Expanded Diet (B & K Universal) (study 2 and 3) and tap water throughout the study. For all studies animals were held under controlled temperature (21 ± 1°C), humidity (55 ± 15%) and light conditions (12 hour light-dark cycle, lights on between 6.00a.m. and 6.00p.m. GMT).

2.2.1.2 Study 1: Effect of Model Inducers on Cytochromes P450, Testosterone Metabolism and Plasma Hormone Levels

Animal Treatment

Dosing of animals and collection of samples were conducted by technical staff at AstraZeneca. Male Sprâgue Dawley rats (8 to 10 weeks old) were randomly assigned to one of the five treatment groups shown in Table 2.1 (n=5 animals per group). Animals were dosed once daily for four consecutive days and each compound was administered in a volume of lOml/kg body weight. Control animals were not dosed, p-naphthoflavone (PNF) and ciprofibrate were administered by oral gavage (p.o.).

Phénobarbital (PB) and pregnenolone-16a-carbonitrile (PCN) were administered by intraperitoneal injection (i.p.). Doses were selected from the published literature (Amacher & Schomaker, 1998; Kocarek & Reddy, 1996; Graham et al, 1996; Hanoika et al, 1995).

62 Table 2.1: Dosing schedules for compounds administered to rats for the model inducers study.

Compound Vehicle Dose (mg/kg body weight/day) Control 0

p-naphthofiavone Com oil 100 Phénobarbital Saline, pH 7.0 80

Pregnenolone-16a- Com oil 60 carbonitrile Ciprofibrate 1% carboxymethylcellulose 10 (CMC)/0.1% Tween 80

Collection of Samples

On day five (9.30 a.m. to 12.00 p.m.), animals were killed by administration of a lethal dose of sodium pentobarbital (60mg/kg body weight) (i.p.). The livers and testes were removed immediately and weighed. The tissues were then snap frozen using liquid nitrogen and stored at -70°C. Blood samples were taken by cardiac puncture and transferred in to lithium heparin tubes. Plasma was prepared by centrifiigation (lOOOg for 10 minutes at 4° C) and stored a t-20° C.

2.2.1.3 Study 2: Effect of Lansoprazole on Cytochromes P450, Testosterone Metabolism and Plasma Hormone Levels

Animal Treatment

Male Sprague Dawley rats (6 to 8 weeks old) were randomly assigned to the control (n=l 1) or experimental group (n=14). Each animal was dosed once daily with vehicle (0.5% (w / v) CMC) (5ml/kg body weight) or lansoprazole (150mg/kg body weight/day) (5ml/kg body weight) by oral gavage for fourteen days. This dose was selected as it has previously been shown to reduce serum testosterone and increase serum LH levels in male rats (Fort et al., 1995).

63 Collection o f Samples

On day 15 (10.00 a.m. to 12.00 p.m.), animals were killed by administration of a lethal dose of sodium pentobarbital (60mg/kg body weight) (i.p.) and the livers and testes were removed immediately. The livers were weighed and then perfused with ice-cold 0.9% (w / v) saline. The testes were weighed and then decapsulated. Livers and testes were transferred in to sterile tubes containing ice-cold 1.15% (w / v) KCl or SET buffer (25OmM sucrose, 5.4mM EDTA, 5OmM Tris, pH 7.4) respectively. Blood samples were taken by cardiac puncture and transferred in to lithium heparin tubes. Plasma was prepared by centrifugation (lOOOg for 10 minutes at 4°C) and stored at

-20°C.

2.2.1.4 Study 3: Effect of Lansoprazole on Gene Expression in the Liver and Testis

Animal Treatment

Male Sprague Dawley rats (6 to 8 weeks old) were randomly assigned to the control or experimental group (n=10 animals per group). Each animal was dosed once daily with vehicle (0.5% (w / v) CMC) (5ml/kg body weight) or lansoprazole (150mg/kg body weight/day) (5ml/kg body weight) by oral gavage for fourteen days.

Collection o f Samples

On day 15 (9.45 a.m. to 11.45 a.m.), animals were killed by administration of a lethal dose of sodium pentobarbital (60 mg/kg body weight) (i.p.). The livers and testes were removed immediately and weighed. For each liver, a part of the left lobe was removed using a sterile scalpel and snap firozen using liquid nitrogen. The remaining liver was perfused with ice-cold 0.9% (w / v) saline and then placed in to a sterile tube containing ice-cold 1.15% (w / v) KCl.

For half of the animals (i.e. n=5 control and n=5 lansoprazole-treated), both testes were snap firozen using liquid nitrogen and stored at -80°C. For the remaining animals, both testes were transferred in to sterile tubes containing 5ml ice-cold buffer (25OmM sucrose, 5OmM Tris, pH 7.4).

64 Blood samples were collected by cardiac puncture and transferred in to lithium heparin tubes. Plasma was prepared by centrifugation (lOOOg for 10 minutes at 4°C) and stored at -20°C.

2.2.1.5 Study 4: Effect of Lansoprazole on Gene Expression in the Pituitary Gland

Animal Treatment

Dosing of animals and collection of samples were performed by technical staff at

AstraZeneca. Male Sprague Dawley rats ( 6 to 8 weeks old) were randomly assigned to the control or experimental group (n=5 animals per group). Each animal was dosed once daily with vehicle (0.5% (w / v) CMC/ 0.1% (v / v) Tween 80) or lansoprazole (150mg/kg body weight/day) (5ml/kg body weight) by oral gavage for fourteen days.

Collection o f Samples

On day 14 (2.00 p.m. to 3.30 p.m.), approximately four hours after the final dose of vehicle or lansoprazole, animals were killed by administration of a lethal dose of sodium pentobarbital (60mg/kg body weight) (i.p.). Pituitary glands were removed and immediately placed in to tubes containing 500pl Trizol® Reagent. These samples were snap frozen using liquid nitrogen and stored at -80°C. Blood samples were collected by cardiac puncture and transferred in to lithium heparin tubes. Plasma was prepared by centrifugation (lOOOg for 10 minutes at 4°C) and stored at -20°C.

2.2.1.6 Study 5: Effect of Lansoprazole on the Plasma Clearance of ^"^C-Testosterone

Animal Treatment

Dosing of animals and collection of samples were performed by technical staff at

AstraZeneca. Male Sprague Dawley rats ( 6 to 8 weeks old) were randomly assigned to the control (n=5) or experimental group (n=4). Each animal was dosed once daily with vehicle (0.5% (w / v) CMC/ 0.1% (v / v) Tween 80) or lansoprazole (I50mg/kg body weight/day) (5ml/kg body weight) by oral gavage for fourteen days.

65 Administration of^'^C-Testosterone

On day 15 (10.45 a.m.), approximately 24 hours after the final dose of vehicle or lansoprazole, each animal was dosed via the right tail vein with 50pCi 4-^"^C- testosterone (0.254mg testosterone) (0.5ml) in 10:90 (v / v) ethanol: 0.9% (v / v) saline. This dose of ^'^C-testosterone was selected based on the limits of detection of the methods used to analyse plasma samples.

Collection o f Blood Samples

Serial blood samples (250pl) were collected at the following timepoints after the dose of ^"^C-testosterone: 5, 15, 20, 30, 40, 50, 60, 90, 120, 180 and 260 minutes. These sampling times were selected based on the half-life of testosterone from the published literature (Pirke et al., 1982; Wang & Bulbrook, 1967; Henderson et al., 1980). Blood samples were collected from the left tail vein except for the final sample (260 minutes), which was taken by cardiac puncture following the administration of a lethal dose of sodium pentobarbital (60mg/kg body weight) (i.p.). Blood samples were transferred in to lithium heparin tubes and plasma was prepared by centrifugation (lOOOg for 10 minutes at 4°C). Samples were snap firozen on dry ice and stored at -

80°C.

2.2.2 Preparation of Microsomes

Model Inducers Study

Livers and testes (with capsules) were thawed and then homogenised in ice-cold 0.154M KCl/ 5OmM Tris (pH 7.4) using a Polytron PT3100 homogeniser. Testes from each treatment group were pooled to give five samples. Cell debris, nuclei and mitochondria were removed by centrifugation, for 20 minutes at 9,000g in a RC28S centrifuge (4°C). The supernatant was transferred to a clean tube and centrifuged for

1 hour at 100,000g in a RC28S centrifuge (4°C). The microsomal pellet was resuspended in 5OmM potassium phosphate buffer containing 20% (w / v) glycerol (pH 7.4) using a volume equivalent to 500pl per gram of original liver weight or lOOpl per gram of original testis weight. Samples were stored at -80°C.

66 Lansoprazole Studies

Chilled livers and decapsulated testes were homogenised in ice-cold 1.15% (w / v) KCl or SET buffer respectively using a Potter-Elvehjem homogeniser. Testes from each treatment group were pooled to give two samples. Cell debris, nuclei and mitochondria were removed by centrifugation for 20 minutes at 10,000g in a MSE High Spin 18 centrifuge (4°C). The supernatant was transferred to a clean tube and centrifuged for 1 hour at 100,000g in a Beckman L5-65 centrifuge (4°C). The microsomal pellet was resuspended in 15ml ice-cold 5OmM potassium phosphate buffer (pH 7.4) containing 20% (w / v) glycerol or 6 ml ice-cold SET buffer containing 20% (w / v) glycerol for liver and testis microsomes respectively. Samples were stored at -80°C.

Preparation o f Washed Microsomes

An experiment was conducted to determine whether it would be advantageous to include a washing step during the preparation of liver and testis microsomes. For this experiment a batch of microsomes were prepared using the protocol described for lansoprazole studies with the following modifications. Following the second centrifugation step the microsomal pellet was resuspended in 1.15% (w / v) KCl or SET buffer, for liver and testes samples respectively, and then centrifuged for 1 hour at 100,000g in a Beckman L5-65 centrifuge (4°C). The microsomal pellet was resuspended in ice-cold 5OmM potassium phosphate buffer (pH 7.4) containing 20% (w / v) glycerol (500pl per gram of original liver weight) or ice-cold SET buffer containing 2 0 % (w / v) glycerol (150pl per gram of original testis weight) for liver and testis microsomes respectively. Samples were stored at -80°C.

2.2.3 Preparation of Testicular Supernatant

Testes from each animal were homogenised in 5ml ice-cold buffer (250mM sucrose, 5OmM Tris, pH 7.4) using a Potter-Elvehjem homogeniser. Samples were centrifuged for 20 minutes at 10,000g in a MSE High Spin 18 centrifuge (4°C). The supernatant was transferred to a clean tube and centrifuged for 1 hour at 100,000g in a Beckman L5-65 centrifuge (4°C). The supernatant obtained after the second centrifugation step was transferred to a clean tube and stored at -20°C.

67 2.2.4 Hormone Assays

Technical staff at AstraZeneca performed all assays to measure hormone levels in plasma and testicular supernatant.

Total testosterone was measured using a solid phase radioimmunoassay method (Coat- A-Count® Total Testosterone) supplied by Diagnostic Products Corporation (DPC). This kit employs a testosterone specific antibody immobilised to the wall of a polypropylene tube. ^^^I-labelled testosterone competes with testosterone present in the test sample for antibody binding sites. Antibody-bound radioactivity is separated. from free by decanting the tube and quantified using a gamma counter. A standard curve is used to determine the amount of testosterone present in test samples. The lower limit of detection of this kit is 0.14 nmol/L.

Plasma prolactin, LH and FSH were measured using radioimmunoassay methods supplied by Amersham (Biotrak kits). These kits are based on the principle previously described for the testosterone assay except that the hormone specific antibody is free in solution. Separation of free and antibody-bound radioactivity is effected by magnetic separation using a secondary antibody that is bound to magnétisable polymer particles. The lower limits of detection of the prolactin, LH and FSH kits are 0.7, 0.8 and 0.9 ng/ml respectively.

2.2.5 Protein Determination

The protein concentration was determined for liver and testis microsomes using the method of Miller (1959). A standard curve was constructed using bovine serum albumin diluted in MilliQ water to give final protein concentrations spanning the 0 to 200pg/ml range (0, 20, 40, 80, 100, 120, 160 and 200pg/ml). Standard samples were assayed in triplicate. Liver and testis microsomes were diluted in MilliQ water (1:100 to 1:500 (v / v) dilution) and assayed in duplicate.

For the assay, 1ml of solution A (1ml 5% (w / v) copper (11) sulphate, 10ml 1% (w / v) potassium sodium (+) tartrate and 100ml 0.5M sodium hydroxide containing 10%

(w / v) sodium carbonate (anhydrous)) was added to 1 ml of the diluted protein sample. The tubes were incubated at room temperature for exactly 10 minutes. Then, 3ml

68 solution B (Folin-Ciocalteu’s phenol reagent diluted 1:10 in MilliQ water) was added and the tubes were incubated at 50°C for 1 0 minutes. The tubes were cooled and the absorbance at 650nm was read on a Kontron Uvicon 932 spectrophotometer.

2.2.6 Microsomal P450 Content Determination

The total cytochrome P450 content of liver microsomes was determined using the method of Omura and Sato (1964). Liver microsomes were diluted 1:10 (v / v) in 5OmM potassium phosphate buffer (pH 7.4) and placed on ice. An excess (approximately lOmg) of sodium dithionite was added and the tube was vortexed. The contents of each tube were divided between two clean spectrophotometer cuvettes and a baseline absorbance measured between 400 and 490nm on a Kontron Uvicon 932 spectrophotometer. Carbon monoxide was then gently bubbled through one of the cuvettes for approximately 30 seconds. The absorbance of the sample was re­ measured across the 400 to 490nm range. The concentration of cytochrome P450 was calculated from the difference in absorbance between 450 and 490nm, employing a molar extinction coefficient of 91mM'^cm'\

2.2.7 High Pressure Liquid Chromatography (HPLC)

The following HPLC method (B.Buckle, Personal Communication) was used to separate and quantify testosterone and individual metabolites present in incubation extracts from the testosterone hydroxylase assay ( 2 .2 .8 ) and plasma samples from study 5 (2.2.1.6).

2.2.7.1 HPLC Apparatus

All analyses were performed using a Hewlett Packard HP 1090 Series HPLC system controlled by ChemStation software (Hewlett Packard) linked to a Berthold LB506 C-1 radioactivity monitor (YG-150 solid detector cell) and a SpectroMonitor 3000 variable wavelength detector (LDC/Milton Roy). Analytes were detected at 240nm and by radiochemical detection. Chromatography was performed on a Supelcosil LC- 18 reverse-phase silica column (150 x 4.6mm, 3pm particle size) preceded by a

Supelguard LC-18 guard column (2cm, 5pm particle size). Berthold HPLC software (version 1.51) was used for data acquisition and analysis.

69 12.1.2 Preparation of HPLC Standards

6 a-, 6 p-, 15a-, 7a-, 16a-, 16p-, lip-, 2a- and 2p-hydroxytestosterone standards (>98% purity) were initially diluted to 2000 nmoles/ml in HPLC-grade methanol. Standards were then diluted in initial HPLC solvent conditions (see 2.2.7.3) to give stock solutions at concentrations of 1000 nmoles/ml (Inmole/pl injected). For testosterone and androstenedione, lOOpM solutions were prepared in initial HPLC solvent conditions (O.lnmole/pl injected).

2.2.7.3 Chromatographic Conditions

Testosterone and individual metabolites were separated using the two step linear solvent gradient described in Table 2.2. Mobile phase A consisted of methanohacetonitrile (90:10 (v / v)) containing 0.5% (v / v) glacial acetic acid and mobile phase B consisted of methanol: 1 OmM ammonium acetate (pH 4.5) (30:70 (v / v)) (see 2.1.2). The mobile phases were degassed for 30 minutes using a gentle stream of helium and then the colunrn was equilibrated with initial solvent conditions for 30 minutes prior to injection of the first sample. All chromatographic separations were performed at room temperature using a flow rate of 0 . 8 ml/minute. Testosterone and individual metabolites present in test samples were identified by co­ chromatography with authentic standards injected on to the column with each sample.

Table 2.2: The solvent gradient used to separate testosterone and metabolites

Time Mobile Phase A Mobile Phase B

(minutes) (%) (%,)

0 1 0 90

15 1 0 90

32.5 2 0 80 55 60 40 60 60 . 40

65 1 0 90

70 2.2.8 Testosterone Hydroxylase Assay: Microsomal Incubations

The following assay (B.Buckle, Personal Communication) was used to measure the rates of testosterone hydroxylation and androstenedione formation catalysed by liver and testis microsomes.

2.2.8.1 Preparation of the Substrate

Microsomes were incubated with a mixture of 4-^"^C-testosterone and cold testosterone substrate, which was prepared as follows. The required volume of ^"^C-testosterone (50pCi/ml in toluene, 57mCi/mmol) was dispensed in to a glass vial, reduced to dryness under a gentle stream of nitrogen and reconstituted in methanol to give a final concentration of 200pCi/ml. A 1.215mM solution of cold testosterone was prepared by weighing 8.76mg testosterone in to a 25ml volumetric flask. This was dissolved in 7.5ml methanol and then slowly made up to volume with 5OmM potassium phosphate buffer (pH 7.4). This solution was sonicated for approximately 10 minutes to aid dissolution. Finally, the ^"^C-testosterone and cold testosterone solutions were mixed to give a solution containing 1.65%:98.35% (v / v) ^ "^C-testosterone : cold testosterone. 200pl of this mixture, containing 239 nmoles cold testosterone and 11.6 nmoles ^"^C-testosterone (0.66|iCi), was added to each incubation.

2.2.8.2 Incubation Conditions

Liver Microsomes

The incubation components listed in Table 2.3 were dispensed in to 10ml conical flasks placed on ice. Incubations were performed in triplicate for each animal. A blank incubation, in which NADPH was replaced by an equivalent volume of 5OmM potassium phosphate buffer (pH 7.4), was also prepared for each animal. The 5a- reductase inhibitor, 4-MA (17p-MA-Diethylcarbamoyl-4-methyl-4-aza-5a-androstan-

3-one) (IpM), was added to all incubations unless otherwise stated. This concentration of 4-MA was selected as it has been previously shown to completely inhibit steroid 5a-reductase enzyme without inhibiting other pathways of testosterone oxidation catalysed by rat liver microsomes (Sonderfan & Parkinson, 1988).

71 Incubation mixtures were pre-incubated for 5 minutes at 37°C with gentle shaking

Reactions were initiated by the addition of 200|il ^"^C-testosterone (see 2.2.8.1) to give a final substrate concentration and incubation volume of 250.6pM and 1ml respectively. Samples were incubated for 5 minutes at 37°C with gentle shaking.

Reactions were terminated by the addition of 200pl 3M HCl to each flask. The flask contents were mixed well and then placed on ice.

Table 2.3: Components of incubation mixtures for the testosterone hydroxylase assay.

Component Stock Solution Volume / Final incubation (pi) Concentration

Potassium phosphate 5 OmM 399 5 OmM ; buffer (pH 7.4)

EDTA lOmM 1 0 0 ImM

Microsomal Protein 2.5mg/ml 2 0 0 0.5mg/ml

NADPH lOmM 1 0 0 ImM

4-MA ImM 1 IpM

All stock solutions and dilutions of microsomal protein were prepared in 50mM potassium phosphate buffer (pH 7.4) with the exception of 4-MA which was dissolved in acetone.

Testis Microsomes

Incubations using testis microsomes were performed as previously described for liver microsomes except that the, microsomal protein concentration was increased to

2 mg/ml.

2.2.S.3 Extraction of Testosterone and Metabolites

Testosterone and metabolites were extracted by the addition of 3ml ice-cold ethyl acetate followed by vortexing for one minute. Samples were then centrifuged at 2000g for 10 minutes at room temperature. The lower aqueous phase was frozen using an acetone/dry ice bath and the upper organic phase transferred to a clean test tube. This extraction procedure was repeated once, the two organic phases were combined and then reduced to dryness under a gentle stream of nitrogen.

72 Extracts were reconstituted in lOOpl HPLC-grade methanol and then made up to a final volume of 200pl with HPLC-grade water. Testosterone and metabolites present in extracts were separated and quantified using the HPLC method described in 2.2.7.

80pl of each extract was injected on to the column along with 2 0 pl of a mixture of authentic testosterone standards (see 22.12).

2.2.5.4 Calculation of the Extraction Efficiency

Aliquots of the ^"^C-testosterone substrate (200pl) and reconstituted incubation extracts

(20pl) were added to scintillation vials containing 3.5ml scintillation fluid and counted for one minute on a Wallac 1410 liquid scintillation counter. Triplicate or duplicate samples were counted for the substrate and extracts respectively. The extraction efficiency was calculated as follows:

dpm / extract % Extraction efficiency = ------— x 100 dpm / incubation

. 2.2.5.5 Calculation of Enzyme Activity

The specific activity for the formation of each metabolite was calculated in picomoles per minute per milligram of microsomal protein (pmol/min/mg protein) as follows. The use of radiolabelled substrate enabled quantification of individual testosterone metabolites injected on to the HPLC column in incubation extracts. Berthold HPLC software calculates a percentage Region of Interest (%ROI) value for each peak, which corresponds to the percentage of total radioactive counts measured during a chromatographic run that were attributable a particular peak.

The %ROI value for a particular metabolite peak was converted to a specific activity using the following calculation: '

Specific _ / o ROI ^ amount of substrate added (nmoles) = nmoles / 5 minute activity incubation / 0.5 mg microsomal protein

The product of this calculation was converted to a specific activity in pmol/min/mg microsomal protein.

73 If metabolism was detected in blank samples, the specific activity was calculated for each metabolite and subtracted from the specific activity calculated for incubations containing cofactor.

2.2.9 Optimisation of the Testosterone Hydroxylase Assay

The incubation protocol described in 2.2.8 represents a method that was optimised as part of this project. The following sections describe the experiments conducted during the optimisation of this protocol for use with microsomal samples from the current project.

2.2.9.1 Incubation Time

An optimisation experiment was performed to define the linear range of metabolite formation with respect to the incubation time.

The incubation protocol described in 2.2.8 was used except for the following modifications. Initially, each flask contained five times the incubation mixture (i.e. 5ml final volume) and Img/ml of microsomal protein. 4-MA was not added to incubation mixtures and was replaced by an equivalent volume of 5OmM potassium phosphate buffer (pH 7.4). Aliquots (1ml) of the incubation mixture were removed from each flask at the following timepoints following addition of the testosterone substrate: 1, 5, 7.5, 10 and 20 minutes. Reactions were then terminated by the addition of 200pl 3M HCl and extractions performed as previously described. All incubations were performed in duplicate. ,

2.2.9.2 Microsomal Protein Concentration

An optimisation experiment was performed to define the linear range of metabolite formation with respect to the microsomal protein concentration.

The incubation protocol described in 2.2.8 was used except for the following modifications. Microsomes were incubated with testosterone at protein concentrations of 0.1, 0.5, 0.75, 1, 1.5, 2 and 4 mg/ml. 4-MA was not added to incubation mixtures and was replaced by an equivalent volume of 5OmM potassium phosphate buffer

74 (pH 7.4). Reactions were terminated by addition of 200pl 3M HCl and extractions performed as previously described. All incubations were performed in triplicate.

2.2.9.3 Stability of Incubation Extracts

A study was performed to determine whether incubation extracts and testosterone standards were sufficiently stable at room temperature to allow the use of a non- refiigerated autosampler.

Incubations and extractions were performed as described in 2.2.8. 4-MA was not added to incubation mixtures and was replaced by an equivalent volume of 5OmM potassium phosphate buffer (pH 7.4). Incubation extracts were reconstituted in lOOpl

HPLC-grade methanol then made up to a final volume of 200pl with HPLC-grade water. 80pl of the extract was injected on to the HPLC column immediately. A further 80pl of the extract was injected on to the column following 24 hours at room temperature. This process was repeated for a number of incubation extracts and for a mixture of testosterone standards. A metabolite was considered to be stable if the amount detected following the second injection (after 24 hours) was no more than

15% lower (or higher) than the amount detected after the first injection (at 0 hours).

2.2.9.4 Inclusion of 4-MA in Incubation Mixtures

An experiment was performed to investigate the effects of the 5a-reductase inhibitor, 4-MA, on testosterone metabolism catalysed by liver microsomes.

The incubation protocol described in 2.2.8 was used. For each microsome sample studied, triplicate incubations were performed in the presence of 4-MA (IpM) and single incubation was performed in which 4-MA was replaced by an equivalent volume of 5OmM potassium phosphate buffer (pH 7.4).

75 2.2.10 Extraction and Quantification of ^'‘C-Testosterone in Plasma Samples

The following section describes analysis of plasma concentrations of ^"^C-testosterone in samples from study 5 (see 2.2.1.6).

2.2.10.1 Extraction of ^"*C-Testosterone from Plasma

Plasma samples (50pl) were added to test tubes containing 450pl HPLC-grade water. 1.5ml ethyl acetate was added to each tube followed by vortexing for two minutes. Samples were then centrifuged at 2000g for 10 minutes at room temperature. The lower aqueous phase was frozen using an acetone/dry ice bath and the upper organic phase transferred to a clean test tube. This extraction procedure was repeated three times and the organic phases were combined. Wherever possible, plasma extractions were performed in duplicate.

2.2.10.2 HPLC Analysis of Plasma Sample Extracts

To investigate whether any ^"^C-testosterone metabolites were present, selected plasma extracts were subjected to HPLC analysis. Samples from one control animal (at every timepoint) and one lansoprazole-treated animal (at 5, 20 and 180 minutes) were analysed.

The combined organic phases following extraction were reduced to dryness under a gentle stream of nitrogen. Extracts were reconstituted in 90pl HPLC-grade methanol and then made up to a final volume of 180pl with HPLC-grade water. Samples were analysed using the HPLC method described in 2.2.7. 95pl of each extract was injected on to the column along with 5jil of an authentic testosterone standard (see 2.2.7.2).

To calculate the extraction efficiency, aliquots of the original plasma sample (5 pi) and reconstituted extract (20pl) were added to scintillation vials containing 3.5ml scintillation fluid and counted for one minute on a Wallac 1410 liquid scintillation counter. All samples were counted in duplicate. The extraction efficiency was calculated as follows:

% Extraction efficiency = ______dpm/extract _____ x 100 dpm / 50fÀ plasma sample

76 2.2.10.3 Quantification of ^'‘C-Testosterone in Plasma Sample Extracts

Aliquots of the combined organic phases (500pl) and post extraction aqueous phase (200pl) were added to scintillation vials containing 3.5ml scintillation fluid and counted for one minute on a Wallac 1410 liquid scintillation counter. The organic phase was counted in triplicate. Due to sample volume limitations, a single determination was performed for each post extraction aqueous phase.

The extraction recovery was calculated as follows. Aliquots of the original plasma sample (5pi) were added to scintillation vials containing 3.5ml scintillation fluid and counted for one minute on a Wallac 1410 liquid scintillation counter. The following calculation was used:

% Extraction recovery = dpm / organic phase + dpm / aqueous phase ^ jgg

dpm / 50jj1 plasma sample

2.2.10.4 Calculation of Kinetic Parameters

Plasma '^C-testosterone concentration-time curves were analysed by non- compartmental analysis (i.v. bolus model) using WinNonlin software (Version 3.2)

(Pharsight Corporation, Mountain View, U.S.A.). The area under the plasma concentration-time curve (AUC) from the time of dosing to the last measured concentration (AUCiast) was calculated using the linear trapezoidal rule. The terminal elimination fate constant (Xz) is the slope of the terminal phase of the plasma concentration-time profile on a semi-logarithmic scale, which was determined by linear regression of at least three data points in the terminal phase. AUCmf, the area under the plasma concentration-time curve from the time of dosing extrapolated to infinity, was calculated using the equation below:

AUCirf = AUC,as, + — EIï L

where, C^st is the plasma concentration at the last measured timepoint.

77 Comparison of the AUCiast and AUCinf values for each animal revealed that the

AUCiast value provided a more accurate estimate of the AUC for this data set. This was due to the fact that the 1 % value is derived from linear regression using the last few data points, which were subject to greater variability due to the low levels of ^"^C-testosterone present in these samples. The Xz value is used to calculate the extrapolated portion of the AUCmf and therefore resulted in inaccurate estimations of this parameter. All kinetic parameters calculated for this data set were therefore based on non-extrapolated area values.

The plasma clearance (CL) of ^"^C-testosterone was calculated using the equation below: Dose CL = AUCiast

The volume of distribution at steady state (Vgg) was estimated using the equation below:

Vss = MRTiast*CL where, MRTiast (mean residence time from the time of dosing to the last measured concentration) was calculated using the equation below:

M R T ,.. . v\LK4wt where, AUMCiast was the area under the moment curve from the time of dosing to the last measured concentration, which was calculated using the linear trapezoidal rule.

78 2.2.11 Western Blotting

Western blotting was used to study the relative levels of CYPIA, 2B, 3A and 4A proteins in microsome samples from different experimental groups. Rat cytochrome P450 ECL™ Western blotting kits (Amersham) were used for all experiments. Each kit contained ECL molecular weight markers, positive control rat liver microsomes, anti-rat cytochrome P450 primary antibody, a species-specific Ig-biotinylated secondary antibody and streptavidin-horseradish peroxidase (HRP) conjugate. The anti-rat cytochrome P450 primary antibodies were polyclonal and showed cross reactivity with CYP forms belonging to the same subfamily (e.g. anti-CYP2B recognises CYP2B1 and 2B2 proteins) (except anti-CYPlA2 which did not cross react with CYPlAl) but not members of other subfamilies (Amersham data sheets). Bound antibodies were visualised using ECL™ Plus Western blotting detection reagents. This detection system is based on streptavidin-HRP catalysed oxidation of the Lumigen PS-3 acridan substrate to form acridinium ester intermediates. Reaction of these intermediates with peroxide produces a high intensity chemiluminescence. Which is captured on film. The following protocol is based on the method supplied by the manufacturer.

2.2.11.1 Preparation of Samples

Liver microsomes were diluted to a final protein concentration of 1 to 2 mg/ml in 5OmM potassium phosphate buffer (pH 7.4). Testis microsomes were not diluted. Samples were then mixed with an equal volume of loading buffer (see 2.1.3), heated to 95-100°C for 5 minutes and then placed on ice.

2.2.11.2 SDS-Polyacrylamlde Gel Electrophoresis (SDS-PAGE)

Proteins were separated by SDS-P AGE using a Bio-Rad Mini-PROTEAN II dual gel system. A 12% resolving gel was prepared by mixing the following reagents together is this order: 3.55ml MilliQ water, 4.5ml 2 x resolving gel buffer (see 2.1.3), 7.5ml 30% (w / v) acrylamide, 2.95ml 2% (w / v) bisacrylamide, 18pl N,N,N’,N’- tetramethyethylenediamine (TEMED) and 0.18ml 10% (w / v) ammonium persulphate. The gel was cast and 1ml MilliQ water was gently layered on top. The gel was allowed to set at room temperature for approximately 1 hour and then the layer of water was carefully removed using filter paper.

79 A 4% stacking gel was prepared by mixing the following reagents together in this order: 6.9ml MilliQ water, 2.5ml 2 x stacking gel buffer (see 2.1.3), 1ml 30% (w / v) acrylamide, 0.4ml 2% (w / v) bisacrylamide, lOpl TEMED and 50pl 10% (w / v) ammonium persulphate. The stacking gel was layered on top of the resolving gel and a comb was inserted to create ten wells. The stacking gel was allowed to set at room temperature for approximately 3 hours.

The gel was placed in to an electrophoresis tank containing 1 x running buffer (see 2.1.3). Two lanes on each gel were reserved for the molecular weight markers and positive control, which were reconstituted and loaded as described in the appropriate kit. Microsomal protein samples (5-lOpl per lane) were loaded in to the remaining six lanes of the gel. Electrophoresis was performed at room temperature at 40mA for 1.5 to 2 hours until the loading dye was approximately 1 cm from the bottom of the gel.

2.2.11.3 Transfer of Proteins to Nitrocellulose Membrane

Following electrophoresis, proteins were transferred on to nitrocellulose membranes as follows. The nitrocellulose membrane, blotting paper and scotchbrite fibre pads were equilibrated with 1 x transfer buffer (see 2.1.3) for 10 minutes. The gel and nitrocellulose membrane were assembled in the transfer cassette in the following order: scotchbrite pad, 2 sheets of filter paper, gel, nitrocellulose membrane, 2 sheets of filter paper, scotchbrite pad. The transfer cassette was placed in, to an electrophoresis tank containing 1 x transfer buffer. Transfer was conducted overnight at 100mA at room temperature.

2.2.11.4 Immunodetection

Following protein transfer, the membranes were placed in to 50ml tubes, which were placed on a roller throughout the immunodetection procedure. The following steps were carried out at room temperature. First, membranes were blocked for 1 hour using 5% (w / v) fat free dried milk powder diluted in TBS-Tween (see 2.1.3). The membranes were then washed once for fifteen minutes and twice for five minutes in fresh changes of TBS-Tween. Next, membranes were incubated for 1 hour with anti­ cytochrome P450 primary antibody diluted 1:1000 (v / v) in diluent (1% (w / v) fat free dried milk powder in TBS-Tween). The membranes were washed in TBS-Tween

80 (as previously described) and then incubated for one hour with species-specific Ig- biotinylated secondary antibody diluted 1:2000 (v / v) in diluent. The membranes were washed in TBS-Tween (as previously described) and then incubated for 30 minutes with streptavidin-HRP conjugate diluted 1:2000 (v / v) in diluent. The membranes were then washed once for fifteen minutes and four times for five minutes in fresh changes of TBS-Tween. The membranes were then removed from the tubes and placed in to plastic petri dishes, protein side uppermost. Membranes were covered with ECL Plus detection reagents (mixed as described in the manual) and incubated for 5 minutes. Membranes were drained to remove excess detection reagents and then exposed to film (Polaroid 667) in a mini camera.

2.2.11.5 Staining of membranes with Ponceau S

Ponceau S staining was carried out to confirm even protein loading and to evaluate protein transfer from the gel to the membrane. Each membrane was immersed in Ponceau S working solution (Ponceau S concentrate diluted 1:10 (v / v) in MilliQ water) for 1 to 2 minutes and then rinsed with 5% (v / v) acetic acid until the protein bands were visible.

2.2.12 ELISA

An enzyme-linked immunosorbant assay (ELISA) was employed as a semi- quantitative assay to compare the relative levels of CYPIA, 2B, 3 A and 4A proteins in microsome samples firom different experimental groups. Primary antibodies (goat anti-rat cytochrome P450 primary antibodies), secondary antibody (rabbit anti-goat IgG-horseradish peroxidase) and positive control rat liver microsomes were purchased firom Gentest. The anti-rat cytochrome P450 primary antibodies were polyclonal and showed cross reactivity with GYP forms belonging to the same subfamily (e.g. anti- CYPlAl recognises CYPlAl and 1A2 proteins) (Gentest data sheets). During the optimisation of this method the binding specificity of each antibody was confirmed by Western blotting (see chapter 3). Bound antibodies were detected using streptavidin- HRP catalysed oxidation of 3,3’,5,5’ tetramethylbenzidine (TMB). Addition of acid to the oxidised TMB forms a yellow colour, which is measured at 450nm. The assay was performed using a Standard Operating Procedure (SOP) obtained from AstraZeneca.

81 A standard curve was constructed using liver microsomes from 3-methylcholanthrene (CYPIA), phénobarbital (CYP2B and 3A) or cloflbrate-treated rats (CYP4A). These microsomes contain a predetermined amount of the appropriate cytochrome P450 protein. Standard microsomes were diluted in solubilisation buffer (see 2.1.4) to give, final microsomal protein concentrations spanning the 0.195 to 100p,g/ml range (0.195,

0.39, 0.78, 1.56, 3.125, 6.25, 12.5, 25, 50 and lOOpg/ml). Test samples were diluted in solubilisation buffer to a final microsomal protein concentration of 0.05mg/ml.

The assay was conducted in 96-well Maxisorp ELISA plates (Fisher). lOpl of each standard microsome dilution was transferred to the assay plate to give the following amounts of microsomal protein per well: 1.95, 3.9, 7.8, 15.6, 31.3, 62.5, 125, 250, 500 and lOOOng/well. Standard samples were assayed in triplicate. A blank well, in which microsomes were replaced with an equivalent volume of solubilisation buffer, was also prepared. lOpl of each test sample was transferred to the assay plate (final microsomal protein concentration of 500ng/well) and each sample was assayed in duplicate.

For the assay, 190pl of carbonate/ bicarbonate buffer (see 2.1.4) was added to each well followed by incubation of the plate at 37°C for one hour. The microsomes were removed and the plate was washed four times with fresh changes of wash buffer (200pl) (see 2.1.4) using an automated 96-well plate washer (Dynatech Laboratories). lOOpl of diluted primary antibody (1:4000 (v / v) in Superblock buffer (a proprietary protein formulation in phosphate buffered saline (pH 7.4)) (Fisher)) was added to each well, followed by incubation of the plate at 37°C for one hour. The primary antibody was removed and four wash cycles performed as previously described. lOOpl of diluted secondary antibody (1:7500 (v / v) in Superblock buffer) was added to each well, followed by incubation of the plate at 37°C for one hour. The secondary antibody was removed and four wash cycles performed as previously described. lOOpl of Turbo-TMB substrate (a proprietary formulation containing TMB and stabilised hydrogen peroxide (Fisher)) was added to each well and the plate was incubated at room temperature for 10 to 30 minutes to allow colour development. lOOpl of sulphuric acid (IM) was added to each well and the absorbance at 450nm measured using a Victor 2 microplate reader (Wallac).

82 The concentration of cytochrome P450 proteins in test samples was interpolated from the appropriate standard curve using Origin software (logistic fit). Mean values were calculated for the control and drug-treated groups. Fold induction was calculated by dividing the mean of the drug-treated group by the mean of the control group.

2.2.13 Optimisation of the ELISA Assay

The ELISA method described in 2.2.12 was the result of the optimisation experiments described below.

2.2.13.1 Specificity of the anti-CYP Primary Antibodies

The binding specificity of the anti-CYP primary antibodies used for the ELISA assay was evaluated by Western blotting using liver microsomes firom control and inducer- treated animals (from study 1). The general protocol described in 2.2.11 was used except for the following modifications to the immunodetection procedure. The primary antibodies (goat anti-rat cytochrome P450 primary antibodies) were diluted 1:500 (v / v) and the secondary antibody (rabbit anti-goat IgG-horseradish peroxidase) was diluted 1:50,000 (v / v). The secondary antibody was conjugated to HRP therefore the step in which the membranes were incubated with streptavidin-HRP was omitted. Finally, the membranes were developed with ECL detection reagents (mixed as described in the manual) and then exposed to film.

2.2.13.2 Linear Range of Microsomal Protein Concentrations

The linear range of the assay was determined with respect to the amount of microsomal protein added to each well. Standard curves were constructed by serial dilution of liver microsomes (from study 1 or purchased from Gentest) in solubilisation buffer and each sample was assayed in triplicate as described in 2 .2 . 1 2 .

83 2.2.14 Preparation of Samples for Real-time PCR (TaqMan™)

For all nucleic acid work, gloves were worn throughout and all glass and plasticware were sterilised by autoclaving at 121°C for 20 minutes at 15 Ib/m^.

2.2.14.1 Total RNA Extraction

Total RNA was isolated from tissue samples using Trizol® Reagent, according to the technical manual provided by the manufacturer (Gibco). Trizol® Reagent is a monophasic solution of phenol and guanidine isothiocyanate which disrupts and dissolves cellular components whilst maintaining the integrity of the RNA. RNA is separated from DNA, proteins and other contaminants by adding chloroform followed by centrifugation. RNA remains exclusively in the aqueous phase and is recovered by precipitation with isopropanol.

Liver and Testis

The following steps were performed at room temperature unless otherwise stated. Frozen tissue samples (« 80mg) were ground under dry ice using a mortar and pestle and then homogenised in 1ml Trizol® Reagent using a glass-Teflon® hand held homogeniser. Samples were incubated for 5 minutes to allow the complete dissociation of nucleoprotein complexes. 200pl of chloroform was added and the tubes were shaken vigorously for 15 seconds. Samples were incubated for 3 minutes and then centrifuged at 12,000g for 15 minutes (4°C). The upper aqueous phase, containing the RNA, was transferred to a clean tube. Samples were incubated for 10 minutes with 500pl of isopropanol and then centrifuged for 10 minutes at 12,000g

(4°C). The supernatant was discarded and the RNA pellet was washed by mixing each sample with 1ml of 75% (v / v) ethanol followed by centrifugation at 7,500g for 5 minutes. The supernatant was discarded and the RNA pellet allowed to air dry. The pellet was resuspended in 40pl nuclease-free water.

Pituitary Glànds

Total RNA was isolated from pituitary glands as previously described for liver and testis samples except for the following modifications. Pituitary glands, in 500pl

Trizol® Reagent (see 2.2.1.5), were removed from the -80°C freezer and thawed on

84 ice. The tissue was disrupted using a Polytron PT3100 homogeniser. The volumes of chloroform and isopropanol added were reduced to lOOpl and 250pl respectively to correct for the smaller volume of Trizol® Reagent used at the start.

2.2.14.2 Purification of RNA (Ambion RNAqueous’^-4PCR Kit)

Total RNA was purified using materials and solutions firom the kit supplied by Ambion, according to the technical manual provided. This kit uses silica-based filters which selectively bind RNA allowing DNA, protein and other contaminants to be removed using a series of washing steps. The pure RNA is then eluted from the filter and treated with DNase 1 to remove any residual DNA contamination.

Liver and Testis Samples

The following steps were performed at room temperature unless otherwise stated. RNA isolated using Trizol® Reagent was added to 1ml of the Lysis/Binding Solution from the kit. An equal volume of 64% ethanol was added, samples were mixed thoroughly and then transferred on to the filter cartridges. Samples were moved through the filter by centrifugation at 12,000g for 1 minute. The RNA was washed once with 700pl of Wash Solution 1 and then twice with 500pl of Wash Solution 2. For each washing step, the Wash Solution was added to the filter cartridge and then moved through the filter by centrifugation as previously described. The RNA was eluted firom the filter in two steps using a total volume of 80pl of hot Elution Solution

(O.lmM EDTA) (100°C). Samples were then incubated at 37°C for 30 minutes with

Ipl DNase 1. The reaction was terminated by adding 9pl of DNase Inactivation Reagent, which was then removed by centrifugation at 10,000g for 1 minute. The purified RNA was transferred to a clean tube and stored at -20°C.

Pituitary Gland Samples

RNA isolated from pituitary gland samples was purified as previously described for liver and testis samples except for the following modifications. The RNA was added to 200pl of the Lysis/Binding Solution and was eluted from the filter using 50pl hot Elution Solution.

85 2.2.14.3 Determination of Total RNA Yield and Quality

Liver and Testis Samples

The concentration and purity of RNA samples were measured using a Genequant™ II spectrophotometer (Pharmacia Biotech). An aliquot of each sample was diluted 1:35 in nuclease-free water and the absorbance was measured at 260nm and 280nm. Nuclease-free water was used as a blank. The concentration of RNA was determined

from the absorbance at 260nm. The ratio of A 2 6 0 to A 2 8 0 values provides a measure of

RNA purity. Pure RNA has an A 2 6 0 • A2 8 0 ratio between 1.8 and 2.1.

The overall quality of RNA. samples was assessed by denaturing gel electrophoresis. A 1% (w / v) agarose gel was prepared by dissolving 0.5g agarose in 50ml 1 x Tris- acetate (TAB) buffer (see 2.1.5) by heating in a microwave oven. Ethidium bromide (0.2pg/ml) was added and the gel was allowed to cool to approximately 50°C. The agarose was poured in to a gel casting tray containing a comb to create loading wells and was allowed to set. The gel was then transferred in to an electrophoresis tank containing 1 x TAE buffer. RNA samples were mixed with Gel Loading Solution (0.05% (w / v) bromophenol blue, 40% (w / v) sucrose, 0.1% (w / v) SDS, O.IM EDTA (pH 8) (Sigma)) (4pl: Ipl RNA (v / v)) and then loaded on to the gel. Electrophoresis was conducted at lOOV until the loading dye had migrated approximately three quarters of the length of the gel. The gel was visualised using a UV transilluminator. Total RNA should give two.fairly sharp bands, corresponding to 28S and 18S ribosomal RNA. The intensity of the 28S band should be approximately twice that of the 18S band for intact RNA.

Pituitary Gland Samples

The concentration and quality of RNA isolated from pituitary gland samples was ■ assessed by denaturing gel electrophoresis using an RNA standard curve. A 1% (w / v) agarose gel was prepared as previously described for liver and testes samples except

that 0.5 X Tris-borate (TBE) buffer (see 2.1.5) was used in place of 1 x TAE buffer.

A standard curve was constructed using material (brain RNA) for which the total RNA content had been accurately quantified. This sample was diluted in nuclease-free

86 water and 2pi Gel Loading Solution to give solutions with the following RNA concentrations: 0.01, 0.025, 0.05, 0.1 and 0.2 pg/pl. Pituitary gland samples were mixed with 2pl Gel Loading Solution and 6pl nuclease-free water. All samples were incubated for 5 minutes at 70°C using a dry heat block and then placed on ice for 5 minutes. lOpl of each sample was loaded on to the gel. Electrophoresis was conducted at 140V until the loading dye had migrated approximately three quarters of the length of the gel. The gel was visualised using an Image Master VDS (Amersham). The concentration of RNA isolated from pituitary gland samples was estimated by comparing the band intensity in each sample to those of standard curve samples.

2.2.14.4 First-Strand cDNA Synthesis

First strand cDNA synthesis was carried for each total RNA sample using a Superscript II Reverse Transcriptase (RT) kit, according to the technical manual provided by manufacturer (Gibco). This kit contained Superscript II RNase H" reverse transcriptase (RT) (200U/pl), 5 x First-Strand buffer (375mM KCl, 15mM MgClz,

250mM Tris-HCl (pH 8.3)) and O.IM dithiothreitol (DTT). Reactions contained 3pg total RNA for liver and testis samples or Ipg total RNA for pituitary gland samples. The reaction was primed using random hexamers, therefore all RNA species in a population will act as templates for first-strand synthesis.

For the reaction, 200ng random primers were added to each aliquot of RNA and nuclease-free water was added to a final volume of 11 pi. The tubes were heated to

70°C for 10 minutes using a dry heat block and then immediately placed on ice. The following reaction components were added to each tube: 8pi 5 x First-Strand buffer, 2pi O.IM DTT, Ipl lOmM dNTP mix and Ipl Superase-In (20 U/pl). Superase-In is an RNase inhibitor, which was added to prevent degradation of the RNA template during cDNA synthesis. The tubes were mixed gently and then heated to 25°C for 10 minutes immediately followed by 42°C for 2 minutes.

Ipl of Superscript II RT (RT+) or nuclease-free water (RT-) was added to each tube. Inclusion of a reaction containing no RT enzyme (RT-) for each sample in addition to

87 the standard reaction (RT+) allowed the identification of any genomic DNA contamination. Following addition of the RT enzyme, the tubes were heated to 42°C for 50 minutes followed by 70°C for 15 minutes and then immediately placed on ice for 2 minutes. cDNA samples were stored at -20°C.

2.2.15 Real-time PCR (TaqMan™) Analysis of Candidate Genes

TaqMan™ was used to study the effects of lansoprazole treatment on mRNA levels for selected candidate genes in liver, testis and pituitary samples. A number of genes of relevance to our hypothesis were selected for quantitation

2.2.15.1 Background

Real-time PCR (TaqMan) is a powerful method that enables measurement of the abundance of a target sequence in a biological sample quantitatively (Holland et al, 1991; Lee et al., 1993; Livak et al., 1995). This technique measures the accumulation of a PCR product in real-time and differs from a standard PCR reaction in the addition of a single reagent, the fluorogenic probe designed to anneal to the target sequence. The probe generates a signal that accumulates during PCR cycling in a manner proportional to the concentration of amplification products (Holland et al., 1991; Lee et al., 1993).

Principle of the assay

A diagrammatic representation of a typical TaqMan reaction is shown in Figure 2.1. This technique uses a specific, non-extendable oligonucleotide probe, which anneals to a portion of the target sequence located between the two primer binding sites. The probe is labelled with a 5’-reporter dye (e.g. 6 -carboxy-fluorescein (FAM)) and a 3’- quencher dye (e.g. 6 -carboxy-tetramethyl-rhodamine (TAMRA)). When the probe is intact, the proximity of the quencher dye absorbs the fluorescent emission from the reporter dye (Livak et al., 1995).

During PCR cycling, the probe anneals to the template and is then cleaved by the 5’ to 3’ exonuclease activity of Taq DNA polymerase (Holland et al., 1991). This separates the reporter dye from the probe, resulting in an increase in reporter fluorescence Figure 2.1: Diagrammatic Representation of the steps in a typical TaqMan reaction

Step 1: DNA dénaturation 3’ 5’

5 ’ ------3 ’

3 ’

Step 2: Probe annealing y ------

5’ ------3 ’

Forward primer Step 3: Primer annealing y y and extension ______

Reverse primer

Step 4: 5’ exonuclease 5’ cleavage o f probe 3’ 5’

O I Step 5: Continuing y polymerisation y ^

(&) Reporter dye Quencher dye

Adapted from Lie & Petropoulos (1998)

89 without affecting emission of the quencher dye (Heid et al., 1996). This process occurs during each PCR cycle and the increase in reporter fluorescence is directly proportional to the amount of PCR product accumulated (Heid et al., 1996). Fluorescence emission is monitored in real-time during the PCR amplification using a sequence detector, which combines a 96-well thermal cycler, a laser for excitation of the fluorescent dyes and a system for data collection and analysis.

Data Analysis

The sequence detector software automatically creates an exponential amplification plot for each sample by plotting the normalised reporter (R„), a value which reflects the amount of annealed probe that has been cleaved, against the cycle number. The quencher dye emission intensity varies only minimally during PCR cycling and is used to normalise reporter dye emission (Heid et al., 1996). The cycle at which reporter fluorescence rises above baseline is known as the threshold cycle or C?. The Ct value is inversely proportional to the starting template copy number (Heid et al, 1996).

The Ct value of a test sample is used to determine the initial target copy number by interpolation from a standard curve amplified in the same PCR run. Levels of expression of a target sequence can be compared across experimental samples by expressing the data relative to an endogenous control gene, which is expressed at a constant level in all samples. Normalisation of data in this way compensates for factors such as intersample variation in efficiency of the reverse transcriptase reaction and inaccuracies in estimation of the concentration and quality of total RNA samples. A constitutively expressed “housekeeping” gene, such as p-actin, GAPDH or 18S ribosomal RNA (rRNA), is often used as an endogenous control.

90 2.2.15.2 Primer and Probe Design

Rat sequences for target genes were obtained from the NIH gene database (http:/www.ncbi.nlm.nih.gov/). TaqMan primers and probes were designed for each gene using Primer Express"^*^ software (Version 1) (Applied Biosystems) according to the guidelines sununarised in Table 2.4. The nucleotide sequences of primers and probes used for this study are shown in Table 2.5. Probes designed for target genes were labelled with the reporter dye FAM (5’ end) and the quencher dye TAMRA (3’ end). Primers and probes for target genes were purchased from Oswel DNA Services (University of Southampton). Primers and probes for 18S rRNA were purchased as a kit (Applied Biosystems). The probe was labelled with the reporter dye VIC (5’ end) and the quencher dye TAMRA (3’ end).

Table 2.4: Guildelines for the design of TaqMan primers and probes

Prim ers Probe

20 - 80% GC content

Length 9 - 40 bases

Tm 58 -60°C Tm 10°C higher than primer Tm

<2°C difference in Tm between the two No G on 5’ end primers Maximum of 2 G or C bases within the <4 contiguous G’s 5 nucleotides at the 5’ end 3’ end of the primer as close as possible Select the strand that gives the probe to the probe without overlapping it more G’s than C’s

91 o

m 2 A h

I 3 73ID u I g ' u M lU (4-, 0

D 1 I Vi o (D g o vo m LA 0 \ 7f- C\ On 73 o 00 vo ro ON V3 o 00 vo CN 0 O CN s V) Tf m s 00 o 5 O o s 1 % % 3 % 1 ûi m ri &) V3 Q

PCR reactions were performed in optical 96-well reaction plates (Applied Biosystems). Expression of 18S rRNA was used as an endogenous control to normalise data. Each plate was divided in to two halves: the left-hand side was used for the target gene and the right-hand side for 18S rRNA. Each side of the plate contained a standard curve, buffer blanks (containing no DNA) and RT+ and RT- cDNA for each test sample. Each reaction was performed in duplicate.

A standard curve was constructed by serial dilution of rat genomic DNA (1:10, 1:100, 1:400 and 1:1000 (v / v)). Five standard curve points were used and were designated arbitrary values of 0.1, 0.25, 1, 10 and 100. The standard curve was used to compare the relative levels of expression of the target gene between samples, therefore it was not necessary to determine the exact amount of genomic DNA present in each standard sample. cDNA samples were diluted 1:1000 (v / v) for 18S rRNA and 1:2 to 1:100 (v/ v) for the target gene. All dilutions were made using nuclease-ffee water.

Each PCR reaction contained the following components: 12.5pl qPCR™ Mastermix (proprietary mixture containing dNTPs, Hot Goldstar DNA polymerase, 5mM MgCl], Uracil-N-glycosylase, stabilisers and ROX passive reference dye) (Eurogentec), 300nM forward primer, 300nM reverse primer, lOOnM probe, Ipl template (genomic DNA dilution or cDNA sample) and nuclease-fi-ee water to a final volume of 25pl. - A mastermix, containing all reaction components except for the template, was prepared for each primer and probe set on each plate. 24pl of mastermix was added to each well followed by Ipl of template (genomic DNA dilution or cDNA sample). The plate was sealed with an optical adhesive cover (Applied Biosystems) and then centrifuged briefly. Real-time PCR analysis was performed with an ABI Prism II 7700, Sequence Detector (Applied Biosystems), using the universal cycling conditions shown in Table 2.6.

93 Table 2.6: The universal thermal cycling conditions used for TaqMan PCR reactions

Times and Temperatures

Initial Steps Each of 40 cycles Melt Anneal/Extend HOLD HOLD CYCLE

2 minutes 1 0 minutes 15 seconds 1 minute 50°C 95°C 95°C 60°C

2.2.15.4 Data Analysis

At the end of the PCR reaction, the sequence detector plots an exponential amplification plot for each well. The default baseline setting was adjusted to accommodate the earliest amplification plot and then the threshold was set just above the baseline. For each test sample, the initial copy number of the target gene or 18S rRNA was interpolated from the appropriate standard curve using the Ct value. The copy number of the target gene was then divided by the copy number of 18S rRNA to give a normalised target value. Fold changes were calculated by dividing the mean normalised target value in the drug-treated group by that in the control group.

2.2.16 Statistical Analysis

Unless otherwise stated, all statistical analysis was performed using a Student’s t-test (two-tailed).

94 2.3 Experimental Strategy

The overall aim of this project was to investigate the existence of a liver-testis axis wherein microsomal enzyme inducers enhance the metabolic clearance of testosterone resulting in a drop in circulating hormone levels. The consequent increase in serum LH levels may be responsible for the appearance of Leydig cell tumours (LCTs). This hypothesis was addressed by designing a series of animal studies, which are summarised in the following experimental plan.

95 Study 1: Effect of Model Inducers on Cytochromes P450, Testosterone Metabolism and Plasma Hormone Levels

Aims: The aim of this study was to characterise the effects of dosing rats with model inducers of the major GYP families (P-naphthoflavone (p-NT), phénobarbital (PB), pregnenolone-16a-carbonitrile (PCN) and ciprofibrate) on hepatic and testicular testosterone metabolism and the endocrine control of the testis.

Experimental plan:

Male Sprague- Dawley rats

Treatment with a model inducer for 4 days

Collection of samples

Liver Testes Plasma

Microsome Microsome preparation preparation

• Protein assay Protein assay Radioimmunoassay: • Total cytochrome P450 Testosterone hydroxylase Testosterone, LH • Testosterone hydroxylase assay FSE and prolactin assay , • Western blotting

General methods sections of relevance to this study: 2.2.1.1, 2.2.1.2, 2.2.2, 2.2.4, 2.2.5, 2.2.6, 2.2.7, 2.2.8 and 2.2.11.

96 Study 2: Effect of Lansoprazole on Cytochromes P450, Testosterone Metabolism and Plasma Hormone Levels

Aims: Lansoprazole induces hepatic cytochromes P450 and produces Leydig cell tumours in rats (Atkinson et al., 1990; Masubuchi et al., 1997). The aim of this study was to investigate the effects of lansoprazole treatment on GYP expression (CYPIA, 2B, 3A and 4A) in the liver and testis, hepatic and testicular testosterone metabolism and the endocrine control of the testis. Effects of lansoprazole were compared to the model inducers to help identify any changes in CYP-dependent testosterone metabolism that may be important in the induction of LCTs.

Experimental plan: Male Sprague- Dawley rats

Treatment with lansoprazole for 14 days

Collection of samples

Liver PlasmaTestes

Microsome Microsome preparation preparation

• Protein assay Protein assay Radioimmunoassay: • Total cytochrome P450 Testosterone hydroxylase Testosterone, LH • Testosterone hydroxylase assay FSH and prolactin assay Western blotting • Western blotting • ELISA

General methods sections of relevance to this study: 2.2.1.1,2.2.1.3, 2.2.2, 2.2.4, 2.2.5, 2.2.6, 2.2.7, 2.2.8, 2.2.11 and 2.2.12

97 Study 3: Effect of Lansoprazole on Gene Expression in the Liver and Testis

Aims: This study was designed to further characterise the effects of lansoprazole on the liver and testis with the following objectives: - further characterise effects on CYP expression in the liver and testis (CYP2À and CYP2C) - investigate effects on the expression of selected genes of relevance to our hypothesis (e.g. steroidogenic enzymes, LH-regulated protein) - further investigate effects on the HPT axis by measuring intratesticular testosterone levels

Experimental plan: Male Sprague- Dawley rats

Treatment with lansoprazole for

Collection of Plasma samples

Radioimmunoassay: Testosterone, LH Liver Testes FSH and prolactin

Snap frozen tissue

Microsome Supernatant preparation preparation RNA extraction & cDNA synthesis

• Protein assay TaqMan analysis of ' Radioimmunoassay: • Total cytochrome P450 gene expression Testosterone • ELISA

General methods sections of relevance to this study: 2.2.1.1, 2.2.1.4, 2.2.2, 2.2.3, 2.2.4,2.2.5, 2.2.6, 2.2.12, 2.2.14 and 2.2.15

98 Study 4: Effect of Lansoprazole on Gene Expression in the Pituitary Gland

Aim: The aim of this study was to further investigate the effects of lansoprazole treatment on the endocrine control of the testis by measuring plasma hormone levels and LH and prolactin mRNA levels in the pituitary gland.

Experimental plan: Male Sprague- Dawley rats

Treatment with lansoprazole for 14 days

Collection of samples

Pituitary gland Plasma

Snap frozen tissue

RNA extraction & cDNA synthesis

TaqMan analysis of Radioimmunoassay: gene expression Testosterone, LH FSH and prolactin

General methods sections of relevance to this study: 2.2.1.1, 2.2.1.5, 2.2.4, 2.2.14 and 2.2.15

99 Study 5: Effect of Lansoprazole on the Plasma Clearance of ^^C-Testosterone

Aim: The aim of this study was to investigate the effect of lansoprazole treatment on the plasma clearance of ^"^C-testosterone in vivo.

Experimental plan:

Male Sprague- Dawley rats

Treatment with lansoprazole for 14 days

Administration of ^"^C-testosterone

Collection of serial blood samples

Extraction of C- testosterone

HPLC analysis ' Quantitation of ^"^C- testosterone by liquid scintillation counting

General methods sections of relevance to this study: 2.2.1.1, 2.2.1.6, 2.2.7 and 2.2.10

100 Chapter 3

Method Optimisation 3.1 Introduction

This chapter will describe experiments that were performed in order to optimise and validate two of the techniques that were subsequently used as part of this project, namely the HPLC-based testosterone hydroxylase assay and the ELISA assay.

3.2 HPLC-based testosterone hydroxylase assay

3.2.1 Introduction

Testosterone hydroxylase assays have been widely used to monitor the levels of expression of hepatic cytochromes P450. This technique is based upon the fact that individual cytochromes P450 form characteristic and unique patterns of hydroxylated metabolites when incubated with steroids such as testosterone (Waxman, 1988). Certain hydroxylation reactions are catalysed by a single CYP form and can be used as a diagnostic probe to monitor the levels of expression of that enzyme in rat liver

microsomes (e.g. 2 a-hydroxylation catalysed by CYP2C11) (Morgan et al., 1985; Waxman, 1984; Waxman et al., 1987). Other hydroxylation reactions are catalysed by multiple CYP forms (e.g. 16a-hydroxylation catalysed by CYP2B1/2 and CYP2C11) (Wood et al., 1983; Waxman et al., 1987). These properties enable the simultaneous analysis of the levels of expression of multiple cytochromes P450 by measuring changes in testosterone hydroxylation.

For the current project, a HPLC-based testosterone hydroxylase assay was employed to study the effects of chemical inducers on CYP-dependent testosterone metabolism in the liver and testis. This assay involves incubation of microsomal samples with radiolabelled testosterone followed by separation and quantification of metabolites using a reversed-phase HPLC system. Metabolites are quantified by radiochemical detection and the identity of each peak is confirmed by co-chromatography with authentic testosterone standards injected on to the column with each sample. The following section will describe experiments performed to optimise this assay for use with microsomal samples firom the current project.

102 3.2.2 High pressure liquid chromatography (HPLC)

3.2.2.1 Separation of authentic testosterone metabolites

Figure 3.1 shows a representative separation of a mixture of authentic testosterone standards using the solvent cycle described in 2.2.7.3. The retention times of the standards during this experiment are summarised in Table 3.1. The retention times of testosterone standards were found to vary over repeated chromatographic runs but crucially the position of each metabolite on the chromatogram relative to the other metabolites remained constant. Such shifts in retention times were not entirely unexpected and are probably due to variations in the ambient temperature, as a column oven was not used. This finding was of little consequence because identification of metabolites is based on co-chromatography with authentic standards injected on to the column with each sample.

Figure 3.1: A representative chromatogram showing the separation of a mixture of authentic testosterone metabolites by HPLC

un i t • . 2a Testos. ' lip

16a 16p Absorbance And. lA 15a (240nm) &/ 7a l / . 15P 6 P L/ V ,1 k II © ,UUJ 10 20 40 50 Time (mins)

Standard solutions were prepared as described in 2.2.7.2. Approximately 1.8 nmoles o f each hydroxylated metabolite and 0.18 nmoles of testosterone (Testos.) and androstenedione (And.) were injected on to the column. Elution of analytes was monitored by optical absorbance at 240nm.

103 Another characteristic of the chromatogram shown in Figure 3.1 is the gradual rise in baseline absorbance during the chromatographic run. This is due to the increasing proportion of organic solvent (mainly acetonitrile) which also absorbs at 240nm wavelength. This does not interfere with peak integration and quantification as this is based upon measurement of radiolabelled metabolites whilst the optical absorbance trace is used solely as a means of peak identification.

Good resolution was achieved for all of the metabolites examined except for lip- and

2a-0HT, which were only partially separated using the current HPLC system. Further

optimisation of the solvent cycle was not necessary because lip-OHT was not detected in any of the incubation extracts firom the current project. This finding is consistent with the published literature, which states that lip-OHT is not formed metabolically by rat liver microsomes (Sonderfan et al, 1987).

Table 3.1 : Chromatographic retention times for testosterone and metabolites

Retention Time Standard Componnd Chemical Name (mins) 20.06 6a-0H T 4-Androsten-6a, 17p-diol-3-dione

22.30 15P-0HT 4-Androsten-15p, 17p-diol-3-dione

28.54 6P-0HT 4-Androsten-6p, 17P-diol-3-dione

30.18 15a-0HT 4-Androsten-l 5 a, 17p-diol-3-dione

31.42 7a-0H T 4-Androsten-7a, 17p-diol-3-dione

38.54 16a-0HT 4-Androsten-16a, 17p-diol-3-dione

40.48 16P-0HT 4-Androsten-l 6 p, 17p-diol-3-dione

44.42 lip-O H T 4-Androsten-l Ip, 17p-diol-3-dione

45.18 2a-0H T 4-Androsten-2a, 17p-diol-3-dione

46.12, 2P-0HT 4-Androsten-2p, 17p-diol-3-dione 50.42 Androstenedione 4-Androsten-3,17-dione

53.12 Testosterone 4-Androsten-17p-ol-3-one

Standard solutions were prepared as described in 1 2 .1 2 . Approximately 1.8 nmoles of each hydroxylated metabolite and 0.18 nmoles of testosterone and androstenedione were injected on to the column. Elution of analytes was monitored by optical absorbance at 240nm.

104 3.2.2.2 Linear range of radiochemical detection

A study was performed to confirm that the radioactivity detector produced a linear response with respect to the amount of radioactivity injected on to the HPLC column. Serial dilutions of ^"^C-testosterone were prepared containing amounts of radioactivity

spanning the range expected to be present in incubation extracts (i.e. 1 0 0 to 600,000 dpm/lOOpl injected). Each sample was injected on to the HPLC column in triplicate and was also quantified by liquid scintillation counting (LSC). The results are shown in Figure 3.2.

This study demonstrates that the radioactivity detector displays linear detection between 1 0 0 and 600,000 dpm on column, which spans the range of activities typically analysed in incubation extracts. The limit of detection was approximately 9 pmoles equivalent on colunm.

Figure 3.2: The linear range of detection of the HPLC radioactivity detector

120000

100000 ,Ff=1

80000

60000

5: 40000

20000

0 100000 300000500000 600000 700000 800000 LSC <*ni100 nicroiitres

Serial dilutions of '“^C-testosterone were prepared and lOOpl aliquots of each dilution were counted by HPLC and LSC. Each point represents the mean value of triplicate determinations. Linear regression was performed and the value is shown on the graph.

105 3.2.3 Microsomal incubations using liver microsomes

3.2.3.1 Timecourse

A study was performed to determine the linear range of metabolite formation with respect to incubation time for rat liver microsomes. Liver microsomes from control and lansoprazole-treated animals were incubated with testosterone for various incubation periods as described in 2.2.9.1. A representative time course of formation for three metabolites is shown in Figure 3.3. The linear ranges for all metabolites studied are shown in Table 3.2.

Table 3.2: Linear range of metabolite formation by rat liver rhicrosomes with respect to incubation time

Treatment group M etabolite Linear range (mins)

6P-0HT O-.IO 0.99 Control 16a-0HT 0-7.5 0.98

2a-0H T 0 - 1 0 &98

7a-0H T 0 - 1 0 0.99

2P-0HT 5-20 0.98

6a-0H T 0 - 1 0 0.99 Lansoprazole 6P-0HT 0 - 1 0 0.99

16a-0HT 0 - 1 0 0.97

2a-0H T 0 -10 O j#

7a-0H T 0 - 1 0 0.99

2P-0HT 5-20 0.99

Liver microsomes from control and lansoprazole-treated animals (from Study 2) were incubated with testosterone as described in 2.2.9,1. Each incubation was performed in duplicate. Linear regression analysis was performed and values were calculated.

Metabolite formation was proportional to incubation time up to 10 minutes for 6a-,

6P-, 7a- and 2a-0HT and up to at least 7.5 minutes for 16a-0HT. 2P-0HT was not

106 Figure 3.3: The formation of three testosterone metabolites as a function of the incubation time by rat liver microsomes

25 -,

i

I -6b-OHT -2a-OHT -7a-OHT

0 5 10 15 20 25 Time (mins)

Liver mlCTOSomes from a lansoprazole-treated animal (from Study 2) were incubated with testosterone as described in 2.2.9.I. Each data point represents the average value from duplicate incubations.

Figure 3.4: The formation of three testosterone metabolites as a function of the microsomal protein concentration

25

20

15 -6b-OHT -16a-OHT -2a-OHT 10

5

0 0.5 1.5 2 2.5 3 3.5 4.5 Microsomal protein concentration (mg/ml)

Liver microsomes from a control animal (from Study 2) were incubated with testosterone as described in 2.2.9.2. Each data point represents the average value from duplicate incubations.

107 detected at the one minute timepoint but metabolite formation was linear between 5 and 20 minutes. 15a-0HT and 16p-0HT were not consistently detected in incubation extracts from this study therefore it was not possible to draw firm conclusions for these metabolites. Based on these findings, a standard incubation time of five minutes was selected.

3.2.3.2 Protein Study

A study was performed to determine the linear range of metabolite formation with respect to the microsomal protein concentration for rat liver microsomes. Testosterone was incubated with liver microsomes from control and lansoprazole- treated animals at increasing microsomal protein concentrations as described in 2.2.9.2. The effect of the microsomal protein concentration on the formation of three metabolites is shown in Figure 3.4. The linear ranges for all metabolites studied are shown in Table 3.3.

Table 3.3: Results of linear regression analysis of data from the protein study

Treatment group M etabolite Linear range R' (mg/ml) 6P-0HT 0 - 2 0.99 Control 16a-0HT 0 - 2 0.99

2a-0H T 0 - 2 0.99

7a-0H T 0 - 2 0.98

2P-0HT 0.75-1.5 - 0.96''

6P-0HT 0 - 2 0.95 Lansoprazole 16a-0HT 0 - 2 0.97

2a-0H T 0 - 2 0.97

7a-0H T 0 - 2 0.97

2P-0HT 0.5 -1.5 0.96

Liver microsomes from control and lansoprazole-treated rats (from Study 2) were incubated with testosterone as described in 2.2.9.2. Each incubation was performed in duplicate. Linear regression analysis was performed and values were calculated. #, indicates an value where only three data points were used for linear regression analysis.

108 Metabolite formation was proportional to the protein concentration up 2 mg/ml for

6P-, 16a-, 2a- and 7a-0HT. 2P-0HT was not detected in incubation extracts at lower microsomal protein concentrations (below 0.5 mg/ml), but metabolite formation was linear up to 1.5 mg/ml. 15a-0HT and 16P-0HT were not consistently detected in incubation extracts from this study therefore it was not possible to draw firm conclusions for these metabolites. Based on these findings, a standard microsomal protein concentration of 0.5 mg/ml was selected.

3.2.3.3 Effect of 4-MA on microsomal testosterone metabolism

In addition to cytochromes P450, rat liver microsomes also contain other enzymes that are capable of metabolising testosterone, such as the 5a-R. The 5a-R inhibitor, 4-MA, is often added to incubation mixtures to prevent extensive conversion of testosterone and its metabolites to 5a-reduced forms (Sonderfan & Parkinson, 1988). A study was performed to investigate the effect of 4-MA on testosterone metabolism by rat liver microsomes. Liver microsomes from control and lansoprazole-treated animals were incubated with testosterone in the presence or absence of 4-MA (IpM) as described in 2.2.9.4. This concentration of 4-MA was selected as it has previously been shown to completely inhibit steroid 5a-reductase activity without inhibiting other pathways of testosterone oxidation catalysed by liver microsomes (Sonderfan & Parkinson, 1988). Inclusion of 4-MA in incubation mixtures had no marked effects on testosterone metabolism by rat liver microsomes as shown in Figure 3.5. 4-MA might have produced slight increases in the accumulation of some hydroxylated testosterone metabolites but it was not possible to assess the statistical significance of this effect as single incubations were performed in the absence of 4-MA. In the absence of any adverse effects of 4-MA on testosterone metabolism it was decided to include 4-MA (IpM) routinely in all incubation mixtures.

109 <

4- 0Û c 1 13 •S I i

o \ (N (N •S . Vi 73(U o « 8 I I (U > a

X3 1*o e a00 0 II 1 0 I e § D 1

Ç 1) 00 B0 g I 3 I % 1 ■| I (U 1 a 8 'o cd I 'a , I ■fi •S 73 1 oI

a I Î a (S3|0UJU) pouijoi aiiioqejaui p junoiuv §1 3.2.3 A Stability of incubation extracts

A study was performed to determine whether incubation extracts and authentic testosterone standards were sufficiently stable at room temperature to allow the use of a non-reffigerated autosampler. Aliquots of reconstituted incubation extracts and mixtures of authentic testosterone standards were injected at 0 hours and then following 24 hours incubation at room temperature. An example of the data obtained from stability studies with incubation extracts is shown in Figure 3.6. Mixtures of authentic testosterone standards and metabolites present in incubation extracts were found to be stable at room temperature for at least 24 hours.

3.2.4 Microsomal incubations using testis microsomes

Studies were performed to determine the linear range of metabolite formation with respect to the incubation time and the microsomal protein concentration for testis microsomes. Unfortunately, due to the low levels of testosterone hydroxylase activity present in testis microsomes these experiments failed to yield any useful data. Consequently, the incubation conditions for testis microsomes were based upon the optimisation experiments previously described for liver microsomes. A standard incubation time of 5 minutes was selected as this was previously used for liver microsomes and it was sufficient for the formation of detectable levels of all of the metabolites formed by testis microsomes. A standard microsomal protein concentration of 2mg/ml was used because this concentration was at the top of the linear range for liver microsomes therefore it would be expected to be within the linear range for testis microsomes, which contain less cytochrome P450 per milligram of protein. The use of a higher protein concentration would also increase the absolute amount of metabolites formed and thereby facilitate their detection and accurate quantification. This protein concentration has also been used by other groups measuring testosterone hydroxylase activity in testis microsomes from rats (Sonderfan et al., 1989).

I l l MS mwg

I % % I <- P '.s'» ■ J1.3H. I .s

O., %

a 8 8 «Î .S Ov \ le *0 I (N B

*W^ T | I

T3 #_ '4 1 ------g "44

i 1 1 < I «BBl •s ■s \ (U H

'O rn r U ti If) in o in o in o in o \ & o o G) O) CO 00 M- S 3.2.5 Discussion

The HPLC system has been shown to be capable of resolving at least eleven potential metabolites of testosterone, which are expected to include all of the major and the majority of the minor pathways of testosterone metabolism catalysed by liver microsomes. The linear response of the radioactivity detector has been demonstrated across the range of activities typically analysed in incubation extracts. In addition, studies have confirmed that authentic testosterone standards and metabolites present in incubation extracts are sufficiently stable to allow the use of a non-refirigerated autosampler, enabling a marked increase in sample throughput for the assay.

The microsomal incubation procedure has been optimised with respect to incubation time and microsomal protein concentration for liver microsomal samples. The substrate concentration used in this assay (250pM) would be expected to be saturating with respect to the cytochromes P450 involved in testosterone metabolism based on the published Km value for testosterone metabolism by rat liver microsomes (26pM) (Kuntzman et al., 1965). This substrate concentration is used by the majority of testosterone hydroxylase assays from the published literature (Reinerink et al., 1991; Sonderfan et al., 1987; Wortelboer et al., 1991; Purdon & Lehman-McKeeman, 1997).

The 5a-R inhibitor, 4-MA, is often added to incubation mixtures to prevent extensive conversion of testosterone and its metabolites to 5a-reduced forms. Studies conducted by Sonderfan and Parkinson (1988) indicated that 4-MA increases the accumulation of hydroxylated testosterone metabolites through two mechanisms. Firstly, it inhibits 5a-reduction of metabolites that are substrates for the 5a-R enzyme (e.g. 6P-0HT). Secondly, it inhibits the formation of 5a-dihydrotestosterone, which acts as a competitive inhibitor of CYP-dependent pathways of testosterone metabolism catalysed by rat liver microsomes. In the current study, inclusion of 4-MA in incubation mixtures might have resulted in a slight increase in the accumulation of some hydroxylated testosterone metabolites, consistent with reports that 5a-R activity is low in liver microsomes firom adult male rats (Sonderfan & Parkinson, 1988). In the absence of any adverse effects of 4-MA on testosterone metabolism it was decided to include 4-MA (IpM) routinely in all incubation mixtures. This concentration has previously been shown to completely inhibit 5a-R activity without inhibiting other

113 pathways of testosterone oxidation catalysed by liver microsomes (Sonderfan & Parkinson, 1988).

In conclusion, a HPLC-based testosterone hydroxylase assay has been set up and optimised for use with microsomal samples from our laboratory. This method uses optical absorbance and radiochemical detection to identify and quantify metabolites respectively, which has a number of advantages. Firstly, use of the optical absorbance trace solely as a means of metabolite identification allowed a mixture of authentic testosterone standards to be injected with each sample. Overlaying the optical absorbance and radioactivity traces makes metabolite identification rapid and accurate and avoids potential problems associated with inter-run variability in metabolite retention times. Secondly, the use of radiolabelled substrate allows the construction of a mass balance relating the loss of counts as testosterone is metabolised to the appearance of counts as each metabolite is formed (Arlotto et al., 1991). Finally, the use of radiolabelled substrate enables the calculation of extraction efficiency without the need for an internal standard. Recovery of radioactivity using the current extraction procedure is typically within the range of 85 to 95%. Samples with extraction efficiencies of less than 70% were excluded from the analysis.

3.3 ELISA

All optimisation experiments for the ELISA assay were conducted by Alex Bell at AstraZeneca (Loughborough).

3.3.1 Introduction

An enzyme-linked immunosorbant assay (ELISA) was employed as a semi- quantitative assay to compare the relative levels of CYPIA, 2B, 3 A and 4A proteins in liver microsomes from different experimental groups. For the assay, 96-well plates are coated with solubilised microsomal samples and CYP proteins are detected using the principle of the antigen-antibody interaction. The plates were incubated with an anti-CYP primary antibody followed by a secondary antibody conjugated to streptavidin-HRP. The extent of antibody binding can then be quantified using the HRP catalysed oxidation of 3,3’,5,5’-tetramethylbenzidine (TMB) to form a coloured product, which is measured at 450nm using a spectrophotometer. The following

114 section will describe experiments that were conducted during the optimisation of this assay for use with liver microsomes.

3.3.2 Specificity of the anti-CYP primary antibodies

The binding specificity of the anti-CYP primary antibodies used for the ELISA assay was evaluated by Western blotting using liver microsomes from control and inducer- treated animals (from study 1). These experiments confirmed that the anti-CYPlAl did not cross react with CYP2B, 3A or 4A proteins in liver microsomes from control or inducer-treated animals (Figure 3.7). Similarly, anti-CYP3A2 and anti-CYP4Al did not cross react with CYP proteins belonging to other CYP subfamilies (data not shown). Anti-CYP2B1 showed some cross reactivity with CYP proteins from other subfamilies in liver microsomes from control and inducer-treated animals (data not shown). This was probably due to the high concentration of primary antibody used for Western blotting (1:500) whereas a much higher dilution (1:4000) is used for the ELISA assay. Each of these antibodies is known to cross react with other CYP proteins belonging to the same subfamily (e.g. anti-CYP2Bl cross reacts with CYP2B2) (Gentest data sheets).

3.3.3 Linear range of microsomal protein concentrations

The linear range of the assay was determined with respect to the amount of microsomal protein added to each well. Standard curves were constructed by serial dilution of liver microsomes and the range over which the microsomal protein concentration showed a linear relationship with the absorbance at 450nm was identified. , '

A representative standard curve from this study is shown in Figure 3.8. The linear range of the assay was found to be between approximately 0.1 and Ipg of microsomal protein per well for each of the antibodies used. At microsomal protein concentrations higher than Ipg per well the assay was saturated such that the increase in absorbance at 450nm no longer showed a linear relationship with the protein concentration (Figure 3.8). Based on these results, a microsomal protein concentration of 0.5 pg per well was used for all experiments.

115 Figure 3.7: Western blot of rat liver microsomes developed with anti-CYP 1 Al (Gentest)

Lane 1 2 3 4 5 6 7 8

5|ag microsomal protein was loaded in each lane. Lane 1, C YPlA l positive control liver microsomes from |3-NF-treated rats (Amersham) Lane 2, CYP1A2 positive control liver microsomes from isosaffole-treated rats (Amersham) Lane 3, CYPIA positive control liver microsomes from 3-MC-treated rats (Gentest) Lane 4, Liver microsomes from a P-NF treated animal Lane 5, Liver microsomes from a PB-treated animal Lane 6, Liver microsomes from a PCN-treated animal Lane 7, Liver microsomes from a ciprofibrate-treated animal Lane 8, Liver microsomes from a control animal

Figure 3.8: CYPIA ELISA standard curves constructed using liver microsomes from control and inducer-treated animals

1.6

•>*«• artmja 1 ( * batn-NPf

5.4 artroM tuiilfaaWt -K - CvF 1A sJardsfd c«vo

1.S -

06

06 -

04

02

10 100 lOiW IWOO In^ng pm lain}

Liver microsomes were assayed for CYPIA protein levels as described in 2.2.12. The CYPIA standard curve was constructed using liver microsomes from 3-MC-treated rats (Gentest). Each data point represents the mean ± SD from replicate determinations (n=3 for each protein concentrations).

116 3.3.4 Discussion

The ELISA assay has been optimised and will be used to quantify the levels of CYP proteins present in liver microsomes generated as part of this project. Western blotting experiments indicated that the anti-CYP antibodies would not be expected to cross react with CYP proteins belonging to other subfamilies at the dilutions used in the ELISA assay. However, these results can only be considered as a guide due to potential differences in antibody-antigen interactions between the denatured protein samples used for Western blotting and the solubilised proteins used for the ELISA. The ELISA assay has also been optimised to ensure to that the amount of microsomal protein added to each well is within the linear range of the assay. Overall, this assay has been shown to be a reproducible, sensitive and relatively rapid method for the quantification of CYPIA, 2B, 3 A and 4A proteins in rat liver microsomes.

117 Chapter 4

Effect of model inducers on cytochromes P450, testosterone metabolism and plasma hormone levels 4.1 Introduction

This study was designed to characterise the effects of dosing rats with model inducers of the major CYP families (P-NP-CYPIA, PB-CYP2B, PCN-CYP3A and ciprofibrate -CYP4A) on microsomal testosterone metabolism and the endocrine control of the testis. The quantitative and qualitative effects of these compounds on the hepatic CYP complement have previously been well characterised, however the physiological implications of concomitant changes in steroid hormone metabolism are poorly understood (Guengerich et al., 1982; Waxman et al., 1985; Wortelboer et al., 1991; Parkinson et al., 1992; Makowska et al., 1990; Zangar et al., 1996). The four model compounds selected for the current study all induce hepatic CYPs but there is no evidence that they produce Leydig cell hyperplasia or tumours in rats. It was therefore useful to compare and contrast the effects of these compounds on CYP-dependent pathways of testosterone metabolism to those of CYP inducers known to produce LCTs (e.g. lansoprazole). The overall aim of this study was to provide an insight in to the pathways that might be involved in xenobiotic-induced LCT formation in rats.

The main objective was to characterise the effects of these four compounds on hepatic CYP-dependent testosterone metabolism and plasma hormone levels (testosterone, LH, FSH and prolactin). In addition, experiments were performed to determine whether these compounds also influenced testicular CYP-dependent testosterone metabolism. Such changes could have important local effects on the testis because hydroxylated testosterone metabolites might possess unique biological or endocrine activities that could underlie important physiological functions (e.g. 7a-hydroxylated androgens may regulate testosterone biosynthesis (see section 1.5)).

4.2 Study design

Male Sprague Dawley rats were dosed once daily with P-NP (100 mg/kg/day) (p.o.), PB (80 mg/kg/day) (i.p.), PCN (60 mg/kg/day) (i.p.) or ciprofibrate (10 mg/kg/day) (p.o.) for four days. The dosages were selected as they have previously been shown to cause induction of the appropriate CYP enzymes in rats (Amacher & Schomaker, 1998; Kocarek & Reddy, 1996; Graham et al, 1996; Hanoika et al, 1995). On day five, blood samples were collected and liver and testis microsomes were prepared. More detailed experimental details concerning animal treatment, sample collection and the analytical methods used for this study can be found in ehapter 2 .

119 4.3 Results

4.3.1 Body, liver and testes weights

Final body weights and relative liver and testes weights for control and xenobiotic- treated animals are shown in Table 4.1.

Table 4.1: Effect of xenobiotic pretreatment on relative liver and testes weights

Treatm ent Final body Relative liver Relative testes group weight weight weight

(g) (g /lOOg body weight) (g /lOOg body weight) Control 416.0+10.5 3.89 ±0.29 0.86 ± 0.05

P-NF 416.4 ± 4.6 3.95 ± 0.29 0.87 ± 0.08

PB 415.0 ±10.8 4.65 ±0.26** 0.84 ± 0.06

PCN 418.2 ±10.5 4.32 ± 0.45 0.90 ± 0.06

Ciprofibrate 409.0 ± 6.7 5.27 ±0.23*** 0.89 ±0.08

Results are expressed as mean ± SD for each group of animals (n=5 animals per group). Values from each xenobiotic-treated group were compared to the control group using a Student’s t-test where asterisks (*) represent: * * P<0.01, *** P<0.001.

There were no statistically significant differences in initial (data not shown) or final body weights between control and drug-treated groups. Consistent with the changes in relative liver weights, significant increases in absolute liver weights (data not shown) were observed in PB- and ciprofibrate-treated animals (P<0.01 and P<0.001 respectively) whereas no significant changes were observed in P-NF or PCN-treated groups. There were no significant changes in absolute (data not shown) or relative testis weights in any of the xenobiotic-treated groups.

4.3.2 Plasma hormone levels

Blood samples were collected (~24 hours after the final dose) and assayed for plasma testosterone, LH, FSH and prolactin levels by radioimmunoassay (Figures 4.1 to 4.4). No statistically significant changes in plasma testosterone, LH, FSH or prolactin levels were observed in any of the xenobiotic-treated groups.

120 Figure 4.1: Plasma testosterone levels in control, p-NF, PB, PCN and ciprofibrate- treated animals

30

25

20

^ 10

Control B-NF PB PCN Qprofibrate Treatment Group

Results are expressed as mean ± SEM for each group (n=5 animals per group).

Figure 4.2: Plasma luteinising hormone (LH) levels in control, p-NF, PB, PCN and ciprofibrate-treated animals

Control B-NF PB PCN Qprofibrate TreeÉment Q"oup

Results are expressed as mean ± SEM for each group (n=5 animals per group).

121 Figure 4.3: Plasma follicle-stimulating hormone (FSH) levels in control, p-NF, PB, PCN and ciprofibrate-treated animais

PB Ciprofibrate Treatment Group

Results are expressed as mean ± SEM for each group (n=5 animals per group).

Figure 4.4: Plasma prolactin levels in control, p-NF, PB, PCN and ciprofibrate- treated animals

PB Ciprofilxate Treatment Group

Results are expressed as mean ± SEM for each group (n=5 animals per group).

122 4.3.3 Protein content of liver and testis microsomes

The protein content of liver and testis microsomes prepared from control and xenobiotic-treated animals are shown in Table 4.2. Due to the fact that testis samples were pooled, it was not possible to assess the statistical significance of any differences in microsomal protein yields between the treatment groups.

Table 4.2: Effect of xenobiotic pretreatment on the protein yields of liver and testis microsomes

Treatm ent Liver protein yield Testicular protein yield group (mg/g liver) (mg/g testis) Control 21.31 ±0.67 3.99

P-NF 21.39 ±0.99 4.05

PB 21.45 ± 1.19 3.96

PCN 2 2 . 2 2 ± 1 . 0 1 3.83

Ciprofibrate 23.99 ±0.92 *** 4.04

Results are expressed as mean ± SD from individual liver microsomal preparations (n=5 animals per group). Values from each xenobiotic-treated group were compared to the control group using a Student’s t-test where asterisks (*) represent; * * * P<0.001. Protein yields are shown for pooled testis microsomes from each group of animals.

123 4.3.4 Total CYP content of liver microsomes

The specific CYP content of liver microsomes from control and xenobiotic-treated animals are shown in Table 4.3. Due to low levels of expression in the testis it was not possible to measure the total CYP content of testis microsomes firom this study.

Table 4.3: Effect of xenobiotic pretreatment on the total cytochrome P450 content of liver microsomes

Treatment group Microsomal CYP content (nmoles/mg microsomal protein)

Control 0.47 ± 0.06

P-NF 0.72 ±0.18*

PB 1.02 ±0.15 ***

PCN 0.90 ±0.08***

Ciprofibrate 0.56 ± 0.06

Results are expressed as mean ± SD from individual microsomal preparations (n=5 animals per group). Values for each xenobiotic-treated group were compared to the control group using a Student’s t-test where asterisks (*) represent: * P<0.05, *** P<0.001.

4.3.5 Microsomal testosterone metabolism

Testosterone metabolism by liver and pooled testis microsomes firom control and xenobiotic-treated rats was determined using the optimised HPLC-based testosterone hydroxylase described in Chapter 3.

4.3.5.1 Testosterone metabolism by liver microsomes

Control profile

A representative HPLC profile of the metabolites formed following incubation of testosterone with liver microsomes from a control animal is shown in Figure 4.5. Control liver microsomes metabolised testosterone to form up to fourteen peaks, nine of which were identified by co-chromatography with authentic standards (namely

124 androstenedione, 6 a-, 6 P-, 7a-, 16a-, 16p-, 2a-, 2P- and 15a-0HT). Five unidentified peaks were present in all control samples, three of which eluted after testosterone. The identity of these peaks is currently unknown and they will be referred to using the letters shown in Figure 4.5 throughout this thesis.

The major oxidation products identified in control incubations were androstenedione,

16a- and 2a-0HT. Smaller amounts of 6 p-, 7a-, 6 a- and 2P-0HT were also formed.

Traces of 16p- and 15a-0HT were detected in a small number of control incubation extracts.

Figure 4.5: Representative chromatographic separation of testosterone metabolites formed following incubation of '"^C-testosterone with liver microsomes from a control animal. Testes. B cprn

16a 2a X

And

2 p 6p

7a 6a

1 0 2 0 30 40 Time (mins)

Liver microsomes from a control animal were incubated with C-testosterone as described in 2.2.8. 80gl of the reconstituted incubation extract was injected on to the HPLC column and analytes were quantified by radiochemical detection. Metabolites were identified by co-chromatography with authentic standards (detected by optical absorbance at 240nm, not shown), for example 6a corresponds to 6a-0HT, Testos., testosterone; And., androstenedione. Unidentified peaks are indicated by capital letters.

125 Metabolism in blank samples

A blank incubation containing all of the reaction components except for NADPH was prepared for each microsomal sample to verify that this cofactor was required for metabolism. Measurable amounts of 6a-0HT, 6p-0HT, peak Y and androstenedione were detected in blank incubation extracts (Figure 4.6). 6a-0HT and peak Y were formed in equal amounts in incubations conducted in the absence and presence of NADPH, as shown in Figure 4.7. In contrast, the rates of formation of 6p-0HT and androstenedione were higher in incubations containing NADPH compared to blanks.

Figure 4.6: Representative chromatographic separation of testosterone metabolites detected following incubation of ^"^C-testosterone with liver microsomes in the absence of NADPH Testos.

cpm

. And. 6p Y i

S3. 1 0 20 30 40 50 Time (mins)

Liver microsomes from a control animal were incubated with ^'^C-testosterone in the absence of NADPH as described in 2.2.8. SOpl of the reconstituted incubation extract was injected on to the HPLC column and analytes were quantified by radiochemical detection. Metabolites were identified by co­ chromatography with authentic standards (detected by optical absorbance at 240nm, not shown), for example 6a corresponds to 6a-0HT, Testos., testosterone; And., androstenedione. Unidentified peaks are indicated by capital letters.

126 Figure 4.7: The effects of omission of NADPH from incubation mixtures on testosterone metabohsm by rat liver microsomes from a control animal.

□ (+) NADPH

O {-) NADPH

6 alpha-OHT A nd. Metabolite

Liver microsomes were incubated with testosterone in the presence or absence of NADPH as described in 2.2.8. The rate of formation of each metabolite in the blank incubation (no NADPH) is expressed as a percentage of that in the standard incubation (with NADPH) (displayed as 100%). Standard incubations were performed in triplicate and each bar represents the mean ± SEM. A single incubation was performed for the blank.

Figure 4.8: Comparison of the levels of metabohtes detected in blank incubations with standard or washed liver microsomes

□ S tan d ard microsorres

W a sh e d m icrosom es

e a lp h a O H T

Liver microsomes from a control animal were prepared using the procedure described in 2.2.2, with (washed microsomes) or without (standard microsomes) the inclusion of an additional “washing” step. Microsomes were incubated with 'Y-testosterone in absence of NADPH. Each bar represents the average value from duplicate incubations.

127 Initially experiments were performed to investigate the possibility that these peaks were detected in blank samples due to the presence of residual endogenous NADPH in the microsomal sample. “Washed” liver microsomes were prepared by including an extra resuspension and centrifugation step that would presumably remove any residual cofactor from the sample. The same metabolite peaks were present in blank incubations with washed microsomes, confirming that these peaks were not present due to residual cofactor (Figure 4.8).

Subsequently, it was observed that following injection of ^"^C-testosterone on to the HPLC colunrn at similar concentrations to those routinely analysed in incubation extracts, peaks other than testosterone were present on the chromatogram (data not shown). Peaks corresponding to 6P-0HT and androstenedione were identified by co­ chromatography with authentic standards. In addition, a peak corresponding to peak Y was identified based on the position of this peak on the chromatogram relative to the authentic standards. Traces of 6p-0HT and androstenedione were detected in the ^"^C-testosterone substrate at levels that did not account for the entire peaks detected in the blank samples. In contrast, the amount of peak Y detected in the ^"^C-testosterone substrate fully accounted for the peak detected in blank samples. These findings indicate that peak Y was not formed metabolically during the incubation period but was present as a contaminant in the ^"^C-testosterone. This peak probably represents an impurity or breakdown product of testosterone. At present the pathways responsible for the formation of 6a-0HT, 6P-0HT and androstenedione in the absence of exogenous NADPH are unknown.

For the current project, the primary interest was CYP-dependent pathways of testosterone metabolism, which require the presence of NADPH. The presence of peaks in blank samples was therefore corrected for by subtracting the amount of each metabolite detected in the blank sample from the amount detected in the corresponding standard incubation (containing NADPH) prior to the calculation of enzyme activity. Following the subtraction of blanks, no 6a-0HT or peak Y were detected.

128 Effects o f inducers on hepatic testosterone metabolism

The effects of treating rats with the four model CYP inducers on microsomal testosterone metabolism are shown in Figures 4.9 to 4.12. In addition to the pathways shown in Figures 4.10 and 4.11, PB and PCN induced the formation of unidentified metabolites that were not present in incubation extracts firom control microsomes, as shown in Figure 4.13. PB and PCN induced the formation of the same metabolites. These metabolites were present at low levels and were poorly resolved by the current HPLC system and therefore were not quantified.

Figure 4.13: Chromatographic separation of testosterone metabolites formed following incubation of ^"^C-testosterone with liver microsomes firom a PB-treated animal.

cpm

16a

16P 2P

6a 15P o. 10 2 0 30 40 50 Time (mins)

Liver microsomes from a PB-treated animal were incubated with testosterone as described in 2.2.8. 80gl of the reconstituted incubation extract was injected on to the HPLC column and analytes were quantified by radiochemical detection. Metabolites were identified by co-chromatography with authentic standards (detected by optical absorbance at 240nm, not shown), for example 6a corresponds to 6a-0HT, Testos., testosterone; And., androstenedione. Unidentified peaks that were not present in control incubations are shown in circles.

129 CO ■Q II 0 c T3 00 0

1

1 1 O I CO S e S s s c«o «Ê J § ; li X3 -I | -gO) e B

1 1 :2 o > o 0 I : % 1 £ 1I o u +1 8 S .. * C/5 I V-) 2 o g 2 * o\

* h- t ' II *-H CQ 1—l ...... L L .J s /e^oi f * î 1 h - 3:-a o o T3 1 I— I

e(U c II en0 o s ^ .'2 § üS ■Il > a cilII * X3 ■S I o n I I1 a 1 3 o 3> 0 a o 0 1 1 %I*iS * î I

p H h

go (u ja} O Jd iu/uiiu/sa|oiud) A^jAipe auuAzug 6 lî CO TJ II 0) c « O o c Z in ii o Ü II I Ü 0. 3 g Q i li 1 ilI- CO ®%0, /®;o/ H u) PC .S O g I ^O/ejo/ é i | VO 11 a!) § (ujajoJd Buj/u|iu/se|oiud) Ajiaijob oiuA zug 2 .52 0 00 (g *o S ii 1 g ii IIO CO IÎ cn ; Ii X3 il 2 13 C/3 Ii 1 |l a§ 0 11 "o 1 (U “I+1 3 c O i IH oV) J il§ II go È [u|s)ojd 6ui/ujui/sa|ouid) A)|A|)oe auiAzug II s i CO LO is II il Q. I-§ §1 0Û c/3 .s I I I '«to, if 2 s 1 0 0 c/3 s ? O I t I O o SI

I uiaiojd 6ui/u!Lu/S9|Oiud) Ajiaiidb aiuAzug

I 11O CO h ë Ü Ii

W" 00 i CO Ü CO e 00o ilO (U X § "O c ; (U he < > a X) a-II CM X) 0I 1 <0 o 'S a CM "3 c/3 R> 2 12 I CO o I o

Control profile

A representative HPLC profile of metabolites formed following incubation^ of testosterone with pooled testis microsomes from the control group is shown in Figure 4.14. Incubation extracts from testis microsomes contained the major metabolites androstenedione and 6P-0HT, along with traces 6a-0HT, 7a-0HT and peak Y.

Figure 4.14 : Chromatographic separation of testosterone metabolites formed following incubation of ^"^C-testosterone with pooled testis microsomes jfrom control animals.

Testos.

cpni

And.

6a

40 50 18 Time (mins)

Pooled testes microsomes were incubated with testosterone as described in 2.2.8. 80pl of the reconstituted incubation extract was injected on to the HPLC column and analytes were quantified by radiochemical detection. Metabolites were identified by co-chromatography with authentic standards (detected by optical absorbance at 240nm, not shown), for example 6a corresponds to 6a-0HT, Testos., testosterone; And., androstenedione. Unidentified peaks are indicated by capital letters.

134 Metabolism in blank samples

As previously described for incubations with liver microsomes, measurable amounts of 6a-0HT, 6p-0HT, peak Y and androstenedione were present in blank incubations with testis microsomes. This was corrected for by subtracting the amount of each metabolite detected in the blank sample from the amount detected in the corresponding standard incubation (containing NADPH) prior to the calculation of enzyme activity. Following subtraction of blanks, no 6a-0HT or peak Y were detected. Due to the low levels of metabolism in the testis, only traces of testosterone

6 p-hydroxylase activity remained following blank subtraction. In contrast, androstenedione was the major metabolite produced by testis microsomes and adequate levels of this metabolite remained following blank subtraction.

Effects o f inducers on testicular testosterone metabolism

The effects of treating rats with the four model CYP inducers on testosterone metabolism catalysed by pooled testis microsomes are shown in Figure 4.15.

135 1

■g I I 1

1cl> ■g %

\D en

s s O 0 Cl, X) "Sc/3 1 a 0 1 B (U i

H •O (U|3}OJd 6lU/U|lU/S9|OUld) Â)|Aj}38 3LUÂZU3

SI 4.3.6 Western blotting

Western blotting experiments were performed with liver microsomes from control and xenobiotic-treated animals as described in 2.2.11. Each CYP protein was localised using the appropriate anti-CYP primary antibody and was identified based on the literature molecular weight and co-migration with the positive control sample.

Figure 4.16 shows a representative Western blot from this study and the results of all Western blotting experiments are summarised in Table 4.4. CYP1A2 and 3A proteins were detected in liver microsomes from control animals, whereas CYP2B and 4A proteins were not detected. The effects of ciprofrbrate on CYP3A protein levels were inconclusive because the lanes on the gel appeared to have merged.

Figure 4.16: Western blot of rat liver microsomes developed with anti-CYP 1A2 primary antibody

Lane 1

58100

39800

5)Lig microsomal protein was loaded in to each lane. Lane 1, Molecular weight markers Lane 2, Positive control liver microsomes from isosafrole-treated rats Lanes 3, Liver microsomes from a control animal Lanes 4-8, Liver microsomes from five individual P-NF-treated animals

137 Table 4.4: Effect of xenobiotic pretreatment on CYP proteins levels in rat liver microsomes

Treatm ent CYP1A2 CYP2B CYP3A CYP4A1 group

p-NF + - -—

PB - + + -

PCN - + + —

Ciprofrbrate - - ? +

+, induction of the isoform was observed; - , no induction was observed; ?, results were inconclusive. Western blotting for CYP4A1 was performed by Alex Bell at AstraZeneca, Loughborough.

138 4.4 Discussion

4.4.1 Effects on hepatic CYPs and testosterone metabolism

4.4.1.1 Testosterone metabolism by control liver microsomes

The optimised HPLC-based testosterone hydroxylase assay described in Chapter 3 produced good resolution of metabolites present in incubation extracts, the majority being identified by co-chromatography with authentic standards. The profile of metabolites detected in incubation extracts firom control liver microsomes is consistent with previously published data for male Sprague Dawley rats (Purdon & Lehman- McKeeman, 1997; Parkinson et al., 1992). The absolute rates of testosterone hydroxylation catalysed by liver microsomes in the current study tended to be lower than values reported in the published literature, particularly with respect to testosterone 6 P-hydroxylation. Such differences might be due to differences in animal age and environmental factors (e.g. housing conditions, diet).

In the current study, androstenedione and up to eight hydroxytestosterone metabolites

(6 a-, 6 p-, 15a-, 7a-, 16a-, 16p-, 2a-, and 2p-0HT) were detected in control incubation extracts. In addition to these metabolites, some groups have detected traces of ISp-OHT and 16-ketosterone in control incubations (Purdon & Lehman- McKeeman; Parkinson et al., 1992). The generally lower testosterone hydroxylase activity of liver microsomes fi*om the current study may have resulted in levels of these metabolites that were below the limits of detection of our assay.

Four unidentified peaks were detected in control incubation extracts that were not present in blank incubations (Figure 4.5). Firstly, an unidentified peak was detected that eluted between androstenedione and testosterone (peak X). Based on the published literature this peak might correspond to 17p-hydroxy-4,6-androstadiene-3- one (A^-T), which has previously been shown to elute from a C l 8 HPLC column between androstenedione and testosterone (Levin et al., 1987). Nagata et al. (1986) characterised this metabolite, which is formed by dehydrogenation of testosterone to form a double bond between the C - 6 and C-7 positions. Nagata et al. (1986) provided evidence that conversion of testosterone to A^-T by rat liver microsomes is mediated by CYP-dependent pathways, possibly involving a CYP3A form. Consistent with our

139 current findings for peak X, the formation of A^-T was previously shown to be enhanced in liver microsomes from PB- and PCN-treated rats (Nagata et ah, 1986). In the current study, peak X showed negligible absorbance at 240nm, which is consistent with the fact that A^-T has a maximum absorbance at 284nm (Nagata et ah, 1986).

Secondly, three unidentified peaks (peaks A, B and C) were present that eluted from the HPLC column after testosterone. The formation of these metabolites was negligible in the absence of exogenous NADPH. The elution pattern of these metabolites along with the fact that they were not detected on the UV-absorbance trace (monitored at 240nm) suggests that these peaks might correspond to reduced derivatives of testosterone. Reduced, metabolites known to be formed by rat liver microsomes include dihydrotestosterone (DHT) (catalysed by 5a-R), 3a-diol

(catalysed by 3a-hydroxysteroid dehydrogenase (HSD)) and 3p-diol (catalysed by 3P- HSD) (Edwards et al, 1992). However, 4-MA was added to all incubations at a concentration that would be expected to completely inhibit the activity of the 5a-R enzyme therefore DHT would not be expected to be present in these samples (Sonderfan & Parkinson, 1988). The identity of these three metabolites is currently unknown. One approach that could be used to confirm the identity of these peaks would be to collect each peak using a fraction collector and then analyse each fraction by mass spectrometry or NMR spectroscopy.

4.4.1.2 Effects of model inducers on hepatic testosterone metabolism

/3-Naphthoflavone (P-NF)

Treatment of rats with p-NF had no significant effect on liver weight or microsomal protein content but produced a significant increase in the specific microsomal CYP content (1.5 fold), consistent with the published literature (Amacher & Schomaker, 1998; Guengerich et al., 1982). Western blotting confirmed that P-NF treatment- produced a marked induction of CYP1A2 protein with no effects on CYP2B, 3A or 4A1 protein levels. These findings are consistent with the published literature (Guengerich et al., 1982; Waxman et al., 1985; Wortelboer et al., 1991). Although not investigated in the current study, treatment of rats with p-NF has been reported to suppress hepatic CYP2C11 (-63% of control) and 2C6 (- 6 6 % of control) protein

140 levels and to have no effect or cause a small induction of CYP2A1 protein (Guengerich et al, 1982; Waxman et al., 1985).

p-NF produced a marked reduction in the formation of androstenedione, 2a- and 16a- OHT, indicating suppression of CYP2C11 (and possibly CYP2C6) protein levels (Waxman, 1984). p-NF also caused a significant reduction in the formation of two unidentified metabolites, peaks A and B. No statistically significant effects were observed on the remaining pathways of testosterone metabolism, consistent with the lack of effect of this compound on CYP2B or 3 A protein levels. The overall rate of formation of hydroxylated testosterone metabolites was significantly reduced in liver microsomes from p-NF-treated animals (78% of control).

The effects of P-NF on microsomal testosterone metabolism observed in the current study are generally in good agreement with the published literature, except that some

groups report increases in 7a-hydroxylase activity and decreases in 6 p-hydroxylase activity (Lee & Park, 1989; Tredger et al., 1984; Wortelboer et al.,-1991). Such differences between the current data and the published literature is probably related to differences in strain, age, dosing regimen (e.g. dose, route, duration) and environmental factors (e.g. housing conditions, diet) which can influence the effects of inducers on testosterone metabolism.

Phénobarbital (PB)

Treatment of rats with PB was associated with significant increases in liver weight and specific microsomal CYP content (2.2 fold), consistent with the published literature (Guengerich et al., 1982; Amacher & Schomaker, 1998; Tredger et al., 1984). This compound had no significant effect on the hepatic microsomal protein yield. Western blotting confirmed that PB treatment caused a marked induction of CYP2B and a smaller increase in CYP3A proteins, consistent with the published literature (Guengerich et al., 1982; Waxman et al., 1985; Parkinson et al., 1992). Although not investigated in the current study, PB has been reported to induce hepatic CYP2A1 (~2 fold) and CYP2C6 (-2.5 fold) and decrease CYP2C11 (-50% of control) protein levels in male rats (Guengerich et al., 1982; Waxman et al., 1985; Parkinson et al., 1992).

141 Consistent with the results of Western blotting experiments, PB treatment produced a marked induction of microsomal testosterone 16p-hydroxylase activity (CYP2B) along with significant increases in 6 p-, ISp- and 2p-hydroxylase activity (CYP3A).

The significant increase in testosterone 7a-hydroxylase activity and decrease in 2a- hydroxylase activity is consistent with the ability of this compound to induce CYP2A1 and suppress CYP2C11 respectively. A significant increase in the formation of the unidentified metabolite, peak X, was also observed. There was a trend towards an increase in the total rate of hydroxylated testosterone metabolite formation in the PB- treated group (1.3 fold), but this did not reach statistical significance.

PB-treatment also induced the formation of metabolites that were not previously detected in control incubations (Figure 4.13). These metabolite peaks were present at relatively low levels and were poorly resolved using the current HPLC method, therefore they were not quantified. Based on their position on the chromatogram and comparison with the published literature, the peak(s) eluting between 16p- and 2 a-

OHT might correspond to la-/p-OHT and/or 18-OHT (Purdon & Lehman- McKeeman, 1997). These metabolites are often poorly resolved by HPLC resulting in a composite peak and are reportedly induced following PB treatment (Purdon & Lehman-McKeeman, 1997). The identity of the second peak that elutes between 2^- OHT and androstenedione is currently unknown.

The effects of PB on microsomal testosterone metabolism observed in the current study are generally in good agreement with the published literature, except that some other groups have reported increases in 16a-hydroxylase activity and androstenedione formation (Parkinson et al., 1992; Lee & Park, 1989). In the current study, there was a trend towards an increase in androstenedione formation but this did not reach statistical significance.

142 Pregnenolone-16a-carbonitrile {PCN)

PCN-treated rats showed a trend towards increased liver weights but this did not reach statistical significance. PCN had no effect on the hepatic microsomal protein yield but produced a significant increase in the specific microsomal CYP content (1.9 fold), consistent with the published literature (Amacher & Schomaker, 1998; Guengerich et al., 1982; Hanoika et al., 1995). Western blotting confirmed that PCN treatment caused marked induction of CYP3A and a smaller increase in CYP2B protein levels, consistent with the published literature (Guengerich et al., 1982; Waxman et al., 1985). Although not investigated in the current study, PCN has been reported to suppress hepatic CYP2A1 (-54% of control), CYP2C11 (-63% of control) and

CYP2C6 (-57% of control) protein levels in male rats (Waxman et al., 1985).

Consistent with the results of Western blotting experiments, PCN treatment produced

a marked induction of testosterone 6 p-, 15^- and 2p-hydroxylase activities (CYP3A)

and a smaller induction of 16p-hydroxylase activity (CYP2B) although the latter did not reach statistical significance. Significant induction of 15a-hydroxylation was also observed, suggesting induction of CYP2A2. There was a trend towards reduced formation of androstenedione, 2a- and 16a-0HT, suggesting suppression of CYP2C11 (and possibly CYP2C6). PCN treatment also induced the formation of the unidentified peak X and reduced the fomiation of peak A. There was a statistically significant increase in the total rate of hydroxylated testosterone metabolite formation in the PCN-treated group (1.6 fold). PCN treatment also induced the formation of metabolites that were not previously detected in control incubations as discussed above for PB. The effects of PCN treatment on microsomal testosterone metabolism observed in the current study are consistent with the published literature (Duffy et al., 1995; Sonderfan et al., 1987; Tredger et al., 1984).

Ciprofîbrate

Treatment of rats with ciprofibrate produced a marked increase in liver weight and hepatic microsomal protein yield, consistent with the published literature (Graham et al., 1994; Makowska et al., 1990). In the current study this compound had no significant effect on the specific microsomal CYP content. Makowska et al. (1990)

143 reported that treatment of male rats with ciprofibrate for 14 days was associated with an increase in the specific microsomal CYP content of liver microsomes at a low dose

( 2 mg/kg/day) with no significant effect observed at the higher dose ( 2 0 mg/kg/day). •

Western blotting confirmed that ciprofibrate treatment caused marked induction of microsomal CYP4A1 levels, which was not found with any of the other model inducers and is consistent with the published literature (Makowska et ah, 1990; Zangar et al., 1996). No effect was observed on the expression of CYP1A2 or 2B proteins. The effect of ciprofibrate on CYP3A was inconclusive due to technical difficulties with this Western blot (the lanes on the gel had merged). Several studies have reported reductions in CYPIA mRNA and associated catalytic activities (EROD and methoxyresorufin 0-dealkylase (MROD)) following treatment of Sprague Dawley rats with ciprofibrate (Makowska et al., 1990; Canivenc-Lavier et al., 1996; Gallagher et al., 1995). Similarly, reductions in CYP3A2 (-80% of control) and CYP2C11 (-20% of control) mRNA levels have been reported following short-term (3 days) dietary exposure of rats to this compound (Gallagher et al., 1995). Conflicting data has been presented concerning the effects of ciprofibrate on CYP2B expression in male Sprague Dawley rats. Makowska et al. (1990) reported that treatment of rats with ciprofibrate (2 or 20 mg/kg/day) for 14 days was associated with a significant reduction in benzphetamine deaUcylation, a marker for CYP2B1/2. In contrast, a subsequent study reported significant increases in benzoxylresorufin 0-dealkylation (BROD) with no effect on CYP2B mRNA levels following 3 days dietary exposure to ciprofibrate (0.025% w/w) (Gallagher et al., 1995). Zangar et al. (1996) reported a significant increase in CYP2B mRNA and protein levels following treatment with ciprofibrate for 5 days (20 mg/kg/day i.p.). These differences between studies are probably related to differences in animal age, dosing regimen (e.g. dose, route, duration) and environmental factors (e.g. housing conditions, diet) along, with the different end­ points measured (mRNA, protein, enzyme activity).

In contrast to the other three compounds, the effects of ciprofibrate on microsomal testosterone metabolism have not been well characterised. The most prominent effect of ciprofibrate observed in the current study was induction of CYP4A1, which is mainly involved in fatty acid metabolism. The major effect of ciprofibrate on

144 microsomal testosterone metabolism was the marked reduction in the formation of androstenedione, 16a- and 2a-0HT, suggesting repression of CYP2C forms. Ciprofibrate also produced a significant increase in testosterone 6|3-hydroxylase activity (CYP3A) and a decrease in the formation of the unidentified metabolite, peak A. Ciprofibrate treatment was associated with a significant reduction in the total rate of hydroxytestosterone metabolite formation ( 6 6 % of control).

4.4.2 Effects on testicular CYPs and testosterone metabolism

4.4.2.1 Organ weights

Treatment of rats with the four model compounds had no significant effect on testis weights. Fahim et al. (1970) and Levin et al. (1974) also reported no significant changes in testis weights following treatment of adult male rats with PB for 60 days (50 mg/kg/day i.p.) or 4 weeks (0.05% or 0.1 % in the diet) respectively.

4.4.2.2 Testosterone metabolism

Measurement o f testosterone metabolism by testis microsomes

Due to the low levels of CYP expression in the testis, pooled testis microsomes were prepared fi*om each treatment group to enrich the preparation for CYP enzymes. The major disadvantage of this approach is that subtle changes might be lost due to interanimal variability in the levels of CYP expression and responses to drug treatment.

Testis microsomes showed low levels of testosterone hydroxylase activity, which presented a problem for the accurate quantification of individual testosterone metabolites. Two factors determined the limit of detection of the current testosterone hydroxylase assay. The first factor involves discrimination of radioactive counts associated with a metabolite from the background counts (by radiochemical detection). The second factor involves discrimination of metabolites that were formed by NADPH-dependent pathways from the peaks that were present in blank samples.

145 The following metabolites were detected in incubation extracts from testis microsomes: 6a-0HT, 6P-0HT, 7a-0HT, peak Y and androstenedione. Androstenedione formation was the major pathway catalysed by testis microsomes and this metabolite was formed at levels that were at least three times higher than the amount detected in the blank incubation, resulting in reliable quantification of this metabolite. In contrast, 6P-0HT was present in incubation extracts at levels that provided a good signal to noise ratio with respect to radiochemical detection, but following subtraction of the blank only traces of this metabolite remained. This would mean that the quantification of 6p-0HT would be less accurate than that of androstenedione. Consistent with data from incubations with liver microsomes, following blank subtraction no 6a-0HT or peak Y were present in incubation extracts from testis microsomes. 7a-0HT was not detected in blank samples but was present in incubation extracts at levels that were close to the limits of detection with respect to radiochemical detection, resulting in less accurate quantification of this metabolite. These factors must be considered when interpreting the results from incubations with testis microsomes. The sensitivity of the current assay might be improved by measuring testosterone metabolism using a testicular fraction enriched with Leydig cells (e.g. purified using a percoll gradient). Most of the testicular CYP content is localised the Leydig cell population, which comprises less than 3% of the total testis volume (Mori & Christensen, 1980).

The profile of metabolites detected in incubation extracts from control testis microsomes was consistent with previously published data for male Sprague Dawley rats (Sonderfan et al., 1989). In addition to the metabolites detected in the current study, Sonderfan et al. (1989) reported low levels of testosterone 16a-hydroxylase activity in testis microsomes. The testicular enzymes responsible for testosterone 7a- hydroxylation and androstenedione formation are CYP2A1 and 17p-HSD respectively (Sonderfan et al., 1989). In contrast, the testicular enzymes that catalyse the formation of 6 p- and 16a-0HT are unknown (Sonderfan et al., 1989).

146 Effects of model inducers on testicular testosterone metabolism

Following treatment of rats with the four model inducers no significant effect was observed on testicular testosterone 6p-hydroxylase activity, although reliable conclusions cannot be drawn due to the sensitivity problems discussed above. The apparent ability of PB and PCN to markedly induce hepatic but not testicular 6P- hydroxylase activity suggests that this pathway may exhibit tissue-specific sensitivity to the effects of chemical inducers. Indeed, much of the published literature indicates that testicular CYP enzymes are less sensitive to exogenous inducers than their hepatic counterparts (Goldstein & Linko, 1984; Omiecinski, 1986). Hydroxylation of testosterone at the 6p-position is predominantly catalysed by CYP3A forms in the liver but it is not known whether enzymes belonging to this subfamily also catalyse this pathway in the testis (Waxman et al., 1985; Sonderfan et al., 1989). When evaluating the effects of inducers on the testis, consideration must be given to the cellular localisation of the enzyme of interest because the blood-testis barrier regulates the entry of substances in to the tubular compartment. However, testicular CYPs are predominantly localised to the Leydig cells which reside in the interstitial compartment, which is in direct contact with substances entering the testis in the blood or lymph (Mukhtar et al., 1978).

Treatment of rats with p-NF or PB had no statistically significant effects on testicular testosterone 7a-hydroxylase activity. In contrast, 7a-hydroxylase activity showed a small but statistically significant increase in ciprofibrate-treated animals and decrease in the PCN-treated group, although firm conclusions cannot be drawn due to the sensitivity problems discussed above. These compounds had no significant effects on the activity of this pathway catalysed by liver microsomes firom these animals. Hydroxylation of testosterone at the 7a-position is a developmentally-regulated pathway, which is predominantly catalysed by CYP2A1 in both the liver and testis (Levin et al., 1987; Wood et al., 1983; Sonderfan et al., 1989). Evidence has been presented supporting a potential role for 7a-hydroxylated androgens in the regulation of testosterone biosynthesis therefore xenobiotic-induced alterations in the local formation of these metabolites might have important effects on the testis (Inano et al., 1973; Seng et al., 1991). In addition, there is evidence to support a potential role for

147 high testicular CYP2A1 expression in the aetiology of spontaneous Leydig cell hyperplasia in rodents (see 1.5.3) (Seng et al., 1996).

Treatment of rats with all four compounds was associated with significant increases in androstenedione formation catalysed by testis microsomes. The conversion of testosterone to androstenedione by liver microsomes is catalysed by CYP-dependent (CYP2B and CYP2C forms) and CYP-independent (17p-HSD) pathways (Wood et al., 1983; Sonderfan et al., 1987; Martel et al., 1992). 17P-HSD has also been reported to catalyse this pathway in the testis (Sonderfan et al., 1989). Several distinct 17P-HSD enzymes have been identified, which differ in their substrate specificity, cofactor requirements, preference for oxidative or reductive reactions and tissue distribution (reviewed by Peltoketo et al., 1999). The forms that catalyse the oxidation of testosterone to form androstenedione preferentially use NAD"^ as the cofactor. For the current study, exogenous NAD(P)^ was not added to microsomal incubations therefore 17P-HSD would not be expected to make a significant contribution to the formation of androstenedione. The majority of the androstendione formation was dependent upon the addition of exogenous NADPH, suggesting that CYP enzymes might catalyse this pathway in testis microsomes. Androgens show higher activity in the 17p-hydroxy form therefore the conversion of testosterone to androstenedione is predominantly considered to represent a catabolic pathway. Consequently, the ability of the model inducers to enhance the activity of this pathway might reduce the local concentration of active androgen in the testis.

4.4.3 Effects on plasma hormone levels

No statistically significant changes in plasma hormone levels were observed following treatment of rats with the four, model inducers. This may reflect the true situation but subtle changes might not have been detected due to a number of factors. Firstly, the timing of blood samples relative to the last dose drug would be expected to affect the magnitude of changes in serum hormone levels observed. In the current study, blood samples were collected approximately 24 hours after the final dose, by which time plasma hormone levels might have been returning towards the normal range depending on the half-life of the drug.

148 Secondly, significant “time of day” variations in circulating hormone levels occur in rats. Circulating testosterone levels show a diurnal rhythm with the most pronounced peak in hormone levels occurring during the dark periods (-23 3Oh) with a smaller peak during the light period (-1300h) (Keating & Tcholakian, 1979). The least variations in testosterone concentrations are reportedly observed during the morning (Keating & Tcholakian, 1979). Circulating FSH levels show a distinct circadian rhythm with relatively stable concentrations between 0130 and 1330h and a pronounced peak between 1630 and 2130h (Kalra & Kalra, 1977). Clearly, the timing of blood samples would influence the levels of testosterone and FSH measured, therefore for the current study (and all other studies conducted as part of this project) blood samples were collected between 1000 and 1200h to minimise these effects. In addition, circulating LH levels measured at a single timepoint show a high degree of variability due to the pulsatile pattern of LH secretion from the pituitary gland (Ellis & Desjardins, 1982).

Finally, circulating hormone levels show marked interanimal variability in rodents, which often confounds the detection of statistically significant changes following xenobiotic exposure. Wilson et al. (1999) postulated that due to this problem, subtle disturbances of steroid hormone homeostasis might remain undetected until more profound irreversible effects are manifested. A high degree of intragroup variability in plasma hormone levels was apparent in the current study, particularly with respect to plasma testosterone concentrations (Figure 4.1). This variability along with the small sample sizes used, might have obscured the detection of statistically significant changes in plasma hormone levels.

There is a lack of information in the published literature concerning the effects of the four inducers on steroid homeostasis in rats. Consistent with the current findings, Yeowell et al. (1987) reported no significant changes in serum testosterone levels in adult male Sprague Dawley rats following treatment with PB (80 mg/kg/day i.p.) for one, three or five consecutive days. This study was similar to the current study in that the blood samples were collected 24 hours after the final dose of PB.

149 4.4.4 Conclusions

The effects of dosing rats with the four model inducers on hepatic CYP-dependent testosterone metabolism have been characterised and the current findings are generally in good agreement with the published literature. Each model compound exerted differential effects on the pathways of testosterone oxidation catalysed by liver microsomes, inducing the activity of certain pathways whilst suppressing or having no effect on other pathways. Interestingly, all four compounds caused suppression of testosterone 2a-hydroxylase activity, albeit to different extents. This pathway is catalysed by CYP2C11, which is the major constitutive CYP (- 40% total hepatic CYP) found in male rat liver (Guengerich et al., 1982). Yeowell et al. (1987) reported that the reduction in hepatic CYP2C11 levels observed following treatment of rats with PB or 3-methylcholanthrene was due to a decrease in protein synthesis. The physiological functions of constitutive CYP forms have not been fully elucidated. Waxman (1984) postulated that CYP2C11 probably makes a substantial contribution to the overall rate of testosterone oxidation in rat liver but might also be responsible for the formation of specific hydroxylated metabolites with important physiological functions. Changes in CYP2C11 expression following xenobiotic exposure might therefore have important physiological implications for the animal.

Each model inducer altered the total rate of hydroxylated testosterone metabolite formation catalysed by liver microsomes, but no significant impact was observed on circulating testosterone levels. This might be due to compensatory changes in the activity of other pathways of testosterone metabolism (e.g. reductive pathways, ST and UDP-GT), resulting in no effect of these compounds on the overall rate of testosterone clearance in vivo. Consistent with this hypothesis, treatment of humans with Y-phenylbarbital produces a significant increase (> 60%) in the urinary excretion of polar testosterone metabolites but has no significant effect on the plasma clearance of testosterone (Bammel et al., 1992). Alternatively, initial effects of these compounds on circulating testosterone levels may be rapidly compensated for through changes in the rate of testosterone biosynthesis and/or secretion from the testis. Such homeostatic regulation might be sufficient to maintain normal circulating hormone levels despite the xenobiotic-induced changes in the rate of metabolic clearance of testosterone. Indeed, Stripp et al. (1974) reported that treatment of rats with PB had

150 no effect on seminal vesicle weight suggesting that the ability of this compound to increase hepatic testosterone hydroxylation could be counterbalanced by changes in androgen biosynthesis and/or other pathways of testosterone metabolism.

The effects of the model inducers on the pathways of testicular CYP-dependent testosterone metabolism have been studied, although the low levels of testosterone hydroxylase activity present in this tissue confounded the accurate quantification of testosterone metabolites. The magnitude of the changes in testicular testosterone hydroxylase activity observed in the current study would not be expected to have a significant impact on the overall rate of androgen clearance from the body, but might have important local implications for androgen homeostasis within the testis.

151 Chapter 5

Effect of lansoprazole on cytochromes P450, testosterone metabolism and plasma hormone levels 5.1 Introduction

Lansoprazole was selected as the model compound for studies designed to investigate the existence of a liver-testis axis as it induces hepatic CYPs involved in testosterone metabolism and chronic exposure produces LCTs in rats (Masubuchi et ah, 1997a; Atkinson et al., 1990). This chapter describes initial work to characterise the effects of dosing male rats with lansoprazole on CYP expression, microsomal testosterone metabolism and the endocrine control of the testis. The main objective of this study was to investigate the effects of lansoprazole on hepatic CYP-dependent testosterone metabolism and plasma hormone levels. In addition, experiments were performed to determine whether this compound also influenced testicular CYP-dependent testosterone metabolism, which might play a role in LCT induction.

5.2 Study design

Male Sprague Dawley rats were dosed once daily with lansoprazole (150mg/kg/day) or vehicle (0.5% (w / v) CMC) by oral gavage for 14 days. This dose was selected as it has previously been shown to reduce serum testosterone levels and increase circulating LH levels in male rats (Fort et al., 1995). On day 15, blood samples were collected and liver and testis microsomes were prepared. More detailed experimental details concerning animal treatment, sample collection and the analytical methods used for this study can be found in chapter 2 .

.153 5.3 Results

5.3.1 Body, liver and testes weights

Final body weights and relative liver and testes weights for control and lansoprazole- treated animals are shown in Table 5.1.

Table 5.1 : Effect of lansoprazole treatment on relative liver and testes weights

Control Lansoprazole-treated

Final body weight (g) 353.6 ±17.5 328.9 ±16.0 **

Relative liver weight 4.14 ±0.24 4.53 ± 0.42 * (g/lOOg body weight)

Relative testes weight 0 . 8 6 ±0.06 0.94 ±0.07* (g/lOOg body weight)

Results are expressed as mean ± SD for control (n=l 1) and lansoprazole-treated (n=14) animals. Statistical analysis was performed using a Student’s t-test where asterisks (*) represent: * P<0.05, **P<0.01.

There were no statistically significant differences in initial body weights between experimental groups (data not shown) but final body weights of lansoprazole-treated animals were significantly lower than the control group (P<0.01). During the 14 day dosing period, body weight gain was significantly lower (P<0.01) in drug-treated animals compared to controls from day 4 of dosing onwards (data not shown). There were no significant differences in absolute organ weights (data not shown) between the experimental groups, but significant increases in relative liver and testes weights were observed in lansoprazole-treated animals compared to the control group (P<0.05).

5.3.2 Plasma hormone levels

Blood samples were collected (-18 to 20 hours after the final dose) and assayed for plasma testosterone, LH, FSH and prolactin levels by radioimmunoassay (Figures 5.1 and 5.2). Lansoprazole treatment was associated with a significant reduction in plasma testosterone levels compared to control group (P<0.05). No statistically significant changes in plasma FSH, LH or prolactin levels were observed.

154 Figure 5.1: Plasma testosterone levels in control and lansoprazole-treated rats

Cortrol(n=11) Lansopra2Dle4r60ted (rfe14)

Each bar represents the mean ± SEM from control (n=l 1) or lansoprazole-treated (n=14) animals. Statistical analysis was performed using a Student’s t-test where an asterisk (*) represents; * P<0.05.

Figure 5.2: Plasma prolactin, follicle-stimulating hormone and luteinising hormone levels in control and lansoprazole-treated rats

E3 C ontrol

Lansoprazole-treated

Hormone

Each bar represents the mean ± SEM from control (n=l 1) or lansoprazole-treated (n=14) animals.

155 5.3.3 Protein content of liver and testis microsomes

The protein content of liver and testis microsomes prepared from control and lansoprazole-treated animals are shown in Table 5.2.

Table 5.2: Effect of lansoprazole treatment on the protein yields of liver and testis microsomes

Treatment group Liver protein yield Testicular protein yield

(mg/g liver) (mg/g testis) Control 16.02 ±3.12 3.53

Lansoprazole 14.53 ±2.47 3.39

Results are expressed as mean ± SD from individual liver microsomal preparations from control (n=l 1) and lansoprazole-treated (n=l4) animals. Protein yields are shown for pooled testis microsomes from control and lansoprazole-treated groups. ,

There were no statistically significant differences in hepatic protein yields between experimental groups. Due to the fact that the testis samples were pooled, it was not possible to assess the statistical significance of any differences in microsomal protein yields between experimental groups.;

5.3.4 Total CYP conteiit of liver microsomes

The specific CYP contents of liver microsomes from control and lansoprazole-treated animals are shown in Table 5.3. Due to low levels of expression in the testis it was not possible to measure the total CYP content of testis microsomes from this study.

Table 5.3: Effect of lansoprazole treatment on the total cytochrome P450 content of liver microsomes

Treatment group Microsomal P450 content (nmol/mg microsomal protein)

Control 1.36 ±0.18 .

Lansoprazole 1.62 ± 0.18**

Results are expressed as mean ± SD from individual microsomal preparations from control (n=l 1) and lansoprazole-treated (n=13) groups. Statistical analysis was performed using a Student’s t-test where asterisks (*) represent: ** P<0.01.

156 5.3.5 Microsomal testosterone metabolism

Testosterone metabolism by liver and pooled testis microsomes from control and lansoprazole-treated rats was determined using the optimised HPLC-based testosterone hydroxylase assay described in chapter 3.

5.3.5.1 Testosterone metabolism by liver microsomes

Control profile

Control liver microsomes metabolised testosterone to form up to fourteen peaks, nine of which were identified by co-chromatography with authentic standards (namely androstenedione, 6 a-, 6(3-, 7a-, 16a-, 16p-, 2a-, 2p- and 15a-0HT). Five unidentified peaks (peak X, Y, A, B and C) were detected in all control samples, as previously discussed in Chapter 4. The major oxidation products identified in control incubations were androstenedione, 2a-, 16a- and 6p-0HT. Smaller amounts of 7a-,

2p- and 6a-0HT were also detected. 16P-0HT was detected in the majority of samples and traces of 15a-0HT were detected in a small number of control incubation extracts.

Peaks corresponding to androstenedione, peak Y, 6 a- and 6P-0HT were detected in blank incubations, consistent with data from the model inducers study (chapter 4). This was corrected for by subtracting the amount of each metabolite detected in the blank sample from the amount detected in the corresponding standard incubation (containing NADPH) prior to the calculation of enzyme activity. Following the subtraction of blanks, no 6a-0HT or peak Y were detected.

Effect o f lansoprazole treatment on hepatic testosterone metabolism

The effects of treating rats with lansoprazole on testosterone metabolism by liver microsomes are shown in Figure 5.3. Lansoprazole treatment produced a statistically significant increase in the activity of several pathways of testosterone metabolism. No significant effects were observed on the formation of any of the unidentified metabolites. Lansoprazole treatment significantly enhanced the total rate of formation of hydroxytestosterone metabolites by liver microsomes (P<0.001).

157 0 O I 00 II

Q. T3 •§ I £ §1

jâ -g S I ' l ^ i o I

c/5I § o o o o o o o o o fî o o o O e o o o o o O CM o 00 CD ■M- CM II

(uj9)Ojd 6ui/u!UJ/se|OUJd) ÀiiAi^oe aiuA zug 2g ^ X3 • B c/5 II a 0 1c d “ (U 1 S (U § ê S I vn00 11 I g "S S o IG

ex|: 3s IC/5 1 2 I Is i l cd I .£ : ir Il§ s : 00I +1 0 s I1§ I 1 X a o d,c -o>« n o o o o o (U S 2 m o o o o o o m o m o co CM CM à So {uiejojd Biu/uiiii/saïoiud) Aîîaiiob euiAzug I 5.3.S.2 Testosterone metabolism by testis microsomes

Control profile

Pooled testis microsomes metabolised testosterone to form the major metabolites androstenedione and 6p-0HT, along with traces of 6 a-, 7a-, 16a- and 2a-0HT. The unidentified peak Y was present in all samples, however previous studies indicate that^ this compound probably represents a contaminant that is present in the ^"^C-testosterone (chapter 4).

Peaks corresponding to androstenedione, peak Y, 6 a- and 6P-0HT were detected in blank incubations, consistent with data from the model inducers study (chapter 4). This was corrected for by subtracting the amount of each metabolite detected in the blank sample from the amount detected in the corresponding standard incubation (containing NADPH) prior to the calculation of enzyme activity. Following the subtraction of blanks no 6a-0HT or peak Y were detected.

Effect o f lansoprazole on testicular testosterone metabolism

Treatment of rats with lansoprazole produced no statistically significant effects on the pathways of testosterone metabolism catalysed by testis microsomes, as shown in Figure 5.4. Testosterone 16a- and 2a-hydfoxylase activities were very low in testis microsomes and consequently these metabolites were not detected in all incubation extracts. Consequently, it was not possible to draw any conclusions concerning the effects of lansoprazole on these pathways.

159 Figure 5.4: The effect of treatment of rats with lansoprazole on testosterone metabolism by testis microsomes

140

120

100

E Control

60 B ■ Lansoprazole- treated

6beta-OHT 7alpha-0HT 16alpha-0HT 2alpha-0HT And.

Metabolite

Each bar represents the mean ± SEM from triplicate incubations with pooled testis microsomes from control or lansoprazole-treated animals. 16a- and 2a-0H T were not detected in all three incubations using microsomes form lansoprazole-treated animals therefore SEM values could not be calculated.

160 5.3.6 ELISA

The levels of CYP proteins in liver microsomes from control and lansoprazole-treated animals were quantified by ELISA and the results are shown in Table 5.4,

Table 5.4: ELISA quantitation of CYP proteins in liver microsomes from control and lansoprazole-treated animals

CYP content (pmoles/mg protein) Anti-CYP Fold change Control Lansoprazole CYPIA 5.2 ±1.3 51.8 ±24.6 *** 9.9

CYP2B 31.2 ±3.2 47.5 ± 9.6 *** 1.5

CYP3A 19.6 ±4.0 27.3 ±8.0 ** 1.4

CYP4A 1 1 . 2 ± 2 . 0 21.9 ±6.5 *** 2 . 0

Results are expressed as mean ± SD from individual microsomal preparations from control (n=l 1) and lansoprazole-treated (n=13) groups. Fold changes were calculated by dividing the mean value in the lansoprazole-treated group by the mean value in the control group. Statistical analysis was performed using a Student’s t-test where asterisks (*) represent: ** P<0.01, ***P<0.001.

5.3.7 Western blotting

5.3.7.1 Liver microsomes

The levels of CYP proteins in liver microsomes from control and lansoprazole-treated rats were compared by Western blotting, as described in 2.2.11. Expression of CYPIA, 2B and 3A proteins were examined as these forms show testosterone hydroxylase activity. The identity of each protein was confirmed based on the literature molecular weight and . co-migration with the positive control sample. Ponceau S staining of nitrocellulose membranes confirmed even loading of protein on to each gel (data not shown). .

CYPIA

CYPlAl protein was not detected in liver microsomes from control animals but induction of this protein was observed in liver microsomes from the lansoprazole-

161 Figure 5.5: Western blot of rat liver microsomes developed with anti-rat CYPl A2 primary antibody.

Lane 1 2 3 4 5 6 7

58100

39800

5|ag microsomal protein was loaded in each lane. Lane 1, Molecular weight markers Lane 2, Positive control liver microsomes from isosafrole-treated rats Lanes 3-5, Liver microsomes from three individual control animals Lanes 6-8, Liver microsomes from three individual lansoprazole-treated animals

Figure 5.6: Western blot of rat liver microsomes developed with anti-rat CYP2B primary antibody.

Lane 1

58100

39800

5pg microsomal protein was loaded in each lane. Lane 1, Molecular weight markers Lane 2, Positive control liver microsomes from phenobarbital-treated rats Lanes 3-5, Liver microsomes from three individual control animals Lanes 6-8, Liver microsomes from three individual lansoprazole-treated animals

162 treated group (data not shown). Constitutive expression of CYP1A2 protein was detected in control liver microsomes and this protein was clearly induced in lansoprazole-treated animals (Figure 5.5). The positive control sample used for CYP1A2 showed a faint or undetectable band in this experiment. The reasons for this are unknown.

CYP2B

Anti-CYP2B detected two distinct bands in liver microsomes from control and lansoprazole-treated animals (Figure 5.6). The higher molecular weight protein was detected as a faint band present in all samples with only minor iriteranimal variability in levels of expression. The lower molecular weight protein was detected as a faint band in control animals and was induced in lansoprazole-treated rats. This protein appeared to show interanimal variability in the both the constitutive levels of expression and extent of induction. The smaller protein co-migrated with the positive control sample (PB-induced rat liver microsomes). These findings together with the published literature suggest that the higher and lower molecular weight proteins correspond to CYP2B2 and CYP2B1 respectively (Wilson et al., 1987). The polyclonal anti-CYP2B primary antibody used in the current study would recognise CYP2B1 and 2B2 as these proteins share 97% amino acid sequence homology (Yuan et al., 1983). CYP2B1 and 2B2 can be distinguished due to slight differences in migration on SDS-polyacrylamide gels, depending on the specific conditions used (Wilson et al, 1987).

CYP2A

A number of problems were encountered when Western blots were performed to detect CYP3A protein in liver microsomes from the current study. Initially, anti- CYP3A detected a single band present in microsomal samples from control and lansoprazole-treated animals (Figure 5.7). This band had a higher molecular weight than the positive control sample therefore the identity of this protein was unknown. No bands co-migrating with the positive control were detected. These bands might have been detected due to the anti-CYP3A primary antibody cross-reacting with another CYP protein. This polyclonal primary antibody would be expected to cross- react with other CYP3A proteins that are expressed in rat liver (CYP3A9, 3A18 and

163 Figure 5.7: Western blot of rat liver microsomes developed with anti-rat CYP3A primary antibody.

Lane 1 2 3 4 5 6 7

58100

39800

5|ag microsomal protein was loaded in each lane. Lane 1, Molecular weight markers Lane 2, Positive control liver microsomes from dexamethasone-treated rats Lanes 3-5, Liver microsomes from three individual control animals Lanes 6-8, Liver microsomes from three individual lansoprazole-treated animals

Figure 5.8: Western blot of rat liver microsomes developed with anti-rat CYP3A primary antibody.

Lane 1 2 3 4 5 6 7

58100

39800 ------►

lOpg microsomal protein was loaded in each lane. Lane 1, Molecular weight markers Lane 2, Positive control liver microsomes from dexamethasone-treated rats Lanes 3 & 4, Liver microsomes from two individual control animals Lane 5, Liver microsomes from a PCN-treated animal (model inducers study) Lanes 6-8, Liver microsomes from three individual lansoprazole-treated animals

164 3A23). However, this band is unlikely to correspond to a CYP protein as it has a molecular weight greater than 58,00 daltons. The possibility that the higher molecular weight bands were detected due to non-specific binding of the secondary antibody was investigated by stripping the blots of primary antibody and re-probing with secondary antibody and streptavidin-HRP complex alone. No bands were detected under these conditions (data not shown).

The CYP3A blots were repeated to determine whether the higher molecular weight bands could be detected reproducibly. In this experiment, the positive control band was present, confirming that the immunodetection procedure had worked, but no CYP3A protein was detected in any of the microsomal samples (data not shown). The higher molecular weight bands were not detected in this experiment, therefore the identity of the protein(s) detected in the previous experiment was not investigated further.

The amount of microsomal protein loaded on to the gel was increased firom 5pg to lOpg per lane to determine whether there was a sensitivity problem. A lane containing liver microsomes from a PCN-treated animal (firom the model inducer study) was included on each blot as an additional positive control. CYP3A protein was detected as a bright band in liver microsomes from the PCN-treated animal (Figure 5.8). Faint bands corresponding to CYP3A protein were detected in microsomes firom control and lansoprazole-treated animals, although this protein was present at very low levels and was not detected in some samples. These Western blots indicate that lansoprazole did not have a marked effect on the expression of CYP3A protein. However, subtle effects of this compound on CYP3A protein levels might have been missed due to the low and sometimes undetectable levels of this protein in microsome samples firom this study.

Finally, an experiment was performed to test an anti-CYP3A2 primary antibody from a different supplier (Gentest). Gels were loaded with positive control samples and liver microsomes from both the lansoprazole study and model inducer study (Figure 5.9). CYP3A2 was easily detected in liver microsomes firom a control animal from the model inducers study and this protein was markedly induced in PCN-treated animals.

165 Figure 5.9: Western blot of rat liver microsomes developed with anti-rat CYP3A2 primary antibody from Gentest.

Lane 1 7

lOjLig microsomal protein was loaded in each lane. Lane 1, Molecular weight markers Lane 2, Positive control liver microsomes from dexamethasone-treated rats (Amersham) Lane 3, Positive control liver microsomes from PB-treated rats (Gentest) Lane 4, Liver microsomes from a control animal from the lansoprazole study Lane 5, Liver mierosomes from a lansoprazole-treated animal Lane 6, Liver mierosomes from a eontrol animal from the model inducers study Lane 7, Liver microsomes from a PCN-treated animal from the model inducers study.

Figure 5.10: Western blot of rat testis microsomes developed with anti-rat CYP4A1 primary antibody.

Lane 1 2 3 4 5 6 7 8

58100

39800 ------►

The amount of microsomal protein loaded in to each lane is shown in brackets. Lane 1, Molecular weight markers Lane 2, Positive control liver microsomes from clofibrate-treated rats Lane 3 & 4, Pooled testis microsomes from control group (98pg protein) Lane 5, Pooled testis mierosomes from control group (40pg protein) Lane 6, Pooled testis microsomes from lansoprazole-treated group (45pg protein) Lanes 7 & 8, Pooled testis microsomes from lansoprazole-treated group (113pg protein)

166 In contrast, faint bands corresponding to CYP3A2 protein were detected in liver microsomes from control and lansoprazole-treated animals, only following long exposures of the blot to the film (5 minutes). It was not possible to draw reliable conclusions concerning the effects of lansoprazole on CYP3A levels from these blots due to the low levels of expression of this protein and the small number of samples investigated (one control and one lansoprazole-treated animal). In addition, further experiments would be required to optimise the concentrations of primary and secondary antibodies used.

5.3.7.2 Testis microsomes

Western blotting with testis microsomes failed to detect any CYP3A or CYP2B- immunoreactive protein despite loading maximum amounts of protein on to the gels (data not shown). CYP4A1 was detected in testis microsomes from control and lansoprazole-treated animals (Figure 5.10). Due to uneven protein loading it was not possible to conclude whether testicular CYP4A1 had been induced by lansoprazole treatment.

167 5.4 Discussion

5.4.1 Final body and organ weights

Treatment of rats with lansoprazole was associated with lower body weight gain during the dosing period leading to a statistically significant reduction in final body weights compared to the control group (P<0.01). Lansoprazole treatment (> 50 mg/kg/day) has previously been found to produce a dose-dependent decrease in body weight gain associated with reduced food consumption in short- and long-term toxicity studies (Unpublished reports by Takeda Chemical Industries Ltd., Report numbers: A-29-167; A-29-452; A-29-438). In contrast, other studies have reported no consistent drug-related changes in these parameters at the same dosages (Unpublished reports by Takeda Chemical Industries Ltd., Report numbers: A-29-168; A-29-345). In the current study, all animals were examined daily for clinical signs related to treatment and appeared healthy throughout the dosing period.

Relative liver weights were significantly higher in lansoprazole-treated animals compared to the control group (-109% of control) (P<0.05), however no change in absolute liver weights were observed. This apparent increase in relative liver weights might be an artefact due to the significant reduction in final body weights of drug- treated animals. However, lansoprazole treatment (>50 mg/kg/day for 4 weeks) has previously been shown to cause a dose-dependent increase in relative liver weights in male Wistar rats (Unpublished report by Takeda Chemical Industries Ltd., Report number A-29-167). Histopathological examination of livers from lansoprazole-treated animals revealed a small increase in smooth endoplasmic reticulum of the centrilobular hepatocytes (Unpublished reports by Takeda Chemical Industries Ltd., Report number A-29-168). In the current study, lansoprazole had no significant effect on the hepatic microsomal protein yield but produced a significant increase in the specific microsomal CYP content (-1.2 fold) (P<0.01). These findings are consistent with the effects reported in female rats following 7 days of lansoprazole administration (300 mg/kg/day) (Masubuchi et al., 1997a).

Relative testes weights were significantly higher in lansoprazole-treated animals compared to the control group (P<0.05), but this was probably related to the reduced final body weights of drug-treated animals as previously discussed; ^ contrast to the

168 current data, Fort et al. (1995) reported a significant reduction in testicular weights (-20% decrease) in rats treated with lansoprazole (150 mg/kg/day) for four weeks.

5.4.2 Effect of lansoprazole on hepatic CYP proteins

Two complementary techniques, namely ELISA and Western blotting, were employed to compare the levels of expression of CYP proteins in liver microsomes from control and lansoprazole-treated animals. The ELISA method was useful as it provided numerical data allowing statistical comparisons to be performed. The major disadvantage associated with this technique is that it is not possible to confirm that the antibody is only cross-reacting with the specific protein of interest. During the optimisation of the ELISA method the binding specificity of the anti-CYP primary antibodies was confirmed by Western blotting (see chapter 3). However, these results can only be used as a guide due to potential differences in antibody-antigen interactions between the denatured protein samples used for Western blotting and the solubilised proteins used for the ELISA. Western blotting was used in the current study to provide qualitative data concerning the effects of lansoprazole on CYP expression with the advantage that the identity of the protein ' of interest could be confirmed based on the molecular weight.

Lansoprazole treatment was associated with statistically significant increases in the expression of proteins belonging to the CYPIA (9.9 fold), 2B (1.5 fold), 3 A (1.4 fold) and 4A (2 fold) subfamilies as determined by ELISA. Lansoprazole treatment had the ; most prominent effect on CYPIA expression and Western blotting experiments confirmed that this compound induced both CYPlAl and 1A2 protein levels. Western blotting also confirmed that lansoprazole induced CYP2B1, with little effect on CYP2B2 protein levels.

Liver microsomes from this study contained low or undetectable levels of CYP3A protein, as determined by Western blotting (Figures 5.8 and 5.9). The band corresponding to CYP3A that was detected in liver microsomes using the polyclonal anti-CYP3A antibody is likely to be composed of at least two proteins, namely CYP3A1 and 3A2, which have previously .been shown to co-migrate on SDS- polyacrylamide gels (Gemzik et al., 1992). According to the published literature

169 CYP3A2 is constitutively expressed at measurable levels whereas CYP3A1 is difficult to detect in liver microsomes from control animals (Cooper et al., 1993; Debri et al., 1995). Western blotting data indicated that lansoprazole treatment had no marked effect on CYP3A protein levels, although subtle effects might have been missed due to the low levels of expression in liver microsomes from this study. These findings are inconsistent with the results obtained using the testosterone hydroxylase assay where lansoprazole treatment produced a statistically significant increase in microsomal testosterone 6 p-hydroxylase activity (1.4 fold), which is a more sensitive indicator of CYP3A expression. In addition, a significant increase in CYP3A protein levels (1.4 fold) was detected in liver microsomes from lansoprazole-treated animals compared to the control group by ELISA. Together these findings indicate that lansoprazole treatment caused a modest induction of hepatic CYP3A proteins and associated catalytic activity, which was not detected by Western blotting. One possible explanation for the inconsistencies in the data obtained using the testosterone hydroxylase assay and the ELISA compared to Western blotting might be related to differences in the microsome processing required for these methods. On a Western blot, the primary antibody interacts with a denatured protein sample. Consequently, the processing required to denature the liver microsomes might have resulted in loss of the epitope(s) in the CYP3A proteins leading to reduced antibody binding. In contrast, the microsomal proteins are retained in a native conformation for the testosterone hydroxylase assay and the ELISA.

In conclusion, lansoprazole induced several CYP proteins across the CYPIA, 2B, 3A and 4A subfamilies and data from the testosterone hydroxylase assay suggests that this compound might also induce CYP2A (7a-hydroxylase activity) and CYP2C forms

(2a-hydroxylase activity) in male rats. These findings are in agreement with previously published data for female Sprague Dawley rats (Masubuchi et al., 1997a). Treatment of female rats with lansoprazole (300 mg/kg/day) for 7 days increased the expression of CY PlA l (9.3 fold), 1A2 (6.3 fold), 2B1 (2.7 fold), 2B2 (2.4 fold), 2C6 (2 fold), 3A2 (3 fold) and 4A1 (1.8 fold) proteins in liver microsomes (Masubuchi et al., 1997a). However, most of these CYP forms show sex-dependent expression therefore it is not possible to directly compare male and female derived data. The ability of lansoprazole to induce the expression of CYPs across at least four families

170 suggests that this compound might represent a distinct type of “mixed inducer”, as no other published examples of compounds with this property could be identified.

A number of agents have been identified that induce CYP forms across three families, such as phenothiazine, indole-3-carbinol (13C), (methylthio) methylpyrazine (MTMP) and flavastatin. Phenothiazine and 13 C both induce hepatic CYPs belonging to the CYPIA, 2B and 3 A subfamilies in male rats, however 13 C itself might not be entirely responsible for the effects observed on the hepatic CYP complement (Tateishi et al., 1999; Stresser et al., 1994). Following oral administration, 13C is transformed at gastric pH to a number of condensation products (e.g. diindolylmethane and 2,3-bis (3-indolylmethyl)indole) and there is evidence to support a role for these derivatives in mediating at least some of the effects on the hepatic CYP complement (De Kruif et al., 1991; Wortelboer et al., 1992; Renwick et al., 1999). Treatment of male rats with the pyrazine derivative, MTMP, is associated with induction of CYP2B1, 3A and 2E1 (Japenga et al., 1993). Fluvastatin treatment has been shown to induce the hepatic expression of CYPlAl, 2B1/2 and 4A proteins with no effect on CYP3A levels in male rats (Kocarek & Reddy, 1996).

5.4.3 Effect of lansoprazole on hepatic testosterone metabolism

The profile of metabolites detected in incubation extracts from control liver microsomes from the current study was in good agreement with previously published data for male Sprague Dawley rats (Purdon & Lehman-McKeeman, 1997; Parkinson et al., 1992). Liver microsomes from this study showed higher absolute rates of testosterone hydroxylation compared to samples from the model inducers study

(Chapter 4), particularly with respect to 6 p-hydroxylase activity.

Treatment of rats with lansoprazole produced significant increases in the activity of several pathways of testosterone oxidation catalysed by liver microsomes. Increases in testosterone 6 p* and 2p-hydroxylase activities were consistent with CYP3A induction as previously discussed. Increased testosterone 16p-hydroxylase activity reflects induction of CYP2B, consistent with Western blotting and ELISA data. This compound also increased the rate of hydroxylation at the 7a- and 2a- positions.

171 suggesting induction of CYP2A and CYP2C forms respectively. This compound did not exert a large effect (< 2 fold induction) on any of the pathways studied but produced a generalised increase in the activity of several pathways of testosterone oxidation. This resulted in a significant increase in the total rate of hydroxytestosterone metabolite formation (-30% higher than controls) by rat liver microsomes from lansoprazole-treated animals (P<0.001), suggesting that this compound might have the potential to increase the rate of metabolic clearance of testosterone m vfvc>.

5.4.4 Effect of lansoprazole on testicular CYP proteins

Western blotting with testis microsomes resulted in the detection of low levels of CYP4A1 protein. Due to uneven protein loading on the gel it was not possible to draw firm conclusions as to whether lansoprazole treatment altered the expression of testicular CYP4A1. However, the differences in protein loading were relatively small (98pg and 113pg/lane for control and lansoprazole groups respectively) therefore lansoprazole treatment did not appear to have any marked effects on CYP4A1 expression in the testis.

Consistent with the published literature, CYP2B and CYP3A were not detected in testis microsomes by Western blotting (Bengtsson et al., 1990; Seng et al., 1996). This may be due to the fact that CYP2B and CYP3A proteins are not expressed in the testis or these forms might be expressed at levels that are below the limits of detection on a Western blot. CYP2B1 mRNA has been detected in testis from Sprague Dawley rats, although this doesn’t necessarily mean that the corresponding protein will be expressed (Omiecinski, 1986). Western blotting with a testicular fraction enriched with Leydig cells (e.g. purified using a percoll gradient) might increase the sensitivity of detection of CYP proteins.

5.4.5 Effect of lansoprazole on testicular testosterone metabolism

The profile of metabolites formed by pooled testis microsomes from the current study was consistent with previously published data for male Sprague Dawley rats (Sonderfan et al., 1989; Seng et al., 1996). Androstenedione formation was the major pathway of testosterone oxidation catalysed by testis microsomes. Smaller amounts of

172 6 p-, 7a-, 6 a-, 16a- and 2a-0HT were also detected. As previously discussed in Chapter 4, the levels of metabolites present in testis incubations were generally close to the limits of detection of the assay resulting in less accurate quantification of metabolites. In the current study traces of testosterone 16a- and 2a-hydroxylase activity were detected in testis microsomes. These metabolites were not detected in blank incubations but were present in incubation extracts at levels that were close to the limits of detection with respect to radiochemical detection. Consequently, these metabolites were not detected in all incubation extracts therefore it was not possible to draw conclusions concerning the effects of lansoprazole treatment on these pathways.

Lansoprazole treatment had no marked effects on the pathways of testosterone metabolism catalysed by testis microsomes. This was not entirely unexpected as lansoprazole produced relatively modest increases in the activity of CYP-dependent pathways of testosterone metabolism in the liver and testicular CYPs appear to be less sensitive to the effects of chemical inducers (Goldstein & Linko, 1984; Omiecinski, 1986).

5.4.6 Plasma hormone levels

Testosterone

Lansoprazole treatment was associated with a significant reduction ( 6 6 % of control values) in plasma testosterone levels (P<0.05). These results confirm the findings of Fort et al. (1995) who reported significant reductions in serum testosterone levels (maximal decrease of 50%) between 3.5 and 5 hours postdose in male rats following treatment with lansoprazole (150 mg/kg/day) for four weeks. Plasma testosterone levels reflect the balance between the rate of hormone biosynthesis in the testis and clearance via the liver and kidneys. The reduced circulating testosterone levels observed following lansoprazole treatment are probably, at least in part, related to the ability of this compound to inhibit testosterone biosynthesis (Fort et al., 1995). However, our current data from the testosterone hydroxylase assay demonstrate that lansoprazole increases the activity of several CYP-dependent pathways of testosterone metabolism. This might lead to enhanced metabolic clearance of testosterone, which may represent an important mechanism involved the depression of circulating testosterone levels by lansoprazole.

173 LH

Lansoprazole had no statistically significant effects on plasma LH levels in the current study. In contrast, Fort et al. (1995) reported significant elevations in serum LH levels at 4 and 5 hours postdose in lansoprazole-treated rats (150 mg/kg/day for 4 weeks). In the current study, failure to detect an increase in plasma LH levels despite the significant reduction in circulating testosterone levels might have been related to factors such as the timing of blood sample collection and interanimal variability in hormone levels as discussed previously in Chapter 4. For the current study, blood samples were collected between 18 and 2 0 hours after the last dose of lansoprazole, by which time the majority of the drug would have been cleared (half-life < 3 hours) and homeostatic regulation may have meant that plasma hormone levels were returning to the normal range. It would be useful to study the effects of lansoprazole on plasma hormone levels at a number of timepoints following the final dose of drug.

FSH

Lansoprazole-treated animals showed a trend towards higher plasma FSH levels compared to the control group, but this did not reach statistical significance. FSH has no direct effects on the Leydig cell but stimulates the Sertoli cells to secrete a variety of paracrine factors, which may influence Leydig cell structure and function. Administration of FSH to immature hypophysectomised rats has previously been shown to have a positive effect on the Leydig cell population causing hypertrophy and hyperplasia and enhanced hCG-stimulated testosterone production in vitro (Kerr & Sharpe, 1985; Vihko et al., 1991). \

Prolactin

Lansoprazole-treated animals showed a trend towards lower plasma prolactin levels compared to the control group, but this did riot reach statistical significance. Prolactin appears to be essential for normal testicular function, maintaining Leydig cell LH receptor numbers and potentiating the effects of LH on the male reproductive system (Welsh et al., 1986; Bartke et al., 1978; Zipf et al., 1978a). Prolactin also has indirect effects on testicular function by suppressing LH and FSH secretion fi*om the pituitary gland (McNeilly et al., 1978). The trend towards lower plasma prolactin levels in lansoprazole-treated animals might be of particular interest as dopamine agonists are

174 believed to induce LCTs in rats through their ability to reduce circulating prolactin levels (see 1.3.5).

5.4.7 Conclusions

Lansoprazole treatment produced effects on the liver consistent with an enhanced metabolic capacity, including significant increases in relative liver weight and total microsomal CYP content. Lansoprazole induced the expression of several hepatic CYP proteins (including CYPlAl, CYP1A2, CYP2B1, CYP3A and CYP4A1), suggesting that this compound may represent a distinct type of “mixed inducer”. This was associated with significant increases in the activity of several CYP-dependent pathways of testosterone metabolism catalysed by liver microsomes leading to an increase in the total rate of hydroxytestosterone metabolite formation in vitro. These findings suggest that lansoprazole might have the potential to enhance the metabolic clearance of testosterone in vivo. Indeed, circulating testosterone levels were significantly lower in lansoprazole-treated animals compared to the control group. No significant changes in plasma LH, FSH or prolactin levels were detected in lansoprazole-treated animals.

The effects of lansoprazole on testicular CYP enzymes have been studied, although the low levels of expression in the testis confounded the detection and accurate quantification of CYP proteins in this tissue. Lansoprazole treatment was not associated with any marked effects on CYP-dependent pathways of testosterone metabolism catalysed by testis microsomes.

175 Chapter 6

Effect of lansoprazole on gene expression in the liver and testis 6.1 Introduction

The previous study demonstrated that lansoprazole treatment induces the expression of proteins belonging to the CYPIA, 2B, 3A and 4A subfamilies in liver microsomes (chapter 5). This was associated with increases in the activity of several CYP- dependent pathways of testosterone metabolism catalysed by liver microsomes in vitro, indicating that lansoprazole might also induce enzymes belonging to the CYP2A (7a-hydroxylase activity) and 2C (2a-hydroxylase activity) subfamilies. The aim of the current study was to further characterise the effects of lansoprazole on the liver and testis.

The major objective of this study was to examine the effects of lansoprazole on the hepatic and testicular expression of selected genes of relevance to the current hypothesis. The expression of C7P2^7, CYP2C11, CYP2C13, CYP2C62Lnd l7f-HSD were studied in the liver and testis. In addition, selected genes involved in testosterone biosynthesis were examined in the testis, namely CYPscc (CYPllAJ), CYP17A1 and StAR. Further studies were also performed to characterise the effects of lansoprazole treatment on the hypothalamic-pituitary-testis axis by measuring intratesticular testosterone and plasma hormone levels (testosterone, LH, FSH and prolactin).

6.2 Study design

Male Sprague Dawley rats were dosed once daily with lansoprazole (150 mg/kg/day) or vehicle (0.5% (w / v) CMC) by oral gavage for 14 days. On day 15, blood samples were collected and livers and testes were removed. For each liver, a part of the left lobe was frozen in liquid nitrogen for total RNA extraction and the remaining tissue was used to prepare microsomes. Testes from soriie animals (n=5 per group) were frozen in liquid nitrogen for total RNA extraction whereas those from the remaining animals (n=5 per group) were used to measure intratesticular testosterone levels. More detailed experimental details concerning animal treatment, sample collection and the analytical methods used for this study can be found in chapter 2 .

177 6.3 Results

6.3.1 Body, liver and testes weights

Final body weights and relative liver and testes weights for control and drug-treated

animals are shown in Table 6 .1 .

Table 6.1: Effect of lansoprazole treatment on relative liver and testes weights

Control Lansoprazole-treated

Final body weight (g) 324.7 ± 27.7 309.7 ±17.6

Relative liver weight 4.16 ±0.23 4.24 ±0.53 (g/lOOg body weight)

Relative testes weight 0.87 ±0.15 0.85 ± 0.30 (g/lOOg body weight)

Results are expressed as mean ± SD for control and lansoprazole-treated animals (n=10 animals per group).

There were no statistically significant differences in initial (data not shown) or final body weights between the experimental groups. No statistically significant differences in absolute (data not shown) or relative liver and testes weights were observed between control and lansoprazole-treated animals.

6.3.2 Total CYP content of liver microsomes

The specific CYP contents of liver microsomes from control and lansoprazole-treated animals are shown in Table 6.2.

Table 6.2:, Effect of lansoprazole treatment on the total cytochrome P450 content of liver microsomes '

Treatment group Microsomal P450 content (nmol/mg microsomal protein)

Control 0.79 ± 0.06

Lansoprazole 0.96 ±0.17**

Results are expressed as mean ± SD from individual microsomal preparations from control and lansoprazole-treated animals (n=10 animals per group). Statistical analysis was performed using a Student’s t-test where asterisks (*) represent: ** P<0,01.

178 6.3.3 ELISA

The levels of CYP proteins in liver microsomes from control and lansoprazole-treated animals were quantified by ELISA and the results are shown in Table 6.3.

Table 6.3: ELISA quantitation of CYP proteins in liver microsomes from control and lansoprazole-treated animals

CYP content (pmoles/mg protein) Anti-CYP Fold change Control Lansoprazole

CYPIA 2.5 ±1.2 80.3 ±38.5 *** 32.1

CYP2B 34.0 ±4.1 47.0 ±11.9** 1.4

CYP3A 21.0 ±5.2 28.5 ±4.9** 1.4

CYP4A 14.5 ±2.8 28.0 ±7.3*** 1.9

Results are expressed as mean ± SD from individual microsomal preparations from control and lansoprazole-treated animals (n=10 animals per group). Fold changes were calculated by dividing the mean value in the lansoprazole-treated group by the mean value in the control group. Statistical analysis was performed using a Student’s t-test where asterisks (*) represent: ** P<0.01, *** P<0.001.

6.3.4 Plasma hormone levels

Blood samples were collected (~ 18 to 20 hours after the final dose) and assayed for plasma testosterone, LH, FSH and prolactin levels by radioimmunoassay (Figures 6.1 and 6.2). Lansoprazole-treated animals showed a trend towards lower plasma testosterone levels compared to the control group, but this did not reach statistical significance. Lansoprazole treatment was associated with a significant decrease in plasma prolactin levels (P<0.05). No statistically significant changes in plasma FSH or LH levels were detected.

179 Figure 6.1: Plasma testosterone levels in control and lansoprazole-treated rats

g 8

o 6

Control Lansoprazde-treated

Each bar represents the mean ± SEM from control (n=10) and lansoprazole-treated animals (n=10).

Figure 6.2: Plasma prolactin, follicle-stimulating hormone and luteinising hormone levels in control and lansoprazole-treated rats.

□ Contro

Lansoprazole treated

Prolactin

Hormone

Each bar represents the mean ± SEM from control (n=10) and lansoprazole-treated animals (n=10). Statistical analysis was performed using a Student’s t-test where asterisks (*) represent, * P<0.05.

180 6.3.5 Intratesticular testosterone levels

The effect of lansoprazole treatment on intratesticular testosterone levels was investigated in a subgroup of animals from this study (n=5 animals per group). Testes from each animal were homogenised and the supernatant was assayed for testosterone levels by radioimmunoassay.

The lansoprazole-treated group showed a trend towards lower intratesticular testosterone levels compared to controls, but this did not reach statistical significance (Figure 6.3). To enable comparison of the effects of lansoprazole treatment on testicular and plasma hormone levels in these animals, plasma testosterone levels in this subgroup are shown in Figure 6.4. It was observed that intratesticular and plasma testosterone levels were markedly reduced in four of the lansoprazole-treated animals. In contrast, the fifth drug-treated animal exhibited intratesticular and plasma testosterone concentrations that were within the control range. If this animal was omitted from the statistical analysis, intratesticular testosterone ’ levels were significantly lower in the lansoprazole-treated group (37.7 ±1.0 ng/g testis (mean ±

SD)) compared to the control group (70.8 ± 19.1 ng/g testis) (P<0.05).

181 Figure 6.3; Intratesticular testosterone levels in control and lansoprazole-treated animals

Lansoprazde-treated

Each bar represents the mean ± SEM from control (n=5) and lansoprazole-treated (n=5) animals.

Figure 6.4: Plasma testosterone levels in the subgroup of control and lansoprazole- treated animals used to measure intratesticular hormone levels

Lansoprazole

Each bar represents the mean ± SEM from control (n=5) and lansoprazole-treated (n=5) animals. Statistical analysis was performed using a Student’s t-test where asterisks (*) represent, * P<0.05.

182 6.3.6 TaqMan analysis

6.3.6.1 Preparation of cDNA

Total RNA was extracted from each liver and testis sample and then purified using the RNAqueous™-4PCR kit as described in 2.2.14. Each total RNA sample was quantified by speetrophotometry and the quality of the sample was assessed by separation on a 1% agarose gel. Figure 6.5 shows a representative total RNA sample from this study. Two distinct bands corresponding to 28S and 18S ribosomal RNA (rRNA) with a ratio of approximately 2:1 are present, indicating that no significant degradation of the RNA has occurred. 3pg of each total RNA sample was used for cDNA synthesis as described in 2.2.14.4.

Figure 6.5: Total RNA extracted from rat testis

Lane 1 Lane 2

^ ______28S rRNA

18SrRNA

The total RNA sample was mixed with Gel Loading Solution (4pl:lpl RNA) and then loaded on to a 1% agarose gel. Electrophoresis was conducted at lOOV for approximately 45 minutes as described in 2.2.14.3.

Lane 1, 1Kb DNA ladder (6pl) Lane 2, Total RNA (2 pi) extracted from testis from a control animal

6.3.6.2 Real-time PCR (TaqMan^^ quantitatation

Real-time PCR (TaqMan™) was used to study the levels of expression of the selected genes in liver and testis samples from control and lansoprazole-treated animals as described in 2.2.15. For each sample, the initial copy number of the target mRNA or 18S rRNA was interpolated from the appropriate standard curve. The copy number of

183 the target mRNA was then divided by the copy number for 18S rRNA to give a normalised target value. Fold changes in target mRNA levels were calculated by dividing the mean normalised target value in the drug-treated group by that in the control group. The data obtained for liver and testis samples is presented below.

Liver samples

Initially, the effect of lansoprazole on hepatic CYPlAl mRNA levels was studied as a positive control because this compound is known to produce a large induction of CYPl A1 protein levels (chapter 5 and Table 6.3). Lansoprazole treatment produced a significant increase in hepatic CYPlAl mRNA levels (16.8 fold induction) as shown

in Figure 6 .6 .

The effects of lansoprazole treatment on the five candidate genes studied in the hver are shown in Figure 6.7.

Figure 6.6: Effect of lansoprazole treatment on liver C YPlAl mRNA levels as determined by TaqMan

** 45

40

35

30

25

Control Lansoprazole

Each bar represents the mean ± SEM from control (n=4) and lansoprazole-treated (n=4) animals. Statistical analysis was performed using a Student’s t-test, where asterisks (*) represent: ** P<0.01.

184 Figure 6.7: Fold changes in hver mRNA levels of selected genes following treatment of rats with lansoprazole as determined by TaqMan

24

22

o> 1-8 c (0 .co jO I? 1.6

1.4

1.2

çyp2ai C/P2C6 C/R2C13 C/P2A1 UbetaHSD

Each bar represents the mean ± SEM fold change in mRNA levels of each gene compared to the control group (n=5 animals per group). Statistical analysis was performed using a Student’s t-test, where asterisks (*) represent; * P<0.05.

Testis samples

The effects of lansoprazole treatment on the expression of the eight candidate genes studied in the testis are shown in Figure 6 .8 . Data is not presented for CYPscc because the primer and probe set designed for the current study failed to amplify the target sequence, despite attempts to optimise primer and probe concentrations in the reaction mixture and the use of a cDNA standard curve.

185 g co a

IC/ü c3 g a. I

cd o

S i3 W) c "S ooVO 0 3 8 s E a I %

1 ia < < W) 1 1 2 4 2

eng -H bû §

2 (£ cîô \ 6 I I eBueijo p|oj I 6.4 Discussion

6.4.1 Final body and organ weights

Lansoprazole treatment had no statistically significant effects on body weight gain or final body weights in the current study, in contrast to the significant reductions observed in the previous study (chapter 5). In addition, lansoprazole had no significant effects on liver and testes weights in the current study. In contrast, lansoprazole treatment produced a significant increase in relative liver weights and a reduction in relative testes weights in the previous study (chapter 5).

6.4.2 Effect of lansoprazole on hepatic CYPs

In the Current study, lansoprazole treatment produced a significant increase in the specific microsomal CYP content ( 1 . 2 fold) which was associated with induction of CYPIA, 2B, 3A and 4A proteins as determined by ELISA. These findings were in good agreement with data firom the previous study except that the fold increase in CYPIA levels calculated for liver microsomes firom the current study (32.1 fold) was markedly higher than for samples from chapter 5 (9.9 fold). This mainly resulted from differences in CYPIA levels in control samples (2.5 ± 1.2 and 5.2 ±1.3 pmoles/mg microsomal protein (mean ± SD) for control liver microsomes from the current study and chapter 5 respectively). The levels of CYPIA proteins present in control liver microsomes were low and despite being within the range of the standard curve the levels of CYPIA proteins were close to the limits of detection of the ELISA assay. This might have resulted in less accurate quantitation of CYPIA levels in control samples. In addition, these differences between studies might have been related to interanimal variability in CYPIA levels and responses to drug treatment.

In conclusion, these results provide confirmation of the previously observed effects of lansoprazole treatment on the hepatic CYP complement and also act as a surrogate marker for drug exposure in the current study.

187 6.4.3 Plasma hormone levels

Lansoprazole-treated animals showed a trend towards lower plasma testosterone levels compared to the control group, but this did not reach statistical significance. The magnitude of the reduction in circulating testosterone levels in the current study (80%

of control) was smaller than that observed in the previous study ( 6 6 % of control) (chapter 5). Lansoprazole treatment had no statistically significant effects on plasma LH levels, consistent with data from the previous study (chapter 5). Lansoprazole- treated animals showed a trend towards lower plasma FSH levels compared to the control group, but this did not reach statistical significance. In contrast, a trend towards higher plasma FSH levels was observed in the previous study (chapter 5).

In the current study, the most marked effect of lansoprazole treatnient was the significant reduction in plasma prolactin levels (32% of control) (P<0.05). In the previous study, there was a trend towards lower plasma prolactin levels in the drug- treated group, but this did not reach statistical significance (chapter 5). The ability of lansoprazole to reduce circulating prolactin levels might be involved in the mechanism of LCT induction by this compound. Prolactin appears to play an important role in regulation of the male reproductive system through direct (e.g. maintenance of Leydig cell LH receptors) and indirect effects (suppression of gonadotropin secretion from the pituitary gland) on testicular function (Zipf et ah, 1978a; McNeilly et ah, 1978; Winters & Loriaux, 1978; Smith & Bartke, 1987; Cheung, 1983). Therefore changes in plasma prolactin levels might lead to changes in circulating gonadotropin levels and modulation of the sensitivity of the Leydig cells to ÉH stimulation. Indeed, induction of hypoprolactinaemia by treatment of rats with 2-bromo-a-ergocriptine (BR) has been reported to produce a reduction in testicular LH receptor numbers compared to control animals (Huhtaniemi & Catt, 1981; Pakarinen et ah, 1994). Huhtaniemi & Catt (1981) reported that this was not associated with a reduction in testicular steroidogenesis or circulating testosterone levels, which might be explained by the presence of a large excess of “spare” LH receptors. In contrast, Pakarinen et ah

(1994) reported a significant reduction in serum testosterone levels following 8 days of BR-treatment in male rats. The apparent differences in the effects of BR-induced hypoprolactinaemia on circulating testosterone levels between these two studies might have been related to differences in the experimental conditions used such as the animal

188 age (60 days v 90 days old for Huhtaniemi & Catt (1981) and Pakarinen et ah, (1994) respectively).

Prolactin has previously been implicated in the aetiology of. LCT formation in rats. Dopamine agonists (or agents that enhance dopamine levels) may produce LCTs through their ability to reduce circulating prolactin levels. It has been proposed that this might cause downregulation of Leydig cell LH receptors and a partial inhibition of testosterone biosynthesis leading to a compensatory increase in LH secretion (Prentice et al., 1992). Consistent with this hypothesis, the dopamine agonist mesulergine has been reported to reduce circulating prolactin levels and Leydig cell LH receptor numbers, which was associated with sustained increases in circulating LH levels in rats (Prentice et ah, 1,992). Chronic treatment of rats with mesulergine is associated with an increased incidence of Leydig cell hyperplasia and tumours in rats (Dirami et ah, 1996). Interestingly, induction of chronic hyperprolactinaemia has been reported to markedly reduce the incidence of spontaneous LCTs in Fischer 344 rats and the authors postulated that this was due to suppression of circulating LH levels (Bartke et ah, 1985).

6.4.4 Intratesticular testosterone levels

Lansoprazole-treated animals showed a trend towards lower intratesticular testosterone levels (72% of control), but this did not reach statistical significance. These results confirm the findings of Fort et ah (1995) who reported significant reductions in intratesticular testosterone levels (67% of control) at 4 hours postdose in male Sprague Dawley rats following treatment with lansoprazole (I50mg/kg/day) for four weeks. Plasma testosterone levels were also significantly reduced in drug-treated animals firom this subgroup ( 6 6 % of control) (P<0.05). The overall magnitude of the decrease in plasma and intratesticular hormone levels was reduced due to the fact that one of the lansoprazole-treated animals did not show marked changes in testosterone levels. The reasons for the apparent lack of effect of drug treatment on testosterone levels in this animal are unknown, but this is unlikely to be due to inadequate drug exposure as the levels of CYP induction measured in liver microsomes (by ELISA) from this animal were within the range of the lansoprazole-treated group.

189 Measurement of intratesticular hormone levels provides an indicator of the biosynthesis, secretion and bioavailability of testosterone within the testis. In contrast, plasma testosterone levels reflect the balance between the rate of hormone biosynthesis in the testis and clearance via the liver and kidneys. Lansoprazole has previously been shown to inhibit several sites in the testosterone biosynthetic pathway in vitro, the most sensitive site being transport of cholesterol to the CYPscc enzyme (Fort et al., 1995). Furthermore, treatment of rats with lansoprazole (50 or 150 mg/kg/day) for one or two weeks was associated a reduction in hCG- stimulated testosterone secretion in vivo (Fort et al., 1995). The ability of lansoprazole to inhibit testosterone biosynthesis is probably the major factor responsible for the reduction in intratesticular hormone levels in drug-treated animals. .Inhibition of testosterone biosynthesis would be expected to contribute to the decrease in circulating testosterone levels observed in lansoprazole-treated animals, but changes in other processes (e.g. metabolic clearance) might also contribute as previously discussed in chapter 5.

The ability of lansoprazole to reduce the local concentration of testosterone might have important effects on the testis. Testosterone is essential for the maintenance of tubular spermatogenesis therefore sustained reductions in intratesticular hormone levels might be associated with suppression of spermatogenesis. Indeed, treatment of Sprague Dawley rats with lansoprazole has been reported to produce testicular atrophy and epididymal hypospermia in a 3 month study (> 300 mg/kg/day) and an increased incidence of seminiferous tubule atrophy in a 2 year study (> 75 mg/kg/day) (Unpublished reports by Takeda Chemical Industries Ltd., Report numbers: TA-90- 151 and TA-91-024). There is evidence to suggest that the seminiferous tubules might have a regulatory influence on Leydig cell morphology and function. In rats, Leydig cells located adjacent to seminiferous tubules at certain stages of germ cell development (stages VII to VIII) have been reported to be larger and contain more smooth endoplasmic reticulum compared those lying close to tubules at other stages (Bergh, 1982; Fouquet, 1987). Since the volume of smooth endoplasmic reticulum is correlated to the steroidogenic capacity of the Leydig cell, stage VII and VIII seminiferous tubules may modulate the ability of adjacent Leydig cells to produce testosterone (Russell et al., 1992). Interestingly, Sharpe et al. (1992) identified several

190 androgen-regulated proteins that were produced by stage VI to VIII seminiferous tubules, although the identity and cellular source (i.e. germ cells, Sertoli cells or peritubular cells) of these proteins was not confirmed in this study. These findings indicate that reciprocal regulatory pathways may exist between the seminiferous tubules and the Leydig cells. Furthermore, these pathways might be altered following lansoprazole treatment due to the ability of this compound to reduce the local testosterone concentration and produce tubular atrophy.

There is evidence that testosterone may modulate the production of several potential regulatory molecules that are produced within the testis. PModS (Peritubular factor that Modulates Sertoli cell function) is produced by the peritubular cells and has been reported to act as a potent stimulator of Sertoli cell differentiated function in vitro (e.g. production of transferrin, ABP and inhibin) (Norton & Skinner, 1989; Skinner & Fritz, 1985; Skinner et ah, 1989). The peritubular cells express androgen receptors and there is evidence to suggest that androgens have a stimulatory effect on PModS secretion (Anthony et al., 1989; Skinner & Fritz, 1985; Verhoeven & Cailleau, 1988). At present this regulatory factor has only been partially characterised but it has been suggested that it might mediate some of the effects of androgens on the Sertoli cells (Skinner & Fritz, 1985). Other examples of potential testicular paracrine factors that might be modulated by testosterone include P-endorphin and oxytocin (Fabbri & Dufau, 1988; Nicholson et al., 1986). The ability of lansoprazole to reduce intratesticular testosterone levels might have important effects on the paracrine environment of the testis, which could play a role in thé formation of LCTs.

6.4.5 Effect of lansoprazole on gene expression in the liver and testis

Selection and quantitation o f candidate genes

The major aim of the current study was to investigate the effect of drug treatment on the expression of selected genes to further characterise the effects of lansoprazole on the liver and testis. Eight candidate genes of relevance to our hypothesis were chosen: CYP2A1, CYP2C11, CYP2C6, CYP2C13, 17P-HSD, CYPscc, CYP17A1 and StAR. CYP2C11, 2C6, 2C13 and 2A1 were selected because data firom the testosterone hydroxylase assay indicated that in addition to induction of CYPIA, 2B, 3A and 4A proteins, lansoprazole might also induce CYP2A (7a-hydroxylase activity) and

191 CYP2C (2a- and 16a-hydroxylase activity) forms (see chapter 5). Effects of drug- treatment on CYP2C forms might be of particular interest, as this subfamily accounts for a major proportion of the total hepatic CYP content in untreated rats (Guengerich et ah, 1982). In addition, although the physiological roles of constitutive CYPs have not been fully elucidated it has been suggested that CYP2C11 makes a substantial contribution to the overall rate of hepatic testosterone oxidation in rats and might even be responsible for the formation of specific hydroxylated metabolites with important physiological functions (Waxman, 1984).

17p-HSD enzymes appear to play an important role in androgen homeostasis by regulating the interconversion of high activity 17p-hydroxysteroids (e.g. testosterone) and low activity 17-ketosteroids (e.g. androstenedione). Several distinct 17p-HSD enzymes have been identified, which differ in their substrate specificity, cofactor requirements, preference for oxidative or reductive reactions and subcellular and tissue distribution (reviewed by Peltoketo et ah, 1999). For the current study, 17p-HSD type

2 was selected because this form is known to be expressed at high levels in the liver and preferentially catalyses the oxidative pathway (i.e. conversion of testosterone to androstenedione) (Akinola et ah, 1996). It has been suggested that 17p-HSD type 2 might play a protective role in the tissues by lowering the intracellular concentration of active sex steroids (Akinola et ah, 1996).

As previously discussed, lansoprazole is known to inhibit testosterone biosynthesis therefore we decided to further investigate the effects on this compound on key components of this pathway. CYPscc (product of the CYPllAl gene) catalyses the conversion of cholesterol to pregnenolone, which is the first and rate-limiting enzymatic step in the testosterone biosynthetic pathway (Simpson, 1979). In mice, the ability of the Leydig cells to synthesise androgens is significantly correlated to CYPscc protein levels therefore factors that regulate the expression of this protein would be expected to influence the steroidogenic capacity of the testis (Nolan & Payne, 1990). LH is required for maximal expression of testicular CYPscc but a component of this activity appears to be constitutive (Anakwe & Payne, 1992; Payne et al., 1992; Anderson & Mendleson, 1985). CYP 17AI is considered to be rate limiting in the conversion of pregnenolone to testosterone and LH is essential for the

192 maintenance of normal amounts of this enzyme in the testis (Anakwe & Payne, 1992; Payne et ah, 1992). In addition to gonadotropins, intratesticular autocrine and paracrine factors may modulate the expression of these steroidogenic enzymes (reviewed by Cook et al., 1999; Saez, 1994).

The StAR protein mediates the transfer of cholesterol to the inner mitochondrial membrane, which is the true rate-limiting step in steroid biosynthesis during periods of acute stimulation (Crivello & Jefcoate, 1980; Privalle et al., 1983; Clark et al., 1994; Luo et al., 1998). StAR mRNA and protein levels are correlated to the acute increase in steroidogenesis following LH stimulation (Luo et al., 1998; Clark et al., 1995). For the current study it was of interest to examine the effects of lansoprazole treatment on StAR mRNA levels as an indicator of effects on testosterone biosynthesis. In addition, stimulation of the Leydig cells by LH appears to induce StAR gene transcription therefore changes in StAR mRNA levels might provide a marker for the levels of endogenous LH stimulation (Clark et al., 1995; Caron et al., 1997). However, StAR gene expression is also influenced by autocrine and paracrine factors such as IGF-1, TGF-P and TNF-a (reviewed by Christensen & Strauss, 2000).

For the current study it was decided to quantify mRNA levels for the selected genes using real-time PCR (TaqMan™) for a number of reasons. Firstly, the TaqMan assay has high specificity due to the fact that signal generation requires the binding of the two primers and the probe within the target sequence. In contrast, quantification of CYPs using immunochemical techniques, such as Western blotting, is often hindered by the high degree of amino acid sequence similarity between individual CYP forms. This is particularly true for the discrimination of individual CYP2C forms for which it is difficult to reproducibly obtain good resolution by SDS-PAGE and primary antibodies show a high degree of cross reactivity with other members of this subfamily (Waxman, 1991). Secondly, the TaqMan assay shows high sensitivity, which was particularly important with respect to the low levels of CYP expression in testis samples. Finally, this assay is rapid, reproducible and relatively simple to perform, allowing the expression of the eight candidate genes to be measured in samples from the current study. The timescale of the current project would not have been sufficient to allow the optimisation of Western blotting for the eight candidate proteins. Instead, the effects of drug treatment were investigated at the mRNA level

193 with the intention of following up any significant changes with protein studies (e.g. Western blotting, immunohistochemistry).

Liver samples

In liver samples, the effect of lansoprazole treatment on the expression of CYP2C11, CYP2C6, CYP2C13, CYP2A1 and lip-H SD was studied. In addition, CYPlAl mRNA levels were measured as a positive control because lansoprazole is known to produce a marked induction of CYPlAl protein in the liver (see chapter 5 and Table 6.3). Indeed, a significant increase in hepatic C YPlAl mRNA levels (16.8 fold) was observed in the drug-treated group, compared to controls (P<0.01). In contrast, lansoprazole treatment had no statistically significant effects on hepatic CYP2A1, CYP2C11, CYP2C13 or 17p-HSD mRNA levels. CYP2C6 mRNA levels were significantly higher (P<0.05) in the lansoprazole-treated animals compared to the control group. However, the mean fold change did not reach the threshold value of two fold, which is generally accepted to be the minimum fold change that represents a true effect using the TaqMan assay.

The lack of effect of lansoprazole on the expression of CYP2A1 md CYP2C forms was not consistent with the ability of this compound to increase testosterone 7a- and

2a-hydroxylase activity respectively. However, the extent of the induction of these testosterone hydroxylase activities was relatively small (less than two fold) therefore it might be expected that any associated increases in mRNA levels would be of a similar magnitude. The statistically significant increase in CYP2C6 mRNA levels might therefore account for the increase in testosterone 2 a-hydroxylase activity in liver microsomes from lansoprazole-treated animals, although reliable conclusions cannot be drawn as discussed above. In addition, induction of many cytochromes P450 involves increased gene transcription but other mechanisms have been identified. Such as mRNA stabilisation (e.g. CYPlAl, CYP2B1/2, CYP2C12), changes in translation (e.g. CYP2B1) and protein stabilisation (e.g. CYP2E1, CYP3A1/2) (reviewed by Lewis, 2001). Lansoprazole might alter the levels of CYP proteins by influencing processes other than gene transcription, highlighting the importance of combining studies performed at the mRNA level with protein studies.

194 Testis samples

In testis samples, the effect of lansoprazole treatment on the expression of CYPlAl, CYP2C11, CYP2C6, CYP2C13, CYP2A1, 17P-HSD, CYPscc, CYP17A1 and StAR was studied. Unfortunately it was not possible to measure CYPscc expression because the primer and probe set designed for the current study did not amplify the target sequence despite attempts to optimise the assay. As lansoprazole treatment produced a marked induction of C YPlAl mRNA levels in the liver we investigated the effects of this compound on the expression of C YPlAl in the testis. CTPM7 was detected at very low levels in testis samples and was not significantly affected by lansoprazole treatment. These findings provide further evidence that CYPs in the testis might be less sensitive to the effects of exogenous inducers than their hepatic counterparts.

CYP2C11, CYP2C13, CYP2C6, CYP2A1, 17f-HSD, CYP17A1 and StAR transcripts were detected in testis samples, however lansoprazole treatment did not significantly affect the expression of any of these genes. Consistent with the detection of CYP2C11 and CYP2A1 transcripts in testis samples from the current study, other investigators have reported the presence of the corresponding proteins in rat testis microsomes (Ryan et al., 1993; Seng et al., 1996; Sonderfan et al., 1989). Testicular CYP2A1 levels are of particular interest with respect to the current project due to the potential physiological roles of 7a-hydroxylated androgens and the possible involvement of high CYP2A1 expression in the aetiology of spontaneous Leydig cell hyperplasia in rats (see section 1.5.3). Lansoprazole treatment did not produce significant changes in CYP2A1 levels in the testis, which was consistent with the previously observed lack of effect of this compound on testicular testosterone 7a-hydroxylase activity (chapter 5). No other studies reporting the detection of CYP2C13 or CYP2C6 mRNA or proteins in rat testis could be identified in the published literature. Consistent with the detection of 17p-HSD type 2 mRNA in testis samples from the current study, other investigators have reported that testis preparations can catalyse the oxidation of testosterone to form androstenedione (Martel et al., 1992). In the current study, 17p- HSD showed marked interanimal variation in levels of expression in testis samples, which was confirmed by repeat experiments.

Lansoprazole might be expected to influence the expression of enzymes involved in

195 testosterone biosynthesis, either directly or as a consequence of the reduction in testicular and circulating testosterone levels. In the current study, lansoprazole treatment had no effect on testicular C Y P llA l or StAR mRNA levels, although this compound might have influenced processes other than gene transcription (e.g. translation or protein stability). However, in human tissues (including the testis) StAR mRNA levels have been reported to correlate well with protein levels as determined by immunohistochemistry (Pollack et al., 1997). The lack of effect of lansoprazole on StAR mRNA levels was consistent with the fact that no significant changes in plasma LH levels were detected in these animals.

6.4.6 Conclusions

The current study has further characterised the effects of lansoprazole on the liver and testis. Lansoprazole treated animals exhibited a significant increase in total microsomal CYP content and induction of hepatic CYPIA, CYP2B, CYP3A and CYP4A proteins, confirming the findings of the previous study (chapter 5). In the liver, no significant changes in CYP2C11, CYP2Ç12, CYP2C6, CYP2A1 or 17J3-HSD mRNA levels were observed in lansoprazole-treated animals compared to the control group. No significant changes in CYPlAl, CYP2C11, CYP2C13, CYP2C6, CYP2A1, lip-HSD, CYPllAl or StAR were detected in testis samples firom lansoprazole-treated animals compared to the control group. However, lansoprazole might alter the levels of the corresponding proteins by influencing processes other than gene transcription.

The effects of lansoprazole on the endocrine control of the testis were also studied. Lansoprazole-treated animals showed a trend towards lower plasma testosterone levels, but this did not reach statistical significance. No statistically significant changes in plasma LH or FSH levels were detected in lansoprazole-treated animals compared to the control group, consistent with previous findings (chapter 5). Interestingly, lansoprazole-treated animals exhibited a marked reduction in plasma prolactin levels, which might play a role in the mechanism of LCT induction by this compound. Lansoprazole treatment was associated with a reduction in intratesticular testosterone levels, which was probably mainly due to the ability of this compound to inhibit testosterone biosynthesis (Fort et al., 1995). The ability of lansoprazole to reduce the local concentration of testosterone might have important effects on the paracrine environment of the testis, which could play a role in the formation of LCTs.

196 Chapter 7

Effect of lansoprazole on gene expression in the pituitary gland 7.1 Introduction

Previous studies have shown that treatment of rats with lansoprazole produces a significant reduction in plasma and testicular testosterone concentrations but we have

failed to detect any significant changes in plasma LH levels (chapters 5 and 6 ). As previously discussed, this might be due to factors such as the timing of blood sampling relative to the final dose of drug and the marked interanimal variability in circulating hormone levels in rodents.

The aim of the current study was to further characterise the effects of lansoprazole treatment on the endocrine control of the testis. Firstly, plasma hormone levels (testosterone, LH, FSH and prolactin) were measured in blood samples collected approximately four hours after the final dose of lansoprazole. Fort et al. (1995) reported a significant reduction in serum testosterone levels and a significant increase in LH levels at four and five hours after the final dose of lansoprazole in Sprague Dawley rats that had been treated once daily with this compound (150 mg/kg/day) for four weeks. Secondly, the effect of lansoprazole treatment on LH and prolactin mRNA levels in the pituitary gland was studied.

7.2 Study design

Male Sprague Dawley rats were dosed once daily with lansoprazole (150 mg/kg/day) or vehicle (0.5% (w / v) CMC/ 0.1% (v / v) Tween 80) by oral gavage for 14 days. On day 14 (~ 4 to 5 hours after the final dose of lansoprazole), blood samples were collected and liver, testes and pituitary glands were removed. Liver and testes were weighed and then snap frozen in liquid nitrogen. Pituitary glands were immediately placed in to tubes containing Trizol® reagent for total RNA extraction. More detailed experimental details concerning animal treatment, sample collection and analytical methods used for this study can be found in chapter 2 .

198 7.3 Results

7.3.1 Body, liver and testes weights

Final body weights and relative liver and testes weights for control and drug-treated animals are shown in Table 7.1.

Table 7.1 : Effect of lansoprazole treatment on relative liver and testes weights

Control Lansoprazole-treated

Final body weight (g) 353.3 ±28.0 348.2 ±20.8

Relative liver weight 3.79 ±0.30 4.39 ±0.13 ** (g/lOOg body weight)

Relative testes weight 0.93 ± 0.07 0.83 ± 0.08 (g/lOOg body weight)

Results are expressed as mean ± SD for each group o f animals (n=5 animals per group). Statistical analysis was performed using a Student’s t-test where asterisks (*) represent: ** P<0.01.

There were no statistically significant differences in initial (data no shown) or final body weights between experimental groups. Relative liver weights were significantly higher in lansoprazole-treated animals compared to the control group, but no significant changes in absolute organ weights (data not shown) were observed. No significant changes in relative or absolute (data not shown) testes weights were observed between experimental groups.

7.3.2 Plasma hormone levels

Blood samples were collected (~ 4 to 5 hours postdose) and assayed for plasma testosterone, LH, FSH and prolactin levels by radioimmunoassay (Figures 7.1 and 7.2). Lansoprazole-treated animals showed a trend towards lower plasma testosterone and prolactin levels, but this did not reach statistical significance. No statistically significant changes in plasma FSH or LH levels were detected.

199 Figure 7.1: Plasma testosterone levels in control and lansoprazole-treated rats

k m #

Ccrtrd Lenscprazdo^rested

Each bar represents the mean ± SEM from control and lansoprazole-treated animals (n=5 animals per group).

Figure 7.2: Plasma prolactin, follicle-stimulating hormone and luteinising hormone levels in control and lansoprazole-treated rats.

□ Control

□ Lansoprazole- treated

y ■

Prolactin

Each bar represents the mean ± SEM from control and lansoprazole-treated animals (n=5 animals per group).

200 7.3.3 TaqMan analysis

Total RNA was extracted from each pituitary gland and then purified using the RNAqueous™-4PCR kit as described in 2.2.14. The quahty of each total RNA sample was assessed by separation using a 1% agarose gel. The samples were run alongside a standard curve that was constructed using material (brain RNA) for which the total RNA content had been accurately quantified (data not shown). This allowed the concentration of each total RNA sample to be estimated. Approximately Ipg of each total RNA sample was used for cDNA synthesis as described in 2.2.14.4.

Real-time PCR (TaqMan™) was used to study the expression of LH and prolactin in pituitary glands taken from control and lansoprazole-treated animals and the results are shown in Figure 7.3.

Figure 7.3: Fold changes in pituitary gland mRNA levels for LH and prolactin following treatment of rats with lansoprazole as determined by TaqMan

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

LH Prolactin

Each bar represents the mean ± SEM fold change in mRNA levels of each gene compared to the control group (n=5 animals per group).

201 7.4 Discussion

7.4.1 Final body and organ weights

Lansoprazole treatment had no statistically significant effects on body weight gain or final body weights but produced a significant increase in relative liver weights (116% of control) (P<0.01). This indicates that the increase in relative liver weight is not solely an artefact due to the ability of lansoprazole to reduce final body weights. Lansoprazole treatment had no significant effects on relative or absolute testes weights in the current study.

7.4.2 Plasma hormone levels

To further characterise the effects of lansoprazole on the endocrine control of the testis, plasma hormone levels were measured in blood samples collected approximately four hours after the final dose of drug. Lansoprazole-treated animals showed a trend towards lower plasma testosterone levels (58% of control), but this did not reach statistical significance. Failure to attain statistical significance was probably due to the marked interanimal variation in plasma testosterone levels and the' small numbers of animals used in this study. Fort et al. (1995) previously reported significant reductions in serum testosterone levels at 3.5, 4 and 5 hours after the final dose of lansoprazole in Sprague Dawley rats that had been treated once daily with lansoprazole (150 mg/kg/day) for four weeks. However, at 4.5 hours postdose the reduction in testosterone levels did not reach statistical significance due to the high degree of interanimal variability in serum hormone levels (Fort et al., 1995). The magnitude of the reduction in plasma testosterone levels at four hours postdose (58% of control) appeared to be slightly larger than that previously observed at 18 to 2 0 hours postdose ( 6 6 % and 80% of control in chapter 5 and 6 respectively), suggesting that homeostatic regulation might have been beginning to restore normal circulating hormone levels by the later timepoint.

Lansoprazole treatment had no statistically significant effects on plasma LH levels, consistent with findings from my previous studies (chapters 5 and 6 ). In contrast. Fort et al. (1995) reported significant increases in serum LH levels in lansoprazole-treated animals at 4 and 5 hours postdose. Lansoprazole-treated animals showed a trend towards lower plasma prolactin levels (79% of control), but this did not reach

202 statistical significance. No significant effects were observed on plasma FSH levels in the current study. These findings are consistent with data from previous studies

(chapters 5 and 6 ).

7.4.3 Effect of lansoprazole on gene expression in the pituitary gland

Real-time PCR (TaqMan™) was used to examine the effects of lansoprazole treatment on the expression of selected genes in the pituitary gland, namely LH and prolactin. LH was selected due to the central role of this hormone in the hypothalamic-pituitary- testis axis and the failure of previous studies to detect statistically significant changes in plasma LH levels in lansoprazole-treated animals. LH is a heterodimeric glycoprotein composed of an a-subunit that is common to all pituitary glycoproteins

(LH, FSH, TSH) and a p-subunit, which is responsible for the unique properties of this hormone (Pierce & Parsons, 1981). Separate genes encode the two subunits and for the current study TaqMan primers and probes were designed to recognise the hormone-specific P-subunit of LH. Lansoprazole treatment had no significant effect on LH mRNA levels in pituitary gland samples from this study, which was consistent with the lack of effect on plasma LH levels observed in these animals. In contrast. Fort et al. (1995) reported significant increases in pituitary and serum LH levels four hours after the final dose of lansoprazole in Sprague Dawley rats that had been treated once daily with lansoprazole (150 mg/kg/day) for four weeks.

Prolactin was selected as the second gene of interest due to the fact that lansoprazole has previously been shown to produce a marked reduction in plasma prolactin levels

(chapter 6 ), which might play a role in, the mechanism of LCT induction. Lansoprazole treatment had no significant effects on prolactin mRNA levels in pituitary gland samples. There was a trend towards lower plasma prolactin levels in the lansoprazole-treated group firom this study, but this did not reach statistical significance. LH and prolactin are both stored in secretory granules in the pituitary gland therefore short-term changes in hormone secretion might occur in the absence of changes in hormone biosynthesis.

203 7.4.4 Conclusions

The current study was designed to further characterise the effects of lansoprazole on the endocrine control of the testis. Lansoprazole treatment was associated with a significant increase in relative liver weight, consistent with findings from previous studies (chapter 5). Lansoprazole-treated animals showed a trend towards lower plasma testosterone and prolactin levels compared to the control group in blood samples collected approximately four hours after the final dose of drug. No statistically significant changes in plasma LH or FSH levels were detected in lansoprazole-treated animals, consistent with previous findings (chapter 5 and 6 ). No significant changes in LH or prolactin mRNA levels were observed in pituitary gland samples from lansoprazole-treated animals, therefore the decrease in plasma prolactin levels might have resulted from changes in hormone secretion and/or elimination.

204 Chapter 8

Effect of lansoprazole on the plasma clearance of ^'‘C-testosterone 8.1 Introduction

In previous studies, treatment of rats with lansoprazole induced the expression of several hepatic CYP P450 proteins from the CYPIA, 2B, 3A and 4A subfamilies

(chapter 5 and 6 ). This was associated with significant increases in the activity of several CYP-dependent pathways of testosterone metabolism and in the overall rate of testosterone hydroxylation catalysed by liver microsomes in vitro (chapter 5). The ability of lansoprazole to induce hepatic steroid metabolising enzymes might contribute to the reduction in plasma testosterone levels previously observed in drug- treated rats. The aim of the current study was to determine whether treatment of rats with lansoprazole is associated with an increase in the rate of /"^C-testosterone clearance from the plasma m vzvo.

8.2 Study design

Male Sprague Dawley rats were dosed once daily with lansoprazole (150 mg/kg/day) (n=4) or vehicle (0.5% (w / v) CMC/ 0.1% (v / v) Tween 80) (n=5) by oral gavage for 14 days. On day 15, each animal was administered 50pCi ^"^C-testosterone (0.254mg testosterone) as an intravenous bolus injection in to the right tail vein. This dose of ^"^C-testosterone was selected based on the limits of detection of the methods used to analyse plasma samples. Serial blood samples were collected at the following timepoints after the dose of ^"^C-testosterone: 5, 15, 20, 30, 40, 50, 60, 90, 120, 180 and 260 minutes. These sampling times were selected based on the half-life of testosterone in rats obtained from the published literature (Pirke et al., 1982; Wang & Bulbrook, 1967; Henderson et al., 1980). Blood samples were collected from the left tail vein except for the final sample (260 minutes) which was taken by cardiac puncture. More detailed experimental details concerning animal treatment, sample collection and analytical methods used for this study can be found in chapter 2 .

206 8.3 Results

8.3.1 Body and liver weights

Final body weights and relative liver weights for control and lansoprazole-treated animals are shown in Table 8 .1 .

Table 8.1 : Effect of lansoprazole treatment on relative liver weights

Control Lansoprazole-treated

Final body weight (g) 367.4 ±18.6 351.2 ±18.9

Relative liver weight 3.84 ±0.23 3.79 ±0.14 (g/lOOg body weight)

Results are expressed as mean ± SD for control (n=5) and lansoprazole-treated animals (n=4).

There were no statistically significant differences in initial (data not shown) or final body weights between experimental groups. No statistically significant differences in absolute (data not shown) or relative liver weights were observed between control and lansoprazole-treated animals.

8.3.2 Analysis of plasma samples by HPLC

Selected plasma samples were subjected to HPLC analysis to determine whether any metabolites of ^"^C-testosterone were present at any of the timepoints. Samples from one control animal (at every timepoint) and one lansoprazole-treated animal (at 5, 20 and 180 minutes) were analysed. Plasma samples were extracted using ethyl acetate (as described in 2.2.10.1) and 95pl of the reconstituted extract was injected on to the

HPLC column with 5pi of an authentic testosterone standard. Each chromatogram was then examined for the presence of peaks other than ^'^C-testosterone.

Representative HPLC profiles of plasma extracts from a control animal from this study are shown in Figure 8.1. For all plasma samples analysed, no peaks other than ^"^C-testosterone were identified by radiochemical detection.

207 Figure 8.1: Representative HPLC profiles of serial plasma samples collected following an intravenous injection of ^"^C-testosterone in a control rat

a) 5 minutes cprii Testosterone

AU 1 ■ tL J —i AAr/j. nIiJtt.. JX. . ».V-'AA .Mjiâr. AOK yiLju^Jj A

10 - 20 30 40 50 Time (mins) o b) 40 minutes g. Testosterone

10 20 40 50 Time (mins) c) 180 minutes o cpra

' Testosterone

Time (mins)

Each plasma sample (SOpl) was extracted using ethyl acetate (as described in 2.2.10.1) and 95pi o f the reconstituted extract was injected on to the HPLC column. Analytes were detected by radiochemical detection. Testosterone was identified by co-chromatography with an authentic standard injected on to the column with each sample (5pi) (detected by optical absorbance at 240nm (data not shown)).

208 8.3.3 Quantification of ^"^C-testosterone in plasma samples

HPLC analysis of plasma extracts confirmed that metabolites of ^"^C-testosterone were not present at detectable levels in plasma samples firom this study. It was therefore decided to quantify the amount of ^"^C-testosterone in plasma samples by liquid scintillation counting (LSC), which allowed a higher sample throughput compared to analysing the samples by HPLC. Plasma samples were extracted using ethyl acetate

(as described in 2 .2 .1 0 . 1 ) and aliquots of the ethyl acetate (i.e. extractable radioactivity) and aqueous phases (i.e. unextractable radioactivity) were counted by LSC.

The levels of ^"^C-testosterone present in the ethyl acetate phase (i.e. extractable radioactivity) following extraction of plasma samples firom control and lansoprazole- treated animals are shown in Figure 8.2. The mean plasma concentrations of ^"^C- testosterone were generally lower in the lansoprazole-treated group compared to the control group.

A representative profile of the levels of radioactivity remaining in the aqueous phase (i.e. unextractable radioactivity) following extraction of plasma samples is shown in Figure 8.3. In general, the absolute amount of radioactivity remaining in the postextraction aqueous phase reached a maximum between 30 and 40 minutes after the injection of ^"^C-testosterone and declined thereafter (Figure 8.3a). The unextractable proportion of the total plasma radioactivity increased with time eventually reaching a plateau (Figure 8.3b).

209 (!) ^ ■I+1 Ti­ ll C N 11 2 5 II "ô 0n "°0) (0 -î-î c c (0 0 ca (D ü - J iz l{ ü o i l coo i t s i 11

1 2 s ^ G S i 1 1

( /3 îi . 2 coi 7 3 I t3 T3 H H D O § e s I 1. CD a: E î i - ilA.E I h-«H a il g I— $— I hOH !1 4-, 7 3 1 o o i ü K34 ^II s § i

18000

2 14000

12000

8000

6000

2000

50 100 200

Time (mins) b) Unextractable percentage of total plasma radioactivity

35

150 200 250

Time (mins)

Serial blood samples were collected from a control animal following administration o f an intravenous dose of ''*C- testosterone (SOpCi). Plasma samples were extracted three times using ethyl acetate (as described in 2.2.10.1) and aliquots of the postextraction aqueous phase (i.e. unextractable radioactivity) were counted by LSC. Each data point represents the average value derived from duplicate extractions of each plasma sample.

211 8.3.4 Calculation of kinetic parameters

Plasma ^"^C-testosterone concentration-time curves were analysed by non- compartmental analysis (i.v. bolus model) using WinNonlin software. The non- compartmental approach does not require a specific compartmental model for the system and therefore enables the estimation of various pharmacokinetic parameters without making assumptions about any of the pharmacokinetic properties of the behaviour of the drug in the body.

Table 8.2 summarises the relevant kinetic parameters that were derived from the testosterone plasma concentration-time curves from control and lansoprazole-treated animals.

Table 8.2: ^^C-testosterone plasma elimination kinetics in control and lansoprazole- treated rats

Parameter Control Lansoprazole-treated

AUCjast 7499 ± 2492 4423 ±1418* (dpm/min/ml)

Plasma Clearance 46.7 ±18.3 76.4 ±19.7* (ml/min/kg)

3869 ±1234 5931 ± 1427* (ml/kg)

*‘^C-testosterone plasma concentration-time curves were analysed by non-compartmental analysis using WinNonlin software as described in 2.2.10.4, AUCiast, area under the plasma concentration-time curve from the time of dosing to the last measured concentration; Vg;, volume o f distribution at steady state. Results are expressed as mean ± SD from individual control and lansoprazole-treated animals (n=4 animals per group): Statistical analysis was performed using a Student’s t-test where asterisks (*) represent: * P<0,05.

212 8.4 Discussion

8.4.1 Analysis of plasma samples

Initially, selected plasma samples were extracted and the reconstituted extracts subjected to HPLC analysis to determine whether any metabolites of ^"^C-testosterone were present at any of the timepoints. For all plasma samples analysed, no peaks other than ^"^C-testosterone were identified by radiochemical detection. In contrast, some other groups have reported the presence of testosterone metabolites in plasma samples collected from rats following intravenous administration of ^H-testosterone (Henderson et al., 1980; Bruchovsky & Wilson, 1968; Hobbs et al., 1992). Henderson et al. (1980) reported that the percentage of the total extractable plasma radioactivity associated with the parent compound decreased from 6 8 % at one minute postdose to 23% at 64 minutes following intravenous administration of ^H-testosterone. Unfortunately, the identity of the remaining extractable radioactivity was not reported in this paper. In contrast, Hobbs et al. (1992) reported that testosterone constituted the majority (80%) of the total extractable plasma radioactivity with smaller amounts of dihydrotestosterone (9%) and estradiol (11%) being present in blood samples collected 45 minutes after administration of ^H-testosterone.

There are a number of factors that might explain the differences between data from the current study and that reported by other groups. Firstly, one crucial difference between the current study and much of the published literature is related to the radioisotope employed. ^H-testosterone is used in the majority of studies reported in the published literature due to factors such as the higher specific activity and relative cheapness of tritiated material. However, the major problem associated with the use of tritium is the possibility of loss of the isotope from the compound as a result of, metabolism and chemical exchange. In contrast, ^"^C-testosterone was used for the current study because the metabolic stability of this compound is more predictable. Consequently, the type of radioisotope employed might have an impact on the proportion of the total plasma radioactivity that is detected in association with the parent compound (i.e. testosterone). Secondly, animals were anaesthetised during the experiments conducted by Henderson et al. (1980) and Hobbs et al. (1992) whereas no anaesthetic was used in the current study. Treatment of rats with an anaesthetic might have significant qualitative and quantitative effects on the pathways of steroid

213 metabolism. Indeed, Tapper & Brown-Grant (1975) reported markedly lower

0 estradiol plasma clearance rates in female rats under ether anaesthesia compared to animals that had been anaesthetised with tribromoethanol.

Thirdly, each study used different methods to process and analyse plasma samples, which might have contributed to the differences in the data obtained. One possible explanation for the apparent absence of metabolites in plasma extracts from the current study is that the ethyl acetate extraction procedure might have been less efficient at recovering neutral transformation products of testosterone (e.g. dihydrotestosterone) compared to the parent compound. This would result in these metabolites largely remaining in the postextraction aqueous phase and consequently not being detected during HPLC of plasma extracts. In addition, Henderson et al. (1980) and Bruchovsky & Wilson (1968) used thin layer chromatography (TLC) followed by LSC to separate and quantify radiolabelled compounds present in plasma extracts. In contrast, Hobbs et al. (1992) analysed plasma extracts by HPLC. Interestingly, the two studies employing TLC yielded similar data with respect to the percentage of extractable plasma radioactivity that was associated with testosterone (10% parent at 60 minutes postdose (Bruchovsky & Wilson, 1968) compared to 23% parent at 64 minutes postdose (Henderson et al., 1980)). In contrast, Hobbs et al. (1992) reported that testosterone constituted a much larger proportion (80%) of the total extractable plasma radioactivity at 45 minutes postdose. In the current study, it cannot be completely ruled out that plasma extracts contained traces of metabolites that were below the limits of detection of the HPLC method. However, such trace amounts of metabolites would not be expected to contribute significantly to the total amount of radioactivity measured in the plasma extracts.

Following confirmation by HPLC that metabolites of ^"^C-testosterone were not present at detectable levels in plasma samples from this study, it was , decided to quantify the amount of ^"^C-testosterone present in plasma extracts by liquid scintillation counting (LSC). This method enabled a higher sample throughput compared to analysing samples by HPLC. Each plasma sample was extracted using ethyl acetate and the levels of extractable (i.e‘ ^"^C-testosterone) and unextractable radioactivity were measured by LSC.

214 8.4.2 Extractable radioactivity

8.4.2.1 ^'^C-Testosterone plasma concentration-time profiles

Plasma concentrations of extractable ^"^C-testosterone showed an exponential decline following intravenous administration to rats (Figure 8.2). When this data was plotted semi-logarithmically (data not shown) it became apparent that the decline in plasma ^"^C-testosterone concentrations was biphasic, indicating that testosterone is distributed to more than one pharmacokinetic “compartment” within the body. The biphasic plasma concentration decay curves suggest that testosterone is distributed in to two compartments: a central (rapidly equilibrating) and a peripheral (more slowly equilibrating) compartment. In addition to the systemic circulation, the central compartment might include tissues such as the liver, prostate and seminal vesicles as these tissues show rapid uptake (from 1 minute postdose) of testosterone following administration of the radiolabelled steroid to rats (Bruchovsky & Wilson, 1968; Lee et al., 1974). In a two-compartment model the first phase of the plasma concentration­ time plot (i.e. steeper slope) includes the distribution of drug from the central to the peripheral compartment and the second phase (i.e. shallower slope) represents elimination from the central compartment. In reality, both of these processes occur simultaneously and the relative rates of distribution and elimination determine what each phase represents. Following intravenous administration, testosterone is rapidly cleared from the systemic circulation therefore the first phase in the ^"^C-testosterone plasma concentration decay curve probably represents a significant proportion of the elimination of this steroid.

The two-compartment pattern of testosterone clearance observed in the current study is consistent with previously published literature for rats (Henderson et al., 1980; Wang & Bulbrook, 1967; Bruchovsky & Wilson, 1968). However, it is difficult to compare data between studies due to different experimental designs, particularly with respect to blood sampling regimens. In the current study, blood samples were collected at various timepoints between 5 and 260 minutes following the administration of ^"^C-testosterone. In contrast, Henderson et al. (1980) collected blood samples at 1, 2, 4, 8 , 16, 32 and 64 minutes following an intravenous injection

3 ' 3 of H-testosterone. This group reported a biphasic decline in plasma H-testosterone concentrations with a rapid decline in plasma radioactivity up to 4 minutes postdose

215 and a slower decline thereafter. These findings indicate that distribution may occur very rapidly following intravenous administration of testosterone and might therefore have occurred significantly before the first blood sample was collected in the current study (i.e. 5 minutes postdose). This example illustrates how experimental design can influence the number of compartments that are used to describe drug disposition profiles.

LSC provided a sensitive method for quantification of the radioactivity present in plasma samples. However, the levels of '"^C-testosterone present in plasma samples from lansoprazole-treated animals at the later timepoints (i.e. 180 and 260 minutes) were close to the limits of detection and therefore subject to greater analytical error. This might explain the apparent increase in plasma ^"^C-testosterone concentrations observed in some animals between 180 and 260 minutes. In addition, the final blood sample (260 minutes) was collected by a different route (cardiac puncture) compared to the other samples (tail vein), which may also have had some effect.

S.4.2.2 Kinetic analysis

Two different approaches can be used to estimate pharmacokinetic parameters from plasma drug concentration-time profiles, namely compartmental and non- compartmental analysis. For the current study the non-compartmental approach was used to analyse ^"^C-testosterone plasma concentration-time curves. Non- compartmental analysis does not require a specific compartmental model for the system and therefore enables the estimation of various pharmacokinetic parameters without making assumptions about any of the properties of the pharmacokinetic behaviour of the drug in the body. In addition, this approach does not require the use of complicated non-linear regression processes. Alternatively, a two-compartment model could have been applied to this data set based on the biphasic appearance of the semi-log ^"^C-testosterone plasma concentration-time profiles. However, this approach would not have allowed non-extrapolated area values to be calculated, which was necessary for this data set (see below).

Initial studies confirmed that for the current data it was most appropriate to calculate pharmacokinetic parameters using non-extrapolated area values (e.g. AUChst). This

216 was necessary because extrapolation of the terminal portion of the ^"^C-testosterone plasma concentration-time curves tended to lead to overestimation of the AUChf. This was probably due to the fact that the last few data points were subject to greater analytical error due to the low levels of ^"^C-testosterone present in plasma samples collected at the later timepoints, particularly in the lansoprazole-treated group.

Three pharmacokinetic parameters were estimated from the ^"^C-testosterone plasma concentration-time profiles, namely AUCiast, clearance and Vgg. The area under the plasma concentration-time curve (AUCiast) is an important pharmacokinetic parameter in quantifying the extent of drug exposure and clearance. Plasma clearance is defined as the volume of plasma cleared of drug per unit time and is a measure of the efficiency of drug elimination from the body. The AUC is inversely related to the extent of systemic clearance. In the current study, lansoprazole-treated animals showed a significant reduction in AUCiast and a significant increase in plasma clearance compared to control animals following intravenous administration of ^"^C-testosterone (P<0.05). The plasma clearance calculated for control animals from the current study is consistent with previously published data for male Sprague Dawley rats (Wang & Bulbrook, 1967).

The volume of distribution at steady state (Vss,) represents the ratio between the amount of drug in the body and its plasma concentration once distribution equilibrium is achieved between the plasma and tissues. This parameter provides a measure of the extent of drug distribution from the plasma in to the tissues and is influenced by factors such as the extent of binding to blood (e.g. plasma proteins, blood cells) and tissue components. Two important observations were made concerning the Vss of ^"^C-testosterone in rats from the current study. Firstly, ^"^C-testosterone has a relatively large volume of distribution (Vss) in rats, indicating that this hormone is extensively distributed from the plasma in to the tissues and may become concentrated in certain tissues. This finding is consistent with the physiological roles of testosterone and the fact that this hormone has previously been reported to be rapidly taken up by tissues such as the liver, prostate and seminal vesicles in rats (Bruchovsky & Wilson, 1968; Lee et al., 1974).

217 Secondly, the Vss of ^"^C-testosterone was significantly higher in lansoprazole-treated animals compared to the control group (P<0.05). The reason for the observed increase in Vss following lansoprazole treatment is currently unknown but it could be due to some effect of this compound on the binding of ^"^C-testosterone to blood and tissue components. Lansoprazole might have an effect within the tissues leading to an increase in tissue binding or uptake of ^"^C-testosterone. Previous studies have demonstrated that this compound produces effects on the liver consistent with an increase in metabolic capacity (e.g. increase in relative organ weight, induction of cytochromes P450), which might be associated with enhanced hepatic uptake of testosterone. In addition, treatment of rats with lansoprazole has previously been shown to deplete endogenous testosterone levels (plasma and intratesticular). This might lead to increased uptake of ^"^C-testosterone by androgen-dependent tissues (e.g. testis, prostate) in an attempt to restore normal hormone levels. Lansoprazole could also have an effect on the extent of binding of ^"^C-testosterone to plasma proteins. For example, a reduction in the plasma albumin concentration might lead to a decrease in the amount of testosterone retained in the systemic circulation with a consequent increase in the volume of distribution. However, there is currently no evidence to suggest that treatment of rats with lansoprazole produces changes in plasma protein concentrations (Atkinson et al., 1990).

Finally, it is worthwhile to discuss the reasons why the terminal half-life of ^"^C-testosterone was not used to interpret data from the current study. Firstly, using the non-compartmental approach the terminal half-life of a drug is estimated using the equation below:

0.693 Terminal half-life =

The terminal elimination rate constant, Xz, is estimated by linear regression of the last few data points in the terminal portion of the plasma concentration-time curve, which were subject to greater analytical error as previously discussed. In addition, following intravenous administration the elimination of testosterone is an exponential process such that the rate of elimination is proportional to the concentration of testosterone at any given time (i.e. first order process). At the later timepoints, plasma ^"^C-testosterone concentrations were lower in the lansoprazole-treated animals

218 compared to the control group therefore the terminal elimination rate would also be expected to be slower. These factors would result in the generation of terminal half- life values that do not reflect the data. Finally, following intravenous administration testosterone is rapidly cleared from the systemic circulation therefore the first phase in the ^"^C-testosterone plasma concentration decay curve (the so-called “distribution” phase) probably represents a significant proportion of the elimination of this steroid, which would not be reflected in the terminal half-life value. Consequently, plasma clearance was considered to provide a more accurate index of testosterone elimination for the current data.

8.4.3 Unextractable radioactivity

The amount of radioactivity remaining in the aqueous phase following extraction of plasma samples increased rapidly until 30 to 40 minutes after the dose of testosterone, declining thereafter (Figure 8.2). This unextractable radioactivity constituted up to 40% of the total plasma radioactivity at the later timepoints (> 90 minutes). Failure of ethyl acetate to extract all of the radioactivity from plasma samples was probably mainly due the presence of steroid conjugates. In the liver, testosterone is conjugated with glucuronic acid or sulphate, either directly or subsequent to hydroxylation, in reactions catalysed by UDP-GT and ST enzymes respectively. These pathways produce more polar testosterone conjugates, facilitating excretion in the urine or bile. Extraction of plasma samples using ethyl acetate would be expected to yield all of the free testosterone whereas any conjugated steroids would remain in the aqueous phase. Indeed, when plasma samples were spiked with testosterone the extraction efficiency was approximately 95% using the current procedure (data not shown).

The observed decline in the extractable proportion of the total plasma radioactivity with time is consistent with the published literature (Bruchovsky & Wilson, 1968; Lee et al., 1974; Henderson et al., 1980). Bruchovsky & Wilson (1968) reported that following treatment of male Sprague Dawley rats with an intravenous injection of ^H-testosterone, the proportion of the plasma radioactivity that was recovered by chloroform-methanol extraction decreased with time in normal animals. In contrast, the entire plasma radioactivity was recovered in the chloroform-methanol phase at all

219 timepoints in functionally hepatectomised animals. These findings confirm that testosterone is rapidly metabolised by the liver to form polar derivatives that appear in the plasma.

In the current study no attempts were made to identify the ^"^C-testosterone derivatives present in the postextraction aqueous phase. To determine whether the unextractable radioactivity was composed of testosterone conjugates, aliquots of the aqueous phase could have been incubated with p-glucuronidase/sulfatase, which would enzymatically . cleave glucuronyl or sulfate groups. These samples could then have been re-extracted using ethyl acetate and the amount of extractable and unextractable radioactivity quantified by LSC. Analysis of the ethyl acetate phase by HPLC would also enable identification of liberated ^"^C-testosterone and metabolites (e.g. hydroxytestosterones).

In the current study, it was not possible to conclude whether there were any significant differences in the levels of unextractable radioactivity present in plasma samples fi*om control and lansoprazole-treated animals. This was due to the fact the absolute levels of radioactivity measured in the postextraction aqueous phase might contain a relatively high degree of experimental error for a number of reasons. Firstly, the aqueous phase was contaminated with varying amounts of residual ethyl acetate, which was not removed when the samples were decanted. Secondly, the aqueous phase was not a homogenous solution due to the presence of precipitated plasma proteins, meaning that it was difficult to pipette this phase consistently. Finally, due to the small volume of the aqueous phase it was not possible to count replicate samples by LSC. Due to these factors this data could only be used to identify trends in the levels of radioactivity measured in the postextraction aqueous phase during the timecourse of this study.

220 8.4.4 Conclusions

In the current study, lansoprazole-treated rats exhibited a significantly smaller AUCiast and significantly higher plasma clearance and volume of distribution (Vgg) following intravenous administration of ^"^C-testosterone (P<0.05). These findings indicate that lansoprazole increases the capacity of the body to eliminate testosterone firom the systemic circulation, consistent with our hypothesis. However, further studies need to be conducted to confirm that this increase in ^"^C-testosterone plasma clearance is a consequence of increased activity of hepatic CYP-dependent pathways of testosterone metabolism. This could be investigated by performing a mass balance study, which would involve intravenous administration of ^"^C-testosterone followed by measurement of the rate of elimination of radioactivity in the urine and faeces. Firstly, this would confirm that the increase in plasma ^"^C-testosterone clearance is associated with an enhanced rate of elimination of testosterone and its metabolites in the urine and faeces. Furthermore, identification and quantification of testosterone metabolites present in the urine and/or faeces would provide further information regarding the quantitative and qualitative effects of lansoprazole treatment on pathways of testosterone metabolism in vivo. If these studies confirmed that lansoprazole treatment was associated with an increase in the rate of elimination of hydroxylated testosterone metabolites, this would provide convincing evidence that induction of hepatic cytochromes P450 was responsible for the increase in plasma ^"^C-testosterone clearance observed in the current study.

9

221 Chapter 9

Final Discussion 9.1 Introduction

Leydig cell hyperplasia and adenomas are frequently observed in rodents during chronic toxicity and carcinogenicity studies with new therapeutic agents. A diverse range of chemicals have been shown to increase the incidence of Leydig cell hyperplasia and adenomas in rodent bioassays, however the significance of this effect to humans is currently a matter of debate. A common mechanism of tumour induction appears to involve disturbance of the hypothalamic-pituitary-testis (HPT) axis at various points leading to a sustained increase in circulating LH levels. However, for many compounds their mode of tumour induction is currently unknown. Elucidation of a biochemical mechanistic explanation for this effect would aid human risk assessment.

A common property of chemicals that cause Leydig cell hyperplasia and tumours is that they induce hepatic cytochromes P450 (e.g. oxazepam, felbamate, lansoprazole and methyl ^er/-butyl ether (MTBE)) but it is currently unclear whether these two phenomena are causally related (Diwan et al., 1986; reviewed by Glue et al., 1997; Masubuchi et al., 1997a; Williams & Borghoff, 2000). The aim of the current project was to investigate the existence of a liver-testis axis wherein microsomal enzyme inducers enhance the metabolic clearance of testosterone causing a compensatory increase in LH secretion. It is widely accepted that chronic elevation of circulating LH levels is associated with the formation of Leydig cell hyperplasia and tumours in rodents (Christensen & Peacock, 1980; Chatini et al., 1990; reviewed by Cook et al., 1999). Lansoprazole was selected as the model compound for studies designed to investigate the existence of a liver-testis axis because it induces hepatic cytochromes P450 and chronic exposure produces Leydig cell hyperplasia and tumours in rats (Masubuchi et al., 1997a; Atkinson et al., 1990; Unpublished report by Takeda Chemical Industries Ltd., A-29-681). A series of experiments were conducted to investigate the effects of lansoprazole on the liver, testis and the endocrine control of the testis.

9.2 Effects of lansoprazole on the liver

Treatment of rats with lansoprazole produced effects on the liver consistent with an enhanced metabolic capacity, including significant increases in relative liver weight and total microsomal CYP content. Lansoprazole induced the expression of several

223 hepatic CYP proteins, including CYPlAl, CYP1A2, CYP2B1, CYP3A and CYP4A1, as determined by ELISA and Western blotting. This was associated with significant increases in the activity of several CYP-dependent pathways of testosterone metabolism (i.e. formation of 6(3-, 7a-, 16a-, 16(3-, 2a- and 2P-0HT and androstenedione) and a significant increase in the total rate of hydroxytestosterone metabolite formation catalysed by liver microsomes in vitro. Such effects on the pathways of testosterone metabolism indicate that lansoprazole might also induce enzymes belonging to the CYP2A (7a-hydroxylase activity) and CYP2C (2a- hydroxylase activity) subfamilies. Lansoprazole treatment was not associated with significant changes in hepatic CYP2A1, CYP2C6, CYP2C11 or CYP2C13 mRNA levels compared to the control group, however this compound might increase the levels of the corresponding proteins by influencing processes other than gene transcription and mRNA stability.

The ability of lansoprazole to induce the expression of CYP proteins belonging to four families suggests that this compound may represent a distinct type of “mixed inducer” as no other published examples of chemicals with this property could be identified. Furthermore, the effects of lansoprazole treatment on microsomal testosterone metabolism were distinct from those observed following treatment of rats with the four model CYP inducers (P-NF, PB, PCN and ciprofibrate). Firstly, each of the model inducers exhibited differential effects on the pathways of hepatic testosterone metabolism, inducing the activity of certain pathways whilst suppressing or having no effect on other pathways. Secondly, the model inducers tended to exert marked effects on the activity of specific pathways (e.g. 1 2 fold increase in testosterone 6 p~ hydroxylase activity in PCN-treated animals). In contrast, lansoprazole treatment did not exhibit a large effect on any of the pathways studied (< 2 fold change) but rather produced a generalised increase in the activity of several pathways of testosterone oxidation. Consequently, the increase in the total rate of hydroxytestosterone metabolite formation produced by lansoprazole treatment (1.3 fold) was of a similar magnitude to that observed in PB- (1.3 fold) and PCN-treated (1.6 fold) animals.

Finally, a common finding in livers from rats treated with model CYP inducers is reduced expression of CYP2C11 protein and associated catalytic activities (Waxman

224 et al., 1985; Yeowell et al., 1987). Indeed, treatment of rats with each of the model inducers caused suppression of hepatic testosterone 2 a-hydroxylase activity, albeit to different extents. In contrast, treatment of rats with lansoprazole was associated with a significant increase in testosterone 2 a-hydroxylase activity, indicative of induction of CYP2C11 or another CYP2C form. CYP2C11 is the major constitutive CYP (~ 40% total hepatic CYP) found in male rat liver and this form probably makes a substantial contribution to the overall rate of testosterone oxidation by rat liver (Guengerich et al., 1982; Waxman, 1984). Furthermore, it has been postulated that CYP2C11 might be responsible for the formation of specific hydroxylated metabolites with important physiological functions (e.g. 7a-hydroxylated androgens may regulate testosterone biosynthesis) (Waxman, 1984; Inano et al., 1973; Inano & Tamaoki, 1971; Sunde et al., 1982; Rosness et al., 1977; Mittler, 1985). Consequently, xenobiotic-induced changes in the expression of CYP2C11 and associated testosterone hydroxylase activities might have an important impact on androgen homeostasis.

In conclusion, treatment of rats with lansoprazole was associated with a unique profile of hepatic CYP induction and enhanced CYP-dependent testosterone metabolism, consistent with the current hypothesis. The different effects of lansoprazole and the model inducers on hepatic testosterone hydroxylase activities might result in these compounds having diverse effects on testosterone elimination and ultimately, androgen homeostasis within the animal.

9.3 Effect of lansoprazole on the plasma clearance of ^"*C-testosterone

Following the observation that lansoprazole treatment enhanced the capacity of the 1 _ liver to metabolise testosterone in vitro, a study was conducted to determine whether this was associated with an increase in the plasma clearance of ^"^C-testosterone in vivo. Indeed, lansoprazole-treated rats exhibited a significantly smaller AUCiast and a significantly higher plasma clearance and volume of distribution (Vss) following the intravenous administration of '"^C-testosterone. These findings indicate that lansoprazole treatment increases the capacity of the body to eliminate testosterone from the systemic circulation, consistent with the current hypothesis. However, further studies are required to confirm that this increase in ^"^C-testosterone plasma clearance is a direct consequence of the enhanced activity of hepatic CYP-dependent pathways of testosterone metabolism (see Future Work).

225 9.4 Effects of lansoprazole on plasma hormone levels

9.4.1 Testosterone

Treatment of rats with lansoprazole was associated with a reduction in plasma testosterone levels. Circulating testosterone levels, reflect the balance between the rate of hormone biosynthesis in the testis and elimination via the liver and kidneys. The reduced plasma testosterone levels observed following lansoprazole treatment are probably, at least in part, related to the ability of this compound to inhibit testosterone biosynthesis (Fort et al., 1995). However, the ability of lansoprazole to enhance hepatic CYP-dependent testosterone metabolism and ^^C-testosterone plasma clearance suggests that enhanced metabolic clearance of testosterone might also play a role in lowering circulating testo.sterone levels. At present, the relative contributions of inhibited biosynthesis and enhanced metabolic clearance to the observed decrease in plasma testosterone levels are unknown.

In contrast, treatment of rats with the model CYP inducers did not have a significant impact on plasma testosterone levels despite the marked effects of these compounds on hepatic testosterone hydroxylase activities. Such differences between the effects of lansoprazole and the model inducers might be related to the divergent effects of these compounds on the pathways of testosterone hydroxylation or could be a consequence of effects on other pathways of testosterone metabolism (e.g. reductive pathways, glucuronidation, sulfation). Consistent with the latter hypothesis, treatment of humans with A-phenylbarbital produces a significant increase (> 60%) in the urinary excretion of polar testosterone metabolites but has no significant effect on the plasma clearance of testosterone (Bammel et al., 1992). Alternatively, initial effects of these compounds on circulating testosterone levels may be rapidly compensated for through changes in testosterone biosynthesis and/or secretion from the testis. Such homeostatic regulation might be sufficient to maintain normal circulating hormone levels despite the xenobiotic-induced changes in the metabolic clearance of testosterone.

To date, a limited number of studies have been performed to investigate the impact,of microsomal enzyme inducers on androgen homeostasis in rodents. Wilson & LeBlanc (1998) examined the effects of the organochloride pesticide, endosulfan, on steroid

226 metabolism in female CD-I mice. Endosulfan treatment increased the total rate of hydroxytestosterone metabolite formation catalysed by liver microsomes and enhanced the urinary elimination of ^"^C-testosterone (-3.6 fold). No significant changes in the rate of direct glucuronic acid or sulfate conjugation to testosterone were observed in endosulfan-treated mice (Wilson & LeBlanc, 1998). However, this marked increase in androgen elimination was associated with a small, non-significant decrease in serum testosterone levels, suggesting that homeostatic feedback mechanisms were able to compensate for the effects of this compound in female mice (Wilson & LeBlanc, 1998). Wilson & LeBlanc (1998) postulated that xenobiotic- induced increases in testosterone biotrarisformation and elimination might not be sufficient to perturb androgen homeostasis within the animal. However, simultaneous exposure to a compound that enhances testosterone elimination (e.g. endosulfan) and an agent that interferes with the feedback regulation of testosterone biosynthesis (e.g. ketoconazole) might have a significant impact on circulating testosterone levels. Accordingly, the reduced plasma testosterone levels observed in lansoprazole-treated animals might be a consequence of the combined ability of this compound disrupt the HPT axis at more than one site through inhibition of testosterone biosynthesis and enhanced testosterone clearance. A similar hypothesis has been proposed to explain the reduction in circulating testosterone levels observed following chronic exposure of animals to PCBs, which also enhance hepatic testosterone hydroxylase activity and inhibit testosterone biosynthesis (Machala et al., 1998). ^

9.4.2 Luteinising hormone (LH)

In the current project, no statistically significant changes in plasma LH levels were detected despite the reduction in circulating testosterone levels in lansoprazole-treated ' animals. Failure to detect a statistically significant increase in plasma LH levels in lansoprazole-treated animals might reflect the true situation, however subtle changes might not have been detected due to factors such as the timing of blood sample collection and interanimal variability in plasma hormone levels. Plasma LH levels measured at a single timepoint will be subject to a high degree of variability due to the pulsatile pattern of LH secretion from the pituitary gland (Ellis & Desjardins, 1982). Indeed, Wilson et al. (1999) postulated that the marked interanimal variability in circulating hormone levels observed in rodent models might result in subtle

227 disturbances of steroid hormone homeostasis remaining undetected until more profound irreversible effects are manifested. The studies conducted during this project were not primarily designed to detect subtle changes in plasma hormone levels therefore it would be beneficial to perform a specialised study to further characterise the effects of lansoprazole on circulating hormone levels (see Future Work).

In contrast to the current findings, Fort et al. (1995) provided evidence to support the involvement of LH in Leydig cell tumorigenesis following treatment of rats with lansoprazole. This group reported statistically significant increases in serum LH levels in blood samples collected four and five hours after the final dose from rats treated with lansoprazole (150mg/kg/day) for four weeks (Fort et al., 1995). Furtherrhore, providing testosterone supplementation to lansoprazole-treated Fischer 344 rats lowered serum LH levels and completely suppressed the induction of LCTs (Fort et al., 1995). These findings indicate that increased LH secretion, secondary to reduced negative feedback of testosterone at the hypothalamus and pituitary gland, might be responsible for the induction of LCTs by lansoprazole.

Fort et al. (1995) proposed a compensation-decompensation mechanism to explain the sequence of events leading to the formation of LCTs in lansoprazole-treated rats. Lansoprazole-treated animals exhibit significant reductions in circulating testosterone levels in short-term studies (< 1 month), however no significant changes are observed in longer-term studies (> 3 months) (Fort et al., 1995; Unpublished reports by Takeda Chemical Industries Ltd., A-90-151; A-29-I863). Leydig cell hyperplasia and adenomas are observed following approximately one year of lansoprazole treatment (Atkinson et al., 1990). Fort et al. (1995) postulated that Leydig cells might initially be able to compensate for the partial inhibition of testosterone biosynthesis produced by lansoprazole treatment through increasing the net responsiveness of the Leydig cell population to LH stimulation (i.e. compensation). Indeed, repeated treatment of rats with low levels of LH enhances LH-stimulated testosterone biosynthesis in vivo, despite downregulation of LH receptor numbers (Zipf et al., 1978b). This compensation process might involve changes in LH receptor signalling, steroidogenic enzyme induction or recruitment of an initially unresponsive population of Leydig cells in to the responsive pool (Lin et al., 1982; Payne et al., 1980a,b). Normal

.228 deterioration of these homeostatic mechanisms during ageing would result in loss of the ability to maintain circulating testosterone concentrations without an increase in Leydig cell mass, resulting in the formation of Leydig cell hyperplasia and tumours (i.e. decompensation). This compensation-decompensation model might explain the apparent absence of elevated plasma LH levels in lansoprazole-treated animals from the current project. Homeostatic regulation might have meant that plasma testosterone and LH levels were returning towards their normal ranges due to an increase in the steroidogenic responsiveness of the Leydig cell population to LH stimulation. However, no significant changes in StAR or CYP17A1 mRNA levels were observed in testis samples from lansoprazole-treated animals compared to the control group. In addition, the duration of lansoprazole treatment ( 2 weeks) in the current studies might not have been sufficient to attain this apparent homeostatic state. Indeed, Fort et al. (1995) reported that rats treated with lansoprazole (50 or 150mg/kg/day) for one or two weeks exhibited reductions in hCG-stimulated testosterone secretion m vivo (Fort et al., 1995).

In conclusion, failure to detect a significant increase in plasma LH levels in lansoprazole-treated animals from the current project might be a consequence of the inherent variability in circulating hormone levels or the result of adaptive responses occurring at the level of the testis. There is evidence in the published literature to support the involvement of LH in the mechanism of LCT induction by lansoprazole, therefore further studies are required to characterise the effects of lansoprazole treatment on plasrna LH levels (Fort et al., 1995).

9.4.3 Other hormones

The effects of lansoprazole treatment on other hormones involved in the endocrine control of the testis were examined. Treatment of rats with lansoprazole did not have a significant impact on plasma FSH levels but was associated with a reduction in plasma prolactin levels (reduced to 32% control values in study 3). No significant changes in prolactin mRNA levels were detected in pituitary gland samples from lansoprazole-treated animals, therefore the decrease in plasma prolactin levels might have resulted from changes in hormone secretion and/or elimination.

229 The ability of lansoprazole to reduce circulating prolactin levels might play a role in the mechanism of LCT induction by this compound. Prolactin appears to play an important role in , regulation of the male reproductive system through direct (e.g. maintenance of Leydig cell LH receptors) and indirect effects (e.g. suppression of gonadotropin secretion from the pituitary gland) on testicular function (Zipf et al., 1978a; McNeilly et al., 1978; Winters & Loriaux, 1978; Smith & Bartke, 1987; Cheung, 1983). Consequently, changes in plasma prolactin levels might lead to changes in circulating gonadotropin levels and modulation of the sensitivity of the Leydig cells to LH stimulation. Indeed, prolactin has previously been implicated in the aetiology of LCT formation in rats. Dopamine agonists, such as mesulergine, might produce LCTs through their ability to reduce circulating prolactin levels leading to downregulation of LH receptors (Prentice et al., 1992). This might result in a partial inhibition of testosterone biosynthesis and a consequent increase in LH secretion (Prentice et al., 1992). Furthermore, induction of chronic hyperprolactinaemia has been reported to markedly, reduce the incidence of spontaneous LCTs in Fischer 344 rats and the authors postulated that this was due to suppression of circulating LH levels (Bartke et al., 1985). It is currently unknown whether the ability of lansoprazole to reduce plasma prolactin levels contributes to the mechanism of LCT induction in rats.

9.5 Effects of lansoprazole on the testis

The main objective of the current project was to investigate the effects of lansoprazole treatment on the liver and the endocrine control of the testis. In addition, studies were performed to characterise the effects of lansoprazole treatment on testicular CYP expression, testosterone metabolism and intratesticular testosterone concentrations. The contribution of the testis to the overall rate of testosterone metabolism is probably relatively small, but the local concentration of hydroxylated testosterone metabolites might be important for testicular androgen homeostasis. For example, there is evidence to support a role for 7a-hydroxylated androgens in the regulation of testicular testosterone biosynthesis (Inano et al., 1973; Inano & Tamaoki, 1971; Sunde et al., 1982; Rosness et al., 1977; Mittler, 1985). Consequently, xenobiotic-induced changes in testicular testosterone metabolism might have important local effects on the testis.

230 Cytochromes P450 are expressed at low levels in the testis hence Western blotting was not sensitive enough to detect GYP proteins other than CYP4A1 in pooled testis microsomes. Consequently, real-time PCR (Taqman) was employed as sensitive and specific method to examine the effects of lansoprazole treatment on the expression of CYPlAl, CYP2A1, CYP2C11, CYP2C6, CYP2C13, IJP-HSD, CYP17A1 and StAR in the testis. Each mRNA was present at detectable levels in testis samples but no significant changes in gene expression were detected in lansoprazole-treated animals compared to the control group.

Experiments were performed to characterise the effects of lansoprazole treatment on testicular testosterone metabolism. However, the low levels of testosterone hydroxylase activity present in testis microsomes confounded the accurate quantification of testosterone metabolites present in incubation extracts. Consequently, the derived enzyme activities were less reliable and subtle effects of drug treatment might not have been detected due to the use of pooled testis microsomes. Treatment of rats with lansoprazole did not exert any marked effects on the pathways of testicular testosterone metabolism studied, consistent with the absence of changes in testicular CYP mRNA levels. This was not entirely unexpected as lansoprazole treatment produced relatively modest effects on the activity of CYP- dependent pathways of testosterone metabolism in the liver and testicular CYPs appear to be less sensitive to the effects of chemical inducers (Goldstein & Linko, 1984; Omiecinski, 1986).

The most pronounced testicular effect of lansoprazole treatment was a trend towards lower intratesticular testosterone concentrations in drug-treated animals compared to the control group (72% of control value). Measurement of intratesticular hormone levels provides an indicator of the biosynthesis, secretion and bioavailability of testosterone within the testis. Fort et al. (1995) reported that lansoprazole has direct effects on the rat testis, inhibiting testosterone biosynthesis by cultured Leydig cells and reducing hCG-stimulated testosterone secretion in vivo. Hence, the reduction in intratesticular testosterone levels in drug-treated animals is probably mainly due to the ability of lansoprazole to inhibit testosterone biosynthesis.

231 The ability of lansoprazole to reduce the local concentration of testosterone might have important effects on the testis. Testosterone is essential for the maintenance of normal spermatogenesis therefore sustained reductions in intratesticular hormone levels are probably responsible for the increased incidences of epididymal hypospermia and seminiferous tubule atrophy reported in lansoprazole-treated rats (Unpublished reports by Takeda Chemical Industries Ltd., Report numbers: TÀ-90- 151 and TA-91-024). Such effects might influence Leydig cell morphology and function through disturbance of the reciprocal regulatory pathways that appear to exist between the seminiferous tubules and the Leydig cells. In addition, testosterone may modulate the production of several potential regulatory molecules that are produced within the testis (e.g. PModS, p-endorphin, oxytocin). Consequently, the ability of lansoprazole to reduce intratesticular testosterone concentrations might have important effects on the paracrine environment of the testis, which could play a role in the formation of LCTs. .

232 9.6 Conclusions

The current project has generated data that significantly increases our current understanding of the effects of lansoprazole on the liver and testis in male rats. Lansoprazole treatment induced hepatic CYP proteins belonging to four families, suggesting that this compound might represent a distinct type of “mixed inducer”. Lansoprazole-treated rats exhibited significant increases in hepatic CYP-dependent testosterone metabolism in vitro and enhanced plasma clearance of ^"^C-testosterone in vivo. These findings demonstrate that lansoprazole treatment increases the capacity of the body to eliminate testosterone from the systemic circulation, however further studies are required to confirm that this effect was a direct consequence of enhanced hepatic CYP-dependent testosterone metabolism. Reductions in plasma testosterone levels were observed in lansoprazole-treated animals compared to the control group, confirming that this compound has the potential to perturb androgen homeostasis within the animal. This effect has previously been attributed to the ability of lansoprazole to inhibit testosterone biosynthesis, but data from the current project suggests that enhanced clearance might also play a role in lowering circulating testosterone levels. At present, the relative contribution of these two mechanisms to the depletion of circulating testosterone is unknown.

Despite the reduction in circulating testosterone levels, no significant changes in plasma LH levels were detected in lansoprazole-treated animals from the current project. However, there is evidence in the published literature to support the involvement of elevated LH levels in the mechanism of LCT induction by lansoprazole (Fort et al., 1995). As the current studies were not primarily designed to detect subtle changes in plasma hormone levels, it would be beneficial to perform a specialised study to further characterise the effects of lansoprazole on plasma LH levels.

In conclusion, the current project has generated data to support the hypothesis that hepatic CYP induction contributes to the reduction in circulating testosterone levels observed in lansoprazole-treated rats. Hence, this effect might play a role in the mechanism by which lansoprazole interferes with the HPT axis and produces Leydig cell hyperplasia and tumours in rats. These findings are not inconsistent with the

233 existence of a liver-testis axis, however further studies are required to determine whether enhanced metabolic clearance of testosterone alone is sufficient to perturb androgen homeostasis in rats.

9.7 Future Work

The current project has generated data to support the hypothesis that hepatic CYP induction contributes to the reduction in circulating testosterone levels observed in lansoprazole-treated rats. The following studies would provide additional information concerning the mechanism of LCT induction in lansoprazole-treated animals.

The following experiments would be of interest to further characterise the effects of lansoprazole on testosterone biotransformation and elimination processes and the endocrine control of the testis.

a) Mass balance study This study would confirm whether the increase in plasma ^"^C-testosterone clearance observed in lansoprazole-treated animals is a, direct consequence of enhanced hepatic CYP-dependent testosterone metabolism. The elimination of radiolabelled testosterone would be measured in the urine and faeces following administration of an intravenous injection of ^"^C-testosterone to control and lansoprazole-treated rats. Identification and quantification of testosterone metabolites present in the urine and faeces would provide further information concerning the quantitative and qualitative effects of lansoprazole treatment on the pathways of testosterone metabolism in vivo. If enhanced elimination of hydroxylated testosterone metabolites was observed in the lansoprazole-treated group, this would provide convincing evidence that hepatic CYP induction was responsible for the increase in plasma ^"^C-testosterone clearance.

b) Other pathways of testosterone metabolism It would be of interest to examine the effects of lansoprazole treatment on other pathways that contribute to testosterone biotransformation and elimination (e.g. reductive pathways, glucuronidation, sulfation). These studies could be performed in vitro using liver microsomal or cytosolic fractions prepared from control and lansoprazole-treated animals.

234 c) Endocrine study It would be beneficial to perform a well-designed study to further characterise the effects of lansoprazole treatment on the endocrine control of the testis, particularly plasma LH levels. The design of these studies might include larger numbers of animals, pair-matched controls (i.e. blood samples collected at exactly the same time) and collection of serial blood samples from each animal.

In addition, other potentially interesting endocrine changes were identified in lansoprazole-treated animals from the current project. It would be of interest to further investigate whether these effects play a role in the mechanism of LCT induction by lansoprazole as follows. d) Decrease in plasma prolactin levels There is some evidence that changes in circulating prolactin levels might play a role in aetiology of LCT formation in rats, therefore it would be of interest to further investigate the significance of the reduction in plasma prolactin levels in lansoprazole-treated animals (Prentice et al., 1992; Bartke et al., 1985). For example, a prolactin supplementation study could be conducted to investigate the effects of the reduction in plasma prolactin levels on plasma hormone levels (e.g. LH, testosterone), the sensitivity of the Leydig cells to LH-stimulation and the incidence of Leydig cell hyperplasia and tumours in lansoprazole-treated animals. e) Reduction in intratesticular testosterone levels It would be interesting to investigate whether the reduced intratesticular testosterone concentrations in lansopraozle-treated animals are associated with changes in the paracrine environment of the testis. This might include comparing the expression of various paracrine factors in testis samples from control and

I ' ' . lansoprazole-treated animals (e.g. using real-time PCR (TaqMan)). The significance of any changes could be further investigated by examining the effects of the paracrine factor on Leydig cell numbers and function.

Finally, to further investigate the existence of a liver-testis axis, it would be beneficial to treat rats with a series of compounds known to induce hepatic cytochromes P450 and produce Leydig cell tumours in rodents. The effects of these compounds on

235 hepatic CYP proteins, microsomal testosterone metabolism, ^"^C-testosterone plasma clearance and elimination in vivo and circulating hormone levels could be examined. Such studies might enable the identification of a subset of compounds that induce LCTs as a consequence of enhanced metabolic clearance of testosterone. In addition, studying the quantitative and qualitative effects of these compounds on the pathways of testosterone metabolism and elimination might indicate whether compounds that induce LCTs alter specific pathways or metabolites.

236 References AHMAD, N., HALTMEYER, G.C. & EIK-NES, K.B. (1975). Maintenance o f spermatogenesis with testosterone or dihydrotestosterone in hypophysectomized rats. J. Reprod.Fertil, 44(1), 103-107.

AKINOLA, L.A., POUTANEN, M. & VIHKO, R. (1996). Cloning o f rat 17p-hydroxysteroid dehydrogenase type 2 and characterisation of tissue distribution and catalytic activity of rat type 1 and type 2 enzymes. Endocrinol., 137(5), 1572-1579.

AMACHER, D.E. & SCHOMAKER, S J. (1998). Ethinyl morphine N-demethylase activity as a marker for CYP3 A activity in rat hepatic microsomes. Toxicol.Letters, 94, 115-125.

AMADOR, A., STEGER, R.W., BARTKE, A., JOHNS, A., SILER-KHODR, T.M., PARKER, C.R & SHEPHERD, A.M. (1985). Testicular LH receptors during ageing in Fisher 344 rats. J.Androl, 6, 61- 64.

ANAKWE, 0 .0 & PAYNE, A.H. (1987). Noncoordinate regulation of de novo synthesis of cytochrome P450 cholesterol side-chain cleavage and cytochrome P450 17 alpha-hydroxylase/C17-20 lyase in mouse Leydig cell cultures. Mol.EndocrinoL, 1(9), 595-603.

ANDERSON, C.M. & MENDLESEN, C.R. (1985). Regulation of steroidogenesis in rat Leydig cells in culture: effects of human chorionic gonadotropin and dibutyryl cyclic AMP on the synthesis of cholesterol side chain cleavage cytochrome P450 and adrenodoxin. Arch.Biochem.Biophys., 238, 378- 387.

ANDERSSON, S. & MOGHRABI, N. (1997). Physiology and molecular genetics o f 17P- hydroxysteroid dehydrogenases. Steroids, 62(1), 143-147.

ANTHONY, C.T., KOVACS, W.J & SKINNER, M.K. (1989),. Analysis of the androgen receptor in isolated testicular cell types with a microassay that uses an affinity ligand. Endocrinol, 125(5), 2628- 2635.

AOYAMA, T., YAMANO, S., WAXMAN, D.J., LAPENSON, D.P., MEYER, U.A., FISCHER, V., TYNDALE, R., INABA, T., KALOW, W., GELBOIN, H.V. & GONZALEZ, F.J. (1989). Cytochrome P450 hPCN3, a novel cytochrome P450 IIIA gene product that is differentially expressed in adult human liver. 264(18), 10388-10395.

ARIMORI, K., YASUDA, K., KATSUKI, H. & NAKANO, M. (1998). Pharmacokinetic differences between lansoprazole enantiomers in rats. y.PAarw.P^armaco/., 50, 1241-1245.

ARLOTTO, M.P., TRANT, J.M. & ESTABROOK, R.W. (1991). Measurement of steroid hydroxylation reactions by high-performance liquid chromatography as indicator of P450 identity and function. Methods in Enzymology,2Q6, 454-462.

ARLOTTO, M.P., GREENWAY, D.J. & PARKINSON, A. (1989). Purification of two isozymes of rat liver microsomal cytochrome P450 with testosterone 7a-hydroxylase activity. Arch.Bioch.Biophys., 210(2), 441-457.

ATKINSON, I.E., DALY, I.W., BOLTE, H.F., MORISHIMA, H. & SASAKI, S. (1990). One-year oral gavage toxicity study of lansoprazole (AG-1749) in rats. Jpn.PharmacolTher (Suppl), 18, 59-91.

BAKER, J., HARDY, M.P., ZHOU, J., BONDY, C., LUPU, F., BELLVE, A.R. & EFSTRATIADIS, A. (1996). Effects of an IGF-1 gene null mutation on mouse reproduction. MolEndocrinoL, 10, 903- 918. ' .

BAMMEL, A , VAN DERMEE, K., OHNHAUS, E.E. & KIRCH, W. (1992). Divergent effects of different enzyme inducing agents on endogenous and exogenous testosterone. Eur.J.Cliri.Pharmacol., 42, 641-644.

BANDIERA, S., RYAN, D.E., LEVIN, W. & THOMAS, P.E. (1986). Age- and sex-related expression of cytochromes P450f and P450g in rat liver. Arch.Biochem.Biophys., 248(2), 658-676.

238 BARDIN, C.W., CHENG, c.Y., MUSTO, N.A. & GUNSALUS, G.L. (1994). The Sertoli Cell. In. The Physiology of Reproduction, eds. Knobil, E. & Neill, J.D. Raven Press. New York.

BARONI, C., MAGRINI, U., MARTINAZZI, M. & BERTOLI, G. (1966). Testicular Leydig cell tumorigenesis by diethylstilbestrol in the BALB/c mouse. Eur.J.Cancer, 7, 211-220.

BARTKE, A. (1980). Role of prolactin in reproduction in male mammals. Fed.Proc., 39, 2577-2581.

BARTKE, A , SWEENEY, C.A., JOHNSON, L., CASTRACANE, V.D. & DOHERTY, P.O. (1985). Hyperprolactinaemia inhibits development of Leydig cell tumours in aging Fischer rats. Exp. Aging Jîej., 11, 123-128.

BARTKE, A., HAFIEZ, A.A. & BEX, F.J., DALTERIO, S. (1978). Hormonal interactions in regulation of androgen secretion. Biol.Reprod., 18,44-54.

BARTKE, A , SMITH, M.S., MICHAEL, S.D., PERON, F.G. & DALTERIO, S. (1977). Effects of experimentally-induced chronic hyperprolactinaemia on testosterone and gonadotropin levels in male rats and mice. E'nt/ocrmo/., 100(1), 182-186.

BARTLETT, J.M., KERR, J.B. & SHARPE, R.M. (1986). The effect of selective destruction and regeneration of rat Leydig cells on the intratesticular distribution of testosterone and morphology of the seminiferous epithelium J.Androl., 7, 240-253.

BELL, D.R., BARS, R.G. & ELCOMBE, C.R. (1992). Differential tissue-specific expression and induction of cytochrome P450IVAI and acyl-CoA oxidase. Eur. J.Biochem., 206(3), 979-986.

BELPOGGI, F., SOFFRITTI, M. & MALTONI, C. (1995). Methyl-tertiary-butyl ether(MTBE)-a gasoline additive-causes testicular and lymphohaematopoietic cancers in rats. Toxicol.Ind.Health, 11, 119-149.

BENGTSSON, M., HALLBERG, E., GEORGELLIS, A. & RYDSTROM, J. (1990). On the identity of the xenobiotic-metabolising form(s) of cytochrome P450 in endocrine organs. Cancer Letters, 52, 235-241.

BERGH, A. (1982). Local differences in Leydig cell morphology in the adult rat testis: evidence for a local control of Leydig cells by adjacent seminiferous tubules. Int.J.AndroI., 5,325-330.

BICSAK, T.A., VALE, W., VAUGHAN, J., TUCKER, M., CAPPEL, S. & HSUEH, W. (1987). Hormonal regulation of inhibin production by cultured Sertoli cells. Mol. Cell.EndocrinoL, 49, 211- 217. '

BIEGEL, L.B., HURTT, M.E., FRAME, S.R., O’CONNOR, J.C. & COOK, J.C. (2001). Mechanisms of extrahepatic tumour induction by peroxisome proliferators in male CD rats. Toxicol.Sci., 60,44-55.

BIEGEL, L.B., LIU, R.C., HURTT, M.E. & COOK, J.C. (1995a). Effects of ammonium perfluorooctanoate on Leydig cell function. Toxicol.Appl.Pharmacol., 134, 18-25.

BIEGEL, L.B., COOK, J.C., O’CONNOR, J.C., ASCHIERO, M., ARDUENGO, A.J. & SLONE, T.W. (1995b). Subchronic toxicity study in rats with 1-methyl-3-propylimidazole-2-thione (PTI):effects on the thyroid. 27, 185-194.

BIEGEL, L.B., HURTT, M.E., FRAME, S.R., APPLEGATE, M., O’CONNOR, J.C. & COOK, J.C. (1992). Comparison of the effects of Wyeth-14,643 in CrhCD BR and Fisher-344 rats. Fundam.Appl.Toxicol., 19, 590-597.

BIRD, M.G., BURLEIGH-FLAYER, H.D., CHUN, J.S., DOUGLAS, J.F., KNEISS, J.J. & ANDREWS, L.S. (1997). Oncogenicity studies of inhaled methyl tertiary-butyl ether (MTBE) in CD- 1 mice and F344 rats. J.Appl. Toxicol., 17, S45-S55.

239 BRIMBLECOMBE, R.W. & LESLIE, G.B. (1984). Cimetidine. In. Testing of New Drugs, eds. Laurence, D.R., McLean, A.E. & Weatheral, M. Academic Press. London.

BROWN, G.E., WARREN, S., CHUTE, R.N., RYAN, K.J. & TODD, R.B. (1979). Hormonally induced tumours of the reproductive system of parabiosed male rats. Cancer Res., 39, 3971-3976.

BRUCHOVSKY, N. & WILSON, J.D. (1968). The conversion of testosterone to 5 a-androstan-17 P - ol-3-one by rat prostate m v/vo and m viYra. y.5ro/.C/iew., 243(8), 2012-2021.

BUTERS, J.T., SAKAI, S., RICHTER, T., PINEAU, T., ALEXANDER, D.L., SAVAS, U., DOEHMER, J., WARD, J.M., JEFCOATE, C.R. & GONZALEZ, F.J. (1999). Cytochrome P450 CYP IB 1 determines susceptibility to 7, 12-dimethylbenz(a)anthracene-induced lymphomas. Proc.Natl.Acad.Sci., 96(5), 1977-1982.

CANIVENC-LAVIER, M-C., VERNEVAUT, M-F., TOTIS, M., SIESS, M-H., MAGDALOU, J. & SUSCHETET, M. (1996). Comparative effects of flavenoids and model inducers on drug- metabolising enzymes in rat liver. Toxicol, 114, 19-27.

CARON, K.M., IKED A, Y., SOO, S.C., STOCCO, D.M., PARKER, K.L. & CLARK, B.J. (1997). Characterisation of the promoter region of the mouse gene encoding the steroidogenic acute regulatory protein. Mol Endocrinol, 11(2), 138-147.

CARTER, J.N., TYSON, I.E., TOLIS, G., VAN VLIET, S., FAIMAN, C. & FRIESEN, H.G. (1978). Prolactin-secreting tumours and hypogonadism in 22 men. New Eng. J. Med., 299, 847-852.

CHANDRASHEKAR, V., BARTKE, A. & WAGNER, T.E. (1991). Interactions of human growth hormone and prolactin on pituitary and Leydig cell function in adult transgenic mice expressing the human growth hormone gene. Biol Reprod., 44, 76-82.

CHATINI, F., NONOYAMA, T., SUDO, K., MIYAJIMA, H., TAKEYAMA, M., TAKATSUKA, D., MORI, H. & MATSUMOTO, K. (1990). Stimulatory effect o f luteinising hormone on the development and maintenance of 5-alpha-reduced steroid-producing testicular cell tumours in Fischer 344 TdXs. Anticancer Res.,10, 337-342.

CHEUNG, C.Y. (1983). Prolactin suppresses luteinising hormone secretion and pituitary responsiveness to luteinising hormone-releasing hormone by a direct action at the anterior pituitary. Endocrinol, 113(2), 632-638.

CHRISTENSEN, L.K. & STRAUSS, J.F. (2000). Steroidogenic acute regulatory protein (StAR) and the intramitochondrial translocation of cholesterol. Rioczw. yfc/n, 1529, 175-187.

CHRISTENSEN, A.K. & PEACOCK, K.C. (1980). Increase in Leydig cell numbers in testes o f adult rats treated chronically with an excess of human chorionic gonadotropin. BiolReprod., 22, 383-391.

CHRISTENSEN, A.K. & MASON, N.R. (1965). Comparative ability of seminiferous tubules and interstitial tissues of rat testes to synthesise androgens from progesterone-4-'"‘C in vitro. Endocrinol, 76, 646-656.

CLARK, B.J., SOO, S.C., CARON, K.M., IKEDA, Y., PARKER, K.L. & STOCCO, D.M. (1995). Hormonal and developmental regulation of the steroidogenic acute regulatory (StAR) protein. Mol. Endocrinol, 9, 1346-1355.

CLARK, B.J., WELLS, J., KING, S.R. & STOCCO, D.M. (1994). The purification, cloning, and expression of a novel luteinizing hormone-induced mitochondrial protein in M A-10 mouse Leydig tumor cells. Characterization of the steroidogenic acute regulatory protein (StAR). J.BiolChem., 269(45), 28314-28322.

CLAYTON, R.M. & HUHTANIEMI, L.T. (1982). Absence of gonadotropin-releasing hormone receptors in human gonadal tissue. Nature, 299, 56-59.

240 CLEGG, E.D., COOK, J.C., CHAPIN, G.P., FOSTER, P.M.D. & DASTON, G.P. (1997). Leydig cell hyperplasia and adenoma formation: mechanisms and relevance to humans. Reproductive Toxicol, 11(1), 107-121.

COOK, I.e., KLIENEFELTER, G.R., HARDISTY, J.F., SHARPE, R.M. & FOSTER, P.M.D. (1999). Rodent Leydig cell tumorigenesis: a review of the physiology, pathology, mechanisms and relevance to humans. Crit.Rev.Toxicol, 29{2), \69-26\.

COOK, I.e., HURTT, M.E., FRAME, S.R. & BIEGEL, L.B. (1994). Mechanisms of extrahepatic tumor induction by peroxisome proliferators in CrhCDBR rats. Toxicologist (Abstract), 14, 301.

COOK, I.e., MULLIN, L.S., FRAME, S.R. & BIEGEL, L.B. (1993). Investigation of a mechanism for Leydig cell tumorigenesis by linuron in rats. Toxicol Appl Pharmacol 119(2), 195-204.

COOK, I.e., MURRAY, S.M., FRAME, S.R. & HURTT, M. (1992). Induction o f Leydig cell adenornas by ammonium perfuorooctanoate: a possible endocrine-related mechanism. Toxicol Appl.Pharmacol, 113,209-217.

COOKE, B.A., LINDH, L.M. & JANSZEN, F.H.A. (1976). Correlation of protein kinase activation and testosterone production after stimulation of Leydig cells with luteinising hormone. Biochem. J., 160,439-446. '

COOPER, K.O., REIK, L.M., JAYYOSI, Z., BANDIERA, S., KELLEY, M., RYAN, D.E., DANIEL, R., McCLUSKEY, S.A., LEVIN, W. & THOMAS, P.E. (1993). Regulation of two members of the steroid-inducible cytochrome P450 subfamily (3A) in rats. Arch.Biochem.Biophys., 301(2), 345-354.

CREASEY, D.M. (1998). The Male Reproductive System. In. Target Organ Pathology, ezfe. Turton, J. & Hooson, J. Taylor & Francis. London.

CRIVELLO, J.F. & JEFCOATE, C.R. (1980). Intracellular movement of cholesterol in rat adrenal cells: kinetics and effects of inhibitors. J.5zo/. C^em., 255, 8144-8151.

CURI-PEDROSA, R., DAUJAT, M., PICHARD, L., OURLIN, J.C., CLAIR, P., GERVOT, L., LESCA, P., DOMERGUE, J., JOYEUX, H., FOURTANIER, G. & MAUREL, P. (1994). Omeprazole and Lansoprazole are mixed inducers of CYPIA and CYP3A in human hepatocytes in primary culture. y.P^arwaco/.Exj^r.r/zer., 269, 384-392.

CURRAN, P.G. & DEGROOT, L.J. (1991). The effect o f hepatic enzyme-inducing drugs on thyroid hormones and the thyroid gland. i?ev., 12(2), 135-150.

CZERWIEC, F.S., MELNER, M.H. & PÜETT, D. (1989). Transiently elevated levels of c-fos and c- myc oncogene messenger ribonucleic acids in cultured murine Leydig tumor cells after addition of human chorionic gonadotropin. A/b/.F/îz/ocrwo/., 3, 105-109.

DARNEY, K.J., ZIRKJN, B.R. & EWING, L.L. (1996). Testosterone autoregulation of its biosynthesis in the rat testis: inhibition of 17-hydroxylase activity. J.Androl, 17, 137-142.

DAUGHADAY, W. & ROTWEIN, P. (1989). Insulin-like growth factors I and II. Peptide, messenger ribonucleic acid and gene stmcture and tissue concentration. Endocrin.Rev., 10, 68-91.

DEBRI, K., BOOBIS, A.R., DAVIES, D.S. & EDWARDS, R.J. (1995). Distribution and induction of CYP3A1 and 3A2 in rat liver and extrahepatic tissues. Biochem.Pharmacol, 50(12), 2047-2056.

DE HERDER, W.W., BONTHUIS, F., RUTGERS, M., OTTEN, M.H., HAZENBERG, M.P. & VISSER, T.J. (1988). Effects of inhibition of type I iodothyronine deiodinase and phenol sulfotransferase on the biliary clearance of triiodothyronine in rats. Endocrinol, 122, 153-157.

241 DE KRUIF, C.A., MARSMAN, J.W., VENEKAMP, J.C., FALKE, H.E., NOORDHOEK, J., BLAAUBOER, B J. & WORTELBOER, H.M. (1991). Stmcture elucidation of acid reaction products of indole-3-carbinol: detection and in vivo and enzyme induction in vitro. Chem.Biol.Int., 80(3), 303- 315.

DIBASIO, K.W., SILVA, M.H., SHULL, L.R., OVERSTREET, J.W., HAMMOCK, B.D. & MILLER, M.G. (1991). Xenobiotic metabolising activities in rat, mouse, monkey and human testes. DrugMetab.Disp., 19{\), 227-232.

DIRAMI, G., TEERDS, K.J. & COOKE, B.A. (1996). Effect of a dopamine agonist on the development of Leydig cell hyperplasia in Sprague-Dawley rats. Toxicol. Appl. Pharmacol, 141(1), 169-177.

DIWAN, B.A., RICE, J.M. & WARD, J.M. (1986). Tumor-promoting activity o f benzodiazepine tranquillisers, diazepam and oxazepam, in mouse liver. Carcinogenesis, 7(5), 789-794.

DONAUBAUER, H.H., KRAMER, M., KRIEG, K., MAYER, D., VON RECHENBERG, W., SANDOW, J. & SCHÜTZ, E. (1987). Investigations of the carcinogenicity of the LHRH analog buserelin (HOE 766) in rats using the subcutaneous route of administration. Fundam. Appl Toxicol, 9(4), 738-752.

DRAPER, A.J., MAD AN, A., SMITH, K. & PARKINSON, A. (1998). Development of a non-high pressure liquid chromatography assay to determine testosterone hydroxylase (CYP3A) activity in human liver microsomes. D w/7., 26(4), 299-304.

DUFFY, O., BERTHOU, F., BARDOU, L.G. & MENEZ, J-F. (1995). Effects o f various cytochrome P450 inducers on testosterone metabolism and several monooxygenase activities in Sprague Dawley (SpD), high alcohol sensitivity (HAS) and low alcohol sensitivity (LAS) rats. Alcohol & Alcoholism, 30(3), 329-335.

DUNN, J.F., NISULA, B.C., & RODBARD, D. (1981). Transport of steroid hormones; binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma. J.Clin.Endocrinol.Metab., 52, 5S-68.

DYM, M., RAJ, H.G., LIN, Y.C., CHERNES, H.E., KOTITE, N.J., NAYFEH, S.N. & FRENCH, F.S. (1979). Is FSH required for maintenance of spermatogenesis in adult rats? J.Reprod. Fertil, Suppl.., 26,175-181.

EDWARDS, R.J., FREELING, A.B., WATSON, D. & DAVIES, D. (1992). The effect o f budesonide and triamcinolone acetonise on hepatic microsomal testosterone metabolism in the rat. Biochem.Pharmacol, 43(2), 271-282.

EECHAUTE, W., LACROIX, E. & LEUSEN, I. (1974). The conversion o f testosterone to 7a- hydroxy testosterone by incubated rat testes. Aerozzù, 24, 753-754.

EELKMAN ROOD A, S.J., OTTEN, M.H., VAN LOON, M.A.C., KAPTEIN, E. & VISSER, T.J. (1989). Metabolism of triiodothyronine in rat hepatocytes. Endocrinol, 125, 2187-2197.

ELLIS, G.B. & DESJARDINS, C. (1982). Male rats secrete luteinising hormone and testosterone episodically. Endocrinol, 110(5), 1618-1627.

FABBRI, A. & DUE AU, M.L. (1988). Hormonal regulation of beta-endorphin in the testis. J.Steroid Biochem., 30(1-6), 347-352.

FABBRI, A., TSAI-MORRIS, C.H., LUNA, S., FRAIOLI, F. & DUFAU, M.L. (1985). Opiate receptors are present in the rat testis:identifîcation and localisation in Sertoli cells. Endocrinol, 117(6), 2544-2546.

FAHIM, M.S., DEMENT, G., HALL, D.G. & FAHIM, Z. (1970). Induced alterations in the hepatic metabolism o f androgens in the rat. Am.J.Obstet.Gynecol, 107(7), 1085-1091.

242 FAHMY, O.G. & FAHMY, MJ. (1976). Mutagenic selectivity of carcinogenic nitroso compounds: n,a-acetoxymethyl-N-methylnitrosamine. Chem-Biol. Int., 14, 21-35.

FANG, V.S., REFETOFF, S. & ROSENFIELD, R.L. (1974). Hypogonadism induced by a transplantable, prolactin-producing tumor in male rats. Endocrinol, 95,991-998.

FAWCETT, D.W., NEAVES, W.B. & FLORES, M.N. (1973). Comparative observations on intertubular lymphatics and the organisation of the interstitial tissue o f the mammalian testis. BiolReprod., 9, 500-532.

FELDMAN, D. (1986). Ketoconazole and other imidazole derivatives as inhibitors of steroidogenesis. EndocriolRev., 7, 409-420.

FLOWERS, N.L., O’DONNELL, J.P. & COLBY, H.D. (1989) Spirinolactone metabolism in target tissues: characteristics of deacetylation in kidney, liver, adrenal cortex and testes. Drug Metab.Disp., 17(2), 186-189.

FOOTE, R.H. & BERNDTSON, W.E. (1992). The Germinal Epithelium. In. Reversibility in Testicular Toxicity Assessment, eds. Scialli, A.R. & Clegg, E.D. Ann.Arbor, Boca Raton, FL CRC Press.

FORD, E. & HUGGINS, C.B. (1963). Selective destruction in testis is induced by 7,12- dimethylbenz(a)anthracene. J.Exp.Med., 118, 27-40.

FORT, F.L., MIYAJIMA, H., ANDO, J., SUZUKI, T., YAMAMOTO, M., HAMASHIMA, T., SATO, S., KITAZAKI, T., MAHONY, M.C. & HODGEN, G.D. (1995). Mechanism for species- specific induction of leydig cell tumors in rats by Lansoprazole. Fundam.ApplToxicol, 26(2), 191- 202.

FOSTER, J.R., ELCOMBE, C.R., BOOBIS, A.R., DAVIES, D.S. et al. (1986). Immunochemical localisation o f cytochrome P450 in hepatic and extrahepatic tissues of the rat with a monoclonal antibody against cytochrome P-450c. Biochem.Pharmacol, 35(24), 4543-4554.

FOUQUET, J.P. (1987). Ultrastructural analysis of a local regulation o f Leydig cells in the adult monkey (Macaca fascicularis) and rat. J.i?epràz/.Ferù7., 79, 49-56.

FRANCHIMONT, P., CHARI, S. & DEMOULIN, À. (1975). Hypothalamus-pituitary-testis interaction. J.Reprod.Fertil, 44, 335-350.

FUCHS, W., SENNEWALD, R., KLOTZ, U. (1994). Lansoprazole does not affect the bioavailability of oral contraceptives. Br.J.Clin.Pharmacol, 38, 376-380.

GAETANI, M., DE GIORGIO, R., BURATTI, P., PASQUALI, R., CAPELLI, M., STANGHELLINI, V. & CORINALDESI, R. (1995). Chronic oral administration of lansoprazole does not affect the hypothalamic pituitary gonadal axis in healthy young men. Eur.J. Gastroenterology & Hepatology, 7(3), 211-213.

GALLAGHER, E.P., BUETLER, T.M., STAPLETON, P.L., WANG, C., STAHL, D.L. & EATON, D.L. (1995). The effects of diquat and ciprofibrate on mRNA expression and catalytic activities of hepatic xenobiotic-metabolising and antioxidant enzymes in rat liver. ToxicolApplPharmacol, 134, 81-91. ' .

GAY, V.L. & DEVER, N.W. (1971). Effects of testosterone proprionate and estradiol benzoate, alone or in combination, on serum LH and FSH in orchidectomised rats. Endocrinol, 89, 161-168.

GELBER, S., HARDY, M., MENDIS-HANDAGAMA, S. & CASELLA, S. (1992). Effect of insulin-like growth factor-1 on androgen production by highly purified pubertal and adult Leydig cells. J.Androl, 13, 125-130.

243 GEMZIK, B., GREENWAY, D., NEVINS, c. & PARKINSON, A. (1992). Regulation of two electrophoretically distinct proteins recognised by antibody against rat liver cytochrome P450 3A1. J.Biochem.Toxicol.,l{\), 43-52.

GEORGELLIS, A. & RYDSTROM, J. (1989). Cell-specific metabolic activation of 7,12- dimethylbez(a)anthracene in rat testis. Chem.Biol.Int., 12, 65-78.

GEORGELLIS, A , MONTELIUS, J. & RYDSTROM, J. (1987). Evidence for a ffee-radical- dependent metabolism of 7,12-dimethylbenz(a)anthracene in rat testis. Toxicol.Appl.Phanm.<:o\., 87, 141-154.

GIBSON, G.G. & SKETT, P. (2001). Enzymology and molecular mechanisms of drug metabolism reactions. In. Introduction to Drug Metabolism, eds. Gibson, G.G. & Skett, P. Nelson Thornes Ltd. United Kingdom.

GIBSON, G. & LAKE, B. (1993). PeroxisomesiBiology and Importance in Toxicology and Medicine. Gibson, G & Lake, B. eds. Taylor & Francis. Washington.

GIBSON, J.P., NEWBERNE, J.W., KUHN, W.L. & ELSEA, J.R. (1967). Comparative chronic toxicity of three oral estrogens in rats. Toxicol.Appl.Pharmacol., 11, 489-510.

GILLILAND, F.D. & KEY, C.R. (1995). Male genital cancers. Cancer, Suppl.l, 295-315.

GLUE, P., BANFIELD, C.R., PERHACH, J.L., MATHER, G.G., RACHA, J.K. & LEVY, R.H. (1997). Pharmacokinetic interactions with felbamate: m vz>o-zn vivo correlation. Clin.Pharmacokinet., 33(3), 214-224.

GNESSI, L., BASCIANI, S., MARIANI, S., ARIZZI, M., SPERA, G., WANG, C., BONDJERS, C., KARLSSON, L. & BETSHOLTZ, C. (2000). Leydig cell loss and spermatogenic arrest in platelet- derived growth factor (PDGF)-A-deficient mice. J.CellBiol., 149, 1019-1026.

GNESSI, L., FABBRI, A. & SPERA, G. (1997). Gonadal peptides as mediators of development and functional control of the testis: An integrated system with hormones and the local environment. Endocrine Rev., 18(4), 541-609.

GOLDSTEIN, J.A. & LINKO, P. (1984). Differential induction of two 2,3,7,8,-tetrachlorodibenzo-p- dioxin-inducible forms of cytochrome P450 in extrahepatic versus hepatic tissues. Mol.Pharmacol, 25, 185-191.

GONZALEZ, F.J., SONG, B.J. & HARDWICK, J.P. (1986). Pregnenolone 16 alpha-carbonitrile- inducible P-450 gene family: gene conversion and differential regulation. Mol. Cell Biol., 6, 2969- 2976.

GRAHAM, M.J., WINHAM, M.A., OLD, S.L. & GRAY, T.J.B. (1996). Comparative hypolipidaemic and peroxisomal effects of ciprofibrate, clofibric acid and their respective difluorocyclopropyl and 4-fluoro- substituted analogues in rat. Xenobiotica, 26(7), 695-707.

GRAHAM, M.J., WINHAM, M.A., HARPUR, E.S., BONNER, F.W. & GRAY, T.J.B. (1994). The relative hypolipidaemic activity and hepatic effects of ciprofibrate enantiomers in the rat. Biochem.Pharmacol., 48(\2), 2163-2111.

GRISWOLD, M.D. (1993). Actions of FSH on mammalian Sertoli cells. In. The Sertoli Cell. ezfe. Russell, L.D. & Griswold, M.D. Cache River Press. Clearwater, FI.

GUENGERICH, F.P. (1988). Roles of cytochrome P450 enzymes in chemical carcinogenesis and cancer chemotherapy. Cancer Res., 48, 2946-2954.

244 GUENGERICH, P.P., DANNON, G.A., WRIGHT, ST ., MARTIN, M.V. & KAMINSKY, L.S. (1982). Purification and characterisation of liver microsomal cytochrome P450: electrophoretic, spectral, catalytic and immunochemical properties and inducibility of eight isozymes isolated from rats treated with phénobarbital and (3-naphthoflavone. Biochem., 21(23), 6019-6030.

GUSTAFSSON, J.A., MODE, A., NORSTEDT, G. & SKETT, P. (1983). Sex steroid induced changes in hepatic enzymes. Ann.Rev.Physiol., 45, 51-60.

HABERT, R., LEJEUNE, H. & SAEZ, J.M. (2001). Origin, differentiation and regulation o f fetal and adult Leydig cells. Mol.Cell. Endocrinol, 179, 47-74.

HAFIEZ, A. A., LLOYD, C.W. & BARTKE, A. (1972). The role of prolactin in the regulation of testis function; the effects of prolactin and luteinising hormone on the plasma levels of testosterone and androstenedione in hypophysectomised rats. J.Endocrinol, 52, 327-332.

HALL, P.P. (1994). Testicular steroid synthesis; organisation and regulation. In. The Physiology of Reproduction. Knobil, E. & Niell, J.D. eds. Raven Press. New York.

HALL, S.H., BERTHELON, M-C., AVALLET, O. & SAEZ, J.M. (1991). Regulation of c-fos, c-jun, jun-~Q and c-myc messenger ribonucleic acids by gonadotropin and growth factors in cultured pig Leydig cell. Endocrinol, 129(3), 1243-1249.

HALL., P.F., CHARBONNIER, C., NAKAMURA, M. & GABBIANI, G. (1979). The role of microfilaments in the response of Leydig cells to luteinising hormone. J.Steroid Biochem., 11, 1361- 1366.

HALL, P.P., IRBY, D.C. & DE KRETSER, D.M. (1969). Conversion of cholesterol to androgens by rat testes: comparison of interstitial cells and seminiferous tubules. Endocrinol, 84, 488-496.

HAMAD A, Y & FUTAMURA, Y. (1998). Induction of Leydig cell tumours by lacidipine via up- regulation of the LHRH receptor on Leydig cells in rats. J.Toxzco/.^cz., 23,35-52.

HANOIKA, N., JINNO, H., TAKAHASHI, A., NAKANO, K., YOADA, R., NISHIMURA, T .& ANDO, M. (1995). Interaction of trichloroethylene with rat hepatic microsomal cytochrome P450- dependent monooxygenases. WeModzorica, 25(2), 151-165.

HARD, G.C. (1998). Recent developments in the investigation o f thyroid regulation and thyroid carcinogenesis. Env.Health Persp., 106(8), 427-436.

HARDY, M.P., ZIRKIN, B.R. & EWING, L.L. (1989). Kinetic studies on the development o f the adult population of Leydig cells in testes of the pubertal rat. Endocrinol, 124, 762-770.

HEID, C.A., STEVENS, J., LIVAKK, K.J. & WILLIAMS, P.M. (1996). Real-time quantitative PCR. Genome Research, 6, 986-994.

HELLER, C. & LEACH, D. (1971). Quantification of Leydig cells and measurement of Leydig cell size following administration of human chorionic gonadotropin to normal men. J.Reprod.Fertil, 25, 185-192.

HENDERSON, S.B ., CIACCIO, L.A. & KINCL, F.A. (1980). Neonatal sterilisation of rodents with steroid hormones: Metabolic clearance rate of testosterone and estradiol in 60 day old male rats influenced by neonatal steroid exposure. J.Steroid Biochem., 13, 297-304.

HILL-SAMLI, M. & MACLEOD, R.M. (1974). Interaction of thyrotropin-releasing hormone and dopamine on the release of prolactin from the rat anterior pituitary in vitro. Endocrinol, 95(4), 1189- 1192.

HOBBS, C.J., JONES, R.E. & PLYMATE, S.R. (1992). Effects of sex hormone binding globulin (SHBG) on testosterone transport in to the cerebrospinal fluid. J.Steroid Biochem. Mol Biol, 42(6), 629-635.

245 HOLLAND, P.M., ABRAMSON, R.D., WATSON, R. & GELFAND, D.H. (1991). Detection of specific polymerase chain reaction product by utilising the 5’-3’ exonuclease activity of Thermus aquaticus DNA polymerase. Proc. Natl. Acad. Sci., 88, 7276-7280.

HUCKINS, C. (1971). The spermatogonial stem cell population in adult rats: their morphology, proliferation and maturation. Anatomical Record, 169, 533-558.

HUHTANIEMI, I.T. (1983). Gonadotropin receptors: correlates with normal and pathological functions of the human ovary and testis. In. Clinical Endocrinology and Metabolism:Receptors in Health and Disease, ez/. Clayton, W.B. Saunders. Philadelphia.

HUHTANIEMI, I.T. & CATT, K.J. (1981). Induction and maintenance of gonadotropin and lactogen receptors in hypoprolacinemic rats. Endocrinol., 109(2), 483-490. •

HUNTER, M.G., SULLIVAN, M.H., DIX, C.J., ALDRED, L.F. & COOKE, B.A. (1982). Stimulation and inhibition by LHRH analogues of cultured rat Leydig cell function and lack o f effect on mouse Leydig cells. M?/. Ce//.F«zfz?crmo/., 27(1), 31-44.

HUSEBY, R.A. (1980). Demonstration of a direct carcinogenic effect of estradiol on Leydig cells of the mouse. Cancer Res., 40, 1006-1013.

HUSEBY, R.A. (1976). Estrogen-induced Leydig cell tumour in the mouse: a model system for the study of carcinogenesis and hormone dependency. J.Toxicol.Environ.Health, Suppl., 1, 177-192.

IMAOKA, S., FUJITA, S. & FUNAE, Y. (1991). Age-dependent expression of cytochrome P450s in rat liver. Biochim. Biophys. Acta, 1091(3), 181-\92.

IMAOKA, S; TERANO, Y. & FUNAE, Y. (1988). Constitutive testosterone 6|3-hydroxylase in rat liver. y.RiocAew., 104,481-487.

INANO, H & TAMAOKI, B. (1971). Regulation o f testosterone biosynthesis in rat testes by 7 alpha- hydroxylated C 19 steroids. Biocim.Biophys.Acta, 239, 482-493..

INANO, H., SUZUKI, K., WAKABAYASHI, K. & TAAMAOKI, B. (1973). Biological activities of 7a-hydroxylated Cl 9-steroids and changes in rat testicular 7a-hydroxylase activity with gonadal status. Endocrinol., 92,22.

INANO, H., TSUNO, K. & TAMAOKI, B. (1970). Identification of 7a-hydroxylated androgens as the metabolites of androstenedione by testicular microsomal fraction o f rats. Biochem., 9(11), 2253- 2260. ^

ISWARAN, T.J., IMAI, M., BETTON, G.R. & SIDDALL, R.A. (1997). An overview of animal toxicology studies withbicalutamide (ICI 176,334). Vi TbxzcoZ. Jcz., 22(2), 75-88.

JAPENGA, A.C., DAVIES, S., PRICE, R.J. & LAKE, B.G. (1993). Effect of treatment with pyrazine and some derivatives on cytochrome P450 and some enzyme activities in rat liver. Xenobiotica., 23(2), 169-179.

JANSSON, I, MOLE, J. & SCHENKMAN, J.B. (1985a). Purification and characterisation of anew form (RLM2) of liver microsomal cytochrome P450 from untreated rat. J.BiolChem., 260, 7084- 7093.

JANSSON , J.O., EDEN, S. & ISAKSSON, O. (1985b). Sexual dimorphism in the control of growth hormone secretion. E’nz/ocn'zze Rev., 6(2), 128-150.

JIANG, Y., KUO, C.L., PERNECKY, S.J. & PIPER, W.N. (1998). The detection of cytochrome P4502E1 and its catalytic activity in rat testis. Biochem. Biophys. Res. Comm., 246(3), 578-583.

JONES, E.J. & RIDDICK, D.S. (1996). Regulation of constitutive rat hepatic cytochromes P450 by 3-methylcholanthrene. Xenobiotica, 26(10), 995-1012.

246 JUNKER-WALKER, U. & NOGUES, V. (1994). Changes induced by treatment with aromatase inhibitors in testicular Leydig cells of rats and dogs. Exp. Toxicol.PathoL, 46, 211-213.

KALLA, N.R., NISULA, B.C., MENARD, R. & LORIAUX, D.L. (1980). The effect of estradiol on testicular testosterone biosynthesis. Endocrinol, 106(1), 35-39.

KALRA, P.S. & KALRA, S.P. (1977). Circadian periodicities on serum androgens, progesterone, gonadotropins and luteinising hormone-releasing hormone in male rats. Endocrinol, 101(6), 1821- 1827.

KAY, L., KAMPMANN, J.P., SVENDSEN, T.L. et al. (1985). Influence of rifampicin and isoniazid on the kinetics of phenytoin. y.C/m.R/zarwaco/., 20, 323-326.

KEATING, RJ. & TCHOLAKIAN, R.K. (1979). In vivo patterns of circulating steroids in adult male rats. I. Variations in testosterone during 24- and 48- hour standard and reverse light/dark cycles. . Endocrinol, 104,184-188.

KEDDERIS, G.L. & MUGFORD, C.A. (1998). Sex-dependent metabolism of xenobiotics. CUT Archives, 18(7-8), 1-7. -

KEENEY, D.S. & MASON, J.I. (1992). Expression of testicular 3 beta-hydroxysteroid dehydrogenase/delta 5-4 isomerase: regulation by luteinising hormone and forskolin in Leydig cells of adult rats. Endocrinol, 130(4), 2001-2015.

KEENEY, D.S., MENDIS-HANDAGAMA, S.M., ZIRKIN, B.R. & EWING, L.L. (1988). Effect of long term deprivation of luteinising hormone on Leydig cell volume, Leydig cell number and ' steroidogenic capacity of the rat testis. Endocrinol, 113(6), 2906-2915.

KERR, J.B. & SHARPE, R.M. (1985). Follicle-stimulating hormone induction of Leydig cell maturation. Endocrinol, 116(6), 2592-2604.

KHAN, S., TEERDS, K. & DORRINGTON, J. (1992). Growth factor requirements for DNA synthesis by Leydig cells from the immature rat. BiolReprod., 46, 335-341.

KLEMCKE, H.G. & BARTKE, A. (1981). Effects of chronic hyperprolactinaemia in mice on plasma gonadotropin concentrations and testicular human chorionic gonadotropin binding sites. Endocrinol, , 108, 1763-1768.

KOBAYASHI, Y., MOTOHASHI, Y., MIYAZAKI, Y., YATAGAI, M .& TAKANO, T. (1991). Immunohistochemical study of temporal variations in cytochrome P450 isozymes in rat testis and their modifications by the inductive effects of cadinenes. Int.J.Biometeorol, 35, 234-238.

KOCAREK, T.A. & REDDY, A.B. (1996). Regulation of cytochrome P450 expression by inhibitors of hydroxymethylglutaryl-coenzyme A reductase in primary cultured rat hepatocytes and in rat liver. Drug Metab.Disp.,14(\\),\\91-1204.

KOKKINAKIS, D.M., KOOP, D.R., SCARPELLI, D.G., COON, M.J. & HOLLENBERG, P.P. (1985). Metabolism of N-nitroso-2,6-dimethylmorpholine by isozymes of rabbit liver microsomal cytochrome P450. Cancer Re^., 45(2), 619-624.

KOKUFU, T., IHARA, N., SUGIOKA, N., KOYAMA, H., OHTA, T., MORI, S. & NAKAJIMA. (1995). Effects of lansoprazole treatment on pharmacokinetics and metabolism of theophylline. Eur.J.Clin.Pharmacol, 48,391-395.

KUNTZMAN, R., LAWRENCE, D. & CONNEY, A.H. (1965). Michaelis constants for the hydroxylation of steroid hormones and drugs by rat liver microsomes. MolPharmacol, 1, 163-167.

LACROIX, D., SONNIER, M., MONCION, A , CHERON, G. & CRESTEIL, T. (1997). Expression of CYP3A in the human liver: evidence that the shift between CYP3A7 and CYP3A4 occurs immediately after birth. Eur. J. Biochem, 247(2), 625-634.

247 LACROIX, E., EECHAUTE, W. & LEUSEN, L (1975). Influence of age on the formation of 5a- androstanediol and 7a-hydroxy testosterone by incubated rat testes. Steroids, 25, 649-661.

LANG, P.L. (1995). Spontaneous neoplastic lesions in the Crl:CD-l BR mouse. Charles River Laboratory, Raleigh.

LEBLOND, C.P. & CLERMONT, Y. (1952). Definition of the stages of the cycle of the seminiferous cpiXlaéivammXhQxzX. Annals New York Acad. Sci., 55,548-513.

LEE, D.W. & PARK, K.H. (1989). Testosterone metabolism by microsomal cytochrome P450 in the liver of rats treated with some inducers. Int. J.Biochem., 21(1), 49-57.

LEE, L.G., CONNELL, C.R. & BLOCH, W. (1993). Allelic discrimination by nick-translation PCR with fluorogenic probes. Amc/czc yfdzfe Rc 5 ., 21, 3761-3766.

LEE, I.P., SUZULI, K., MUKHTAR, H. & BEND, J.R. (1980). Hormonal regulation of cytochrome P450 dependent monooxygenase activity and epoxide-metabolising enzyme activity in the testis of hypophysectomized rats. Cancer Research, 40, 2486-2492.

LEE, D.K.H., JANIKOWSKY, A., BIRD, C.E. & CLARK, A.F. (1974). Kinetics of (1,2-% - testosterone metabolism in normal adult male rats: effects of estrogen administration. J.Steroid Biochem., 5, 27-32.

LEJEUNE, H., SANCHEZ, P., CHUZEL, F., LANGLOIS, D. & SAEZ, J.M. (1998). Time-course of human recombinant luteinising hormone on porcine Leydig cell specific differentiated function. Mol.Cell Endocrinol., 144, 59-69.

LEPHART, E.D. & SIMPSON, E.R. (1991). Assay of aromatase activity. Methods Enzymol., 206, 477-483.

LESLIE, G.B., NOAKES, D.N., POLLITT, F.D., ROE, F.J. & WALKER, T.F. (1981). A two-year study with cimetidine in the rat: assessment for chronic toxicity and carcinogenicity. ToxicolApplPharmacol., 61, \\9-\31.

LEVIN, W., THOMAS, P.E., RYAN, D.E. & WOOD, A.W. (1987). Isozyme specificity of testosterone 7a-hydroxylation in rat hepatic microsomes: is cytochrome P-450 the sole catalyst? Arch.Biochem.Biophys., 258(2), 630-635.

LEVIN, W., WELCH, R.M. & CONNEY, A.H. (1974). Increased liver microsomal androgen metabolism by phénobarbital. J.Pharmacol.Expt.Ther., 188(2), 281-292.

LEWIS, D.F.V. (2001). Substrate selectivity and metabolism. In. Cytochrome P450: Structure and Function, ed. Lewis, D.F.V. Taylor & Francis. London.

LEWIS, D.F.V. (1996). P450 substrate specificity and metabolism. 7n Cytochrome P450: structure, function and mechanisms, ed. Lewis, D.F.V. Taylor & Francis. London.

LIE, Y.S. & PETROPOULOS, C.J. (1998). Advances in quantitative PCR technology: 5’ nuclease assays. Current Opinion in Biotechnology, 9 ,43-48.

LIN, T., LINCOLN, T.M., BROWN, N., MURONO, E.P., OSTERMAN, J. & NANKIN, H.R. (1982). Protein kinase activity o f purified Leydig cells: Low protein kinase activity causes impaired steroidogenesis in band two cells. Endocrinol, 111(4), 1391-1393.

LIU, R., HURTT, M.E., COOK, J.C. & BIEGEL, L.B. (1996). Effect of peroxisome proliferator ammonium perfluorooctanoate (C 8 ) on hepatic aromatase activity in adult male CRLCD rats. Fundam.Appl.Toxicol, 30(2), 220-228.

248 LIVAK, K.J., FLOOD, SJ., MARMARO, L, GIUSTI, W. & DEETZ, K. (1995). Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridisation. PCR Methods Applic., 4, 357-362.

LONGCOPE, C., PRATT, J.H., SCHNEIDER, S.H. & FINEBERG, S.E. (1978). Aromatisation of androgens by muscle and adipose tissue in vitro. J.Clin.Endocrinol. Metab., 46, 146-152.

LUBET, R.A., DRAGNEV, K., CHAUHAN, D.P., NIMS, R.W., DIWAN, B.A., WARD, J.M., JONES, C.R., RICE, J.M. & MILLER, M.S. (1992). A pleiotropic response to phenobarbital-type enzyme inducers in the F344/NCr rat. Biochem. Pharmacol., 43, 1067-1078.

LUO, L., CHEN, H. & ZIRKIN, B.R. (2001). Leydig cell ageing: steroidogenic acute regulatory protein (StAR) and cholesterol side-chain cleavage enzyme. 7.^ndrol., 22 (1), 149-156.

LUO, L., CHEN, H., STOCCO, D.M. & ZIRKIN, B.R. (1998). Leydig cell protein synthesis and steroidogenesis in response to acute stimulation by luteinsing hormone in rats. Biol. Reprod., 59, 263- 270.

MACHALA, M., NECA, J., DRABEK, P., ULRICH, R., SABATOVA, V., NEZVEDA, K., RASZYK, J. & GAJDUSKOVA, V. (1998). Effects of chronic exposure to PCBs on cytochrome P450 systems and steroidogenesis in liver and testis of bulls {Bos taurus). Comp.Biochem.PhysioL, 120, 65-70.

MACLEOD, R.M. & LEHMEYER, J.E. (1974). Studies on the mechanism o f dopamine-mediated inhibition o f prolactin secretion. Rnz/ocrmo/., 94(4), 1077-1085.

MAENPAA, J., PELKONEN, O., CRESTEIL, T. & RANE, A. (1993). The role of cytochrome P4503A (CYP3A) isofrom(s) in the oxidative metabolism of testosterone and benzphetamine in human Sidulttind fetül liver. J.Steroid.Biochem.Mol.Biol., 44{l), 61-67.

MAGNANTI, M., GIULIANI, L., GANDINI, O., GAZZANIGA, P., SANTIEMMA, V., CIOTTI, M., SACCANI, G., FRATI, L. & AGLIANO, A.M. (2000). Follicle-stimulating hormone, testosterone and hypoxia differentially regulate UDP-glucuronsyltransferase I isoforms expression in rat Sertoli and Tpcntubular myoid cells. J.Steroid Biochem.Mol.Biol., 74, 149-155.

MAKOWSKA, J.M., ANDERS, C., GOLDFARB, P.S., BONNER, F. & GIBSON, G.G. (1990). Characterisation of the hepatic responses to the short-term administration of ciprofibrate in several rat strains. Biochem.Pharmacol., 40{5), 10S3-1093.

MANN, D.R., ADAMS, S.R., GOULD, K.G., ORR, T.E & COLLINS, D.C. (1989). Evaluation of the possible direct effects of gonadotropin-releasing hormone analogues on the monkey {Macaca mulatto) testis. J.Reprod.Fetil., 85, 89-95.

MARIEB, E.N. (1995). The reproductive system. In. Human Anatomy and Physiology (3^** Edition). Benjamin Cummings. California.

MARTEL, C., RHEAUME, E., TAKAHASHI, M., TRUDEL, C., COUET, J., LUU-THE, V., SIMARD, J. & LABRIE, F. (1992). Distribution of 17(3-hydroxysteroid dehydrogenase gene expression and activity in rat and human tissues. J.Steroid Biochem. Mol. Biol., 41(3-8), 597-603.

MASUBUCHI, N., LI, A.P. & OKAZAKI, 0. (1998). An evaluation of the cytochrome P450 induction potential o f pantoprazole in primary human hepatocytes. Chem.Biol.Int., 114, 1-13.

MASUBUCHI, N., HAKUSUI, H. & OKAZAKI, O. (1997a). Effects of pantoprazole on xenobiotic metabolising enzymes in rat liver microsomes: a comparison with other proton pump inhibitors. Drug Metab.Disp., 25(5), 584-589.

MASUBUCHI, N., HAKUSUI, H. & OKAZAKI, O. (1997b). Effects of proton pump inhibitors on thyroid hormone metabolism in rats. Biochem.Pharmacol., 54, 1225-1231.

249 MATSUI, M., KINUYAMA, Y. & HAKOZAKI, M. (1974). Biliary metabolites of testosterone and testosterone glucosiduronate in the rat. Steroids, 24(4), 557-573.

MATSUMOTO, A.M., KARPAS, A.E. & BREMNER, W.J. (1986). Chronic human chorionic gonadotropin administration in normal men: evidence that follicle-stimulating hormone is necessary for the maintenance of quantitatively normal spermatogenesis in man. J. Clin. Endocrinol.Metab., 62(6),1184-1192.

MATSUNAGA, T., NAGATA, K., HOLSZTYNSHA, E.J., LAPENSON, D.P., SMITH, A., KATO, R., GELBOIN, H.V., WAXMAN, D.J. & GONZALEZ, F.J. (1988). Gene conversion and differential regulation in the rat P-450 HA gene subfamily. Purification, catalytic activity, cDNA and deduced amino acid sequence, and regulation of an adult male-specific hepatic testosterone 15 alpha- hydroxylase. J.5zo/. C/zew., 263, 17995-18002.

McCLAIN, R.M. (1992). Thyroid gland neoplasia: non-genotoxic mechanisms. Toxicol. Letters, 64, 397-408.

McCLAIN, R.M., LEVIN, A.A., POSCH, R. & DOWNING, J.C. (1989). The effect o f phénobarbital on the metabolism and excretion of thyroxine in rats. Toxicol.Appl.Phamracol., 99, 216-228.

McCLAIN, R.M., POSCH, R.C., BOSAKOWSKI, T. & ARMSTRONG, J.M. (1988). Studies on the mode of action for thyroid gland tumor promotion in rats by phénobarbital. Toxicol.Appl.Pharmacol., 94, 254-265.

McCLELLAN-GREEN, P., WAXMAN, D.J., CAVERNESS, M. & GOLDSTEIN, J.A. (1987): Phenotypic differences in expression of cytochrome P450g but not its mRNA in outbred male Spragaa-'DavAQyrais. Arch.Biochem.Biophys.,lS3,\5-15.

McCo n n e l l , r .f ., w e s t e n , h .h ., u l l a n d , b .m ., b o s l a n d , m .c . & w a r d , j.m . (1992). Proliferative lesions of the testes in rats with selected examples from mice. URG-3, Guidelines for Toxicologic Pathology, STP/ARP/AFIP, Washington.

McNEILLY, A.S., SHARPE, R.M., DAVIDSON, D.W. & FRASER, H.M. (1978). Inhibition of gonadotropin secretion by induced hyperprolactinaemia in the male rat. J.Endocrinol., 79, 59-68. -

MEDCLE, A.W., SANDERS, S.W., TOLMAN, K.G., JENNINGS, D.E., KAROL, M.D. & RINGHAM, G.L. (1994). Effect of Lansoprazole on male hormone function. Drug.Invest., 8(4), 191- 202.

MENARD, R.H. & PURVIS, J.L. (1973). Studies of cytochrome P450 in testis microsomes. Arch.Biochem.Biophys., 154, 8-18.

MENON, K.M., DORFMAN, R.I., & FORCHIELLI, E. (1967). Influence of gonadotropins on the side chain cleavage reaction by rat testis mitochondrial preparations. Biochim. Biophys. Acta, 148, 486-494.

MILLER, G.L. (1959). Protein determination for a large number of samples Anal. Chem., 31, 964- 965.

MITSUMORI, K. & EL WELL, M.R. (1988).“ Proliferative lesions in the male reproductive system of F344 rats and B6C3F1 mice: incidence and classification. Env. Health Pers^., 77, 11-21.

MITTLER, J.C. (1985). Studies on the kinetics of the interaction of 7a-hydroxytestosterone with the 5a-reductase. Steroids, 45(2), 135-142.

MIYATA, M., KUDO, G., LEE, Y., YANG, T.J., GELBOIN, H.V., FERNANDEZ-SALGUERO, P., KIMURA, S. & GONZALEZ, F.J. (1999). Targeted disruption of the microsomal epoxide hydrolase gene. y.Rzo/.C/zem., 274(34), 23963-23968.

250 MOORE, J.W. & BULBROOK, R.D. (1988). The epidemiology and function o f sex hormone-binding globulin. Oxford Reviews of Reproductive Biology, 10,180-236.

MOORE, R.J. & WILSON, J.D. (1972). Localisation of the reduced nicotinamide adenine dinucleotide phosphate: A4-3-ketosteroid 5a-oxidoreductase in the nuclear membrane of the rat ventral prostate. y.5zo/.C/zem., 247, 958-967.

MORGAN, E.T., MACGEOCH, C. & GUSTAFSSON, J.A. (1985). Hormonal and developmental regulation of expression o f the hepatic microsomal steroid 16 alpha-hydroxylase cytochrome P450 apoprotein in the rat. y.5zo/.C^em., 260, 11895-11898.

MORI, H. & CHRISTENSEN, A.K. (1980). Morphometric analysis of Leydig cells in the normal rat testis. J.Cell.BioL, 84,340-354.

MORI, H., KADOTA, A., FUKUNISHI, R., KUKITA, H., TAKEUCHI, N. & MATSUMOTO, K. (1980). Effects of a cholesterol-rich diet and a hypolipidaemic drug (cloflbrate, CPIB) on Leydig cells in rats: stereological and biochemical analysis. Andrologia, 12(3), 281-291.

MUKHTAR, H., LEE, I.P., FOUREMAN, G.L. & BEND, J.R. (1978). Epoxide metabolising enzyme activities in rat testis: postnatal development and relative activity in interstitial and spermatogenic cell compartments. Chem.Biol.Int., 22, 153-165.

NAGATA, K., MATSUNAGA, T., GILLETTE, J., GELBOIN, H.V. & GONZALEZ, F.J. (1987). Rat testosterone 7a-hydroxylase: isolation, sequence and expression o f the cDNA and its developmental regulation and induction by 3-methylcholanthrene. J.BiolChem., 262(6), 2787-2793.

NAGATA, K., LIBERATO, D.J., GILLETTE, J.R. & SASAME, H.A. (1986). An unusual metabolite of testosterone: 17P-hydroxy-4,6-androstadiene-3-one. Drug Metab.Disp., 14(5), 559-565.

NAKAJIN, S., SHIVLEY, J.E., YUAN, P-M. & HALL, P.P. (1981). Microsomal cytochrome P450 from neonatal pig testis: two enzymatic activities (17a-hydroxylase and C17-20 lyase) associated with one protein. Biochem., 20,4037-4042.

NEBERT, D.W., NELSON, D.R. & FEYEREISEN, R. (1989). Evolution of the cytochrome P450 genes. Xenobiotica., 19, \\49-l\60.

NELSON, D.R., KOYMANS, L., KAMATAKI, T., STEGEMAN, J.J., FEYEREISEN, R., WAXMAN, D.J., WATERMAN, M.R., GOTOH, 0., COON, M.J., ESTABROOK, R.W., GUNSALUS, I.e. & NEBERT, D.W. (1996). P450 superfamily: update on new sequences, gene mapping, accession numbers and nomenclature. P/zamacogenetics, 6 , 1-42.

NICHOLSON, H.D., WORLEY, R.T., CHARLTON, H.M. & PICKERING, B.T. (1986). LH and testosterone cause the development of seminiferous tubule contractile activity and the appearance of - testicular oxytocin in hypogonadal mice. /.Paz/ocrzzzo/., 110(1), 159-167.

NISHIHARA, M. & TAKAHASHI, M. (1983). Effects of active immunization against estradiol-17 beta on luteinizing hormone and testosterone in male rats. BiolReprod., 29(5),1092-1097.

NOLAN, C.J. & PAYNE, A.H. (1990). Genotype at the P450scc locus determines differences in the amount of P450scc protein and maximal testosterone production in mouse Leydig cells. MolEndocrinol, 4(10), 1459-1464.

NORMINGTON, K. & RUSSELL, D.W. (1992). Tissue distribution and kinetic characteristics of rat steroid 5 alpha-reductase isozymes. Evidence for distinct physiological functions. J.BiolChem., 267(27), 19548-19554.

NORTON, J.N. & SKINNER, M.K. (1989). Regulation of Sertoli cell function and differentiation through the actions of a testicular paracrine factor PModS. Endocrinol, 124(6), 2711-2719.

251 OMIECINSKI, C.J. (1986). Tissue-specific expression of rat mRNAs homologous to cytochromes P450b and P-450e. Nucleic Acids Research, \4{y),\525-\539.

OMIECINSKI, C.J., REDLICH, C.A. & COSTA, P. (1990a). Induction and developmental expression of cytochrome P4501A1 messenger RNA in rat and human tissues:detection by polymerase chain reaction. Cancer Res., 50, A3\5-432\.

OMIECINSKI, C.J., HASSETT, C. & COSTA, P. (1990b). Developmental expression and in situ localisation of the phenobarbital-inducible rat hepatic mRNAs for cytochromes CYP2B1, CYP2B2, CYP2C6 and CYP3A1. M?/.P^armaco/., 38, 462.

OMURA, T. & SATO, R. (1964). The carbon monoxide binding pigment of liver microsomes. J.5zo/.C/zew., 239 (7), 2370-2378.

ONODA, M. & HALL, P.P. (1981). Inhibition of testicular microsomal cytochrome P-450 (17a- hydroxylase/ 017,20 lyase) by estrogens. Endocrinol., 109(3), 763-767.

ORADELL, N.J. (1995a). Pendil® (Felodipine). 7zz. Physician’s desk reference (11* Edition). Medical Economics Data.

ORADELL, N.J. (1995b). Eulexin (Flutamide). In. Physician’s desk reference (11* Edition). Medical Economics Data.

ORADELL, N.J. (1995c). Proscar® (Finasteride). In. Physician’s desk reference (11* Edition). Medical Economics Data.

ORADELL, N.J. (1995d). Supprelin® (Histrelin). In. Physician’s Desk Reference (11* Edition). Medical Economics Data.

ORADELL, N.J. (1995e). Lupron (Leuprolide). In. Physician’s Desk Reference (11* Edition). Medical Economics Data.

ORADELL, N.J. (1995f). Atromid-S® (Cloflbrate). In. Physician’s desk reference (11* Edition). Medical Economics Data..

ORTH, J.M. (1986). FSH-induced Sertoli cell proliferation in the developing rat is modified by beta- endorphin produced in the testis. Endocrinol., 119(4), 1876-1878.

O’SHAUGNESSY, P.J. & MURPHY, L. (1991). Steroidogenic enzyme activity in the rat testis . following Leydig cell destruction by ethylene-1,2-dimethanesulphonoate and during subsequent Leydig cell regeneration, y. E'zzz/ocrizzz?/., 131, 451-457.

O’SHAUGNESSY, P.J. & PAYNE, A.H. (1982). Differential effects of single and repeated administration of gonadotropins on testosterone production and steroidogenic enzymes in Leydig cell preparations. y.5zo/.CAem., 257(19), 11503-11509.

OTTO, S., BHATTACHARYYA, K.K. & JEFCOATE, C.R. (1992). Polycyclic aromatic hydrocarbon metabolism in rat adrenal, ovary and testis microsomes is catalysed by the same novel cytochrome P-450 (450RAP). Endocrinol., 131(6), 3067-3076.

PAKARINEN, P., NIEMIMAA, T., HUHTANIEMI, I.T. & WARREN, D.W. (1,994). Transcriptional and translational regulation of LH, prolactin and their testicular receptors by hCG and bromocriptine treatments in adult and neonatal rats. Mol.Cell Endocrinol., 101, 37-47.

PAPADOPOULOS, V. (1998). Structure and function of the peripheral-type benzodiazepine receptor in steroidogenic cells. Proc.Soc.Expt.Biol.&Med., 211{2), 130-142.

PARKINSON, A., CLEMENT, R.P., CASCIANO, C.C. & CAYEN, M.N. (1992). Evaluation of Loratadine as an inducer of liver microsomal cytochrome P450 in rats and mice. Biochem.Pharmacol., 43(10), 2169-2180.

252 PAYNE, A.H., YOUNGBLOOD, G.L., SHA, L., BURGOS-TRINIDAD, M. & HAMMOND, S.H. (1992). Hormonal regulation o f steroidogenic enzyme gene expression in Leydig cells. J.Steroid Biochem., 42{8), 895-906.

PAYNE, A.H., WONG, K-L. & VEGA, M.M. (1980a). Differential effects of single and repeated administration of gonadotropins on luteinising hormone receptors and testosterone synthesis in two populations of Leydig cells. J.BiolChem., 1S5{\5),1\\8-1\21.

PAYNE, A.H., DOWNING, J.R. & WONG, K-L. (1980b). Luteinising hormone receptors and testosterone synthesis in two distinct populations of Leydig cells. Endocrinol, 106(5), 1424-1429.

PEARCE, R.E., RODRIGUES, A.D., GOLDSTEIN, J.A. & PARKINSON, A. (1996). Identification o f the human P450 enzymes involved in lansoprazole metabolism. J.PharmacolExp.Ther., 277(2), 805-816.

PEDERSEN, R.C. & BROWNIE, A.C. (1987). Steroidogenesis-activator polypeptide isolated from a rat Leydig cell tumour. iSczence, 236, 188-190.

PELTOKETO, H., LUU-THE, V., SIMARD, J. & ADAMSKI, J. (1999). 17|3-Hydroxysteroid dehydrogenase (HSD)/17-ketosteroid reductase (KSR) family: nomenclature and main characteristics of the 17HSD/KSR enzymes. J.MolEndocrinol, 23, l-ll.

PFEIFFER, C.A. & HOOKER, C.W. (1943). Testicular changes resembling early stages in the development of interstitial cell tumours in mice of the A strain after long-continued injections of pregnant mare serum. Cancer Res., 3,162-166.

PIERCE, J.G. & PARSONS, T.F. (1981). Glycoprotein hormones: structure and ftmction. Ann.Rev.Biochem., 50, 465-495.

PICHARD, L., FABRE, L, FABRE, G., DOMERGUE, J., SAINT AUBER, B., MOURAD, G. & MAUREL, P. (1990). Cyclosporin A dmg interactions. Screening for inducers and inhibitors of cytochrome P450 (cyclosporin A oxidase) in primary cultures of human hepatocytes and in liver microsomes. Drug Metab.Disp., \8{5), 595-606.

PIRKE, K.M., BARANO, J.L., CALANDRA, R., LUTHY, I. & SPYRA, B. (1982). Influence of starvation on the dihydrotestosterone-luteinizing hormone feedback in the male rat. J.Steroid Biochem., 16,403-406.

POLLACK, S.E., FURTH, E.E., KALLEN, C.B., ARAKANE, F., KIRIAKIDOU, M., KOZARSKY, K.F. & STRAUSS, J.F. (1997). Localization of the steroidogenic acute regulatory protein in human tissues. J.Clin.EndocrinolMetab.,82{\2),A243-A25\.

PRAHALADA, S., MAJKA, J.A., SOPER, K.A., NETT ,T.M., BAGDON, W.J., PETER, C.P., BUREK, J.D., MACDONALD, J.S. & VAN ZWIETEN, M.J, (1994). Leydig cell hyperplasia and adenomas in mice treated with finasteride, a 5 alpha-reductase inhibitor: a possible mechanism. Fundam. Appl Toxicol., 22{2), 211-219.

PRENTICE, D.E., SIEGEL, R.A., DONATSCH, P., QURESHI, S. & ETTLIN, R.A. (1992). Mesulergine induced Leydig cell tumours, a syndrome involving the pituitary-testicular axis of the rat. Arch. Toxicol, Suppl. 15, 197-204.

PRIVALLE, C.T., CRIVELLO, J.F. & JEFCOATE, C.R. (1983). Regulation of intramitochondrial cholesterol transfer to side chain cleavage cytochrome P450 in rat adrenal gland. Proc. Natl Acad. Scl USA., 80, 702-706.

PURDON, M.P. & LEHMAN-McKEEMAN, L.D. (1997). Improved high-performance liquid chromatographic procedure for the separation and quantification of hydroxytestosterone metabolites. JPM., 37(2), 67-73.

253 PURVIS, K. & HANSSON, V. (1978). Hormonal regulation o f Leydig cell function. Mol.CellEndocrinoL, 12, 123-138.

PURVIS, J.L., CANICK, J.A., LATIF, S.A., ROSENBAUM, J.H., HOLOGGITAS, J. & MENNARD, R.H. (1973). Lifetime of microsomal cytochrome P450 and steroidogenic enzymes in rat testis as influenced by human chorionic gonadotropin. Arch.Biochem.Biophys., 159, 39.

QUIGLEY, C.A., DE BELLIS, A., MARSCHKE, K.B., EL-AWADY, M.K., WILSON, E.M. & FRENCH, F.S. (1995). Androgen receptor defects: historical, clinical and molecular perspectives. Endocr.Rev.,16,21\-32\.

RAUCY, J.L., KRNAER, J.C., LASKER, J.M. (1993). Bioactivation of halogenated hydrocarbons by cytochrome P4502E1. Crit.Rev.Toxicol.,13{\),\-20.

REINERINK, J.M., DOORN, L., JANSSEN, E.H.J.M. &N VAN lERSEL, A.A.J. (1991). Measurement of enzyme activities o f cytochrome P450 isozymes by high-performance liquid chromatographic analysis of products. y.C^romatog., 533,233-241.

RENWICK, A.B., MISTRY, H., BARTON, P.T., MALLET, F., PRICE, R.J., BEAMAND, J.A. & LAKE, B.G. (1999). Effect of some indole derivatives on xenobiotic metabolism and xenobiotic- induced toxicity in cultured rat liver slices. Food. Chem. Toxicol., 37, 609-618.

RISBRIDGER, G.P. (1993). Discrete stimulatory effects of platelet-derived growth factor (PDGF- BB) on Leydig cell steroidogenesis. Mb/.CeZZ Enz/ocrmoA, 97,125-128.

RISLEY, M.S., TAN, I.P., ROY, C. & Saez, J.C. (1992). Cell-, age-and stage-dependent distribution of connexin43 gap junctions in testes. 7. Ce//5cz., 103(1), 81-96. .

ROBERTS, S.A., ROBINSON, R.L., VAN RYZIN, R.J. & STOLL, R.E. (1993). Increased incidence of Leydig cell tumours in male rats with the dopamine agonist, SDZ205-502: correlation with increased semm LH levels. Toxicologist, 13, 292-297.

ROMAN, B.L., POLLENZ, R.S. & PETERSON, R.E. (1998). Responsiveness o f the adult male rat reproductive tract to 2,3,7,8-tetrachlorobenzo-p-dioxin exposure: Ah receptor and ARNT expression, CYPlAl induction and Ah receptor down regulation. ToxicolApplPharmacol., 150, 228-239.

ROMMERTS, F.F.G., DE JONG, F.H., BRINKMAN, A.O. & MOLEN, H.J. (1982). Development and cellular localisation of rat testicular aromatase acHvlty. J.Reprod.Fert., 65, 281-288.

ROSELLI, C.E. & RESKO, J.A. (1990). Regulation of hypothalamic luteinizing hormone releasing hormone levels by testosterone and estradiol in male rhesus monkeys. Brain Research, 509(2), 343- 346.

ROSNESS, P.A., SUNDE, A. & EIK-NES, K.B. (1977). Production and effects of 7a- hydroxytestosterone on testosterone and dihydrotestosterone metabolism in rat testis. Biochem.Biophys.Acta., 448, 55-68.

RUSSELL, L.D., CORBIN, T.J., REN, H.P., AMADOR, A , BARTKE, A. & GHOSH, S. (1992). Structural changes in rat Leydig cells posthypophysectomy: a morphometric and endocrine study. Endocrinol, 131, 498-508.

RUSTIA, M. & SHUBIK, P. (1979). Experimental induction o f hepatomas, mammary tumours, and Other tumours with metronidazole in non-inbred Sas:MRC(WI)BR rats. J. Natl. Cancer Inst., 63, 863- 868 .

RUTGERS, J.L. & SCULLY, R.E. (1991). The androgen insensitivity syndrome (testicular ' féminisation): a clinicopathological study of 43 cases. Int.J.Gynecol.Pathol, 10, 126-144.

254 RYAN, D.E., THOMAS, P.K., LEVIN, W., MAINES, S.L., BANDIERA, S. & REIK, L.M. (1993). Monoclonal antibodies of differential specificities as probes of cytochrome P450h (2C11). J.Biochem.Biophys., 301(2), 282-293.

SAEZ, J.M. (1994). Leydig cells: endocrine, paracrine and autocrine regulation. Endocrine Rev., 15(5), 574-626. \

SAEZ, J.M., HAOUR, F. & CAITHIARD, A.M. (1978a). Human chorionic gonadotropin-induced Leydig cell refractoriness to gonadotropin stimulation. Mol.Pharmacol, 14, 1054-1062.

SAEZ, J.M., HAOUR, F. & CAITHARD, A.M. (1978b). Early hCG-induced desensitisation in 'Leyd.xgzeWs. Biochem.Biophys.Res.Comm.,81, 552-558.

SCHENKMAN, J.B. (1992). Steroid Metabolism by constitutive cytochromes P450. J.Steroid Biochem.Mol.Biol.,42{8),\023-m0.

SCHENKMAN. J.B., THUMMEL, K.E. & FAVREAU, L.V. (1989). Physiological and pathological alterations in rat hepatic cytochrome P450. Drug Metab. Rev., 20(2-4), 557-584.

SCHUUR, A.G., BOEKHORST, P.M., BROUWER, A. & VISSER, T.J. (1997). Extrathyroidal effects of 2,3,7,8,-tetrachlorodibenzo-p-dioxin on thyroid hormone turnover in male Sprague-Dawley rats. Endocrinol, 138, 3727-3734.

SENG, J.E., GANDY, J., TURTURRO, A., LIPMAN, R. et al. (1996). Effects o f calorific restriction on testicular cytochrome P450 enzymes associated with the metabolic activation o f carcinogens. Arch.Biochem.Biophys., 335(1), 42-52.

SENG, J.E., LEAKEY, J.E., ARLOTTO, M.P., PARKINSON, A. & GANDY, J. (1991). Cellular localisation of cytochrome P4502A1 in testes of mature Sprague-Dawley rats. BiolReprod., 45, 876- 882.

SHARPE, R.M. (1994). Regulation of spermatogenesis. In. The Physiology o f Reproduction, eds. Knobif E. & Neil, J.D. Raven Press. New York.

SHARPE, R.M., MADDOCKS, S., MILLAR, M., KERR, J.B., SAUNDERS, P.T.K. & McKINNEL, C. (1992). Testosterone and spermatogenesis: identification of stage-specific, androgen-regulated proteins secreted by adult rat seminiferous tubules. J.vlnz/ro/., 13(2), 172-184.

SHARPE, R.M. & McNEILLY, A.S. (1979). The effect of induced hyperprolactinaemia on Leydig cell function and LH-induced loss of LH receptors in the rat testis. Mol. Cell Endocrinol, 16, 19-27.

SHIMADA, T., YAMAZAKI, H., MIMURA, M., INUI, Y. & GUENGERICH, F.P. (1994). Interindividual variations in human liver cytochrome P450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J.PharmacolExp.Ther., 210{\), 414-423.

SIBINSKI, L.J. (1987). Two-year oral (diet) toxicity/ carcinogenicity study o f flourochemical FC- 143 in rats. Riker Laboratories, Inc/3M Company.

SIMARD, J., LUTHY, I., GUAY, J., BELANGER, A. & LABRIE, F. (1986). Characteristics o f the interaction of the antiandrogen flutamide with the androgen receptor in various target tissues. MolCellEndocrinol, 44,261-210.

SIMPSON, E.R. (1979). Cholesterol side chain cleavage, cytochrome P450 and the control of steroidogenesis. Mol Cell Endocrinol, 13,213-217.

SIMPSON, M.E. & VAN WAGENEN, G. (1954). Persistent nodules in testis o f the monkey associated with Leydig cell hyperplasia induced by gonadotropins. Cancer Res., 14, 289-293.

255 SKINNER, M.K. & FRITZ, LB. (1985). Testicular peritubular cells secrete a protein under androgen control that modulates Sertoli cell functions. Proc.Natl.Acad.Sci., 82(1), 114-118.

SKINNER, M.K., McLACHLAN, R.I. & BREMNER, W.J. (1989). Stimulation of Sertoli cell inhibin secretion by the testicular paracrine factor PModS. Mol.CellEndocrinol., 66(2), 239-249.

SMITH, M.S. & BARTKE, A. (1987). Effects of hyperprolactinaemia on the control of luteinising hormone and follicle-stimulating hormone secretion in the male rat. Biol.Reprod., 36, 138-147.

SONDERFAN, A.J. & PARKINSON, A. (1988). Inhibition of steroid 5a-reductase and its effects on testosterone hydroxylation by rat liver microsomal cytochrome P450. Arch.Biochem.Biophys., 265(1), 208-218.

SONDERFAN, A.J., ARLOTTO, M.P. & PARKINSON, A. (1989). Identification of the cytochrome P450 isozymes responsible for testosterone oxidation in rat lung, kidney and testis. Endocrinol., 125(2), 857-866.

SONDERFAN, A.J., ARLOTTO, M.P., DUTTON, D.R., McMILLEN, S.K. & PARKINSON, A. (1987). Regulation of testosterone hydroxylation by rat liver microsomal cytochromes P450. Arch.Biochem.Biophys.,25S(\),21-4l.

SOUTHREN, A.L., GORDON, G.G., TOCHIMOTO, S., KRIKUN, E., KRIEGER, D., JACOBSON, M. & KUNTZMAN, R. (1969). Effect of iV-phenylbarbital on the metabolism o f testosterone and cor\.iso\mmd.XL J.Clin.Endocrinol.Metab., 29(2), 251-256.

STRESSER, D.M., BAILEY, G.S. & WILLIAMS, D.E. (1994). Indole-3-carbinol and P-naphthoflavone induction o f afltoxin B% metabolism and cytochromes P450 associated with bioactivation and detoxication of aflatoxin B, in the rat. Drug Metab.Disp., 22(3), 383-391.

STRIPP, B., MENARD, R.H. & GILLETTE, J.R. (1974). Effect o f chronic treatment with phénobarbital or 3-methylcholanthrene on the male reproductive system in rats. Life Sciences, 14, 2121-2130.

SUGAWARA, T., HOLT, J.A., DRISCOLL, D., STRAUSS, J.F., LIN, D., MILLER, W.L., PATTERSON, D., CLANCY, K.P., HART, I.M., CLARK, B.J. et al. (1995). Human steroidogenic regulatory protein: functional activity in COS-1 cells, tissue-specific expression, and mapping of the

structural gene to 8 pl 1.2 and a psuedogene to chromosome 13. Proc. Natl. Acad. Sci. USA., 92(11), 4778-4782. .

SUNDE, A., AARESKJOLD, H., HAUG, E. & EIK-NES, K.B. (1982). Synthesis and androgen effects of 7a, 17(3-dihydroxy-5 a-androstan-3 -one, 5 a-androstan-3 a, 17 p-triol and 5a-androstane- P,7a,17P-triol. 16,483-488.

TAPPER, C.M. & BROWN-GRANT, K. (1975). The secretion and metabolic clearance rates of oestradiol in the rat. y.E’«^/ocn«o/., 64(2), 215-227.

TATEISHI, T., NAKURA, H., ASOH, M., WATANABE, M., TANAKA, M., KUMAI, T. & KOBAYASHI, S. (1999). Multiple cytochrome P-450 subfamilies are co-induced with P- glycoprotein by both phenpthiazine and 2-acetylaminofluorene in rats. Cancer Letters, 138, 73-79.

TEERDS, K.J., ROMMERTS, F.F., VAN DER TWEEL, I. & WENSING, C.J. (1989a). Turnover time of Leydig cells and other interstitial cells in testes of adult rats. Arch.Androl., 23, 105-111.

TEERDS, K.J., ROMMERTS, F.F., VAN DER KANT, H.J. & DE GROOT, D.J. (1989b). Leydig cell number and function in the adult cynomolgus monkey is increased by daily hCG treatment but not by daily FSH treatment. J.Reprod. Fertil., 87, 141-146.

TENNISWOOD, M., ABRAHAMS, P., WINTERTON, V., BIRD, G.E., CLARK, A,F. (1982). Binding of testosterone, 5 alpha-dihydrotestosterone and 5 alpha-androstane (3 alpha- and 3 beta-), 17 beta-diols to serum proteins in the rat. J.Steroid Biochem.Mol.Biol, 16(5), 617-620.

256 THOMAS, P.E., BANDIERA, S., MAINES, S.L., RYAN, D.T. & LEVIN, W. (1987). Regulation of a high affinity N-Nitrosodiethylamine demethylase in rat hepatic microsomes. Biochem., 26, 2280- 2289.

THOMAS, P.E., RIEK, L.M., RYAN, D.E. & LEVIN, W. (1983). Induction o f two immunochemically related rat liver cytochrome P-450 isozymes, cytochromes P450c and P450d, by structurally diverse xenobiotics. J.5io/.CAem., 258(7), 4590-4598.

THOMAS, P.E., REIK, L.M., RYAN, D.E. & LEVIN, W. (1981). Regulation of three forms of cytochrome P450 and epoxide hydrolase in rat liver microsomes: effects of age, sex and induction. J.BioI. Chem., 256, 1044-1052.

THUMMEL, K.E., FAVREAU, L.V., MOLE, I.E. & SCHENKMAN, J.B. (1988). Further characterisation of RLM2 and comparison with a related form o f cytochrome P450e, RLM2b. Arch.Biochem.Biophys., 266, 319-333.

THURMAN, J.D., BUCCI, T., HART, R.W. & TURTURRO, A. (1994). Survival, body weight, and spontaneous neoplasms in ad libitum-fed and food-restricted Fischer-344 rats. Toxicol.PàthoL, 22, 1-9.

TREDGER, J.M., SMITH, H.M. & WILLIAMS, R. (1984). Effects of ethanol and enzyme-inducing agents on the monooxygenation o f testosterone and xenobiotics in rat liver microsomes. J.Pharmacol.Expt.Ther., 223, 292-298.

TSAI-MORRIS, C.H., AQUILANO, D.R. & DUFAU, M.L; (1985). Cellular localisation of rat testicular aromatase activity during development. E’n^/ocrmo/., 116,38-46.

TUREK, F.W. & DESJARDINS, C. (1979). Development of Leydig cell tumours and the onset of changes in the reproductive and endocrine system of ageing Fischer 344 rats. l.Natl. Cancer. Institute, 63, 969-975.

UTESCH, D., DIENER, B., MOLITER, E., OESCH, F. & PLATT, K.L. (1992). Characterisation of cryopreserved rat liver parenchymal cells by metabolism of diagnostic substrates and activities of related enzymes. Biochem.Pharmacol., 44(2), 309-315.

VALLADARES, L.E. & PAYNE, A.H. (1981). Effects ofhCG and cyclic AMP on aromatisation in purified Leydig cells of immature and mature rats. Biol.Reprod., 25, 752-758

VANDEN BOSSCHE, H. (1985). Biochemical targets for antifiingal azole derivatives: hypothesis on the mode of action. In. Current Topics in Medical Mycology, ed. McGinnis, M.R. Springer-Verlag, New York.

VERHOEVEN, G. & CAILLEAU, J. (1988). Testicular peritubular cells secrete a protein under androgen control that inhibits induction o f aromatase activity in Sertoli cells. Endocrinol., 123(4), 2100- 2110.

VIGUIER-MARTINEZ, M.C., HOCHEREAU DE REVIERS, M.T., BARENTON, B. & PERREAU, C. (1983a). Endocrinological and histological changes induced by flutamide treatment on the hypothalamo-hypophyseal testicular axis of the adult male rat and their incidences on fertility. Acta Endocrinol, 104(2), 246-252.

VIGUIER-MARTINEZ, M.C., HOCHEREAU DE REVIERS, M.T., BARENTON, B. & PERREAU, C. (1983b). Effect of a non-steroidal antiandrogen, flutamide, on the hypothalamo-pituitary axis, genital tract and testis in growing male rats: endocrinological and histological data. Acta Endocrinol, 102(2), 299-306.

VIHKO, K.K., LA POLT, P.S., NISHIMORI, K. & HSUEH, A.J. (1991). Stimulatory effects of recombinant follicle-stimulating hormone on Leydig cell function and spermatogenesis in immature hypophysectomized rats. 129(4), 1926-1932.

257 VISSER, T.J., KAPTEIN, E. & HARPUR, E.S. (1991). Differential expression and ciprofibrate induction of hepatic UDP-glucuronyltiansferases for thyroxine and triiodothyronine in Fischer rats. Biochem.Pharmacol., 42(2), 444-446.

WAHLSTROM, T., HUHTANIEMI, L, HOVATTA, O. & SEPPALA, M. (1983). Localisation of luteinising hormone, follicle-stimulating hormone, prolactin, and their receptors in human and rat testis using immunohistochemistry and radioreceptor assay. J.Clin.Endocrinol. Metab., 57(4), 825-830.

WALKER, N.J., GASTEL, J.A., COSTA, L.T., CLARK, G.C., LUCIER, G.W. & SUTTER, T.R. (1995). Rat CYPIBI: an adrenal CYP450 that exhibits sex-dependent expression in livers and kidneys ofTCDD-treated animals. Carcinogenesis, 16(6), 1319-1327.

WANG, R.W. & LU, A.Y.H. (1997). Inhibitory anti-peptide antibody against human CYP3A4. Drug Metab. Disp., 25(6), 762-767.

WANG, D.Y. & BULBROOK, R.D. (1967). The metabolic clearance rates of dihydroepiandrosterone, testosterone and their sulphate esters in man, rat and rabbit. J.Endocrinol., 38,307-318.

WANG, H., SEGALOFF, D.L. & ASCOLI, M. (1991). Lutropin/choriogonadotropin down regulates its receptor by both receptor-mediated endocytosis and a cAMP-dependent reduction in receptor messenger RNA. y..5/o/.C^ew., 266, 780-785.

WANG, N.G., SUNDARAM, K., PAVOU, S., RIVIER, J., VALE, W. & BARDIN, C.W. (1983). Mice are insensitive to the anti-testicular effects of luteinising hormone-releasing hormone agonists. Endocrinol., 112, 2)11-21)5.

WATERMAN, M. & SIMPSON, E. (1985). Regulation of the biosynthesis o f cytochrome P450 involved in steroid hormone synthesis. A/b/.Ce//.E’«ifocnno/., 39, 81-89.

WAXMAN, D.J. (1992). Regulation o f liver-specific steroid metabolizing cytochromes P450: cholesterol 7a-hydroxylase, bile acid 6(3-hydroxylase, and growth hormone-responsive steroid hormone hydroxylases. Steroid.Biochem.Mol.BioL,‘11(Z), \Q55-\0n2.

WAXMAN, D.J. (1991). Rat hepatic P450IIA and P450IIC subfamily expression using catalytic, immunochemical and molecular probes. Methods Enzymol., 206, 249-267.

WAXMAN, D.J. (1988). Interactions of hepatic cytochromes P450 with steroid hormones. Biochem.Pharmacol., 37(1), 71-84.

WAXMAN, D.J. (1984). Rat hepatic cytochrome P450 isozyme 2c: identification as a male-specific, developmentally-induced steroid 16 alpha-hydroxylase and comparison to a female-specific cytochrome P450 isoenzyme. y.5fo/.CAcm., 259,15481-15490.

WAXMAN, D.J. & AZAROFF, L. (1992). Phénobarbital induction of cytochrome P450 gene expression. Biochem.J., 281, 577-592.

WAXMAN, D J., PAMPORI, N.A., RAM, P.A., AGRAWAL, A.K. & SHAPIRO, B.H. (1991a). Interpulse interval in circulating growth hormone patterns regulates sexually dimorphic expression of hepatic cytochiomePiSO. Proc. Natl.Acad.Sci.USA.,SS,6S6S-6S72.

WAXMAN, D.J., LAPENSON, D.P., AOYAMA, T., GELBOIN, H.V., GONZALEZ, F.J. & KORZEKWA, K. (1991b). Steroid hormone hydroxylase specificities of eleven cDNA-expressed human cytochrome P 4 5 0 s . 290(1), 160-166.

WAXMAN, D.J., LAPENSON, D.P., NAGATA, K. & CONLON, H.D. (1990). Participation of two structurally related enzymes in rat hepatic microsomal androstenedione 7a-hydroxylation. Biochem. J., 265, 187-194. ,

258 WAXMAN, DJ., LE BLANC, G.A., MORRISSEY, J.J., STAUNTON, J. & LAPENSON, D.P. (1988a). Adult male-specifîc and neonatally programmed rat hepatic P450 forms RLM2 and 2a are not dependent on pulsatile plasma growth hormone for expression. J.Biol.Chem., 263(23), 11396- 11406.

WAXMAN, D.J., ATTISANO, C., GUENGERICH, P.P. & LAPENSON, D.P. (1988b). Human liver microsomal steroid metabolism: identification of the major microsomal steroid hormone 6(3- hydroxylase cytochrome P450 enzyme. Arch.Biochem.Biophys., 263(2), 424-436.

WAXMAN, D.J., LAPENSON, D.P., PARK, S.S., ATTISANO, C., GELBOIN, H.V. (1987). Monoclonal antibodies inhibitory to rat hepatic cytochromes P450: P450 form specificities and use as probes for cytochrome P450-dependent steroid hydroxylations. Mol.Pharmacol, 32, 615-624.

WAXMAN, D.J., DONNAN, G.A. & GUENGERICH, P.P. (1985). Regulation of rat hepatic CYP450: age-dependent expression, hormonal imprinting and xenobiotic inducibility o f sex-specific isoenzymes. 5/oc/ie/M., 24,4409-4417.

WEINBAUER, G.F. & WESSELS, J. (1999). Paracrine control of spermatogenesis. Andrologia, 31, 249-262.

WELSH, T.H., KASSON, B.G., HSEUH, A.J. (1986). Direct biphasic modulation o f gonadotropin- stimulated testicular androgen biosynthesis by prolactin. Biol.Reprod., 34, 796-804.

WEUSTEN, J.J., SMALS, A.G., HOFMAN, J.A., KLOPPENBORG, P.W. & BENRAAD, T.J. (1987). Early time sequence in pregnenolone metabolism to testosterone in homogenates of human , and rat testis. E’nJocrmo/., 120 (5), 1909-1913.

WILLIAMS, T.M. & BORGHOFF, S.J. (2000). Induction of testosterone biotransformation enzymes following oral administration of methyl fert-butyl ether to male Sprague-Dawley rats. Toxicol.Sci., 57, 147-155.

WILLIAMS, T.M., CATTLEY, R.C. & BORGHOFF, S.J. (2000). Alterations in the endocrine responses in male Sprague-Dawley rats following oral administration of methyl tert-h\ity\ ether. Toxicol.Sci., 54, 168-176.

WILSON, V.S. & LE BLANC, G.A. (1998). Endosulfan elevates testosterone biotransformation and clearance in CD-I mice. Toxicol.Appl.Pharmacol.,14^,\5Z-\6%.

WILSON, V.S., McLACHLAN, J.B., FALLS, J.G. & LE BLANC, G.A. (1999). Alteration in sexually dimorphic testosterone biotransformation profiles as a biomarker of chemically-induced androgen disruption in mice. Env.Health Persp., 107(5), 377-384.

WILSON, N.M., CHRISTOU, M. & JEFCQATE, C.R. (1987). Differential expression and function of three closely related phenobarbital-inducible cytochrome P450 isozymes in untreated rat liver. Arch.Biochem.Biophys., 256(2), 407-420.

WINER, L.M., SHAW, M.A. & BAUMANN, G. (1990). Basal plasma growth hormone levels in man: new evidence for rhythmicity of growth hormone secretion. J.Clin.endocrinol.Metab., 70(6), 1678-1686.

WING, T.Y., EWING, L.L. & ZIRKIN, B.R. (1984). Effects o f luteinising hormone withdrawal on Leydig cell smooth endoplasmic reticulum and steroidogenic reactions which convert pregnenlone to testosterone. Endocrinol., 115(6), 2290-2296.

WINTERS, S.J. & LORIAUX, D.L. (1978). Suppression of plasma luteinising hormone by prolactin in the male rat. Endocrinol., 102, 864-868.

WOOD, A.W., RYAN, D.E., THOMAS, P.E. & LEVIN, W. (1983). Regio- and stereoselective metabolites of two C 19 steroids by five highly purified and reconstituted rat hepatic CYP450 isozymes. 258(14), 8839-8847.

259 WORTELBOER, H.M., DE KRUIF, C.A., VAN lERSEL, A.J., FALKE, H.E., NOORDHOEK, J. & BLAAUBOER, B.J. (1992). Acid reaction products of indole-3-carbinol and their effects on cytochrome P450 and phase II enzymes in rat and monkey hepatocytes. Biochem.Pharmacol, 43(7), 1439-1447.

WORTELBOER, H.M., DE KRUIF, C.A., VAN lERSEL, A.J., FALKE, H.E., NOORDHOEK, J. & BLAAUBOER, B.J. (1991). Comparison o f cytochrome P450 isoenzyme profiles in rat liver and hepatocyte cultures. 5ioc^em.PAnrmaco/., 42(2), 381-390.

WRIGHT, K. & MORGAN, E.T. (1991). Regulation of cytochrome P450 2C12 expression by interleukin-la, interleukin-6 and dexamethasone. 39, 468.

YAMADA, T., MAITA, K., NAKAMURA, J., MURAKAMI, M., OKUNO, Y., HOSOKAWA, S., MATSUO, M. & YAMADA, H. (1994a). Carcinogenicity and oncogenicity studies of oxolinic acid in rats and mice. Food Chem. Toxicol., 32, 397-408.

YAMADA, T., NAKAMURA, J., MURAKAMI, M., OKUNO, Y., HOSOKAWA, S. MATSUO, M. & YAMADA, S. (1994b). The correlation of serum luteinizing hormone levels with the induction of Leydig cell tumors in rats by oxolinic acid. Toxicol.Appl.Pharmacol., 129, 146-154.

YAMAMOTO, R., KALLEN, C.B., BABALOLA, G.O., RENNERT, H., BILLHEIMER, J.T. & . STRAUSS, J.F. (1991). Cloning and expression of a cDNA encoding human sterol carrier protein 2. Proc.Natl.Acad.Sci.USA., SS{2), 463-467.

YAMAZAKI, H. & SHIMADA, T. (1997). Progesterone and testosterone hydroxylation by cytochromes P450 2C19, 2C9 and 3A4 in human liver microsomeS. Arch.Biochem.Biophys., 346(1), 161-169.

ŸASUDA, Y., OHARA, I., KONISHI, H. & TANIMURA, T. (1988). Long-term effects on the male reproductive organs of prenatal exposure to ethinyl estradiol. Am. J.Obstet. Gynecol., 159,1246-50.

YEOWELL, H.N., WAXMAN, D.J., WADHERA, A. & GOLDSTEIN, J.A. (1987). Suppression of the constitutive male-specific rat hepatic cytochrome P450 2c and its mRNA by 3,4,5,3’,4’,5’- hexachlorobiphenyl and 3-methylcholanthrene. Mol.Pharmacol., 32, 340-347.

YUAN, P.M., RYAN, D.E., LEVIN, W. & SHIVLEY, I.E. (1983). Identification and localization of amino acid substitutions between two phenobarbital-inducible rat hepatic microsomal cytochromes P450 by micro sequence analyses. Proc.USJ., 80(5), 1169-1173.

ZANGAR, R.C., WOODCROFT, K.J. & NOVAK, R.F. (1996). Differential effects of ciprofibrate on renal and hepatic cytochrome P450 2E1 expression. Toxicol.Appl.Pharmacol., 141,110-116.

ZHOU, Z.X., LANE, M.V., KEMPPAINEN, J.A., FRENCH, F.S. & WILSON, E.M. (1995). Specificity of ligand-dependent androgen receptor stabilisation: receptor domain interactions influence ligand dissociation and receptor stability. A/o/.Pn^/ocrmo/., 9(2), 208-218.

ZIMNIAK, P. & WAXMAN, D.J. (1993). Liver cytochrome P450 metabolism of endogenous steroid hormones, bile acids and fatty acids. In. Cytochrome P450. ed. Schenkman, J.B. & Greim, H. Springer Verlag.

ZIPF, W.B., PAYNE, A.H. & KELCH, R.P. (1978a). Prolactin, growth hormone and luteinising hormone in the maintenance o f testicular luteinising hormone receptors. Endocrinol., 103 (2), 595- 600.

ZIPF, W.B., PAYNE, A.H. & KELCH, R.P. (1978b). Dissociation of lutropin-induced loss of testicular lutropin receptors and lutropin-induced desensitisation o f testosterone synthesis. Biochim.Biophys.Acta., 540, 330-336.

260 ZIRKIN, B.R., SANTULLI, R., AWONIYI, C.A. & EWING, L.L. (1989). Maintenance of advanced spermatogenic cells in the adult testis: quantitative relationship to testosterone concentration within the testis. Endocrinol., 124, 3043-3049.

ZYLBER-HARAN, E.A., GERSHMAN, H., ROSENMANN, E. & SPITZ, I.M. (1982). Gonadotropin, testosterone and prolactin interrelationship in cadmium-treated rats. J.EndcrinoL, 92, 123-130.

261

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.