Investigation of the Involvement of Covalent Binding in Nevirapine-Induced Hepatic and Cutaneous Idiosyncratic Adverse Drug Reactions

Transcription

1 Investigation of the Involvement of Covalent Binding in Nevirapine-Induced Hepatic and Cutaneous Idiosyncratic Adverse Drug Reactions by Amy M. Sharma A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Graduate Department of Pharmaceutical Sciences Faculty of Pharmacy University of Toronto Copyright by Amy M. Sharma, 2013

2 Investigation of the Involvement of Covalent Binding in Nevirapine- Induced Hepatic and Cutaneous Idiosyncratic Adverse Drug Reactions Amy M. Sharma Doctor of Philosophy Abstract Graduate Department of Pharmaceutical Sciences Faculty of Pharmacy University of Toronto 2013 Nevirapine (NVP) can cause serious idiosyncratic drug reactions (IDRs); specifically, skin rash and hepatotoxicity. Treatment of rats or mice with NVP led to covalent binding to hepatic proteins. Studies of this covalent binding including the use of a deuterated analog of NVP leading to a decrease in oxidation of the methyl group indicated that the metabolite responsible for covalent binding in the liver is a quinone methide. Covalent binding in NVP-treated rats was also observed in the epidermis but by a different pathway. Incubation of 12-OH-NVP sulfate with homogenized human and rat skin led to extensive covalent binding. Inhibition of sulfation in the liver significantly decreased 12-OH-NVP sulfate in the blood, but it did not prevent covalent binding in the skin or the rash. In contrast, topical application of a sulfotransferase inhibitor prevented covalent binding in the skin as well as the rash, but only where it was applied. In contrast to rats, treatment of mice with NVP did not result in covalent binding in the skin or skin rash. These findings provide compelling evidence that 12-OH-NVP sulfate formed in the skin is responsible for the skin rash. IL-1β and IL-18 production in the skin of rats treated with NVP were increased. An anti- IL-1ß antibody significantly decreased rash severity. These cytokines were also produced by incubation of human keratinocytes with 12-OH-NVP sulfate. These data indicate that 12-OH- NVP sulfate activates the NLRP3 inflammasome, a pathway known to be responsible for contact hypersensitivity rashes. ii

3 In summary, NVP was found to produce two different reactive metabolites, a quinone methide species in the liver, and a benzylic sulfate in the skin. Significant liver injury did not occur, presumably due to immune tolerance. Although it is usually assumed that reactive metabolites are responsible for most IDRs, this is the first example to actually demonstrate that a specific reactive metabolite is responsible for an IDR. This is also the first study to show that sulfotransferase in the skin is responsible for bioactivation of a drug leading to a skin rash. It is likely that there are other drugs that cause skin rashes by a similar mechanism. iii

4 Acknowledgments I can't go back to yesterday because I was a different person then. - Lewis Carroll, Alice in Wonderland This thesis is truly the culmination of years of work on the nevirapine project, which I have had the privilege to study. If I am to be completely honest, it also represents my blood, sweat and tears. Jack, I want you to know that I have the deepest respect for you, and I am eternally grateful to you for the work you have allowed me to do in your lab. I enjoyed every moment I spent working on my project, and without your support, this work would not be. I was just so happy to do the science. Each day, I would embark on my 1.5 hour commute and think how lucky I was to have such an opportunity, such a beautiful lab to work in, under such renowned supervision. I never once took this journey for granted and each day we became smarter, wiser, better. It will be hard to move on and work in another lab after having you Jack, an amazing person and scientist, as my mentor. The science will of course always be with me but it is the life lessons I will always hold near. Thank you for never wavering in the project and remaining true to the science in the end, we did succeed, but it was not without much struggle. I am evermore thankful for the experience at having failed, failed, and failed again, only to have the hard work and discipline prevail. In my wildest dreams I never thought the project would be this interesting and so rewarding - I guess that s what happens when you work with pure abandon from the gut of your soul. I also want to thank you for the mental attitude I developed through the tough times. They say your attitude dictates 99% of your outcome in life and the resilience, problem solving and patience I developed working for you, I am grateful for. Most of all Jack, thank you for trusting me with this work, and giving me the freedom to pursue science to its fullest. I hope we can always collaborate in the future and share science, advice, and laughter. I would like to thank my committee members, Dr. Peter Wells, Dr. Mario Ostrowski, Dr. Peter Pennefather and Dr. Neil Shear for their valuable guidance over the years. Thanks also to Dr. Tony Hayes, as you first introduced me to the field of IDRs with that assignment on troglitazone that changed my life thank you for your support. Thanks also to Dr. Peter O Brien, Dr. Jeff Henderson, Dr. David Dubins, and Dr. Ray Reilly, who let me teach and lecture for them, through which I gained valuable professional experience. iv

5 To everyone else who played a part in this story to lab members, especially Ms. Maria Novalen, Dr. Yan Li, and Ms. Sandrine Fischer, for their foundational work on the project. The many animals that were sacrificed for this work, your lives were not in vain. To my good and dear friends, Stephanie MacAllister, Lutfiya Miller, Maya Latif, and Raza Mirza, who became my support system thank you deeply. To Dr. Henderson, for taking me in as an honorary member of his lab, where I would go to hide out and just think. Thank you, Jeff. To Dr. Dana Philpott, for her invaluable input on the NLRP3 studies and becoming one of my role models in the field of immunology. Dana you are now also my trusted colleague and friend and I look up to you so much, such a strong female scientist who makes it all so look so easy (Dana, you re a superstar!). When I grow up, I hope I am the kind of mentor and inspiration you have been to me. To Dr. Lance R. Pohl, I have the deepest regard for your scientific style and your work, and I want to thank you for introducing me to the origins of our field. I hope we can collaborate in the future and one day determine the root causes of DILI. To my new mentor and great scientist, a person I respect and admire very much, Dr. Ruslan M. Medzhitov - I have never met such a larger-than-life scientist so full of generosity and humble spirit. You have welcomed me so kindly into your entire world at Yale and I am incredibly excited to begin my next chapter in your group. I hold you in such high esteem; you are a scientist so ahead of his time, recognizing problems in nature before others even begin to realize they exist. I hope that we will accomplish much and I want to thank you for this golden, once-ina-lifetime opportunity! To KD, thank you for your love and support in the toughest of times. You are my best friend and I could not have done this without your encouragement. Finally, to Nika and my parents, for knowing that I could do it, even when I questioned if I could. Mom and Dad you both live for your children, and in turn, our success is your most revered reward. Mom, you are my biggest fan and I am grateful for your care your unconditional acceptance for your children means the world to me. Dad, I know you came here with eight bucks in your pocket but I hope you know it was worth it thank you for the life you have given us. I dedicate this thesis to you both. You are my inspiration for all things good and true in this world, and I love and admire you both more than words could ever capture. v

6 Table of Contents Abstract... ii Acknowledgments... iv List of Figures... xi List of Tables... xxi List of Schemes... xxii List of Abbreviations... xxiii List of Thesis Publications... xxiv CHAPTER 1: INTRODUCTION Adverse Drug Reactions An Overview Types of Adverse Drug Reactions Idiosyncratic Drug Reactions (IDRs) Proposed Mechanisms of Idiosyncratic Drug Reactions Hapten Hypothesis Danger Hypothesis Pharmacological Interaction Hypothesis Idiosyncratic Hepatotoxicity Hepatic Function and Morphology Biotransformation in the Liver Liver Immune System in Relation to Hepatotoxicity Idiosyncratic Cutaneous Toxicity Skin Structure and Function Metabolic Enzymes in Skin Immune Function of the Skin Implications of Cutaneous Biotransformation on Immune-Mediated Skin Rashes Role of Drug Metabolism, Reactive Metabolites, and Covalent Binding in IDRs Sulfotransferase Enzymes in Drug Metabolism and Toxicity Nevirapine Toxicity Animal Model of Nevirapine-Induced Skin Rash Nevirapine Metabolism leading to Skin Rash Role of Sulfation in Nevirapine Skin Rash Nevirapine Metabolism in the Liver and Lack of Hepatotoxicity in Rats Research Hypotheses vi

7 CHAPTER 2: Bioactivation of Nevirapine to a Reactive Quinone Methide: Implications for Liver Injury Abstract Introduction Materials and Methods Chemical Materials Instruments and Software Synthesis of 12-trideutero-NVP (DNVP) Production of Anti-NVP Anti-Serum in Male White New Zealand Rabbits Animal Care Treatment of Animals with NVP, 12-OH-NVP, DNVP, or ABT Incubations with Microsomes or Supersomes Quantification of NVP and its Metabolites from Microsomal Incubations Mass Spectrometry Analysis Analysis of Covalent Binding Using SDS-PAGE and Immunoblotting Analysis of in Vivo Covalent Binding using Immunohistochemistry Plasma Alanine Transaminase and Cytokine Analysis Results Characterization of the Anti-NVP-NAC-KLH Antiserum Covalent Binding of NVP, DNVP, or 12-OH-NVP to Hepatic Microsomes in Vitro and Comparison to in Vivo Hepatic Covalent Binding Covalent Binding of NVP to Expressed Rat CYP2C11 or CYP3A1 Supersomes, or of NVP, DNVP, or 12-OH-NVP to Human Hepatic Expressed CYP3A4 Supersomes Covalent binding of NVP or 12-OH-NVP to Hepatic Proteins from Female BN Rats Treated with NVP or 12-OH-NVP Immunohistochemistry of Liver from NVP- or DNVP-Treated or NVP + ABT Cotreated Female BN Rats Oxidation of NVP or 12-OH-NVP by Rat Liver Microsomes Covalent Binding, Serum ALT levels, INF-γ, and IL-6 Levels in Mice Liver Histology and ALT in Male Cbl-b -/- or C57BL/6 Mice Treated with NVP Comparison of Hepatic Covalent Binding of NVP between Mice and Female BN Rats Discussion vii

8 CHAPTER 3: Nevirapine Bioactivation and Covalent Binding in the Skin Abstract Introduction Materials and methods Chemicals Animal Care Primary and Secondary Treatment of Animals with NVP or 12-OH-NVP Separation of Dermis and Epidermis and Preparation of Homogenates of Skin Fractions or of Whole Rat Skin Preparation of Cytosol, S9, or Microsomes from NVP-treated Rat Epidermal or Dermal Fractions Preparation of Human Skin Dermatome Cytosolic or S9 Fractions from Rat or Human Skin or Liver Incubation of Human or Rat Skin or Fractionated Skin with NVP, 12-OH-NVP, or 12- OH-NVP Sulfate In Vitro Metabolism of 12-OH-NVP and NVP Covalent Binding Using SDS PAGE and Immunoblotting Preparation of BN Rat Skin for Histology Results Attempts to Detect In Vivo Covalent Binding in Whole Skin Covalent Binding of NVP, 12-OH-NVP, or 12-OH-NVP Sulfate to Human or Rat Skin in Vitro Covalent Binding of NVP to Rat Skin in Vivo Early Histological Changes in the Skin in Response to NVP Treatment Covalent Binding of NVP to Mouse Skin in Vivo Covalent Binding of NVP to Subcellular Rat Skin Fractions in Vivo Covalent Binding of 12-OH-NVP to Human or Rat Liver or Skin Proteins in the Presence or Absence of PAPS Covalent Binding of 12-OH-NVP or NVP to Human or Rat Liver or Skin Subcellular Fractions in the Presence or Absence of PAPS or NADPH Sulfation and Oxidation of NVP and 12-OH-NVP in Mouse and Rat Skin Anti-NVP and Autoantibodies in NVP-Treated Rats Discussion Supplemental Material Separation of Dermis and Epidermis of the Ear and Preparation of Homogenates viii

9 3.6.2 Grading of Skin Rash Covalent Binding and Histology in the Ears from NVP- or 12-OH-NVP-Treated Rats CHAPTER 4: 12-OH-Nevirapine Sulfate, Formed in the Skin, is Responsible for Nevirapine-Induced Skin Rash Abstract Introduction Materials and methods Chemical Materials and Reagents Animal Care Quantification of NVP, 12-OH-NVP, 12-OH-NVP Sulfate, and 4-COOH-NVP in Plasma Sulfation Inhibition Studies Separation of Skin Dermis and Epidermis and Preparation of Homogenates Preparation of Human Skin Dermatome Incubation of Human Skin or Expressed Human SULT 1A1*1 with 12-OH-NVP or 12- OH-NVP Sulfate, With or Without PAPS Incubation of Rat Liver or Skin Cytosol or Human Liver Cytosol with 12-OH-NVP and 1-Phenyl-1-hexanol in the Presence and Absence of PAPS Covalent Binding Using SDS-PAGE and Immunoblotting Preparation of BN Rat Skin for Histology Results General Scheme to Study the Role of 12-OH-NVP Sulfation on the Skin Rash Effect of Salicylamide on 12-OH-NVP Sulfate Levels and Rash Effects of DHEA on NVP Metabolism and Skin Rash Effects of 1-Phenyl-1-Hexanol on Covalent Binding and Rash In Vitro Inhibition of Covalent Binding by 1-Phenyl-1-Hexanol Discussion Supplemental Material Quantification of NVP, 12-OH-NVP, 12-OH-NVP Sulfate, and 4-COOH-NVP in Urine Grading of Skin Rash ix

10 CHAPTER 5: Discussion, Conclusions and Summary Hypotheses Revisited Discussion and Limitations Future Directions Bibliography x

11 List of Figures CHAPTER 1 Figure 1-1. Reaction of penicillin with protein nucleophiles via spontaneous ring opening in a hapten type mechanism. Figure 1-2. The Hapten Hypothesis: a reactive, unmetabolized parent drug, or reactive metabolite, conjugates to protein in a covalent manner. Upregulation of appropriate costimulatory molecules allows for the elicitation of an immune response. Adapted from Uetrecht, Figure 1-3. The Danger Hypothesis: stressed or damaged cells release endogenous danger signals which activate APCs, leading to upregulation of costimulatory molecules (B7 on APCs) which interact with CD28 on T cells leading to an immune response. Adapted from Uetrecht, Figure 1-4. The P-I Hypothesis: A reactive parent drug binds directly to the MHC-TCR complex in a reversible manner (signal 1), leading to an immune response. Adapted from Uetrecht, 2007 Figure 1-5. Proposed pathway of drug bioactivation, leading to haptenation and cellular mechanisms of hepatocyte death. Adapted from Kaplowitz, Figure 1-6. Proposed mechanisms of DILI, including drug bioactivation leading to reactive intermediate which may cause hepatocyte damage invoking immune mediated responses. A balance of hepatoprotective and inflammatory factors dictate the potential for toxicity. Adapted from Holt and Ju, Figure 1-7. Cellular composition of the skin. Taken from Feldmeyer et. al, xi

12 CHAPTER 2 Figure 2-1. Bioactivation pathway of NVP leading to liver injury. Figure 2-2. ELISA analysis showing (A) binding of the anti-nvp-nac-klh antiserum to the NVP-NAC-BSA conjugate, KLH, or BSA and (B) the effect of preincubation of the antiserum with NVP or its metabolites on the binding of the antisera to the NVP-NAC-BSA conjugate. Data represent the mean ± s.d. from 3 incubations. Figure 2-3. (A) Comparison of covalent binding of 12-OH-NVP (lanes 2, 5) and DNVP (lane 3, 6) with that of NVP (lane 4, 7) after a 30 or 60 min incubation with male BN rat microsomes (1 mg/ml protein) at a drug concentration of 1 mm. For comparison, covalent binding to hepatic proteins is shown after 8 days of treatment of female rats with 12-OH-NVP (159 mg/kg/day, lane 9) or NVP (150 mg/kg/day, lane 10). Protein loading was 15 µg for lanes 1-7 and 20 µg for lanes (B) Comparison of covalent binding of 12-OH-NVP (lanes 2, 5) and DNVP (lanes 3, 6) with that of NVP (lanes 4, 7) at a concentration of 1 mm after a 30 or 60 min incubation with microsomes (1 mg/ml protein) from male C57BL/6 mice. For comparison, covalent binding to hepatic proteins is shown after 6 weeks of treatment of C57BL/6 mice with NVP at a dose of 950 mg/kg/day in food. Protein loading was 13 µg for lanes 1-7 and 20 µg for lanes 8-9. (C) Comparison of covalent binding of NVP to hepatic microsomes from male C57BL/6 mice (lanes 2-4) or male BN rats (lanes 6-8) after a 15, 30, or 60 min incubation at a drug concentration of 1 mm and microsome concentration of 1 mg/ml protein. Protein loading was 20 µg per lane. The primary antiserum dilution was 1:500 and that of the secondary antisera was 1:5000. Figure 2-4. Covalent binding of NVP to expressed male rat CYP2C11 (A) or CYP3A1 (B) in vitro. Protein concentration for each incubation was 0.8 mg/ml with 0.5 mm of drug. For immunoblots, protein loading was 9 µg and 7.5 µg per lane for A and B, respectively. (+) indicates incubations containing NVP while ( ) indicates incubations lacking NVP. Proteins were resolved on 12% gels with 1:100 dilution of primary anti-serum followed by 1:2000 dilution of secondary antisera. Comparison of covalent binding of 12-OH-NVP (lanes 2, 5), or DNVP (lanes 3, 6) with that of NVP (lanes 4, 7) to human CYP3A4 with a drug concentration of 1 mm and protein concentration in each incubation of 1 mg/ml (C). Proteins (10 µg/lane) were resolved on an 8% gel. Dilutions of antisera were 1:500 for the primary anti-serum and 1:5000 for the secondary antisera. xii

13 Figure 2-5. (A) Covalent binding to hepatic proteins from female BN rats fed NVP (150 mg/kg) or 12-OH-NVP (159 mg/kg) for 8 days. Protein loading was 12 µg per lane. Samples were resolved on an 8% gel. A 1:500 dilution of primary anti-serum was followed by 1:5000 dilution of secondary antisera. (B) Incubation of the anti-nvp serum with 2 mm NVP for 2 h at 37 C blocked most of the binding (left side of panel) to samples from livers of 12-OH or NVP treated rats. Samples for both panels A and B were prepared, run, blocked, incubated with secondary antibody, and imaged at the same time and protein loading was 10 µg/well of protein per lane. Figure 2-6. Immunohistochemistry of liver sections from female BN rats; blank control, NVP treatment (150 mg/kg/day x 7 days in food), DNVP treatment (150 mg/kg/day x 7 days in food), ABT treatment (50 mg/kg/day x 28 days by gavage), or NVP (150 mg/kg/day) + ABT (50 mg/kg/day) x 28 days by gavage. Slides were incubated with 1:100 dilution of primary antisera and 1:2000 dilution of the secondary antisera. The slides were counterstained with Mayer s hematoxylin, magnification 20x. Figure COOH-NVP concentrations from incubations of 12-OH-NVP with microsomes from male (n = 3) and female (n = 1) BN rats. Figure 2-8. (A) Changes in ALT in male C57BL/6 mice treated with NVP (950 mg/kg/day) for 4 weeks. Values are based on the mean of triplicate readings per time point per animal ± S.D, n=5 treated mice or n = 4 control mice. Unpaired t-test, 7 d.f., p<0.05. (B) Corresponding covalent binding of NVP at the same dose in male BALB/c (n=2) or C57BL/6 (n=3) mouse livers after 6 weeks of treatment. Protein loading was 20 µg per lane. Samples were resolved on an 8% gel. Figure 2-9. (A) Plasma ALT levels in male Cbl-b -/- mice fed NVP orally for 14 days (950 mg/kg/day). Values are based on the mean of triplicate readings per time point per animal ± S.D, n=5 treated mice or n = 4 control mice. Unpaired t-test, 7 d.f., p<0.05. (B) Covalent binding of NVP in the livers of the same Cbl-b -/- mice. (C) Plasma ALT levels in NVP-treated (950 mg/kg/day) female Cbl-b -/- mice, n=4 treated or n=4 control mice. Values are based on the mean of triplicate readings per time point per animal ± S.D, n=5 treated mice or n = 4 control mice. Unpaired t-test, 6 d.f., p<0.05. (D) Covalent binding of NVP in the livers of the same mice. Protein loading was 25 µg per lane. Samples were resolved on 10-20% gradient gels. A 1:500 dilution of primary antisera followed by 1:5000 dilution of secondary antisera was used. xiii

14 Figure Serum IL-6 (A) or IFN- (B) from control and NVP-treated Cbl-b -/- mice at day 7 of NVP treatment. Animals showing gross necrosis are displayed separately. Figure H&E staining of livers from Cbl-b -/- mice treated with NVP for 2 weeks. (A) Untreated control liver with normal ALT; (B) liver from a NVP-treated mouse with gross necrosis and an ALT of 271 U/L, and (C) liver from another NVP-treated mouse with gross necrosis and ALT of 313 U/L. Areas of massive hepatocyte necrosis surrounded by viable hepatocytes are shown in (B) and (C). Figure H&E staining of livers from male C57BL/6 mice treated with NVP for 3 weeks. (A) Untreated control liver with a normal ALT; (B) liver from a NVP-treated mouse with very mild necrosis (appearing as the thin band around the capsule) and ALT of 94 U/L, and (C) liver from another NVP-treated mouse with an ALT of 75 U/L. Changes to the liver parenchyma due to enlargement of hepatocytes in the periacinar regions and extensive expansion of the endoplasmic reticulum are also present in both (B) and (C). Figure Comparison of covalent binding of NVP to hepatic proteins in mice and rats. NVP was fed to rats in a time course manner from 1 to 8 days at 150 mg/kg orally in food. Mice were given 950 mg/kg/day for 2 weeks or 10 weeks. C57BL/6 males given NVP for 2 weeks are represented by C57.1 and C57.2. Each lane was loaded with 20 µg of protein. Samples were resolved on a 4-20% gradient gel. A 1:500 dilution of primary antisera followed by 1:5000 dilution of secondary was used. xiv

15 CHAPTER 3 Figure 3-1. (A) Immunoblot showing covalent binding of NVP (150 mg/kg) to whole rat skin in vivo with a major artifact band in each lane indicated by the arrows. From left to right: primary treatment days 22, 24, 25, rechallenge (RCH, 7 days), or untreated control. Each lane was loaded with 25 µg of protein. Exposure duration in the imager was 3 min. (B) Epidermis floating above dermis from trypsin-separated skin from a rat treated for 21 days with NVP (left panel); isolated epidermal layer (right panel). Figure 3-2. (A) Immunoblot showing in vitro covalent binding of 1 mm each NVP, 12-OH-NVP, or 12-OH-NVP sulfate to rat whole skin homogenate containing both dermis and epidermis after incubation for 30 or 60 min. (B) Covalent binding of 1 mm NVP, 12-OH-NVP, or 12-OH-NVP sulfate to isolated epidermal or dermal homogenates prepared from a control rat. (C) Covalent binding of 1 mm NVP, 12-OH-NVP, or 12-OH-NVP sulfate to isolated epidermal or dermal homogenate prepared from a control rat showing that preincubation of the primary antisera with 1 mm NVP for 2 h at 37 C blocked binding of the antibody. Proteins (7.5 µg/well) were loaded in immunoblots A-C. (D) Human dermatome skin incubated with 1 mm each of 12-OH-NVP, NVP, or 12-OH-NVP sulfate compared to 1 mm 12-OH-NVP +/- 0.3 mm PAPS (1 mg/ml protein). (E) Covalent binding of 1 mm NVP, 12-OH-NVP, or 12-OH-NVP sulfate to isolated human dermatome homogenate showing that preincubation of the primary antiserum with 1 mm NVP for 2 h at 37 C blocked binding of the antibody. Protein (12 µg/well) was loaded for blots D-E. Figure 3-3. Immunoblots showing epidermal covalent binding in vivo after treatment with NVP or 12-OH-NVP for either (A) 7 days or (B) 21 days. (C) Immunoblots showing that preincubation of the primary antiserum with 1.5 mm NVP for 2 h at 37 C blocked binding of the antibody (right panel) to the drug-modified proteins after treatment with NVP for 21 days or after rechallenge with NVP (RCH), left panel. Epidermal protein loading was 15 µg/well. Figure 3-4. (A) Immunoblot showing covalent binding in the epidermis of NVP-treated female BN rats on days 10, 15, or 21 (n = 2 animals per time point) of NVP treatment. Each lane represents an individual animal with 15 µg/well of protein loaded in each well. (B) Representative H&E stained rat skin sections comparing the early infiltration of immune cells into the dermis or xv

16 epidermis of NVP- or 12-OH-NVP-treated rats. Marked acanthosis (thickening of the epidermis) combined with early lymphocyte infiltrate at the dermal-epidermal junction can be observed by day 7 of 12-OH-NVP-treated animals. By day 21 there is an increase in the cellular infiltrate with areas of detachment of the epidermis. Magnification 20x. Figure 3-5. Skin histology of PD-1 -/- knockout mice. No immune infiltrate or acanthosis was observed as was seen with NVP-treated rats. Figure 3-6. (A) Comparison of covalent binding to S9 with that to the cytosolic fractions from the skin of NVP-treated rats, isolated from either the dermis or epidermis. (B) Comparison of covalent binding to S9 with that to microsomal fractions from the skin of NVP-treated rats, isolated from either the dermis or epidermis. All animals were treated for 21 days. Protein loading was 10 µg/well. Figure 3-7. (A) Immunoblot of rat skin S9 or human female (second right lane) or human male (most right lane) liver cytosol after incubation with 12-OH-NVP in the presence and absence of PAPS. (B) Immunoblot of rat skin S9 or cytosol, or female human or rat liver cytosol after incubation with 12-OH NVP in the presence of absence of PAPS. (C) Covalent binding of 12- OH-NVP to human liver S9, human dermatome skin, or rat skin S9 in the presence and absence of PAPS. Protein loading was 12 µg/well. Figure 3-8. (A) Immunoblot comparing covalent binding of 12-OH-NVP to rat liver S9 versus rat skin S9 in the presence or absence of either PAPS or a NADPH-regenerating system (NRS). (B) Comparison of covalent binding of NVP to rat liver S9 versus rat skin S9 incubated in the presence of absence of either PAPS or a NADPH-regenerating system (NRS). (C) Comparison of covalent binding of 12-OH-NVP to rat liver S9 versus human liver S9 incubated in the presence or absence of either PAPS or NRS. (D) Comparison of covalent binding of NVP to human liver S9 versus rat liver S9 incubated with or without either PAPS or NRS. (E) Covalent binding of NVP or 12-OH-NVP to human skin in the presence or absence of PAPS or NRS. Protein loading was 12 µg/well. 1 mm of 12-OH-NVP or NVP was used for each incubation. xvi

17 Figure 3-9. (A) Immunoblot comparing the covalent binding of NVP to mouse vs. rat liver S9 in the presence of absence of either PAPS or an NADPH-generating system (NRS). (B) Comparison of the covalent binding of 12-OH-NVP to either mouse or rat liver S9 in the presence or absence of either PAPS or NRS. (C) Comparison of covalent binding of NVP to mouse vs rat skin S9 either in the presence or absence of PAPS or NRS. (D) Comparison of covalent binding of 12- OH-NVP to mouse vs. rat skin S9 either in the presence or absence of PAPS or NRS. Protein loading was 12 µg/well. 1 mm of 12-OH-NVP or NVP was present in each incubation. Figure Detection of anti-nvp and autoantibodies in the serum of a rat after rechallenge with NVP (A) Liver homogenate (10 µg/lane) from an untreated (control) rat, NVP-treated rats, or 12-OH NVP-treated rats run on SDS PAGE and stained with serum (diluted 1:500) from a rat that had been rechallenged with NVP after earlier development of a NVP-induced rash (left panel) or with serum from an untreated control rat (right panel). A 1:4000 dilution of goat anti-rat HRP linked antibody was used as the secondary antibody to visualize the binding. Blots were imaged for 3 minutes on medium exposure. (B) Using serum from the same rechallenged rat, an analogous experiment was performed using fractionated skin protein (20 µg/lane) for the epidermis, designated E, or dermis, marked D ) from untreated (control) or NVP-treated rats. Blots were run, blocked, incubated with secondary, and imaged together. Supplemental Figure 3S-1. (A) Method to fractionate ear using dorsal-ventral axis separation is shown. Ear pieces were floated on 0.625% trypsin overnight at 4 C to ensure complete epidermal-dermal separation. (B) Immunoblot experiments comparing the epidermis from the neck or ear from NVP- or 12-OH-NVP-treated female BN rats; 12 µg protein/well. Lane designations are as follows: 1 & 2 = epidermis from the neck of control rats; 3 = epidermis from the neck of a 12-OH-NVP-treated rat; 4 = epidermis from the neck of a NVP-treated animal; 5 = epidermis from the ear of a 12-OH-NVP-treated rat; 6 = epidermis from the ear of a NVP-treated rat. (C) H&E images of ear sections taken from each treatment group (representative slide from 1 of 4 rats per group shown). Magnification 20x. xvii

18 CHAPTER 4 Figure 4-1. (A) Incidence of skin rash, (B) plasma concentrations of NVP, (C) 12-OH-NVP, and (D) 12-OH-NVP sulfate in female Brown Norway rats treated with NVP only (100 mg/kg/day, n = 4), in combination with oral DHEA (50 and 100 mg/kg/day) or in combination with oral salicylamide (274 mg/kg/day). Figure 4-2. (A) Immunoblot of the epidermis comparing individual 12-OH-NVP-treated rats to NVP + oral salicylamide cotreated rats (N+Sal) or NVP only-treated rats, against 0.5% methyl cellulose gavaged controls. Protein loading was 15 µg/lane. (B) Skin histology of NVP + oral salicylamide cotreated rats, n = 4. (C) Skin histology compared between various treatment groups: normal and gavaged controls are normal without a cellular infiltrate in the dermis, while NVP, 12- OH-NVP and NVP + oral salicylamide treated rats display keratinocyte necrosis within the epidermis, with marked inflammatory infiltrate at the dermal-epidermal junction. A representative photo from one of four animals per group is shown. All rats represented in this figure were treated for 21 days. Magnification was 20x for all slides in this figure. Figure 4-3. (A) Diagram of the preliminary sites for administration of topical DHEA or topical 1- phenyl-1-hexanol to determine their effect on the NVP-induced skin rash. In 2/2 animals tested, the rash was slightly milder with DHEA, but it was completely prevented in 1-phenyl-1-hexanoltreated areas only (photos not shown). (B) Diagram of sites employed in 2 independent trials to test the effect of topical 1-phenyl-1-hexanol on the NVP-induced skin rash. Five animals in total were treated with NVP (150 mg/kg/day) in food and 1-phenyl-1-hexanol (20 mg/kg/day) on the skin. In 100% of the animals, the rash was prevented by topical 1-phenyl-1-hexanol. One representative rat from each study is shown above. Photos showing (C) skin from the back of a control rat, (D) skin from the back of the NVP only-treated rat, (E) vehicle versus 1-phenyl-1- hexanol-treated areas from an inhibitor-treated rat (topical treatment). Figure 4-4. Using skin isolated from rats cotreated with NVP and topical 1-phenyl-1-hexanol using the schematic shown in Figure 3B, epidermal immunoblot analysis was performed. (A) Immunoblot of epidermis from rash areas versus vehicle areas from the epidermis of inhibitortreated rats cotreated with NVP. (B) Immunoblot of epidermis from topical 1-phenyl-1-hexanol areas versus vehicle areas from epidermis of inhibitor-treated rats compared with that of an xviii

19 untreated control and a NVP-treated control. 15 µg of protein per lane was loaded for each of A and B. Figure 4-5. Representative histology of rat skin isolated from rats cotreated with NVP and topical 1-phenyl-1-hexanol using the schematic shown in Figure 3B. (A) H&E stained sections from upper neck/rash area from control (Ct), nevirapine-treated (NVP), or 1-phenyl-1-hexanol (Ph) rat number 1 shown in 3C-E; (B) H&E stained sections from left shoulder/vehicle area from control (Ct), nevirapine-treated (NVP), or 1-phenyl-1-hexanol (Ph) rat number 1 shown in 3C-E; (C) H&E stained sections from right shoulder/1-phenyl-1-hexanol-treated area from control (Ct), nevirapine-treated (NVP), or 1-phenyl-1-hexanol (Ph) rat number 1 shown in 3C-E. Magnification 20x for all slides in this panel. Figure 4-6. (A) Second topical schematic used to test the inhibitor 1-phenyl-1-hexanol. (B) Immunoblot of epidermis from areas of vehicle or inhibitor treated areas using the second schematic shown in 6A. The control is epidermis from the untreated control rat; Ph1 or Ph2 are topical inhibitor-treated epidermal areas from rat # 1 or 2, respectively; Vh1 or Vh2 are vehicle treated epidermal areas for each rat, and RA1 or RA2 are from rash areas with no topical treatment. NVP is from the epidermis of the back of the neck for the NVP-treated positive control rat. Protein loading was 15 µg/lane. Figure 4-7. Histology with H&E staining of skin isolated from rats cotreated with NVP and topical 1-phenyl-1-hexanol using the schematic shown in Figure 6A. The upper left slide of each panel is from a control animal without NVP treatment, the upper right slide is from a NVP-treated animal with no topical treatment, and the lower two slides are from animals with NVP + topical treatment. (A) skin from upper back with no topical treatment representing the typical rash; (B) midback where the vehicle was applied in inhibitor-treated animals only (lower two slides); (C) the lower back were 1-phenyl-1-hexanol was applied in inhibitor-treated animals only (lower 2 slides). Magnification 20x. (D) Preincubation of the primary anti-nvp serum with 1.5 mm NVP dissolved in DMSO for 2 h at 37 C prevented covalent binding of the anti-serum to epidermal samples from the samples shown in Figure 3C-E, except for one artifact band. The DMSO control (right most lane) where the primary anti-serum was incubated with DMSO alone. Protein loading was 15 µg/lane. xix

20 Figure 4-8. (A) Immunoblot of isolated rat liver cytosol or rat skin cytosol incubated with 12- OH-NVP or a combination of 12-OH-NVP (12-OH) and 1-phenyl-1-hexanol in vitro, in the presence and absence of PAPS. (B) Immunoblot of rat skin cytosol versus female human liver cytosol incubated with or without PAPS and 12-OH-NVP or 12-OH-NVP and 1-phenyl-1- hexanol. (C) Human skin dermatome homogenized and incubated with 12-OH-NVP or 12-OH- NVP + 1-phenyl-1-hexanol to show the same phenomenon exists in human skin. (D) Human SULT 1A1*1 incubated with 1 mm 12-OH-NVP (12-OH) or 12-OH-NVP and 1-phenyl-1- hexanol +/- 0.3 mm PAPS, or 12-OH-NVP sulfate (12-Sulfate). 12 µg/well protein was loaded for each blot. Supplemental Figure 4S-1. Urinary excretion of (A) 12-OH-NVP, (B) 4-COOH-NVP, (C) 12- OH-NVP sulfate, (D) 2-OH-NVP, and (E) 3-OH-NVP from rats treated with NVP (100 mg/kg/day), NVP + DHEA (100 mg/kg/day) each or NVP + salicylamide (274 mg/kg/day; n = 4 in each group). Data depicts the mean ± SD. Supplemental Figure 4S-2. H&E stained sections comparing the histology of rat skin in response to (A) NVP treatment, (B) NVP + topical DHEA cotreatment, or (C) control rats. Magnification = 20x. xx

21 List of Tables CHAPTER 1 Table 1-1. Human Sulfotransferase Isoforms and Expression. CHAPTER 3 Table 3S-1. Day 7 Skin Rash Grading. Table 3S-2. Day 10 and 15 Skin Rash Grading. Table 3S-3. Day 21 Skin Rash Grading. CHAPTER 4 Table 4-1. Inhibitors of sulfation and dosing method. Table 4-2. Comparison of results obtained from 12-OH-NVP sulfate inhibitor studies. Table 4S-1. Day 21 Skin Rash Grading. xxi

22 List of Schemes CHAPTER 2 Scheme 2-1. Synthetic pathway of the immunogen used for induction of anti-nvp antiserum. CHAPTER 3 Scheme 3-1. Proposed chemical mechanism of NVP-induced skin rash resulting from covalent binding of 12-OH-NVP sulfate in the skin. Scheme 3-2. Proposed bioactivation pathway of NVP leading to immune-mediated skin rash. CHAPTER 4 Scheme 4-1. Depiction of schematic used to prevent rash in this study. xxii

23 List of Abbreviations 12-OH-NVP 12-hydroxynevirapine 12-OH-NVP sulfate 12-sulfatenevirapine 2-OH-NVP 2-hydroxynevirapine 3-OH-NVP 3-hydroxynevirapine 4-COOH-NVP 4-carboxynevirapine ADR adverse drug reaction ALT alanine aminotransferase APC antigen presenting cell BN rat Brown Norway rat CTL cytotoxic T-lymphocyte DHEA dehydroepiandosterone DILI drug-induced liver injury DNA deoxyribonucleic acid FMO flavin monoxygenase HIV human immunodeficiency virus HLA human leukocyte antigen HPLC high performance liquid chromatography IDILI idiosyncratic drug-induced liver injury IDR idiosyncratic drug reaction IL interleukin KC keratinocyte LC Langerhans cell LC/MS liquid chromatography/mass spectrometry LTT lymphocyte transformation test MHC major histocompatibility complex NADPH nicotinamide adenine dinucleotide phosphate (reduced form) NLR nod-like receptor NVP nevirapine papc professional antigen presenting cell P450 cytochrome P450 PAPS 3 -phosphoadenosine-5 -phosphosulfate P-I hypothesis pharmacological interaction hypothesis PRR pattern recognition receptor SA salicylamide SJS Steven s-johnson syndrome SLE systemic lupus erythematous SN 2 substitution nucleophilic 2 SULT sulfotransferase TEN toxic epidermal necrolysis Th1 t-helper cell 1 TLR toll-like receptor UV ultraviolet xxiii

24 List of Thesis Publications First Author: Bioactivation of Nevirapine to a Reactive Quinone Methide: Implications for Liver Injury. Amy M. Sharma, Yan Li, Maria Novalen, M. Anthony Hayes, and Jack Uetrecht. Chemical Research in Toxicology, , Demonstration of Nevirapine Bioactivation and Covalent Binding in Skin. Amy M. Sharma, Klaus Klarskov, Jack Uetrecht. Chemical Research in Toxicology, , OH-Nevirapine Sulfate, Formed in the Skin, is Responsible for Nevirapine-Induced Skin Rash. Amy M. Sharma, Maria Novalen, Tadatoshi Tanino, Jack P. Uetrecht. Chemical Research in Toxicology, , Bioactivation of Drugs in the Skin and its Relationship to Skin Rashes. Amy M. Sharma and Jack Uetrecht. Drug Metabolism Reviews, 2013 (in submission). A Mechanism for Cutaneous Drug-Induced Hypersensitivity: Role of the NLRP3 Inflammasome in Nevirapine-Induced Skin Rash. Amy M. Sharma, Dana J. Philpott, Jack Uetrecht. Journal of Investigative Dermatology, 2013 (in preparation). Co-author: Animal Models of Idiosyncratic Drug Reactions. Amy M. Sharma (co-author), Winnie Ng (first author), Alexandra R. M. Lobach, Xu Zhu, Xin Chen, Feng Liu, et al. Advances in Pharmacology, , Identification of Danger Signals in Nevirapine Induced Skin Rash. Amy M. Sharma (second author), Xiaochu Zhang (first author), Jack Uetrecht. Chemical Research in Toxicology, 2013 (in submission). Methanol Embryopathies and Protein Oxidation in Mouse Embryo Culture Following Pre-treatment with a Free Radical Spin Trapping Agent and Inhibitors of Prostaglandin H Synthase and NADPH Oxidases: A Role for NADPH Oxidase-Derived ROS. Amy M. Sharma (second author), Lutfiya Miller (first author), Peter G. Wells (in submission). Direct Activation of Antigen Presenting Cells by Nevirapine and its Metabolites. Amy M. Sharma (second author), Xin Chen (first author), Jack Uetrecht (in preparation). xxiv

25 CHAPTER 1 INTRODUCTION No amount of experimentation can ever prove me right; a single experiment can prove me wrong. - Albert Einstein 1

26 1.1 Adverse Drug Reactions An Overview Much progress has been made in the past 50 years in the treatment of disease and the control of ailments and afflictions through the use of numerous drugs but intimately tied to this medical success remains the dark side: adverse effects to xenobiotica. Adverse drug reactions (ADRs) are, according to the World Health Organization, any noxious, unintended, and undesired effect of a drug, which occur at doses used in humans for prophylaxis, diagnosis, or therapy. Out of all hospitalized patients each year, approximately 2,216,000 experienced a serious ADR and approximately 106,000 per year died from an ADR. 1 Fatal ADRs rank 4 th to 6 th among the leading causes of death in the United States. 1 Clearly, ADRs represent a huge burden on the health care system in North America and elsewhere. The occurrence of ADRs may result in a black box warning label for the offending drug or cause the drug to be removed entirely from the market. 2 This has many implications, not only on the drug companies who spend upwards of 10 years and millions of dollars to see a drug through to medical practice, but also to the majority of patients who do not suffer from the ADR(s) and who may then be limited from using a potentially beneficial drug. It is therefore pertinent to understand and develop ways to test or screen for the occurrence of ADRs, both in preclinical and clinical settings Types of Adverse Drug Reactions ADRs may be classed into seven distinct types, each based on the mode or mechanism of occurrence. The classification system used to describe specific ADRs is as follows: 3 Type A: Augmented response to a drug Type B: Bizarre or idiosyncratic effect Type C: Chemical effect Type D: Delayed effect Type E: End of treatment effect Type F: Failure of therapy Type G: Genetic basis effect 2

27 The original classification system containing simply Type A and Type B ADRs proved insufficient to encompass a variety of side effects and has therefore evolved over time. Type A and Type B were first introduced in 1977 by Rawlins and Thompson. 4 Type A adverse reactions represent an augmented or exaggerated response to a drug, and their occurrence is predictable based on the known pharmacology of the drug. Typically, a clear dose-response relationship is associated with Type A ADRs, and discontinuation of the drug is often able to halt the adverse effect. In contrast, Type B ADRs do not display a simple dose-dependency relationship, and these cannot be predicted based upon the known pharmacology of the drug. The Type B bizarre reactions are also known as idiosyncratic drug reactions (IDRs). Type A effects represent ~ 80% of all ADRs while Type B represents ~ 20% of all ADRs; however, the Type B effects are often very severe and can be fatal. This, along with their unpredictable nature make them the most problematic form of ADRs. 5 IDRs are the major focus of this thesis and will be discussed extensively in the following sections. 1.2 Idiosyncratic Drug Reactions (IDRs) Idiosyncratic drug reactions are rare and unpredictable side effects of drugs, for which an exact definition is subject to debate. In the context of this work, the term will be used to describe any reaction which does not usually occur in humans within clinically used dosages, and does not involve the known pharmacologic properties of the drug. 6 Due to their rare and unpredictable nature, IDRs are usually not identified during clinical trials because the sample size is insufficient to allow for their detection. Approximately 10% of drugs released onto the American market between the years were withdrawn or received a black box warning label due to adverse side effects, which did not appear during their respective clinical trials. 2 The significant financial and patient-care burden caused by IDRs has led to the need for biomarkers or identifiers for the occurrence of IDRs in preclinical and clinical settings. Unfortunately, reproduction or modeling of an IDR in experimental animals is difficult because the reactions occur just as rarely in animals as they do in humans. Therefore, understanding IDRs is challenging, and to date few animal models exist that successfully capture clinical features of IDRs as they occur in people. IDRs can affect almost any organ system; however, the liver and skin are common targets. 7 IDRs can also often affect the bone marrow or blood cells, which possess enzymes such 3

28 as myeloperoxidase that, although never designed to metabolize drugs, are capable of doing so. Below is a short description of IDRs as they affect each major target organ type. The liver represents one of the most common organs afflicted by IDRs, and it is also a primary reason for both preclinical drug candidate failure and drug withdrawal from the market. Toxic drug effects on the liver may manifest as asymptomatic mild increases in alanine transaminase (ALT) or other hepatic enzymes, which typically resolve over time despite continued treatment (termed adaptation or tolerance ); however, in some cases, continuation of the drug may lead to fulminant liver failure and ultimately death. Halothane-induced hepatitis is one of the best characterized examples of drug-induced idiosyncratic hepatotoxicity. The skin is another commonly afflicted organ, with drug-induced skin toxicity representing ~3% of hospitalized patients. 8 In the case of cutaneous IDRs, they may manifest as a mild rash, with minor irritation, few lesions and scaling, or progress to anywhere from Steven s- Johnson syndrome (SJS) to toxic epidermal necrolysis (TEN), with complete destruction of cutaneous integrity. SJS and TEN are associated with high mortality: 5% for SJS, ~ 30% for TEN). 9 Most adverse skin toxicities include perivascular lymphocytic infiltration into upper dermal layers with the presence of activated macrophages and epidermal alterations, which supports the hypothesis that drug-induced rashes are immune mediated. 10 Bloor disorders or dyscrasias including aplastic anemia, thrombocytopenia, and more commonly, agranulocytosis, are examples of drug-induced IDRs affecting the blood and bone marrow. 8 Numerous unrelated drugs can induce toxic effects on blood. Myeloperoxidase enzymes present in granulocytes such as neutrophils are capable of metabolizing drugs. Agranulocytosis is a major IDR where granulocytes, mostly neutrophils, are diminished to < 500 cells/µl, where the normal neutrophil levels are 5,000 10,000 cells/µl in the blood. 11 In patients treated with aminopyrine who develop agranulocytosis, anti-neutrophil antibodies have been detected. 12 Clozapine is the most commonly prescribed drug that is associated with a high incidence of idiosyncratic agranulocytosis. Myeloperoxidase present in neutrophils metabolize clozapine to a reactive nitrenium ion capable of binding to neutrophils. 13 Rechallenge in patients who previously experienced clozapine-induced agranulocytosis usually leads to a recurrence of agranulocytosis; however, the recurrence is delayed and does not occur any faster than on first exposure. 14 This suggests a non-immune mechanism. It may also be consistent with an autoimmune mechanism, but this has yet to be proven. 11 4

29 Multi-organ effects, although less common, can occur. Clinically, there is a wide range of severity in which IDRs may present themselves. Specific factors governing the severity of IDRs, or involvement of a single organ versus multi-organs, are not known to date. There exist a few clinical characteristics that can be used to describe a typical IDR. Delayed onset of the adverse effect is a common characteristic of IDRs, with the adverse effect appearing one week or more after starting the drug. 15 This is presumably because these reactions are immune mediated, 16 and it takes time to activate the few T cells that have the appropriate specificity and have them proliferate to sufficient numbers for the immune response to become clinically evident. On rechallenge, the reaction usually occurs more rapidly because of memory T cells. It can present with a more severe toxicity than on first exposure, and it can also lead to more systemic effects or effects on other organs as well. 6 Certain IDRs also involve the production of autoantibodies, such as in the case of hydralazine- or isoniazid-induced systemic lupus erythematosus (SLE). Autoantibodies to specific proteins have also been linked to drug hepatotoxicity; examples of this include halothane-induced hepatitis 17 or dihydralazine-induced hepatitis. 8 The fundamental mechanisms of IDRs are to date unknown. Circumstantial evidence suggests that most IDRs involve the production of reactive metabolites generated by drug metabolism, 18 and the major hypothesis is that IDRs are immune mediated. 16 Host-specific factors (gender, age, weight, concomitant disease, etc.) and genetics (HLA gene associations) appear to play a role in dictating the occurrence of IDRs; however, much work is still required to confirm their specific contributions, and their associations may not exist for all drugs causing IDRs. 1.3 Proposed Mechanisms of Idiosyncratic Drug Reactions Although the exact mechanism(s) for IDRs are unknown, three major working hypotheses attempt to explain how IDRs may occur. 6 These include the Hapten Hypothesis, the Danger Hypothesis, and the Pharmacological Interaction (P-I) Hypothesis. Each is described in detail below Hapten Hypothesis In 1936, Landsteiner and Jacobs coined the term hapten in reference to low molecular weight (< 1000 Da) chemical allergens that are by themselves non-immunogenic unless they are bound to larger carrier macromolecules such as proteins. 22,23 Pro-haptens, in a similar manner, are chemical entities that must be metabolized to compounds capable of irreversibly binding to such 5

30 macromolecules. 23 The hapten hypothesis is built on the classical self-nonself immunological framework, and can be applied to drugs, most of which are less than 500 Da. A reactive, unmetabolized drug, or more often as circumstantial evidence suggests in the case of IDRs, a reactive metabolite of a drug, may become antigenic by binding to endogenous proteins or other macromolecules. Following this drug-protein conjugation step is the uptake of the modified protein(s) via professional antigen presenting cells (papcs), leading to antigen processing into peptide fragments, which are then presented in the groove of the major histocompatibility complex (MHC) on antigen-presenting cells (APCs) to T cells. If the T cell receptor of this cell matches the peptide, it generates a signal, which is referred to as Signal 1. In the presence of other costimulatory molecules, which are referred to as Signal 2, this can lead to activation of the T cell. Signal 1 in the absence of Signal 2 results in immune tolerance. Penicillin allergy is an example of a reactive parent drug, where in the absence of metabolism, protein haptenation occurs, sometimes leading to an IgE-mediated IDR. The sp 2 hybridized carbonyl group within the β lactam ring is forced from its ideal 120º to be 90. This inherent ring strain of the β lactam ring makes it susceptible to nucleophilic attack by nitrogen and sulfur nucleophiles, thus opening the ring and relieving the strain. Antibodies have been detected against the penicillin-modified proteins that are associated with penicillin allergy that occurs in a small percentage of patients. These antibodies can go on to cause mast cell degranulation, leading to the release of inflammatory mediators such as histamine and leukotrienes. Anaphylactic shock can result and re-exposure can be life-threatening. β-lactam ring Benzylpenicillin Figure 1-1. Reaction of penicillin with protein nucleophiles via spontaneous ring opening in a hapten-type mechanism. 6

31 There are numerous examples of drugs known to cause hapten formation through reactive metabolites, and this is believed to lead to immune-mediated toxicity; some of these include aminopyrine-induced agranulocytosis, 12 halothane-induced hepatotoxicity, 24 and tienelic acidinduced hepatotoxicity. 25 Although numerous drugs causing IDRs form reactive metabolites capable of haptenating proteins, the degree of protein haptenation and risk of IDRs is not well defined or understood, although several groups have tried to correlate this relationship. Presumably, organ-specific responses, host-immune and enzymatic factors, and basic principles underlying the organisms response to noxious stimuli produced by the degree of binding all play a role. Numerous proteins are also typically adducted by a single drug, yet no concrete association has been established between the type of protein adducted and IDR risk. Therefore, all of these factors mentioned must be pieced together to understand the role of haptenation in the occurrence of IDRs. Figure 1-2. The Hapten Hypothesis: a reactive parent drug or the reactive metabolite of a drug covalently binds to protein. This modified protein can elicit an immune response; however, it is 7

32 now known that additional factors are required as discussed below. Adapted from Uetrecht, Danger Hypothesis In 1994, Polly Matzinger challenged the field of immunology with her radical Danger theory, 26 which opposed the classical self-nonself immunological framework. She proposed that foreign proteins do not typically elicit an immune response unless cell stress or cellular perturbations lead to the release of danger or stress signals from damaged or dying cells. These signals, as the theory posits, cause upregulation of appropriate co-stimulatory molecules, such as B7 on APCs, which interact with T cells (i.e. through CD28), stimulating the T cells. This allows APCs to produce Signal 2, which is necessary for an immune response. Without Signal 2, the result is immune tolerance. Signal 1 in this context is recognition of a peptide by a T cell when the peptide has been presented in the grove of MHC II on an APC. The Danger theory also implies that the type or nature of an immune response is dictated by the affected tissue, and in this way an unnecessary systemic immune response is avoided. The origins of Signal 1 are not addressed by the Danger Hypothesis, and therefore it does not exclude an immune response against a drug itself, drug-protein conjugate, or autoantigen. The Danger theory is attractive in the case of IDRs because reactive metabolites have the potential to cause cellular damage and cause the release of stress signals. 27 This could explain organ or tissue-specific effects where covalent binding is present in different tissues yet an adverse response develops in only one. Identification of danger or stress signals is still in its initial stages, but these molecules should be endogenous compounds such as S100, HMGB1, or heat shock proteins 28 released from damaged cells. There are numerous other intracellular proteins that may also act as danger signals, and in the case of IDRs their release may serve as a biomarker for drug toxicity; more work is needed to confirm which molecules or patterns of response can translate cell stress into an immune-mediated IDR. 8

33 Figure 1-3. The Danger Hypothesis: stressed or damaged cells release endogenous danger signals that activate APCs, leading to upregulation of costimulatory molecules (B7 on APCs) which interact with CD28 on T cells leading to an immune response. Adapted from Uetrecht,

34 1.3.3 Pharmacological Interaction Hypothesis In 1998 Werner Pichler found that isolated T cells from patients who had developed an IDR proliferated in the absence of drug metabolism when incubated with the drug that had induced the IDR. 29 The Pharmacological Interaction (P-I) hypothesis posits that certain drugs can reversibly bind to the MHC-T cell receptor complex leading to an immune response, which in some cases may lead to an IDR. Metals such as nickel and beryllium are examples where reversible, although very tight binding to MHC is able to induce an allergic reaction. Specific drugs such as sulfamethoxazole, carbamazepine, lamotrigine, lidocaine, etc., 30 represent a weaker class of reversible binding to MHC than metals; these drugs can stimulate T cells in their unmetabolized form. 30 Sulfamethoxazole is the best characterized of these drugs. It becomes bioactivated to form a hydroxylamine metabolite, which is further oxidized to a reactive nitroso compound. T cells isolated from patients who developed a sulfamethoxazole-induced IDR responded to the parent drug instead of the reactive metabolite, 31 supporting the P-I hypothesis. However, a more recent study found that many more lymphocytes from patients who developed an sulfamethoxazole-induced IDR responded to the reactive nitroso metabolite than the parent compound, which supports the role of the reactive metabolite in sulfamethoxazole-induced toxicity. 32 This highlights the mechanistic complexity of studying IDRs, and the inability of a single hypothesis to capture such intricacy. In addition, it must be noted that the P-I hypothesis assumes that what T cells respond to in vitro is also what induced the immune response. We have found in the nevirapine model that this is not the case what T cells recognize in the lymphocyte transformation test (LTT) in vitro is not what induced the response in vivo. Thus the basis for the P-I hypothesis is false. Nevertheless, the P-I hypothesis may be useful for small peptide-like compounds or compounds that do not involve reactive metabolite formation, such as ximelegatran

35 Figure 1-4. The P-I Hypothesis: A parent drug binds reversibly to the MHC-TCR complex (Signal 1) leading to an immune response. Adapted from Uetrecht,

36 1.4 Idiosyncratic Hepatotoxicity In order to understand the role that reactive metabolites may play in causing idiosyncratic toxicity and how the aforementioned mechanistic hypothesis may apply to this thesis, a brief review of the enzymatic processes leading to the production of reactive metabolites is appropriate. Both the liver and skin will be discussed in detail. The hepatic morphology and immunological nature of the liver is also discussed in order to understand the unique distribution of hepatic enzymes as well as mechanisms by which toxicants interact with the liver s own immune system Hepatic Function and Morphology The liver is the primary site of xenobiotic metabolism and biotransformation in the body, and the cells of the liver are exposed to significant amounts of various chemicals. Located down stream to the intestinal tract, the liver is the first to receive nutrients, drugs, vitamins, etc., from the portal blood. 34 Metabolic and nutrient homeostasis, synthesis of clotting factors and proteins, lipid metabolism, and production of bile and biliary secretion are only a few of the major functions of the liver. Toxic insults have the ability to damage the functions of the liver through acute or chronic exposure. The structural organization in the liver is designed to facilitate the many important functions it must perform. The liver consists of hepatic lobules, which are further divided into the centrilobular, midzonal, and periportal regions. Hepatic zonation is important when considering toxicological effects on the liver. For example, the oxygen saturation of hepatocytes in zone 3, which are closest to the central vein, is only 4-5% compared to those in zone 1, which is 9-13%. 34 Drug metabolizing enzymes exhibit a similar phenomenon, with Phase I enzymes situated predominantly in zone 3, and Phase II enzymes closer to Zone I. These enzymes are described in the following section Biotransformation in the Liver The primary purpose of xenobiotic metabolism is to terminate biological activity and produce more water soluble compounds through the introduction of a polar group. This occurs primarily by oxidation or conjugation reactions, allowing for easier excretion from the body. However, this biotransformation can sometimes lead to the formation of a chemically reactive 12

37 metabolite of a drug. Therefore a basic understanding of drug metabolism is necessary to understand how this may occur. Drug metabolism occurs through enzymatic biotransformation primarily in the liver via enzymes that are classified as either Phase I or Phase II reactions. Phase I functionalization occurs mostly through a superfamily of heme-containing enzymes called cytochromes P450 (abbreviated P450 or CYP), which are the most important and common drug metabolizing enzymes. P450s are present to the greatest degree in the centrilobular region of the liver and membrane bound in the endoplasmic reticulum of hepatocytes. The P450 family consists of many subfamilies and isoforms, each with preferential substrate activity. Fifty seven human P450 enzymes have been identified to date; however, this does not mean each human expresses all isoforms. 35,36 P450 enzymes directly catalyze oxidation reactions. Specifically, they catalyze four main types of oxidation: hydroxylation, epoxidation, dehydrogenation, and heteroatom oxidation. 37 The general formula can be summarized as follows, where R = substrate and RO = product: NADPH + + H + + R + O 2 NADP + + H RO Although there exist many types of P450 enzymes, the 3A4 family is the most significant from a dug metabolism perspective. CYP3A4 is the most highly expressed P450 enzyme in human liver it is also in the human gut 38 - and exhibits a great range of substrate specificity because the active site can accommodate very large substrates (>1000 g/mol). 37 Approximately half of all marketed drugs are metabolized by hepatic CYP3A4 enzymes. 37,38 Next in metabolic significance are CYP2C9, 2D6, and 2C19, followed by 2E1, 1A2, and 2B6, all of which have fewer substrates than CYP3A4. 39 Other enzymes can also oxidize drugs including peroxidases (myeloperoxidases, prostaglandin H synthase, horseradish peroxidase), flavin monoxygenases (FMOs), aldehyde dehydrogenases, xanthine oxidases, aldehyde oxidase, and monoamine oxidases. Following traditional Phase I oxidation by P450 enzymes, predominantly in the liver, further metabolism known as Phase II may occur. Phase II metabolism involves conjugating enzymes such as sulfotransferases and glucuronosyl transferase. The focus will be on the sulfating enzymes which are described in detail in section All of the enzymes described in this section function to facilitate excretion of xenobiotics and drugs from the body. However, it is an imperfect process and chemically reactive compounds can be produced. Reactive metabolites are electrophilic in nature (electron seeking) and react with nucleophilic groups on proteins (typically a lone pair of electrons or negative charge) or even 13

38 DNA, leading to the formation of neoantigens. Some common reactive species include Michael acceptors, epoxides/arene oxides, and nitroso amines. 38 This process can occur in any organ, and numerous cell types possess metabolizing enzymes capable of drug bioactivation. As the hapten hypothesis stipulates, the resulting covalent binding is irreversible and creates more immunostimulatory compounds, which in turn, activate the immune system. However, the liver is a unique organ in that it is exposed to numerous xenobiotics and forms many neoantigens on a daily basis; how then has it found ways to regulate its immune response? This is explored in the following section, and later in Section 1.5, contrasted to what occurs in skin following the same covalent binding. Figure 1-5. Proposed pathway of drug bioactivation leading to haptenation and cellular mechanisms of hepatocyte death. Adapted from Kaplowitz, Liver Immune System in Relation to Hepatotoxicity The dominant immune response in the liver is immune tolerance; this is due to a unique hepatic microenvironment and ultrastructure. The tolerogenic response is thought to be a key reason why covalent binding in the liver does not usually lead to hepatotoxic responses to drugs. For example, architecture in the liver allows for T cell interaction with resident APCs, which facilitates APCs, and especially dendritic cells (DCs), which are the major professional antigen presenting cell, to become capable of inducing a immunogenic or tolerogenic response. Cytokines 14

39 (and other molecules) produced by hepatic APCs (DC s, Kupffer cells, liver sinusoidal endothelial cells, hepatic stellate cells, etc) such as interleukin-10 (IL-10), transforming growth factor (TGF-β), prostaglandin (PG)E 2, and granulocyte macrophage-colony stimulating factor (GMCSF) influence tolerogenic outcomes in the liver. 41 In addition, DCs and other hepatic APCs express death ligands, 41 which may contribute to apoptosis of activated T cells within the liver; this has been suggested as a reason for the ease of hepatic transplantations as compared to other organs such as skin. In addition, hepatic T cells undergoing apoptosis have been shown to release TGF-β and IL-10, which further promotes a tolerogenic microenvironment in the liver. 41 Furthermore, activation of regulatory T cells is believed to down-regulate MHC-II and costimulatory molecules on DCs, which maintains them in an immature phenotypic state. 42 Given that the liver has developed mechanisms to down regulate inflammatory immune mediated responses, it has been very difficult to develop animal models of idiosyncratic druginduced liver injury. In many cases the drug is metabolized by P450 to reactive electrophiles which, if very electrophilic, may bind directly to the enzyme from which it was formed, acting as a suicide inhibitor for that enzyme (i.e. P450). This P450 adduct or other hepatic adducts, in theory, could act as a hapten, inducing an immune response, and certainly in many cases where drug-p450 haptenation occurs, a transient increase in ALT is seen in animal models. However, the transient injury is cleared, and animals, as well as the majority of humans, adapt or tolerize to the insult. Further, not all drugs that modify hepatic proteins induce an immune response. No clear or convincing mechanisms specific to idiosyncratic hepatotoxicity exist; there does appear to be a clear link to host-dependent factors which are poorly characterized. 34 Thus, a good understanding of drug-induced immune-mediated liver injury does not exist. 15

40 Figure 1-6. Proposed mechanisms of DILI, including drug bioactivation leading to reactive intermediates, which may cause hepatocyte damage and invoke immune-mediated responses. A balance of hepatoprotective and inflammatory factors dictate the potential for toxicity. Adapted from Holt and Ju, Idiosyncratic Cutaneous Toxicity The skin is a dynamic organ and truly an environment unto its own. In order to understand how cutaneous IDRs may arise, a discussion of the integument is necessary. Little is known regarding the relationship between the skin structure, enzymatic processes, and immunological activity in the skin as applied to idiosyncratic cutaneous toxicity. The purpose of this mini review is to integrate current and emerging findings into proposed mechanisms of drug bioactivation in the skin leading to rashes Skin Structure and Function The skin provides the body with a structural barrier to the outside world, and it is the first line of defense against any external stimuli. A highly complex tissue, the structure of skin can 16

41 easily be understood through its primary protective role. Composed of three main layers: the epidermis, dermis, and hypodermis or subcutis, the skin maintains internal homeostasis and participates in a myriad of important physiological and cellular functions. These include, but are not limited to, thermal regulation, electrolyte and hormonal balance, metabolic and immune regulation, as well as defense against invading pathogens, micro-organisms, chemicals, ultraviolet radiation, and physical-mechanical insults. 34 The skin possesses a variety of sophisticated mechanisms to perform its numerous functions. Although the structure of skin is similar in all areas of the body, the thickness varies depending on the specific organ and location, reflecting the requirement for increased barrier functions in certain anatomical sites (i.e. the soles of the feet). The hypodermis is the lowest layer of the skin, therefore in closest contact to internal organs. This portion of skin provides cushioning and insulation because it is composed primarily of adipocytes arranged in a lobular fashion used for fat storage. Other cell types found in the hypodermis include fibroblasts and macrophages. Blood and lymphatic vessels, nerves, and fibrous tissues connecting the skin to the deep fasica are also components of the hypodermis. Above the hypodermis lies the dermis, which is structurally the largest and thickest portion of skin, comprising 90% of the entire integument. 34 The dermis provides structural integrity as well as mechanical and tensile strength due to the dense collagen and elastin connective tissue networks produced by resident dermal fibroblasts. The collagen alone accounts for approximately 75% of the skin s dry weight. 44 Although fibroblasts are the major cell type found in the dermis, other cell types such as mast cells, macrophages, and T cells, are also found here. These cells reside mostly in the surrounding vasculature and within the papillary dermis (upper dermis). 44 The reticular dermis comprises the mid to lower dermal layers, and it is much thicker. Other key properties of the dermis include the presence of nerves and receptors, hair follicles, sebaceous glands, eccrine sweat glands, apocrine glands, lymphatic vessels, and allimportant blood vessels. 44 The epidermis is situated above the dermis and is structurally a very different tissue type from the rest of the skin. While the dermis serves a primarily supportive function and is structurally very sound, the epidermis is almost completely cellular and very metabolically and immunologically active. This is due to the cell types that form the epidermis, namely keratinocytes (KCs), Langerhans cells (LCs), and melanocytes. 17

42 Keratinocytes comprise approximately 95% of the epidermis and are considered the most important in maintaining structural epidermal integrity. Keratinocytes originate as stem cells in the basal epidermal layers (stratum basale), and as the cells mature, they differentiate and migrate to form the upper epidermal layers. Thus the cells divide from the stratum basale to the stratum spinosum, granulosum, lucidum, and corneum, with the stratum corneum being the outermost layer of skin (composed of simple keratinized cells). Desmosomes are the chief epidermal intracellular adhesion molecule, and they provide the keratinocytes with a structural framework. 44 Keratinocytes possess numerous metabolic and immunologic capabilities that will be reviewed in the following sections, but these cells also produce keratins, which provide mechanical and cellular strength. It must be noted that mucous membrane epithelia lack the stratum corneum (outer keratinized layer) in order to provide a lubricating lining. Langerhans cells and melanocytes are both bone marrow-derived cells, which migrate into the epidermis during embryogenesis. Langerhans cells are specialized cells that account for only ~ 5% of epidermal cells, and are considered professional antigen presenting cells. Their role is considered in detail in Melanocytes synthesize pigment via melanosomes and are also found in hair follicle bulbs Metabolic Enzymes in Skin Biotransformation in the skin occurs primarily in epidermal cells such as keratinocytes and Langerhans cells, which are metabolically more active than dermal or hypodermal cells. Nonetheless the skin does possess both Phase I and Phase II enzymes, and their presence is described here. Phase I. Classical Phase I metabolism accounts for only ~ 2% that of the liver (on a perbody weight basis). 34 CYP1A has been identified in both rodent and human skin; however, the actual protein has only been identified in keratinocytes. 45 Data on the identification CYP2A6 has been mixed, with mrna initially undetectable in KCs, but identified in fibroblasts and melanocytes. 45 However recent data has identified CYP2A6 mrna and CYP2B6 protein in KCs. 45 Results for CYP2C enzymes have also yielded mixed results, but CYP3A5 has been identified in all skin samples and biopsies examined to date. 45 Mixed results are thought to be due to changes in cells with culture conditions i.e. loss of gene expression, and differences in protocols or methods for identification of enzymes. 45 Other enzymes capable of oxidation in the 18

43 skin that have been confirmed in biopsy samples include flavin monooxygenases, types 1 and 3, with the mrna detected for the former and both the mrna and protein for the latter. 45 Phase II. Various Phase II enzymes have been identified in rodent and human skin. These include isoforms of epoxide hydrolase, UDP-glucuronosyl transferase, quinone reductase, β-glucuronidases, N-acetyl transferases, esterases, reductases, and sulfotransferases. 34 Although these enzymes have been identified, amounts are significantly much less than expression in the liver, and human skin has far lower expression of all Phase I and Phase II enzymes than rodent skin overall Immune Function of the Skin The skin was previously thought to be an innocent bystander in immune-mediated hypersensitivity responses. This view has since changed, and it has become increasingly clear that the skin is immunologically active and able to respond to various challenges. Keratinocytes, for example, although non-professional immune cells, have been termed the adjuvant of the skin due to their unique ability to shape innate and adaptive cutaneous immune responses. 46 This is made possible in part by keratinocyte expression of numerous functional immune molecules such as cytokines, chemokines, MHC-II molecules, and co-stimulatory molecules. 44 In addition, KCs are known to constitutively express specific cytokines and may produce others when activated. These include interleukins 1α/1β, 3, 6, 7, 8, 10, 12, 15, 18, and Other factors also produced by KCs include IL-1ra (receptor antagonist), TNF-α, IFN-α and IFN-β, TGF-α, TGF-β, chemokine receptor 3 (CCR3), eotaxin, RANTES, macrophage-colony stimulating factor (M- CSF), granulocyte macrophage colony stimulating factor (GM-CSF), and CD14, CD40, and tolllike receptors 1 to 6, and Along with TLRs, another pattern recognition receptor found on keratinocytes is the nucleotide-binding oligomerization domain receptor (NOD-like receptors or NLRs). 47 NLRs represent a platform for innate immune sensing, and keratinocytes have been shown to express a large multimeric form of NLRs termed inflammasomes. The cytosolic NLRP3 inflammasome activates caspase-1, which in turn causes cleavage of pro-il-1β and pro-il-18 into their mature states. 48 Both cytokines are critical for the instruction of T cell responses, providing a link of innate instruction of adaptive immunity. 48 NLRP3 responds to a variety of cell stress stimuli and 19

44 has been shown to be involved in contact hypersensitivity conditions, 49 where covalent binding of small chemical molecules can activate the inflammasome. It is probable that drugs can also act in this way i.e. undergo metabolic biotransformation in the skin where they may form adducts, and the resulting cell perturbations may activate the inflammasome. Keratinocytes do not normally express co-stimulatory molecules such as CD80/86, and therefore, they are unlikely to be able to constitutively prime naïve T cells. 50 They are also unable to classically process and present antigen, and they cannot prime new T cell responses. However, non-classical presentation by keratinocytes has been suggested such as presentation of glycolipids by CD1 molecules. 50 CD1d can also be induced by poison ivy, and expression of CD1d on cultured KCs can activate natural killer T cells. 50 This suggests lipid-derived ligand presentation by KCs may occur. Keratinocytes can also induce Th1-type responses via release of IFN-γ from T cells, monocytes, and macrophages. Similarly, IL-18 can also be up-regulated by keratinocytes, and this is an important cytokine for Langerhans cell migration in mouse models of contact hypersensitivity. 50 IL-18, IL-1β, IL-6, and IL-12 appear to play key roles in initiation and development of immune responses in the skin. 51 Langerhans cells are localized primarily in the lower epidermal layers and function as papcs, capable of epidermal antigen processing. Following the internalization of antigen, LCs migrate to the regional lymph nodes with the help of integrin molecules to present processed antigen through their surface MHC-II and co-stimulatory molecules. 44 Recent studies have tried to determine the mechanisms by which LC migration occurs, and it appears that proinflammatory cytokines stimulate LC migration; specifically, allergens that cause contact hypersensitivity induce proinflammatory cytokines such as IL-1α, IL-1β, and TNF-α, which have been associated with hapten-induced LC migration. 51 It has been proposed that these cytokines diminish E- cadherin-mediated contacts between KCs and LCs

45 Figure 1-7. Cellular composition of the skin. Taken from Feldmeyer et. al, Implications of Cutaneous Biotransformation for Immune-Mediated Skin Rashes Although the skin is clearly capable of producing reactive metabolites, it is quite limited relative to the liver. However, once formed, the skin is very responsive to haptens through both LC and KC activation. The ability of drugs to form haptens in the skin presumably has much more serious implications for toxicity due to the lower amounts of metabolic enzymes in the skin available for drug clearance. In addition, activated T cells expressing skin-homing receptors 45 ultimately result in inflammatory responses, which often leads to drug-induced detachment of the skin due to massive keratinocyte death. When this occurs, conditions such as toxic epidermal 21

46 necrolysis or Stevens Johnson syndrome may result. 53 However, the mechanistic details of these rashes remain to be determined. Upregulation of Fas ligand (FasL) on keratinocytes has been observed in patients who develop drug-induced skin rashes, and the histological portrait in these patients is one of widespread keratinocyte apoptosis, sub-epidermal blistering, and lymphocytic/mononuclear cell infiltrate. 53 Natural killer T cells, also present during episodes of TEN, have been implicated, along with CD8+ cytotoxic T cells (CTLs) and natural killer (NK) cells. 53 Immunophenotypes of cells present in blister fluids from five patients who developed SJS-TEN induced by carbamazepine, phenytoin, and amoxicillin were analyzed in one study, and regardless of the offending drug, the majority of the cells in the blister fluids of these patients were CD3 + T cells (predominantly CD8 + CTL subset ; 33 70%), CD56 + NK cells (48 100%), and CD8 + CD56 + NKT cells. 54 Evidence for the involvement of CTLs in drug-induced skin killing include the observation that injection of granulysin directly into the skin of C3H mice induced significant dermal and epidermal necrosis and inflammatory infiltrates in the skin that was not seen when a similar experiment was performed with granzyme b and lysozyme. 53 Granulysin is produced by CTLs, which have been implicated in drug-induced detachment of epidermal-dermal layers. 53 Despite the myriad of clues regarding the activation of drugs in the skin leading to cutaneous immune-mediated reactions, much remains to be determined: what is the basis for specific HLA gene associations; how does drug activation lead to granulysin secretion; how does a rash progress from moderate severity to SJS-TEN; what is the difference between local versus systemic drug activation in inducing skin rash, and which is more important; can granulysin be used as a biomarker to screen for potentially bioactive drugs; and fundamentally, how do keratinocytes sense danger? 53 22

47 1.6 Role of Drug Metabolism, Reactive Metabolites, and Covalent Binding in IDRs As described above, both the liver and skin are capable of the biotransformation of drugs into reactive metabolites. The ability of reactive metabolites to conjugate to self-proteins forming novel antigenic compounds has been explored in drug toxicity; however, the relationship between covalent adducts and toxicity is, to date, not well understood. Whereas drugs such as halothane covalently bind to hepatic proteins, and resulting anti-drug antibodies have been detected in patients, acetaminophen also covalently binds in the liver and yet it does not induce an immunemediated hepatotoxicity. 34 The skin may be different in that there is a great degree of immunological activity, and the dominant response is not tolerance. Covalent binding of a drug may be necessary but not sufficient to induce and sustain an immune response. Correlation between the amount of binding and resulting toxicity has been attempted; one study, correcting for the total daily dose of the drug to reflect hepatic exposure, found a rough correlation in the amount of binding and risk that a drug would cause drug-induced liver injury. 55 The same study, when results were kept unadjusted, found a large overlap in DILI and non-dili inducing drugs regarding the amount of binding. 55 Clearly, other factors exist to influence immune mediated toxicity. One strategy that has been employed for screening drug candidates is the use of radiolabeled drug analogs to test for their ability to covalently bind to proteins. 38 Although relatively simple to perform in vitro (in vivo is difficult due to the large amount of radiolabeled drug required), false negatives may result if an enzyme capable of metabolizing the compound is absent from the system. Another problem with screening drug candidates for potential toxicity is the lack of metabolic similarity between rodents and humans (rodents metabolize and clear drugs/xenobiotics much faster than humans), thus interpretation based on these studies could lead to underestimation of risk Sulfotransferase Enzymes in Drug Metabolism and Toxicity One often overlooked pathway in the mechanism of drug toxicity and covalent binding is that of the Phase II reaction, sulfation. Sulfation occurs where sulfotransferase enzymes (SULTs) - catalyze the transfer of SO 3 to appropriate substrates. 37 The sulfate donor is PAPS (3-23

48 phosphoadenosine-5 -phosphosulfate), produced from APS kinase and ATP sulfurylase, which exists as a single bi-functional enzyme in mammals. 37 Serum sulfate levels in humans are limited to approximately 0.3 mm, 37 and sulfation is considered to be a high affinity, low capacity metabolic pathway. SULT enzymes exist in plants and all vertebrate classes examined (fish, birds, amphibians), but limited reports of sulfation in insects exist. 37 Sulfotransferases are classed into two groups, the membrane-bound Golgi complex SULTs, which act on endogenous steroid hormones, neurotransmitters, heparins, glucosaminoglycans, etc., and the cytosolic xenobiotic metabolizing SULTs of ~ 300 amino acid residues, which will be discussed here. 37 Table 1-1. Human Sulfotransferase Isoforms and Expression. Isozyme Typical Substrates Expression SULT 1A1 4-nitrophenol Liver, skin*, GI tract SULT 1A2 4-nitrophenol 2-napthol SULT 1A3 Dopamine Platelets, GI tract SULT 1B1 3,3,5 -Triiodothyronine Liver, GI tract SULT 1C2 4-nitrophenol Fetal lung, kidney SULT 1C4 4-nitrophenol nonylphenol SULT 1E1 17-β-estradiol Liver, lung, kidney SULT 2A1 dehydroepiandosterone Liver, adrenals, skin* SULT 2B1 dehydroepiandosterone Skin, prostate SULT 4A1 unknown brain Modified from P. Josephy 37 and F. Oesch. 56 The * indicates isoforms also found in rats. Sulfotransferases typically sulfate alcohol or phenolic groups; the nitrogen of N- substituted aryl and alicyclic chemicals; or pyridine N-oxides. 38 Cytosolic SULT enzymes are quite promiscuous, and there is a large overlap in substrates metabolized by various isoforms within subfamilies. 57 Although sulfation is generally regarded as a detoxification pathway, it can lead to bioactivation to reactive intermediates, which may or may not be harmful. For example in the case of the pyrimidine N-oxide drug minoxidil, sulfation must occur to form the N, O-sulfate ester in order to produce the desired effects on alopecia. In other cases, sulfation is implicated in activation of several classes of toxicants, such is seen with aryl hydroxylamines and hydroxamic acids, which are sulfated to metabolites involved in aromatic amine carcinogenesis. 37 The PPARγ agonist type-ii anti-diabetic drug, troglitazone, was first released onto the market in It was withdrawn only three years later due to idiosyncratic hepatotoxicity, which 24

49 appears to be multifaceted in nature. Although troglitazone liver injury is primarily and more seriously hepatocellular, cholestatic injury also occurs. Male rats were observed to sulfate the drug to a much greater degree than females the male rats were also observed to be more sensitive to troglitazone toxicity. 37 The sulfate metabolite is now recognized as one toxic metabolite of troglitazone, and is a potent inhibitor of the bile salt export pump in hepatocytes. 37,58 Benzylic and allylic alcohols can also undergo sulfation leading to the production of toxic metabolites. In this scenario, the sulfate can be cleaved heterolytically producing resonancestabilized electrophilic carbocation or nitrenium ions. Safrole is a good example where loss of SO 2-4 leads to a resonance-stabilized allylic cation, capable of binding to DNA in rat hepatocytes. This resonance stabilization of the remaining cation following loss of the sulfate from benzylic and allylic alcohols produces a reactive electrophilic intermediate. 1.7 Nevirapine Toxicity Nevirapine (NVP; Viramune TM ) is a non-nucleoside reverse transcriptase inhibitor indicated for the treatment of active HIV-1 infections. 59 NVP was released onto the market in June 1996, and prescribed first in its class as part of the highly-active anti-retroviral therapy regimen (HAART), given along with a protease inhibitor and a nucleoside reverse transcriptase inhibitor. Nevirapine is a highly effective drug, and a single dose can limit the vertical transmission of the virus from the mother to the fetus. Although efficacious at controlling HIV infections, NVP is associated with two adverse toxicities: hepatic damage and rash. 59 In 2000, the FDA placed a black box warning label on NVP due to hepatotoxicity, which occurs in 6% of patients and can be life threatening. 59 Liver injury normally resolves when the drug is stopped, but it can lead to fulminant liver failure and death. In 8-18% of patients, NVP hepatotoxicity can manifest as asymptomatic elevated serum alanine transaminase (ALT) levels, which is the first indication of liver injury and typically occurs within the first six weeks of treatment. 60 There also exists evidence for increased risk of liver injury in non-hiv patients, which may be due to higher CD4+ T cell counts. 61 The initial therapeutic dose of NVP was 400 mg/kg/day; at that dose it induced skin rashes, most of which were mild to moderate in nature, in 32-48% of patients. In contrast, when a lead in dose of 200 mg/kg/day for the first two weeks of treatment was introduced, the incidence 25

50 of skin rash was decreased to 17%. Currently the incidence of NVP skin rash is 9%; however, 16% of skin rash patients develop very severe rash in the form of SJS or TEN. 59 Numerous risk factors have been reported for the development of the NVP-induced toxicities. For example, the HLA-DRB*01 allele has been associated with the development of hepatotoxicity in NVP patients. 62,63 Risk factors associated with development of the skin rash have been better characterized, and they include, but are not limited to, ethnicity (Chinese populations are more sensitive), female gender, co-therapy with antihistamines and corticosteroids, and a higher pre-therapy CD4+ T cell count. 64 The first 6 weeks of NVP treatment is the riskiest time for patients on NVP to develop liver injury or rash. 59 Our laboratory has developed and well-characterized an animal model of the NVP-induced skin rash in the female Brown Norway (BN) rat. 65 This model represents the proportion of the patients who only develop a skin rash, and it is to date one of only two excellent models of an idiosyncratic reaction which occurs in humans (the other is D-penicillamine-induced autoimmunity). Although patient samples would be ideal when attempting to study IDRs, for ethical and other reasons, collection of these sample types is not always feasible. It is also not feasible to methodically control specific variables to test hypotheses in humans. Therefore, animal models such as this represent the next best approach in trying to understand the mechanisms of IDRs. Using animal models as screening tools may also become viable in the future if they can be developed. The NVP model is reviewed in the following section Animal Model of Nevirapine-Induced Skin Rash Shenton et al. from our group first characterized the BN rat model of NVP skin rash in The skin rash in rats occurs over a period of about 3 weeks, with the ears turning red on day 7, and the development of lesions and sloughing of skin between days of treatment. 65 The incidence of rash in female BN rats is 100% whereas only 20% of female Sprague-Dawley rats develop a rash at 3 weeks or later. 65 The rash in rats ranges from mild to moderate severity, and as in humans, there appears to be quite an individual response of the rats to NVP. When rats are removed from NVP and re-challenged, the rash occurs must faster (within 7 days) and the animals develop a systemic sickness (lethargic, uninterested in surroundings, etc). Partial depletion of CD4+ T cells was protective in rats, and depletion of CD8+ T cells, if anything, appeared to make the rash worse. 66 In addition, splenocytes harvested from re-challenged rats 26

51 were able to transfer susceptibility to NVP in naïve recipients through i.v. injection (adoptive transfer studies). 66 Pre-treatment with the immunosuppressants, tacrolimus and cyclosporine, was able to prevent and even resolve the rash during treatment. 66 Other characteristics include increased incidence in female BN rats, increased incidence of rash with increased dose, and tolerance induction through low dose pre-treatment. 59 These characteristics are very similar to the occurrence of rash in humans and strongly support the role of the immune system in induction of the rash. Morphological features of the rash are similar in humans and rats; for example, both develop maculopapular raised lesions and general redness of the skin; some animals develop sloughing of the skin. It is a generalized rash and can even affect mucous membranes in the animals as well as humans. In rats, crusting of eyes and scabbing around the footpad as well as nose have been observed. Histologically, the dermal infiltrate is also similar in humans and rats, with a mild perivascular lymphocytic infiltrate which progresses to a more severe infiltrate later on. A mononuclear infiltrate has been observed in both the dermis and epidermis of patients who develop SJS and TEN; 67 this infiltrate has also been seen predominantly in the dermis of rats. 65 Given the numerous similarities between the animal model and humans, this model has provided us with the unique opportunity to study the mechanism of at least one IDR in detail. It is true that the mechanisms of IDRs for different drugs are likely very different; therefore, this is but one piece in a large puzzle Nevirapine Metabolism leading to Skin Rash A fundamental question in the understanding of IDRs is whether it is the parent drug or a reactive metabolite that is responsible for causing toxicity. The NVP model has allowed us to determine which metabolic biotransformation pathway is responsible for the rash. In 2008, Jie Chen from our group found that the 12-hydroxylation pathway is responsible for causing the skin rash. 68 This pathway is a major route of NVP metabolism in both humans and rats. 69 NVP undergoes Phase I oxidation by P450 enzymes in the liver in both humans and rats, and it is further metabolized to glucuronide conjugates, which are the major metabolites in urine. It can also be further oxidized to a carboxylic acid metabolite. 69 CYP3A4 and CYP3A5 are, respectively, the major and minor P450 isoforms responsible for converting NVP to the 12- hydroxynevirapine (12-OH-NVP) metabolite in the liver of humans

52 Although there are many potential reactive metabolites of NVP, 68 the 12-hydroxy pathway was shown to be responsible through two key lines of evidence. The first is that feeding 12-OH- NVP at half the standard dose of NVP required for 100% incidence (75 mg/kg vs 150 mg/kg) was able to induce the same degree of rash in female BN rats. 68 The second was using the deuterium isotope effect. If the rate limiting step in forming the reactive metabolite is the oxidation of the methyl group to form 12-OH-NVP, then replacement of the methyl hydrogens with deuterium should decrease the rate of oxidation and decrease the production of the reactive metabolite, thus decreasing the incidence of rash. The deuterated analog did cause a lower incidence of rash; however, given at the same dose, the blood level of the deuterated analog was much lower than that of nevirapine. 68 This was unexpected, and it is proposed that the reason for this is that the carbon radical intermediate produced from P450 oxidation can partition between oxygen rebound to form 12-OH-NVP, or undergo hydrogen atom loss to form a reactive quinone methide species. 68 The quinone methide is a strong electrophile and inactivates P450 enzymes. Because deuteration decreases the rate of methyl oxidation, less reactive metabolite is formed and there is less P450 inactivation and more metabolism of the deuterated analog through other oxidative pathways. Despite determining that formation of 12-OH-NVP was required to produce the rash, it was not known which chemical species ultimately induces the rash. 12-OH-NVP is the same oxidation state as the quinone methide and cannot spontaneously rearrange into the reactive quinone methide species. This thesis proposes the 12-OH-NVP metabolite produced from hepatic metabolism reaches the skin through the general circulation where it is sulfated by skin-resident sulfotransferase (SULT) enzymes, such as those present in keratinocytes (KCs). Sulfate is a good leaving group, and attack by nucleophilic groups on cutaneous proteins could lead to cutaneous adducts and initiate an immune response Role of Sulfation in Nevirapine-Induced Skin Rash Both human and rat liver and skin tissue contain sulfotransferase enzymes 71 capable of sulfating the 12-OH-NVP metabolite. Production of the sulfate metabolite was confirmed by various studies performed previously by the Uetrecht group, where the circulating plasma sulfate levels were found to be between 1-8 µg/ml in NVP-treated rats. Previous projects had tried to determine mechanistically if the sulfate metabolite of NVP was responsible for causing skin rash 28

53 through metabolic manipulation studies. Unfortunately, at the time these studies were performed, methods to detect covalent adducts in the skin had not been successful. An overview of these previous experiments is presented here and they were the starting point for the current work. In previous studies, salicylamide (SA) and molybdenum were utilized to interfere with production of 12-OH-NVP sulfate. Salicylamide is a chemical that undergoes sulfation and depletes PAPS, 72 while molybdenum interferes with the synthesis of PAPS. 73 Salicylamide and molybdenum were shown to deplete circulating 12-OH-NVP sulfate levels to below the limit of detection for the mass spectrometer; however, rats still developed typical rash and in the case of molybdenum, became very sick and died. It was not known at that time if binding in skin had been prevented. Follow-up studies utilized dehydroepiandosterone (DHEA), a competitive inhibitor of SULT 2A1, an isoform found in rat skin for which female rats exhibit a 10-fold greater expression than males. DHEA was partially successful in preventing the skin rash; however, it interfered with metabolism of NVP and 12-OH-NVP, making it very difficult to interpret the results. Again, at this time it was not known what was occurring in the skin. 39 All three of these compounds were administered via gavage and presumably, at least in the case of salicylamide and molybdenum, inhibited the production of hepatic 12-OH-NVP sulfate leading to decreased systemic concentrations. Previous studies also indicated that the 12-OH-NVP sulfate metabolite exhibited very low reactivity to nucleophiles in vitro, even in the addition of base; 11 all of these results taken together suggested the sulfate metabolite was not involved. However, nucleophilic groups on cutaneous proteins may behave quite differently, and this was also examined in this work. A simple explanation for the discrepancies in these earlier studies may be that what is important is the formation of the sulfate in the skin. If the turnover of salicylamide in the skin is slow it may not deplete PAPS in the skin. The skin possesses SULT enzymes and the 12-OH- NVP can reach the skin through the systemic circulation. Therefore, the focus of the current work was to determine if the 12-OH-NVP sulfate formed in the skin was responsible for the rash. Repetition of some of these previous experiments with the addition of examination of events occurring in the skin were performed, as were direct testing of sulfate reactivity with cutaneous proteins. Methods were also developed for isolating skin fractions. It was also determined if the events occurring in skin correlated to the events occurring systemically in rats. 29

54 1.7.4 Nevirapine Metabolism in the Liver and Lack of Hepatotoxicity in Rats The standard dose of NVP required to induce a 100% incidence of rash in female BN rats is 150 mg/kg/day, and this dose produces a trough plasma level of 40 µg/ml. 68 However, at this dose or lower, there is no occurrence of asymptomatic ALT increases or hepatotoxicity in rats even though metabolism to the quinone methide occurs, and binding to expressed P450 was previously identified by Dr. Yan Li from the Uetrecht group. 74 As mentioned previously, the dominant response in the liver is immune tolerance, and this could be the key reason for lack of hepatotoxicity. Another reason may the issue of cell stress; revisiting the Danger Hypothesis, it could be that binding in the liver does not induce danger, while in the skin, it does. Because of previous failed attempts at producing liver injury in rats, another focus of this thesis was to develop a mouse model of NVP-induced liver injury. There are numerous strains of knock-out mice available and many more reagents, stains, and procedures for use with mice. Two knock-out strains of mice tested for increased liver toxicity in this work were the casitas-blineage-lymphoma-b (Cbl-b) and programmed-cell death 1 (PD-1) knock-outs. The Cbl-b is null for the E3 ubiquitin ligase and exhibits impaired immune tolerance through interference with T cell receptor (TCR) and transforming growth factor-beta (TGF-β) signaling. 75 They also exhibit impaired T cell anergy and development of peripheral Foxp 3+ regulatory T cells (Treg). 75 Additionally, if covalent adducts are involved with causing liver injury, than impaired protein degradation should increase the level of adducts; in this way the Cblb mice may also be more sensitive. The PD-1 knock-outs lack the ability to control and maintain peripheral T cell tolerance; 76 therefore, tolerance might be overcome to initiate an immune response. Both strains, as well as numerous wild type strains, were tested in this work to determine the involvement of hepatic covalent adducts in NVP-induced liver damage. 1.8 Research Hypotheses The overall hypothesis is that NVP-induced toxicity is mediated by the formation of a reactive metabolite of NVP; in the case of the liver, a quinone methide species and in the case of skin, the sulfate metabolite; which each go on to covalently modify self-proteins leading to an immune response. There are five overall objectives from this hypothesis: 30

55 1. To detect hepatic covalent adducts in rats and mice as well as human hepatic microsomes and examine their involvement in hepatotoxicity. 2. To develop a mouse model of NVP-induced hepatotoxicity. 3. To determine if the 12-OH-NVP sulfate metabolite can covalently bind to cutaneous proteins. 4. To determine if covalent binding occurs in the skin of animals that develop a rash. 5. To determine if inhibiting sulfation prevents covalent binding in the skin and the skin rash. 31

56 CHAPTER 2 Bioactivation of Nevirapine to a Reactive Quinone Methide: Implications for Liver Injury Science does not know its debt to imagination. - Ralph Waldo Emerson This work has been published in the following journal and is reproduced with permission: Amy M. Sharma, Yan Li, Maria Novalen, M. Anthony Hayes, and Jack Uetrecht. Bioactivation of Nevirapine to a Reactive Quinone Methide: Implications for Liver Injury. Chemical Research in Toxicology, , Epub 2012 July 13. Reprinted with permission. Copyright 2012 American Chemical Society. All rights reserved. In this chapter, all experiments were performed by Amy M. Sharma except Figures 2-2, 2-4 A and B, 2-6 and

57 2.1 Abstract Nevirapine (NVP) treatment is associated with a significant incidence of liver injury. We developed an anti-nvp antiserum to determine the presence and pattern of covalent binding of NVP to mouse, rat, and human hepatic tissue. Covalent binding to hepatic microsomes from male C57BL/6 mice and male Brown Norway rats was detected on western blots; the major protein had a mass of ~55 kda. Incubation of NVP with rat CYP3A1 and 2C11 or human CYP3A4 also led to covalent binding. Treatment of female Brown Norway rats or C57BL/6 mice with NVP led to extensive covalent binding to a wide range of proteins. Co-treatment with 1-aminobenzotriazole dramatically changed the pattern of binding. The covalent binding of 12-hydroxy-NVP, the pathway that leads to a skin rash, was much less than that of NVP, both in vitro and in vivo. An analog of NVP in which the methyl hydrogens were replaced by deuterium also produced less covalent binding than NVP. These data provide strong evidence that covalent binding of NVP in the liver is due to a quinone methide formed by oxidation of the methyl group. Attempts were made to develop an animal model of NVP-induced liver injury in mice. There was a small increase in ALT in some NVP-treated male C57BL/6 mice at 3 weeks that resolved despite continued treatment. Male Cbl-b -/- mice dosed with NVP had an increase in ALT of >200 U/L, which also resolved despite continued treatment. Liver histology in these animals showed focal areas of complete necrosis while most of the liver appeared normal. This is a different pattern than the histology of NVP-induced liver injury in humans. This is the first study to report hepatic covalent binding of NVP and also liver injury in mice. It is likely that the quinone methide metabolite is responsible for NVP-induced liver injury. 2.2 Introduction Nevirapine (NVP, Viramune TM, Figure 2-1) is a non-nucleoside reverse transcriptase inhibitor used for the treatment of HIV-1 infections. Treatment with NVP is associated with a significant incidence of idiosyncratic skin rashes and/or liver toxicity. 66 The incidence of skin rashes is approximately 9%. They are usually mild to moderate in nature; however, 16% of NVPinduced rashes are very severe, including Stevens-Johnson syndrome and toxic epidermal necrolysis. 59 In 2000, the FDA placed a black box warning on NVP due to hepatotoxicity, which occurs in 6% of patients and can be life threatening. 59 The incidence of elevated serum alanine transaminase (ALT) in NVP-treated patients, which is the first indication of liver injury, is 33

58 between 8-18% and typically occurs within the first six weeks of treatment. 77 Liver injury normally resolves when the drug is stopped, but it can lead to fulminant liver failure and death. There also exists evidence for increased risk of liver injury in non-hiv patients, which may be due to higher CD4 cell counts. 61 The mechanisms of idiosyncratic liver injury and skin rashes are currently unknown, but most idiosyncratic drug reactions appear to be mediated by reactive metabolites. We developed an animal model of NVP-induced skin rash in Brown Norway (BN) rats that is clearly immunemediated and has characteristics very similar to the rash in humans; however, the rats did not develop liver toxicity. 65,66 We postulated that the 12-hydroxylation pathway was involved in the induction of the skin rash; therefore we replaced the hydrogens on the methyl group with deuterium to slow down the rate of 12-hydroxylation (Figure 2-1). We found that this analog (DNVP) did not cause a skin rash as predicted, but instead of higher blood levels because one of the major metabolic pathways was inhibited, we found that the blood levels of DNVP were actually much lower than those of NVP at the same dose. 68 Although the reason for this was not immediately obvious, we ultimately concluded that, in addition to oxygen rebound to form 12- OH-NVP, the intermediate free radical in the P450-mediated oxidation could also lose a hydrogen atom to form a reactive quinone methide (Figure 2-1). A glutathione conjugate consistent with the quinone methide intermediate has been reported 78,79 ; however, it could also come from a sulfate conjugate of the 12-OH-NVP. In this study we used an antiserum against NVP to study the covalent binding of NVP, DNVP, and 12-OH-NVP to hepatic proteins in mice, rats, and humans. We also studied the effects of chronic administration of NVP to various strains of mice to determine if it causes liver injury. In addition to C57BL/6 and BALB/c we included the Casitas B-lineage lymphoma-b (Cbl-b) knockout mouse (Cbl-b -/- ), which is bred on a C57BL/6 background. The Cbl gene is a mammalian gene that encodes a variety of proteins, specifically those involved in cell signaling and protein ubiquitination. Lack of ubiquitination of NVP protein adducts could lead to more persistent covalent binding and possibly toxicity. This also impairs immune tolerance; therefore, if the liver injury is immune-mediated these animals should be at increased risk. These animals also express a mouse isoform of CYP3A4 (CYP3A11); therefore, oxidative metabolism of NVP should occur 80, and this has the potential to lead to liver injury. 34

59 Figure 2-1. Bioactivation pathway of NVP leading to liver injury. 35

60 2.3 Materials and Methods Chemical Materials. NVP was kindly supplied by Boehringer-Ingelheim Pharmaceuticals Inc. (Ridgefield, CT). The majority of chemical reagents (1-aminobenzotriazole (ABT), tris(hydroxymethyl)aminomethane base, methanol, DMSO, phosphate-buffered saline (PBS, ph 7.4), glycerol, silica gel, etc) were obtained from Sigma-Aldrich (Oakville, ON) unless otherwise noted in the methods. Ammonium persulfate was obtained from Fisher Scientific (Fair Lawn, NJ). Sodium dodecyl sulphate and Tween-20 were obtained from BioShop (Burlington, ON). Stock acrylamide/bis solution (29:1, 3.3% C), non-fat blotting grade milk powder, and nitrocellulose membrane (0.2 µm) were purchased from Bio-Rad (Hercules, CA). Ultra pure tetramethylethylenediamine was purchased from Invitrogen (Carlsbad, CA). Amersham ECL Plus Western Blotting Detection System was obtained from GE Healthcare (Oakville, ON). Horseradish peroxidise-conjugated goat anti-rabbit IgG (H + L chains) and monoclonal GAPDH were purchased from Sigma-Aldrich (St. Louis, Mo). Normal goat serum was obtained from Invitrogen (Grand Island, NY). Expressed human CYP3A4, rat CYP3A1, and rat CYP2C11 (each with P450 reductase and cytochome b 5 ), 0.5 M potassium phosphate ph 7.4, and NADPH regenerating system solutions A and B were purchased from BD Biosciences (Woburn, MA) Instruments and Software. AlphaEaseFC (FluorChem 8800) manufactured by Alpha Innotech, now Cell Biosciences Santa Clara, California, USA was used to image blots. Integrated density values were obtained using the SPOT DENSO function on the FluorChem 8800 Imager Synthesis of 12-trideutero-NVP (DNVP). Synthesis of DNVP was carried out using the method described by Chen et al., H NMR (CDCl 3 ): δ (m, 2H), (m, 2H), (m, 1H), 7.06 (d, J = 4.8 Hz, 1H), 7.19 (dd, J = 4.8, 7.5 Hz, 1H), 8.01 (dd, J = 2.1, 6.6 Hz, 1H), 8.08 (d, J = 4.8 Hz, 1H), 8.50 (dd, J = 1.8, 4.8 Hz, 1H), 9.90 (bs, 1H). ESI-MS; m/z (%) 270 (MH +, 100%). The ratio of the 36

61 peaks at m/z 267:268:269:270 as determined by mass spectrometry was 0:0.007:0.124:0.869, indicating only trace amounts of NVP Production of Anti-NVP Anti-Serum in Male White New Zealand Rabbits. Scheme 2-1. Synthetic pathway of the immunogen used for induction of anti-nvp antiserum. Synthesis of NVP-NAC Conjugate. The synthesis of the immunogen is outlined in Scheme 1. The first step in producing the anti-nvp antiserum was to synthesize 12-OH-NVP (2) and convert this to the benzylic chloride (12-Cl-NVP, 3). The method to produce 12-OH-NVP followed the protocol described previously 81 with minor modifications. ESI-MS; m/z (%) 283 (MH+, 100%). 37

62 To convert 12-OH-NVP to 12-Cl-NVP, the method of Kelly et al. 82 was followed. To 12-OH- NVP (200 mg) in dry dichloromethane (10 ml) at 0 ºC was added N,N-diisopropylethylamine (0.14 ml) followed by thionyl chloride (3 ml) and stirred under argon at room temperature for 3 h after which the thionyl chloride was evaporated by rotary evaporation. The reaction mixture was then extracted with ethyl acetate (3 10 ml). The ethyl acetate layer was washed with water (10 ml), dried over anhydrous sodium sulfate, and concentrated to yield crude product, which was purified with open column chromatography (silica gel, pore size 60 Å, mesh, column dimensions mm) eluted with 50% ethyl acetate/hexanes to yield g of yellow solid. ESI-MS; m/z (%) 301 (MH +, 100%). The 12-Cl-NVP (1.78 g, 3.55 mmol) was dissolved in 18 ml of tetrahydrofuran and reacted with N-acetylcysteine (NAC, 2.31 g, mmol) in 5 ml of triethylamine under argon reflux for 2 h. The crude mixture was cooled to room temperature, acidified to ph 3-4 by 1N HCl and extracted with CHCl 3. The organic layer was dried over anhydrous sodium sulfate. Chloroform was removed under reduced pressure. Nevirapine-NAC conjugate was obtained as a pale yellow solid (4). Formation of the nevirapine-nac conjugate was confirmed by mass spectrometry ESI-MS; m/z (%) 428 (MH +, 100%). Preparation of NVP-KLH Conjugate. All reagents and glassware were dried in a vacuum at 50 ºC. Activation of the carboxy groups on NAC of the synthesized 12-NAC-NVP occurred as follows: to 61.4 mg 12-NAC-NVP was added mg of N-hydroxysuccinimide and mg of 1-ethyl-3-(3-dimethyllaminopropyl) carbodiimide hydrochloride. Anhydrous DMF (4 ml) was introduced via syringe at 0 ºC. The entire mixture was sealed with a rubber stopper and stirred at 0 ºC for 2 h under N 2. Methylene chloride (8 ml) was added, followed by washing with water (3 8 ml) and then the organic layer was partially evaporated in vacuo to yield a pale yellow solution (0.5 ml, 5). DMF (4 ml) was added followed by Keyhole limpet hemocyanin (KLH, 8 mg) and the mixture was stirred for 1 h at 4 ºC. The reaction mixture was then concentrated under a N 2 stream and 1 ml water was added. Centrifugal filtration was performed to collect the protein solution, which was then lyophilized. A final white powder (10.4 mg) was obtained (6) and stored at -20 ºC. The same method was used to prepare a conjugate with bovine serum albumin (BSA) MALDI MS; m/z 67,139-68,569. The hapten density of the BSA conjugate was approximately 4.5 molecules of NVP-NAC/BSA as determined by the increase in mass on mass spectrometry. 38

63 Production of Anti-NVP-NAC-KLH-Antiserum. Polyclonal anti-nvp-nac-klh antibodies were raised in two individual 2 kg, male, pathogen-free New Zealand White rabbits (Charles River, Quebec) housed in the animal care facility at The Division of Comparative Medicine, University of Toronto. Each animal was immunized with the NVP-NAC-KLH conjugate (1 mg antigen µl glycerol in 1.8 ml of phosphate buffered saline emulsified with an equal volume of Freund s complete adjuvant) subcutaneously at multiple sites. Injections with 500 µg of NVP-NAC-KLH in Freund s incomplete adjuvant divided into six to eight subcutaneous sites were repeated 4, 6, 8, and 12 weeks after the initial immunization. The animals were exsanguinated while under pentobarbital anesthesia 10 days after the final immunization. The serum was heat-inactivated at 56 C for 30 min before being stored at -80 C. ELISA. NVP-NAC-BSA, BSA, or KLH (100 µl, 10 µg/ml in carbonate-bicarbonate coating buffer) were coated into the wells of a flat-bottom 96-well plate (Costar, Cambridge, MA) and the plate was incubated overnight at 4 C. The plates were washed with ELISA wash buffer (50 mm tris(hydroxymethyl)aminomethane-buffered saline, ph 8.0, 0.05% Tween-20) three times and blocked by the addition of 100 µl of post-coat solution (50 mm Tris-buffered saline, ph 8.0, 1% BSA) for 30 min at room temperature. Following the blocking step, the wells were washed three times and various dilutions of the anti-nvp-nac-klh antiserum or pre-immune serum were added to the plates, which were then incubated at room temperature for 2.5 h. The plates were subsequently washed three times with ELISA wash buffer and horseradish peroxidaseconjugated goat anti-rabbit IgG (diluted 1:5000 in post-coat solution; 100 µl) was added to each well. The ELISA plates were incubated at room temperature for 2 h. Plates were then washed three times with ELISA wash buffer. Enzyme substrate (3,3,5,5 -tetramethylbenzidine peroxidase substrate and peroxidase solution B, Kirkegaard & Perry Laboratories) was mixed in equal volumes and 100 µl of the enzyme substrate was added to each well. The plate was incubated in the dark at room temperature for 10 min. Sulfuric acid (2M, 100 µl) was added to each well to quench the reaction. Absorbance was measured with the Basic Endpoint Option of SoftMax Pro 5 Software, using the SPECTRA maxplus384 plate reader (Molecular Devices Technologies) set at 450 nm. 39

64 2.3.5 Animal Care. Male ( g) or female BN rats ( g) were obtained from Charles River (Montreal, Quebec). Rats were housed in pairs in standard cages in a 12:12 h light/dark cycle with access to water and Agribrands powdered lab chow diet (Leis Pet Distribution, Inc. Wellesley, Ontario) ad libidum. Following a one week acclimatization period, rats were either maintained on control chow or started on drug containing diet (treatment groups). Drug was mixed thoroughly with powdered lab chow if it was to be administered orally. The amount of drug administered to animals was calculated based on body weight of the rats and their daily food intake. Rats were sacrificed via CO 2 asphyxiation. Male Balb/c or C57BL/6 mice (6-8 weeks age) were obtained from Charles River (Montreal, Quebec). Cbl-b -/- knockout mice were bred in house from animals first developed by Dr. J. Penninger at the Institute of Molecular Biotechnology of the Austrian Academy of Science, Vienna, with his kind permission. Mice were kept 4 per cage. The average weight gain was approximately 0.75 g per week (data not shown). NVP was administrated in lab chow following a one week acclimatization period. Animal experiments were approved by the University of Toronto animal care committee in accordance with guidelines of the Canadian Council on Animal Care Treatment of Animals with NVP, 12-OH-NVP, DNVP, or ABT. Female BN rats were treated with NVP or DNVP at 150 mg/kg/day, or 12-OH-NVP at 159 mg/kg/day (equimolar dose) orally in standard rat chow for either 8, 10, or 21 days. Dosages were based on previous work showing induction of rash at these levels. 65 Treatment of NVP or DNVP by s.c. injection lasted 21 days with a dose of 75 mg/kg/day of either compound. ABT was dissolved in water (20 mg/ml) and administered via gavage at a dose of 50 mg/kg/day. If ABT was to be given to animals, the dose of NVP was 50 mg/kg/day via gavage. Methylcellulose (0.5%) was used to suspend NVP or metabolites given to rats by gavage or s.c. injection. All mice were started on NVP at 950 mg/kg/day in standard chow after preliminary studies showing no apparent toxicity or mortality of mice at either 550 or 950 mg/kg/day. 40

65 2.3.7 Incubations with Microsomes or Supersomes. Livers were homogenized in ice-cold 1.15% KCl using a Polytron 2100 homogenizer and centrifuged at 26,400 g for 10 min at 4 C. The supernatant was then centrifuged at 100,000 g for 50 min at 4 C. The pellet was homogenized in 4 volumes of glycerol-phosphate-kcl buffer and aliquots were stored at -80 C. The protein concentration of the prepared microsomes was quantified using a BCA protein assay kit (Novagen, EMD Biosciences Inc.). All incubations were performed at 37 ºC. NVP, 12-OH-NVP or DNVP stock solutions were prepared in methanol and the final methanol concentration in the reactions did not exceed 1% for any incubation. 83 The microsomal incubations consisted of 100 mm potassium phosphate buffer (ph 7.4), a NADPHregenerating system (Solution A final concentrations: 1.3 mm NADP, 3.3 mm glucose-6- phosphate, 3.3 mm MgCl 2 ; Solution B final concentration: 0.4 Units/mL glucose-6-phosphatedehydrogenase), and microsomal homogenate (final protein concentration varying from 0.3 mg/ml to 15 mg/ml). EDTA 2Na (0.4 mm) was added to rat CYP3A1 and 2C11 incubations and water was added to each incubation to reach a final volume of 400 µl for rat and mouse or 200 µl for human 3A4 incubations. 84 Incubations consisting of all reaction components except the NADPH regenerating system or drug were preincubated for 5 min. The NADPH-regenerating system or drug was added to each of the test and control tubes after the 5 min preincubation. Reactions were stopped by placing the sample vials on dry ice and stored at -80 C. 84 microsomal incubations were to be analyzed via LC/MS, 250 µl ice cold acetonitrile was used to quench the reaction and internal standard (ethyl-nvp a NVP derivative in which the cyclopropyl group has been replaced with an ethyl group, 5.4 µg/ml, 50 µl) was added to each tube, contents were centrifuged, separated by solid phase extraction (Strata solid phase extraction column C18-E, 100 mg, by Phenomenex), evaporated in vacuo at 50 ºC, and reconstituted to 50 µl prior to analysis Quantification of NVP and its Metabolites from Microsomal Incubations. Samples were re-constituted to 50 µl with mobile phase (16% acetonitrile and 84% water with 2 mm ammonium acetate and 1% acetic acid). The samples were separated by HPLC and analyzed by mass spectrometry. The separation was performed on an Ultracarb C18 30 X 2.0 mm, 5 µm column (Phenomenex) under isocratic conditions with a mobile phase consisting of 16% If 41

66 acetonitrile and 84% water with 2 mm ammonium acetate and 1% acetic acid. The flow rate was 0.2 ml/min Mass Spectrometry Analysis. Mass spectrometry was carried out using a PE Sciex API 3000 quadrupole system with an electrospray ionizing source. The ion pairs used for this analysis were: 267.0/226.1 for NVP, 283.1/223.1 for 12-OH-NVP, 297.1/210.1 for 4-COOH-NVP, 283.1/161.0 for 2-OH-NVP, 283.1/214.0 for 3-OH-NVP, 255.1/227.2 for ethyl-nvp (positive ionization mode). Standard curves prepared for 2-OH-NVP ( µg/ml), 3-OH-NVP ( µg/ml), 12-OH- NVP ( µg/ml), 4-COOH-NVP ( µg/ml), and NVP ( µg/ml) had R 2 values of > Analysis of Covalent Binding Using SDS-PAGE and Immunoblotting. Livers homogenized in working cell lysis buffer (Cell Signaling Technologies, Pickering, ON) containing 1X HALT Protease Inhibitor Cocktail (Pierce, Rockford, IL) with a Polytron 2100 homogenizer and centrifuged at 1000 g for 15 min and supernatant was collected and again centrifuged at g for 30 min. The supernatant was mixed with Pierce reducing sample loading buffer in a 4:1 protein to buffer ratio and boiled for 5 min. Sodium dodecyl sulfate polyacrylamide gel electrophoresis was performed using the Protean-3 minigel system (BioRad, Mississauga, ON). Gels were hand-cast (8%) or bought from Bio-Rad Canada (12%), and were run at 130 V. Electrophoresis running buffer (Bio-Rad) consisted of 25 mm Tris base, 192 mm glycine, and 0.1% sodium dodecyl sulfate, ph 8.3. Transfer to nitrocellulose membrane (0.2 µm, BioRad) occurred at 0.13 ma for 90 min at 4 ºC using the same Protean-3 minigel system (BioRad, Mississauga, ON). Tris-glycine transfer buffer (Bio-Rad) consisted of 25 mm Tris, 192 mm glycine, and 20% methanol at ph 8.5. Membranes were washed twice in trisbuffered saline tween-20 (TBST) wash solution for 5 min. Membranes were then blocked in 5% non-fat milk blocking solution in TBST. Blocking was done for 90 min at room temperature. Membranes were then rinsed with three changes of TBST for 5 min each and incubated with a 1:100 or 1:500 dilution of primary anti-nvp antiserum and 10% normal goat serum in TBST overnight at 4 ºC. A 20 min wash (three changes) in TBST after overnight blocking was followed by a 90 min incubation in secondary antisera (1:2000 or 1:5000 dilution) in TBST containing 10% goat serum. The secondary antisera was goat anti-rabbit horseradish peroxidase antisera. 42

67 Membranes were washed 3 times for 20 min with TBST. All blots were incubated with enhanced chemiluminescence stain for 5 min and analyzed with a FluorChem8800 imager. To probe for the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) loading control, membranes were stripped of primary anti-nvp anti-serum using Pierce Restore Plus buffer (Pierce, Rockford, IL) for 15 to 20 min at room temperature followed by a one h blocking step. Membranes were then incubated in mouse monoclonal anti-gapdh antisera (1:40,000) and processed as above except the secondary antisera was goat anti-mouse horseradish peroxidase antisera diluted 1:10,000 (Jackson ImmunoResearch, Baltimore Pike, West Grove, PA.) Analysis of in Vivo Covalent Binding using Immunohistochemistry. The liver samples were fixed in 10% formalin and the paraffin block, hematoxylin/eosin slides, or unstained sections were prepared at the Toronto Hospital for Sick Children. For immunohistochemical staining, non-specific sites were blocked with 10% goat serum for 1 h. At this point the anti-nvp antiserum was diluted to 1:100 in 10% goat serum and applied to each section overnight. Following a washing step, slides were submerged in 0.3% hydrogen peroxide in methanol for 10 min to block endogenous peroxidases followed by a washing step. Secondary antiserum (goat anti-rabbit IgG-HRP conjugated antisera) was applied to the sections at a dilution of 1:3000 in 10% goat serum. Sections were incubated with secondary antiserum for 2 h. After a final washing step, Vector NovaRED stain was added as a substrate for the peroxidases following package protocol. Sections were then counterstained with Mayer s hematoxylin (Sigma), dehydrated by sequential immersion in increasing concentrations of ethanol, cleared in xylenes, and mounted using Permount mounting medium (Fisher, Markham, ON) Plasma Alanine Transaminase and Cytokine Analysis. Alanine transaminase (ALT) was assayed using the Infinity TM ALT (glutamic pyruvate transaminase) Liquid Stable Reagent kit by Thermo Scientific. Screening for cytokines was performed using a Luminex immunoassay mouse cytokine/chemokine kit from Millipore Corporation (Milliplex Map Kit). Homogenized liver tissue or serum samples prepared to the kit specifications were plated and analyzed following the manufacturer s instructions. 43

68 2.4 Results Characterization of the Anti-NVP-NAC-KLH Antiserum. ELISA analysis showed the anti-nvp-nac-klh antiserum recognized the NVP-NAC- BSA conjugate or KLH, but not BSA alone (Figure 2-2A). The binding of the antisera to the NVP-NAC-BSA conjugate was inhibited by preincubating the anti-nvp-nac-klh antiserum with NVP or its metabolites (Figure 2-2B). Inhibition was much less with 2-OH-NVP, 3-OH- NVP, and 4-COOH-NVP (the metabolite in which the methyl group has been oxidized to a carboxylic acid). Binding could still be detected at an anti-serum dilution of 1/1,000,

69 Figure 2-2. ELISA analysis showing (A) binding of the anti-nvp-nac-klh antiserum to the NVP-NAC-BSA conjugate, KLH, or BSA and (B) the effect of preincubation of the antiserum with NVP or its metabolites on the binding of the antisera to the NVP-NAC-BSA conjugate. Data represent the mean ± s.d. from 3 incubations Covalent Binding of NVP, DNVP, or 12-OH-NVP to Hepatic Microsomes in Vitro and Comparison to in Vivo Hepatic Covalent Binding. When microsomes produced from male BN rats were incubated with NVP, 12-OH-NVP, or DNVP, the greatest covalent binding observed was with NVP, and the stongest band was at ~55 kda (Figure 2-3A), which corresponds to the mass of the male dominant P450 2C11/3A1 isoforms. 85 Incubation of mouse liver microsomes with NVP produced a band of slightly higher mass, ~57 kda (Figure 2-3B), corresponding to the mass of the dominant murine P450 3A Significant covalent binding of 12-OH-NVP was not observed with rat microsomes, and DNVP produced a much fainter band at 55 kda than NVP in both rodent species tested. In vivo experiments with either species displayed a wide range of covalently-modified bands that were much more intense than from in vitro experiments. The covalent binding of DNVP to both rat and mouse hepatic microsomes was also much less than that of NVP by almost 5 fold as determined by densitometry (data not shown). The amount of binding did not increase significantly beyond 15 min (Figure 2-3C). 45

70 46

71 Figure 2-3. (A) Comparison of covalent binding of 12-OH-NVP (lanes 2, 5) and DNVP (lane 3, 6) with that of NVP (lane 4, 7) after a 30 or 60 min incubation with male BN rat microsomes (1 mg/ml protein) at a drug concentration of 1 mm. For comparison, covalent binding to hepatic proteins is shown after 8 days of treatment of female rats with 12-OH-NVP (159 mg/kg/day, lane 9) or NVP (150 mg/kg/day, lane 10). Protein loading was 15 µg for lanes 1-7 and 20 µg for lanes (B) Comparison of covalent binding of 12-OH-NVP (lanes 2, 5) and DNVP (lanes 3, 6) with that of NVP (lanes 4, 7) at a concentration of 1 mm after a 30 or 60 min incubation with microsomes (1 mg/ml protein) from male C57BL/6 mice. For comparison, covalent binding to hepatic proteins is shown after 6 weeks of treatment of C57BL/6 mice with NVP at a dose of 950 mg/kg/day in food. Protein loading was 13 µg for lanes 1-7 and 20 µg for lanes 8-9. (C) Comparison of covalent binding of NVP to hepatic microsomes from male C57BL/6 mice (lanes 2-4) or male BN rats (lanes 6-8) after a 15, 30, or 60 min incubation at a drug concentration of 1 mm and microsome concentration of 1 mg/ml protein. Protein loading was 20 µg per lane. The primary antiserum dilution was 1:500 and that of the secondary antisera was 1: Covalent Binding of NVP to Expressed Rat CYP2C11 or CYP3A1 Supersomes, or of NVP, DNVP, or 12-OH-NVP to Human Hepatic Expressed CYP3A4 Supersomes. Incubation of NVP with expressed rat CYP2C11 (Figure 2-4A) or CYP3A1 Supersomes (Figure 2-4B), the dominant forms of P450 in male rats led to covalent binding with major bands produced at ~50 kda and ~52 kda, respectively. In the absence of NVP as indicated in the figures, there is a small artifact band. Binding to 2C11 and 3A4 was strongest at 30 min; a decrease in the intensity of the P450 band was observed from 30 to 120 min. The incubation of expressed human 3A4 displayed the greatest binding with NVP (Figure 2-4C) versus 12-OH-NVP or DNVP. However, 12-OH-NVP did bind to human CYP3A4 more than expected, although less than NVP and there was much less binding of DNVP. The NVPmodified band had a mass of ~57 kda, which is the mass of CYP3A

72 Figure 2-4. Covalent binding of NVP to expressed male rat CYP2C11 (A) or CYP3A1 (B) in vitro. Protein concentration for each incubation was 0.8 mg/ml with 0.5 mm of drug. For immunoblots, protein loading was 9 µg and 7.5 µg per lane for A and B, respectively. (+) indicates incubations containing NVP while ( ) indicates incubations lacking NVP. Proteins were resolved on 12% gels with 1:100 dilution of primary anti-serum followed by 1:2000 dilution of secondary antisera. Comparison of covalent binding of 12-OH-NVP (lanes 2, 5), or DNVP (lanes 3, 6) with that of NVP (lanes 4, 7) to human CYP3A4 with a drug concentration of 1 mm and protein concentration in each incubation of 1 mg/ml (C). Proteins (10 µg/lane) were resolved on an 8% gel. Dilutions of antisera were 1:500 for the primary anti-serum and 1:5000 for the secondary antiserum. 48

73 2.4.4 Covalent binding of NVP or 12-OH-NVP to Hepatic Proteins from Female BN Rats Treated with NVP or 12-OH-NVP. Female BN rats were treated with NVP or 12-OH-NVP for a period of 8 days at doses of 150 mg/kg/day or 159 mg/kg/day, respectively (Figure 2-5A). The pattern of covalent binding was different for NVP and 12-OH-NVP; this difference was most prominent for the lower molecular mass proteins (30 60 kda). NVP-treated female BN rats exhibited greater covalent binding than 12-OH-NVP-treated rats at an equimolar dose, but there was a prominent artifact band in the 12-OH-NVP blot at about 60 kda. Preincubation of the anti-nvp serum with NVP blocked almost all of the binding (Figure 2-5B). 49

74 Figure 2-5. (A) Covalent binding to hepatic proteins from female BN rats fed NVP (150 mg/kg) or 12-OH-NVP (159 mg/kg) for 8 days. Protein loading was 12 µg per lane. Samples were resolved on an 8% gel. A 1:500 dilution of primary anti-serum was followed by 1:5000 dilution of secondary antisera. (B) Incubation of the anti-nvp serum with 2 mm NVP for 2 h at 37 C blocked most of the binding (left side of panel) to samples from livers of 12-OH or NVP treated rats. Samples for both panels A and B were prepared, run, blocked, incubated with secondary antibody, and imaged at the same time and protein loading was 10 µg/well of protein per lane. 50

75 2.4.5 Immunohistochemistry of Liver from NVP- or DNVP-Treated or NVP + ABT Co-treated Female BN Rats. Hepatic covalent binding of NVP and DNVP was greatest in the centrilobular area (Figure 2-6). The pattern of binding was dramatically different in rats treated with a combination of NVP and the P450 inhibitor aminobenzotriazole; specifically, treatment with ABT blocked binding in the centrilobular area and shifted it to the periportal area. Co-treatment with ABT also changed the pattern of binding by western blot although there was still significant binding (data not shown). Clearance of NVP depends on oxidative metabolism and so even if P450 is inhibited, it causes an increase in blood levels, but ultimately NVP is oxidized. Figure 2-6. Immunohistochemistry of liver sections from female BN rats; blank control, NVP treatment (150 mg/kg/day x 7 days in food), DNVP treatment (150 mg/kg/day x 7 days in food), ABT treatment (50 mg/kg/day x 28 days by gavage), or NVP (150 mg/kg/day) + ABT (50 mg/kg/day) x 28 days by gavage. Slides were incubated with 1:100 dilution of primary antisera and 1:2000 dilution of the secondary antisera. The slides were counterstained with Mayer s hematoxylin, magnification 20x. 51

76 2.4.6 Oxidation of NVP or 12-OH-NVP by Rat Liver Microsomes. The carboxylic acid (4-COOH-NVP) of NVP was detected in the incubation of 12-OH- NVP with NADPH and hepatic microsomes from both male and female BN rats (Figure 2-7). No aldehyde intermediate was detected in these reactions. Figure COOH-NVP concentrations from incubations of 12-OH-NVP with microsomes from male (n = 3) and female (n = 1) BN rats Covalent Binding, Serum ALT levels, INF-γ, and IL-6 Levels in Mice. There was no change in plasma ALT in BN rats treated with NVP (data not shown). Various strains of mice were treated with NVP to determine if it causes liver damage, covalent binding, and/or histological changes. Male BALB/c mice treated with NVP had no increase in ALT (data not shown) while there was an increase in ALT in male C57BL/6 mice at 3 weeks followed by normalization of ALT levels (Figure 2-8A). Immunoblots revealed no significant differences between the pattern and degree of binding in these two strains (Figure 2-8B). ALT levels in both male and female Cbl-b -/- mice increased at week 2, with a somewhat greater increase in male mice (Figure 2-9A) than female mice (Figure 2-9C). The animals with the largest ALT increase displayed areas of gross hepatic necrosis evident as areas of white on the surface of the liver upon sacrifice at 2 weeks. Immunoblot analysis showed the presence of a wide range of 52

77 modified hepatic protein in both male (Figure 2-9B) and female (Figure 2-9D) mice. Animals with gross necrosis appeared to have slightly more binding of NVP to lower molecular mass proteins (Figure 2-9B, D). Luminex analysis for a broad range of cytokines performed on serum of mice from the 2 week study on days 1, 7, and 14 of NVP treatment revealed an increase in interferon-gamma (IFN-γ) in plasma samples of male mice on day 7 (Figure 2-10B), both in animals that developed significant necrosis and those that did not, but the level was highest in an animal that did develop necrosis. IL-6 was also increased at day 7 versus day 14 of NVP treatment in plasma of male mice (Figure 2-10A). By day 14 of NVP treatment, the cytokine levels had decreased to or close to baseline (data not shown) for the majority of animals. Changes in cytokines were less clear for serum samples from female Cbl-b -/- mice and no inferences could be made (data not shown). No significant changes in cytokines were observed for GM-CSF, IL-10, 1L-12(p70), IL-13, IL-17, 1L-1β, IL-2, IL-4, IL-5, IL-7, IL-9, MCP-1, or TNFα. 53

78 Figure 2-8. (A) Changes in ALT in male C57BL/6 mice treated with NVP (950 mg/kg/day) for 4 weeks. Values are based on the mean of triplicate readings per time point per animal ± S.D, n=5 treated mice or n = 4 control mice. Unpaired t-test, 7 d.f., p<0.05. (B) Corresponding covalent binding of NVP at the same dose in male BALB/c (n=2) or C57BL/6 (n=3) mouse livers after 6 weeks of treatment. Protein loading was 20 µg per lane. Samples were resolved on an 8% gel. 54

79 55

80 Figure 2-9. (A) Plasma ALT levels in male Cbl-b -/- mice fed NVP orally for 14 days (950 mg/kg/day). Values are based on the mean of triplicate readings per time point per animal ± S.D, n=5 treated mice or n = 4 control mice. Unpaired t-test, 7 d.f., p<0.05. (B) Covalent binding of NVP in the livers of the same Cbl-b -/- mice. (C) Plasma ALT levels in NVP-treated (950 mg/kg/day) female Cbl-b -/- mice, n=4 treated or n=4 control mice. Values are based on the mean of triplicate readings per time point per animal ± S.D, n=5 treated mice or n = 4 control mice. Unpaired t-test, 6 d.f., p<0.05. (D) Covalent binding of NVP in the livers of the same mice. Protein loading was 25 µg per lane. Samples were resolved on 10-20% gradient gels. A 1:500 dilution of primary antisera followed by 1:5000 dilution of secondary antisera was used. 56

81 Figure Serum IL-6 (A) or IFN- (B) from control and NVP-treated Cbl-b -/- mice at day 7 of NVP treatment. Animals showing gross necrosis are displayed separately. 57

82 2.4.8 Liver Histology and ALT in Male Cbl-b -/- or C57BL/6 Mice Treated with NVP. Liver histology of Cbl-b -/- mice sacrificed after 2 weeks of NVP treatment at the time of maximal ALT elevation is shown in Figure The presence of gross necrosis, which was visible on the surface of the liver as white areas was observed in 4 of 7 treated animals. Three NVP-treated males with gross liver necrosis had ALT values 200 U/L. Two of 8 NVP-treated female mice with minor liver necrosis had ALT values of 286 and 80 U/L. Histology in female mice did not demonstrate as much injury as in males (data not shown). Histology of the livers of affected males showed the presence of focal subcapsular areas of massive liver necrosis (Figure 2-11B, 2-11C) sharply demarcated from the adjacent viable liver. Necrotic areas were surrounded by and infiltrated by mononuclear cells, macrophages, and neutrophils. This pattern of liver necrosis suggests an ischemic injury, but no evidence of thrombi or vasculitis was observed. Multifocal necro-inflammatory hepatitis with neutrophil-rich inflammatory response was observed in the absence of gross necrotic lesions in male Cbl-b -/- mouse livers. Lower doses of NVP were also tested with Cbl-b -/- mice, but no injury was seen (data not shown). In contrast, hepatic histology of C57BL/6 mice treated with NVP and sacrificed at 4 weeks displayed hepatocyte death on the edge of the lobe in one animal, as well as small focal areas of necrosis (2-12B). Induction of smooth endoplasmic reticulum (Figure 2-12C), presumably including P450 induction, was present in the histology of all mice strains tested, but was most prominent for Cblb -/- male mice. This marked induction may have led to greater reactive metabolite formation contributing to the greater toxicity in this strain, and this appeared to be the case although the difference is subtle (Figure 2-13). 58

83 Figure H&E staining of livers from Cbl-b -/- mice treated with NVP for 2 weeks. (A) Untreated control liver with normal ALT; (B) liver from a NVP-treated mouse with gross necrosis and an ALT of 271 U/L, and (C) liver from another NVP-treated mouse with gross necrosis and ALT of 313 U/L. Areas of massive hepatocyte necrosis surrounded by viable hepatocytes are shown in (B) and (C). 59

84 Figure H&E staining of livers from male C57BL/6 mice treated with NVP for 3 weeks. (A) Untreated control liver with a normal ALT; (B) liver from a NVP-treated mouse with very mild necrosis (appearing as the thin band around the capsule) and ALT of 94 U/L, and (C) liver from another NVP-treated mouse with an ALT of 75 U/L. Changes to the liver parenchyma due to enlargement of hepatocytes in the periacinar regions and extensive expansion of the endoplasmic reticulum are also present in both (B) and (C) Comparison of Hepatic Covalent Binding of NVP between Mice and Female BN Rats. Female BN rats treated with NVP for 1, 2, 4, or 8 days were sacrificed and covalent binding was determined (Figure 2-13). In comparison with Cbl-b -/- knockout mice at 2 or 10 weeks of treatment, or male C57BL/6 mice at 2 weeks of treatment, rats had significantly greater binding from day 4 onwards. In all animals the presence of a modified P450 band at ~55 kda was prominent and represents the largest modified band in each lane. While modified proteins in rats range from 20 to 100 kda, it appeared that lower molecular weight proteins were modified in 60

85 mice (up to 70 kda). Treatment of Cbl-b -/- mice with NVP for 2 weeks led to greater binding than at 10 weeks, and C57BL/6 mice displayed the least binding of the species tested. Figure Comparison of covalent binding of NVP to hepatic proteins in mice and rats. NVP was fed to rats in a time course manner from 1 to 8 days at 150 mg/kg orally in food. Mice were given 950 mg/kg/day for 2 weeks or 10 weeks. C57BL/6 males given NVP for 2 weeks are represented by C57.1 and C57.2. Each lane was loaded with 20 µg of protein. Samples were resolved on a 4-20% gradient gel. A 1:500 dilution of primary antisera followed by 1:5000 dilution of secondary was used. 61

86 2.5 Discussion An anti-nvp antiserum was produced and used to demonstrate that NVP covalently binds to hepatic proteins, both in vitro and in vivo. Binding occurred directly to P450 as demonstrated by covalent binding to expressed P450s, both rat and human. We have shown that the skin rash requires oxidation of NVP to 12-OH-NVP, 68 and most recently we have shown that covalent binding of the benzylic sulfate of this metabolite formed by sulfotransferase in the skin that is responsible for the rash. 89 In contrast, the majority of binding in the liver must involve direct oxidation by P450 as evidenced by the marked shift in the pattern of binding from the centrilobular region to the portal region caused by the P450 inhibitor ABT as shown in Figure 2-6. There is also less covalent binding of 12-OH-NVP than NVP in the liver. Furthermore, substitution of the methyl hydrogens with deuterium (DNVP) led to a marked decrease in covalent binding. Given that oxidation of the methyl group is involved in the covalent binding, but it does not involve 12-OH-NVP, these data provide strong evidence that the chemical species responsible for the covalent binding in the liver is a quinone methide formed by the loss of a hydrogen atom from the P450-generated free radical (Figure 2-1). Others have found evidence for an epoxide reactive metabolite, 79 but these data suggest that it is less important with respect to covalent binding than the quinone methide. Some covalent binding of 12-OH-NVP was detected in the in vitro experiments where phase II pathways such as sulfation would not occur, and the pattern of binding was somewhat different than that of NVP. This suggests that oxidation of 12-OH-NVP can lead to a reactive metabolite, although the binding is less than for NVP. This could be due to oxidation of the benzylic alcohol to an aldehyde or oxidation of some other part of the molecule. Oxidation of 12- OH-NVP by rat hepatic microsomes led to the carboxylic acid (Figure 2-1), but the intermediate aldehyde was not observed (Figure 2-7). This suggests that 12-OH-NVP is oxidized all the way to the carboxylic acid by P450 without release of the intermediate aldehyde; there is precedent for this. 68 It is conceivable that some of the aldehyde could become covalently bound to P450 and be responsible for the observed covalent binding; however, the pattern of binding was broader than that of NVP; specifically, most of the binding was to proteins with masses different from P450. Therefore, the aldehyde seems unlikely to be responsible for a significant amount of the covalent binding of 12-OH-NVP. To re-emphasize, the data strongly implicate the quinone methide as being the major species responsible for covalent binding of NVP in the liver. 62

87 Although we had previously observed some strange inclusion bodies in the livers of rats treated with NVP, we did not observe an increase in ALT even though there was a significant degree of covalent binding. This suggests that covalent binding may be necessary but not sufficient to produce liver injury; this is consistent with an immune mechanism. We attempted to develop an animal model of NVP-induced liver toxicity in mice. Mice metabolize NVP much faster than rats, and even higher doses did not produce easily detectable blood levels of NVP or outward signs of toxicity. However, even with higher doses and more rapid metabolism, the amount of covalent binding in mice was less than in BN rats. Treatment of C57BL/6 mice with NVP led to a small increase in ALT in some animals that resolved despite continued treatment. This is the pattern of adaptation frequently observed in humans treated with a drug that can cause more severe idiosyncratic liver injury. Liver histology in these mice revealed moderate inflammatory nodules and areas of mild focal necrosis (Figure 2-12B, C). Although the covalent binding in BALB/c mice was similar to that in C57BL/6 mice, no increase in ALT was observed in BALB/c mice. We then treated Cbl-b knockout mice with NVP. Cbl-b -/- mice lack E3 ubiquitin ligase, which leads to impaired immune tolerance; however, the animals are phenotypically normal. This deficiency could also lead to increased covalent binding if ubiquitin ligase is required for clearance of modified proteins and this appeared to be the case (Figure 2-13). We found that there was a much greater increase in ALT in some of the Cbl-b -/- mice than in the C57BL/6 mice, but the ALT also returned to normal despite continued treatment with the drug. Histology performed at the time of peak ALT (14 days) showed areas of complete necrosis with a local inflammatory response. These appeared to represent ischemic lesions because cells close to the liver capsule were spared presumably because they could benefit from diffusion through the liver capsule. However, no vascular lesions were evident histologically. Luminex analysis of cytokines performed on serum samples from Cbl-b -/- mice sacrificed at the time of ALT peak displayed a significant increase in serum IFN-γ and IL-6 in some of the animals (Figure 2-10A, B). This increase was most prominent on day 7 rather than day 1 or 14 in the majority of mice, and it occurred before the ALT increase at day 14. An elevation in cytokines or immune factors that occurs earlier than increases in other toxicity markers (i.e. ALT) is consistent with an immune response. At the study end point of 14 days, IFN-γ in liver samples of male Cbl-b -/- mice was also elevated to ~100 pg/ml for two mice (data not shown) with gross necrosis compared with 39 pg/ml for control mice. One mouse with elevated IFN-γ in the liver 63

88 (130 pg/ml) on day 14 also had markedly elevated plasma IFN-γ (866 pg/ml) on day 7 of treatment. This cytokine is considered a pro-hepatotoxic mediator leading to inflammation and tissue injury through activation of macrophages and natural killer cells. 43 This is consistent with a clinical study performed by Keane et al. that found that incubation of NVP with T cells from a patient with NVP-induced skin rash led to the production of IFN-γ by T cells. 90 Reviews regarding the difficulties with production of animal models of idiosyncratic drug reactions are available elsewhere, but the major obstacle appears to be the development of immune tolerance. 91 This is consistent with the delayed onset of liver injury and resolution despite continued treatment observed in these mice. We suspect that the liver injury in humans is immune-mediated and that the reason that most humans and rats do not develop liver injury is that the dominant response is immune tolerance. It is known that the dominant immune response in the liver is tolerance, 43 and that it is presumably why liver transplantations are relatively easy compared to transplantation of, for example, skin. Co-treatment of Cbl-b -/- mice with polyinosinic:polycytidylic acid, imiquimod, and even γ-irradiation to deplete circulating regulatory T cells was used in an attempt to break the immune tolerance and induce sustained liver damage. All of these attempts were unsuccessful in both male and female mice (data not shown). A clear picture regarding the specific types of proteins covalently modified by hepatotoxic drugs and the outcome of liver injury does not exist. Therefore, even though mice and rats display a relatively similar pattern of NVP-induced covalent binding, other individual or species-specific factors must play a role in the development of liver injury. In support of this, a recent clinical study demonstrated that patients who carried the HLA-DRB*01 allele were at increased risk of developing NVP-induced liver toxicity (the alleles associated with the risk of skin rash were different), but there was no association with the CYP2B6 genotype, which is polymorphic and one of the P450s involved in the metabolism of NVP. 19,62 In conclusion, we have clearly demonstrated that NVP covalently binds to hepatic proteins in mice, rats, and humans. The major chemical species responsible for this covalent binding is a quinone methide metabolite. We have shown a mild delayed-onset liver injury in C57BL/6 mice that may be the basis for an animal model if a method can be found to increase the liver injury. More significant injury was observed in Cbl-b -/- mice, but the histology suggests that the mechanism may be different. 64

89 FUNDING SUPPORT. This work was supported by grants received from the Canadian Institutes of Health Research, grant numbers ACKNOWLEDGMENTS. The authors thank Boehringer-Ingelheim for supplying nevirapine. A.S. is the recipient of a University of Toronto Pharmaceutical Sciences Fellowship. J.U. is the recipient of the Canada Research Chair in Adverse Drug Reactions. The work was supported by grants from the Canadian Institutes of Health Research. Portions of this work were parts of presentations given by A.M. Sharma and J.P.Uetrecht at the Society of Toxicology International Meetings in Salt Lake City, U.T., U.S.A, 2010, and Washington, D.C., USA, ABBREVIATIONS: 1-aminobenzotriazole, ABT; Brown Norway, BN; bovine serum albumin, BSA; 12-hydroxynevirapine, 12-OH-NVP; 12-trideutero-nevirapine, DNVP; cytochrome P450, P450; drug-induced liver injury, DILI; glutathione, GSH; glyceraldehyde 3-phosphate dehydrogenase, GAPDH; human immunodeficiency virus, HIV; idiosyncratic drug reaction, IDR; interferon-gamma, IFN-γ; liquid chromatography/mass spectrometry, LC/MS; nevirapine, NVP; tris-buffered saline tween-20, TBST. 65

90 CHAPTER 3 Nevirapine Bioactivation and Covalent Binding in the Skin Creativity takes courage. - Henri Matisse This work has been published in the following journal and is reproduced with permission: Amy M. Sharma, Klaus Klarskov, and Jack Uetrecht. Nevirapine Bioactivation and Covalent Binding in the Skin. Chemical Research in Toxicology, (3), Epub 2013 February 13. Reprinted with permission. Copyright 2013 American Chemical Society. All rights reserved. In this chapter, all experiments were performed by Amy M. Sharma. 66

91 3.1 Abstract Nevirapine (NVP) treatment is associated with serious skin rashes that appear to be immunemediated. We previously developed a rat model of this skin rash that is immune-mediated and very similar to the rash in humans. Treatment of rats with the major NVP metabolite, 12-OH- NVP, also caused the rash. Most idiosyncratic drug reactions are caused by reactive metabolites; 12-OH-NVP forms a benzylic sulfate, which was detected in the blood of animals treated with NVP or 12-OH-NVP. This sulfate is presumably formed in the liver; however, the skin also has significant sulfotransferase activity. In this study, we used a serum against NVP to detect covalent binding in the skin of rats. There was a large artifact band in immunoblots of whole skin homogenates that interfered with detection of covalent binding; however, when skin was separated into dermal and epidermal fractions, covalent binding was clearly present in the epidermis, which is also the location of sulfotransferases. In contrast to rats, treatment of mice with NVP did not result in covalent binding in the skin or skin rash. Although the reaction of 12- OH-NVP sulfate with nucleophiles such as glutathione is slow, incubation of this sulfate with homogenized human and rat skin led to extensive covalent binding. Incubations of 12-OH-NVP with the soluble fraction from a 9,000xg centrifugation (S9) of rat or human skin homogenate in the presence of 3 -phosphoadenosine-5 -phosphosulfate (PAPS) produced extensive covalent binding, but no covalent binding was detected with mouse skin S9, which suggests that the reason mice do not develop a rash is that they lack the required sulfotransferase. This is the first study to report covalent binding of NVP to rat and human skin. These data provide strong evidence that covalent binding of NVP in the skin is due to 12-OH-NVP sulfate, which is likely responsible for NVP-induced skin rash. Sulfation may represent a bioactivation pathway for other drugs that cause a skin rash. 67

92 3.2 Introduction The basic mechanisms of idiosyncratic drug reactions (IDRs) are currently not well understood. Circumstantial evidence suggests that most IDRs are caused by the formation of reactive metabolites rather than the parent drug; however, without a valid animal model, this is difficult to rigorously test. One such model that has allowed us to study in the mechanism of an idiosyncratic toxicity in detail is the rat model of nevirapine (NVP)-induced skin rash. 65 NVP (Viramune TM, Scheme 3-1) is a nonnucleoside reverse transcriptase inhibitor indicated for the combination treatment of HIV-1 infections. Although effective, NVP was found to induce a high incidence of skin rash or liver toxicity, and sometimes both occur in the same patient. The incidence of skin rash is approximately 9%, most of which are mild to moderate maculopapular rashes. 59 However, 16% of NVP-induced rashes are very severe and lifethreatening, including Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN). 59 Certain risk factors for development of rash have been identified, such as female gender and higher pretreatment CD4 + T cell counts. 92 Our group has developed and characterized a novel animal model of NVP-induced skin rash in female Brown Norway (BN) rats. This model shares many characteristics of the rash that occurs in humans. For example, in both humans and the rat model, evidence suggests CD4 + T cells mediate the rash. 65 Additionally, there is a delay in onset of the rash upon primary NVP treatment, but a rapid onset with secondary rechallenge in both humans and rats. 65 Higher incidence in females, increased incidence of rash with increased dose, and range in severity of rash are all features shared by both the animal model and humans. Furthermore, lymphocytes taken from both patients and animals after NVP-induced skin rash produce interferon-. 90,93 Using the BN rat model, we were able to show that the 12-hydroxylation pathway is involved in the induction of the skin rash. This is based on the observation that substitution of the NVP methyl hydrogens with deuterium markedly decreased formation of 12-OH-NVP as well as the incidence and severity of the rash. 68 Additionally, treatment with a lower dose of 12-OH-NVP induced the same degree of skin rash as treatment with NVP itself. 68 Although we know that oxidation of NVP to 12-OH-NVP is required to induce the rash, it is not clear how it does so. 12- OH-NVP is not chemically reactive; therefore, if the rash is caused by a reactive metabolite, 12- OH-NVP would require further bioactivation. The oxidation state of 12-OH-NVP and the quinone methide, which is the major species involved in covalent binding in the liver, is the same; 68

93 therefore, the quinone methide cannot be formed by oxidation of 12-OH-NVP. 12-OH-NVP is oxidized to the corresponding carboxylic acid, which forms a glucuronide, but inhibition of this oxidation does not decrease the incidence of the rash. 68 The most likely candidate is the benzylic sulfate conjugate. Sulfate is a good leaving group and numerous sulfate metabolites are known to be reactive metabolites. 94 We detected the 12-OH-NVP sulfate in the blood of BN rats treated with NVP, which was presumably formed in the liver, and there are also sulfotransferases in the skin. 95 In the case of the 12-OH-NVP sulfate, not only is the sulfate on a benzylic position, there is also an adjacent amide hydrogen that could be lost to form the same quinone methide as formed by direct oxidation of the methyl group without the formation of a carbocation intermediate. However, the 12-OH-NVP sulfate was synthesized and found to be less reactive than expected; specifically, it reacted only very slowly with glutathione (the reaction occurred over a period of days; unpublished results). In addition, initial attempts to detect covalent binding of NVP in the skin of treated animals were unsuccessful. The present study was an extension of the previous studies to test the hypothesis that 12-OH-NVP sulfate is a plausible candidate for causing NVPinduced skin rashes. 69

94 Scheme 3-1. Proposed chemical mechanism of NVP-induced skin rash resulting from covalent binding of 12-OH-NVP sulfate in the skin. 70

95 3.3 Materials and methods Chemicals. NVP was kindly supplied by Boehringer-Ingelheim Pharmaceuticals Inc. (Ridgefield, CT). The majority of chemical reagents: 3 -phosphoadenosine 5 -phosphosulfate (PAPS), Tris, methanol, DMSO, PBS (ph 7.4), glycerol, silica gel, amido black stain, were obtained from Sigma-Aldrich (Oakville, ON) unless otherwise noted in the Methods. Ammonium persulfate was obtained from Fisher Scientific (Fair Lawn, NJ). SDS and Tween-20 were obtained from BioShop (Burlington, ON). Stock acrylamide/bis solution (29:1), non-fat blotting grade milk powder, and nitrocellulose membranes (pore size 0.2 µm) were purchased from Bio-Rad (Hercules, CA). Ultra-pure tetramethylethylenediamine and 2.5% trypsin were purchased from Invitrogen (Carlsbad, CA). Amersham ECL Plus Western Blotting Detection System was obtained from GE Healthcare (Oakville, ON). Horseradish peroxidase-conjugated goat anti-rabbit IgG (H + L chains) and monoclonal GAPDH were purchased from Sigma-Aldrich (St. Louis, Mo). Normal goat serum was obtained from Invitrogen (Grand Island, NY). The synthesis of 12-OH-NVP, 12- OH-NVP sulfate, and preparation of NVP antiserum were described previously. 68,74 Protein concentrations were determined using a BCA protein assay kit (Novagen, EMD Biosciences Inc.) Animal Care. Female BN rats ( g; between 8-10 weeks of age) were obtained from Charles River (Montreal, QC). Rats were housed in pairs in standard cages in a 12:12 h light/dark cycle with access to water and Agribrands powdered lab chow diet (Leis Pet Distribution, Inc. Wellesley, ON) ad libitum. Following a 1 week acclimatization period, rats were either maintained on control chow or started on drug-containing diet (treatment groups). Drug was mixed thoroughly with powdered lab chow if it was to be administered orally. The amount of drug administered to animals was calculated based on body weight of the rats and their daily food intake. Rats were sacrificed via CO 2 asphyxiation. Balb/c or C57BL/6 mice (6-8 weeks age) were obtained from Charles River (Montreal, Quebec). E3 ubiquitin ligase casitas-b-lineage-lymphoma (Cbl-b -/- ) knockout mice were bred in house from animals first developed by Dr. J. Penninger at the Institute of Molecular Biotechnology of the Austrian Academy of Science, Vienna, with his kind permission. Programmed cell death-1 (PD-1 -/- ) knockout mice were bred in house from animals first 71

96 developed by Dr. Tasuku Honjo at the Department of Immunology and Genomic Medicine, Kyoto, Japan, with his kind permission. Mice were kept 4 per cage. NVP was administrated in lab chow following a 1 week acclimatization period. All animal experiments were approved by the University of Toronto Animal Care Committee in accordance with guidelines of the Canadian Council on Animal Care Primary and Secondary Treatment of Animals with NVP or 12-OH- NVP. Female BN rats were treated orally with NVP (150 mg/kg/day), or an equimolar dose of 12-OH-NVP (159 mg/kg/day) mixed thoroughly in rat chow for up to 21 days. For rechallenge (secondary exposure), rats treated with NVP having a moderate to severe skin rash were removed from drug, fed a diet of rat chow for 4 weeks so that the rash resolved, and then NVP was resumed at the same dose until the animals developed a rash and systemic effects such as weight loss, which usually occurred after 7-10 days Separation of Dermis and Epidermis and Preparation of Homogenates of Skin Fractions or of Whole Rat Skin. At sacrifice, hair was removed from the rats using an electric shaver, and the skin was cleaned of remaining hair using PBS (1x phosphate buffered saline, ph 7.4) and Kimwipes. Skin from the back was then excised and placed on dry ice over aluminum foil. BN rats develop deeply pigmented skin when their fur is going through the anagen phase of the hair growth cell cycle; the epidermis and dermis from this skin was found not to separate. Only pink (telogen phase) skin was used to obtain proper separation of epidermis from dermis. Care was taken to remove all hypodermis and connective tissue using blunt-tip forceps. Sections (~200 mg) of whole skin were stretched using sharp-tipped forceps on clean petri dishes for a maximum of 2 h at 4 C. Once stretched, skins were floated individually overnight at 4 C in a solution of 0.25% trypsin. 96 On day 2 the epidermis was lifted off of the dermis, the dermis was wiped clean of any residual epidermis using Kimwipes, washed with double distilled water, and each fraction was homogenized individually in cell lysis buffer (Cell Signaling Technologies, Pickering, ON) with protease inhibitor (1X HALT Protease Inhibitor Cocktail, Pierce, Rockford, IL) in a 10:1 ratio using a Polytron 2100 homogenizer (~1.5 ml working cell lysis buffer per fraction). In order to clarify samples, dermal and epidermal fractions were centrifuged at 13,000 rpm for 2 min. The 72

97 supernatant was separated from insoluble debris and stored at -80 C. Whole rat skin tissue was prepared using the same method except dermal and epidermal layers were not separated and skin was not floated overnight in trypsin Preparation of Cytosol, S9, or Microsomes from NVP-treated Rat Epidermal or Dermal Fractions. Skin from 4 rats that had been treated with NVP for 21 days was excised from the back of the neck and separated after overnight treatment with trypsin as described above. To examine covalent binding in each subcellular fraction, the separated epidermis and dermis were used to make either cytosolic, 9,000xg supernatant (S9), or microsomal fractions. Homogenates of epidermis and dermis were produced as described above. The homogenates were then centrifuged at 9,000xg for 20 min at 4 C to obtain the S9 fraction, which was set aside for 2 of 4 rats in each group. The S9 fractions obtained from the remaining two rats per group were further centrifuged at 105,000xg for 1 h at 4 C. The supernatant from these fractions contained the cytosol while the pellet, which was homogenized in 4 volumes of glycerol-phosphate-kcl buffer (20% glycerol, 0.4% KCl, 50 nm KH 2 PO4 ph 7.4), contained the microsomes. All fractions were aliquoted and stored at -80 C until use for immunoblots Preparation of Human Skin Dermatome. Human dermatomed skin was obtained from XenoTech LLC (Lenexa, KS, USA) 9.1 h postmortem. The donor was a 56 year old Caucasian male who died from heart disease without any skin diseases or infectious diseases. Skin (10X10 cm) was excised from the abdomen and snap frozen. Skin was prepared to contain only the epidermal layers ( dermatomed skin) by XenoTech LLC. Dermatomed skin was then chopped into fine pieces, immersing in cell lysis buffer with protease inhibitor for 15 min on ice, homogenized, and centrifuged as described above. The supernatant was separated and stored at -80 C Cytosolic or S9 Fractions from Rat or Human Skin or Liver. Human liver cytosol (pool of 10, mixed gender) or S9 (pool of 50, mixed gender); IGS Sprague-Dawley rat liver cytosol (pool of 115); skin cytosol (pool of 50); skin S9 (pool of 50); or liver S9 (pool of 100); B6C3F1 mouse liver S9 (pool of 400); or CD1 mouse skin S9 (pool of 100) were all purchased from XenoTech LLC (Lenexa, KS, USA). 73

98 3.3.8 Incubation of Human or Rat Skin or Fractionated Skin with NVP, 12- OH-NVP, or 12-OH-NVP Sulfate. Stock solutions of each compound were prepared in methanol fresh at the time of use. Before incubation with skin, methanol was partially removed by nitrogen evaporation in order to limit the amount of solvent in the incubation to <0.05%. 83 Following this, an equal volume of naïve human or rat whole skin, rat dermal, or rat epidermal homogenate was added to a final concentration of either 0.8 or 1 mg/ml (see results for specific details) in each tube and vortexed. The final concentration of all compounds tested was 1 mm. Samples were incubated at 37 C in a water bath and aliquots taken at time 0 (before the start of the incubation), 30, and 60 min, and reactions were terminated by placing tubes on dry ice. Negative controls were the skin fractions incubated at 37 C without addition of any drug. All samples were stored at -80 C until use for immunoblotting experiments In Vitro Metabolism of 12-OH-NVP and NVP. For testing sulfation, NVP or 12-OH-NVP (50 mm stock solution in methanol) was added to Dulbecco s PBS with MgCl 2 and CaCl 2 (Invitrogen, Carlsbad, CA) to a final concentration of 1 mm. Total incubation volume was 400 µl. The methanol was partially removed by nitrogen evaporation in order to limit the amount in the incubation to <0.05% 83. Following a 5 min preincubation with between mg/ml of protein from skin, liver S9 or cytosol (human, rat, or mouse; XenoTech LLC, Lenexa, KS) at 37 C in a water bath, PAPS was added to a final concentration of 0.3 mm. Tubes were vortexed thoroughly and incubated for 1 h. Control samples contained all components except PAPS. Samples were frozen on dry ice to halt the reaction and stored at -80 C until used for immunoblotting experiments. For incubations examining P450-mediated bioactivation the procedure was the same except the media was 100 mm potassium phosphate buffer (ph 7.4) and a NADPH-regenerating system (Solution A final concentrations: 1.3 mm NADP, 3.3 mm glucose-6-phosphate, 3.3 mm MgCl 2 ; Solution B final concentration: 0.4 Units/mL of glucose-6-phosphate dehydrogenase) replaced PAPS. 74

99 Covalent Binding Using SDS PAGE and Immunoblotting. The supernatant from homogenized whole skin or the dermal or epidermal fractions was mixed with Pierce 5x stock reducing sample loading buffer in a 4:1 protein to buffer ratio and boiled for 5 min. SDS PAGE was performed using a Protean-3 minigel system (BioRad, Mississauga, ON). Gels were hand-cast (stacking gel, 5% bisacrylamide; resolving gel, 8% bisacrylamide) and were run at ~ 110 V. Electrophoresis running buffer (Bio-Rad) consisted of 25 mm Tris, 192 mm glycine, and 0.1% SDS, ph 8.3. Transfer to nitrocellulose membrane (pore size 0.2 µm, BioRad) was performed at 0.13 ma for 90 min at 4 ºC using the same Protean-3 minigel system (BioRad, Mississauga, ON). Tris-glycine transfer buffer (Bio-Rad) consisted of 25 mm Tris, 192 mm glycine, and 20% methanol at ph 8.5. Membranes were washed twice in Trisbuffered saline tween-20 (TBST) wash solution for 5 min. Membranes were then blocked in 5% nonfat milk blocking solution in TBST for 90 min at room temperature. Membranes were then rinsed with three changes of TBST for 5 min each and incubated with a 1:100 or 1:500 dilution of primary NVP antiserum and 10% normal goat serum in TBST overnight at 4 ºC. A 20 min wash (three changes) in TBST after overnight blocking was followed by a 90 min incubation in secondary antiserum (1:2000 or 1:5000 dilution) in TBST containing 10% goat serum. In some experiments, the primary serum was blocked by preincubation with 1 mm NVP (dissolved in DMSO) for 2 h at 37 C in a water bath as described previously. 74 The secondary antiserum was goat anti-rabbit horseradish peroxidase antisera. Membranes were washed 3 times for 20 min with TBST. All blots were incubated with enhanced chemiluminescence stain for 5 min and analyzed with a FluorChem8800 imager. To probe for the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) loading control, membranes were stripped of primary NVP antiserum using Pierce Restore Plus buffer (Pierce, Rockford, IL) for 15 to 20 min at room temperature followed by a 1 h blocking step. Membranes were then incubated in mouse monoclonal anti-gapdh antiserum (1:40 000) and processed as above except that the secondary antiserum was goat anti-mouse horseradish peroxidase conjugated (Jackson ImmunoResearch, Baltimore Pike, West Grove, PA.) diluted x in TBST. Amido black (Sigma) staining and de-staining was performed at room temperature following the prescribed protocol and membranes were air-dried and visualized immediately after. 75

100 Preparation of BN Rat Skin for Histology. Ear or skin samples were fixed in 10% formalin. The paraffin block, hematoxylin/eosin slides, or unstained sections were prepared at the Hospital for Sick Children in Toronto, ON Canada. 3.4 Results Attempts to Detect In Vivo Covalent Binding in Whole Skin. Detection of covalent binding in the skin was a challenge because when whole skin was analyzed by immunoblotting, there was a large artifact band. With much effort we were able to produce convincing evidence of binding, but we were unable to eliminate the large artifact band (Figure 3-1A). A number of methods to reduce background and remove the artifact band were tried. Preincubation of the primary antiserum with skin homogenate or strips of the artifact band obtained from a large western blot of skin from control rats failed to prevent nonspecific binding even after a 24 h incubation. We were finally able to eliminate the artifact band by separation of the epidermis from the dermis; binding is in the epidermis, and the artifact comes from the dermis. A variety of methods were used to separate the skin layers, but treatment with trypsin was found to yield the most consistent results (optimization data not shown). Therefore, when the epidermis was separated from the dermis by overnight treatment with 0.25% trypsin to yield a clean separation (Figure 3-1B), there was clear binding in the epidermis of NVP or 12-OH-NVPtreated rats. However, due to the trypsin-separation method, all loading controls (GAPDH, actin, laminin, etc) did not work, despite the fact that they work with whole skin (Figure 3-1A). Amido black staining was therefore used to show even protein loading in each lane (data not shown). 76

101 Figure 3-1. (A) Immunoblot showing covalent binding of NVP (150 mg/kg) to whole rat skin in vivo with a major artifact band in each lane indicated by the grey arrows. From left to right: primary treatment days 22, 24, 25, rechallenge (RCH, 7 days), or untreated control. Each lane was loaded with 25 µg of protein. Exposure duration in the imager was 3 min. (B) Epidermis floating above dermis from trypsin-separated skin from a rat treated for 21 days with NVP (left panel); isolated epidermal layer (right panel). 77

102 3.4.2 Covalent Binding of NVP, 12-OH-NVP, or 12-OH-NVP Sulfate to Human or Rat Skin in Vitro. In order to determine if NVP, 12-OH-NVP, or 12-OH-NVP sulfate reacted with skin proteins in vitro, each compound was incubated with skin homogenates from female BN rats or humans. Figure 3-2 demonstrates clear covalent binding to both whole skin and isolated epidermal homogenate prepared from a control rat after incubation with 12-OH-NVP sulfate. Most of the proteins that were modified in the whole rat skin homogenate had masses between 40 to 70 kda (Figure 3-2A), while modified protein from isolated epidermal homogenate varied from 20 to ~ 60 kda (Figure 3-2B). Preincubation of the NVP antiserum with 1 mm NVP for 2 h at 37 C blocked almost all of the covalent binding to the epidermis, showing that the antisera is specific for proteins that have been covalently-modified by NVP metabolites (Figure 3-2C). Incubation with 12-OH-NVP sulfate also caused covalent modification of human skin dermatome proteins (Figure 3-2D) with modified proteins ranging from kda. When compared to 12- OH-NVP in the presence of PAPS on the same blot, abundant binding was observed to proteins ranging from kda. As before, when the primary antiserum was preincubated with NVP, covalent binding of 12-OH-NVP sulfate was blocked save for the nonspecific band (Figure 3-2E), which indicates that the binding to human dermatome proteins is specific for NVP. 78

103 79

104 Figure 3-2. (A) Immunoblot showing in vitro covalent binding of 1 mm each NVP, 12-OH-NVP, or 12-OH-NVP sulfate to rat whole skin homogenate containing both dermis and epidermis after incubation for 30 or 60 min. (B) Covalent binding of 1 mm NVP, 12-OH-NVP, or 12-OH-NVP sulfate to isolated epidermal or dermal homogenates prepared from a control rat. (C) Covalent binding of 1 mm NVP, 12-OH-NVP, or 12-OH-NVP sulfate to isolated epidermal or dermal homogenate prepared from a control rat showing that preincubation of the primary antisera with 1 mm NVP for 2 h at 37 C blocked binding of the antibody. Proteins (7.5 µg/well) were loaded in immunoblots A-C. (D) Human dermatome skin incubated with 1 mm each of 12-OH-NVP, NVP, or 12-OH-NVP sulfate compared to 1 mm 12-OH-NVP +/- 0.3 mm PAPS (1 mg/ml protein). (E) Covalent binding of 1 mm NVP, 12-OH-NVP, or 12-OH-NVP sulfate to isolated human 80

105 dermatome homogenate showing that preincubation of the primary antiserum with 1 mm NVP for 2 h at 37 C blocked binding of the antibody. Protein (12 µg/well) was loaded for blots D-E Covalent Binding of NVP to Rat Skin in Vivo. Once covalent binding of 12-OH-NVP sulfate to skin and isolated epidermal fractions was established in vitro, treatment of female BN rats was undertaken in order to study covalent binding to rat skin in vivo. Treatment of female BN rats with NVP or 12-OH-NVP for 7 days led to an average grade of skin rash of 0 in controls, 0.75 for NVP-treated rats, and 2 for 12-OH- NVP-treated rats (see Supplemental Data for the criteria used to grade the rash). Covalent binding in the epidermis of these animals on day 7 was minimal (Figure 3-3A), as was binding in the dermis (data not shown). On day 21 of NVP or 12-OH-NVP treatment, the average grade of rash was 0 for controls, 3.25 for NVP-treated rats, and 3.5 for 12-OH-NVP-treated rats. Covalent binding for each of the NVP and 12-OH-NVP groups was marked in the epidermis on day 21 (Figure 3-3B), with modified proteins spanning from kda. Binding in the dermis was less apparent as only a single prominent band was present at ~55 kda (data not shown). Covalent binding was also present in the epidermis of rechallenged rats, and preincubation of the antiserum with NVP again blocked specific binding (Figure 3-3C). 81

106 82

107 Figure 3-3. Immunoblots showing epidermal covalent binding in vivo after treatment with NVP or 12-OH-NVP for either (A) 7 days or (B) 21 days. (C) Immunoblots showing that preincubation of the primary antiserum with 1.5 mm NVP for 2 h at 37 C blocked binding of the antibody (right panel) to the drug-modified proteins after treatment with NVP for 21 days or after rechallenge with NVP (RCH), left panel. Epidermal protein loading was 15 µg/well. 83

108 3.4.4 Early Histological Changes in the Skin in Response to NVP Treatment. Based on early time point experiments, it was determined that submaximal NVP covalent binding was detectable by day 10 (Figure 3-4A). We sought to determine if the immune response had a similar time course and this is shown in Figure 3-4B. 12-OH-NVP-treated rats appeared to have a more advanced stage of skin degeneration at day 7 than did NVP-treated rats; it was not until day 15 that NVP-treated rat skin displayed the same acanthosis, edema, and early lymphocyte infiltration along the dermal-epidermal junction and papillary dermis as the 12-OH- NVP-treated animals did on day 7. 84

109 Figure 3-4. (A) Immunoblot showing covalent binding in the epidermis of NVP-treated female BN rats on days 10, 15, or 21 (n = 2 animals per time point) of NVP treatment. Each lane represents an individual animal with 15 µg/well of protein loaded in each well. (B) Representative H&E stained rat skin sections comparing the early infiltration of immune cells into the dermis or epidermis of NVP- or 12-OH-NVP-treated rats. Marked acanthosis (thickening of the epidermis) combined with early lymphocyte infiltrate at the dermal-epidermal junction can be observed by day 7 of 12-OH-NVP-treated animals. By day 21 there is an increase in the cellular infiltrate with areas of detachment of the epidermis. Magnification 20x. 85

110 3.4.5 Covalent Binding of NVP to Mouse Skin in Vivo. NVP was administered to 4 different strains of mice (C57BL/6, BALB/c, programmed cell death-1 knockouts, or casitas-b-lineage-lymphoma-b knockouts) at a high dose of 950 mg/kg day in food, but all failed to develop a skin rash. The latter 2 strains were used because they have impaired immune tolerance, and it was thought they would be more likely to develop an immune response. Both genders of each strain were tested. In order to examine whether the lack of skin rash was due to lack of covalent binding in skin, the anti-nvp serum was blotted against epidermal and dermal proteins of programmed cell death-1 (PD-1 -/- ) knockouts bred on a C57 background. Despite the use of a higher dose than in rats, no binding was observed in either fraction when compared to rat epidermis (data not shown). Histological analysis in these mice displayed no obvious alterations in the skin upon NVP treatment (Figure 3-5). Figure 3-5. Skin histology of PD-1 -/- knockout mice. No immune infiltrate or acanthosis was observed as was seen with NVP-treated rats. 86

111 3.4.6 Covalent Binding of NVP to Subcellular Rat Skin Fractions in Vivo. BN rat skin cytosolic and S9 fractions displayed marked binding in the fractions isolated from the epidermis (Figure 3-6A). When rat skin S9 was compared to the skin microsomal fraction, covalent binding was observed only in the skin S9 fraction from the epidermis (Figure 3-6B). Therefore, stronger binding was present only in the fractions containing sulfotransferases (SULTs), i.e. S9 and cytosol from the epidermis. 87

112 Figure 3-6. (A) Comparison of covalent binding to S9 with that to the cytosolic fractions from the skin of NVP-treated rats, isolated from either the dermis or epidermis. (B) Comparison of covalent binding to S9 with that to microsomal fractions from the skin of NVP-treated rats, isolated from either the dermis or epidermis. All animals were treated for 21 days. Protein loading was 10 µg/well Covalent Binding of 12-OH-NVP to Human or Rat Liver or Skin Proteins in the Presence or Absence of PAPS. The covalent binding of 12-OH-NVP to rat skin S9 was compared with that to human liver cytosol (from a female and a male) in the presence and absence of PAPS (Figure 3-7A). Binding of 12-OH-NVP was only observed in the presence of PAPS, which indicates that it is the sulfate that is responsible for the binding in both. The immunoblot of rat skin S9 or cytosol compared to the immunoblot of human and rat liver cytosol gave the same result (Figure 3-7B). Binding of 12- OH-NVP in the presence of PAPS was also observed to proteins from incubations with human liver S9, human dermatome skin homogenates, and rat skin S9 (Figure 3-7C). 88

113 89

114 Figure 3-7. (A) Immunoblot of rat skin S9 or human female (second right lane) or human male (most right lane) liver cytosol after incubation with 12-OH-NVP in the presence and absence of PAPS. (B) Immunoblot of rat skin S9 or cytosol, or female human or rat liver cytosol after incubation with 12-OH NVP in the presence of absence of PAPS. (C) Covalent binding of 12- OH-NVP to human liver S9, human dermatome skin, or rat skin S9 in the presence and absence of PAPS. Protein loading was 12 µg/well Covalent Binding of 12-OH-NVP or NVP to Human or Rat Liver or Skin Subcellular Fractions in the Presence or Absence of PAPS or NADPH. In order to examine the ability of skin to perform oxidation or sulfation of 12-OH-NVP or NVP, various subcellular fractions of rat or human skin or liver were tested in the presence or absence of PAPS or an NADPH-regenerating system (NRS). Covalent binding of 12-OH-NVP was observed when incubated with rat liver or skin S9 in the presence of PAPS, but not in the presence of NRS, as expected (Figure 3-8A). Conversely, only hepatic incubations of NVP with NRS produced covalent binding (Figure 3-8B). The lack of covalent binding in skin subfractions after incubation with NVP suggested that there is minimal P450 oxidase activity because oxidation and covalent binding of NVP in the liver is extensive. 74 Figure 3-8C and 3-8D produced 90

115 a similar pattern when comparing human liver S9 to rat liver S9. Covalent binding of NVP or 12- OH-NVP to human skin in the presence or absence of PAPS or NRS displayed covalent binding only with 12-OH-NVP in the presence of PAPS (Figure 3-8E), again consistent with 12-OH-NVP sulfate being the major metabolite responsible for covalent binding in the skin. 91

116 92

117 93

118 Figure 3-8. (A) Immunoblot comparing covalent binding of 12-OH-NVP to rat liver S9 versus rat skin S9 in the presence or absence of either PAPS or a NADPH-regenerating system (NRS). (B) Comparison of covalent binding of NVP to rat liver S9 versus rat skin S9 incubated in the presence of absence of either PAPS or a NADPH-regenerating system (NRS). (C) Comparison of covalent binding of 12-OH-NVP to rat liver S9 versus human liver S9 incubated in the presence or absence of either PAPS or NRS. (D) Comparison of covalent binding of NVP to human liver S9 versus rat liver S9 incubated with or without either PAPS or NRS. (E) Covalent binding of NVP or 12-OH-NVP to human skin in the presence or absence of PAPS or NRS. Protein loading was 12 µg/well. 1 mm of 12-OH-NVP or NVP was used for each incubation Sulfation and Oxidation of NVP and 12-OH-NVP in Mouse and Rat Skin. Mice do not develop a skin rash in response to NVP treatment and also have no observable levels of covalent binding in skin. We sought to determine if this was due to differences in the ability of mouse skin to form 12-OH-NVP sulfate or some other difference in metabolism or distribution of 12-OH-NVP. As a positive control, rat liver S9 and mouse liver S9 incubated with NVP in the presence of NRS displayed covalent binding, indicating formation of the quinone methide species as we had previously shown (Figure 3-9A). 74 No binding of NVP was observed to liver S9 in the presence of PAPS, or of 12-OH-NVP to liver S9 in the presence of NRS (Figure 3-9A, B). NVP incubated with skin S9 in the presence or absence of PAPS or NRS produced no detectable binding in rat or mouse samples, indicating a minimal role for oxidation by P450 in the 94

119 skin (Figure 3-9C). Incubation of 12-OH-NVP and PAPS with rat skin S9 resulted in significant covalent binding; in contrast, although there is a very small band in the immunoblot of mouse skin, minimal binding to mouse skin S9 was observed (Figure 3-9D). A pool of skins from 100 CD1 mice were used for Figures 3-9C & 3-9D because CD1 mice are outbred and more representative of the general population of all mice strains, preferable in toxicology studies. 95

120 96

121 Figure 3-9. (A) Immunoblot comparing the covalent binding of NVP to mouse vs. rat liver S9 in the presence of absence of either PAPS or an NADPH-generating system (NRS). (B) Comparison of the covalent binding of 12-OH-NVP to either mouse or rat liver S9 in the presence or absence of either PAPS or NRS. (C) Comparison of covalent binding of NVP to mouse vs rat skin S9 either in the presence or absence of PAPS or NRS. (D) Comparison of covalent binding of 12- OH-NVP to mouse vs. rat skin S9 either in the presence or absence of PAPS or NRS. Protein loading was 12 µg/well. 1 mm of 12-OH-NVP or NVP was present in each incubation. 97

122 Anti-NVP and Autoantibodies in NVP-Treated Rats. Previously our group had found that topical administration of NVP to NVP-sensitized animals led to the development of a generalized rash as opposed to rash only where NVP was applied. Absorption through the skin leading to significant circulating concentrations of NVP may have occurred. However, if the amount of NVP applied to the skin was decreased, eventually no rash was induced, but there was no dose at which only a local response was found. This suggests that there exists an autoimmune component to the rash so that activation of cells leads to a systemic reaction, not just an antidrug reaction confined to drug-exposed cells. The sera from rats rechallenged with NVP were found to contain autoantibodies in addition to anti-drug antibodies (Figure 3-10). Comparison of binding of sera from NVP rechallenged animals and naïve animals to hepatic proteins from naïve and NVP-treated animals demonstrates an artifact band at about 51 kda, an antidrug antibody binding to a protein at about 35 kda and probably 90 kda, and an autoantibody binding to a protein at about 42 kda (Figure 3-10A). An analogous experiment using epidermal proteins showed an artifact band at about 45 kda and antidrug antibodies that bind to proteins at about 40 kda and 52 kda but no antibodies to dermal proteins (Figure 3-10B). 98

123 Figure Detection of anti-nvp and autoantibodies in the serum of a rat after rechallenge with NVP (A) Liver homogenate (10 µg/lane) from an untreated (control) rat, NVP-treated rats, or 12-OH NVP-treated rats run on SDS PAGE and stained with serum (diluted 1:500) from a rat that had been rechallenged with NVP after earlier development of a NVP-induced rash (left panel) or with serum from an untreated control rat (right panel). A 1:4000 dilution of goat anti-rat HRP linked antibody was used as the secondary antibody to visualize the binding. Blots were imaged for 3 minutes on medium exposure. (B) Using serum from the same rechallenged rat, an analogous experiment was performed using fractionated skin protein (20 µg/lane) for the epidermis, designated E, or dermis, marked D ) from untreated (control) or NVP-treated rats. Blots were run, blocked, incubated with secondary, and imaged together. 99

124 Scheme 3-2. Proposed bioactivation pathway of NVP leading to immune-mediated skin rash. 100

125 3.5 Discussion Most idiosyncratic drug reactions appear to involve reactive metabolites that covalently bind to proteins. 16 Consistent with previous findings, treatment of rats with 12-OH-NVP was found to induce earlier covalent binding and histological changes, as well as a more intense rash, than NVP itself. This again indicates that the 12-hydroxylation pathway is involved in the induction of the rash. The most likely candidate for a reactive metabolite of 12-OH-NVP is the benzylic sulfate. However, its chemical reactivity was less than we expected; specifically, its reaction with glutathione was extremely slow, even under alkaline conditions. It is also less polar than would be expected for a sulfate, and under some conditions its retention time on a reverse phase HPLC column is actually longer than that of NVP (data not shown). This may be because of internal hydrogen bonding of the sulfate to the adjacent amide hydrogen. In the present study, we investigated the ability of NVP and its metabolites to covalently bind to proteins in the skin. In contrast to its reactivity with glutathione, incubation of 12-OH- NVP sulfate with skin homogenates led to extensive covalent binding. This presumably involves some neighboring group effect of the proteins involved in the binding such as removal of the adjacent amide hydrogen or neutralization of the negative charge to facilitate attack of an anionic nucleophile. Incubation of 12-OH-NVP with the skin homogenates from rats or humans also led to covalent binding, but only in the presence of PAPS, which indicates that it is the sulfate that is responsible for the binding. Covalent binding of the 12-OH-NVP sulfate involved a wider range of epidermal proteins in vitro than was observed in vivo (Figure 3-2A-E). This suggests that 12- OH-NVP sulfate covalently binds to proteins close to where it is formed even though 12-OH- NVP sulfate is present in significant concentrations in the blood of rats treated with NVP. In contrast to rat and human skin homogenates, there was minimal binding of 12-OH-NVP in the presence of PAPS to skin homogenates from mice, and this is presumably why mice do not develop a rash when treated with NVP. Covalent binding was also observed in the skin of rats treated with NVP, but only in the epidermal layer, which is the location of sulfotransferases. All of these data are consistent with 12-OH-NVP sulfate being responsible for the skin rash associated with NVP treatment. Based on covalent binding results, the epidermis appears to be the key component of skin affected by NVP treatment. The epidermis is primarily composed of keratinocytes (~95%), key in the production of local immune responses. This is not surprising given that, despite their limited 101

126 metabolic capacity, keratinocytes express a range of enzymes such as SULTs and numerous transporters, allowing for the uptake of drugs. 70,95 In addition, keratinocytes are known to be able to act as non-professional antigen presenting cells. 50 Given that the covalent binding is in the epidermis, it is somewhat surprising that the cellular infiltrate is in the dermis. However, the epidermis in rodents is only about 2-3 cells thick, and the majority of the infiltrate in rats is found at the dermal-epidermal junction (Figure 3-4B). The histology of the NVP-induced rash in humans does affect the epidermis, which is much thicker than the rat epidermis. Havlir et al. 67 examined skin biopsies of three patients who developed rash following NVP treatment and found a perivascular lymphocytic infiltrate in the papillary dermis in two patients, and a milder, nonspecific infiltrate in the third. This may simply be because it is difficult for leukocytes to migrate very far from the vasculature, and the rodent epidermis does not include blood vessels. One of the patients who developed a perivascular infiltrate also displayed endothelial swelling, which was also observed on both the back of the neck and the ears of NVP-treated rats (Figure 3-4B; 3S-1C). In the studies performed to establish the progression of covalent binding and induction of the immune response, the epidermis again appeared to play a key role. Covalent binding was not readily detected at early time points, was relatively low at 10 days, but it was clearly present after 15 days of treatment (Figure 3-4A). This presumably reflects the half-life of the modified proteins. The delay in the onset of an idiosyncratic drug reaction is usually explained on the basis of the time required to develop an adaptive immune response, but in the case of NVP-induced skin rash, the slow accumulation of modified proteins in the skin may also play a role. On day 7 of the 12-OH-NVP- and day 15 of NVP-treated groups, the observed pathology involved acanthosis (thickening of stratum corneum, which is a nonspecific result of epidermal inflammation/irritation or injury), dermal-epidermal edema, and early presence of lymphocytes along the dermal-epidermal interface (Figure 3-4B). The clear spaces observed between epidermal cells in the basal layer is termed intercellular edema or spongiosis, which is also nonspecific but indicates that fluid from dermal edema has leaked into the epidermis. This is seen in many forms of dermatitis, but especially if the skin is actively inflammed (e.g. bacterial dermatitis, severe acute allergies, etc). Therefore, the early changes in skin appear to be due to inflammation, primarily affecting the basal epidermis. 102

127 Based on current evidence, a proposed scheme for NVP bioactivation leading to skin rash is presented in (Scheme 2). Oxidation of NVP in the liver leads to 12-OH-NVP, which is carried to the skin in the general circulation. Upon arrival in the skin, 12-OH-NVP metabolism by sulfotransferase enzymes leads to production of 12-OH-NVP sulfate in the epidermis, and subsequent covalent binding leads to initiation of an active immune response. Acanthosis, vacuolization, and edema are ultimately followed by a full-blown immune response. Anti-NVP antibodies, detected in the sera of rechallenged rats (Figure 3-10) may also play a role in the pathogenesis. The presence of antibodies against hepatic proteins may also be an indication of a more general immune response. This is consistent with the observation that upon rechallenge, rats present with a more systemic sickness and lethargy, which is also observed in other more generalized immune responses such as drug-induced autoimmunity. 97 The female BN rat model is a unique tool for in depth studies of one IDR. In this study we demonstrated that the skin is capable of the bioactivation of a drug that causes a high incidence of skin rash. Studies are currently underway to test which sulfotransferase isoforms are involved in forming 12-OH-NVP sulfate and the specific role the sulfate conjugate has in the induction of a skin rash. 103

128 FUNDING SUPPORT. This work was supported by grants received from the Canadian Institutes of Health Research (MPO84520). ACKNOWLEDGMENTS. The authors thank Boehringer-Ingelheim for supplying nevirapine. The authors also thank Dr. Jeff Caswell, D.V.M., D.V.Sc., Ph.D., from the University of Guelph, for his review of the skin pathology. A.S. is the recipient of a University of Toronto Pharmaceutical Sciences Fellowship. J.U. is the recipient of the Canada Research Chair in Adverse Drug Reactions. The work was supported by grants from the Canadian Institutes of Health Research. Portions of this work were parts of presentations given by A.M. Sharma and J.P.Uetrecht at the ISSX meeting in Japan, 2011, and the Society of Toxicology International Meetings in Washington, D.C., USA, 2011, and San Francisco, CA., USA, ABBREVIATIONS: Brown Norway, BN; 12-hydroxynevirapine, 12-OH-NVP; glyceraldehyde 3- phosphate dehydrogenase, GAPDH; human immunodeficiency virus, HIV; idiosyncratic drug reaction, IDR; NADPH-regenerating system, NRS. nevirapine, NVP; tris-buffered saline tween- 20, TBST; sulfotransferase, SULT; 3 -phosphoadenosine 5 -phosphosulfate, PAPS; keratinocyte, KC; soluble fraction from a 9,000xg centrifugation, S9; Steven s-johnson syndrome, SJS; toxic epidermal necrolysis, TEN. 104

129 3.6 Supplemental Material Separation of Dermis and Epidermis of the Ear and Preparation of Homogenates. At sacrifice, the entire ear was excised from the base of the skull and placed in PBS. Each ear was sectioned into 3 parts (base of the ear, right quadrant, and left quadrant; Figure 5A) in order to obtain full separation of dermis from epidermis. Pieces from each ear were floated overnight at 4 C in a solution of 40 ml of 0.625% trypsin (0.25% was found not to work well). On day 2, the epidermis was peeled and scraped off of the dermis. The dermis was washed and wiped clean of any residual epidermis, and each fraction was homogenized and centrifuged as described above. The supernatant was separated and stored at -80 C Grading of Skin Rash. The grading scheme according to the AIDS Clinical Trial Group Protocol Management Handbook Table for Grading Severity of Cutaneous Eruptions was used (where applicable to rats). Grading was performed at the time of sacrifice using an area of shaved 1 x 1 inch skin on the upper neck/back area of each rat. 92,98 (0) grade 0, normal integrity of skin is maintained (1) grade 1, erythema with or without pruritus; (2) grade 2, a diffuse erythematous macular or maculopapular cutaneous eruption or dry desquamation with or without pruritus or typical target lesions without blistering, vesicles, or ulcerations in the lesions; (3) grade 3, 1 of the following clinical presentations: urticaria; diffuse erythematous macular or maculopapular cutaneous eruption or moist desquamation with or without pruritus together with any of the 4 constitutional findings possibly related to the drug (i.e., blistering, vesiculation, or both of cutaneous eruptions; or any site of mucosal lesions considered related to study drug without other etiology, such as herpes simplex or aphthous ulcer); angioedema; exfoliative dermatitis (defined as severe widespread erythema and dry scaling of the skin and generalized superficial lymphadenopathy, with other constitutional findings possibly related to study drug such as fever or weight loss); or diffuse rash and serum sickness-like reactions defined as clinical symptom complex manifested as fever, lymphadenopathy, edema myalgia, arthralgia, or a combination; and (4) grade 4, diffuse cutaneous eruptions usually starting on the face, trunk, or back, often with prodromal symptoms plus one of the following: cutaneous bullae, sometimes confluent with widespread sheet like detachment of skin (Nikolsky s sign), Stevens-Johnson syndrome, erythema multiforme major, or toxic epidermal necrolysis, or 2 or more anatomically distinct sites of mucosal erosion or ulceration not due to another cause. Severe rash was defined as grade 3 and grade 4 cutaneous eruptions when we used this grading scheme. Skin biopsies were not required for categorization of rash 105

130 Grading Tables Table 3S-1. Day 7 Skin Rash Grading Key: UB = upper back; MB = mid-back; LS = left upper shoulder. Rat Treatment Area of Skin Description Rash Grade 1x1 Inch Day 7 Control Rat 1 UB No visible skin abnormality 0 Day 7 Control Rat 2 UB No visible skin abnormality 0 Day 7 NVP Rat 1 UB 1 1 Day 7 NVP Rat 2 UB 1 1 Day 7 NVP Rat 3 UB No visible skin abnormality 0 Day 7 NVP Rat 4 UB 1 1 Day 7 12-OH Rat 1 UB 3 SMALL 2 Day 7 12-OH Rat 2 UB 1 LARGE/DEEP 2 106

131 Table 3S-2. Day 10 and 15 Skin Rash Grading Rat ID Area of Skin Description Rash Grade 1x1 Inch Day 10 Control UB No visible skin abnormality 0 Day 10 NVP Rat 1 UB SLIGHT SCALING 1 Day 10 NVP Rat 2 UB SLIGHT RED 1 Day 15 Control UB CLEAN 0 Day 15 NVP Rat 1 UB 1 SMALL 2 Day 15 NVP Rat 2 UB 2 SMALL, SLIGHT REDNESS 2 107

132 Table 3S-3. Day 21 Skin Rash Grading Rat ID Area of Skin Description Rash Grade 1x1 Inch Day 21 Control Rat 1 UB No visible skin abnormality 0 Day 21 Control Rat 2 UB No visible skin abnormality 0 Day 21 Control Rat 3 UB No visible skin abnormality 0 Day 21 Control Rat 4 UB No visible skin abnormality 0 Day OH-NVP Rat 1 LS/UB DEEP LESIONS 3 Day OH-NVP Rat 2 UB VERY BAD; PELT LIKE; DEEP 4 Day 21 NVP Rat 1 UB/MB VERY RED, VERY DEEP 3 Day 21 NVP Rat 2 UB LESS LESIONS, VERY RED 3 Day 21 NVP Rat 3 UB/MB BLOODY/SLOUGHING 4 Day 21 NVP Rat 4 UB PEELING/LESS RED 3 108

133 H Covalent Binding and Histology in the Ears from NVP- or 12-OH-NVP- Treated Rats. The ear is the first organ to turn red in response to NVP treatment in female BN rats (day 7). 65 In severe cases, the rash may develop on the ear. Dorsal-ventral axis separation (Figure S- 1A) of ear epidermis and dermis from NVP- or 12-OH-NVP-treated rats (n = 4 per group) displayed covalent binding (Figure S-1B) and a marked cellular infiltrate in the ear (Figure S- 1C). 109

134 Supplemental Figure 3S-1. (A) Method to fractionate ear using dorsal-ventral axis separation is shown. Ear pieces were floated on 0.625% trypsin overnight at 4 C to ensure complete epidermal-dermal separation. (B) Immunoblot experiments comparing the epidermis from the neck or ear from NVP- or 12-OH-NVP-treated female BN rats; 12 µg protein/well. Lane designations are as follows: 1 & 2 = epidermis from the neck of control rats; 3 = epidermis from 110

135 the neck of a 12-OH-NVP-treated rat; 4 = epidermis from the neck of a NVP-treated animal; 5 = epidermis from the ear of a 12-OH-NVP-treated rat; 6 = epidermis from the ear of a NVP-treated rat. (C) H&E images of ear sections taken from each treatment group (representative slide from 1 of 4 rats per group shown). Magnification 20x. 111

136 CHAPTER 4 12-OH-Nevirapine Sulfate, Formed in the Skin, is Responsible for Nevirapine- Induced Skin Rash We're all mad here. - Lewis Carroll, Alice in Wonderland This work has been published in the following journal and is reproduced with permission: Amy M. Sharma, Maria Novalen, Tadatoshi Tanino, Jack P. Uetrecht. 12-OH-Nevirapine Sulfate, Formed in the Skin, is Responsible for Nevirapine-Induced Skin Rash. Chemical Research in Toxicology, , Epub 2013 April 16. Reprinted with permission. Copyright 2013 American Chemical Society. All rights reserved. In this chapter, all experiments were performed by Amy M. Sharma, except Figures 4-2, 4S-1, 4S

137 4.1 Abstract Nevirapine (NVP) treatment is associated with a significant incidence of skin rash in humans, and it also causes a similar immune-mediated skin rash in Brown Norway (BN) rats. We have shown that the sulfate of a major oxidative metabolite, 12-OH-NVP, covalently binds in the skin. The fact that the sulfate metabolite is responsible for covalent binding in the skin does not prove that it is responsible for the rash. We used various inhibitors of sulfation to test whether this reactive sulfate is responsible for the skin rash. Salicylamide (SA), which depletes 3 -phosphoadenosine- 5 -phosphosulfate (PAPS) in the liver, significantly decreased 12-OH-NVP sulfate in the blood, but it did not prevent covalent binding in the skin or the rash. Topical application of 1-phenyl-1- hexanol, a sulfotransferase inhibitor, prevented covalent binding in the skin as well as the rash, but only where it was applied. In vitro incubations of 12-OH-NVP with PAPS and cytosolic fractions from the skin of rats or from human skin also led to covalent binding that was inhibited by 1-phenyl-1-hexanol. Incubation of 12-OH-NVP with PAPS and sulfotransferase 1A1*1, a human isoform that is present in the skin, also led to covalent binding, and this binding was also inhibited by 1-phenyl-1-hexanol. We conclude that salicylamide did not deplete PAPS in the skin and was unable to prevent covalent binding or the rash, while topical 1-phenyl-1-hexanol inhibited sulfation of 12-OH-NVP in the skin and did prevent covalent binding and the rash. These results provide definitive evidence that 12-OH-NVP sulfate formed in skin is responsible for NVP-induced skin rashes. Sulfotransferase is one of the few metabolic enzymes with significant activity in the skin, and it may be responsible for bioactivation of other drugs that cause skin rashes. 113

138 4.2 Introduction Idiosyncratic drug reactions (IDRs) are unpredictable adverse events that significantly impact drug development and use. Many drugs that cause IDRs form reactive metabolites, and it is usually assumed that these reactive metabolites are responsible for the IDR associated with the drug involved. 99,100 However, IDRs are difficult to study and it has not been possible to definitively demonstrate that reactive metabolites are causal. In addition, many drugs form several reactive metabolites and it is very difficult to test which, if any, is responsible for a specific IDR. Furthermore, Pichler has proposed the p-i hypothesis in which a reversible interaction between the T cell receptor major histocompatibility complex is sufficient to trigger an IDR without the formation of reactive metabolite. 101 This is especially attractive for skin rashes because the skin has very limited drug metabolism capacity; specifically, the levels of cytochromes P450 are very low. 102 At one point early in the NVP studies we thought that the p-i hypothesis might be relevant for NVP-induced skin rash because we had eliminated several possible reactive metabolites; however, we used the NVP animal model to show that the basis for the p-i hypothesis is false, and certainly it is not the mechanism of NVP-induced skin rash. 93 That does not mean that the hypothesis itself is false; the p-i hypothesis probably is relevant for some compounds, especially small peptidomimetic drugs such as ximelagatran. 33 We have developed an animal model of nevirapine (NVP, Viramune TM, TOC graphic)-induced skin rash, which is clearly immunemediated and has characteristics very similar to the rash that occurs in humans. 65 These characteristics include a similar time to onset, higher incidence in females, and the observation that a low CD4 + T cell count decreases rash incidence. We used this model to test the involvement of a reactive metabolite in the mechanism of NVP-induced skin rash. NVP, a non-nucleoside reverse transcriptase inhibitor indicated for the treatment of HIV-1 infections, causes idiosyncratic hepatotoxicity and mild-to-severe skin rashes. 61 We have demonstrated that NVP is oxidized to a reactive quinone methide, which covalently binds in the liver. 74 However, we have not been able to demonstrate that this reactive metabolite is responsible for the idiosyncratic liver injury caused by NVP because we were not able to produce liver injury in animals with characteristics similar to the liver injury that occurs in humans. In contrast, we identified the 12-hydroxylation pathway to form 12-OH-NVP is responsible for induction of the skin rash because substitution of the methyl hydrogens of NVP with deuterium significantly decreased the incidence and severity of the rash and treatment with 12-OH-NVP also caused a 114

139 skin rash. 68 Because 12-OH-NVP is the same oxidation state as the quinone methide species responsible for covalent binding in the liver, a quinone methide species could not be produced by oxidation of 12-OH-NVP. In a recent paper we demonstrated that covalent binding of NVP in the skin is mediated by a benzylic sulfate formed by first oxidation of NVP in the liver to 12-OH- NVP followed by the formation of a benzylic sulfate (12-OH-NVP sulfate), which has sufficient chemical reactivity to covalently bind to proteins. 89 Both the liver and skin contain sulfotransferases, and although chemically reactive, we were able to detect 12-OH-NVP sulfate in the blood of rats treated with NVP. It remained to be determined if 12-OH-NVP sulfate is responsible for NVP-induced skin rash, and if so, whether it is sulfation in the liver or skin that is most important. We used our animal model and various inhibitors of sulfation to answer this question. Scheme 4-1: Depiction of schematic used to prevent rash in this study. 115

140 4.3 Materials and methods Chemical Materials and Reagents. NVP and ethyl-nvp (a NVP analogue where the cyclopropyl group has been replaced by an ethyl group) were kindly supplied by Boehringer-Ingelheim Pharmaceuticals Inc. (Ridgefield, CT). Common chemical reagents (3 -phosphoadenosine 5 -phosphosulfate (PAPS), β- glucuronidase type IX-A, Tris, methanol, DMSO, PBS (phosphate buffered saline, 1.47 mm KH 2 PO 4 and 8.06 mm Na 2 HPO 4-7H2O, ph 7.4), glycerol, silica gel, etc) were obtained from Sigma-Aldrich (Oakville, ON) unless otherwise noted in the Methods. Ammonium persulfate was obtained from Fisher Scientific (Fair Lawn, NJ). SDS and Tween-20 were obtained from BioShop (Burlington, ON). Stock 30% acrylamide/bisacrylamide solution (29:1), nonfat blotting grade milk powder, and nitrocellulose membrane (0.2 µm) were purchased from Bio-Rad (Hercules, CA). Ultra pure tetramethylethylenediamine, frozen 2.5% trypsin, and normal goat serum were purchased from Invitrogen (Carlsbad, CA). Amersham ECL Plus Western Blotting Detection System was obtained from GE Healthcare (Oakville, ON). Horseradish peroxidase-conjugated goat anti-rabbit IgG (H + L chains) was purchased from Sigma-Aldrich (St. Louis, Mo). 1-Phenyl- 1-hexanol was obtained from Tokyo Chemical Industry (Toshima, Japan) and micronized dehydroepiandrosterone (DHEA) was obtained from PCCA (London, ON). The synthesis of 12- OH-NVP, 12-OH-NVP sulfate, 4-carboxy-NVP (4-COOH-NVP), and the NVP anti-serum were described previously. 68,74 Human liver cytosol (pool of 10, mixed gender) or a 9,000 X supernatant (S9) fraction containing cytosol and microsomes (pool of 50; mixed gender); rat liver cytosol (pool of 115; female Sprague Dawley rats) or rat skin cytosol (pool of 50; female); and recombinant human sulfotransferase (SULT) 1A1*1 expressed in Escherichia coli (E. coli) were purchased from XenoTech LLC (Lexena, KS) Animal Care. Female BN rats ( g; 8 to 10 weeks of age) were age-matched and obtained from Charles River (Montreal, QC). Rats were housed in pairs in standard cages in a 12:12 h light/dark cycle with access to water and Agribrands powdered lab chow diet (Leis Pet Distribution, Inc. Wellesley, ON) ad libidum. Following a 1 week acclimatization period, rats were either maintained on control chow or started on a drug-containing diet (treatment groups). For chronic experiments, drug was mixed thoroughly with powdered lab chow to produce a NVP dose of

141 mg/kg/day, or an equimolar dose of 12-OH-NVP (159 mg/kg/day) for a maximum of 21 days. The amount of drug administered to rats was calculated based on their body weight and daily food intake. For experiments examining blood metabolite levels, drugs were ground to obtain fine particles and NVP was gavaged at a dose of 100 mg/kg/day in 0.5% methyl cellulose. The dose was scaled up from 50 mg/kg/day over a period of 3-5 days to avoid central nervous system toxicity associated with high peak plasma levels of NVP. Rats were sacrificed via CO 2 asphyxiation. Animal experiments were approved by the University of Toronto animal care committee in accordance with guidelines of the Canadian Council on Animal Care Quantification of NVP, 12-OH-NVP, 12-OH-NVP Sulfate, and 4- COOH-NVP in Plasma. Plasma (50 µl) was mixed with internal standard (ethyl-nvp, 5.4 µg/ml, 50 µl) and concentrated with a Strata solid phase extraction column (C18-E, 100 mg, Phenomenex, Torrance, CA). The column was washed with 1 ml of water and the metabolites eluted with 1 ml of methanol. The methanol was collected, dried, and reconstituted with 50 µl of the HPLC mobile phase. The samples were separated on HPLC and analyzed by mass spectrometry. The separation was carried out on an Ultracarb C18 30 X 2.0 mm, 5 µm column (Phenomenex) under isocratic conditions with a mobile phase consisting of 16% acetonitrile and 84% water with 2 mm ammonium acetate and 1% acetic acid and the flow rate of 0.2 ml/min. Mass spectrometry was carried out using a PE Sciex API 3000 quadrupole system with an electrospray ionizing source. The ion pairs used for quantitation in the multiple reaction monitoring/positive ion mode were: 267.0/226.1 for NVP, 283.1/223.1 for 12-OH-NVP, 297.1/210.1 for 4-COOH-NVP, 283.1/161.0 for 2-OH-NVP, 283.1/214.0 for 3-OH-NVP, 255.1/227.2 for ethyl-nvp. Standard curves prepared for 2-OH-NVP ( µg/ml), 3-OH-NVP ( µg/ml), 12-OH-NVP ( µg/ml), 4-COOH-NVP ( µg/ml), and NVP ( µg/ml) had R 2 values of > Quantification of 12-OH-NVP sulfate was done in a similar manner, the major difference being that it was performed in the negative ion mode. Because of this the internal standard was changed to naproxen (2.5 µg/ml, 50 µl added to the plasma). The ion pairs used for the analysis were 361.0/96.0 for 12-OH-NVP sulfate and 229.0/169.8 for naproxen. The HPLC column used for the separation was an Ultracarb C18 column (100 x 2 mm, 5 µm, Phenomenex) with a 117

142 gradient elution of 20 80% acetonitrile over a period of 10 min. The second solvent was water with 2 mm ammonium acetate and 1% acetic acid. The flow rate was 0.2 ml/min. Standard curves prepared for 12-OH-NVP sulfate ( µg/ml) had R 2 values of > Sulfation Inhibition Studies. The effects of 3 sulfation inhibitors on covalent binding and skin rash were studied; see Results for further information on specific experiments. Table 4-1. Inhibitors of sulfation and dosing method. Inhibitor Name Dose Vehicle Application Method Salicylamide 274 mg/kg/day 0.5% methyl cellulose oral; 1x/ day after 5 PM DHEA 20 mg/kg/day 50:50 oil:acetone topical or oral gavage; topical; (topical); 1x/day 50 or % methyl cellulose mg/kg/day oral (oral) 1-phenyl-1-hexanol 20 mg/kg/day topical, or 20 mg/kg/day oral 50:50 oil:acetone (topical); 0.5% methyl cellulose (oral) topical or oral gavage; 1x/day For topical inhibition studies the desired area of skin was shaved using an electric shaver, and administration of the inhibitors was started 3 days after shaving to allow the skin barrier to heal from any nicks that might have occurred. The skin was painted with either DHEA or 1- phenyl-1-hexanol using a 200 µl pipette. For oral administration inhibitors were given as a cotreatment with NVP. Controls were either fed standard lab chow or administered vehicle by gavage if they were controls for salicylamide cotreatment groups. For systemic inhibition studies, female BN rats were treated with NVP together with salicylamide, DHEA, or 1-phenyl-1-hexanol. All drugs were ground, suspended in 0.5% methylcellulose, and administered by gavage with minimal time between NVP and the inhibitor. If NVP was to be given by gavage, the dose was escalated during the first 4 to 5 days of the study from 50 mg/kg/day to the full dose of 100 mg/kg/day (see results for specific details). The lower dose was given in the beginning of the study to avoid central nervous system toxicity associated with high peak plasma levels of NVP. 12-OH-NVP was coadministered at a dose of 100 mg/kg/day by oral gavage with DHEA at doses of 25, 50, or 100 mg/kg/day by oral gavage

143 phenyl-1-hexanol was administered orally by gavage while NVP was fed in food at 150 mg/kg/day. All inhibition studies were carried out for a maximum of 28 days Separation of Skin Dermis and Epidermis and Preparation of Homogenates. At sacrifice, hair was removed from the rats using an electric shaver, and the skin was cleaned of remaining hair using PBS and Kimwipes. Skin from the back was then excised and placed on dry ice over aluminum foil. Care was taken to remove all hypodermis and connective tissue using blunt-tip forceps. Sections (~200 mg) of whole skin were stretched using sharp-tipped forceps on clean petri dishes for a maximum of 2 h at 4 C. Once stretched, skins were floated individually overnight at 4 C in a solution of 0.25% trypsin (approx. 50 ml per skin section). On day 2, the epidermis was lifted off of the dermis, the dermis was wiped clean of any residual epidermis using Kimwipes, and each fraction was homogenized separately in cell lysis buffer (Cell Signaling Technologies, Pickering, ON) with protease inhibitor (HALT Protease Inhibitor Cocktail, Pierce, Rockford, IL) in a 10:1 ratio using a Polytron 2100 homogenizer (~1.5 ml working cell lysis buffer per fraction). In order to clarify samples, dermal and epidermal fractions were centrifuged at 13,000 rpm each for 2 min. The supernatant was separated and stored at -80 C. The protein concentration of the prepared homogenates was quantified using a BCA protein assay kit (Novagen, EMD Biosciences Inc., Mississauga, ON); bovine serum albumin was used as the standard. Whole rat skin tissue was prepared using the same method except dermal and epidermal layers were not separated and skin was not floated overnight in trypsin Preparation of Human Skin Dermatome. Human skin (9 g) was obtained from XenoTech LLC (Lenexa, KS, USA) from the abdomen of a 56 year old Caucasian male 9.1 h after death due to heart disease without any skin or infectious diseases. Skin was prepared to contain the epidermal layers only. Dermatomed skin was then prepared by chopping into fine pieces, immersion in cell lysis buffer with protease inhibitor for ~15 min on ice followed by homogenization via a Polytron 2100 homogenizer. Skin was clarified of debris via centrifugation as described above. The supernatant was separated and stored at -80 C. 119

144 4.3.7 Incubation of Human Skin or Expressed Human SULT 1A1*1 with 12- OH-NVP or 12-OH-NVP Sulfate, With or Without PAPS. Stock solutions of each compound were prepared in methanol fresh on the day of use. Before incubation with skin or isolated SULT 1A1*1, methanol was partially removed by nitrogen evaporation in order to limit the amount of solvent in the incubation to <0.05%. 83 For human dermatomed skin, a 5 min preincubation consisting of each chemical and Dulbecco s PBS with MgCl 2 and CaCl 2 at 37 C in a water bath was followed by addition of dermatomed skin (final concentration 1 mg/ml). SULT 1A1*1 pre-incubates consisted of the SULT 1A1*1 protein, skin (final concentration 1 mg/ml), Dulbecco s PBS with MgCl 2 and CaCl 2, and drug. Following the 5 min preincubation, PAPS was added to a final concentration of 0.3 mm and the samples were vortexed. The final concentration of all drugs tested was 1 mm. Samples were taken at time 0 (before the start of the incubation), 30, and 60 min, and reactions were terminated by placing the samples on dry ice. Negative controls were the skin fractions incubated at 37 C without addition of any drug or PAPS (see results for specific details). All samples were stored at -80 C until use for immunoblotting experiments Incubation of Rat Liver or Skin Cytosol or Human Liver Cytosol with 12-OH-NVP and 1-Phenyl-1-hexanol in the Presence and Absence of PAPS. To an Eppendorf tube containing Dulbecco s PBS with MgCl 2 and CaCl 2 was added 12- OH-NVP dissolved in methanol (stock 50 mm) and 1-phenyl-1-hexanol to a final concentration of 1 mm each. Methanol was partially removed by nitrogen evaporation in order to limit the amount of solvent in the incubation to <0.05%. 83 Following a 5 min preincubation with between mg/ml each of rat skin S9 or male or female human liver cytosol (XenoTech LLC, Lenexa, KS) at 37 C in a water bath, PAPS was added to a final concentration of 0.3 mm. Tubes were vortexed and incubated for 1 h. Negative controls contained all components except PAPS while positive controls did not contain 1-phenyl-1-hexanol. All samples were frozen on dry ice to halt the reaction and stored at -80 C until used for immunoblotting experiments Covalent Binding Using SDS-PAGE and Immunoblotting. Whole skin, dermal, or epidermal homogenate samples, or in vitro incubates, were mixed with Pierce reducing sample loading buffer in a 4:1 protein to buffer ratio and boiled for 5 min. SDS 120

145 PAGE was performed using the Protean-3 minigel system (BioRad, Mississauga, ON). Gels were hand-cast (8% bisacrylamide) or bought from Bio-Rad Canada (12% bisacrylamide), and were run at 130 V. Electrophoresis running buffer (Bio-Rad) consisted of 25 mm Tris, 192 mm glycine, and 0.1% SDS, ph 8.3. Transfer to nitrocellulose membrane (0.2 µm, BioRad) occurred at 0.13 ma for 90 min at 4 ºC using the same Protean-3 minigel system (BioRad, Mississauga, ON). Tris-glycine transfer buffer (Bio-Rad) consisted of 25 mm Tris, 192 mm glycine, and 20% methanol at ph 8.5. Membranes were washed twice in Tris-buffered saline tween-20 (TBST) wash solution for 5 min. Membranes were then blocked in 5% non-fat milk blocking solution in TBST for 90 min at room temperature. Membranes were then rinsed with 3 changes of TBST for 5 min and incubated with a 1:100 or 1:500 dilution of primary anti-nvp antiserum and 10% normal goat serum in TBST overnight at 4 ºC. A 20 min wash (3 changes) in TBST after overnight blocking was followed by a 90 min incubation in secondary antisera (1:2000 or 1:5000 dilution) in TBST containing 10% goat serum. The secondary anti-serum was goat anti-rabbit horseradish peroxidase antisera. Membranes were washed 3 times for 20 min with TBST. All blots were incubated with enhanced chemiluminescence stain for 5 min and analyzed with a FluorChem8800 imager. To probe for the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) loading control, membranes were stripped of primary NVP antiserum using Pierce Restore Plus buffer (Pierce, Rockford, IL) for 15 to 20 min at room temperature followed by a 1 h blocking step. Membranes were then incubated in mouse monoclonal anti-gapdh antisera (1:40,000) and processed as above except the secondary antisera was goat anti-mouse horseradish peroxidase antisera diluted 1:10,000 (Jackson ImmunoResearch, Baltimore Pike, West Grove, PA.) Preparation of BN Rat Skin for Histology. Skin samples were flattened on paper towels to prevent curling of the skin and fixed in 10% formalin. The paraffin block, hematoxylin/eosin slides, or unstained sections were prepared at the Hospital for Sick Children in Toronto, ON Canada. 121

146 4.4 Results General Scheme to Study the Role of 12-OH-NVP Sulfation on the Skin Rash. The sulfation of 12-OH-NVP can occur in both the liver and skin, and it is important to know which, if either, is involved in the NVP-induced skin rash. Salicylamide depletes PAPS and was administered by oral gavage; it is likely to only decrease sulfation in the liver. DHEA and 1- phenyl-1-hexanol are sulfotransferase inhibitors and were administered both orally and topically. Previous work with 1-phenyl-1hexanol has been limited to in vitro studies, and it is likely that it is cleared too rapidly to significantly decrease hepatic sulfation Effect of Salicylamide on 12-OH-NVP Sulfate Levels and Rash. Cotreatment with oral salicylamide markedly decreased circulating 12-OH-NVP sulfate levels (Figure 4-1D). It also decreased the amount of 12-OH-NVP sulfate in a 24 hour urine sample (Figure 4S1-C, Supporting Information). However, it did not prevent covalent binding in the skin or the skin rash (Figure 4-2). The average grade of rash for the salicylamide cotreated group was 3.75 versus 3.25 for NVP alone (Table 4S-1, see Supporting Information for criteria used to grade the rash and grading results). Covalent binding of NVP in the epidermis (Figure 4-2A) and the cellular infiltrate in the skin (Figure 4-2B) were also not decreased by cotreatment with oral salicylamide. The NVP and NVP + salicylamide cotreatment groups displayed similar histological alterations in the skin, signifying no effect of salicylamide (Figure 4-2C). It appeared that animals that presented with the most severe grade of skin rash (3 or 4) displayed the most intense covalent binding from all groups tested (12-OH-NVP, NVP, or NVP + salicylamide). As reported previously, 89 essentially all of the covalent binding was in the epidermis with virtually none in the dermis (data not shown). 122

147 12-OH-NVP sulfate ( g/ml) 12-OH-NVP ( g/ml) NVP ( g/ml) A NVP + salicylamide NVP + DHEA 100 NVP + DHEA 50 NVP B Incidence skin rash C Treatment day 30 NVP NVP + DHEA 50 NVP + DHEA 100 NVP + salicylamide NVP NVP + DHEA 50 NVP + DHEA 100 NVP + salicylamide D Treatment day Treatment day NVP NVP + DHEA 50 NVP + DHEA 100 NVP + salicylamide Figure 4-1. (A) Incidence of skin rash, (B) plasma concentrations of NVP, (C) 12-OH-NVP, and (D) 12-OH-NVP sulfate in female Brown Norway rats treated with NVP only (100 mg/kg/day, n = 4), in combination with oral DHEA (50 and 100 mg/kg/day) or in combination with oral salicylamide (274 mg/kg/day). 123

148 124

149 Figure 4-2. (A) Immunoblot of the epidermis comparing individual 12-OH-NVP-treated rats to NVP + oral salicylamide cotreated rats (N+Sal) or NVP only-treated rats, against 0.5% methyl cellulose gavaged controls. Protein loading was 15 µg/lane. (B) Skin histology of NVP + oral salicylamide cotreated rats, n = 4. (C) Skin histology compared between various treatment groups: normal and gavaged controls are normal without a cellular infiltrate in the dermis, while NVP, 12- OH-NVP and NVP + oral salicylamide treated rats display keratinocyte necrosis within the epidermis, with marked inflammatory infiltrate at the dermal-epidermal junction. A representative photo from one of four animals per group is shown. All rats represented in this figure were treated for 21 days. Magnification was 20x for all slides in this figure Effects of DHEA on NVP Metabolism and Skin Rash. Coadministration of oral DHEA decreased plasma concentrations of NVP, 12-OH-NVP, and 12-OH-NVP sulfate, although it did not appear to decrease the blood levels of 12-OH-NVP sulfate as much salicylamide (Figure 4-1). It also decreased the urinary excretion of 12-OH-NVP sulfate and other metabolites (Figure 4S-1); however, oral DHEA did prevent the skin rash. Because of the complex effects of oral DHEA on NVP metabolism we tried topical administration of DHEA to inhibit sulfation in the skin. It decreased the severity of the rash, but it did not completely prevent it. This was reflected in the histology, where the dermal infiltrate and epidermal changes resembled those of the NVP only group (Figure 4S-2A-C). Even topical DHEA led to decreased plasma 12-OH-NVP sulfate levels near the end of treatment (day 21) 125

150 when compared to NVP only treated animals (data not shown). The results using oral and topical DHEA to prevent sulfation were complicated by other effects on NVP metabolism and did not provide a definitive answer; therefore, we abandoned its use Effects of 1-Phenyl-1-Hexanol on Covalent Binding and Rash. In contrast to DHEA, topical 1-phenyl-1-hexanol did not significantly affect blood levels of NVP or its metabolites (data not shown), but it completely prevented the rash where it was applied. Specifically, topical administration of 1-phenyl-1-hexanol was performed as outlined in Figure 3B using the left and right shoulders of the animals in order to maintain symmetry. Both the skin rash (Figure 4-3, C-E) and covalent binding (Figure 4-4, A-B) were prevented in the 1- phenyl-1-hexanol-treated areas in the epidermis (although there is an artifact band in the control), and these areas had much less cellular infiltrate than vehicle-treated areas (Figure 4-5, A-C). In order to obtain more skin, the upper and lower backs of the animals were employed in another scheme (Figure 4-6A) to test inhibition of skin rash and covalent binding. This schematic for application was found to be better in that rats had a hard time accessing these areas to scratch or lick. Both skin rash and covalent binding (Figure 4-6B; photos not shown) were prevented in the topical 1-phenyl-1-hexanol-treated areas, and these areas had much less cellular infiltrate in the dermis than vehicle-treated areas (Figure 4-7, A-C). Oral administration of 1-phenyl-1- hexanol did not prevent covalent binding or skin rash (data not shown). It is likely that this lack of effect is the result of rapid clearance of 1-phenyl-1-hexanol, but we did not develop an analytical method so that this hypothesis could be tested. In order to show specificity of the anti-serum for NVP-modified proteins, it was preincubated with NVP (Figure 4-7D). The preincubated anti-serum was then used as the primary antibody in an immunoblot using the same epidermal samples shown in Figure 4-3, C-E. Covalent binding to each of the epidermal skin fractions taken from vehicle, topical 1-phenyl-1-hexanol, or rash areas was prevented except for the artifact band. 126

151 127

152 Figure 4-3. (A) Diagram of the preliminary sites for administration of topical DHEA or topical 1- phenyl-1-hexanol to determine their effect on the NVP-induced skin rash. In 2/2 animals tested, the rash was slightly milder with DHEA, but it was completely prevented in 1-phenyl-1-hexanoltreated areas only (photos not shown). (B) Diagram of sites employed in 2 independent trials to test the effect of topical 1-phenyl-1-hexanol on the NVP-induced skin rash. Five animals in total were treated with NVP (150 mg/kg/day) in food and 1-phenyl-1-hexanol (20 mg/kg/day) on the skin. In 100% of the animals, the rash was prevented by topical 1-phenyl-1-hexanol. One representative rat from each study is shown above. Photos showing (C) skin from the back of a control rat, (D) skin from the back of the NVP only-treated rat, (E) vehicle versus 1-phenyl-1- hexanol-treated areas from an inhibitor-treated rat (topical treatment). 128

153 Figure 4-4. Using skin isolated from rats cotreated with NVP and topical 1-phenyl-1-hexanol using the schematic shown in Figure 3B, epidermal immunoblot analysis was performed. (A) Immunoblot of epidermis from rash areas versus vehicle areas from the epidermis of inhibitortreated rats cotreated with NVP. (B) Immunoblot of epidermis from topical 1-phenyl-1-hexanol areas versus vehicle areas from epidermis of inhibitor-treated rats compared with that of an 129

154 untreated control and a NVP-treated control. 15 µg of protein per lane was loaded for each of A and B. 130

155 Figure 4-5. Representative histology of rat skin isolated from rats cotreated with NVP and topical 1-phenyl-1-hexanol using the schematic shown in Figure 3B. (A) H&E stained sections from upper neck/rash area from control (Ct), nevirapine-treated (NVP), or 1-phenyl-1-hexanol (Ph) rat number 1 shown in 3C-E; (B) H&E stained sections from left shoulder/vehicle area from control (Ct), nevirapine-treated (NVP), or 1-phenyl-1-hexanol (Ph) rat number 1 shown in 3C-E; (C) H&E stained sections from right shoulder/1-phenyl-1-hexanol-treated area from control (Ct), nevirapine-treated (NVP), or 1-phenyl-1-hexanol (Ph) rat number 1 shown in 3C-E. Magnification 20x for all slides in this panel. 131

156 Figure 4-6. (A) Second topical schematic used to test the inhibitor 1-phenyl-1-hexanol. (B) Immunoblot of epidermis from areas of vehicle or inhibitor treated areas using the second schematic shown in 6A. The control is epidermis from the untreated control rat; Ph1 or Ph2 are topical inhibitor-treated epidermal areas from rat # 1 or 2, respectively; Vh1 or Vh2 are vehicle treated epidermal areas for each rat, and RA1 or RA2 are from rash areas with no topical treatment. NVP is from the epidermis of the back of the neck for the NVP-treated positive control rat. Protein loading was 15 µg/lane. 132

157 133

158 Figure 4-7. Histology with H&E staining of skin isolated from rats cotreated with NVP and topical 1-phenyl-1-hexanol using the schematic shown in Figure 6A. The upper left slide of each 134

159 panel is from a control animal without NVP treatment, the upper right slide is from a NVP-treated animal with no topical treatment, and the lower two slides are from animals with NVP + topical treatment. (A) skin from upper back with no topical treatment representing the typical rash; (B) midback where the vehicle was applied in inhibitor-treated animals only (lower two slides); (C) the lower back were 1-phenyl-1-hexanol was applied in inhibitor-treated animals only (lower 2 slides). Magnification 20x. (D) Preincubation of the primary anti-nvp serum with 1.5 mm NVP dissolved in DMSO for 2 h at 37 C prevented covalent binding of the anti-serum to epidermal samples from the samples shown in Figure 3C-E, except for one artifact band. The DMSO control (right most lane) where the primary anti-serum was incubated with DMSO alone. Protein loading was 15 µg/lane. Table 4-2. Comparison of results obtained from 12-OH-NVP sulfate inhibitor studies. Schematic: Treatment group: NVP in food NVP in food + salicylamide via gavage NVP in food + topical 1- phenyl-1-hexanol treatment Effect of treatment on 1-8 µg/ml Below limit of quantification 1-8 µg/ml blood 12-OH-NVP sulfate levels Is covalent binding present with this treatment? Yes; epidermis Yes; epidermis Markedly decreased in inhibitor-treated areas Is rash present with this treatment? Yes Yes (same as with NVP only) None in inhibitor-treated areas 135

160 4.4.5 In Vitro Inhibition of Covalent Binding by 1-Phenyl-1-Hexanol. In order to examine the ability of 1-phenyl-1-hexanol to prevent covalent binding in vitro, a series of studies comparing the covalent binding of 12-OH-NVP +/- PAPS alone with that of 12- OH-NVP in combination with 1-phenyl-1-hexanol +/- PAPS were performed. As shown in Figure 4-8A, 1-phenyl-1-hexanol significantly inhibited covalent binding that occurs in the presence of PAPS and 12-OH-NVP in vitro. This was true in cytosolic fractions of rat skin and liver. A similar pattern was observed when rat skin cytosol was compared with human liver cytosol (Figure 4-8B), and human skin S9 compared to human liver S9 (Figure 4-8C). Human skin incubates were then compared to human E.coli expressed SULT 1A1*1 in order to test if SULT 1A1*1, which is found in human skin, can metabolize 12-OH-NVP. Figure 4-8D displays marked covalent binding of 12-OH-NVP to SULT 1A1*1 in the presence of PAPS, indicating that metabolism to the sulfate had occurred. In the absence of PAPS, no binding was observed. This binding was also prevented by 1-phenyl-1-hexanol. SULT 1A1*1 has a mass of ~ 35 kda, and the multiple bands observed on the immunoblot are due to the presence of E. coli cytosol, which contains other proteins. When binding to SULT 1A1*1 was compared to human skin on the same immunoblot, it was observed that the same 35 kda band was modified in both samples in the presence of PAPS, while only background binding remained in the control lanes. 136

161 137

162 138

163 Figure 4-8. Immunoblot of isolated rat liver cytosol or rat skin cytosol incubated with 12-OH- NVP or a combination of 12-OH-NVP (12-OH) and 1-phenyl-1-hexanol in vitro, in the presence and absence of PAPS. (B) Immunoblot of rat skin cytosol versus female human liver cytosol incubated with or without PAPS and 12-OH-NVP or 12-OH-NVP and 1-phenyl-1-hexanol. (C) Human skin dermatome homogenized and incubated with 12-OH-NVP or 12-OH-NVP + 1- phenyl-1-hexanol to show the same phenomenon exists in human skin. (D) Human SULT 1A1*1 incubated with 1 mm 12-OH-NVP (12-OH) or 12-OH-NVP and 1-phenyl-1-hexanol +/- 0.3 mm PAPS, or 12-OH-NVP sulfate (12-Sulfate). 12 µg/well protein was loaded for each blot. 139

164 4.5 Discussion We previously demonstrated that 12-hydroxylation of NVP was required for the induction of a skin rash in BN rats. 68 We proposed that this was due to the formation of 12-OH-NVP sulfate, which was expected to be chemically reactive. However, two observations argued against this hypothesis: the chemical reactivity of the sulfate was very low, and in the first experiments to test the involvement of the sulfate, we found that inhibition of sulfation by depletion of PAPS with salicylamide decreased the blood levels of the circulating sulfate, but it did not prevent the skin rash. However, in a recent paper, 89 we demonstrated that 12-OH-NVP sulfate readily binds to proteins in the epidermis, both in vitro and in vivo in rats and also in human skin homogenates. With this ability to determine covalent binding of 12-OH-NVP sulfate in the skin, we found that although salicylamide was able to markedly decrease blood levels of 12-OH-NVP sulfate, it did not affect covalent binding in the skin. This is presumably because the mechanism by which salicylamide inhibits sulfation involves depletion of PAPS, and although this works to decrease sulfation in the liver, the turnover of salicylamide in the skin is likely to be too low to deplete PAPS in the skin and prevent covalent binding there. Although this could be considered a negative study, this result is important because it indicates that 12-OH-NVP sulfate formed in the liver is not responsible for covalent binding in the skin. We tried the sulfotransferase inhibitor, DHEA, which did prevent the skin rash, but it also affected blood levels of NVP and 12-OH- NVP; therefore, it was impossible to determine the mechanism by which DHEA prevented the rash. For this reason, the studies with DHEA were not pursued further. We had previously shown that binding of 12-OH-NVP to a skin homogenate requires PAPS but not NADPH. 11 There was no binding of NVP with or without a NADPH-generating system. Both 1-phenyl-1-hexanol, a known sulfotransferase inhibitor, and 12-OH-NVP are benzylic alcohols and may be substrates for hydroxysteroid SULTs, which are major SULTs in female rat skin. 95 We also found that 1-phenyl-1-hexanol inhibited covalent binding of 12-OH- NVP mediated by SULT 1A1*1, which is a polymorphic sulfotransferase found in human skin. 71,103 Treatment of rats with oral 1-phenyl-1-hexanol did not prevent covalent binding or the rash. Although we did not develop an assay to study the metabolism of 1-phenyl-1-hexanol, it is likely that it was cleared rapidly and did not reach inhibitory concentrations in the skin. In contrast, topical administration of 1-phenyl-1-hexanol did prevent covalent binding of NVP in the 140

165 skin as well as the rash and histological changes, but only where it was applied. This, along with the previous study showing that it is 12-OH-NVP sulfate that is responsible for NVP covalent binding in the skin 89 provides conclusive evidence that this is the species responsible for the rash. The finding that binding also occurs to the human SULT 1A1 provides a strong link to the NVPinduced skin rash in humans. It is also important to note that as opposed to the liver, covalent binding in the skin may have a direct effect on organ-specific immune responses; i.e., the skin is not a major site of xenobiotic biotransformation and does not have the same tolerogenic mechanisms in place as the liver. 104 Rather, the epidermis is immunologically active and has been termed an adjuvant, with keratinocytes able to modulate numerous immune activities. 50 Additionally, if covalent binding in the skin were involved as we hypothesize, the resulting immune response and rash would take time to develop, which is true for NVP-induced skin rash. 65 This is the first study to use a valid animal model to demonstrate that a reactive metabolite is responsible for an idiosyncratic drug reaction, in this case a reactive metabolite formed in the skin. This mechanism is an alternative to the p-i hypothesis for the mechanism of an idiosyncratic drug reaction in the skin. Sulfotransferase is one of the few metabolic enzymes with significant activity in the skin, and it may be responsible for bioactivation of other drugs that cause skin rashes. 141

166 FUNDING SUPPORT. This work was supported by a grant from the Canadian Institutes of Health Research (MPO84520). ACKNOWLEDGMENTS. The authors thank Boehringer-Ingelheim for kindly supplying nevirapine. A.M.S. is the recipient of a University of Toronto Pharmaceutical Sciences Doctoral Fellowship. J.U. is the recipient of the Canada Research Chair in Adverse Drug Reactions. Portions of this work were parts of presentations given by A.M. Sharma and J.P.Uetrecht at the ISSX meeting in Japan, 2011, and the Society of Toxicology International Meetings in San Francisco, CA., USA, 2012 and in San Antonio, TX., USA, ABBREVIATIONS: Brown Norway, BN; 12-hydroxynevirapine, 12-OH-NVP; 12-OH-NVP sulfate, 12-OH-NVP Sulfate; cytochrome P450, P450; human immunodeficiency virus, HIV; idiosyncratic drug reaction, IDR; liquid chromatography/mass spectrometry, LC/MS; nevirapine, NVP; tris-buffered saline tween-20, TBST; dehydroepiandosterone, DHEA; 9,000 X g supernatant, S9; salicylamide, SA; sulfotransferase, SULT; 3 -phosphoadenosine-5 - phosphosulfate, PAPS; keratinocyte, KC; Steven s-johnson syndrome, SJS; Toxic epidermal necrolysis, TEN. 142

167 4.6 Supplemental Material Quantification of NVP, 12-OH-NVP, 12-OH-NVP Sulfate, and 4- COOH-NVP in Urine. Urine, 50 µl from a 24-h sample was mixed with 100 µl internal standard (ethyl-nvp, 27 µg/ml in the mobile phase) and 10 µl of β-glucuronidase (approximately 10,000 U/mL in 100 mm KH 2 PO 4 buffer, ph 7.4) and incubated overnight at 37 C prior to concentrating on an Strata solid phase extraction column as per above. The samples were separated using the same HPLC conditions as described for plasma samples. As with plasma samples, for quantification of 12-OH-NVP sulfate, the internal standard was naproxen and no preincubation with β- glucuronidase was performed Grading of Skin Rash. The grading scheme according to the AIDS Clinical Trial Group Protocol Management Handbook Table for Grading Severity of Cutaneous Eruptions was used (where applicable to rats). 98 Grading was performed at the time of sacrifice using an area of shaved 1 x 1 inch skin on the upper neck/back area of each rat. (0) grade 0, normal integrity of skin is maintained (1) grade 1, erythema with or without pruritus; (2) grade 2, a diffuse erythematous macular or maculopapular cutaneous eruption or dry desquamation with or without pruritus or typical target lesions without blistering, vesicles, or ulcerations in the lesions; (3) grade 3, 1 of the following clinical presentations: urticaria; diffuse erythematous macular or maculopapular cutaneous eruption or moist desquamation with or without pruritus together with any of the 4 constitutional findings possibly related to the drug (i.e., blistering, vesiculation, or both of cutaneous eruptions; or any site of mucosal lesions considered related to study drug without other etiology, such as herpes simplex or aphthous ulcer); angioedema; exfoliative dermatitis (defined as severe widespread erythema and dry scaling of the skin and generalized superficial lymphadenopathy, with other constitutional findings possibly related to study drug such as fever or weight loss); or diffuse rash and serum sickness-like reactions defined as clinical symptom complex manifested as fever, lymphadenopathy, edema myalgia, arthralgia, or a combination; and (4) grade 4, diffuse cutaneous eruptions usually starting on the face, trunk, or back, often with prodromal symptoms plus one of the following: cutaneous bullae, sometimes confluent with widespread sheet like detachment of skin (Nikolsky s sign), Stevens-Johnson syndrome, erythema multiforme major, or toxic epidermal necrolysis, or 2 or more anatomically distinct sites of mucosal erosion or ulceration not due to another cause. Severe rash was defined as grade 3 and grade 4 cutaneous eruptions when we used this grading scheme. Skin biopsies were not required for categorization of rash 143

168 Grading Table Table 4S-1: Day 21 Skin Rash Grading Key: UB = upper back; MB = mid-back; LS = left upper shoulder. Rat Treatment Area of Skin Description Rash Grade 1x1 Inch Day 21 Control Rat 1 UB No visible skin abnormality 0 Day 21 Control Rat 2 UB No visible skin abnormality 0 Day 21 Gavaged Control Rat 1 UB No visible skin abnormality 0 Day 21 Gavaged Control Rat 2 UB No visible skin abnormality 0 Day OH Rat 1 LS/UB DEEP LESIONS 3 Day OH Rat 2 UB VERY BAD; PELT LIKE; DEEP LESIONS 4 Day 21 NVP + Salicylamide Rat 1 UB/MB MANY LESIONS; PEELING OF EPIDERMIS 4 Day 21 NVP + Salicylamide Rat 2 UB VERY RED, MANY LESIONS 4 Day 21 NVP + Salicylamide Rat 3 UB LARGER/MANY LESIONS 4 144

169 Day 21 NVP + Salicylamide Rat 4 UB/MB REDNESS; LESIONS 3 Day 21 NVP Rat 1 UB/MB VERY RED, VERY DEEP 3 Day 21 NVP Rat 2 UB LESS LESIONS, VERY RED 3 Day 21 NVP Rat 3 UB/MB BLOODY/SLOUGHING 4 Day 21 NVP Rat 4 UB PEELING/LESS RED 3 145

170 12-OH-NVP sulfate ( g/24h) 4-COOH-NVP ( g/24h) 3-OH-NVP ( g/24h) 12-OH-NVP ( g/24h) 2-OH-NVP ( g/24h) A NVP NVP + salicylamide NVP + DHEA D NVP NVP + salicylamide NVP + DHEA B Treatment day 5000 NVP NVP + salicylamide 4000 NVP + DHEA C Treatment day 1500 NVP NVP + salicylamide NVP + DHEA 1000 E Treatment day NVP NVP + salicylamide NVP + DHEA Treatment day Treatment day Figure 4S-1. Urinary excretion of (A) 12-OH-NVP, (B) 4-COOH-NVP, (C) 12-OH-NVP sulfate, (D) 2-OH-NVP, and (E) 3-OH-NVP from rats treated with NVP (100 mg/kg/day), NVP + DHEA (100 mg/kg/day) each or NVP + salicylamide (274 mg/kg/day; n = 4 in each group). Data depicts the mean ± SD. 146

171 Figure 4S-2. H&E stained sections comparing the histology of rat skin in response to (A) NVP treatment, (B) NVP + topical DHEA cotreatment, or (C) control rats. Magnification = 20x. 147