Investigation of Penicillamine-Induced Autoimmunity

Mechanistic Investigation of Penicillamine-Induced Autoimmunity: Covalent Binding of Penicillamine to Macrophages, Involvement of Th17 cells, and Its Relation to Idiosyncratic Drug-induced Liver Injury

Jinze Li

A thesis submitted in conformity with the requirements for the degree of DOCTOR OF PHILOSOPHY Graduate Department of Pharmaceutical Sciences University of Toronto ©Copyright by Jinze Li 2009

Library and Archives Bibliothèque et Canada Archives Canada

Published Heritage Direction du Branch Patrimoine de l’édition

395 Wellington Street 395, rue Wellington Ottawa ON K1A 0N4 Ottawa ON K1A 0N4 Canada Canada

Your file Votre référence ISBN: 978-0-494-61128-9 Our file Notre référence ISBN: 978-0-494-61128-9

NOTICE: AVIS:

The author has granted a non- L’auteur a accordé une licence non exclusive exclusive license allowing Library and permettant à la Bibliothèque et Archives Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par télécommunication ou par l’Internet, prêter, telecommunication or on the Internet, distribuer et vendre des thèses partout dans le loan, distribute and sell theses monde, à des fins commerciales ou autres, sur worldwide, for commercial or non- support microforme, papier, électronique et/ou commercial purposes, in microform, autres formats. paper, electronic and/or any other formats. . The author retains copyright L’auteur conserve la propriété du droit d’auteur ownership and moral rights in this et des droits moraux qui protège cette thèse. Ni thesis. Neither the thesis nor la thèse ni des extraits substantiels de celle-ci substantial extracts from it may be ne doivent être imprimés ou autrement printed or otherwise reproduced reproduits sans son autorisation. without the author’s permission.

In compliance with the Canadian Conformément à la loi canadienne sur la Privacy Act some supporting forms protection de la vie privée, quelques may have been removed from this formulaires secondaires ont été enlevés de thesis. cette thèse.

While these forms may be included Bien que ces formulaires aient inclus dans in the document page count, their la pagination, il n’y aura aucun contenu removal does not represent any loss manquant. of content from the thesis.

ABSTRACT

By Jinze Li

Faculty of Pharmacy, University of Toronto 2009

The mechanisms of idiosyncratic drug reactions (IDRs) are unknown; however, most appear to be immune-mediated. Their idiosyncratic nature and the paucity of animal models make mechanistic studies very difficult. One of the few animal models is penicillamine- induced autoimmunity in Brown Norway rats. The major focus of this thesis was the use of this model to study the interaction between penicillamine and macrophages, the involvement of Th17 cells, and extension of this model to idiosyncratic drug-induced liver injury. One of the costimulatory signals leading to T cell activation appears to be reversible Schiff-base formation between an amine on T cells and an aldehyde on macrophages. We hypothesized that penicillamine binds to these aldehydes leading to macrophage activation and autoimmunity. By using biotinylated aldehyde-reactive agents such as ARP, we demonstrated the existence of aldehydes on the surface of macrophages. We synthesized biotinylated-penicillamine and it also binds to macrophages. Several proteins to which ARP binds were identified providing clues to the signal transduction pathways leading to macrophage activation. Biological consequences of this binding were investigated with a microarray study. ARP binding was also observed in the macrophage cell line, RAW264.7, and incubation with penicillamine stimulated the production of TNF-α, IL-6, and IL-23.

II Hydralazine and isoniazid, which are known to cause a lupus-like syndrome in humans and irreversibly bind to aldehyde groups, were also found to activate RAW264.7 cells. Th17 cells are prominent in autoimmune syndromes and Th17-associated cytokines such as IL-17 were elevated in the penicillamine-treated animals that developed autoimmunity. We have hypothesized that some drug-induced liver injury has an autoimmune component. A pilot study quantified serum concentrations of 26 cytokines/chemokines in patients with various forms of acute liver failure (ALF): idiosyncratic drug-induced ALF, acetaminophen-induced ALF, and viral hepatitis. IL-17 was elevated in 60% of patients with idiosyncratic drug-induced ALF, which supports an autoimmune component in these patients; however, it was also elevated in many cases of acetaminophen-induced ALF, presumably released by the innate immune system. These studies provide important insights into the mechanism of penicillamine-, hydralazine-, and isoniazid-induced autoimmunity and also provide clues to other IDRs that may have an autoimmune component.

III ACKNOWLEDGEMENT

It still feels like yesterday when Jack interviewed me in his office, his face beaming while making the remark “learning science in graduate school is fun!”. Four years have already passed by and I am writing this concluding note with a very mixed feeling. This four-year graduate school in Jack’s lab is a truly exciting adventure and one of the most important steps in my career. I would regret my doctoral education if I did not join Jack’s lab in which I benefited from the generous help of many friends and colleagues to whom I want to express my sincere gratitude. First of all, I would like to whole-heartedly thank my supervisor, Dr. Jack Uetrecht for his great mentorship and friendship throughout this program. Being a mentor, he gives trust, encouragement, and guidance that have allowed me to complete this challenging work in a timely fashion. It is really a great honor and pleasure to work with him. Being a friend, he cares about my well-being and shares with me his thoughts and experience that makes me feel belong. To quote what Jack said at his conference of adverse drug reactions a couple of years ago, “the students do become in a sense like your family and like a family, you see them develop and go off and succeed”, to me this four-year experience does feel more like a family journey that I would cherish forever. I would also like to thank my advisory committee, Dr. Allan Okey, Dr. Micheline Piquette-Miller, Dr. Robert Inman, the internal examiner, Dr. Robert Macgregor, and the external examiner of my thesis, Dr. Dan Wierda from the Lilly Research Labs. I very much value the opinions and time taken by them throughout my research program and thesis preparation. Very importantly, I give my regards and blessings to all my lab buddies in Jack’s group, particularly Wei Lu, Julia Ip, Jie Chen, Baskar Mannargudi, Tharsika Tharmanathan, Xu Zhu, Feng Liu, Xin Chen, Ping Cai, Xiaochu Zhang, Stephanie Pacitto, and Jacintha Shenton for their intellectual inputs, friendship, and the happy hours of gossiping together, which is certainly one of major sources of fun during my PhD education. Meanwhile, I

IV want to give my special thanks to Hong Gou as a great friend who has been supporting and inspiring me in many respects. Finally, I would like to take this opportunity to thank my wife, Yanmei Chen whose constant encouragement and love I have relied on throughout my PhD program. Also, I am indebted to my daughter, mom, and dad for their care, understanding, and love. This work would not have been possible without my family. It is to them that I dedicate this dissertation.

V TABLE OF CONTENTS

ABSTRACT...... II

ACKNOWLEDGEMENT...... IV

TABLE OF CONTENTS...... VI

LIST OF THESIS PUBLICATIONS AND ABSTRACTS...... IX

LIST OF ABBREVIATIONS ...... X

LIST OF FIGURES ...... XIV

LIST OF TABLES...... XVI

CHAPTER 1...... 1

GENERAL INTRODUCTION ...... 1

1.1. ADVERSE DRUG REACTIONS...... 2

1.2. IDIOSYNCRATIC DRUG REACTIONS...... 5

1.2.1. CLINICAL MANIFESTATIONS OF IDRS...... 6 1.2.2. RISK FACTORS FOR IDRS...... 6

1.3. WORKING HYPOTHESIS OF MECHANISMS OF IDRS ...... 7

1.3.1. OVERVIEW OF IMMUNOLOGICAL MODELS...... 7 1.3.1.1. SELF-NONSELF MODELS...... 8 1.3.1.2. DANGER MODEL...... 9 1.3.2. WORKING HYPOTHESIS OF IDRS ...... 11 1.3.2.1. HAPTEN HYPOTHESIS...... 11 1.3.2.2. DANGER HYPOTHESIS ...... 11 1.3.2.3. PHARMACOLOGICAL INTERACTION (PI) HYPOTHESIS...... 12

1.4. ANIMAL MODELS OF IDRS ...... 14

1.4.1. NEVIRAPINE-INDUCED SKIN RASH IN BN RATS ...... 14 1.4.2. PENICILLAMINE-INDUCED AUTOIMMUNITY IN BN RATS...... 15

1.5. AUTOIMMUNITY AND IMMUNE TOLERANCE...... 18

1.5.1. MECHANISMS OF AUTOIMMUNE DIEASES...... 18

VI 1.5.2. IMMUNOLOGICAL TOLERANCE TO SELF TISSUES...... 20 1.5.3. DRUG-INDUCED AUTOIMMUNITY...... 21 1.5.4. PROTEIN CARBONYLATION IN AUTOIMMUNE DISEASES...... 24

1.6. DRUG-INDUCED LIVER INJURY ...... 26

CHAPTER 2...... 30

COVALENT BINDING OF PENICILLAMINE TO MACROPHAGES: IMPLICATIONS FOR PENICILLAMINE-INDUCED AUTOIMMUNITY ...... 30

2.1. ABSTRACT...... 31

2.2. INTRODUCTION...... 32

2.3. MATERIALS AND METHODS...... 35

2.4. RESULTS...... 41

2.5. DISCUSSION ...... 50

CHAPTER 3...... 54

D-PENICILLAMINE-INDUCED AUTOIMMUNITY: RELATIONSHIP TO MACROPHAGE ACTIVATION ...... 54

3.1. ABSTRACT...... 55

3.2. INTRODUCTION...... 56

3.3. MATERIALS AND METHODS...... 59

3.4. RESULTS...... 62

3.5. DISCUSSION ...... 74

CHAPTER 4...... 76

TH17 INVOLVEMENT IN PENICILLAMINE-INDUCED AUTOIMMUNE DISEASE IN BROWN NORWAY RATS...... 76

4.2. INTRODUCTION...... 78

4.3. MATERIALS AND METHODS...... 80

4.4. RESULTS...... 82

4.5. DISCUSSION ...... 90

VII CHAPTER 5...... 92

CYTOKINE AND AUTOANTIBODY PATTERNS IN ACUTE LIVER FAILURE ...... 92

5.1. ABSTRACT...... 93

5.2. INTRODUCTION...... 94

5.3. PATIENTS AND METHODS...... 96

5.4. RESULTS...... 98

5.5. DISCUSSION ...... 107

CHAPTER 6...... 110

OVERALL CONCLUSIONS AND FUTURE DIRECTIONS ...... 110

6.1. SUMMARY ...... 111

6.2. IMPLICATIONS AND FUTURE DIRECTIONS...... 117

REFERENCES...... 120

APPENDIX: SUPPLEMENTAL DATA: LUMINEX DATA FOR ALL CYTOKINES/CHEMOKINES ...... 137

VIII LIST OF THESIS PUBLICATIONS AND ABSTRACTS

Articles 1. Jinze Li, Baskar Mannargudi, and Jack Uetrecht. Covalent binding of D-penicillamine onto macrophages. Chemical Research in Toxicology 2009 July; 22 (7): 1277-84. 2. Jinze Li and Jack Uetrecht. D-penicillamine-induced autoimmunity: relationship to macrophage activation. Chemical Research in Toxicology 2009 July 6. 3. Jinze Li, Carron Sanders, William M. Lee, and Jack Uetrecht. Cytokine and autoantibody patterns in acute liver failure. Submitted. 4. Jinze Li, Xu Zhu, and Jack Uetrecht. Involvement of Th17 pathway in D-penicillamine- induced autoimmune disease. In submission.

Book chapter

1. Jinze Li and Jack Uetrecht. Danger hypothesis. Mechanisms of Adverse Drug Reactions Handbook of Experimental Pharmacology 2009.

Abstracts 1. Jinze Li, Xu Zhu, Feng Liu, Ping Cai, Carron Sanders, William M. Lee, and Jack Uetrecht. Idiosyncratic drug-induced liver injury is characterized by variable patterns of cytokines, chemokines, and autoantibodies. 60th Annual meeting of the American Association for the Study of Liver Diseases Boston 2009. 2. Jinze Li, Xu Zhu, and Jack Uetrecht. Th17 involvement in penicillamine-induced autoimmunity. Society of Toxicology 48th Annual Meeting Baltimore 2009. 3. Jinze Li and Jack Uetrecht. Investigation of mechanism of D-penicillamine-induced autoimmunity: Is it mediated by activation of macrophages? 3rd Drug Hypersensitivity Meeting Paris 2008. 4. Jinze Li and Jack Uetrecht. Global gene expression profiling of macrophages in response to D-penicillamine treatment. Society of Toxicology 46th Annual Meeting Charlotte 2007.

IX LIST OF ABBREVIATIONS

Chapter 1 AD, Alzheimer’s disease ADR, Adverse drug reaction AGE, Advanced glycation end product ALE, Advanced lipoxidation end product ALT, Alanine aminotransferase ANA, Antinuclear antibody APC, Antigen presenting cell ARP, Aldehyde reactive probe ATP, Adenosine triphosphate BN, Brown Norway CD, Cluster of differentiation CTLA, Cytotoxic T-Lymphocyte Antigen DAMP, Damage associated molecular pattern DILI, Drug induced liver injury DNA, Deoxyribonucleic acid EAE, Experimental autoimmune encephalomyelitis FOXP3, Forkhead box P3 HIV, Human immunodeficiency virus HLA, Histocompatibility leukocyte antigen IDILI, Idiosyncratic drug induced liver injury IDO, Indoleamine 2,3-dioxygenase IDR, Idiosyncratic drug reaction IgE, Immunoglobulin E IL, Interleukin

X I.M., Intramuscular INS, Infectious-nonself I.P., Intraperitoneal IPEX, Immunodysregulation polyendocrinopathy enteropathy X-linked syndrome LFA, Lymphocyte function associated antigen LTT, Lymphocyte transformation test MHC, Major histocompatibility MOMP, Mitochondrial outer membrane permeabilisation mPT, Mitochondrial permeability transition mRNA, Messenger ribonucleic acid NSAIDs, Non-steroidal anti-inflammatory drugs PAMP, Pathogen associated molecular pattern p-ANCA, Antineutrophil cytoplasmic antibody PI, Pharmacological interaction Poly I:C, Polyinosinic-Polycytidylic acid PRR, Pathogen recognition receptor RCS, Reactive carbonyl species RM, Reactive metabolite ROS, Reactive oxygen species S.C., Subcutaneous SD, Sprague-Dawley SLE, Systemic lupus erythematosus SNS, Self-nonself TCR, T cell receptor TGF, Transforming growth factor Th, T helper cell TNF, Tumor necrosis factor WHO, World health organization

XI Chapter 2 ARP, Aldehyde reactive probe FBS, Fetal bovine serum HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid RPMI, Roswell Park Memorial Institute MS, Mass spectrometry NMR, Nuclear magnetic resonance SDS-PAGE, Sodium dodecyl sulfate polyacrylamide gel electrophoresis STAT, Signal transducer and activator of transcription Streptavidin-APC, Streptavidin-Allophycocyanin

Chapter 3 ALOX5AP, Arachidonate 5-lipoxygenase activating protein ALOX12, Arachidonate 12-lipoxygenase CCL, Chemokine ligand CCR/CXCR, Chemokine receptor DMEM, Dulbecco’s modified eagle’s medium DUSP, Dual specificity phosphatase ELISA, Enzyme-linked immunosorbent assay GCOS, Genechip operating system GM-CSF, Granulocyte macrophage colony-stimulating factor IFN, Interferon MACS, Magnetic cell separation technology MIP, Macrophage inflammatory protein NK cell, Natural killer cell SBP, Selenium binding protein

Chapter 4

XII CRO/KC, Cytokine growth-related oncogene G-CSF, Granulocyte colony-stimulating factor MCP, Monocyte chemotactic protein PMA, Phorbol myristate acetate ROR, Orphan nuclear receptor VEGF, Vascular endothelial growth factor

Chapter 5 ALF, Acute liver failure ALFSG, Acute liver failure study group APAP, Acetaminophen AST, Aspartate aminotransferase BAFF, B-cell activating factor IDILI, Idiosyncratic drug induced liver injury INR, International normalized ratio MPO, Myeloperoxidase Scl-70, Anti-scleroderma antibody shRNA, Small hairpin RNA SSB/La, Sjogren syndrome antigen B

XIII LIST OF FIGURES

FIGURE 1. DRUG WITHDRAWAL FROM THE US MARKET DUE TO ADRS BETWEEN 1976 AND 2005...... 4 FIGURE 2. PROGRESSION OF IMMUNOLOGICAL MODELS...... 10 FIGURE 3. MODES OF THE INTERACTION BETWEEN DRUGS AND IMMUNE SYSTEM...... 13 FIGURE 4. DOSE DEPENDENCE OF PENICILLAMINE-INDUCED AUTOIMMUNITY IN BN RATS...... 17 FIGURE 5. ADVANCED GLYCATION END PRODUCTS...... 25 FIGURE 6. MECHANISTIC HYPOTHESES OF PATHOGENESIS OF DILI. MODIFIED FROM ABBOUD G AND...... 29 FIGURE 7. BINDING OF PENICILLAMINE AND ARP TO ALDEHYDES ON THE SURFACE OF MACROPHAGES...... 34 FIGURE 8. SYNTHETIC SCHEME FOR BIOTIN-D-PENICILLAMINE...... 40

FIGURE 9. DOSE-RESPONSE CURVES OF BINDING OF ARP (A, N=3) AND BIOTIN-HYDRAZINE (B, N=3) TO SPLENOCYTES AND MACROPHAGES OF BN RATS...... 43

FIGURE 10. DOSE-RESPONSE CURVE OF RAW 264.7 MURINE MACROPHAGES (0.25 MILLION CELLS) STAINING WITH ARP (N=3)...... 44

FIGURE 11. DOSE-RESPONSE CURVES OF SPLENOCYTES AND MACROPHAGES STAINING WITH BIOTIN- PENICILLAMINE (N=3)...... 45 FIGURE 12. DECREASE IN ARP STAINING OF SPLENIC MACROPHAGES BY PRE-INCUBATION WITH PENICILLAMINE

OR HYDRALAZINE...... 46 FIGURE 13. DOSE RESPONSE OF ARP STAINING OF SPLENIC MACROPHAGES FROM BN, SPRAGUE-DAWLEY, AND

LEWIS RATS (N=3)...... 47 FIGURE 14. SDS-PAGE IMAGE OF PROTEIN TARGETS OF ARP OR BIOTIN-PENICILLAMINE...... 48

FIGURE 15. HYPOTHESIS THAT COVALENT BINDING OF PENICILLAMINE TO MACROPHAGES LEADS TO MACROPHAGE ACTIVATION...... 58

FIGURE 16. COMPARISON OF TRANSCRIPTOME OF MACROPHAGES WITHIN THE CONTROL AND PENICILLAMINE GROUPS...... 65 FIGURE 17. VALIDATION OF EXPRESSION OF DIFFERENTIALLY REGULATED GENES BY QRT-PCR...... 66 FIGURE 18. MRNA EXPRESSION PROFILE OF CYTOKINES IN NK CELLS AT 6 H POST-DOSAGE OF PENICILLAMINE..67

FIGURE 19. DOSE-RESPONSE CURVE OF RAW264.7 MACROPHAGES (1 MILLION CELLS) STAINING WITH ARP (N=3)...... 68 FIGURE 20. INDUCTION OF CYTOKINE PRODUCTION IN RAW 264.7 CELLS BY PENICILLAMINE (N=3)...... 69 FIGURE 21. IL-6 PRODUCTION IN RAW264.7 MACROPHAGES INCUBATED WITH PENICILLAMINE, HYDRALAZINE,

OR ISONIAZID FOR 24 H (N=3)...... 70 FIGURE 22. SERUM CONCENTRATION OF IL-6: D-PENICILLAMINE VS. CONTROL (N=3)...... 84 FIGURE 23. A REPEAT OF SERUM IL-6 DETERMINATION IN PENICILLAMINE TREATMENT...... 85

XIV FIGURE 24. PHENOTYPE OF SPLENIC CD4+ T CELLS FROM PENICILLAMINE-TREATED RATS AT THE END OF...... 86 FIGURE 25. CHANGES OF BODY WEIGHT AND CUMULATIVE INCIDENCE OF AUTOIMMUNITY...... 87 FIGURE 26. COMPARISON OF THE NUMBER OF SPLENOCYTES BETWEEN SICK (N=15) AND NON-SICK ...... 87 FIGURE 27. SERUM CYTOKINE/CHEMOKINE PATTERN: SICK (N=15) VS. NON-SICK (N=5)...... 88 FIGURE 28. BIOCHEMICAL PARAMETERS OF LIVER FAILURE PATIENTS...... 100 FIGURE 29. SERUM CYTOKINE/CHEMOKINE COMPARISON BETWEEN PATIENT GROUPS...... 101 FIGURE 30. SERUM LEVELS OF B-CELL ACTIVATION FACTOR (BAFF)...... 102 FIGURE 31. SERUM LEVELS OF ANA...... 102 FIGURE 32. SERUM LEVELS OF ANTI-MPO ANTIBODIES...... 103 FIGURE 33. WORKING HYPOTHESIS OF THE PATHOGENESIS OF PENICILLAMINE-INDUCED AUTOIMMUNITY...... 116

XV LIST OF TABLES

TABLE 1. CLASSIFICATION OF ADVERSE DRUG REACTIONS ...... 4 TABLE 2. IMMUNE-MEDIATED ADVERSE DRUG REACTIONS...... 5 TABLE 3. INFLUENCE OF IMMUNOMODULATORS ON PENICILLAMINE-INDUCED AUTOIMMUNITY ...... 17 TABLE 4. EXAMPLES OF LUPUS-INDUCING DRUGS...... 23 TABLE 5. DRUGS THAT ARE ASSOCIATED WITH BOTH IDILI AND AUTOIMMUNITY ...... 28 TABLE 6. APPARENT ARP-BINDING PROTEINS ...... 49

TABLE 7. DIFFERENTIALLY EXPRESSED MACROPHAGE GENES IN BROWN NORWAY RATS AT 6 H POST-DOSAGE OF PENICILLAMINE...... 71 TABLE 8. PRIMER SEQUENCES FOR QRT-PCR...... 73 TABLE 9. PRIMER SEQUENCES FOR QRT-PCR...... 89 TABLE 10. CORRELATION OF IL-17 WITH OTHER ANALYTES...... 104 TABLE 11. IL-17, IL-21, IL-6, IP-10, ANA, AND ANTI-MPO, IN IDILI PATIENTS ...... 105

XVI

CHAPTER 1

GENERAL INTRODUCTION

1 1.1. ADVERSE DRUG REACTIONS

Law of unintended consequences: any purposeful action will produce some unintended consequences. The unintended side effect can potentially be more significant than any of the intended effects. In medicine, unwanted effects generally refer to adverse effects, and more specifically when referring to effects of drugs, adverse drug reactions (ADRs). The World Healthy Organization (WHO) defines ADRs as harmful, unintended reactions to medicines that occur at doses normally used for treatment (1). Clinical manifestations of ADRs have significant variance in severity, ranging from slight uncomfortable situations such as dryness of mouth, to serious reactions such as cardiac failure. In the practice of ADRs surveillance, health care practitioners and researchers tend to focus on reactions that have a potential to cause serious damage or even death to patients, which usually requires effective treatment including hospitalization. By excluding minor reactions, the American Society of Consultant of Pharmacists redefine ADRs as any unexpected, unintended, undesired, or excessive response to a drug that (2): Requires discontinuing the drug (therapeutic or diagnostic); Requires changing the drug therapy; Requires modifying the dose (except for minor dosage adjustments); Necessitates admission to a hospital; Prolongs stay in a health care facility; Necessitates supportive treatment; Significantly complicates diagnosis; Negatively affects prognosis; Results in temporary or permanent harm, disability, or death. Serious ADRs are a common and significant problem in health care. In America, over 2 million serious ADRs are reported yearly, of which 100,000 patients directly die from ADRs. ADRs are the 4th leading cause of mortality and morbidity ahead of pulmonary disease, diabetes, and AIDS etc (3). They account for about 7% of hospital admissions (4). The drugs most commonly reported in cases of ADRs are: antidiabetic agents, anticoagulants,

2 anticonvulsants, beta-blockers, and non-steroidal anti-inflammatory drugs (NSAIDs). Meanwhile, drug-related deaths from ADRs costs more than $136 billion a year. To date, about 10% of drugs have been withdrawn from the market or received a Black Box warning (5, 6), which presents a significant challenge to the development of new drugs (Figure1). ADRs are conventionally classified into six types: A (dose-dependent, namely an enhanced pharmacological effect), B (bizarre or idiosyncratic, with unknown mechanisms but most likely involving the immune system), C (chronic or time-related), D (delayed effects), E (end-of-treatment effects), and F (failure of therapy) (Table 1) (7). In the past decade, many efforts have been applied to develop an effective surveillance system for ADRs in many countries, and significant progress has been made on mechanistic understanding of most ADRs except type B reactions. In contrast to other ADRs, type B reactions or idiosyncratic drug reactions (IDRs) are still poorly understood and hence unpredictable. However, clinical features of many IDRs and numerous studies of several critical animal models suggest that the immune system is involved in most IDRs (8). If this is correct, IDRs could also be referred to as immune-mediated adverse drug reactions that can be further classed as IgE-mediated, cytotoxic, immune-complex-mediated, and cell- mediated, etc (9) (Table 2).

Figure 1. Drug withdrawal from the US market due to ADRs between 1976 and 2005. Adapted from Wilke RA et al (10). “Other” refers to haemolytic anaemia, skin disease, immune toxicity, gastrointestinal toxicity, respiratory toxicity, fatal, neurotoxicity, blood-related toxicity, and birth defects.

Table 1. Classification of adverse drug reactions Type of reaction Clinical characteristics Example Solutions Dose dependent Predictable Digoxin toxicity Reduce dose Pharmacology related Low mortality Bizarre Unpredictable Penicillin hypersensitivity Discontinue the drug Pharmacology unrelated treatment High mortality Chronic Corticosteroids Reduce dose Delayed Usually dose related Carcinogenesis Withdrawal Shortly after discontinuation of β-blocker withdrawal Restart the medication medication (myocardial ischaemia) Failure therapy Dose related1 Inadequate dosage of an Increase dose Usually caused by drug –drug oral contraceptive interactions 1. In some cases the patient simply does not have the receptor or there is some other reason that the patient does not respond no matter what the dose is, in which increasing dose will not help on therapeutic effect much.

4 1.2. IDIOSYNCRATIC DRUG REACTIONS

Idiosyncratic drug reactions (IDRs) refer to adverse drug reactions (ADRs) that do not occur in most patients at any dose used clinically and in which the mechanism does not involve the known pharmacological properties of the drug (11). Although rare, with a typical incidence from 1/100 to 1/100000, because of the total number of drugs involved and the number of patients treated, such reactions are common, accounting for 6-10% of all ADRs (12, 13). IDRs represent a major clinical problem in that most IDRs are very serious, even life threatening. At present, they are impossible to predict, largely due to limited understanding of involved mechanisms, which adds marked uncertainty to new drug development and, hence, present a big challenge to pharmaceutical industry. It is unlikely that much progress will be made in preventing such reactions until their mechanisms are well understood. Nevertheless, the clinical features of IDRs provide us with clues that the immune system is involved in most cases, for instance, drug-induced autoimmune syndromes (i.e. lupus) (14). Additional evidence supports an immune-mediated mechanism for IDRs: in general, there is a delay between starting drug treatment and the onset of adverse reactions but there is usually a rapid onset on rechallenge. Such characteristics have served as the theoretical basis to focus on the interaction between parent drug or reactive metabolite and immune system leading to a pathogenic immune reaction.

Table 2. Immune-mediated adverse drug reactions Type Features Example IgE-mediated Anaphylactic reactions: hypotension etc. Penicillin-induced anaphylaxis Immune complex-mediated Allergic reactions: rashes, fever etc. Serum sickness Fas/Fas ligand-mediated Serious epidermal necrolysis Stevens-Johnson syndrome T cell-mediated Allergic reactions in skin Abacavir-induced hypersensitivity Less clear Arthritis, skin lesions etc. Drug-induced autoimmune syndrome

5 1.2.1. CLINICAL MANIFESTATIONS OF IDRS

IDRs can present with involvement of any organ system while skin, liver, and bone marrow are most commonly affected. Examples include hydralazine-induced lupus, procainamide-induced hypersensitivity syndrome, clozapine-induced agranulocytosis, nevirapine-induced skin rash, penicillamine-induced autoimmune syndromes, and minocycline-induced liver injury etc. Despite the variance in the manifestations of IDRs for each drug or patient, there is usually a delay between starting the drug and onset of clinical symptoms, especially on primary exposure (9, 15, 16). This is one of golden characteristics of an immune-mediated reaction in which it always takes some time (days to weeks or even months) for the immune cells to proliferate into sufficient numbers and differentiated into pathogenic cell clones. This is particularly true for adaptive immunity in which lymphocyte activation is multi-signal dependent. Another signature characteristic of adaptive immune responses is the production of memory T or B lymphocytes on primary exposure so that immune system is able to deliver a much more efficient response to an antigenic stimulation on re-exposure. Therefore, the observation of a rapid onset on rechallenge to drugs associated with IDRs provides strong support for an immune-mediated mechanism. However, exceptions are occasionally observed, especially in case of drug-induced autoimmunity. Thus, a shortened time delay is not sufficient to exclude the involvement of immune system because, by definition, these autoimmune reactions are immune-mediated.

1.2.2. RISK FACTORS FOR IDRS

To date, many factors have been suggested by epidemiologic data to be risk factors for specific IDRs, such as gender, age, disease state, and genetic predisposition, etc (8). Older people seem to have an increased risk of IDRs, possibly due to drug-drug interactions from their multiple prescriptions. Also, older people are more sensitive to many drugs because many physiological functions are changed significantly with aging (17). This is particularly true for idiosyncratic drug-induced liver injury (18). Besides age, female gender has been found to carry an increased risk of some IDRs such as halothane-induced hepatitis

6 (19) and clozapine-induced agranulocytosis (20), etc. In search of genetic risk factors for IDRs, establishing an association between polymorphisms of drug metabolism and the basis of idiosyncrasy has usually been unsuccessful. However, certain phenotypes of human leukocyte antigen (HLA) have been found to be associated with an enhanced likelihood of IDRs, such as the strong association between HLA-B*1502 and carbamazepine-induced Stevens-Johnson syndrome (21-23), the strong association between HLA-B*5701 and abacavir-induced hypersensitivity reactions (24-27) and also flucloxacillin-induced liver injury (28), and a strong association between HLA-B*3505 and nevirapine-induced skin rash in a Thai population (HLA-B*3505 was observed in 17.5 % of the patients with nevirapine-induced skin rash while only 1.1% of nevirapine-tolerant patients) (29), etc. Another very important risk factor for IDRs is the disease state of patients. Studies found that some infectious diseases such as HIV infections (30) and liver diseases (31), appear to increase the risk of IDRs.

1.3. WORKING HYPOTHESIS OF MECHANISMS OF IDRS

All major current working hypotheses of the mechanisms of IDRs have an immune basis. Therefore, a good comprehension of the immune system is essential in understanding IDRs. An overall review of immunological models is presented as follows.

1.3.1. OVERVIEW OF IMMUNOLOGICAL MODELS

Survival is a vital issue for any form of life. Over the long span of evolution, biological systems have evolved a set of elaborate, dynamic, and well-regulated machinery called the immune system to closely guard themselves and defend against any pathogen that could potentially damage it. An in-depth understanding of this sophisticated system helps to provide a solid basis for dealing with a wide range of immune-related problems that influence virtually all areas of medicine. Rigorous studies in the past few decades have significantly expanded our knowledge of the immune system; however, it has become routine for new data to overthrow longstanding concepts and there remain many unknowns.

7 1.3.1.1. SELF-NONSELF MODELS The first immunological model to address the specificity of the immune system, known as the self-nonself (SNS) model, was proposed by Burnet in 1959 (32, 33). Ever since, it has been widely accepted as one of the most fundamental theories of modern immunology. Simply stated, it suggested that an immune response is mediated by lymphocyte surface receptors specific to foreign substances, and negative selection is programmed early in life to delete self-reactive lymphocytes to differentiate self from nonself. The key principle of the SNS model is that the exclusive determinant of what the immune system responds to is the recognition of nonself by immune cells. The SNS model was accepted until immunologists began to realize that T cell responses depend on a second activation signal delivered by other cells known as antigen presenting cells (APCs). Several major modifications were made to the original SNS model eventually resulting in the birth of Janeway’s infectious-nonself (INS) model in 1989 (34). The two signals required for T cell activation were defined as signal 1, which consists of processed antigen presented by major histocompatibility complex II molecules (MHC II) on APCs to T cell receptors (TCRs), and signal 2, which consists of costimulatory interaction between B7 molecules of APCs and CD28 of T cells, respectively. The gate-keeping step suggested by the INS model is the recognition of a particular pathogen-associated molecular pattern (PAMP) on pathogens by pathogen recognition receptors (PRRs) on APCs. This recognition activates APCs and up-regulates their surface expression of B7 and other costimulatory molecules. On receiving two signals from APCs, T cells are activated and differentiate into specific types of helper cells to facilitate either cell-mediated (Th1 or Th17) or antibody-mediated (Th2) immune reactions. Each pathway is characterized by the cytokines and chemokines that are released. Although Janeway proposed that it is PRRs on APCs instead of lymphocytes that discriminate between self and nonself, both the SNS and INS models are based on the recognition of foreignness. Over more than 50 years, the self-nonself concept dominated immunology. It is true that lymphocytes with a high affinity for self-molecules are deleted in the thymus making it more difficult to mount a strong immune response

8 against self-molecules. However, this hypothesis does not address several other issues such as what causes autoimmunity and why there is no immune response to tumors even though they express neoantigens, etc. Thus, further refinement of this concept was needed.

1.3.1.2. DANGER MODEL In 1994, Polly Matzinger proposed the danger model that posits it is cell damage rather than nonself that determines whether an immune response will occur (35). Injured cells (i.e. stress, necrosis) release danger/alarm signals (damage-associated molecular pattern, DAMP) that activate APCs resulting in increased expression of costimulatory molecules. The danger signals are also referred to as signal 3. According to this model, the immune system is more concerned with potential danger than foreignness. This can explain why a wide variety of nonself exposures do not trigger an immune response in the absence of significant cell damage. In addition, the danger model offers an explanation of how endogenous molecules can induce immune reactions. Therefore, independent of whether a molecule is an exogenous pathogen, chemical, or endogenous intracellular molecule released from necrotic cells, they all must cause damage or cell stress in order to elicit an immune response. Although the danger model is difficult to rigorously test, and it was quite controversial at first, it now appears to have become part of accepted immune theory. Figure 1 illustrates the progression of immunological models from the original SNS model to the current danger model.

Figure 2. Progression of immunological models.

10 1.3.2. WORKING HYPOTHESIS OF IDRs

At present, there are three major working hypotheses proposed to explain the interaction between drugs and/or reactive metabolites and the immune system causing pathogenic immune reactions: the hapten hypothesis, the danger hypothesis, and the pharmacological interaction hypothesis (11). Figure 3 is a schematic description of hapten and danger hypothesis. Although suggesting different triggering events, all three hypotheses center on an immunological mechanism. They are not mutually exclusive, and one or more might be useful in explaining a specific IDR.

1.3.2.1. HAPTEN HYPOTHESIS A basic principle of immunology postulated by Landsteiner over 70 years ago (Landsteiner and Jacobs 1935) is that small molecules with a molecular mass of less than 1000 Daltons are unable to induce an immune response unless they are bound to a macromolecule such as a protein. The term given to a small molecule that leads to an immune response after binding to a macromolecule is hapten. This provides a good explanation for the allergic reactions caused by penicillin and other ß-lactam antibiotics. The ß-lactam ring is reactive and penicillin binds to proteins. Many of the allergic reactions associated with penicillin are mediated by IgE antibodies against penicillin-modified proteins; thus, penicillin is acting as a hapten (36, 37). Although there are several other examples, such as halothane-induced hepatitis in which a reactive metabolite covalently binds to proteins and induces antibodies against the metabolite-modified proteins (38), it is not clear that these antibodies mediate the liver damage. There are few other examples where the covalent binding of a drug so clearly causes an IDR as in the case of penicillin.

1.3.2.2. DANGER HYPOTHESIS If the danger model is correct, simply binding to proteins to make them foreign would not be sufficient to induce an immune response (11, 39-41). In addition, it would require the activation of the immune system by damaged/stressed cells, which is mediated

11 by proteins or other molecules acting as danger signals by binding to certain receptors on innate immune cells such as macrophages (42, 43). This hypothesis could explain why many drugs that form reactive metabolites and covalently bind to proteins are not associated with a significant incidence of IDRs. It may be that the drug, or more likely a reactive metabolite, must also cause cell damage in order to cause IDRs. A follow up question is whether the danger signal must come from the drug or whether other sources of tissue injury such as infection, surgery, or other inflammatory conditions act as risk factors for IDRs. In the past few years, several studies have been done that have implications for the danger hypothesis and IDRs: 1) Identification of potential danger/alarm signals released from cells or tissues, 2) Investigation of the correlation between danger molecules and the induction of IDRs.

1.3.2.3. PHARMACOLOGICAL INTERACTION (PI) HYPOTHESIS Based on the finding that T cells from patients with a history of an IDR to a specific drug (e.g. sulfamethoxazole, lidocaine) proliferated in response to the drug involved in the IDR in the absence of metabolism (44), Pichler proposed that nonreactive drugs reversibly bind to the complex of MHC-T cell receptor, much like a superantigen, and this interaction can stimulate an immune response leading to an IDR. He named this the pharmacological interaction (PI) hypothesis (45). It is a fairly new hypothesis and is being tested in animal models in our lab. The essential assumption of the PI hypothesis is T cells only respond to whatever caused the IDR. Our most recent study has demonstrated that the 12- hydroxylation pathway of nevirapine is responsible for nevirapine-induced skin rash in BN rats (46). However in the lymphocyte transformation test (LTT), T cells isolated from sick animals responded to parent drug much more than 12-hydroxynevirapine (unpublished data), even when the rash was induced by treatment with 12-hydroxynevirapine and the animal had never been exposed to nevirapine. This strongly argues against the general application of the PI hypothesis in explaining the pathogenesis of most IDRs.

Figure 3. Modes of the interaction between drugs and immune system. (1). Complex of drug/reactive metabolite and endogenous proteins as hapten. (2). Cellular injury (generation of danger signals). RM: reactive metabolite.

13 1.4. ANIMAL MODELS OF IDRs

Due to the idiosyncratic nature of adverse reactions, it is practically impossible to carry out prospective studies in humans to investigate the mechanisms of IDRs. Also, since most IDRs appear to involve the immune system, it is naïve to believe that it is possible to mimic the complexity of the complete immune system in an in vitro system. In most biomedicine research, animal models represent a very important tool for mechanistic studies and this is also true for understanding IDRs. However, not many good animals models are currently available for testing mechanistic hypotheses of IDRs due to either a very low incidence of the adverse reaction and the practical difficulty of performing experiments on some species of animals (47) or because the IDRs do not reflect the IDRs that occur in humans. According to a statement by Scarpelli, “The usefulness of an animal model depends on how closely it resembles the disease or condition to which it is compared” (48). The animal models of penicillamine-induced autoimmunity in BN rats and nevirapine- induced skin rash in BN rats currently represent the best models for the mechanistic studies of IDRs because clinical symptoms developed in rats closely mimic those in humans; however, they represent a limited spectrum of IDRs.

1.4.1. NEVIRAPINE-INDUCED SKIN RASH IN BN RATS

Known as a non-nucleoside reverse transcriptase inhibitor used in the treatment of human immunodeficiency virus (HIV) infections, nevirapine causes skin rashes or liver toxicity in 9-16 % and 2.8 % of patients, respectively. When given nevirapine at a dose of 150 mg/kg/day, all female BN rats develop significant skin lesions after 2-4 weeks of treatment with the redness of ears occurring around day 7-10 of dosing. The reactions appear to be strain specific in that the incidence of skin rash in female Sprague-Dawley rats, female Lewis rats, male Sprague-Dawley rats, and male BN rats is 21%, 0%, 0%, and 0%, respectively. In addition, the characteristics of the reactions observed in female BN rats are very similar to those in humans, which strongly suggests that this animal model is a very good model for our mechanistic exploration of IDRs (49). Further studies have shown that

14 this reaction is immune-mediated because: 1). A delay in onset on primary exposure; 2). A rapid onset in less than 24 h of treatment on rechallenge; 3). Susceptibility transferable to naïve animals by spleen cells of rechallenged animals. Meanwhile, a more recent study in our lab has pinpointed that one metabolism pathway, 12-hydroxylation, which involves the oxidation of an exocyclic methyl group, is responsible for the observed skin rashes (46).

1.4.2. PENICILLAMINE-INDUCED AUTOIMMUNITY IN BN RATS

Since L-penicillamine is easily incorporated into proteins and has high reactivity with vitamin B6 (50, 51), the D isomer is the form commonly used in medicine so when the term penicillamine is used in this thesis it refers to the D-penicillamine. Due to its three major chemical behaviors: 1) formation of disulfide bonds; 2) formation of thiazolidine rings; and 3) formation of metal chelates (50), penicillamine is active in treatment of a variety of human diseases such as Wilson disease, cystinuria, rheumatoid arthritis (RA), palindromic rheumatism, scleroderma, primary biliary cirrhosis, heavy metal removal, morphea, keloid, keratosis follicularis, and hyperviscosity syndrome etc. (52). However, many adverse reactions have been reported to be associated with its clinical use, with the majority being autoimmune disorders such as a lupus-like syndrome, myasthenia gravis, membranous glomerulopathy, and pemphigus, etc. These autoimmune reactions appear to be idiosyncratic in nature in that many patients taking penicillamine were found to develop antinuclear antibodies while only a very small percentage of them eventually progressed to clinically evident symptoms. In parallel, studies showed that penicillamine induced the production of anti-nuclear antibodies in several animal strains such as A-SW mice and BN rats etc., but only BN rats developed evident signs of autoimmunity. Interestingly, autoimmune manifestations caused by penicillamine in BN rats were found to be very similar to those of mercury chloride-induced autoimmunity. In past a decade, there have been extensive studies done in our lab to study the penicillamine model and this has furthered our understanding of IDRs even though the exact mechanism is still unknown. In general, when given at more than 20 mg/day, penicillamine causes an autoimmune syndrome in 50-80 % of BN rats (weight ~200 g). There is a delay of 2-3 weeks between

15 starting the treatment and onset of the signature sign that the animal will develop autoimmunity, red ears. Ultimately the animals progress within a week to a more serious state of autoimmunity characterized by clinical features of swollen, red, and arthritic limbs, skin lesions, and rapid weight loss. There is no change in the time delay in onset on rechallenge. In addition, other changes of pathology and serology have been identified including splenomegaly, polyclonal B cell activation, autoreactive T cells, and a rapid rise in IgE serum levels with a concomitant increase in IL-4 mRNA expression. In contrast, Sprague-Dawley and Lewis rats are resistant to this autoimmune syndrome. Continuously increasing the dose from 20 mg/day to 50 mg/day does not significantly change the incidence in BN rats. However, treatment of penicillamine at doses lower than 20 mg/day dramatically decreases the incidence; for example, the incidence is 0% at a dose of 5 mg/day. Interestingly, pretreatment at a low dose of 5 mg/day for 2 weeks completely prevented autoimmunity caused by a subsequent high dose penicillamine treatment (20 mg/day). The tolerance induced by low dose treatment has been shown to be immune tolerance because it can be transferred through splenocytes. Moreover, our previous studies found that a number of immunomodulators are able to modify both the incidence and severity of autoimmunity caused by penicillamine in BN rats. In search of the initial events in pathogenesis of this model, we found a significant infiltration of macrophages in the gut and other tissues shortly after treatment (96 hours). The increase of activated macrophages in the spleen appears to be regulated by immunomodulators. Also, depletion of macrophages by clodronate liposomes significantly decreased the incidence of autoimmunity.

Figure 4. Dose dependence of penicillamine-induced autoimmunity in BN rats.

Table 3. Influence of immunomodulators on penicillamine-induced autoimmunity Time to Macrophage Type of modulation Incidence Severity onset infiltration A single dose i.p. injection of Poly I:C 100% ↓ ↑ ↑ A single dose i.p. injection of ketoprofen 100% ↓ ↑ ↑ A 2-week pretreatment with penicillamine of 5 mg/mL 0% - - ↓ A single dose s.c. injection of misoprostol 0% - - ↓ A single dose i.p. injection of aminoguanidine 0% - - ↓ Daily i.m. injection of tacrolimus for 2 weeks 0% - - ↓

17 1.5. AUTOIMMUNITY AND IMMUNE TOLERANCE

A challenging question that immunologists have always faced is what allows the immune system to attack self-antigens leading to autoimmune diseases. First of all, there is a very important conceptual difference between autoimmunity and pathogenic autoimmune reactions. Back in the early 1900s, autoimmunity was considered all bad and defined as “horror autotoxicus” by Nobel Laureate Paul Ehrlich. Detection of any amount of autoimmunity in a healthy person would be indicative of potential abnormalities. However, over time, it was found that there are always some autoreactive T and B cells in peripheral blood, and most of time, they are inhibited from causing any problems by a precise vigilance system (53). This means autoimmunity is not always harmful until an organism loses control over these autoreactive cells and they become pathogenic. A good balance between autoimmunity and immune tolerance is of great importance in maintaining health because if the immune system eliminated all autoreactive cells, there would be no protection against pathogens that have some cross reactivity with autoantigens. Nevertheless, this balance is always being challenged by many biological events such as pathogens, medication, aging, etc. Extensive studies of autoimmune diseases have provided significant insights into the cellular and genetic mechanisms of autoimmunity; however, many unknowns persist such as why some autoimmunity is organ-specific, etc (54).

1.5.1. MECHANISMS OF AUTOIMMUNE DIEASES

Despite the many puzzles in the basic mechanisms of autoimmune disorders that remain, it has been widely recognized that a breakdown in the balance between autoimmunity and self-tolerance results in the onset of autoimmune reactions eventually progressing to disease states. Although the innate immune system is also involved, activation of self-reactive T cells and reactivation of self-reactive B cells are central to the pathogenesis of autoimmune diseases. The following provides a general overview of the major theories underlying the mechanisms of autoreactive lymphocyte activation. Release of sequestered antigens Many endogenous molecules such as constitutive

18 proteins of organelles in the cytoplasm and nucleus are normally hidden from the immune effector cells. Therefore, deletion of self-reactive T cell clones specific for these molecules in thymus does not occur. However, some physiological events (i.e. apoptosis and necrosis) could potentially release these novel autoantigens in greater quantities than phagocytic cells can clean up and thus lead to autoimmune reactions (55). Modification Some small reactive molecules, such as drugs or their reactive metabolites, can covalently bind to endogenous proteins, which make these self-proteins neoantigens that can be recognized by the immune system as non-self (56). Molecular mimicry Some endogenous molecules may share structural similarities with exogenous antigens. Hence, in theory, any effector cells or immunoglobulins produced by immune system primarily against exogenous antigens could also attack host antigens to initiate autoimmune reactions (57). Epitope spreading An epitope is an antigenic determinant, or a site, on the surface of an antigenic macromolecule that is specifically recognized by antibodies or effector T cells. Epitope spreading occurs when the immune reaction expands from its primary epitope to other epitopes by unknown mechanisms, which in the case of autoimmunity, involves self-molecules. There is increasing evidence to demonstrate this phenomenon in the pathogenesis of autoimmune diseases (58, 59). Meanwhile, autoreactive B cells could be involved in the pathogenesis of autoimmune reactivity in many ways: production of harmful autoantibodies, formation of immune complexes, and the release of cytokines and chemokines that are critical for autoreactive T cell growth (60, 61). Studies have shown that abnormal B cell activity is associated with many different kinds of autoimmune disorders, and thus B cell depletion is one of the major therapies for autoimmune diseases (62-64). Also, immunogenetic studies have identified several genetic risk factors for autoimmune diseases (65). The presence of sequence variants of several genes is associated with an increased susceptibility to pathogenic autoimmunity, e.g. CTLA-4 in type 1 diabetes (66), C4 complement in SLE (67), and FOXP3 in IPEX syndrome (68).

19 1.5.2. IMMUNOLOGICAL TOLERANCE TO SELF TISSUES

The reason that there are no pathogenic autoimmune reactions most of time is because our immune system “chooses” to ignore self tissues or neutralize self-reactive lymphocytes to avoid self-attack, which is known as immunological tolerance or self tolerance (69). Loss of this tolerance has been reported in a number of autoimmune diseases. At present, there are two major mechanistic hypotheses for the genesis of immunological tolerance: clonal deletion and clonal anergy (70). Clonal deletion This is the first step that immune system takes to generate tolerance at the T-cell level during maturation of T cells in thymus. At this stage, autoreactive T cells with a high affinity for self-antigens are detected and deleted via programmed cell death. Any mistake in this selection process would lead to the escape of autoreactive T cells into peripheral circulation with a potential to trigger an autoimmune reaction. Thus, the immune system has evolved another delicate and specific regulatory machinery to constantly check for escaped autoreactive T cells so that they do not cause damage to self-tissues (71). This is usually generalized as the theory of clonal anergy. Meanwhile, at the B-cell level, the central tolerance takes place in bone marrow where the production of harmful autoantibodies is prevented by three mechanisms: receptor editing, deletion, and anergy (61). Clonal anergy and maintenance of peripheral self-tolerance There is a wide spectrum of checkpoints at which autoreactive T cells are kept inactivated in the periphery. First of all, when there is no costimulatory signal (B7.1 or B7.2) sent to CD28 molecules on autoreactive T cells from APC during their interaction, it leads to anergy of these T cells instead of activation. Or if B7.1 or B7.2 binds to CTLA-4 instead of CD28 on T cells, it will also lead to inhibition of IL-2 production and anergy (72). These two situations together are called clonal anergy. More recently, CTLA-4 has also been suggested to be involved in the systemic tolerance mediated by regulator T cells (73). Second, even though autoreactive T cells escape clonal deletion in the thymus and migrate to the periphery, they may be sequestered in some tissues where they cannot access the original self-antigens to which they are reactive. Therefore, without encountering appropriate antigens, these autoreactive

20 T cells will eventually die because of the lack of a stimulus, which is a relatively new theory called clonal ignorance. The third theory is centered on regulatory T cells (Tregs), which have been extensively studied and demonstrated to be essential effector cells for maintaining peripheral tolerance to self-antigens. Identification of a key transcription factor, forehead box P3 (FOXP3), for Tregs significantly advanced our understanding of Tregs (74, 75). Despite some phenotypic differences between human and mouse Tregs, CD4+CD25+FOXP3+ cells are widely accepted as thymus-derived, naturally occurring Tregs. Three major modes of Treg-mediated immunological suppression are (76, 77): 1). Production of inhibitory cytokines such as TGF-β, IL-10, and IL-35 etc; 2). Inhibition of dendritic cell maturation and normal function via direct cell-cell contact; 3). Killing of responder T cells or APCs by releasing granzyme-A,B or perforin in cell-cell contact. In addition, several immunosuppressive molecules have been recently identified to prevent an immune response; for example, indoleamine 2,3-dioxygenase (IDO). IDO is mainly produced by APCs (i.e. activated macrophages) and it is known to be a key enzyme in the catabolism of tryptophan, an essential amino acid for T cell growth (78-80). Therefore, depletion of tryptophan by IDO can potentially inhibit the proliferation of T cells in response to stimuli. Inhibition of IDO activity by 1-methyl tryptophan has been shown to be able to modulate the incidence and severity of autoimmunity in several animal models such as EAE (81). Meanwhile, several checkpoints have been developed to minimize the autoreactivity of B cells in the periphery (82, 83).

1.5.3. DRUG-INDUCED AUTOIMMUNITY

The first case of drug-induced lupus caused by sulfadiazine was reported in 1945. Since then more than 80 drugs (e.g. procainamide, hydralazine, isoniazid, minocycline, methyldopa) have been implicated in the induction of autoimmune disorders, particularly the development of a lupus-like syndrome (84). The most frequent clinical features include arthralgias, myalgias, arthritis, serositis, hepatosplenomegaly, skin rash, and weight loss (14). In addition, drug-induced autoimmunity is always associated with positivity of serum autoantibodies, primarily consisting of antinuclear antibodies (ANA), and antihistone

21 antibodies, antineutrophil cytoplasmic antibodies (p-ANCA) (85). However, many patients develop autoantibodies, while only a minority develop clinical signs that require discontinuing the medication. Thus, development of autoantibodies in the absence of clinical features is not sufficient to make a diagnosis of drug-induced autoimmunity. Although the pathogenic mechanisms of drug-induced autoimmunity are still unknown, multiple working hypotheses have been used to study the pathogenesis of drug-induced autoimmunity; these hypotheses are not mutually exclusive (85, 86): 1). A drug or its reactive metabolite covalently binds to endogenous proteins and consequently makes the self-proteins look foreign and antigenic. As a result, an immune response against the altered self-proteins is elicited and leads to an autoimmune-like syndrome; 2). DNA methylation is very important for T-cell function. A growing body of evidence showed that failure to maintain normal DNA methylation in mature T cells could generate autoreactivity in T cells and lead to autoimmune disorders. One of the most common causes of defective DNA methylation is drug treatment such as procainamide (a competitive inhibitor of DNA methyl transferase) and hydralazine (inhibition of DNA methyl transferase expression by decreasing extracellular signal-regulated kinase) (87, 88). DNA hypomethylation of T cells changes the expression of many molecules, for instance, it results in the over expression of several adhesion molecules (i.e. lymphocyte function- associated antigen-1, LFA-1) that have been shown to play an essential role in pathogenesis of autoimmunity (89); 3). As mentioned in general mechanisms of autoimmunity, self-antigens released from apoptotic or necrotic cells have the potential to induce immune reactions because they are normally well hidden and have not been encountered by T cells before. Although most parent drugs are quite inert, their metabolites could be very reactive and cause significant cytotoxicity. Many different kinds of lupus-inducing drugs have been found to form reactive metabolites that certainly could exhibit cytotoxicity, which could potentially explain cytopenias, which are a very common feature in drug-induced autoimmunity (14, 90);

22 4). Breaking self-tolerance by reactive metabolites is another attractive hypothesis of drug- induced autoimmunity. Studies have found that the injection of the reactive metabolite of procainamide, procainamide hydroxylamine, directly into the thymus of mice impaired central T cell tolerance by interfering with the induction of tolerance to self-antigens (91).

Table 4. Examples of lupus-inducing drugs

Medications Therapeutic category Clinical features ∗ Risk Infliximab TNF inhibitors Arthralgia Very low Interferon-α Biologicals Myalgia Very low Interleukin-2 Biologicals Arthritis Very low Isoniazid Antibiotics Fever Low Hydralazine Antihypertensives Malaise High Methyldopa Antihypertensives Anorexia Low Minocycline Antibiotics Hypertension Low D-penicillamine Anti-inflammatories Weight loss Low Procainamide Antiarrhythmics Pleuritis High Pericarditis Simvastatin Anticholesterolemics Very low Hepatosplenomegaly ∗ Clinical abnormalities are usually milder than those in idiopathic SLE. Initial symptoms are usually mild and gradually worsen over a period of weeks or even months. Generally, the symptoms recede after the discontinuation of the medication. Clinical manifestations of lupus caused by each drug in this table are a mixture of all features listed above.

23 1.5.4. PROTEIN CARBONYLATION IN AUTOIMMUNE DISEASES Generation of reactive oxygen species (ROS) is a very common event occurring in many biological processes (92, 93). Under normal physiological conditions, ROS are maintained at a low level because they could be very toxic and cause significant cellular damage. Overproduction of ROS or defects in ROS-scavenging system would be pathogenic and lead to many disease states such as Alzheimer’s disease (AD) and autoimmune disorders, etc (94, 95). Although ROS could affect various cellular components, proteins are often the primary targets of ROS, potentially leading to altered protein structure and modulation of its biological functions (96). In particular, carbonylation of proteins by low molecular weight reactive carbonyl or aldehyde species (RCS), which are mainly produced during lipid peroxidation (i.e. 4-hydroxynonenal, acrolein, and glyoxal) and glycolysis process (i.e. methylglyoxal) causes irreversible oxidative damage and results in dysfunction of the affected proteins (97, 98). Compared to most ROS and oxidizing intermediates, RCS are relatively long-lived (99). They are the basis for the introduction of carbonyl groups into proteins to generate reactive carbonyl derivatives that are known as advanced lipoxidation end products (ALEs) and advanced glycation end products (AGEs). Figure 5 is the summary of formation of AGEs. Studies have shown that ALEs and AGEs are quite immunogenic and have the potential to trigger unwanted immune reactions (100). Increased levels of protein carbonylation have been found in many inflammatory diseases such as diabetes, juvenile rheumatoid arthritis, and chronic renal failure etc.; therefore, the presence of protein carbonylation has been used as a marker of inflammatory activity (101- 103). In addition to reacting with RCS, the side chains of lysine, arginine, proline, and threonine can be directly oxidized by ROS to carbonylated forms (104). Due to the substantial impact of carbonylation on protein functions and normal physiological activities, the generation of RCS is closely monitored and cleaned up by a wide spectrum of endogenous scavengers such as actin filaments (105-107). As a result, a certain level of carbonylated proteins, especially membrane proteins (i.e. membrane-associated actin) always exists under normal physiological condition. In theory, these carbonylated proteins

24 represent potential targets for drugs such as penicillamine to interact with, which have a potential to lead to an IDR.

Figure 5. Advanced glycation end products. Adapted from Wautier J and Schmidt A (100).

25 1.6. DRUG-INDUCED LIVER INJURY Everything absorbed from the intestine has to go through the liver, and being the predominant site of biotransformation, the liver is under constant challenge by a wide variety of agents including drugs. Hence, drugs affect the liver more frequently than any other organ, and drugs are a very important and common cause of hepatic injury (108). Epidemiological studies have shown that drug-induced liver injury (DILI) is responsible for more than 50% of acute hepatitis, with 39% due to overdoses of acetaminophen and 13% being idiosyncratic reactions (IDILI) caused by other drugs (109). To date, more than 1000 drugs have been reported to cause hepatotoxicity, which has led to the withdrawal of some drugs from the market or “black box” warnings for others (31, 110, 111). DILI represents a serious problem in the clinic because about 1 in 100 patients receiving medication develops DILI during hospitalization (112). In addition, DILI presents a big challenge to the pharmaceutical industry in that it is one of the most frequent causes of drug candidate failure. Clinical manifestations of DILI include a wide spectrum of abnormalities from minor nonspecific derangements to fulminant hepatic necrosis. The two most common types of liver injury are hepatic necrosis (hepatitis) characterized by an increase in ALT and cholestasis characterized by an increase of alkaline phosphatase and bilirubin (113). These injuries have been extensively investigated in the framework of several major hypotheses of DILI pathogenesis as follows (114): Drug-induced hepatitis Hepatocellular death is an essential clinical feature of drug-induced hepatitis, which can be mediated through either apoptosis or necrosis. Apoptotic cell death o Interaction of death ligands and their receptors at the surface of hepatocytes (i.e. TNF/TNF-R1, FasL/Fas), which is immune system-dependent because TNF-R1 and Fas are activated and released by the innate and adaptive immune system, respectively. o Direct cytotoxicity caused by drugs or their reactive metabolites. Necrotic cell death

26 o Profound loss of mitochondrial function with ATP depletion, loss of ion homeostasis, and necrotic cell lysis caused by severe oxidative stress. Drug-induced cholestasis Unlike drug-induced death of hepatocytes, cholestasis refers to bile duct injury or inflammation caused by drugs such as rifampicin and cyclosporine that inhibit the bile salt excretory proteins (115). At present, it is still very difficult to predict the relative risk of DILI in the early phase of new drug development because of the lack of in-depth understanding of the mechanisms involved. Like most IDRs, DILI is associated with risk factors of advanced age, female sex, other illness, environmental factors, and genetic predisposition (10). A very recent study showed a strong association between HLA-B*5701 and the risk of DILI caused by flucloxacillin (80-fold) (28), which significantly advanced our understanding of genetic susceptibility in DILI, at least to this one drug. Based on their clinical manifestations, idiosyncratic DILI are usually divided into two types: metabolic idiosyncrasy and immune idiosyncrasy (8, 111). The presence of fever, skin rash, eosinophilia, anti-drug antibodies, and rapid onset on rechallenge are traditionally used to diagnose immune-mediated DILI. In contrast, DILI cases that lack these signs are then considered to be metabolic idiosyncrasy. However, no genetic polymorphism of a drug-metabolizing enzyme or any other metabolic pathway has ever been shown to be responsible for the idiosyncratic nature of DILI. Some cases of DILI do not occur more rapidly on rechallenge, and this observation has been the strongest argument against an immune-mediated mechanism; however, rapid onset on rechallenge is not always observed in cases of immune-mediated reactions (8). This is particularly true in some cases of drug-induced autoimmunity. For example, by definition penicillamine-induced autoimmunity in BN rats is immune-mediated, but the autoimmune syndrome occurs with the same time delay on rechallenge. A reasonable explanation for this observation is that after the discontinuation of the offending drug, the immune system takes immediate action to delete or inhibit the autoreactive T cells that were responsible for the autoimmune reaction. As mentioned earlier, drug-induced autoimmunity usually resolves rapidly after the causal drug is stopped even though the immune response is against an

27 autoantigen. This would lead to a similar time and process to develop sufficient autoreactive T cell clones to cause autoimmune disorders on the rechallenge (116). In fact, there are a number of drugs that have been shown to be associated with both the induction of idiosyncratic liver injury and an autoimmune syndrome (Table 5).

Table 5. Drugs that are associated with both IDILI and autoimmunity Medications Therapeutic category Refs Isoniazid Antibiotics (117) Minocycline Antibiotics (118) α-methyldopa Antihypertensives (119, 120) Hydralazine Antihypertensives (120, 121) Nitrofurantoin Antibiotics (122, 123) Propylthiouracil Antithyroid (124, 125) Methimazole Antithyroid (126, 127) Aminoglutethimide Antisteroid (128, 129) Diclofenac Anti-inflammatories (130, 131) Allopurinol Antihyperuricemia (132, 133) Phenylbutazone Anti-inflammatories (134, 135) Phenytoin Antiepileptic (136, 137) Carbamazepine Anticonvulsant (138, 139) Sulfonamides Antibiotics (140) Phenothiazines Antipsychotic (141, 142) Terbinafine Antifungal (143, 144) Statins Anticholesterolemics (145, 146) Leflunomide Anti-inflammatories (147, 148) Zafirlukast Anti-inflammatories (149-151) Adapted from Uetrecht J (116).

Figure 6. Mechanistic hypotheses of pathogenesis of DILI. Modified from Abboud G and Kaplowitz N (114). RM: reactive metabolite. A drug or its reactive metabolite induces cellular stress leading to apoptosis or necrosis that is probably due to the increased permeabilisation of the mitochondrion. Cell death in turn leads to the recruitment of members of innate immune system such as natural killer cells, natural killer T cells and neutrophils. Participation of the innate immune system augments cellular death and results in severe liver injury. Meanwhile, a drug or its reactive metabolite may also act as a hapten and stimulate the adaptive immune response, leading to the engagement of death receptors and apoptosis, thereby contributing to severe liver injury.

CHAPTER 2

COVALENT BINDING OF PENICILLAMINE TO MACROPHAGES: IMPLICATIONS FOR PENICILLAMINE- INDUCED AUTOIMMUNITY

Reproduced with permission from Jinze Li, Baskar Mannargudi, and Jack P. Uetrecht Chemical Research in Toxicology 2009 July; 22(7): 1277-84.

30 2.1. Abstract

Idiosyncratic drug reactions (IDRs) represent a major clinical problem, and at present, the mechanisms involved are still poorly understood. One animal model that we have used for mechanistic studies of IDRs is penicillamine-induced autoimmunity in Brown Norway (BN) rats. Previous work in our lab found that macrophage activation preceded the clinical autoimmune syndrome. It is thought that one of the interactions between T cells and macrophages involves reversible Schiff base formation between an amine on T cells and an aldehyde on macrophages, but the identity of the molecules involved is unknown. It is also known that penicillamine reacts with aldehyde groups to form a thiazolidine ring, which unlike a Schiff base, is essentially irreversible. Such binding could lead to macrophage activation. Generalized macrophage activation could lead to the observed autoimmune reaction. Hydralazine and isoniazid also react with aldehydes to form stable hydrazones, and they also cause an autoimmune lupuslike syndrome. In this study, isolated spleen cells from male BN rats were incubated with biotin-aldehyde-reactive probe (ARP, a hydroxylamine), biotin-hydrazide, or D-penicillamine. At all concentrations, ARP, hydrazide, and penicillamine preferentially “stained” macrophages relative to other spleen cells. In addition, preincubation of cells with penicillamine or hydralazine decreased ARP staining of macrophages, which further indicates that most of the ARP binding to macrophages involves binding to aldehyde groups. This provides support for the hypothesis that the interaction between aldehyde-containing signaling molecules on macrophages and penicillamine could be the initial event of penicillamine-induced autoimmunity. Several of the proteins to which ARP binds were identified, and some such as moesin are attractive candidates to mediate macrophage activation.

31 2.2. Introduction

Idiosyncratic drug reactions (IDRs) refer to a group of adverse drug reactions that do not occur in most patients within the therapeutic dose range and cannot be explained by the known pharmacological properties of the responsible drug (152). IDRs can be severe, even life-threatening, and therefore represent a significant clinical problem. At present, it is impossible to predict which drug candidates will cause IDRs or which patients are at greatest risk for such reactions, largely because the mechanisms involved are unknown. Therefore, they also add a significant degree of uncertainty to new drug development. Nevertheless, clinical manifestations such as the delay between starting the drug and the onset of the adverse reactions suggest that most IDRs are immune-mediated (11). Therefore, an important goal is to understand how drugs can induce an immune response. Animal models represent a very powerful tool for mechanistic studies in virtually all fields of biomedical research. One animal model that we have used for mechanistic studies of IDRs is D-penicillamine-induced autoimmunity in Brown Norway (BN) rats, which mirrors the variety of autoimmune reactions that it causes in humans (153, 154). Penicillamine-induced autoimmunity in rats is also idiosyncratic because it is strain-specific: Lewis and Sprague-Dawley rats are tolerant to the same dose that causes autoimmunity in BN rats (154). Moreover, even though BN rats are highly inbred and syngeneic, autoimmunity only occurs in a little over 50% of treated rats. Therefore, penicillamine- induced autoimmunity in BN rats provides a very important model for mechanistic studies of at least one type of IDR. By definition, penicillamine-induced autoimmunity is an immune-mediated adverse reaction. Previous work in our lab found that macrophage activation and infiltration occurred as early as 96 h after penicillamine treatment even though the autoimmunity does not become clinically apparent for 3 weeks (155). In addition, different immunomodulators that modify the incidence and severity of the disease have a parallel effect on infiltration of activated macrophages. Moreover, partial depletion of macrophages decreased the incidence

32 of autoimmunity in BN rats. These results suggested an important role of macrophages in the initiation of penicillamine-induced autoimmunity. Studies have shown that pretreatment of rat peritoneal macrophages with penicillamine can enhance their ability to modulate the lymphocyte response to specific and nonspecific stimuli (156-158). There is also evidence that one of the interactions between macrophages and lymphocytes involves a covalent but rapidly hydrolyzed imine bond (Schiff base) between an aldehyde group on macrophages and an amino group on T cells as shown in Figure 7 (159-162). (Note: Although it was claimed that the Schiff base was formed between an amino group and an aldehyde group, the chemistry of ketones is similar although they are less reactive. Therefore, when we refer to an aldehyde, it is implied that a ketone is also a possibility.) Moreover, it was reported that treatment of macrophages with sodium metaperiodate or sequential neuraminidase and galactose oxidase to generate aldehydes markedly enhanced their binding to lymphocytes and led to increased macrophage-dependent T lymphocyte activation and proliferation (160, 161, 163-165). However, these aldehyde-bearing membrane proteins on macrophages have not been identified. It is well known that penicillamine reacts with aldehyde groups to form a thiazolidine ring (Figure 7), which is more stable than a Schiff base (166). If, as it has been proposed, there are aldehyde groups on the cell membrane of macrophages, it is likely that penicillamine would react with these aldehydes and this reaction could potentially modulate functions of macrophages (86), including becoming primed or fully activated. Generalized activation of macrophages has the potential to lead to autoimmunity. In this study, we used a commercial aldehyde reactive probe (ARP, a biotin-hydroxylamine) and a biotin- penicillamine (which we synthesized) to test the presence of aldehyde groups on macrophages and the chemical interaction between penicillamine and macrophages. In addition, using avidin-biotin chromatography and mass spectrometry, we identified aldehyde-containing proteins to which ARP and penicillamine bind.

Figure 7. Binding of penicillamine and ARP to aldehydes on the surface of macrophages.

34 2.3. Materials and Methods

Animals. Male rats (150-175 g) were purchased from Charles River (Montreal, Quebec) and doubly housed in standard cages in a 12:12 h light:dark cycle at 22 °C. The rats were given free access to standard rat chow (Agribrands, Purina Canada, Strathroy, Ontario) and water for a weeklong acclimatization period before starting experiments. Analytical. NMR was performed with a Varian 300 MHz spectrometer. Mass spectra were performed using a Sciex-API III mass spectrometer (Concord, ON) using electrospray ionization in the positive ion mode. Chemicals, Kits, and Solutions. D-Penicillamine was purchased from Richman Chemical Inc. (Lower Gwynedd, PA). Aldehyde-reactive probe (ARP) (N- (aminooxyacetyl)-N'-(D-biotinoyl)-hydrazine, trifluoroacetic acid salt), phosphate buffered saline (PBS), and fetal bovine serum (FBS) were purchased from Invitrogen Canada (Burlington, ON). Biotin-hydrazide and hydralazine were purchased from Sigma-Aldrich Canada (Oakville, ON). Roswell Park Memorial Institute (RPMI)-1640 medium with 4-(2- hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) modification was also purchased from Sigma-Aldrich Canada (Oakville, ON). Phycoerythrin-conjugated mouse anti-rat macrophage HIS36 monoclonal antibody (ED2 or CD163) was purchased from BD Pharmingen Canada (Mississauga, ON). Streptavidin-allophycocyanin was purchased from Cedarlane (Burlington, ON). MACS streptavidin magnetic microbeads and magnetic μ- columns were purchased from Miltenyi Biotec (Auburn, CA). The membrane protein extraction kit was purchased from BioVision (Mountain View, CA). Intracellular fixation buffer was purchased from eBiosciences (San Diego, CA). Gradient polyacrylamide gels were purchased from Bio-Rad Laboratories Canada (Mississauga, ON). All chemicals and anhydrous solvents used for synthesis of biotin-penicillamine were purchased from Sigma- Aldrich. The murine macrophage RAW 264.7 cell line was purchased from American Tissue Culture Collection (ATCC USA). Synthesis of Biotin-Penicillamine Conjugate. The scheme for the synthesis of biotin-penicillamine is shown in Figure 8. Details are provided below.

35 2-tert-Butoxycarbonylamino-3-mercapto-3-methylbutyric acid (2). To D- penicillamine (1, 0.50 g, 3.35 mmol) in tetrahydrofuran:H2O (50 mL:5 mL) was added

NaOH (0.13 g, 3.35 mmol). After they dissolved, di-tert-butyl-dicarbonate (Boc2O, 0.72 g, 3.35 mmol) was added and the reaction was stirred further at room temperature for 16 h after which the reaction mixture was acidified to pH 2 at 0 oC, using 1N HCl and then extracted with ethyl acetate (50 mL). The ethyl acetate layer was then dried over anhydrous sodium 1 sulfate and concentrated to yield 0.80 g of 2 as an oil in 96% yield. H NMR (CDCl3) δ 1.42 (s, 3H), 1.46 (s, 9H), 1.52 (s, 3H), 3.49 (s, 1H), 4.32 (d, J = 9.6 Hz, 1H), 5.32 (d, J = 9.3 Hz, 1H); ); 13C NMR δ 27.63, 28.50, 29.83, 31.03, 46.47, 50.89, 62.53, 80.78, 85.43, 155.96, 174.63; ESI-MS, MH+ m/z 250. Biotin-carbamic acid-tert-butyl ester (4). To biotin (3, 0.92 g, 3.60 mmol) dissolved in dry dimethylformamide (20 mL) were added 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (EDAC, 0.77 g, 4.05 mmol), 1-hydroxybenzotriazole (0.54 g, 4.05 mmol), followed by triethylamine (1.00 mL, 7.20 mmol). The mixture was stirred at 0 oC for 10 min after which was added N-1,5-diaminopentane-di-tert- butylcarbonate (0.72 mL, 3.60 mmol) and stirred for another 16 h. The crude reaction mixture was then evaporated under high vacuum and triturated with water to obtain a precipitate, which was washed with water: hexane (50 mL: 50 mL). The solid obtained was purified on a silica gel column (70-230 mesh) eluted with 8% methanol/CHCl3 to obtain 1 1.15 g of compound 4 as a white fluffy solid in 75% yield. H NMR (CD3OD) δ 1.27- 1.39 (m, 3H), 1.44 (s, 9H), 1.47- 1.52 (m, 5H), 1.61- 1.71 (m, 4H), 2.20 (t, J = 7.5 Hz, 2H), 2.74 (d, J = 12.9 Hz, 1H), 2.93 (dd, J = 4.8, 12.9 Hz, 1H), 3.06- 3.09 (m, 2H), 4.32 (dd, J = 4.5, 7.8 Hz, 1H), 4.52 (dd, J = 4.5, 7.8 Hz, 1H), 5.9 (bs, 1H); 13C NMR δ 25.01, 26.70, 28.91, 29.25, 29.57, 29.88, 30.42, 36.71, 40.17, 40.30, 41.03, 41.09, 56.79, 61.36, 63.14, 78.43, 78.80, 79.29, 79.78, 158.23, 165.74, 175.68, 175.77; ESI-MS, MH+ m/z 430. (2-mercapto-2-methyl-1-{5-(5-(2-oxo-hexahydro-thieno(3,4-o)imidazol-4-yl)- pentanoylamino)-pentylcarbamoyl}-propyl) carbamic acid- tert-butyl ester (5). To 4 (1.24 g, o 2.89 mmol) at 0 C was added 30% trifluoroacetic acid (TFA)/CH2Cl2 (40 mL) and stirred

36 further at room temperature overnight after which the solvent was evaporated and the residue dissolved in CH2Cl2 (30 mL)/methanol (30 mL) and neutralized with solid KHCO3. The solid was filtered and the filtrate was evaporated to give the pure deprotected product (0.94 g, 2.89 mmol), which was dissolved in dry dimethylformamide (30 mL) and added to 2 (0.72 g, 2.89 mmol), containing EDAC (0.60 g, 3.17 mmol), 1-hydroxybenzotriazole (0.42 o g, 3.17 mmol), and triethylamine (1.20 mL, 8.67 mmol) in dry CH2Cl2 (10 mL) at 0 C. The mixture was stirred further for 16 h at room temperature after which the crude reaction mixture was diluted with water (50 mL), extracted with CHCl3 (50 mL)/ethyl acetate (50 mL) and the organic layer was dried over anhydrous sodium sulfate and concentrated to yield crude product, which was purified with a silica gel column eluted with 5%-10% 1 methanol/CHCl3 to give 0.49 g of 5 as white fluffy solid in 31% yield. H NMR (CD3OD) δ 1.28- 1.75 (m, 27H), 2.19 (t, J = 7.2 Hz, 2H), 2.70 (d, J = 12.9 Hz, 1H), 2.93 (dd, J = 5.1, 12.9 Hz, 1H), 3.13- 3.25 (m, 5H), 4.08 (bs, 1H), 4.30 (dd, J = 4.5, 7.8 Hz, 1H), 4.48 (dd, J = 4.5, 7.8 Hz, 1H); 13C NMR δ 25.26, 27.04, 28.82, 28.95, 29.65, 29.93, 30.06, 30.16, 31.23, 36.97,40.35, 41.19, 46.91, 57.16, 61.76, 63.53, 64.46, 81.05, 166.24,172.24, 176.10, 215.06; ESI-MS, MH+ m/z 560. 5-(2-Oxo-hexahydro-thieno(3,4-o)imidazol-4-yl)-pentanoic acid (5-(2-amino-3- mercapto-3-methyl-butyryl amino)- pentyl)-amide (biotin-penicillamine, 6). To 5 (0.48 g, o 0.85 mmol) at 0 C was added 30% TFA/CH2Cl2 (20 mL) and stirred further for 3 h, after which the solvent was evaporated to give 0.48 g of pure 6 as its yellow TFA salt in 99% 1 yield. H NMR (CD3OD) δ 1.28-1.77 (m, 18H), 2.20 (t, J = 7.5 Hz, 2H), 2.70 (d, J = 12.9 Hz, 1H), 2.86-2.99 (m, 2H), 3.15-3.27 (m, 4H), 3.77 (s, 1H), 4.31 (dd, J = 4.5, 7.8 Hz, 1H), 4.50 (dd, J = 4.5, 7.8 Hz, 1H); 13C NMR δ 25.36, 26.94, 28.65, 29.53, 29.79, 30.10, 30.55, 40.14, 40.69, 41.08, 45.03, 57.02, 61.72, 63.45, 63.87, 167.49, 176.11; ESI-MS, MH+ m/z 460. Determination of Constitutive Aldehyde Groups on the Surface of Macrophages. The experimental procedures for the aldehyde binding experiments were based on a protocol kindly provided by Dr. Kevin Yarema’s lab at the Johns Hopkins University (167). The

37 spleen of a male BN rat was isolated and made into single cell suspensions using a 70 μm nylon mesh cell strainer. To lyse red blood cells, cells were incubated in a 0.17 M ammonium chloride solution for 5 min with occasional shaking. Two million cells were aliquoted into each well of a 96-well plate and washed two times with biotin-staining buffer (PBS pH 6.5 containing 1% FBS). Experiments were also performed at pH 7.4 with almost identical results (data not shown). Splenocytes were incubated with ARP or biotin- hydrazide in PBS pH 6.5 containing 1% FBS for 1.5 h at room temperature. A series of 2- times diluted concentrations of ARP or biotin-hydrazide was used to determine the dose- response curve of the binding. The cells were then washed four times with labeling buffer (PBS pH 7.4 containing 1% FBS) at 4 °C and incubated with streptavidin-allophycocyanin (0.2 μg/L × 106 cells) and rat macrophage mAb CD163 (0.2 μg/L × 106 cells) in labeling buffer for 15 min in the dark. The cells were then washed 3 times with labeling buffer and resuspended in 300 μL of fluorescent activated cell sorting buffer (HEPES modified RPMI- 1640 medium containing 10% FBS) and stored on ice for flow cytometry analysis. The same experimental procedure was performed for ARP binding to RAW 264.7 murine macrophages. Blocking of ARP Binding to Macrophages by Penicillamine and Hydralazine. Splenocytes were preincubated with penicillamine or hydralazine (0.55, 1.1, or 2.2 mM) in PBS pH 6.5 containing 0.5% FBS for 1 h at room temperature immediately before the ARP staining. Flow cytometry analysis was performed and the results were compared to those in which the splenocytes were not preincubated with penicillamine or hydralazine. In addition, another experiment was performed in which the ARP was replaced with the synthesized biotin-penicillamine reagent. Identification of ARP and Penicillamine Binding Proteins. Spleen cells, 5 or 10 × 106, were incubated with ARP or biotin-penicillamine (2.2 mM) in PBS pH 6.5 containing 1% FBS for 1 h at room temperature. Then a membrane protein extraction kit from BioVision was used to extract the total cellular membrane proteins. Specifically, cells were washed twice in ice cold PBS by centrifugation at 700g for 5 min at 4 °C. After adding 500

38 μL of homogenization buffer mix, the resulting cell suspension was homogenized by mechanical disruption using a 27-G syringe. Intact cells, cell debris, and major organelles were removed by centrifugation at 700g for 10 min at 4 °C. The supernatant containing cytosolic proteins and membrane proteins was centrifuged at 10,000g for 30 min at 4 °C. The pellet containing the membrane protein fraction, was dissolved in 50 μL of 0.5% Triton X-100 in PBS pH 7.4. Streptavidin magnetic microbeads (50 µL) were added and the mixture was incubated for 1 min at room temperature and then applied to the magnetic μ- column. After 5 washes in high salt washing buffer and 1 wash in low salt washing buffer, 20 μL of pre-heated 95 °C sodium dodecyl sulfate gel loading buffer was added to the column and allowed to incubate for 5 min at room temperature. Then 50 μl of pre-heated (95 °C) sodium dodecyl sulfate gel loading buffer was used to elute the column to collect the ARP or penicillamine bound membrane protein fraction which was subsequently analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). A 10-20% gradient SDS-PAGE gel and coomassie blue staining were used for protein separation and staining. After gel electrophoresis separation, an in-gel tryptic digestion was performed on the protein bands. The resulting digestion mixture was subjected to mass spectrometry at the Ontario Cancer Biomarker Network (Toronto, ON) using a Thermo Finnigen LTQ in the LC/MS mode. MS/MS was performed on peaks with > 1000 counts and the MASCOT 2.0 data base search engine was used to match peptides to known proteins. Comparison of ARP Binding to Macrophages from Different Rat Strains. Single spleen cell suspension was prepared from male BN, Lewis, or Sprague-Dawley rats as described above. The same ARP staining procedure was performed on splenocytes with CD163 mAb being used to specifically identify ARP-stained macrophages.

Figure 8. Synthetic scheme for biotin-D-penicillamine. Reagents and conditions i) Boc2O, NaOH; ii) 1,5 diaminopentane di-tert-butyl carbonate, EDAC; iii) 30% TFA/CH2Cl2; iv) 2 + EDAC; v) 30% TFA/CH2Cl2.

40 2.4. Results

Dose-Response Curve of ARP/Hydrazide/Penicillamine Binding to Aldehydes on Macrophages. The dose-response curves of ARP binding to total splenocytes and macrophages are shown in Figure 9A. ARP selectively bound to macrophages with about 90% being stained at 200 µM; in contrast the maximal staining of total spleen cells was < 60% even at high ARP concentrations. The results with the biotin-hydrazide reagent were similar (Figure 9B). A similar dose-response of ARP binding was also observed in RAW 264.7 murine macrophages (Figure 10). The binding curve for the biotin-penicillamine adduct is shifted far to the left relative to that for ARP or the biotin-hydrazide reagent and there is also greater binding to other types of cells, presumably because it is not specific for aldehyde groups and can also bind to thiols (Figure 11). A roughly 30% and 65% decrease in ARP staining was observed when macrophages were pre-incubated with penicillamine and hydralazine, respectively (Figure 12), which suggested that both penicillamine and hydralazine bound to aldehyde groups on macrophages and blocked subsequent ARP binding. Comparison of ARP Binding in Different Rat Strains. The dose-response relationship of ARP binding to splenic macrophages was compared among BN, Lewis, and Sprague-Dawley rats (Figure 13). At low concentrations there is significantly higher binding to macrophages from BN rats than from the other two strains. Identification of the Proteins to which ARP and Penicillamine Bind. As shown in Figure 14A, compared to the control in which cells were incubated with PBS buffer, bands of proteins isolated from spleen cells that were incubated with either ARP or penicillamine represent the target of ARP and penicillamine binding, respectively. There were a few extra bands in the penicillamine lane in the range of 54-210 kDa. These probably represent thiol-containing proteins capable of forming disulfide links with penicillamine but not with ARP. Even when the number of cells for ARP staining was increased from 5 × 106 to 10 × 106 to enhance membrane protein yield, we obtained a

41 similar band pattern as shown in Figure 14B in which there were still no visible bands in the control lane. All of these bands should be aldehyde-containing proteins although we cannot be certain that there are absolutely no contaminating proteins, especially if they are high abundance proteins. Subsequently, in-gel tryptic digestion and MS analysis were applied on each individual protein band for further characterization. When comparing the MS data with the rat protein database, two criteria were used to refine the list of protein candidates found by database searching: 1). The molecular mass of the protein had to match the molecular mass estimated form the position of the band on the gel; 2). There had to be at least two peptides generated by trypsin digestion that match the database. The search and comparison led to about 40 proteins, either cytoplasmic or membrane proteins (Table 6).

Figure 9. Dose-response curves of binding of ARP (A, n=3) and biotin-hydrazine (B, n=3) to splenocytes and macrophages of BN rats. The concentration of ARP is given above the upper right corner of each corresponding flow cytometry plot figure.

Figure 10. Dose-response curve of RAW 264.7 murine macrophages (0.25 million cells) staining with ARP (n=3). The concentration of ARP is given above the upper right corner of each corresponding flow cytometry plot figure.

Figure 11. Dose-response curves of splenocytes and macrophages staining with biotin- penicillamine (n=3).

A 75 * p < 0.05 ** p < 0.01 *** ** * p < 0.001 60 *** ***

30 ** ** *

Decrease of ARP staining (%) 15

0 D-pen Hydralazine

0.55mM 1.1mM 2.2mM 0.55mM 1.1mM 2.2mM

B 75 ** p < 0.01 *** ** * p < 0.001 60 *** ***

** ** 30 **

Decrease of ARP staining (%) 15

0 D-pen Hydralazine

0.55mM 1.1mM 2.2mM 0.55mM 1.1mM 2.2mM

Figure 12. Decrease in ARP staining of splenic macrophages by pre-incubation with penicillamine or hydralazine. The concentrations of ARP used in (A) and (B) are 0.0345 mM and 0.069 mM, respectively (n=3).

110 p < 0.05 BN * SD ** p < 0.01 ** * Lewis ** 100 ** * ** ** ** ** 90 ** ** **

80 ** ** 70

50 0.0173 0.0345 0.0690 0.1380 0.2750 0.5500 1.1000 2.2000

Concentration of ARP (mM)

Figure 13. Dose response of ARP staining of splenic macrophages from BN, Sprague- Dawley, and Lewis rats (n=3).

Figure 14. SDS-PAGE image of protein targets of ARP or biotin-penicillamine. MPF stands for the total membrane protein fraction of spleen cells. The total number of cells used for ARP or biotin-penicillamine incubation was 5 million and 10 million cells in upper and lower gel images, respectively. The concentration of ARP or biotin-penicillamine used was 2.2 mM in 100 μl. The negative control is from cells incubated with buffer instead of ARP or biotin-penicillamine. Except MPF, all protein samples were purified by an avidin magnetic column first and then separated via SDS-PAGE.

48 Table 6. Apparent ARP-binding proteins Protein Category Molecular Mass (Daltons) Peptides matched Membrane & Signaling STAT 1alpha 87249 17 STAT 3 87983 4 STAT 5B 90222 5 Annexin A6 75622 21 Tyrosine-protein kinase Syk B 71528 3 Tyrosine-protein kinase Lyn B 56001 3 Moesin 67607 6 Heat shock protein 90-beta 83184 43 Heat shock protein 86 84814 28 Heat shock protein cognate 71 70870 17 UDP-glucose:glycoprotein glucosyltransferase 1 176587 10 Ras GTPase-activating-like protein IQGAP1 200484 19 Tyrosine-protein phosphatase non-receptor type 6 69577 8 Phosphatidylinositol-4-phosphate 5-kinase type-2 alpha 46209 3 Antigen peptide transporter 1 (TAP1) 79149 2 Antigen peptide transporter 2 (TAP2) 77664 3 Cytoskeleton Actin 41736 17 Myosin-9 226204 13 CORO1A protein 51065 17 Vimentin 53601 22 Desmin 53325 2 Gelsolin 86285 7 Lamin A 74323 13 Lamin B-1 66474 42 Others Elongation factor 2 95152 11 Nucleoporin 93 93301 9 Importin beta-1 97183 9 Transketolase 67601 21 Tapasin 50044 3 Proteasome 26S 100187 2 Ribophorin I 68400 5 ATP synthase subunit alpha 59753 13 Protein disulfide-isomerase A3 57078 11 Transketolase 71158 13 Hisone 2a, Histone 2b, Histone 3 13700-16000 ~10

49 2.5. Discussion

The formation of the immunological synapse is critical for communication between antigen presenting cells (which includes macrophages) and T cells, ensuring efficient T cell activation under the right conditions. There are a number of pairs of molecules that have been found to be involved in the formation of the immunological synapse, including TCR/MHC-peptide, B7/CD28, and many types of cytokines and their corresponding receptors. Schiff base formation between amines and aldehyde groups on T cells and APCs, respectively, has been identified as one of the interactions between macrophages and T cells and therefore presumably is part of the immune synapse; however, the sources of the amine and aldehyde groups have not been identified. Schiff base formation between cells would seem ideal for signaling because the Schiff base is readily hydrolyzed and therefore would not hold two cells together if they randomly collide; however, in the context of an immune synapse where there were several interactions between the cells holding them together, a long-lasting covalent bond would be formed that could be involved in signal transduction. An agent such as a hydrazine, hydroxylamine, or penicillamine that forms a stable bond with aldehydes could mimic the more stable Schiff base interaction that occurs in the context of an immune synapse. This has the potential to lead to activation of macrophages and other antigen presenting cells in the absence of an immune synapse. Furthermore, generalized activation of such cells could lead to autoimmunity. Consistent with this hypothesis is the fact that treatment of patients with penicillamine and the hydrazines, hydralazine and isoniazid, is associated with a lupuslike autoimmune syndrome. Penicillamine also causes an autoimmune syndrome in BN rats, but the same is not true for hydralazine and isoniazid, possibly because of a short half-life in rodents. In this study, we found that penicillamine binds to macrophages, but it also binds to other spleen cells. This is presumably because penicillamine can bind to both thiols and aldehydes; therefore, we used ARP, which is selective for aldehyde groups to determine if macrophage cell membranes contain aldehyde groups. The binding of ARP to macrophages

50 was much more selective for macrophages than penicillamine, and its binding to macrophages was partially blocked by pretreatment with penicillamine or hydralazine. This appears to be a dynamic system with turnover of aldehyde group-containing membrane molecules because the longer the time lag between washing the cells and analysis by flow cytometry is, the lower the degree of inhibition is (data not shown). In addition, the reaction between the aldehyde group and the penicillamine or hydralazine is not instantaneous, and although we have not attempted to measure the kinetics, it is likely that the reaction with penicillamine is slower than that of a hydroxylamine or hydrazine because penicillamine has to be in the correct orientation for the second reaction to form the thioazolidine ring to make the binding irreversible. Thus, the lack of complete inhibition by hydralazine and penicillamine may be due to turnover of cell membrane molecules and the lesser degree of blocking by penicillamine may be due to a slower reaction with aldehyde groups. Lewis and Sprague-Dawley rats are resistant to the autoimmune syndrome caused by penicillamine. Our previous study showed that, unlike in BN rats, there was no increase of number of activated splenic macrophages in Lewis rats one week after penicillamine treatment (155). The finding that binding of ARP to macrophages from Lewis and Sprague- Dawley rats was significantly less than that to macrophages from BN rats was somewhat surprising, but it could contribute to the resistance of these strains to penicillamine-induced autoimmunity. We identified several proteins that appear to contain aldehyde groups. LC/MS analysis of protein bands from splenocytes incubated with ARP generated a list of potential target membrane proteins. However, due to limitations of the membrane protein extraction kit, the membrane proteins were contaminated by some cytoplasmic proteins, although even some of these may also be expressed on the cell membrane. Out of almost 40 proteins identified, STAT-related proteins are the most interesting. Moesin, STAT1, STAT3, STAT5, and tyrosine protein kinases Lyn B and Syk B are all well-known signaling molecules in a number of crucial biological signal transduction pathways. Out of these signaling molecules, moesin seems to best fit the original hypothesis in that it is one of the three macrophage

51 membrane proteins called ERM proteins (ezrin, radixin, moesin) that are well known to be involved in the formation of the immune synapse between T cells and macrophages (168).

In addition, Lyn kinase is also very interesting because the first step in HgCl2-induced autoimmunity syndrome in BN rats, which is very similar to the autoimmunity caused by penicillamine, appears to involve binding of HgCl2 to Lyn (169). Binding to these signaling proteins could potentially regulate macrophage functions that can lead to the systemic activation of the immune system. However, at this point we cannot be sure which, if any, of these proteins is involved in the activation of macrophages by penicillamine. We have not yet characterized the chemical origin of the aldehyde on the surface of macrophages. None of the normal amino acids have an aldehyde group, although there are examples of posttranslational oxidation of lysine to form an aldehyde (170). As mentioned in the Introduction, although the carbonyl group involved has been referred to as an aldehyde, other reactive carbonyl groups, in particular ketones, would lead to similar results. Non-enzymatic oxidation of proteins by reactive oxygen species can also lead to reactive carbonyl groups (97, 98). In addition, glycation of proteins can lead to reactive carbonyl groups. This can either occur through the enzymatic posttranslational addition of a sugar to a protein, or it can be due to addition of a reactive molecule such as methylglyoxal; such products are referred to as advanced glycation end products (AGEs) (100). AGEs are known to stimulate an immune response (171-173). We have ongoing studies to determine the exact chemical source of the aldehyde groups, especially on the most interesting proteins. In summary, our results have demonstrated that aldehyde-reactive compounds such as penicillamine and hydralazine are able to covalently bind to aldehydes (or other reactive carbonyl groups) present on the cell membrane of macrophages. In spite of limited understanding the source of these aldehyde groups, we did identify some potential target proteins, either membrane or cytoplasmic proteins that ARP/penicillamine bind to, especially some known to be involved in signal transduction. This provides an attractive hypothesis for how penicillamine induces autoimmunity in both humans and BN rats.

52 Acknowledgement. J.U. holds a Canada Research Chair in Adverse Drug Reactions. This research work was supported by grants from the Canadian Institutes of Health Research.

CHAPTER 3 D-PENICILLAMINE-INDUCED AUTOIMMUNITY: RELATIONSHIP TO MACROPHAGE ACTIVATION

54 3.1. Abstract

Idiosyncratic drug reactions represent a serious health problem, and they remain unpredictable largely due to our limited understanding of the mechanisms involved. Penicillamine-induced autoimmunity in Brown Norway (BN) rats represents one model of an idiosyncratic reaction, and this drug can also cause autoimmune reactions in humans. We previously demonstrated that penicillamine binds to aldehydes on the surface of macrophages. There is evidence that an imine bond formed by aldehyde groups on macrophages and amine groups on T cells is one type of interaction between these two cells that is involved in the induction of an immune response. We proposed that the binding of penicillamine with aldehyde groups on macrophages could lead to their activation and in some patients could lead to autoimmunity. In this study, the transcriptome profile of spleen macrophages 6 h after penicillamine treatment was used to detect effects of penicillamine on macrophages with a focus on 20 genes known to be macrophage activation biomarkers. One biological consequence of macrophage activation was investigated by determining mRNA levels for IL-15 and IL-1β that are crucial for NK cell activation, as well as levels of mRNA for selected cytokines in spleen NK cells. Up-regulation of the macrophage activating cytokines, IFN-γ and GM-CSF, and down-regulation of IL-13 indicated activation of NK cells, which suggests a positive feedback loop between macrophages and NK cells. Furthermore, treatment of a murine macrophage cell line, RAW264.7, with penicillamine increased the production of TNF-R, IL-6, and IL-23, providing additional evidence that penicillamine activates macrophages. Hydralazine and isoniazid cause a lupus-like syndrome in humans and also bind to aldehyde groups. These drugs were also found to activate RAW264.7 macrophages. Together, these data support the hypothesis that drugs that bind irreversibly with aldehydes lead to macrophage activation, which in some patients can lead to an autoimmune syndrome.

55 3.2. Introduction

Idiosyncratic drug reactions (IDRs) are a significant health problem, and they also cause a great deal of uncertainty in drug development. Even though the mechanisms are still unclear, clinical characteristics implicate an immune-mediated mechanism for most IDRs (174). This begs the question of how a drug and/or its reactive metabolites activate the immune system eventually leading to a pathogenic immunological reaction in some patients. Macrophages are an important cell in the initiation of an immune response. As a major phagocytic cell type, macrophages are responsible for the cleanup of dead cells and immune complexes (175-177). In addition, macrophages perform innate immune functions in response to stimulation via pathogen recognition receptors (PRRs), and they are also actively involved in adaptive immunity by presenting antigens to T cells and producing a variety of cytokines and chemokines that are involved in the regulation of the immune response (178-181). Macrophage dysfunction is associated with many different kinds of immune-mediated diseases. For example, uncontrolled macrophage activation appears to be involved in hemophagocytic syndromes, rheumatic disorders, and juvenile systemic lupus erythematosus, etc (182-185). Penicillamine-induced autoimmunity in Brown Norway (BN) rats represents an important animal model for mechanistic studies of IDRs because it closely reflects several of the penicillamine-induced autoimmune reactions that occur in humans (154). Previous work in our lab found that macrophage activation and infiltration were very early events and preceded the clinical syndrome by weeks (153). Different immunomodulators that are able to modify the incidence and severity of the disease had a similar effect on the infiltration of activated macrophages. This suggested an important role of macrophages in the early stages of the pathogenesis of penicillamine-induced autoimmunity. In addition, it has been reported that pretreatment of rat peritoneal macrophages with penicillamine can enhance their ability to modulate the lymphocyte response to specific and non-specific stimuli as indicated by the changes in concanavalin A-stimulated 3H-thymidine incorporation in rat

56 lymph node cells (156-158). Rhodes found evidence that one of the interactions between macrophages and T cells involves a reversible imine bound formed by aldehydes on macrophages and amines on T cells (162, 164). We proposed that irreversible binding to these aldehydes might lead to macrophage activation, and our most recent studies demonstrated that penicillamine irreversibly binds to surface aldehyde-containing molecules on macrophages (86, 186). Hydralazine and isoniazid are hydrazines that also cause autoimmunity in humans and also irreversibly bind to aldehydes. Therefore it is possible that binding to aldehyde-containing proteins on macrophages leading to their activation could represent one mechanism by which a drug could induce an idiosyncratic drug reaction (Figure 15).

Figure 15. Hypothesis that covalent binding of penicillamine to macrophages leads to macrophage activation.

58 3.3. Materials and Methods

Animals. Male BN rats (150-175 g) were purchased from Charles River (Montreal, Quebec) and doubly housed in standard cages in a 12:12 h light:dark cycle at 22 °C. The rats were given free access to standard rat chow (Agribrands, Purina Canada, Strathroy, Ontario) and water for a weeklong acclimatization period before starting experiments. Chemicals, Kits, and Solutions. D-penicillamine was purchased from Richman Chemical Inc. (Lower Gwynedd, PA). Aldehyde-reactive probe (N-(aminooxyacetyl)-N'-D- biotinoyl) hydrazine, trifluoroacetic acid salt; ARP), phosphate buffered saline (PBS), and fetal bovine serum (FBS) were purchased from Invitrogen Canada (Burlington, ON). RNeasy Mini kits were purchased from Qiagen (Mississauga, Ontario, Canada) for purification of total RNA. OmniScript reverse transcriptase kits were purchased from Qiagen. Phycoerythrin-conjugated mouse anti-rat macrophage HIS36 monoclonal antibody (ED2 or CD163) was purchased from BD Pharmingen Canada (Mississauga, ON). Streptavidin-allophycocyanin was purchased from Cedarlane (Burlington, ON). Magnetic cell sorting (MACS) anti-PE magnetic beads were purchased from Miltenyi Biotec (Auburn, CA). RAE 230 2.0 gene chips were purchased from Affymetrix (Santa Clara, CA). LightCycler SYBR Green I kits for quantitative real-time polymerase chain reaction (qRT- PCR) were purchased from Roche (Montreal, Quebec, Canada). The murine macrophage RAW 264.7 cell line was purchased from American Tissue Culture Collection (ATCC USA). Dulbecco’s modified eagle’s medium (DMEM), Roswell Park Memorial Institute (RPMI)- 1640 medium with 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) modification, and the antibiotics, penicillin and streptomycin, were purchased from Sigma- Aldrich, Canada. Mouse ELISA kits for IL-6, IL-23, and TNF-alpha were purchased from R&D system (Minneapolis, MN). Transcriptome analysis of macrophages. In order to examine the gene expression profile of macrophages, BN rats were given 1 mL of penicillamine dissolved in tap water in a single dose of 150 mg/kg by oral gavage. At 6 h post-dosage, spleen macrophages were purified by using phycoerythrin-conjugated anti-rat macrophage ED2 monoclonal antibody

59 and mini-MACS immunomagnetic separation column (Miltenyi Biotec, USA) according to the manufacturer’s instructions. Total RNA was then isolated using RNeasy mini kits as described by the manufacturer. RNA concentration and purity were determined spectrophotometrically. RNA quality was further assessed by capillary electrophoresis using Agilent Bianalyzer. Microarray analysis was performed at the Centre for Applied Genomics, Hospital for Sick Children (Toronto, ON) by using Rat Expression Array 230 2.0 from Affymetrix. Meanwhile, aliquots of RNA were saved to confirm the gene expression changes using qRT-PCR with a Roche LightCycler. Determination of the activation status of NK cells. Natural killer (NK) cells from the spleens of male BN rats 6 h after a single 150 mg oral gavage dose of penicillamine were isolated by using a magnetic column and monoclonal antibodies, CD161 and CD5. The mRNA profile of cytokines from CD5-CD161+ cells was determined by a Roche LightCycler. Determination of ARP binding to aldehydes on mouse Raw 264.7 macrophages. Raw 264.7 macrophages were scraped off the culture flask and passed through a 40 μm cell strainer to produce a single cell suspension. One million cells were aliquoted into each well and washed two times with biotin-staining buffer (PBS pH 6.5 containing 1% FBS). Macrophages were incubated with aldehyde-reactive probe (ARP) in PBS, pH 6.5 containing 1% FBS for 1.5 h at room temperature. A series of 2-times diluted concentrations of ARP was used to determine the dose-response binding curve. The cells were then washed four times with labeling buffer (PBS pH 7.4 containing 1% FBS) at 4 °C and incubated with streptavidin-allophycocyanin (0.2 μg/L × 106 cells) in labeling buffer for 15 min in the dark. The cells were then washed 3 times with labeling buffer and resuspended in 300 μL of fluorescent-activated cell sorting buffer (HEPES modified RPMI-1640 medium containing 10% FBS) and stored on ice for flow cytometry analysis. Treatment of RAW 264.7 macrophages with penicillamine, isoniazid, or hydralazine. Raw 264.7 cells were maintained in D-MEM medium supplemented with 10% fetal heat-inactivated bovine serum and 10-times diluted antibiotics (100 U/mL penicillin and 100 μg/mL streptomycin) at 37 °C and 5% CO2 atmosphere. The culture

60 medium was changed every 2-3 days and cells were passaged when cell density reached 80% of confluency. For drug treatment, 0.5 × 106 cells were seeded into each well of a 24- well tissue culture plate. The cells were allowed to adhere and settle overnight at 37 °C. Non-adherent cells were removed by aspiration, and pre-warmed 1.0 mL of D-MEM medium only or containing different concentrations (a series of 2-fold dilutions starting from 2.2 mM down to 0.0173 mM plus a negative control) of penicillamine, isoniazid, or hydralazine was added to pre-assigned wells. The plate was put back into incubator and medium was collected at certain time points for cytokine analysis using ELISA.

61 3.4. Results

Transcriptome of macrophages at 6 h post-dosage of penicillamine. Genechip Operating System (GCOS) was utilized to analyze the microarray data. First of all, in order to determine whether the experimental procedure caused any significantly inconsistent difference in gene expression, a comparison was performed between the transcriptome files of each pair of animals within both control (C) and penicillamine-treated (T) groups. The results are shown in Figure 16 in which red dots represent the genes expressed in both X- and Y- axis animals; yellow dots represent the genes expressed in neither animal, and blue dots are those genes that were expressed in X-axis animals but not in Y-axis animals. For every individual comparison, the gene dots were all located in a range angled at about 45 degrees, which means a very good consistency of experimental performance throughout animals in each group. Due to the idiosyncratic nature of autoimmune syndromes caused by penicillamine in BN rats, namely only a bit over 50% of animals develop autoimmunity, all three penicillamine-treated rats were separately compared to each control rat to individualize gene expression changes, which resulted in nine GCOS comparisons (T1 vs. C1, C2, and C3; T2 vs. C1, C2 and C3; T3 vs. C1, C2 and C3). The criteria for defining whether genes were up- or down-regulated were if at least comparisons between 2 treated rats with 2 control rats were different and the fold change for up-regulation and down-regulation was no less than 1.2 and no more than 0.7. In total, 324 Affymetrix IDed genes were found to be up-regulated and 273 were found to be down-regulated in treated vs. control. Subsequently, refining these changes through the Affymetrix gene database narrowed the search down to a number of well-studied genes shown to be associated with multiple physiological functions and the activation status of macrophages (Table 7). These genes can be further categorized into ten groups according to gene ontology classes: scavenger receptors, interleukins and receptors, chemokines and receptors, S100 proteins, inflammatory response, complement system, stress response, transcription factors, metabolism, and transporters. A comparison of observed

62 changes in our study with genes that have been reported in literature to represent macrophage activation markers led us to focus on 20 genes: CD14, CD36, CD163, IL1β, IL15, IL13 receptor α1, CCL5, CCR5, CXCR4, S100A8, S100A9, annexin 1 & 4, ALOX5AP, complement component 3, fibronectin 1, ALOX12, DUSP1 & 6, and SBP2. Out of these 20 genes, 14 associated with inflammatory reactions, recruitment of immune cells, and regulation and polarization of macrophages, natural killer cells, and T helper cells were chosen for additional testing by qRT-PCR. Most of the penicillamine-induced changes in mRNA expression by macrophages identified by microarray analysis were confirmed by qRT-PCR (Figure 17). NK cell activation. The mRNA expression profile of each penicillamine-treated animal relative to the average of the control animals is summarized in Figure 18. IFN-γ was approximately 2-fold up-regulated in two out of three treated rats. Also, granulocyte- macrophage colony stimulating factor (GM-CSF) was significantly up-regulated in one treated rat and marginally up-regulated in the other one. In contrast, IL-13 was significantly up-regulated in one rat and down-regulated in the other two. Up-regulation of MIP-1β and IL-10 were found only in one treated rat. There is no differential mRNA expression of IL-4, IL-5, TNF-α, MIP-1α, RANTS, and TGF-β1 in treated animals. Dose-response curve of ARP binding to aldehydes on RAW 264.7 macrophages. The dose-response curve of ARP binding to RAW 264.7 macrophages showed that ARP staining of RAW 264.7 cells is a function of the concentration of ARP, which is similar to what we observed previously in BN rat macrophages (Figure 19). Maximal ARP staining can be obtained at a concentration of approximately 1.1 mM. Activation effect of penicillamine, isoniazid, and hydralazine on RAW 264.7 macrophages. Incubation of penicillamine for 24 h induced the production of cytokines TNF-α, IL-6, and IL-23 (unlike TNF-α and IL-6, detectable levels of IL-23 were not observed until 24 h of incubation) in RAW 264.7 macrophages (Figure 20). For TNF-α and IL-6, a penicillamine concentration of 0.069 mM appeared to cause maximal release, while a concentration of 0.275 mM induced maximal release of IL-23. Meanwhile, macrophages

63 incubated with isoniazid or hydralazine were also found to have an increased production of IL-6, which implies they also activate macrophages (Figure 21). In addition, isoniazid and hydralazine appear to have a greater activation effect on RAW 264.7 cells than penicillamine in terms of IL-6 production. Furthermore, the optimal concentration to promote IL-6 production is much greater for both isoniazid and hydralazine than penicillamine. Except at concentrations of hydralazine higher than 0.55 mM, there was no cellular toxicity observed during treatment of RAW 264.7 cells.

Figure 16. Comparison of transcriptome of macrophages within the control and penicillamine groups. Red dots represent the genes expressed in both X- and Y- axis animals, yellow dots represent the genes expressed in neither of animals, and blue dots are those genes that were expressed in X- axis animals but not in Y-axis animals.

Figure 17. Validation of expression of differentially regulated genes by qRT-PCR.

Figure 18. mRNA expression profile of cytokines in NK cells at 6 h post-dosage of penicillamine. Three penicillamine-treated rats (T1, T2, and T3) were compared to the average of three control rats.

Figure 19. Dose-response curve of RAW264.7 macrophages (1 million cells) staining with ARP (n=3).

Figure 20. Induction of cytokine production in RAW 264.7 cells by penicillamine (n=3). A. TNF-α; B. IL-6; C. IL-23. The time points for the TNF-α and IL-6 data at each concentration of penicillamine treatment are: 3, 6, 12, and 24 h (from left to right) while IL-23 was not detectable until 24 h which is the time point for the IL-23 data.

Figure 21. IL-6 production in RAW264.7 macrophages incubated with penicillamine, hydralazine, or isoniazid for 24 h (n=3). The drug concentrations are a series of 2-fold dilutions starting from 2.2 mM down to 0.0173 mM plus a negative control. Concentrations of hydralazine higher than 0.55 mM led to obvious toxicity and cell death.

Table 7. Differentially expressed macrophage genes in Brown Norway rats at 6 h post-dosage of penicillamine

Fold change Gene name & category Biological process T1 T2 T3 C1 C2 C3 C1 C2 C3 C1 C2 C3 Scavenger receptors CD14 Inflammatory response (with TLR4) NC NC NC 1.41 1.52 1.41 1.32 1.41 1.41 CD163 Antimicrobial humoral response NC NC NC 1.41 1.32 1.41 1.32 1.23 1.41 Interleukins & receptors Interleukin1 beta Inflammatory response, immune response NC NC NC NC 1.41 1.32 NC 2.0 1.87 Interleukin15 Positive regulation of immune response NC NC NC NC 1.41 1.41 NC 1.74 1.41 Interleukin 3 receptor, alpha 1 Dimered with IL4RA for IL4 and IL13 NC NC NC 1.32 1.74 2.30 1.62 2.14 2.83 Chemokines & receptors Chemokine (C-C motif) ligand 5 Inflammatory response, immune response 4.92 1.74 1.52 4.29 1.52 1.41 2.64 NC NC Chemokine (C-C motif) receptor 5 G-protein coupled signaling pathway 0.54 0.76 NC NC 1.32 1.74 NC 1.52 1.87 Chemokine (C-X-C motif) receptor 4 G-protein coupled signaling pathway 2.30 NC 1.73 1.87 0.71 1.52 1.52 0.54 NC Other cell surface receptors CD38 Cell adhesion, signal transduction, and calcium signaling NC NC NC NC 1.32 2.14 1.23 1.74 2.83 CD52 Compliment activation 0.31 0.35 0.2 NC NC 0.54 0.38 0.38 0.27 CD71 (transferrin receptor) Cellular iron ion homeostasis 1.23 1.41 1.23 1.52 1.74 1.52 1.32 1.52 1.23 Prostaglandin E receptor EP2 G-protein coupled prostaglandin E receptor activity NC NC NC NC 1.62 1.87 NC 1.41 1.52 S100 proteins S100 calcium binding protein A4 Calcium ion binding, calcium-dependent protein binding 3.48 2.83 1.41 2.46 1.87 NC 2.14 1.74 NC S100 calcium binding protein A6 Cell cycle, cell proliferation, ion transmembrane transporter activity 3.48 3.48 NC 2.0 2.0 NC 2.142.14NC S100 calcium binding protein A8 6.06 4.60 1.62 3.48 2.00 0.76 2.46 1.87 0.66 S100A8–S100A9 heterodimers promote further recruitment of leukocytes S100 calcium binding protein A9 7.46 4.59 1.52 3.03 2.00 0.62 2.83 2.00 0.62 Complement system Complement component 3 Complement activation 2.00 1.87 NC 1.62 1.52 NC 2.00 2.00 NC Ficolin B Phosphate transport, signal transduction 3.25 4.00 1.41 2.30 2.83 NC 2.00 2.64 NC Galectin-1 (beta-galactoside-binding lectin) Positive regulation of I-kappaB kinase/NF-kappaB cascade 1.74 1.87 1.23 1.41 1.62 NC NC 1.41 NC Inflammatory response Annexin A1 Prostaglandin synthesis regulation 1.87 1.62 1.41 NC NC 0.76 1.52 1.41 NC

71 Annexin A4 Prostaglandin synthesis regulation 0.76 NC NC NC NC 1.41 NC NC 1.41 Arachidonate 5-lipoxygenase activating protein Leukotriene metabolism 2.64 1.74 1.52 1.52 NC NC 2.46 1.62 1.52 Fibronectin 1 Acute phase response 3.03 3.25 1.32 1.62 1.87 NC 1.87 1.87 NC Arachidonate 12-lipoxygenase Prostaglandin and leukotriene metabolism 0.41 0.54 0.25 NC 1.41 0.66 0.35 0.47 0.23 Stress response Dual specificity phosphatase 1 Response to oxidative stress 1.74 1.74 2.14 NC NC NC 2.46 2.46 4.00 Dual specificity phosphatase 6 Regulation of cell cycle NC NC NC NC 1.23 1.23 NC 1.32 1.23 Selenium binding protein 2 Selenium binding 0.57 0.57 0.27 NC NC 0.77 0.47 0.47 0.19 Transcription factor NF-E2-related factor 2 (Nrf2) Regulation of transcription NC 1.23 NC 1.23 1.41 1.23 1.32 1.52 1.32 Kruppel-like factor 4 (KLF4) Regulation of transcription, proinflammatory signal transduction 1.52 2.30 1.23 NC 1.87 NC 1.87 2.64 1.41 Immediate early transcription factors NGFI-B (Nr4a1) Regulation of transcription NC 1.41 NC 1.32 1.52 NC NC 1.52 1.32 Early growth response 1 Regulation of transcription NC 1.62 1.62 NC NC NC 1.52 2.30 2.30 Metabolism L-Arginine:glycine amidinotransferase Response to oxidative stress 1.62 2.30 1.32 NC 1.52 NC 1.23 1.74 NC lysozyme Metabolic process, defense response 1.52 1.41 NC 1.32 1.23 NC 1.41 1.32 NC Carnitine palmitoyltransferase 1 alpha, liver isofor Regulation of fatty acid beta-oxidation, glucose metabolic process NC NC NC 0.71 0.57 0.76 0.76 0.57 0.76 Transporters Solute carrier family 4, member 1 Ion transport and anion transport 0.35 0.35 0.22 NC NC 0.57 0.35 0.41 0.23 Solute carrier family 28, member 2 Nucleoside transporter, 0.71 NC 0.71 1.41 1.51 NC 1.74 1.87 1.51 Solute carrier family 30, member 1 Ion transport, cation transport, zinc ion transport 0.76 NC NC NC 2.00 2.30 NC 1.87 2.46 Solute carrier family 11, member 1 Acute and chronic inflammation 2.00 2.00 NC 2.00 1.87 NC 1.62 1.74 NC Others Serum/glucocorticoid regulated kinase Inflammation, response to DNA damage stimulus 1.32 NC NC 1.32 1.41 NC 1.52 1.62 1.32 Protein tyrosine phosphatase Dephosphorelation 1.74 1.74 1.62 NC NC NC 2.46 2.46 2.30 Syntaxin 3 Intracellular protein transport 0.71 NC NC 1.62 1.87 2.14 1.62 1.62 2.00 Serine protease Proteolysis, cytolysis 4.00 2.46 1.52 2.83 1.41 NC 2.30 NC NC

72 Table 8. Primer sequences for qRT-PCR Gene Forward Primer (5’-3’) Reverse Primer (5’-3’) CD14 antigen AGAACGCTGCTGTAAAGGAAAG TCAAGGGCAGAGACCTGATAAT CD52 antigen AGCTGTTACAGAGCCCAAGAAG TTTTGTCCCAAGACTCCTGTTT CD163 antigen GACATCTGGATGGACAAGGTTT CCCAGATAGCTGACTCATTTCC Interleukin 1 beta AGGCTTCCTTGTGCAAGTGT TGAGTGACACTGCCTTCCTG Interleukin 4 TCAACACTTTGAACCAGGTCAC GCAGCTTCTCAGTGAGTTCAGA Interleukin 5 ATGAGCACAGTGGTGAAAGAGA TCTTGCAGGTAATCCAGGAAAT Interleukin 10 AGGACCAGCTGGACAACATACT TCATTCATGGCCTTGTAGACAC Interleukin 13 ACAGGACCCAGAGGATATTGAA AACTGAGGTCCACAGCTGAGAT Interleukin 13 receptor, alpha 1 AAGTGGGGTCCCAGTGTAGC GTGTTGACCTTCTCTGTGGATG Interleukin 15 CTTCTTAACTGAGGCTGGCATC GTGAAGTTTCTCTCCTCCAGCT MIP-1α (CCL3) AAAGAGACCTGGGTCCAAGAAT TTCAAGTGAAGAGTCCCTGGAT MIP-1β (CCL4) TACGTGTCTGCCTTCTCTCTCC CAAAGGCTGCTGGTCTCATAG CCL5 TCCACAGTCTCTGCTTCAGGTA CTTGAACCCACTTCTTCTCTGG CXC4 TCATCAAGCAAGGATGTGAGTT TTGAGGATTCTGACTCTGTGGA CCR5 TGCTAACAGGGAAGAACCACTT TCAAAGCTGGTACGGTAGGATT IFN-γ ATATCTGGAGGAACTGGCAAAA TAGATTCTGGTGACAGCTGGTG TGF-β1 AACTGTGGAGCAACACGTAGAA GTATTCCGTCTCCTTGGTTCAG TNF-α GAAAACGGAGCTGAACAATAGG GCAAACTTTATTTCTCGCCACT CCR5 TGCTAACAGGGAAGAACCACTT TCAAAGCTGGTACGGTAGGATT GM-CSF GCATGTAGATGCCATCAAAGAA GAAATCCTCAAAGGTGGTGACT Proteoglycan 2, bone marrow GATGGAAGCTCTTGGAATTTTG CCAGGAGAGGATGAATTTGAAC Complement component 3 GGGGAGCCCCATGTACTC TTGTTGTCCACAGTGAAGATCC Fibronectin 1 GGACCAGAGATCTTGGATGTTC CGATTTGGACCTCCTCATCTAC Arachidonate 5-lipoxygenase AAGGTGGAGCTTGAAAGCAA AAGTGGGGTACGCATCTACG activating protein S100 calcium binding protein A9 CCAAAACAGGATCTCAGCTG GGTTGTTCTCATGCAGCTTC S100 calcium binding protein A8 CGACAATGGCAACTGAACTG CCACCCTTATCACCAACACA Arachidonate 12-lipoxygenase GTTCGTGAAACTGCACAAAGAG GGAGGTCATCCTTACAGTCTGC Selenium binding protein 2 TAGTGGTCAAGGGAAAACGAGT AGCCCTCCATTTGCTGTATCTA

73 3.5. Discussion

The global mRNA expression profile of macrophages isolated from BN rats very shortly after penicillamine treatment demonstrated the activation of macrophages. Some of the genes are known to be involved in innate immune reactions (e.g., scavenger receptor CD14 and CD163 (187-191)) while others are more associated with regulation of adaptive immunity (e.g., CCR5, IL-13r). In addition, we found the up-regulation of several transcription factors such as Nrf2, Kruppel-like factor 4, and Nr4a1 etc. Furthermore, gene expression of a group of solute carrier family proteins for transportation of ions were shown to be changed (e.g., solute carrier family 4, 11, 28, and 30), suggesting that change of intracellular level of certain ions could result in macrophage activation in response to penicillamine. The fact that penicillamine caused rapid macrophage activation while the onset of autoimmunity occurs after about 3 weeks of treatment strongly suggests that macrophages play a role in the initiation of the immune response leading to autoimmunity. This is consistent with our previous study in which significant infiltration of activated macrophages into the caecum was found only 96 h after penicillamine treatment (155). In addition to demonstrating that penicillamine led to activation of macrophages, downstream effects of this activation were also observed. Specifically, the microarray data point to NK cell activation: 1) the up-regulation of IL-15 and IL-1β, which are known to play a crucial role in NK cell proliferation, cytotoxicity, and cytokine production (192-194); 2) NK cells play an important immunomodulatory role in adaptive immune responses by releasing chemokines and cytokines such as IFN-γ that appear to play a key role in autoimmune diseases (195-197). The mRNA expression profile of cytokines in NK cells at 6 h with up-regulation of IFN-γ and GM-CSF and down-regulation of IL-13 suggests NK cell activation. IFN-γ and GM-CSF are able to activate macrophages, which suggests a positive feedback regulation between macrophages and NK cells. Taken together, these data provide strong evidence that macrophage activation plays an important role in the early events leading to penicillamine-induced autoimmunity.

74 As in rat macrophages, the existence of aldehyde-containing molecules on the surface of murine macrophages was demonstrated by the binding of ARP to RAW 264.7 cells. This appears to lead to RAW 264.7 macrophage activation because it led to the release of TNF-α, IL-6, and IL-23. The concentration of penicillamine in this study that induced cytokine production in RAW 264.7 cells starts from 17 µM (2.5 µg/mL), which is very close to the peak plasma concentrations of penicillamine (approximately 3 µg/mL) in patients with rheumatoid arthritis treated with penicillamine (198). In contrast to the optimal concentration to induce the release of TNF-α and IL-6 (0.069 mM), it required a penicillamine concentration of 0.275 mM to lead to the maximal release of IL-23. This may be due to the difference in the mechanism by which the production of these cytokines is controlled with the production of TNF-α and IL-6 being transcription factor NF-kappaB dependent, while as a heterodimeric cytokine, overall expression of IL-23 is regulated differently (199-201). Although penicillamine appeared to activate macrophages in vivo and RAW 264.7 cells in vitro, the pattern of cytokine release was somewhat different. The microarray data from in vivo experiments showed that penicillamine led to an increased expression of IL-1β and IL-15, but activation of RAW 264.7 cells in vitro was characterized by the release of IL-6 and TNF-α. Overall, both in vivo and in vitro studies support the hypothesis that penicillamine activates macrophages through binding to aldehyde-containing molecules on macrophages. More importantly, isoniazid and hydralazine, which also bind to aldehydes and induce an autoimmune syndrome in humans, also stimulate IL-6 production in RAW 264.7 cells. This provides further evidence that binding to aldehydes may represent a general mechanism by which penicillamine, isoniazid, and hydralazine activate macrophages leading to lupus-like syndromes. Acknowledgement. J.U. holds a Canada Research Chair in Adverse Drug Reactions. This research work was supported by grants from the Canadian Institutes of Health Research.

CHAPTER 4 TH17 INVOLVEMENT IN PENICILLAMINE-INDUCED AUTOIMMUNE DISEASE IN BROWN NORWAY RATS

Jinze Li, Xu Zhu, and Jack Uetrecht Work from this chapter will be submitted for publication.

76 4.1. Abstract At present, idiosyncratic drug reactions (IDRs) are unpredictable largely due to a lack of mechanistic understanding, although their clinical characteristics suggest that they are immune-mediated. For example, the delay between starting the drug and the onset of an IDR is most easily explained by an immune mechanism. Penicillamine-induced autoimmunity in Brown Norway rats has been utilized as an animal model for mechanistic studies of one type of IDR because it closely mimics the autoimmune syndromes that this drug causes in humans. Our previous work suggested that it is a T cell-mediated immune reaction. It has been shown that Th17 cells play a central role in many types of autoimmune diseases. This study was designed to test whether Th17 cells are involved in the pathogenesis of penicillamine-induced autoimmunity and to establish an overall serum cytokine/chemokine profile for this IDR as a possible template for other types of IDRs. In sick animals, IL-6 and TGF-β1, known to be driving forces of Th17 differentiation, were consistently increased shortly before the onset of autoimmunity and a few days after the treatment, respectively; however, no significant changes in these cytokines were observed in animals that did not develop autoimmunity. In addition, IL-17, one of the most characteristic cytokines produced by Th17 cells, was increased in sick animals at both the mRNA and serum protein level. This strongly suggests that Th17 cells are involved in the autoimmune syndrome caused by this agent. In total, 24 serum cytokines/chemokines were determined by a Luminex assay and revealed a dynamic process. For example, a peak in IL-13 at an early time point predicted which animals would develop autoimmunity. Such data provide important mechanistic clues that may help to predict which drug candidates will cause a relatively high incidence of such autoimmune IDRs.

77 4.2. Introduction

Idiosyncratic drug reactions (IDRs) refer to a specific group of adverse drug reactions (ADRs) that do not occur in most patients within their therapeutic dose range and cannot be explained by the known pharmacological properties of the drug (11). IDRs can be very severe and even life-threatening in some cases, and therefore they represent a significant clinical problem. They also present a challenge to the pharmaceutical industry by adding an additional level of uncertainty to new drug development. At present it is impossible to predict IDRs largely because the mechanisms involved are unknown. Nevertheless, the delay between starting the drug and the onset of the adverse reactions suggests that most are immune-mediated (8). Animal models represent a very powerful tool for mechanistic studies (47). Penicillamine-induced autoimmunity in Brown Norway (BN) rats represents an important model for the mechanistic study of one type of IDR because it closely mirrors the autoimmune reactions that it causes in humans: both involve the presence of antinuclear antibodies, skin rash, a deposit of IgG along the glomerular basement membrane, arthritis, and weight loss (202). By definition, drug-induced autoimmunity is an immune-mediated IDR. Penicillamine-induced autoimmunity in rats is also idiosyncratic as it is strain specific – treatment of Lewis and Sprague-Dawley rats does not induce autoimmunity. Moreover, even though BN rats are highly inbred and syngeneic, autoimmunity only occurs in a little over 50% of male BN rats. One of the most important clinical characteristics of this model is the delay of about 2-3 weeks between starting the drug and the onset of first signature symptom, red ears. In addition, our previous studies demonstrated that the incidence and severity of autoimmunity can be influenced by a number of immunomodulators; for example, treatment with the IL-2 inhibitor, tacrolimus, completely prevents D-penicillamine-induced autoimmunity (155, 203-205). Ever since it was proposed in 1986, the Th1-Th2 hypothesis has been a major aspect of mechanistic theories of T cell-mediated diseases. For example, organ-specific autoimmune

78 diseases were thought to be driven by Th1 cells (206-209). A major part of the evidence supporting the role of Th1 cells in autoimmune diseases was obtained from studies of IL-12 (an essential cytokine in Th1 cell development) in several animal models of autoimmune diseases such as experimental autoimmune encephalomyelitis (EAE). However, the Th1 theory of organ specific autoimmunity was challenged because Th1 cytokines were often found to be protective. As a result, much of the attention has switched from IL-12 to IL-23, which contains a unique p19 subunit while sharing a p40 subunit with IL-12. Additional studies of the involvement of IL-23 in autoimmune diseases led to the discovery of a new helper T cell subset characterized by the production of a proinflammatory cytokine, IL-17, which were therefore called Th17 cells (210-213). In spite of many unknowns in the function of Th17 cells, significant progress has been made in characterizing this new T cell population. A large number of studies have found that a combination of TGF-β and IL-6 are required for the initial commitment of naïve T cells to Th17 cells (214-216); exposure to TGF-β in the absence of IL-6 leads to T regulatory cells, which are believed to play an important role in immune tolerance (217). In contrast, IL-23 was found to play an important role in maintaining the growth and expansion of Th17 cells. In addition, the role of transcription factors or signaling molecules such as STAT3, RORγt, and RORα in regulating the expression of IL-17 was discovered (218-221). Meanwhile, numerous studies in both humans and mice strongly suggest that the Th17 cell is a major determinant of the development of many kinds of autoimmune diseases (222, 223). Our previous studies demonstrated that penicillamine-induced autoimmunity in BN rats is T cell-mediated; therefore, this study designed to examine the involvement of Th17 cells in penicillamine- induced autoimmunity as a model of autoimmune IDRs.

79 4.3. Materials and Methods

Animals. Male BN rats (175-200 g) were purchased from Charles River (Montreal, Quebec, Canada) and doubly housed in standard cages in a 12:12 h light:dark cycle at 22 °C. The rats were given free access to standard rat chow (Agribrands, Purina Canada, Strathroy, Ontario, Canada) and tap water for a weeklong acclimatization period before starting an experiment. Chemicals, Kits, and Solutions. D-Penicillamine was purchased from Richman Chemical Inc. (Lower Gwynedd, PA). MACS anti-rat CD4 magnetic microbeads and magnetic columns were purchased from Miltenyi Biotec (Auburn, CA). All ELISA kits were purchase from R&D Systems (Minneapolis, MN). Luminex kits were purchased from Millipore (St. Charles MO). RNeasy Mini kits and OmniScript reverse transcriptase 285 kits were purchased from Qiagen (Mississauga, Ontario, Canada). Oligo (dT15) primers, RNAse inhibitor, and LightCycler SYBR Green I kits were all purchased from Roche (Montreal, Quebec, Canada) for quantitative real-time PCR (qRT-PCR). HPLC-purified primers for qRT-PCR were designed and obtained from Integrated DNA Technologies (Coralville, IA). D-Penicillamine Treatment. Rats were given D-penicillamine dissolved in tap water at a concentration of 1.0 mg/mL with an average water intake of 25 mL per day. The D- penicillamine solution was prepared fresh every two days because of the slow formation of inactive penicillamine disulfide. Unless otherwise indicated or unless signs of a severe autoimmune syndrome led to sacrifice of the animal, the duration of D-penicillamine treatment was 8 weeks. Determination of Serum IL-6 and TGF-β1. Blood samples were drawn via tail vein on day 0 and at the end of each week of penicillamine treatment. Blood samples were allowed to clot for 2 h at room temperature before centrifuging for 20 min at approximately 2000×g. Sera were aliquoted and stored at -80 ºC. IL-6 and TGF-β1 levels were determined by ELISA.

80 Phenotyping Splenic CD4+ T Cells by qRT-PCR. At the end of penicillamine treatment, splenic CD4+ T cells were isolated using rat CD4 magnetic microbeads according to the manufacturer’s instructions. Total RNA was isolated from CD4+ T cells using RNeasy mini kits as described by the manufacturer. RNA concentrations and purity were determined spectrophotometrically. RNA (0.5 μg) was reverse transcribed to cDNA from each sample. The expression level of Th17-related cytokine mRNAs was determined using real time quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) that was carried out with a LightCycler instrument (Roche). The basic PCR program for all samples was as follows: 95 ◦C for 10 min; 45 cycles of 95 ◦C for 5 sec, annealing temperature (primer-specific, 295 range 55-62 ◦C) for 5 sec, elongation at 72 ◦C for various times (due to difference in PCR product length, range 5-16 sec). Melting curve analysis was performed after amplification and carried out at a temperature transition rate of 0.2 ◦C/sec up to 95 ◦C. Data were normalized by calculating the absolute concentration of the cDNA of interest relative to absolute GAPDH concentration in each cDNA sample. Profiling Serum Cytokines/Chemokines. Male BN rats (20) were treated with penicillamine and blood samples were drawn via the tail vein on day 0 and at the end of each week of treatment. Serum was isolated as described above. A Luminex assay of 24 cytokines/chemokines (Eotaxin, GM-CSF, G-CSF, MCP-1, Leptin, MIP-1α, IFN-γ, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-9, IL-10, IL-12p70, IL-13, IL-17, IL-18, IP-10, GRO/KC, RANTES, TNF-α, VEGF) was performed using the protocol provided by the manufacturer to determine the overall pattern of serum cytokines/chemokines over the course of penicillamine treatment.

81 4.4. Results

Serum Levels of IL-6 during Penicillamine Treatment. In the first group of penicillamine treated rats (n=3), 2 rats developed an autoimmune syndrome (rat # 2 (D2): day 18; rat # 3 (D3): day 20). Serum IL-6 was significantly increased around the time that the animals developed clinical signs of autoimmunity while the serum IL-6 levels in nonsick and control animals remained at non-detectable levels throughout penicillamine treatment (Figure 22). To confirm the association of IL-6 and autoimmunity, we repeated experiments again (n=4 for each group). Similarly, an increase of serum IL-6 was found in the two sick animals (rats # 1 and 3) shortly before the onset of autoimmunity but not in the nonsick and control rats (Figure 23). However, the increase in IL-6 occurred earlier in rat 4 and it returned back to basal level one week after the onset of autoimmunity (day 28). Although the rat kept scratching itself resulting in local dermatitis, its serum IL-6 remained at a basal level after day 28. In addition, the level of serum IL-6 dropped to undetectable levels in rat 6 one week after penicillamine treatment was stopped. Phenotype of Splenic CD4+ T Cells from Treated Animals. By the end of treatment, rats # 1, 2, and 3 were sacrificed and total RNA was isolated from purified splenic CD4+ T cells. Expression of IFN-γ, IL-17, IL-21, IL-4, and IL-10 mRNAs was determined (Figure 24). The results showed that IL-21 increased markedly in all three treated animals. In contrast, a two-fold increase of IL-17 was only observed in the two sick animals. There was no change in IFN-γ. IL-4 was also increased in all treated animals, while IL-10 was just slightly increased in sick animals. Serum Cytokine/Chemokine Pattern during Penicillamine Treatment. During 5 weeks of penicillamine treatment, 15 out of 20 rats developed autoimmunity, and in most cases the time to onset was between 14 and 21 days. The body weight and cumulative incidence are shown in Figure 25. Comparison of total splenocytes between sick and non- sick animals is shown in Figure 26. Of all 24 cytokines/chemokines, serum levels of IL-6, TGF-β1, IL-17, IL-2, IL-9, IL-10, IL-13, IL-18, GRO/KC, MCP-1, leptin, and RANTES

82 were found to be significantly different between sick and non-sick animals (Figure 27), while other analytes were either non-detectable (i.e. IL-1α) in all treated animals or there was no difference between sick and non-sick rats (e.g. IFN-γ).

Figure 22. Serum concentration of IL-6: D-penicillamine vs. control (n=3). Out of three penicillamine treated rats, two developed autoimmunity (D2 and D3). Significant serum IL-6 levels were only detected in the two sick animals, not in non-sick (D1) and control animals (C1, C2, and C3).

Figure 23. A repeat of serum IL-6 determination in penicillamine treatment. Out of four penicillamine treated rats, two developed autoimmunity (D1 and D3). Penicillamine treatment was discontinued at day 35. Serum IL-6 was non-detectable in four control animals over the course of follow-up (data was not list here).

Figure 24. Phenotype of splenic CD4+ T cells from penicillamine-treated rats at the end of penicillamine treatment. The 3 penicillamine treated animals (D1, 2, 3) were those presented in figure 22.

Figure 25. Changes of body weight and cumulative incidence of autoimmunity.

700

** p < 0.01 600 **

500 400 300 200 100

0 Sick Non-sick

Figure 26. Comparison of the number of splenocytes between sick (n=15) and non-sick rats (n=5). Splenocyte counts were determined on the last day of penicillamine treatment.

Figure 27. Serum cytokine/chemokine pattern: Sick (n=15) vs. Non-sick (n=5).

Table 9. Primer sequences for qRT-PCR Gene Forward Primer (5’-3’) Reverse Primer (5’-3’) IFN-γ ATATCTGGAGGAACTGGCAAAA TAGATTCTGGTGACAGCTGGTG Interleukin 4 TCAACACTTTGAACCAGGTCAC GCAGCTTCTCAGTGAGTTCAGA Interleukin 10 AGGACCAGCTGGACAACATACT TCATTCATGGCCTTGTAGACAC Interleukin 13 ACAGGACCCAGAGGATATTGAA AACTGAGGTCCACAGCTGAGAT Interleukin 17 TGGACTCTGAGCCGCATTGA GACGCATGGCGGACAATAGA Interleukin 21 CGAAGCTTTTGCCTGTTTTC GAAGGGCATTTAGCCATGTG

89 4.5. Discussion

There was a consistent increase in serum IL-6 shortly before the onset of penicillamine- induced autoimmunity. This very good concordance between an elevation of serum IL-6 and the development of autoimmunity was further confirmed by the cytokine profiling study (Figure 27), which suggests an immunopathological role of IL-6 in penicillamine-induced autoimmunity in BN rats. In addition, the serum level of TGF-β1, another Th17 pathway related cytokine, was significantly increased in sick animals at early time points before animals developed autoimmunity, while there was no change in non-sick rats. Since IL-6 and TGF-β are the driving force for the differentiation of CD4+ naïve T cells into Th17 cells, the elevation of serum IL-6 and TGF-β suggests an induction of Th17 cells in response to penicillamine treatment. Meanwhile, a 2-fold increase of IL-17 mRNA only in sick animals provided further evidence in support of the involvement of Th17 cells in this animal model (Figure 24). In contrast, a 4-fold increase of IL-21 mRNA in splenic CD4+ T cells was found in all treated animals, which indicates T cell activation, although the formed T cell clones were not pathogenic in non-sick animals. Studies have shown that a combination of IL-21 and IL-4 drives T cells toward the Th2 pathway, while in the presence of IL-6, Th17 cells predominate (224-226). This suggests that in rats # 2 and #3, the Th2/Th17 balance of CD4+ T cells was tilted toward Th17 cells. More direct evidence of Th17 cell involvement was the elevated serum level of IL-17 in sick animals (Figure 27). The development of autoimmunity in this model can be divided into two stages based on the observed cytokine patterns. During the initiation of autoimmunity and prior to clinical manifestations there was an increase in IL-17, TGF-β1, and IL-13. Then coincident with the development of clinical autoimmunity there is an increase in IL-6, IL-9, IL-10, IL- 18, MCP-1, GRO/KC, leptin, and IL-2. Many of these cytokines/chemokines such as MCP- 1 and leptin also appear to be involved in pathogenesis of other autoimmune diseases. In addition to IL-6 and TGF-β, IL-9 and IL-18 are also closely associated with Th17 cells. IL- 18 is mainly produced by macrophages and is able to activate Th17 cells to produce IL-17 in

90 synergy with IL-23 (227, 228). IL-9 is predominantly produced by Th17 cells and regulates the balance between Th17 cells and regulatory T cells (229). IL-10 is also produced by Th17 cells, possibly as part of a system to limit inflammation (229-233). Intriguingly, we observed a very good correlation between a sharp peak of serum IL-13 on day 7 and the later development of autoimmunity. A recent study found that Th17 cells have increased expression of IL-13Rα1 that mediates the regulatory effect of IL-13 on Th17 cells (234). Furthermore, this regulation appeares to be specific to Th17 cells because IL-13 did not increase Th1 and Th2 cytokines (IFN-γ and IL-4, respectively). Hence, a marked increase in IL-13 may represent a biomarker for the induction of an autoimmune reaction. In summary, the cytokine profile in the penicillamine model provides strong evidence for the involvement of Th17 cells in the pathogenesis of this autoimmune IDR. It remains to be determined whether a similar profile exists in drug-induced autoimmunity in humans and whether the ability of a drug to induce cytokines such as IL-21 in animals is a biomarker for their ability to induce autoimmunity in humans.

Acknowledgement. J.U. holds a Canada Research Chair in Adverse Drug Reactions. This research work was supported by grants from the Canadian Institutes of Health Research.

CHAPTER 5 CYTOKINE AND AUTOANTIBODY PATTERNS IN ACUTE LIVER FAILURE

Jinze Li, Xu Zhu, Feng Liu, Ping Cai, Corron Sanders, William M. Lee and Jack Uetrecht Work from this chapter will be submitted for publication.

Author contributions: Jinze Li was the primary author who performed Luminex and ELISA analysis of serum samples and did most of data analysis with the assistance of Xu Zhu and Feng Liu; Ping Cai did the BAFF test; Corron Sanders and William Lee were the key persons from the Acute Liver Study Group in Texas USA, recruiting patients and collecting biological samples.

92 5.1. Abstract

The mechanisms of idiosyncratic drug-induced liver injury (IDILI) are still a matter of dispute. Some of the characteristics of reactions that have been classed as metabolic idiosyncrasy could also be characteristics of an immune-mediated reaction with an autoimmune component. We quantified a number of cytokines, chemokines, and autoantibodies in the serum of patients with acute liver failure due to IDILI and compared the values to those from patients with acetaminophen-induced liver failure, and with acute liver failure due to viral hepatitis. We paid special attention to cytokines such as IL-17 that are associated with Th17 cells and autoimmunity. We found that IL-17 was elevated in about 60% of patients with IDILI; however, we were surprised to find that IL-17 was also elevated in many patients with acetaminophen-induced liver failure as well as in a few patients with viral hepatitis. It is unlikely that acetaminophen-induced liver failure is mediated by the adaptive immune system, and it is now known that IL-17 is also produced by cells of the innate immune system. Although the levels of other cytokines such as IL-21, which are also produced by Th17 cells, were higher in patients with IDILI, there was overlap with acetaminophen DILI. There was also a higher frequency of various autoantibodies (antinuclear antibodies or anti-myeloperoxidase antibodies) in patients in the IDILI group; however, autoantibodies were not detected in most patients. These data provide a general picture of the cytokine/chemokine profile in patients with various types of liver failure. The pattern varies from patient to patient, probably reflecting differences in the underlying disease mechanism; however, interpretation is complicated by the fact that the same cytokine can originate from more than one type of cell.

93 5.2. Introduction

Idiosyncratic drug-induced liver injury (IDILI) has been categorized as being due to either immune idiosyncrasy or metabolic idiosyncrasy. The designation of immune idiosyncrasy is based on the presence of fever, rash, eosinophilia, accelerated onset with rechallenge, and anti-drug antibodies (235). The designation of metabolic idiosyncrasy is based on the event being rare, unpredictable, the lack of allergic features, and typically a long latency period, sometimes more than a year. Although genetic polymorphisms in a metabolic pathway have been suspected to be involved, there are no clearly defined cases where a polymorphism in a metabolic pathway is sufficient to explain the idiosyncratic nature of an IDILI reaction. Certainly not all immune-mediated events have the characteristics that are used to classify IDILI as being immune idiosyncrasy. The most important characteristic that argues against an immune mechanism is the lack of a rapid recurrence on rechallenge, but many immune-mediated reactions, especially drug-induced autoimmunity, do not always occur more rapidly on rechallenge (174). Many drugs that cause IDILI classified as metabolic idiosyncrasy, such as isoniazid, also cause a lupus-like autoimmune syndrome, and the autoimmune IDILI caused by minocycline is characterized by a relatively long latency period (236). It is possible that many cases of IDILI classed as metabolic idiosyncrasy have an autoimmune component even though they are not associated with the classic autoantibodies associated with autoimmune hepatitis. Recently, a new subtype of helper T cells, Th17 cells, has been identified and characterized by a set of proinflammatory cytokines including IL-17, IL-21, and IL-22 (237, 238). The IL-17 receptor is expressed on various epithelial tissues, and Th17 cells are therefore considered a very crucial messenger between immune system and tissues (223, 239). Growing evidence suggests that Th17 cells play a very important role in pathogenesis of many kinds of autoimmune syndromes, especially in organ specific autoimmunity (240-243). Our recent studies also suggested the involvement of Th17 cells in the animal model of penicillamine-induced idiosyncratic

94 autoimmunity, which includes hepatotoxicity (unpublished data). In this study, we set out to investigate the cytokine pattern of patients with various forms of acute liver failure (ALF) to determine if different causes of ALF are associated with characteristic cytokine patterns, and in particular, if any have a Th17-related pattern that suggests an autoimmune component.

95 5.3. Patients and Methods

Patients. The Acute Liver Failure Study Group has studied prospectively more than 1,400 cases of ALF over 11 years and obtained detailed data as well as serum and DNA samples on most of these patients, carefully stored at -80o. ALF is defined as an acute hepatic illness that leads to coagulopathy with an international normalized ratio (INR) ≥ 1.5 accompanied by any degree of hepatic encephalopathy in less than 24 weeks. The focus in the present study was patient samples and data from the ALFSG registry that were diagnosed as having IDILI by the site investigator after a standard set of evaluations; patients with ALF caused by acetaminophen (APAP) and viral hepatitis (either A or B) were used for comparison purposes, as well as sera obtained from a cohort of 10 patients with chronic hepatitis C. Informed consent was obtained from next of kin since patients by definition had altered mentation. In addition to clinical samples, information on each case was available for review. Determination of Serum Cytokine/Chemokine Profile by Luminex or ELISA. Serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, and alkaline phosphatase were determined by the treating hospitals and recorded in the case report forms as noted above. Serum levels of 21 cytokines/chemokines (IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12 (p40), IL-13, IL-15, IL-17, Eotaxin, TNF-α, GM-CSF, IFN-γ, IP-10, MCP-1, and MIP-1α) were determined using a human cytokine/chemokine milliplex luminex kit (Millipore St. Charles, Missouri USA). Serum concentrations of IL-21 were determined by ELISA kit according to the manufacturer’s instructions (eBioscience, CA USA). Serum concentrations of B-cell activating factor (BAFF) were determined with an ELISA kit from R&D system. (Minneapolis, MN) Determination of Serum Autoantibodies. The BINDAZYME ANA screen enzyme immunoassay kit (Binding Site Ltd, Birmingham UK) was used to collectively detect total antinuclear antibodies (ANAs) against dsDNA, histones, SSA/Ro (60 and 52kD), SSB/La, Sm, Sm/RNP, Scl-70, Jo-1, and centromeric antigens. This kit only determines the presence

96 of these antibodies, but without further testing, it is impossible to know which of these autoantibodies is elevated. A human anti-MPO antibody ELISA kit (IMMCO diagnostics Inc, Buffalo NY, USA) was used for semiquantitation of antibodies to myeloperoxidase following the protocol provided by the manufacturer.

97 5.4. Results

Overall Patient Cohort. For this initial study, sera from a total of 70 ALF patients were utilized, obtained between study days 1 and 6. Patients were randomly selected for study from the overall registry and were balanced between the following for categories: IDILI (n=39), acetaminophen (APAP, n=21) and hepatitis A (n=5) and B (n=5); ten patients with chronic hepatitis C, not receiving interferon treatment, were considered as a separate positive disease control group. Biochemical Parameters. The time between the onset of initial symptoms and hospitalization (a measure of acuteness) and biochemical parameters are summarized in Figure 28. In the IDILI patients, a wide variety of medicines were involved, ranging from herbal medicines to drugs such as diclofenac and troglitazone that are well known to cause IDILI. Serum Cytokine Profile of Patients. The lowest concentration of cytokine/chemokine for the standard curves was 3.2 pg/mL so any concentration lower than that should be considered non-detectable. In normal individuals the levels of cytokines such as IL-4, IL-6, and IL-17 are less than 2 pg/mL and so detectable levels can be considered abnormal. In contrast, normal levels of IL-21 are 466 ± 90. A complete list of the cytokine/chemokine data in the current study is presented as Supplemental Data. IL-17 was detectable in ~ 60% of APAP and IDILI patients and a lower fraction in patients with viral hepatitis (Figure 29). Serum levels of IL-6, a critical cytokine for the differentiation of naïve T cells into Th17 cells, were elevated in most patients except those with hepatitis C (Figure 29). The mean IL-21, a Th17 cytokine, and IL-1α levels were significantly higher in IDILI patients than other ALF patients (Figure 29). Serum IP-10/CXCL10, reported to be dramatically elevated in both type 1 diabetes and autoimmune liver diseases, was also highest in IDILI patients (Figure 29). Serum levels of BAFF were also significantly higher in IDILI patients than APAP patients (3106 ± 447.5 vs. 1609 ± 276.8) (Figure 30). In contrast, the mean levels of MCP-1 and IL-15 were higher in the APAP group (Figure 29).

98 In the IDILI group, IL-17 levels significantly correlated with several other cytokines; in contrast, correlations with other cytokines in the APAP group were far fewer (Table 10). Antinuclear Autoantibodies and Antimyeloperoxidase Antibodies. Antinuclear autoantibodies (ANA) were detectable in 14 out of 39 IDILI patients (33%) and markedly elevated in 3, but only elevated in 1 out of 21 APAP patients (Figure 31). Three of the patients with viral hepatitis also had significant elevations of ANA. As for serum levels of anti-MPO autoantibodies, the mean was also significantly higher in DILI patients than APAP patients, but of all the patients, only one IDILI patient had a marked elevation of anti-MPO antibodies (Figure 32). Neither ANA nor anti-MPO antibodies correlated significantly with any serum cytokine.

30 *** 8000 *** A B 7000 ALT *** AST *** p < 0.0001 6000 p < 0.0001 20 *** 5000 4000

3000 10 2000

1000

0 0 APAP IDILI APAP IDILI Hepatitis A Hepatitis B

25 ** 300 C D *** *** *** p < 0.0001 20 p < 0.05 * p < 0.01 ** p < 0.0001 200 15 *** *

10 100

0 0 APAP IDILI Hepatitis A Hepatitis B APAP IDILI Hepatitis A Hepatitis B

Figure 28. Biochemical parameters of liver failure patients. Statistical analysis was done between APAP and every other group in figure B and C, between IDILI and APAP in figure D.

100 1000 IL-17 100000 IL-21 * * p < 0.05 10000 100

1000

100

1 10 APAP IDILI Hepatitis A Hepatitis B Hepatitis C APAP IDILI Hepatitis AHepatitis BHepatitis C

10000 IL-6 10000 IL-1alpha

** ** p < 0.01 1000 1000

100 100

10 10

1 1 APAP IDILI Hepatitis A Hepatitis B Hepatitis C APAP IDILI Hepatitis A Hepatitis B Hepatitis C

100000 IP-10 100000 MCP-1 *** *** p < 0.0001 * * p < 0.05 10000 10000

1000 1000

100 100

10 10 APAP IDILI Hepatitis AHepatitis BHepatitis C APAP IDILI Hepatitis A Hepatitis B Hepatitis C

1000 IL-15 1000 IFN-gamma

*** *** p < 0.0001 100 100

10 10

1 1 APAP IDILI Hepatitis AHepatitis BHepatitis C APAP IDILI Hepatitis A Hepatitis B Hepatitis C

Figure 29. Serum cytokine/chemokine comparison between patient groups.

101

15000 **

** p < 0.01 10000

5000

0 APAP IDILI

Figure 30. Serum levels of B-cell activation factor (BAFF).

100 * * p < 0.05

0 APAP IDILI Hepatitis A Hepatitis B Hepatitis C

Figure 31. Serum levels of ANA.

102

3500 * * p < 0.05 3000

2500

2000

1500

1000 500

0 APAP IDILI Hepatitis AHepatitis BHepatitis C

Figure 32. Serum levels of anti-MPO antibodies.

103 Table 10. Correlation of IL-17 with other analytes

Anti- IL- IL- IL- IL- TNF- GM- IL- IL-12 IL- IL- IL- IL- IL- IL- IL- IP- MCP- MIP- ANA IL-2 Eotaxin IFN-γ MPO 21 6 4 5 a CSF 10 (p40) 13 15 1α 1β 3 7 8 10 1 1α

APAP ns ns ns ns ∗∗ ∗ ns ∗ ns ns ∗∗∗ ns ns ∗ ns ns ns ns ns ns ns ns ∗

IDILI ns ns ns ∗ ∗∗∗ ∗ ∗∗ ns ns ∗∗∗ ∗∗∗ ∗∗ ∗ ∗∗ ns ∗∗ ∗ ∗ ∗∗ ∗ ∗ ns ∗∗∗

HAV ns ns ns ns ns ns ns ns ns ns ∗ ns ns ns ns ns ns ns ns ns ns ns ns

HBV ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns

HCV ns ns ns ∗ ns ns ns ns ∗ ns ∗ ns ns ∗∗ ∗ ns ∗ ns ns ∗ ns ns ∗∗

Ns: no significance; ∗: p<0.05; ∗∗: p<0.01; ∗∗∗: p<0.0001.

104 Table 11. IL-17, IL-21, IL-6, IP-10, ANA, and anti-MPO, in IDILI patients IL-21 IL-17 IL-6 ANA Anti-MPO IP-10 Patient ID Cause of liver failure (pg/ml) (pg/ml) (pg/ml) (EU/ml) (EU/ml) (pg/ml) 12-015-04 ISONIAZID 191 <3.2 53 71 914 169 13-010-04 ISONIAZID 651 <3.2 9 <10 1263 294 13-088-01 ISONIAZID 453 <3.2 33 12 982 3549 13-133-02 ISONIAZID 211 <3.2 14 15 1684 96 14-005-02 ISONIAZID 1789 4.3 19 <10 1217 297 14-067-04 INH, PZA, RIFAMPIN 305 <3.2 60 <10 1214 276 12-020-01 PZA/RIFAMPIN 1864 <3.2 25 <10 1805 403 10-009-06 RIFAMPIN 1240 44.6 76 <10 1043 1162 13-015-02 NITROFURANTOIN 2122 <3.2 18 12 1190 130 13-067-04 NITROFURANTOIN 857 4.8 16 <10 1298 228 13-080-02 BACTRIM 1726 4.9 263 <10 930 1940 10-025-01 SULFADIAZINE 842 <3.2 1127 <10 1032 3419 11-021-04 TRIMETHOPRIM/SULFA 849 12.5 40 63 1166 407 15-002-02 BROMFENAC 1374 <3.2 99 <10 1339 95 13-003-03 TROGLITAZONE 128 4.2 13 <10 1145 188 13-047-01 PHENYTOIN 918 <3.2 197 <10 1016 >10000 14-037-02 PHENYTOIN 1742 5.5 664 11 1591 6369 13-129-02 PROPYLTHIOURACIL 1074 10.2 133 65 795 420 13-064-01 ALLOPURINOL 857 107.8 1042 <10 853 >10000 11-072-04 GEMTUZUMAB 781 3.3 1437 <10 1239 449 11-100-02 KAVA KAVA PHENYTOIN 854 3.3 154 <10 1611 73 13-016-01 PRAVASTATIN 1853 3.5 321 17 1568 455 14-007-04 CERIVASTATIN 4457 <3.2 59 <10 1365 573 15-039-03 ROSIGLITAZONE 1174 <3.2 615 <10 1216 294 13-145-04 DICLOFENAC 496 3.5 278 30 3113 >10000 10-061-01 CIPROFLOXACIN 250 5.8 24 <10 1272 868 11-153-02 ETODOLAC 906 8.2 32 25 1239 326 13-007-03 ZAFIRLUKAST 796 8.6 76 <10 1938 339 11-081-04 DISULFIRAM 790 22.6 70 <10 1101 532 10-027-06 QUETIAPINE 972 39.7 94 11 1347 488

105 14-042-04 HORNY GOAT WEED 2165 <3.2 228 15 1105 341 15-049-02 B6 729 3.3 64 <10 2273 92 16-007-06 ISOFLURANE 858 40.1 935 <10 1783 332 14-078-06 UNKNOWN DRUG 464 65.6 122 <10 2290 530 13-147-02 TAK-559 713 111.0 279 <10 1585 922 13-176-01 UNKNOWN DRUG 2408 <3.2 3.3 12 987 1829 13-039-02 THERMA SLIM 5623 <3.2 37 <10 1282 332 15-041-02 BLACK COHOSH 775 <3.2 37 <10 1577 304 HERBAL MEDS 13-004-02 552 12.2 1140 13 1603 457 (MULTIPLE)

106 5.5. Discussion

Acute liver failure represents the most severe form of liver injury and frequently results in a fatal outcome unless transplantation can be performed. As such, it would be expected that massive inflammation results and this indeed is the principal finding on hepatic biopsy in such patients, along with widespread destruction of hepatocytes. In the present study we focused on cytokine responses in ALF due to drugs. We suspected that markers of activation of the adaptive immune system would be evident and, in particular, we wanted to determine if some IDILI that has been classed as metabolic idiosyncrasy would have evidence of an autoimmune component as evidenced by elevated IL-17 levels. While IL-17 levels were indeed high in many IDILI patients, this was not consistently observed, and we were somewhat surprised to find that IL-17 levels were also elevated in several patients with APAP-induced ALF. However, in addition to Th17 cells, a variety of other cells have recently been found to produce IL-17 including neutrophils, CD8+ T cells, NKT cells, γδ T cells, macrophages, and NK cells (244, 245). Given the acute nature of APAP- induced ALF, it is unlikely that the source of IL-17 in these patients was Th17 cells, which usually require days or even weeks to differentiate and expand from naïve T cells to pathogenic cells. High doses of APAP cause significant cell damage, which is likely to lead to an innate immune response to cleanup the dead and dying cells. In contrast, the level of IL-17 in IDILI patients correlated with the levels of several other cytokines, possibly because in these patients it is measure of activation of the adaptive immune system in which Th17 cells are the major cellular source of IL-17. Of all characteristic cytokines released by Th17 cells, IL-21, which is mainly produced by CD4+ T cells, has been shown to be critical for the development of autoimmunity (246, 247). Therefore, it should be a better measure of Th17 cell-mediated autoimmunity. IL-21 was higher, on average, in the IDILI patients, but only markedly elevated in a few patients. The greatest elevation of IP-10, a cytokine associated with autoimmune liver diseases and type I diabetes(248, 249), was also found in the IDILI patients. BAFF was also elevated in a significantly greater number of IDILI

107 patients. BAFF is a critical factor for the survival and maturation of B cells and has been shown to be involved in the pathogenesis of many kinds of chronic autoimmune diseases such as SLE, rheumatoid arthritis, and autoimmune hepatitis (250-252). Furthermore, a recent study has shown that silencing BAFF gene by shRNA was able to suppress the generation of Th17 cells leading to amelioration of autoimmune arthritis (253). Therefore, the significantly elevated serum levels of BAFF in IDILI patients provide additional evidence to support the autoimmune nature of some IDILI cases. In contrast, MCP-1 and IL-15 were elevated more in the APAP patients than other groups, presumably a reflection of innate immune system activation in response to liver damage (254, 255). Another parameter that would suggest an autoimmune component is the presence of autoantibodies. However, part of the differential diagnosis for ALF is idiopathic autoimmune hepatitis, and therefore a screen for ANA and liver/kidney microsome type 1 antibodies should be part of the workup of these patients and these patients do not have a classic picture of autoimmune hepatitis. In spite of this we found that a few of the IDILI patients had significant elevations of ANA or anti-MPO antibodies. The drugs in this study that were associated with marked elevations in ANA were propylthiouracil, trimethaprim- sulfa, and isoniazid, drugs known to cause a lupus-like syndrome (Table 11). A marked elevation of anti-MPO antibodies was observed in a patient with diclofenac-induced ALF (Table 11). There is no way of knowing whether these autoantibodies were related to the IDILI that occurred. Obviously ANA and anti-MPO antibodies only represent a tiny fraction of the total range of possible autoantibodies that these patients might have, and we are working on a much more comprehensive screen for autoantibodies that may be useful in the future. In addition, there could be auto-reactive T cells in the absence of autoantibodies. Even though there were differences in the parameters measured between the different categories of ALF, there was a large degree of overlap, and there is no simple parameter that could be used diagnostically or that would provide solid evidence that specific cases of IDILI have an autoimmune component. An important limitation of the data is that the patients were not part of a controlled study, and therefore serum was obtained at different

108 time points in the course of the disease and the severity of liver injury was different in different patients. In addition, it is likely that there are significant differences in the mechanism of IDILI in different patients, and it is striking that the exact profile appeared different in each individual patient. However, these data do provide a rough picture of the cytokine profile of patients with different types of ALF.

Acknowledgement. J.U. holds a Canada Research Chair in Adverse Drug Reactions. This research work was supported by grants from the Canadian Institutes of Health Research.

109

CHAPTER 6

OVERALL CONCLUSIONS AND FUTURE DIRECTIONS

110 6.1. SUMMARY

Advancing the mechanistic understanding of IDRs is a prerequisite for any significant progress in identifying which drug candidates are likely to be associated with a high risk of IDRs during drug development and preventing these events in clinical practice. Unfortunately, despite a general agreement on the involvement of immune system in the pathogenesis of many IDRs, our current knowledge of IDRs is insufficient to allow their prediction and prevention. Essentially the only way to test the many potential mechanistic hypotheses of IDRs is with valid animal models. We have used the animal model of penicillamine-induced autoimmunity in BN rats to investigate the mechanism of this idiosyncratic drug-induced autoimmune syndrome. The previous studies in our lab have not only demonstrated several aspects of the immunological pathogenesis of penicillamine- induced autoimmunity, but also and more importantly, narrowed our search for initial events in the interaction between penicillamine and macrophages. This is based on the observation of infiltrations of activated macrophages in several organs 96 hours after drug treatment. This raised the obvious question of exactly how penicillamine led to the activation of macrophages. The hypothesis that reversible Schiff base formation involving the reaction of amines on T cells and aldehydes on APCs represents a basic activation pathway for these cells opened up the possibility that the irreversible reaction of penicillamine with aldehydes on macrophages could lead to generalized macrophage activation, and in some individuals, an autoimmune syndrome. Furthermore, hydralazine and isoniazid, which are hydrazines and also react irreversibly with aldehydes, are also associated with idiosyncratic drug- induced autoimmunity, and this further strengthens the hypothesis that such an irreversible interaction with aldehydes is the basis for the autoimmunity induced by these drugs. We designed experiments in a step-wise fashion to test each essential element of this hypothesis. First of all, two specific aldehyde reactive reagents that are tagged with biotin (ARP and hydrazide) were used to examine the existence of aldehydes on the surface of spleen macrophages from BN rats. Although when Rhodes proposed the idea that Schiff base

111 formation between molecules on macrophages and T cells led to cell activation he always referred to the active groups involved as an aldehyde and amine, ketones undergo similar reactions as aldehydes and therefore it is possible that the actual group is a ketone. Therefore, when the term aldehyde is used in this thesis it should be understood that a ketone is also a possibility even though they are somewhat less reactive than aldehydes; however, other carbonyl-containing groups such as esters and amides do not form Schiff bases. The finding that both ARP and hydrazide reagents bound preferentially to macrophages with similar binding curves was the first strong evidence for the existence of membrane-localized aldehyde-containing molecules specific to macrophages. This binding was partially inhibited by preincubation with hydralazine or penicillamine, which are known to bind to aldehydes. It appears that there is a dynamic turnover of the aldehyde-containing molecules because the observed binding, which is only detected if the molecule is on the surface of the cell, decreases with time after the cells are washed. Hydralazine had a greater inhibitory effect than penicillamine on ARP binding and this could be due to requirement of a specific orientation of the molecule to produce a five-membered ring. If this slows the rate of irreversible binding it could result in a lower fraction of molecules reacting during the period of incubation. The experiment in which the biotin-penicillamine adduct bound to spleen cells provided direct evidence to support the hypothesis that penicillamine binds to all splenocytes at high penicillamine concentrations but selectively to macrophages at low penicillamine concentrations. Additionally, ARP binding to cell surface aldehydes was also observed in the murine RAW 264.7 macrophage cell line with similar kinetics as those observed with rat macrophages. Overall, this series of studies suggest that binding to membrane aldehydes may represent a general event in the pathogenesis of autoimmunity caused by a class of drugs that are capable of covalently binding to aldehydes on immune cells, particularly macrophages. Intriguingly, we found significantly less binding of ARP to splenic macrophages from Lewis and Sprague Dawley rats, which are completely resistant to penicillamine-induced autoimmunity. The different level of membrane aldehydes could

112 partially explain the different response in different rat stains. Since our previous studies found activation of macrophages in penicillamine-treated rats well before the onset of clinical autoimmunity, the question becomes whether the observed binding of penicillamine to macrophages is directly responsible for this activation. The second part of this thesis examined the biological consequences of the covalent binding. First, a microarray study found an increase in the expression of mRNAs that code for macrophage activation biomarkers such as CD14, CD163, IL-1β, IL-15, etc. in splenic macrophages 6 h after penicillamine treatment. This clearly indicated that penicillamine very rapidly led to the activation of macrophages. Consistent with the in vivo data, RAW 264.7 cells were also stained by ARP indicating the existence of aldehyde-containing molecules on their cell membrane. More importantly, incubation of these cells with penicillamine stimulated cytokine production (IL-6, IL-23, TNF-α) in the absence of other cells, which further supported the hypothesis that activation of macrophages was a consequence of penicillamine binding. In addition, IL-1β and IL-15 are essential cytokines for NK cell differentiation and development, and the observation of a 2-fold increase in mRNA expression of IFN-γ in splenic NK cells indicated that there were downstream consequences of macrophage activation. This regulatory loop could be relevant to generalized activation of the immune system resulting in autoimmunity. Moreover, hydralazine and isoniazid were found to have similar activation effects on Raw 264.7 cells. In short, the observation that penicillamine, hydralazine, and isoniazid, which are very different molecules but have in common the ability to bind to aldehydes, all activate macrophages and all induce a lupus- like autoimmune syndrome in humans provides strong support for the hypothesis that the mechanism of this IDR involves covalent binding to aldehydes on macrophages. The covalent binding of these molecules to aldehydes and its activation effect on macrophages provides a mechanism of communication between macrophages and T cells and the induction of some types of IDRs. Nevertheless, there are still many puzzles that need to be solved in order to gain a full understanding of the exact mechanism of penicillamine-induced autoimmunity. The primary question is the identity of the aldehyde-

113 bearing molecules, especially the ones that are involved in signal transduction pathways. By employing biotin-avidin chromatography and mass spectrometry, we were able to identify a list of aldehyde-containing proteins that, in theory, could be the proteins to which ARP and penicillamine bind. However, these are just candidates because the bands that were visualized by ARP may be contaminated with other proteins. In addition to identification of the aldehyde-containing proteins, it is important to locate the exact binding site of penicillamine on proteins. If, for example, they are produced by the reaction of protein with aldehyde-containing molecules produced by oxidative stress, oxidative stress could be a significant risk factor for this type of IDR. The second major part of this thesis was to define and characterize the T cell response in penicillamine-induced autoimmune disease with a focus on Th17 cells, which have been shown in a number of studies to be the main pathogenic immune cells in several autoimmune diseases. We found a very good concordance between development of autoimmunity and elevated serum levels of IL-6, which is the most important cytokine for the initial differentiation of naïve T cells into the Th17 lineage. In addition, increased mRNA expression of IL-17 in splenic CD4+ T cells was only detected in rats that developed autoimmunity. IL-17 is one of the signature cytokines of Th17 cells, and therefore Th17 cells appear to play a key role in the mechanism of penicillamine-induced autoimmunity. A more comprehensive serum cytokine/chemokine pattern was obtained during the development and progression of autoimmunity. Of the 24 analytes tested, several Th17 pathway-related cytokines/chemokines were found to be at much higher concentrations in sick animals than in non-sick animals at either early (i.e. TGF-β, IL-17) or late (i.e. IL-10, IL-9, IL-17) time points. This provides additional support for the involvement of Th17 cells in the pathogenesis of penicillamine-induced autoimmunity. However, due to the recent discovery that IL-17 is also produced by cells of the innate immune system, in order to directly associate an increased serum IL-17 with Th17 cells, a more specific analysis (i.e. flow cytometry, ELISPOT) of cells to determine the source of IL-17 and other cytokines is required. The primary benefit of this type of cytokine profile is the establishment of

114 biomarkers that might predict that a drug candidate will cause a significant incidence of autoimmunity. With compelling evidence that Th17 cells are involved in penicillamine-induced autoimmunity in which liver injury is also observed, we set out to test whether some cases of drug-induced liver injury that do not have classic features of an adaptive immune response might represent a form of drug-induced autoimmunity. We quantified 26 cytokines/chemokines, several autoantibodies, and BAFF in serum samples from 39 patients with idiosyncratic drug-induced liver failure, 31 patients with acetaminophen-induced acute liver failure, and patients with viral hepatitis A, B, and C. We paid special attention to IL-17 and IL-21, which are produced by Th17 cells. The average serum concentrations of IL-21, autoantibodies, and BAFF were highest in IDILI patients, which was consistent with the hypothesis that at least some of the idiosyncratic cases probably involved an autoimmune component. However, many patients with acetaminophen-induced liver failure, which is very unlikely to represent an autoimmune reaction, had similar elevations of serum IL-17. Therefore, it is likely that in the cases of acetaminophen-induced liver failure, the source of IL-17 was the innate immune system with the result that serum levels of cytokines such as IL-17 do not provide a diagnostic marker of autoimmunity. Hence, in the future, we will have to obtain fresh blood from patients in order to determine the phenotype of the cells that are the source of IL-17 and related cytokines.

115

Figure 33. Working hypothesis of the pathogenesis of penicillamine-induced autoimmunity.

116 6.2. IMPLICATIONS AND FUTURE DIRECTIONS

Having demonstrated covalent binding of penicillamine to aldehyde-bearing surface molecules on macrophages and subsequent macrophage activation, this thesis also raised several critical questions that need to be investigated in order to further our mechanistic understanding of how idiosyncratic drug-induced autoimmunity is initiated and progresses to a pathogenic adaptive immune reaction. First of all, what are the exact binding sites of penicillamine on macrophages and through which signaling pathway does this binding activate macrophages? At this time, we do not know if carbonylation of the molecules involved is a specific or a random process. If it is specific, is the reason why SD and Lewis rats are resistant to penicillamine-induced autoimmunity because macrophages in these two rat strains have fewer specific surface carbonylated proteins for penicillamine to react with and therefore macrophages are not fully activated by penicillamine; however, the difference is small. It is more likely that the difference in response of the rat strains is due to basic differences in immune response and only detailed studies of the differences in the response to penicillamine will reveal why BN rats are unique. Meanwhile, we showed that both isoniazid and hydralazine activated RAW 264.7 macrophages in vitro, but up to now treatment with these two drugs has not been able to induce any evident clinical symptoms in BN rats. Could this be because we have not achieved sufficient blood concentration of hydralazine/isoniazid due to their fast metabolism? If so, we will need to modify the dosage or the way of drug administration to boost blood level of both drugs close to the ones used in humans so that we might be able to develop more valid animal models of IDRs. Second, we have shown that macrophages are activated shortly after penicillamine treatment with an increase in the production of cytokines such as IL-6. However, significant elevation of serum IL-6 is only detected at about 2 weeks, which raises the question of what are the early steps by which activation of macrophages appears to drive Th17 cell differentiation. Possibilities to explain this paradox include: 1) Early macrophage activation and interaction with Th17 precursor cells is localized and therefore serum IL-6 level is not a good reflection of the local environment that leads to Th17 differentiation; 2) Because the first time point of

117 serum IL-6 we tested was at day 7, we may have missed an early spike in IL-6. Hence, we need to closely examine serum cytokines in first 7 days in future with a focus on IL-6. Although we have shown that penicillamine and other drugs that bind to aldehydes can induce the production of IL-6 in Raw 264.7 cells in vitro, the demonstration that these drugs can induce an immediate increase in IL-6, either localized or systemic, would provide additional support for the basic hypothesis that this is the mechanism by which penicillamine drives naïve T cells to differentiate into Th17 cells leading to autoimmunity. We also do not know what other immune cells are required to cause the pathology observed in these animals. Experiments are planned in which B cells will be depleted with an antibody against CD20 similar to Rituximab, which should answer part of this question. Another important question is why some Brown Norway rats develop autoimmunity and others do not. Given the genetic homogeneity of this strain of rat, a genetic basis seems unlikely. We have noticed a variation in the incidence with time and which animal facility is used to house the animals and this implies environmental factors, but it would be difficult to clearly define the environmental factors involved. We also do not know why hydralazine and isoniazid activate macrophages in vitro and can cause a lupus-like syndrome in humans but do not induce autoimmunity in Brown Norway rats. One plausible explanation is simply that they are cleared too rapidly in rats and the concentrations required to induce autoimmunity simply cannot readily be sustained in rats. The mechanism that we have investigated in this work is limited to drugs that bind irreversibly to aldehyde groups and so it may have limited implications for the mechanisms of other IDRs. However, our lab has demonstrated that many, if not most, other drugs that cause autoimmunity are oxidized to reactive metabolites by the myeloperoxidase system of macrophages and this could also lead to macrophage activation through a different pathway. These same drugs can also cause other types of IDRs such as liver toxicity and agranulocytosis. Thus it is possible that macrophage activation is a key step in many other types of IDRs. Another hypothesis that is being tested is that many IDRs have an autoimmune component; in fact, most drugs that cause idiosyncratic liver toxicity can also

118 cause a lupus-like syndrome. This could explain the characteristics of some IDRs such as a very long lag between starting the drug and the onset of the IDR and the lack of immune memory. These characteristics have led people to believe that IDRs with these characteristics are not immune-mediated. We believe that testing these hypotheses with well-controlled experiments in valid animal models is the best way to significantly increase our mechanistic understanding of IDRs, which in turn is essential for prediction and prevention of these serious adverse reactions.

119 REFERENCES

(1) WHO (1972) International drug monitoring: role of national centres.

(2) (2003) American Society of Consultant Pharmacists Guidelines on detecting and reporting adverse drug reactions in long-term care environments., (www.ascp.com/public/pr/guidelines/adverse.shtml, Ed.).

(3) (2002) Center for Drug Evaluation and Research Preventable drug reactions; a focus on drug reactions, (www.fda.gov/cder/drug/drugReactions/default.html, Ed.).

(4) Pirmohamed, M. and Park, B. K. (2003) Adverse drug reactions: back to the future. Br J Clin Pharmacol 55, 486-492.

(5) Jefferys, D. B., Leakey, D., Lewis, J. A., Payne, S. and Rawlins, M. D. (1998) New active substances authorized in the United Kingdom between 1972 and 1994. Br J Clin Pharmacol 45, 151-156.

(6) Lasser, K. E., Allen, P. D., Woolhandler, S. J., Himmelstein, D. U., Wolfe, S. M. and Bor, D. H. (2002) Timing of new black box warnings and withdrawals for prescription medications. Jama 287, 2215- 2220.

(7) Edwards, I. R. and Aronson, J. K. (2000) Adverse drug reactions: definitions, diagnosis, and management. Lancet 356, 1255-1259.

(8) Uetrecht, J. (2009) Immune-mediated adverse drug reactions. Chem Res Toxicol 22, 24-34.

(9) Riedl, M. A. and Casillas, A. M. (2003) Adverse drug reactions: types and treatment options. Am Fam Physician 68, 1781-1790.

(10) Wilke, R. A., Lin, D. W., Roden, D. M., Watkins, P. B., Flockhart, D., Zineh, I., Giacomini, K. M. and Krauss, R. M. (2007) Identifying genetic risk factors for serious adverse drug reactions: current progress and challenges. Nat Rev Drug Discov 6, 904-916.

(11) Uetrecht, J. (2007) Idiosyncratic drug reactions: current understanding. Annu Rev Pharmacol Toxicol 47, 513-539.

(12) Goldstein, R. A. and Patterson, R. (1984) Drug allergy: prevention, diagnosis, and treatment. Ann Intern Med 100, 302-303.

(13) Lazarou, J., Pomeranz, B. H. and Corey, P. N. (1998) Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies. Jama 279, 1200-1205.

(14) Borchers, A. T., Keen, C. L. and Gershwin, M. E. (2007) Drug-induced lupus. Ann N Y Acad Sci 1108, 166-182.

(15) Uetrecht, J. P. (1992) The role of leukocyte-generated reactive metabolites in the pathogenesis of idiosyncratic drug reactions. Drug Metab Rev 24, 299-366.

120 (16) (1999) Executive summary of disease management of drug hypersensitivity: a practice parameter. Joint Task Force on Practice Parameters, the American Academy of Allergy, Asthma and Immunology, the American Academy of Allergy, Asthma and Immunology, and the Joint Council of Allergy, Asthma and Immunology. Ann Allergy Asthma Immunol 83, 665-700.

(17) Zeeh, J. and Platt, D. (2002) The aging liver: structural and functional changes and their consequences for drug treatment in old age. Gerontology 48, 121-127.

(18) Ju, C. (2005) Immunological mechanisms of drug-induced liver injury. Curr Opin Drug Discov Devel 8, 38-43.

(19) Walton, B., Simpson, B. R., Strunin, L., Doniach, D., Perrin, J. and Appleyard, A. J. (1976) Unexplained hepatitis following halothane. Br Med J 1, 1171-1176.

(20) Alvir, J. M., Lieberman, J. A., Safferman, A. Z., Schwimmer, J. L. and Schaaf, J. A. (1993) Clozapine- induced agranulocytosis. Incidence and risk factors in the United States. N Engl J Med 329, 162-167.

(21) Ferrell, P. B., Jr. and McLeod, H. L. (2008) Carbamazepine, HLA-B*1502 and risk of Stevens- Johnson syndrome and toxic epidermal necrolysis: US FDA recommendations. Pharmacogenomics 9, 1543-1546.

(22) Locharernkul, C., Loplumlert, J., Limotai, C., Korkij, W., Desudchit, T., Tongkobpetch, S., Kangwanshiratada, O., Hirankarn, N., Suphapeetiporn, K. and Shotelersuk, V. (2008) Carbamazepine and phenytoin induced Stevens-Johnson syndrome is associated with HLA-B*1502 allele in Thai population. Epilepsia 49, 2087-2091.

(23) Yang, C. W., Hung, S. I., Juo, C. G., Lin, Y. P., Fang, W. H., Lu, I. H., Chen, S. T. and Chen, Y. T. (2007) HLA-B*1502-bound peptides: implications for the pathogenesis of carbamazepine-induced Stevens-Johnson syndrome. J Allergy Clin Immunol 120, 870-877.

(24) Lucas, A., Nolan, D. and Mallal, S. (2007) HLA-B*5701 screening for susceptibility to abacavir hypersensitivity. J Antimicrob Chemother 59, 591-593.

(25) Mallal, S., Nolan, D., Witt, C., Masel, G., Martin, A. M., Moore, C., Sayer, D., Castley, A., Mamotte, C., Maxwell, D., James, I. and Christiansen, F. T. (2002) Association between presence of HLA- B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 359, 727-732.

(26) Mallal, S., Phillips, E., Carosi, G., Molina, J. M., Workman, C., Tomazic, J., Jagel-Guedes, E., Rugina, S., Kozyrev, O., Cid, J. F., Hay, P., Nolan, D., Hughes, S., Hughes, A., Ryan, S., Fitch, N., Thorborn, D. and Benbow, A. (2008) HLA-B*5701 screening for hypersensitivity to abacavir. N Engl J Med 358, 568-579.

(27) Rodriguez-Novoa, S., Garcia-Gasco, P., Blanco, F., Gonzalez-Pardo, G., Castellares, C., Moreno, V., Jimenez-Nacher, I., Gonzalez-Lahoz, J. and Soriano, V. (2007) Value of the HLA-B*5701 allele to predict abacavir hypersensitivity in Spaniards. AIDS Res Hum Retroviruses 23, 1374-1376.

(28) Daly, A. K., Donaldson, P. T., Bhatnagar, P., Shen, Y., Pe'er, I., Floratos, A., Daly, M. J., Goldstein, D. B., John, S., Nelson, M. R., Graham, J., Park, B. K., Dillon, J. F., Bernal, W., Cordell, H. J., Pirmohamed, M., Aithal, G. P. and Day, C. P. (2009) HLA-B*5701 genotype is a major determinant of drug-induced liver injury due to flucloxacillin. Nat Genet.

121 (29) Chantarangsu, S., Mushiroda, T., Mahasirimongkol, S., Kiertiburanakul, S., Sungkanuparph, S., Manosuthi, W., Tantisiriwat, W., Charoenyingwattana, A., Sura, T., Chantratita, W. and Nakamura, Y. (2009) HLA-B*3505 allele is a strong predictor for nevirapine-induced skin adverse drug reactions in HIV-infected Thai patients. Pharmacogenet Genomics 19, 139-146.

(30) van der Ven, A. J., Koopmans, P. P., Vree, T. B. and van der Meer, J. W. (1991) Adverse reactions to co-trimoxazole in HIV infection. Lancet 338, 431-433.

(31) Lee, W. M. (2003) Drug-induced hepatotoxicity. N Engl J Med 349, 474-485.

(32) Billingham, R. E., Brent, L. and Medawar, P. B. (1953) Actively acquired tolerance of foreign cells. Nature 172, 603-606.

(33) Burnet, F. M. (1961) Immunological recognition of self. Science 133, 307-311.

(34) Janeway, C. A., Jr. (1989) The priming of helper T cells. Semin Immunol 1, 13-20.

(35) Matzinger, P. (1994) Tolerance, danger, and the extended family. Annu Rev Immunol 12, 991-1045.

(36) Parker, C. W. (1982) Allergic reactions in man. Pharmacol. Rev. 34, 85-104.

(37) Parker, C. W. (1981) Hapten immunology and allergic reactions in humans. Arthritis Rheum 24, 1024- 1036.

(38) Vergani, D., Mieli-Vergani, G., Alberti, A., Neuberger, J., Eddleston, A., Davis, M. and Williams, R. (1980) Antibodies to the surface of halothane-altered rabbit hepatocytes in patients with severe halothane-associated hepatitis. N. Engl. J. Med. 303, 66-71.

(39) Uetrecht, J. P. (1999) New concepts in immunology relevant to idiosyncratic drug reactions: the "danger hypothesis" and innate immune system. Chem Res Toxicol 12, 387-395.

(40) Pirmohamed, M., Naisbitt, D. J., Gordon, F. and Park, B. K. (2002) The danger hypothesis--potential role in idiosyncratic drug reactions. Toxicology 181-182, 55-63.

(41) Seguin, B. and Uetrecht, J. (2003) The danger hypothesis applied to idiosyncratic drug reactions. Curr Opin Allergy Clin Immunol 3, 235-242.

(42) Oppenheim, J. J., Tewary, P., de la Rosa, G. and Yang, D. (2007) Alarmins initiate host defense. Adv Exp Med Biol 601, 185-194.

(43) Miyake, K. (2007) Innate immune sensing of pathogens and danger signals by cell surface Toll-like receptors. Semin Immunol 19, 3-10.

(44) Zanni, M. P., von Greyerz, S., Schnyder, B., Brander, K. A., Frutig, K., Hari, Y., Valitutti, S. and Pichler, W. J. (1998) HLA-restricted, processing- and metabolism-independent pathway of drug recognition by human alpha beta T lymphocytes. J Clin Invest 102, 1591-1598.

(45) Pichler, W. J. (2002) Pharmacological interaction of drugs with antigen-specific immune receptors: the p-i concept. Curr Opin Allergy Clin Immunol 2, 301-305.

122 (46) Chen, J., Mannargudi, B. M., Xu, L. and Uetrecht, J. (2008) Demonstration of the metabolic pathway responsible for nevirapine-induced skin rash. Chem Res Toxicol 21, 1862-1870.

(47) Shenton, J. M., Chen, J. and Uetrecht, J. P. (2004) Animal models of idiosyncratic drug reactions. Chem Biol Interact 150, 53-70.

(48) Scarpelli, D. G. (1997) Animal models of disease: utililty and limitations, in:P.M. Iannaccone, D.G. Scarpelli (Eds.), Biological Aspects of Disease: Contributions from Animal Models, Harwood Academic Publishers, Amsterdam.

(49) Shenton, J. M., Popovic, M., Chen, J., Masson, M. J. and Uetrecht, J. P. (2005) Evidence of an immune-mediated mechanism for an idiosyncratic nevirapine-induced reaction in the female Brown Norway rat. Chem Res Toxicol 18, 1799-1813.

(50) Weigert, W. M., Offermanns, H. and Scherberich, P. (1975) D-Penicillamine--production and properties. Angew Chem Int Ed Engl 14, 330-336.

(51) Tamir, M., Bornstein, B., Behar, M. and Chwat, M. (1964) Mercury Poisoning from an Unsuspected Source. Br J Ind Med 21, 299-303.

(52) Howard-Lock, H. E., Lock, C. J., Mewa, A. and Kean, W. F. (1986) D-penicillamine: chemistry and clinical use in rheumatic disease. Semin Arthritis Rheum 15, 261-281.

(53) Matzinger, P. (2002) The danger model: a renewed sense of self. Science 296, 301-305.

(54) Goodnow, C. C., Sprent, J., Fazekas de St Groth, B. and Vinuesa, C. G. (2005) Cellular and genetic mechanisms of self tolerance and autoimmunity. Nature 435, 590-597.

(55) Rosen, A. and Casciola-Rosen, L. (2001) Clearing the way to mechanisms of autoimmunity. Nat Med 7, 664-665.

(56) Yang, M. L., Doyle, H. A., Gee, R. J., Lowenson, J. D., Clarke, S., Lawson, B. R., Aswad, D. W. and Mamula, M. J. (2006) Intracellular protein modification associated with altered T cell functions in autoimmunity. J Immunol 177, 4541-4549.

(57) Albert, L. J. and Inman, R. D. (1999) Molecular mimicry and autoimmunity. N Engl J Med 341, 2068- 2074.

(58) Deshmukh, U. S., Bagavant, H., Lewis, J., Gaskin, F. and Fu, S. M. (2005) Epitope spreading within lupus-associated ribonucleoprotein antigens. Clin Immunol 117, 112-120.

(59) Kidd, B. A., Ho, P. P., Sharpe, O., Zhao, X., Tomooka, B. H., Kanter, J. L., Steinman, L. and Robinson, W. H. (2008) Epitope spreading to citrullinated antigens in mouse models of autoimmune arthritis and demyelination. Arthritis Res Ther 10, R119.

(60) Grimaldi, C., Nashi, E., Venkatesh, J. and Diamond, B. (2007) B cell hyporesponsiveness and autoimmunity: a new paradigm. Adv Exp Med Biol 596, 181-190.

(61) Azulay-Debby, H. and Melamed, D. (2007) B cell receptor editing in tolerance and autoimmunity. Front Biosci 12 , 2136-2147.

123 (62) Racke, M. K. (2008) The role of B cells in multiple sclerosis: rationale for B-cell-targeted therapies. Curr Opin Neurol 21 Suppl 1, S9-S18.

(63) Roll, P. and Tony, H. P. (2009) [B-cell-targeted therapies in the treatment of autoimmune diseases.]. Z Rheumatol 68, 255-259.

(64) Matthews, R. (2007) Autoimmune diseases. The B cell slayer. Science 318, 1232-1233.

(65) Rioux, J. D. and Abbas, A. K. (2005) Paths to understanding the genetic basis of autoimmune disease. Nature 435, 584-589.

(66) Ueda, H., Howson, J. M., Esposito, L., Heward, J., Snook, H., Chamberlain, G., Rainbow, D. B., Hunter, K. M., Smith, A. N., Di Genova, G., Herr, M. H., Dahlman, I., Payne, F., Smyth, D., Lowe, C., Twells, R. C., Howlett, S., Healy, B., Nutland, S., Rance, H. E., Everett, V., Smink, L. J., Lam, A. C., Cordell, H. J., Walker, N. M., Bordin, C., Hulme, J., Motzo, C., Cucca, F., Hess, J. F., Metzker, M. L., Rogers, J., Gregory, S., Allahabadia, A., Nithiyananthan, R., Tuomilehto-Wolf, E., Tuomilehto, J., Bingley, P., Gillespie, K. M., Undlien, D. E., Ronningen, K. S., Guja, C., Ionescu-Tirgoviste, C., Savage, D. A., Maxwell, A. P., Carson, D. J., Patterson, C. C., Franklyn, J. A., Clayton, D. G., Peterson, L. B., Wicker, L. S., Todd, J. A. and Gough, S. C. (2003) Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 423, 506-511.

(67) Atkinson, J. P. (1989) Complement deficiency: predisposing factor to autoimmune syndromes. Clin Exp Rheumatol 7 Suppl 3, S95-101.

(68) Bennett, C. L., Christie, J., Ramsdell, F., Brunkow, M. E., Ferguson, P. J., Whitesell, L., Kelly, T. E., Saulsbury, F. T., Chance, P. F. and Ochs, H. D. (2001) The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutations of FOXP3. Nat Genet 27, 20-21.

(69) Davies, A. J. (2008) Immunological tolerance and the autoimmune response. Autoimmun Rev 7, 538- 543.

(70) Romagnani, S. (2006) Immunological tolerance and autoimmunity. Intern Emerg Med 1, 187-196.

(71) van Driel, I. R. (2007) Tolerance and autoimmunity: entwined pathways lead to immunological tolerance. Immunol Cell Biol 85, 267-268.

(72) Chikuma, S. and Bluestone, J. A. (2003) CTLA-4 and tolerance: the biochemical point of view. Immunol Res 28, 241-253.

(73) Friedline, R. H., Brown, D. S., Nguyen, H., Kornfeld, H., Lee, J., Zhang, Y., Appleby, M., Der, S. D., Kang, J. and Chambers, C. A. (2009) CD4+ regulatory T cells require CTLA-4 for the maintenance of systemic tolerance. J Exp Med 206, 421-434.

(74) Hori, S., Nomura, T. and Sakaguchi, S. (2003) Control of regulatory T cell development by the transcription factor Foxp3. Science 299, 1057-1061.

(75) Fontenot, J. D., Gavin, M. A. and Rudensky, A. Y. (2003) Foxp3 programs the development and function of CD4+CD25+ regulatory T cells. Nat Immunol 4, 330-336.

(76) Vignali, D. A., Collison, L. W. and Workman, C. J. (2008) How regulatory T cells work. Nat Rev

124 Immunol 8, 523-532.

(77) Sakaguchi, S., Yamaguchi, T., Nomura, T. and Ono, M. (2008) Regulatory T cells and immune tolerance. Cell 133, 775-787.

(78) Takikawa, O. (2005) Biochemical and medical aspects of the indoleamine 2,3-dioxygenase-initiated L-tryptophan metabolism. Biochem Biophys Res Commun 338, 12-19.

(79) Grohmann, U., Fallarino, F. and Puccetti, P. (2003) Tolerance, DCs and tryptophan: much ado about IDO. Trends Immunol 24, 242-248.

(80) Forouzandeh, F., Jalili, R. B., Germain, M., Duronio, V. and Ghahary, A. (2008) Differential immunosuppressive effect of indoleamine 2,3-dioxygenase (IDO) on primary human CD4+ and CD8+ T cells. Mol Cell Biochem 309, 1-7.

(81) Sakurai, K., Zou, J. P., Tschetter, J. R., Ward, J. M. and Shearer, G. M. (2002) Effect of indoleamine 2,3-dioxygenase on induction of experimental autoimmune encephalomyelitis. J Neuroimmunol 129, 186-196.

(82) Jacobi, A. M., Zhang, J., Mackay, M., Aranow, C. and Diamond, B. (2009) Phenotypic characterization of autoreactive B cells--checkpoints of B cell tolerance in patients with systemic lupus erythematosus. PLoS ONE 4, e5776.

(83) Ota, T., Aoki-Ota, M., Tsunoda, K., Nishikawa, T., Koyasu, S. and Amagai, M. (2008) Autoreactive B- cell elimination by pathogenic IgG specific for the same antigen: implications for peripheral tolerance. Int Immunol 20, 1351-1360.

(84) Mongey, A. B. and Hess, E. V. (2008) Drug insight: autoimmune effects of medications-what's new? Nat Clin Pract Rheumatol 4, 136-144.

(85) Olsen, N. J. (2004) Drug-induced autoimmunity. Best Pract Res Clin Rheumatol 18, 677-688.

(86) Uetrecht, J. (2005) Current trends in drug-induced autoimmunity. Autoimmun Rev 4, 309-314.

(87) Richardson, B. (2003) DNA methylation and autoimmune disease. Clin Immunol 109, 72-79.

(88) Deng, C., Lu, Q., Zhang, Z., Rao, T., Attwood, J., Yung, R. and Richardson, B. (2003) Hydralazine may induce autoimmunity by inhibiting extracellular signal-regulated kinase pathway signaling. Arthritis Rheum 48, 746-756.

(89) Yung, R. L. and Richardson, B. C. (1994) Role of T cell DNA methylation in lupus syndromes. Lupus 3, 487-491.

(90) Uetrecht, J. (1991) Metabolism of drugs by activated leukocytes: implications for drug-induced lupus and other drug hypersensitivity reactions. Adv Exp Med Biol 283, 121-132.

(91) Kretz-Rommel, A., Duncan, S. R. and Rubin, R. L. (1997) Autoimmunity caused by disruption of central T cell tolerance. A murine model of drug-induced lupus. J Clin Invest 99, 1888-1896.

(92) Hensley, K., Robinson, K. A., Gabbita, S. P., Salsman, S. and Floyd, R. A. (2000) Reactive oxygen

125 species, cell signaling, and cell injury. Free Radic Biol Med 28, 1456-1462.

(93) Moldovan, L. and Moldovan, N. I. (2004) Oxygen free radicals and redox biology of organelles. Histochem Cell Biol 122, 395-412.

(94) Smith, M. A., Rottkamp, C. A., Nunomura, A., Raina, A. K. and Perry, G. (2000) Oxidative stress in Alzheimer's disease. Biochim Biophys Acta 1502, 139-144.

(95) Kumagai, S., Jikimoto, T. and Saegusa, J. (2003) [Pathological roles of oxidative stress in autoimmune diseases]. Rinsho Byori 51, 126-132.

(96) Davies, M. J., Fu, S., Wang, H. and Dean, R. T. (1999) Stable markers of oxidant damage to proteins and their application in the study of human disease. Free Radic Biol Med 27, 1151-1163.

(97) Dalle-Donne, I., Aldini, G., Carini, M., Colombo, R., Rossi, R. and Milzani, A. (2006) Protein carbonylation, cellular dysfunction, and disease progression. J Cell Mol Med 10, 389-406.

(98) Aldini, G., Dalle-Donne, I., Colombo, R., Maffei Facino, R., Milzani, A. and Carini, M. (2006) Lipoxidation-derived reactive carbonyl species as potential drug targets in preventing protein carbonylation and related cellular dysfunction. ChemMedChem 1, 1045-1058.

(99) Esterbauer, H., Schaur, R. J. and Zollner, H. (1991) Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic Biol Med 11, 81-128.

(100) Wautier, J. L. and Schmidt, A. M. (2004) Protein glycation: a firm link to endothelial cell dysfunction. Circ Res 95, 233-238.

(101) Schleicher, E. D., Wagner, E. and Nerlich, A. G. (1997) Increased accumulation of the glycoxidation product N(epsilon)-(carboxymethyl)lysine in human tissues in diabetes and aging. J Clin Invest 99, 457-468.

(102) Margetis, P. I., Antonelou, M. H., Petropoulos, I. K., Margaritis, L. H. and Papassideri, I. S. (2009) Increased protein carbonylation of red blood cell membrane in diabetic retinopathy. Exp Mol Pathol.

(103) Smerjac, S. M. and Bizzozero, O. A. (2008) Cytoskeletal protein carbonylation and degradation in experimental autoimmune encephalomyelitis. J Neurochem 105, 763-772.

(104) Stadtman, E. R. (1990) Metal ion-catalyzed oxidation of proteins: biochemical mechanism and biological consequences. Free Radic Biol Med 9, 315-325.

(105) Dalle-Donne, I., Rossi, R., Milzani, A., Di Simplicio, P. and Colombo, R. (2001) The actin cytoskeleton response to oxidants: from small heat shock protein phosphorylation to changes in the redox state of actin itself. Free Radic Biol Med 31, 1624-1632.

(106) O'Reilly, P. J., Hickman-Davis, J. M., Davis, I. C. and Matalon, S. (2003) Hyperoxia impairs antibacterial function of macrophages through effects on actin. Am J Respir Cell Mol Biol 28, 443-450.

(107) Dalle-Donne, I., Carini, M., Vistoli, G., Gamberoni, L., Giustarini, D., Colombo, R., Maffei Facino, R., Rossi, R., Milzani, A. and Aldini, G. (2007) Actin Cys374 as a nucleophilic target of alpha,beta- unsaturated aldehydes. Free Radic Biol Med 42, 583-598.

126 (108) Bussieres, J. F. and Habra, M. (1995) Application of International Consensus Meeting Criteria for classifying drug-induced liver disorders. Ann Pharmacother 29, 875-878.

(109) Ostapowicz, G., Fontana, R. J., Schiodt, F. V., Larson, A., Davern, T. J., Han, S. H., McCashland, T. M., Shakil, A. O., Hay, J. E., Hynan, L., Crippin, J. S., Blei, A. T., Samuel, G., Reisch, J. and Lee, W. M. (2002) Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med 137, 947-954.

(110) Temple, R. J. and Himmel, M. H. (2002) Safety of newly approved drugs: implications for prescribing. Jama 287, 2273-2275.

(111) Zimmerman, H. J. (1999) Hepatotoxicity: the adverse effects of drugs and other chemicals on the liver, Lippincott Williams & Wilkins, Philadelphia (PA).

(112) Meier, Y., Cavallaro, M., Roos, M., Pauli-Magnus, C., Folkers, G., Meier, P. J. and Fattinger, K. (2005) Incidence of drug-induced liver injury in medical inpatients. Eur J Clin Pharmacol 61, 135-143.

(113) Benichou, C. (1990) Criteria of drug-induced liver disorders. Report of an international consensus meeting. J Hepatol 11, 272-276.

(114) Abboud, G. and Kaplowitz, N. (2007) Drug-induced liver injury. Drug Saf 30, 277-294.

(115) Stieger, B., Fattinger, K., Madon, J., Kullak-Ublick, G. A. and Meier, P. J. (2000) Drug- and estrogen- induced cholestasis through inhibition of the hepatocellular bile salt export pump (Bsep) of rat liver. Gastroenterology 118, 422-430.

(116) Uetrecht, J. (2009) Immunoallergic Drug-Induced Liver Injury in Humans. SEMINARS IN LIVER DISEASE 29.

(117) Salazar-Paramo, M., Rubin, R. L. and Garcia-De La Torre, I. (1992) Systemic lupus erythematosus induced by isoniazid. Ann Rheum Dis 51, 1085-1087.

(118) Gough, A., Chapman, S., Wagstaff, K., Emery, P. and Elias, E. (1996) Minocycline induced autoimmune hepatitis and systemic lupus erythematosus-like syndrome. Bmj 312, 169-172.

(119) Dupont, A. and Six, R. (1982) Lupus-like syndrome induced by methyldopa. Br Med J (Clin Res Ed) 285, 693-694.

(120) Liu, Z. X. and Kaplowitz, N. (2002) Immune-mediated drug-induced liver disease. Clin Liver Dis 6, 755-774.

(121) Perry, H. M., Jr. (1973) Late toxicity to hydralazine resembling systemic lupus erythematosus or rheumatoid arthritis. Am J Med 54, 58-72.

(122) Selroos, O. and Edgren, J. (1975) Lupus-like syndrome associated with pulmonary reaction to nitrofurantoin. Report of three cases. Acta Med Scand 197, 125-129.

(123) Sharp, J. R., Ishak, K. G. and Zimmerman, H. J. (1980) Chronic active hepatitis and severe hepatic necrosis associated with nitrofurantoin. Ann Intern Med 92, 14-19.

127 (124) Levy, M. (1993) Propylthiouracil hepatotoxicity. A review and case presentation. Clin Pediatr (Phila) 32, 25-29.

(125) Takuwa, N., Kojima, I. and Ogata, E. (1981) Lupus-like syndrome--a rare complication in thionamide treatment for Graves' disease. Endocrinol Jpn 28, 663-667.

(126) Kurtz, M. D. (1968) Migratory polyarthritis occurring with methimazole therapy. N Y State J Med 68, 2810-2811.

(127) Fischer, M. G., Nayer, H. R. and Miller, A. (1973) Methimazole-induced jaundice. Jama 223, 1028- 1029.

(128) McCraken, M., Benson, E. A. and Hickling, P. (1980) Systemic lupus erythematosus induced by aminoglutethimide. Br Med J 281, 1254.

(129) Stuart-Harris, R. C. and Smith, I. E. (1984) Aminoglutethimide in the treatment of advanced breast cancer. Cancer Treat Rev 11, 189-204.

(130) Kramer, M. R., Levene, C. and Hershko, C. (1986) Severe reversible autoimmune haemolytic anaemia and thrombocytopenia associated with diclofenac therapy. Scand J Haematol 36, 118-120.

(131) Scully, L. J., Clarke, D. and Barr, R. J. (1993) Diclofenac induced hepatitis. 3 cases with features of autoimmune chronic active hepatitis. Dig Dis Sci 38, 744-751.

(132) Choi, H. K., Merkel, P. A. and Niles, J. L. (1998) ANCA-positive vasculitis associated with allopurinol therapy. Clin Exp Rheumatol 16, 743-744.

(133) Chawla, S. K., Patel, H. D., Parrino, G. R., Soterakis, J., Lopresti, P. A. and D'Angelo, W. A. (1977) Allopurinol hepatotoxicity. Case report and literature review. Arthritis Rheum 20, 1546-1549.

(134) Ecker, J. A. (1965) Phenylbutazone Hepatitis. Am J Gastroenterol 43, 23-29.

(135) Farid, N. and Anderson, J. (1971) S.L.E.-like reaction after phenylbutazone. Lancet 1, 1022-1023.

(136) Brown, M. and Schubert, T. (1986) Phenytoin hypersensitivity hepatitis and mononucleosis syndrome. J Clin Gastroenterol 8, 469-477.

(137) Ross, S., Ormerod, A. D., Roberts, C., Dwyer, C. and Herriot, R. (2002) Subacute cutaneous lupus erythematosus associated with phenytoin. Clin Exp Dermatol 27, 474-476.

(138) Bateman, D. E. (1985) Carbamazepine induced systemic lupus erythematosus: case report. Br Med J (Clin Res Ed) 291, 632-633.

(139) Morales-Diaz, M., Pinilla-Roa, E. and Ruiz, I. (1999) Suspected carbamazepine-induced hepatotoxicity. Pharmacotherapy 19, 252-255.

(140) Dujovne, C. A., Chan, C. H. and Zimmerman, H. J. (1967) Sulfonamide hepatic injury. Review of the literature and report of a case due to sulfamethoxazole. N Engl J Med 277, 785-788.

(141) Alberti-Flor, J. J. (1983) Chlorpromazine-induced lupus-like illness. Am Fam Physician 27, 151-152.

128 (142) Russell, R. I., Allan, J. G. and Patrick, R. (1973) Active chronic hepatitis after chlorpromazine ingestion. Br Med J 1, 655-656.

(143) Bonsmann, G., Schiller, M., Luger, T. A. and Stander, S. (2001) Terbinafine-induced subacute cutaneous lupus erythematosus. J Am Acad Dermatol 44, 925-931.

(144) Fernandes, N. F., Geller, S. A. and Fong, T. L. (1998) Terbinafine hepatotoxicity: case report and review of the literature. Am J Gastroenterol 93, 459-460.

(145) Ahmad, S. (1991) Lovastatin-induced lupus erythematosus. Arch Intern Med 151, 1667-1668.

(146) Alla, V., Abraham, J., Siddiqui, J., Raina, D., Wu, G. Y., Chalasani, N. P. and Bonkovsky, H. L. (2006) Autoimmune hepatitis triggered by statins. J Clin Gastroenterol 40, 757-761.

(147) Marzano, A. V., Ramoni, S., Del Papa, N., Barbareschi, M. and Alessi, E. (2008) Leflunomide- induced subacute cutaneous lupus erythematosus with erythema multiforme-like lesions. Lupus 17, 329-331.

(148) Sevilla-Mantilla, C., Ortega, L., Agundez, J. A., Fernandez-Gutierrez, B., Ladero, J. M. and Diaz- Rubio, M. (2004) Leflunomide-induced acute hepatitis. Dig Liver Dis 36, 82-84.

(149) Actis, G. C., Morgando, A., Lagget, M., David, E. and Rizzetto, M. (2001) Zafirlukast-related hepatitis: report of a further case. J Hepatol 35, 539-541.

(150) Finkel, T. H., Hunter, D. J., Paisley, J. E., Finkel, R. S. and Larsen, G. L. (1999) Drug-induced lupus in a child after treatment with zafirlukast (Accolate). J Allergy Clin Immunol 103, 533-534.

(151) Soy, M., Ozer, H., Canataroglu, A., Gumurdulu, D. and Erken, E. (2002) Vasculitis induced by zafirlukast therapy. Clin Rheumatol 21, 328-329.

(152) Uetrecht, J. (2007) Idiosyncratic Drug Reactions: Past, Present, and Future. Chem Res Toxicol.

(153) Masson, M. J. and Uetrecht, J. P. (2004) Tolerance induced by low dose D-penicillamine in the brown Norway rat model of drug-induced autoimmunity is immune-mediated. Chem Res Toxicol 17, 82-94.

(154) Donker, A. J., Venuto, R. C., Vladutiu, A. O., Brentjens, J. R. and Andres, G. A. (1984) Effects of prolonged administration of D-penicillamine or captopril in various strains of rats. Brown Norway rats treated with D-penicillamine develop autoantibodies, circulating immune complexes, and disseminated intravascular coagulation. Clin Immunol Immunopathol 30, 142-155.

(155) Masson, M. J., Teranishi, M., Shenton, J. M. and Uetrecht, J. P. (2004) Investigation of the Involvement of Macrophages and T Cells in D-Penicillamine-Induced Autoimmunity in the Brown Norway Rat. J Immunot 1, 79-93.

(156) Arrigoni-Martelli, E. and Binderup, L. (1981) D-penicillamine, lymphocytes, and macrophages: an account of experimental investigations in vivo. J Rheumatol Suppl 7, 62-68.

(157) Binderup, L., Bramm, E. and Arrigoni-Martelli, E. (1978) D-penicillamine and macrophages: modulation of lymphocyte transformation by concanavalin A. Scand J Immunol 7, 259-264.

129 (158) Binderup, L., Bramm, E. and Arrigoni-Martelli, E. (1980) D-penicillamine in vivo enhances lymphocyte DNA synthesis: role of macrophages. Scand J Immunol 11, 23-28.

(159) Chen, H., Hall, S., Heffernan, B., Thompson, N. T., Rogers, M. V. and Rhodes, J. (1997) Convergence of Schiff base costimulatory signaling and TCR signaling at the level of mitogen-activated protein kinase ERK2. J Immunol 159, 2274-2281.

(160) Greineder, D. K. and Rosenthal, A. S. (1975) The requirement for macrophage-lymphocyte interaction in T lymphocyte proliferation induced by generation of aldehydes on cell membranes. J Immunol 115, 932-938.

(161) Greineder, D. K., Shevach, E. M. and Rosenthal, A. S. (1976) Macrophage-lymphocyte interaction. III. Site of alloantiserum inhibition of T lymphocyte proliferation induced by allogeneic or aldehyde- bearing cells. J Immunol 117, 1261-1266.

(162) Rhodes, J. (1996) Covalent chemical events in immune induction: fundamental and therapeutic aspects. Immunol Today 17, 436-441.

(163) Chen, H. and Rhodes, J. (1996) Schiff base forming drugs: mechanisms of immune potentiation and therapeutic potential. J Mol Med 74, 497-504.

(164) Rhodes, J., Chen, H., Hall, S. R., Beesley, J. E., Jenkins, D. C., Collins, P. and Zheng, B. (1995) Therapeutic potentiation of the immune system by costimulatory Schiff-base-forming drugs. Nature 377, 71-75.

(165) Zheng, B., Brett, S. J., Tite, J. P., Lifely, M. R., Brodie, T. A. and Rhodes, J. (1992) Galactose oxidation in the design of immunogenic vaccines. Science 256, 1560-1563.

(166) Siegel, R. C. (1977) Collagen cross-linking. Effect of D-penicillamine on cross-linking in vitro. J Biol Chem 252, 254-259.

(167) Yarema, K. J., Mahal, L. K., Bruehl, R. E., Rodriguez, E. C. and Bertozzi, C. R. (1998) Metabolic delivery of ketone groups to sialic acid residues. Application To cell surface glycoform engineering. J Biol Chem 273, 31168-31179.

(168) Faure, S., Salazar-Fontana, L. I., Semichon, M., Tybulewicz, V. L., Bismuth, G., Trautmann, A., Germain, R. N. and Delon, J. (2004) ERM proteins regulate cytoskeleton relaxation promoting T cell- APC conjugation. Nat Immunol 5, 272-279.

(169) Senga, T., Hasegawa, H., Tanaka, M., Rahman, M. A., Ito, S. and Hamaguchi, M. (2008) The cysteine-cluster motif of c-Src: its role for the heavy metal-mediated activation of kinase. Cancer Sci 99, 571-575.

(170) Grimsrud, P. A., Xie, H., Griffin, T. J. and Bernlohr, D. A. (2008) Oxidative stress and covalent modification of protein with bioactive aldehydes. J Biol Chem 283, 21837-21841.

(171) Lappas, M., Permezel, M. and Rice, G. E. (2007) Advanced glycation endproducts mediate proinflammatory actions in human gestational tissues via nuclear factor-kappaB and extracellular signal- regulated kinase 1/2. J Endocrinol 193, 269-277.

130 (172) Shanmugam, N., Figarola, J. L., Li, Y., Swiderski, P. M., Rahbar, S. and Natarajan, R. (2008) Proinflammatory effects of advanced lipoxidation end products in monocytes. Diabetes 57, 879-888.

(173) Virella, G., Thorpe, S. R., Alderson, N. L., Stephan, E. M., Atchley, D., Wagner, F. and Lopes-Virella, M. F. (2003) Autoimmune response to advanced glycosylation end-products of human LDL. J Lipid Res 44, 487-493.

(174) Uetrecht, J. (2008) Idiosyncratic drug reactions: past, present, and future. Chem Res Toxicol 21, 84-92.

(175) Kavai, M. and Szegedi, G. (2007) Immune complex clearance by monocytes and macrophages in systemic lupus erythematosus. Autoimmun Rev 6, 497-502.

(176) Coleman, D. L. (1986) Regulation of macrophage phagocytosis. Eur J Clin Microbiol 5, 1-5.

(177) Hart, S. P., Dransfield, I. and Rossi, A. G. (2008) Phagocytosis of apoptotic cells. Methods 44, 280- 285.

(178) Rosenthal, A. S. (1980) Regulation of the immune response--role of the macrophage. N Engl J Med 303, 1153-1156.

(179) McCoy, C. E. and O'Neill, L. A. (2008) The role of toll-like receptors in macrophages. Front Biosci 13, 62-70.

(180) Fadok, V. A., McDonald, P. P., Bratton, D. L. and Henson, P. M. (1998) Regulation of macrophage cytokine production by phagocytosis of apoptotic and post-apoptotic cells. Biochem Soc Trans 26, 653-656.

(181) Gordon, S. (1998) The role of the macrophage in immune regulation. Res Immunol 149, 685-688.

(182) Arceci, R. J. (2008) When T cells and macrophages do not talk: the hemophagocytic syndromes. Curr Opin Hematol 15, 359-367.

(183) Zhang, K., Biroschak, J., Glass, D. N., Thompson, S. D., Finkel, T., Passo, M. H., Binstadt, B. A., Filipovich, A. and Grom, A. A. (2008) Macrophage activation syndrome in patients with systemic juvenile idiopathic arthritis is associated with MUNC13-4 polymorphisms. Arthritis Rheum 58, 2892- 2896.

(184) Tristano, A. G. (2008) Macrophage activation syndrome: a frequent but under-diagnosed complication associated with rheumatic diseases. Med Sci Monit 14, RA27-36.

(185) Pringe, A., Trail, L., Ruperto, N., Buoncompagni, A., Loy, A., Breda, L., Martini, A. and Ravelli, A. (2007) Macrophage activation syndrome in juvenile systemic lupus erythematosus: an under- recognized complication? Lupus 16, 587-592.

(186) Li, J., Mannargudi, B. and Uetrecht, J. (2009) Covalent binding of penicillamine to macrophages: implications for penicillamine-induced autoimmunity. Chem Res Toxicol.

(187) Dobrovolskaia, M. A. and Vogel, S. N. (2002) Toll receptors, CD14, and macrophage activation and deactivation by LPS. Microbes Infect 4, 903-914.

131 (188) Polfliet, M. M., Fabriek, B. O., Daniels, W. P., Dijkstra, C. D. and van den Berg, T. K. (2006) The rat macrophage scavenger receptor CD163: expression, regulation and role in inflammatory mediator production. Immunobiology 211, 419-425.

(189) Behrens, E. M. (2008) Macrophage activation syndrome in rheumatic disease: what is the role of the antigen presenting cell? Autoimmun Rev 7, 305-308.

(190) Yeap, S. T., Sheen, J. M., Kuo, H. C., Hwang, K. P., Yang, K. D. and Yu, H. R. (2008) Macrophage activation syndrome as initial presentation of systemic lupus erythematosus. Pediatr Neonatol 49, 39- 42.

(191) van den Berg, T. K., Dopp, E. A. and Dijkstra, C. D. (2001) Rat macrophages: membrane glycoproteins in differentiation and function. Immunol Rev 184, 45-57.

(192) Smyth, M. J., Cretney, E., Kelly, J. M., Westwood, J. A., Street, S. E., Yagita, H., Takeda, K., van Dommelen, S. L., Degli-Esposti, M. A. and Hayakawa, Y. (2005) Activation of NK cell cytotoxicity. Mol Immunol 42, 501-510.

(193) Di Santo, J. P. (2006) Natural killer cell developmental pathways: a question of balance. Annu Rev Immunol 24, 257-286.

(194) Ma, A., Koka, R. and Burkett, P. (2006) Diverse functions of IL-2, IL-15, and IL-7 in lymphoid homeostasis. Annu Rev Immunol 24, 657-679.

(195) Pazmany, L. (2005) Do NK cells regulate human autoimmunity? Cytokine 32, 76-80.

(196) Perricone, R., Perricone, C., De Carolis, C. and Shoenfeld, Y. (2008) NK cells in autoimmunity: a two-edg'd weapon of the immune system. Autoimmun Rev 7, 384-390.

(197) Gordon, S. (2003) Alternative activation of macrophages. Nat Rev Immunol 3, 23-35.

(198) Butler, M., Carruthers, G., Harth, M., Freeman, D., Percy, J. and Rabenstein, D. (1982) Pharmacokinetics of reduced D-penicillamine in patients with rheumatoid arthritis. Arthritis Rheum 25, 111-116.

(199) Kulms, D. and Schwarz, T. (2006) NF-kappaB and cytokines. Vitam Horm 74, 283-300.

(200) Rhodus, N. L., Cheng, B., Myers, S., Bowles, W., Ho, V. and Ondrey, F. (2005) A comparison of the pro-inflammatory, NF-kappaB-dependent cytokines: TNF-alpha, IL-1-alpha, IL-6, and IL-8 in different oral fluids from oral lichen planus patients. Clin Immunol 114, 278-283.

(201) Mise-Omata, S., Kuroda, E., Niikura, J., Yamashita, U., Obata, Y. and Doi, T. S. (2007) A proximal kappaB site in the IL-23 p19 promoter is responsible for RelA- and c-Rel-dependent transcription. J Immunol 179, 6596-6603.

(202) Tournade, H., Pelletier, L., Pasquier, R., Vial, M. C., Mandet, C. and Druet, P. (1990) D- penicillamine-induced autoimmunity in Brown-Norway rats. Similarities with HgCl2-induced autoimmunity. J Immunol 414, 2985-2991.

(203) Seguin, B., Teranishi, M. and Uetrecht, J. P. (2003) Modulation of D-penicillamine-induced

132 autoimmunity in the Brown Norway rat using pharmacological agents that interfere with arachidonic acid metabolism or synthesis of inducible nitric oxide synthase. Toxicology 190, 267-278.

(204) Seguin, B., Masson, M. J. and Uetrecht, J. (2004) D-penicillamine-induced autoimmunity in the Brown Norway rat: role for both T and non-T splenocytes in adoptive transfer of tolerance. Chem Res Toxicol 17, 1299-1302.

(205) Sayeh, E. and Uetrecht, J. P. (2001) Factors that modify penicillamine-induced autoimmunity in Brown Norway rats: failure of the Th1/Th2 paradigm. Toxicology 163, 195-211.

(206) Mosmann, T. R. (1992) T lymphocyte subsets, cytokines, and effector functions. Ann N Y Acad Sci 664, 89-92.

(207) Singh, V. K., Mehrotra, S. and Agarwal, S. S. (1999) The paradigm of Th1 and Th2 cytokines: its relevance to autoimmunity and allergy. Immunol Res 20, 147-161.

(208) Zamvil, S. S. and Steinman, L. (1990) The T lymphocyte in experimental allergic encephalomyelitis. Annu Rev Immunol 8, 579-621.

(209) Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A. and Coffman, R. L. (2005) Two types of murine helper T cell clone. I. Definition according to profiles of lymphokine activities and secreted proteins. 1986. J Immunol 175, 5-14.

(210) Langrish, C. L., Chen, Y., Blumenschein, W. M., Mattson, J., Basham, B., Sedgwick, J. D., McClanahan, T., Kastelein, R. A. and Cua, D. J. (2005) IL-23 drives a pathogenic T cell population that induces autoimmune inflammation. J Exp Med 201, 233-240.

(211) Murphy, C. A., Langrish, C. L., Chen, Y., Blumenschein, W., McClanahan, T., Kastelein, R. A., Sedgwick, J. D. and Cua, D. J. (2003) Divergent pro- and antiinflammatory roles for IL-23 and IL-12 in joint autoimmune inflammation. J Exp Med 198, 1951-1957.

(212) McKenzie, B. S., Kastelein, R. A. and Cua, D. J. (2006) Understanding the IL-23-IL-17 immune pathway. Trends Immunol 27, 17-23.

(213) Weaver, C. T., Harrington, L. E., Mangan, P. R., Gavrieli, M. and Murphy, K. M. (2006) Th17: an effector CD4 T cell lineage with regulatory T cell ties. Immunity 24, 677-688.

(214) Bettelli, E., Carrier, Y., Gao, W., Korn, T., Strom, T. B., Oukka, M., Weiner, H. L. and Kuchroo, V. K. (2006) Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature 441, 235-238.

(215) Mangan, P. R., Harrington, L. E., O'Quinn, D. B., Helms, W. S., Bullard, D. C., Elson, C. O., Hatton, R. D., Wahl, S. M., Schoeb, T. R. and Weaver, C. T. (2006) Transforming growth factor-beta induces development of the T(H)17 lineage. Nature 441, 231-234.

(216) Zhou, L., Ivanov, II, Spolski, R., Min, R., Shenderov, K., Egawa, T., Levy, D. E., Leonard, W. J. and Littman, D. R. (2007) IL-6 programs T(H)-17 cell differentiation by promoting sequential engagement of the IL-21 and IL-23 pathways. Nat Immunol 8, 967-974.

(217) Zhou, L., Lopes, J. E., Chong, M. M., Ivanov, II, Min, R., Victora, G. D., Shen, Y., Du, J., Rubtsov, Y.

133 P., Rudensky, A. Y., Ziegler, S. F. and Littman, D. R. (2008) TGF-beta-induced Foxp3 inhibits T(H)17 cell differentiation by antagonizing RORgammat function. Nature 453, 236-240.

(218) Ivanov, II, McKenzie, B. S., Zhou, L., Tadokoro, C. E., Lepelley, A., Lafaille, J. J., Cua, D. J. and Littman, D. R. (2006) The orphan nuclear receptor RORgammat directs the differentiation program of proinflammatory IL-17+ T helper cells. Cell 126, 1121-1133.

(219) Zhou, L. and Littman, D. R. (2009) Transcriptional regulatory networks in Th17 cell differentiation. Curr Opin Immunol 21, 146-152.

(220) Yang, X. O., Pappu, B. P., Nurieva, R., Akimzhanov, A., Kang, H. S., Chung, Y., Ma, L., Shah, B., Panopoulos, A. D., Schluns, K. S., Watowich, S. S., Tian, Q., Jetten, A. M. and Dong, C. (2008) T helper 17 lineage differentiation is programmed by orphan nuclear receptors ROR alpha and ROR gamma. Immunity 28, 29-39.

(221) Harris, T. J., Grosso, J. F., Yen, H. R., Xin, H., Kortylewski, M., Albesiano, E., Hipkiss, E. L., Getnet, D., Goldberg, M. V., Maris, C. H., Housseau, F., Yu, H., Pardoll, D. M. and Drake, C. G. (2007) Cutting edge: An in vivo requirement for STAT3 signaling in TH17 development and TH17- dependent autoimmunity. J Immunol 179, 4313-4317.

(222) Fossiez, F., Djossou, O., Chomarat, P., Flores-Romo, L., Ait-Yahia, S., Maat, C., Pin, J. J., Garrone, P., Garcia, E., Saeland, S., Blanchard, D., Gaillard, C., Das Mahapatra, B., Rouvier, E., Golstein, P., Banchereau, J. and Lebecque, S. (1996) T cell interleukin-17 induces stromal cells to produce proinflammatory and hematopoietic cytokines. J Exp Med 183, 2593-2603.

(223) Ouyang, W., Kolls, J. K. and Zheng, Y. (2008) The biological functions of T helper 17 cell effector cytokines in inflammation. Immunity 28, 454-467.

(224) Sonderegger, I., Kisielow, J., Meier, R., King, C. and Kopf, M. (2008) IL-21 and IL-21R are not required for development of Th17 cells and autoimmunity in vivo. Eur J Immunol 38, 1833-1838.

(225) Deenick, E. K. and Tangye, S. G. (2007) Autoimmunity: IL-21: a new player in Th17-cell differentiation. Immunol Cell Biol 85, 503-505.

(226) Suto, A., Kashiwakuma, D., Kagami, S., Hirose, K., Watanabe, N., Yokote, K., Saito, Y., Nakayama, T., Grusby, M. J., Iwamoto, I. and Nakajima, H. (2008) Development and characterization of IL-21- producing CD4+ T cells. J Exp Med 205, 1369-1379.

(227) Boraschi, D. and Dinarello, C. A. (2006) IL-18 in autoimmunity: review. Eur Cytokine Netw 17, 224- 252.

(228) Martinez, G. J., Nurieva, R. I., Yang, X. O. and Dong, C. (2008) Regulation and function of proinflammatory TH17 cells. Ann N Y Acad Sci 1143, 188-211.

(229) Elyaman, W., Bradshaw, E. M., Uyttenhove, C., Dardalhon, V., Awasthi, A., Imitola, J., Bettelli, E., Oukka, M., van Snick, J., Renauld, J. C., Kuchroo, V. K. and Khoury, S. J. (2009) IL-9 induces differentiation of TH17 cells and enhances function of FoxP3+ natural regulatory T cells. Proc Natl Acad Sci U S A.

(230) McGeachy, M. J., Bak-Jensen, K. S., Chen, Y., Tato, C. M., Blumenschein, W., McClanahan, T. and Cua, D. J. (2007) TGF-beta and IL-6 drive the production of IL-17 and IL-10 by T cells and restrain

134 T(H)-17 cell-mediated pathology. Nat Immunol 8, 1390-1397.

(231) Vanden Eijnden, S., Goriely, S., De Wit, D., Willems, F. and Goldman, M. (2005) IL-23 up-regulates IL-10 and induces IL-17 synthesis by polyclonally activated naive T cells in human. Eur J Immunol 35, 469-475.

(232) Stumhofer, J. S., Silver, J. S., Laurence, A., Porrett, P. M., Harris, T. H., Turka, L. A., Ernst, M., Saris, C. J., O'Shea, J. J. and Hunter, C. A. (2007) Interleukins 27 and 6 induce STAT3-mediated T cell production of interleukin 10. Nat Immunol 8, 1363-1371.

(233) McGeachy, M. J. and Cua, D. J. (2008) Th17 cell differentiation: the long and winding road. Immunity 28, 445-453.

(234) Newcomb, D. C., Zhou, W., Moore, M. L., Goleniewska, K., Hershey, G. K., Kolls, J. K. and Peebles, R. S., Jr. (2009) A functional IL-13 receptor is expressed on polarized murine CD4+ Th17 cells and IL-13 signaling attenuates Th17 cytokine production. J Immunol 182, 5317-5321.

(235) Zimmerman, H. (1999) Hepatotoxicity: The Adverse Effects of Drugs and Other Chemicals on the Liver, 2nd ed, Lippincott Williams & Wilkins, Philadelphia.

(236) Lawrenson, R. A., Seaman, H. E., Sundstrom, A., Williams, T. J. and Farmer, R. D. (2000) Liver damage associated with minocycline use in acne: a systematic review of the published literature and pharmacovigilance data. Drug Saf 23, 333-349.

(237) Dong, C. (2009) Differentiation and function of pro-inflammatory Th17 cells. Microbes Infect.

(238) Korn, T., Bettelli, E., Oukka, M. and Kuchroo, V. K. (2009) IL-17 and Th17 Cells. Annu Rev Immunol 27, 485-517.

(239) Shen, F. and Gaffen, S. L. (2008) Structure-function relationships in the IL-17 receptor: implications for signal transduction and therapy. Cytokine 41, 92-104.

(240) Aranami, T. and Yamamura, T. (2008) Th17 Cells and autoimmune encephalomyelitis (EAE/MS). Allergol Int 57, 115-120.

(241) Emamaullee, J. A., Davis, J., Merani, S., Toso, C., Elliott, J. F., Thiesen, A. and Shapiro, A. M. (2009) Inhibition of Th17 cells regulates autoimmune diabetes in NOD mice. Diabetes.

(242) Nistala, K. and Wedderburn, L. R. (2009) Th17 and regulatory T cells: rebalancing pro- and anti- inflammatory forces in autoimmune arthritis. Rheumatology (Oxford).

(243) Peck, A. and Mellins, E. D. (2009) Breaking old paradigms: Th17 cells in autoimmune arthritis. Clin Immunol.

(244) Song, C., Luo, L., Lei, Z., Li, B., Liang, Z., Liu, G., Li, D., Zhang, G., Huang, B. and Feng, Z. H. (2008) IL-17-producing alveolar macrophages mediate allergic lung inflammation related to asthma. J Immunol 181, 6117-6124.

(245) Benghiat, F. S., Charbonnier, L. M., Vokaer, B., De Wilde, V. and Le Moine, A. (2009) Interleukin 17- producing T helper cells in alloimmunity. Transplant Rev (Orlando) 23, 11-18.

135 (246) Ettinger, R., Kuchen, S. and Lipsky, P. E. (2008) Interleukin 21 as a target of intervention in autoimmune disease. Ann Rheum Dis 67 Suppl 3, iii83-86.

(247) Sutherland, A. P., Van Belle, T., Wurster, A. L., Suto, A., Michaud, M., Zhang, D., Grusby, M. J. and von Herrath, M. (2009) Interleukin-21 is required for the development of type 1 diabetes in NOD mice. Diabetes 58, 1144-1155.

(248) Feferman, T., Aricha, R., Mizrachi, K., Geron, E., Alon, R., Souroujon, M. C. and Fuchs, S. (2009) Suppression of experimental autoimmune myasthenia gravis by inhibiting the signaling between IFN- gamma inducible protein 10 (IP-10) and its receptor CXCR3. J Neuroimmunol 209, 87-95.

(249) Nishioji, K., Okanoue, T., Itoh, Y., Narumi, S., Sakamoto, M., Nakamura, H., Morita, A. and Kashima, K. (2001) Increase of chemokine interferon-inducible protein-10 (IP-10) in the serum of patients with autoimmune liver diseases and increase of its mRNA expression in hepatocytes. Clin Exp Immunol 123, 271-279.

(250) Matsushita, T., Hasegawa, M., Matsushita, Y., Echigo, T., Wayaku, T., Horikawa, M., Ogawa, F., Takehara, K. and Sato, S. (2007) Elevated serum BAFF levels in patients with localized scleroderma in contrast to other organ-specific autoimmune diseases. Exp Dermatol 16, 87-93.

(251) Migita, K., Abiru, S., Maeda, Y., Nakamura, M., Komori, A., Ito, M., Fujiwara, S., Yano, K., Yatsuhashi, H., Eguchi, K. and Ishibashi, H. (2007) Elevated serum BAFF levels in patients with autoimmune hepatitis. Hum Immunol 68, 586-591.

(252) Bosello, S., Pers, J. O., Rochas, C., Devauchelle, V., De Santis, M., Daridon, C., Saraux, A., Ferraccioli, G. F. and Youinou, P. (2007) BAFF and rheumatic autoimmune disorders: implications for disease management and therapy. Int J Immunopathol Pharmacol 20, 1-8.

(253) Lai Kwan Lam, Q., King Hung Ko, O., Zheng, B. J. and Lu, L. (2008) Local BAFF gene silencing suppresses Th17-cell generation and ameliorates autoimmune arthritis. Proc Natl Acad Sci U S A 105, 14993-14998.

(254) Ohteki, T. (2002) Critical role for IL-15 in innate immunity. Curr Mol Med 2, 371-380.

(255) Matsukawa, A., Hogaboam, C. M., Lukacs, N. W. and Kunkel, S. L. (2000) Chemokines and innate immunity. Rev Immunogenet 2, 339-358.

136 APPENDIX: Supplemental Data: luminex data for all cytokines/chemokines

IL- Patient IL- IL- TN GM- IFN IL- 12 IL- IL- IL- IL- MC MIP Cause of liver failure IL-6 IL-2 Eotaxin IL-4 IL-7 IL-8 IP-10 ID 17 21 Fα CSF γ 10 (p4 13 15 1α 1β P-1 -1α 0) 10-091- APAP 3 152 74 <3.2 36 27 11 47 <3.2 31 17 <3.2 15 33 <3.2 22 57 232 160 6 03 11-029- APAP 4 367 4321 4 144 558 18 903 12 140 159 18 15 50 9 328 202 145 1756 48 06 11-011- APAP 5 271 198 8 135 174 6 538 5 56 396 30 27 198 17 179 220 728 1089 106 06 11-074- APAP 6 1419 650 15 111 178 10 704 4 83 339 47 18 92 16 214 189 365 1481 63 03 11-078- APAP 6 278 29 <3.2 59 <3.2 5 NA 6 <3.2 NA 40 5 13 <3.2 9 10 142 240 15 02 11-006- APAP 10 1045 27 <3.2 180 <3.2 7 91 10 4 65 <3.2 7 21 <3.2 <3.2 175 266 1179 67 04 10-032- APAP 13 1364 606 4 170 <3.2 4 190 9 32 47 5 16 63 <3.2 64 115 147 687 69 02 10-085- APAP 23 191 89 24 153 221 5 994 16 78 138 6 8 193 9 233 156 229 553 97 04 10-055- APAP 24 200 92 56 165 42 6 419 13 38 334 11 34 99 19 13 39 227 758 88 03 13-005- APAP 29 746 6514 23 515 542 10 1224 46 220 157 33 13 111 18 492 675 274 3277 112 02 11-063- APAP 29 3559 1848 3 86 <3.2 8 168 28 95 20 <3.2 14 21 <3.2 21 253 80 4578 177 01 11-102- 110 APAP 29 669 812 180 171 554 12 1760 20 229 79 56 631 91 377 153 289 1993 135 02 4 11-023- APAP 72 100 86 18 184 <3.2 7 205 70 8 62 8 14 80 6 23 68 268 502 177 06

137 10-029- APAP <3.2 0 9 <3.2 65 48 6 310 <3.2 10 17 <3.2 4 10 <3.2 31 28 43 385 22 04 10-060- APAP <3.2 47 87 <3.2 96 <3.2 6 NA <3.2 <3.2 71 <3.2 7 15 16 <3.2 26 79 142 3 05 10-026- APAP <3.2 535 63 2 52 <3.2 31 115 <3.2 28 167 12 10 123 5 14 19 596 6841 173 01 11-024- APAP <3.2 341 77 4 62 7 21 151 <3.2 16 88 <3.2 11 25 <3.2 17 144 209 415 86 06 11-130- APAP <3.2 1242 227 5 107 <3.2 17 197 4 45 76 <3.2 14 70 <3.2 13 92 167 4848 78 02 11-058- <3. APAP <3.2 1386 32 <3.2 143 <3.2 8 50 <3.2 17 <3.2 14 11 <3.2 <3.2 76 207 4353 111 02 2 11-139- APAP <3.2 1472 393 <3.2 47 69 <3.2 358 <3.2 13 38 <3.2 18 26 <3.2 50 314 70 3454 5 03 10-034- >10 APAP <3.2 739 <3.2 332 95 29 877 29 4643 101 5 35 152 5 134 808 1106 5883 32 02 000 13-038- HAV 7 238 37 6 55 100 10 418 5 45 318 12 14 142 10 95 43 417 301 73 04 13-001- HAV 18 740 18 20 61 117 10 261 48 42 117 6 20 43 10 114 34 236 224 97 05 13-019- HAV <3.2 441 26 5 27 <3.2 9 7 <3.2 8 20 <3.2 5 63 <3.2 <3.2 21 555 73 23 04 13-012- HAV <3.2 988 281 <3.2 134 98 8 472 <3.2 58 30 26 6 44 3 139 218 454 595 30 04 11-118- <3. HAV <3.2 211 6 <3.2 58 7 12 43 <3.2 40 88 <3.2 104 <3.2 <3.2 20 1119 262 24 02 2 12-027- <3. HBV 7 330 28 5 79 16 5 157 <3.2 13 103 <3.2 46 3 10 35 178 310 31 02 2 13-081- HBV 21 930 454 11 145 216 19 702 9 352 138 58 10 205 8 174 95 609 3150 188 04 13-032- HBV <3.2 3184 74 <3.2 81 46 14 273 <3.2 63 58 3 5 487 <3.2 55 111 9056 755 61 04

138 11-150- <3. <3. HBV <3.2 424 35 2 126 <3.2 <3.2 33 <3.2 7 <3.2 6 <3.2 <3.2 65 100 546 11 02 2 2 13-054- HBV <3.2 705 18 2 133 312 9 781 <3.2 101 185 8 10 214 8 287 263 1649 560 46 02 MYLOTAG 11-072- (GEMTUZUMAB) 3 781 1437 <3.2 80 36 12 276 7 88 63 <3.2 38 62 <3.2 63 373 449 3554 43 04 P1 12-015- <3. ISONIAZID <3.2 191 53 <3.2 98 <3.2 4 <3.2 4 5 132 <3.2 15 <3.2 5 23 169 194 34 04 2 13-010- <3. <3. ISONIAZID <3.2 651 9 <3.2 91 12 <3.2 89 <3.2 5 <3.2 20 <3.2 10 17 294 579 14 04 2 2 13-088- ISONIAZID <3.2 453 32 <3.2 84 14 10 220 26 31 65 <3.2 11 253 <3.2 <3.2 78 3549 367 18 01 13-133- <3. ISONIAZID <3.2 211 14 <3.2 63 <3.2 3 17 <3.2 <3.2 NA <3.2 NA <3.2 <3.2 33 96 477 21 02 2 14-005- <3. ISONIAZID 4 1789 19 <3.2 191 48 6 271 3 11 121 7 49 3 76 38 297 383 45 02 2 14-067- INH, PZA, <3.2 305 60 <3.2 243 128 4 484 5 39 196 14 12 57 7.3 181 85 276 229 47 04 RIFAMPIN 12-020- 120. PZA/RIFAMPIN <3.2 1864 25 14 105 <3.2 4 NA 14 5 17 <3.2 16 43 13 <3.2 58.4 403.0 992 01 0 10-009- RIFAMPIN 45 1240 76 6 81 56 14 327 27 23 176 4 9 141 4 112 91 1162 671 108 06 13-015- <3. NITROFURONTAIN <3.2 2122 18 <3.2 113 <3.2 <3.2 NA <3.2 10 6 <3.2 3 <3.2 <3.2 14 130 172 NA 02 2 13-067- NITROFURANTOIN 5 857 16 17 95 367 6 859 <3.2 110 167 18 14 53 8 300 93 227 315 44 04 13-080- BACTRIM 5 1726 263 <3.2 109 <3.2 24 148 3 66 20 <3.2 14 185 <3.2 3 242 1940 1008 31 02 10-025- <3. SULFADIAZINE J1 <3.2 842 1127 <3.2 165 <3.2 27 104 4 58 <3.2 15 334 <3.2 19 174 3418 3400 8 01 2 11-021- TRIMETHOPRIM/S 13 848 40 43 122 243 10 576 30 86 361 33 29 144 25 193 147 407 598 129

139 04 ULFA 15-002- <3. BROMFENAC <3.2 1374 99 <3.2 50 6 14 87 <3.2 23 58 <3.2 7 <3.2 19 74 95 578 30 02 2 13-003- TROGLITAZONE 4 127 13 43 37 44 3 164 3 32 40 <3.2 5 24 <3.2 34 62 188 435 30 03 14-037- PHENYTOIN 6 1742 664 <3.2 223 148 14 500 14 97 76 <3.2 4 488 3 124 93 6369 1410 92 02 13-047- <3. >100 PHENYTOIN <3.2 918 197 <3.2 195 5 29 239 8 37 90 79 573 <3.2 <3.2 61 558 30 01 2 00 13-129- PROPYLTHIOURA 10 1074 133 5 171 <3.2 7 96 20 42 92 30 6 86 <3.2 4 20 420 491 115 02 CIL 13-064- >100 ALLOPURINOL 108 857 1042 26 101 61 14 351 132 67 86 17 13 678 118 89 424 2378 232 01 00 11-100- KAVA KAVA 3 854 154 12 82 <3.2 6 67 <3.2 9 42 <3.2 12 18 5 2 28 73 361 23 02 PHENYTOIN 13-016- PRAVASTATIN 3 1853 321 <3.2 102 67 6 295 3 37 111 8 6 60 <3.2 61 36 455 964 49 01 14-007- CERIVASTATIN <3.2 4457 59 <3.2 60 <3.2 6 49 <3.2 15 2 <3.2 4 78 <3.2 <3.2 79 573 768 10 04 15-039- <3. ROSIGLITAZONE <3.2 1174 615 <3.2 194 <3.2 6 50 <3.2 25 <3.2 10 16 <3.2 9 186 294 1141 <3.2 03 2 13-145- >100 DICLOFENAC 4 496 278 <3.2 196 521 35 1045 7 187 205 13 10 699 9 352 283 3454 43 04 00 10-061- CIPROFLOXACIN 6 250 24 6 72 <3.2 12 108 4 14 58 29 5 116 8 <3.2 56 868 761 93 01 11-153- ETODOLAC 8 906 32 6 125 19 4 205 14 15 71 6 5 49 <3.2 48 41 326 575 89 02 13-007- <3. ZAFIRLUKAST 9 796 76 <3.2 108 23 <3.2 25 9 8 NA <3.2 43 <3.2 20 58 339 210 45 03 2 11-081- DISULFIRAM 23 790 70 <3.2 193 272 4 665 12 67 155 12 7 82 5 191 65 532 743 91 04 10-027- QUETIAPINE 40 972 94 5 181 27 6 267 32 27 6 <3.2 <3. 163 <3.2 160 95 488 151 82

140 06 2 14-042- HORNY GOAT <3.2 2165 228 <3.2 175 <3.2 9 20 <3.2 13 134 4 5 59 <3.2 6 37 341 805 30 04 WEED 15-049- B6 3 729 64 12 124 102 4 504 9 49 201 6 15 42 7 195 105 92 551 48 02 13-004- HERBAL MEDS 12 552 1140 9 209 13 19 250 12 196 174 6 7 73 4 56 115 457 2946 69 02 (MULTIPLE) 16-007- ISOFLURANE 40 858 935 48 220 113 14 511 60 100 207 88 10 363 15 119 178 331 810 127 06 14-078- UNKNOWN DRUG 66 464 122 14 283 20 10 104 11 18 15 4 8 178 <3.2 24 49 530 517 69 06 13-147- 141 13 TAK-559 111 713 279 311 271 64 43 1241 97 56 233 914 93 134 101 922 216 254 02 7 4 13-176- UNKNOWN DRUG <3.2 2408 3 3 71 <3.2 22 47 10 16 NA <3.2 5 254 <3.2 <3.2 26 1829 113 77 01 13-039- <3. THERMA SLIM? <3.2 5623 37 <3.2 78 37 5 193 <3.2 13 2 <3.2 29 <3.2 26 78 332 435 22 02 2 15-041- BLACK COHOSH <3.2 775 37 <3.2 214 147 7 512 <3.2 36 155 7 11 57 7 156 67 304 489 39 02 K2 <3. HCV-C2 HCV 4 1261 4 <3.2 118 <3.2 6 <3.2 7 <3.2 42 54 99 <3.2 <3.2 30 760 193 50 2 HCV-C8 HCV 8 881 6 <3.2 119 6 5 83 <3.2 13 219 15 8 59 9 44 15 152 685 69 HCV- HCV 9 898 9 <3.2 180 384 4 1185 4 58 315 10 11 50 14 197 45 76 412 61 C10 HCV-C5 HCV <3.2 1393 7 <3.2 141 332 <3.2 711 <3.2 78 248 10 5 80 7 215 32 215 570 38 <3. HCV-C1 HCV <3.2 741 <3.2 <3.2 48 15 <3.2 <3.2 <3.2 <3.2 65 <3.2 9 <3.2 <3.2 <3.2 236 240 <3.2 2 <3. <3. HCV-C6 HCV <3.2 1110 <3.2 <3.2 93 12 <3.2 <3.2 <3.2 <3.2 45 <3.2 <3.2 <3.2 5 80 298 <3.2 2 2 <3. HCV-C3 HCV <3.2 992 <3.2 <3.2 46 <3.2 <3.2 NA <3.2 <3.2 33 <3.2 37 <3.2 <3.2 4 460 300 <3.2 2 HCV-C9 HCV <3.2 1136 <3.2 <3.2 142 186 <3.2 255. <3.2 19 107 <3.2 <3. 287 <3.2 68 13 64 332 12

141 0 2 <3. <3. HCV-C4 HCV <3.2 1210 <3.2 <3.2 49 50 <3.2 NA <3.2 <3.2 7 <3.2 <3.2 <3.2 <3.2 144 501 <3.2 2 2 <3. HCV-C7 HCV <3.2 1281 <3.2 <3.2 144 <3.2 5 NA <3.2 <3.2 99 <3.2 15 <3.2 <3.2 5 189 911 34 2

142