The Immunopathogenesis of drug hypersensitivity

The Immunopathogenesis of drug

hypersensitivity

Craig Michael Rive – Bsc (Hons), Bforensics.

PhD candidate

This thesis is presented for the degree of

Doctor of Philosophy of Murdoch University.

November, 2016

i

Declaration

I declare that this thesis is my own account of my research and contains as its main

content work which has not previously been submitted for a degree at any tertiary

education institution.

Signed:

_________________________________________

Craig M. Rive

Date submitted for review: 30th of June, 2016

Date submitted after revisions: 21th of November, 2016

ii

iii

Acknowledgments

I would like to acknowledge Professor Elizabeth J. Phillips not only for her insight into

immunology and disease, but for the great work being done under her guidance and

for the opportunity to partake in the research being conducted here at The Institute

for Immunology and Infectious Diseases, Murdoch University, Perth, Australia. I

would also like to thank Professor Simon Mallal for the care and support he has given

my mother, over the past 20 years in her battle with Systemic Lupus Erythematous,

my family and I will always be grateful. I would also like to thank Dr Abha Chopra of

The Institute for Immunology and Infectious Diseases, Murdoch University, Perth,

Australia for her guidance, wisdom, motivating, and encouragement.

A special mention goes to colleagues at The Institute for Immunology and Infectious

Diseases, Murdoch University, Perth, Australia, Dr Elizabeth McKinnon and Dr

Rebecca Pavlos, for their contribution to the nevirapine associated HLA allele

analysis of Chapter 3. Also, thanks to Shay Leary at The Institute for Immunology

and Infectious Diseases, Murdoch University, Perth, Australia who performed the in-

silico analysis of the human herpesvirus, which provided the list of HLA-restricted

predicted peptides to target in the ex-vivo assays used in Chapter 4 and Dr David

Ostrov and colleagues, University of Florida College of Medicine, Gainsville, Florida,

USA for providing the virtual modelling data.

I have learned so much through the insight, knowledge and patience of those

mentioned above and others at The Institute for Immunology and Infectious

Diseases, Murdoch University, Perth, Australia, including Dr Andrew Lucas, Dr Alec

Redwood, Dr Niamh Keene, Dr Monika Tschochner, Associate professor Mark

Watson, and all the staff and students.

Lastly, I would like to thank my ever extending family for putting up with me and for

their continued support throughout the years. This is, in part, dedicated to all my

family and those whom I love.

iv

v

Table of Contents Declaration i

Acknowledgments iii

Abstract ix

List of Abbreviations xi

List of Tables xv

List of Figures xvii

Chapter 1 Introduction: Immunopathogenesis of Drug

Hypersensitivity 1

1.1 Adverse Drug Reactions 3

1.2 Adaptive, Cell-Mediated Immune Response 3

1.3 Immunopathogenesis of Delayed Immunologically Mediated-

Adverse Drug Reactions 11

1.4 Established Models of HLA-driven delayed Immunologically

Mediated-Adverse Drug Reactions 21

1.5 Scope of this Thesis 24

Chapter 2 Methods and Materials 29

2.1 Samples, Ethics and Limitations 31

2.2 Cellular Assays 37

2.3 Virtual Modelling and in-silico Peptide Design 48

2.4 Molecular Assays 49

Chapter 3 Nevirapine-Induced Hypersensitivity 57

3.1 Introduction 59

3.2 Hypothesis and Aims 60

3.3 Methods 61

3.4 Results 63

3.5 Discussion 80

vi

Chapter 4 Carbamazepine-Induced Stevens - Johnson Syndrome

and Toxic Epidermal Necrolysis: Predicted HLA-Restricted Peptide

Screening 85

4.1 Introduction 87


4.3 Methods 88

4.4 Results 92

4.5 Discussion 115


and Toxic Epidermal Necrolysis: Specific T-cell Clonotype and

Detection by Digital Droplet PCR 119

5.1 Introduction 121


5.3 Methods 122

5.4 Results 123

5.5 Discussion 129


and Toxic Epidermal Necrolysis: Granulysin Expression 133



6.3 Methods 135

6.4 Results 137

6.5 Discussion 148

Chapter 7 Summary and Conclusion 151


7.2 Nevirapine Induced Hypersensitivity 153

7.3 Carbamazepine-Induced Stevens-Johnson Syndrome and

Toxic Epidermal Necrolysis 156

7.4 Final Conclusions 161

vii

8 Bibliography 165

9 Appendix 181

9.1 Nevirapine isolation from Viramune tablets 181

9.2 Chapter 5 Specific T-cell clonotype and detection by ddPCR 182

9.3 Future experimentation for HLA-15:02-restricted CBZ-SJS/TEN 186

9.4 Publication and Conference Abstracts 188

viii

ix

Abstract

Introduction: Immunologically mediated-adverse drug reactions threaten the

viability of drugs and the health of patients (1, 2). The mechanistic basis for abacavir

hypersensitivity is understood and preventable through HLA-B*57:01 screening. The

immunopathogenesis of most immunologically mediated-adverse drug reaction

remain unsolved, representing an opportunity to improve drug safety.

Methods: Using a repository of clinically phenotyped and HLA typed human

Donors and Controls, this work aimed to shed light on the immunopathogenesis of

nevirapine and carbamazepine severe cutaneous adverse reactions. Cellular and

molecular, techniques were employed including ex-vivo/in-vitro immunophenotyping,

detection, stimulation and expansion of drug-specific responses, and characterisation

of the specific T-cell receptor using droplet digital PCR. Virtual and Bioinformatics

approaches were used to define potential interactions between nevirapine and

class-I HLA

Results and discussion: In carbamazepine associated Steven-Johnson’s

Syndrome / toxic epidermal necrolysis, increased expression of granulysin was seen

ex-vivo in CD8+ T cells exposed to carbamazepine. To explore the hypothesis that

this is related to a cross-reactive memory cytotoxic T-cell response, we examined

responses to human simplex virus 1 / 2 responses in HLA-B*15:02 carbamazepine

associated Steven-Johnson’s Syndrome / toxic epidermal necrolysis patients and

identified a shared epitope. Common T-cell receptor clonotypes were identified using

droplet digital PCR in HLA-B*15:02 carbamazepine associated Steven-Johnson’s

Syndrome / toxic epidermal necrolysis patients.

Nevirapine severe cutaneous adverse reactions are associated with multiple class-I

HLA alleles across differing ethnicities. The hypothesis that this relates to shared

peptide-binding specificities between these alleles was confirmed using

bioinformatics, statistical, and virtual approaches. This showed the strongest

association across European, Asian and African-American ethnicities to be with two

HLA class C alleles HLA-C*04:01 and C*05:01 defined by a unique F-binding pocket.

Conclusion: Continuing work should build on the results presented and the unique

techniques used, to further understand the underlying immunopathogenesis and the

mechanistic basis for these adverse reactions, leading to the development of

screening strategies to improve drug and patient safety.

x

xi

List of Abbreviations Abacavir…………………………………………………………............................... ABC

Adverse drug reactions………………………………………....…….……............ ADR

Alpha………………………………………………………….…............................... α

Analysis of variance………………………………………………….…..………..... ANOVA

Antigen-presenting cell..………………………………………............................... APC

Antiretroviral……………………………………………………............................... ARV

Basic Local Alignment Search Tool.................................................................... BLAST

Beta 2-Microglobulin ……...……………………………………............................. β2M

Beta…………………………………………………................................................ β

Carbamazepine………………………………………………….............................. CBZ

Cluster of differentiation…………………………………………............................ CD

Comma-separated values……………………………………………………........... CSV

Complementarity-determining region…………………………............................. CDR

Complementary DNA……………………………………………………….............. cDNA

Constant TCR region…………………………………………................................ C

Cross-reactive group.......................................................................................... CREG

Cytomegalovirus………………………………………........................................... CMV

Cytometric Bead Array…………………………………………………....…............ CBA

Cytotoxic T-lymphocyte-associated protein 4..................................................... CTLA4

Dendritic cell……………………………………………………............................... DC

Deoxyribonucleic acid…………………………………………............................... DNA

Deoxyribonucleotide triphosphates………………………………………..…........ dNTP

Dimethyl sulfoxide…………………………………………………………............... DMSO

Dimethylformamide……………………………………………..…….…….............. DMF

Diverse TCR region…………………………………………................................... D

Droplet digital Polymerase Chain Reaction ……………………..……................. ddPCR

xii

Drug reaction with eosinophilia and systemic symptoms………….………........ DRESS

Drug-Induced Liver Injury................................................................................... DILI

Enzyme-Linked ImmunoSpot……………………………………….………........... ELISpot

Epstein-Barr virus………………………………………......................................... EBV

Ethylenediaminetetraacetic acid……………………………………...…................ EDTA

Foetal calf serum……………………………………………...……………….......... FCS

Gamma Interferon……………………………………………………......……......... IFN-γ

Genome-wide association studies…………………………………...…………..... GWAS

Helper T cells…………………………………………………….............................. Th

Herpes simplex viruses type 1…………………………………………..……........ HSV-1

Herpes simplex viruses type 2………………………………………….................. HSV-2

Herpes Zoster..................................................................................................... HZ

Highly active antiretroviral therapy..................................................................... HAART

Human herpesvirus.………………………………………...................................... HHV

Human immunodeficiency virus………………………………………….….......... HIV

Human leukocyte antigen………………………………………............................. HLA

Hypersensitivity reaction………………......……….............................................. HSR

Hypersensitivity syndrome…………………………………….…........................... HSS

Immunoglobulin……………………………………………..................................... Ig

Immunologically mediated-adverse drug reactions............................................ IM-ADR

Interleukin………………………………………..;;………………………………...... IL

Intracellular cytokine staining ……………………………;;…………………..….... ICS

Joining TCR region……………………………………......…................................. J

Lymphoblastic cell lines……………………………;.………………….....……...... LCL

Maculopapular eruption...................................................................................... MPE

Major histocompatiblity complex………………………………............................. MHC

Natural killer cell…………………………………………………...……………........ NK cell

xiii

Natural killer T cell………………………………………………….......………........ NK-T cell

N-Ethylmaleimide…………………………………………………...…....………..... NEM

Nevirapine………………………………………………………............................... NVP

Non-steroidal anti-inflammatory.......................................................................... NSAID

Peptide-binding................................................................................................... pb-

Peripheral blood mononuclear cell………………………………..…………......... PBMC

Pharmological interaction………………………………………...……..………...... P-I

Phenylmethanesulfonylfluoride……………….………………………………........ PMSF

Phosphate-buffered saline.................................................................................. PBS

Post-herpetic neuralgia………………………………………………...………….... PHN

Programmed Death 1......................................................................................... PD-1

Programmed Death Ligand 1.............................................................................. PD-L1

Prospective Randomised Evaluation of DNA Screening in a Clinical Trail....... PREDICT-1

Regulatory T cell.………………………………………………............................... Treg

Relative centrifugal force.................................................................................... RPM

Revolutions per minute....................................................................................... RCF

Ribonuclease P RNA component H1……………………….........………………... RPPH1

Ribonucleic acid……………………….....................................………………....... RNA

Severe cutaneous adverse reaction……………………………..……....…........... SCAR

Staphylococcal enterotoxin B............................................................................. SEB

Stevens-Johnson syndrome……………………………………............................ SJS

Study of Hypersensitivity to ABC and Pharmacogenetic Evaluation……......... SHAPE

T-cell receptor.…………………………………………………............................... TCR

Tosyl-L-lysine chloromethyl ketone………………………………………….......... TLCK

Tosyl-L-phenylalanine chloromethyl ketone…………………………………........ TPCK

Toxic epidermal necrolysis……………………………………............................... TEN

Tris-EDTA buffer................................................................................................. TE buffer

xiv

Variable TCR DNA region…………………………………….…........................... V

Varicella zoster virus.......................................................................................... VZV

Amino acids........................................................................................................ AA

Alanine…………………………………………………………………...................... A

Arginine………………………………………………………………….................... R

Asparagine…………………………………………………………………............... N

Aspartic acid…………………………………………………………………............ D

Cysteine…………………………………………………………………................... C

Glutamic acid……………………………………………………………….….......... E

Glutamine…………………………………………………………………................. Q

Glycine…............................................................................................................ G

Histidine………………………………………………..……………….................... H

Isoleucine………………………………………………………………..….............. I

Leucine…………………………………………………………………..................... L

Lysine…………………………………………………………………...................... K

Methionine……………………………………………………………….…............. M

Phenylalanine………………………………………………………………...…..... F

Proline…………………………………………………………………..................... P

Serine…………………………………………………………………...................... S

Threonine…………………………………………………………………............... T

Tryptophan………………………………………………………………….…........ W

Tyrosine.…………………………………………………………………................. Y

Valine…………………………………………………………………....................... V

xv

List of Tables Table 1.1: Delayed IM-ADR and their significant features. Adapted from Rive C et al.

2013, Phillips EJ et al. 2010, and Pavlos R et al. 2012 (2, 67, 74) ............................ 13

Table 1.2: Genetic (HLA alleles) and tissue specificity variation seen in different

ethnic groups associated with the NVP-HSRs. Table adapted from Rive C et al. 2013

and Phillips EJ et al. 2011(2, 150) ............................................................................. 21

Table 2.1: List of samples used in Diminishing ex-vivo PBMC’s responses and NVP-

HSR Immunopathogenesis ....................................................................................... 34

Table 2.2: List of samples used in CBZ-HSR immunopathogenesis associated

studies. ...................................................................................................................... 36

Table 2.3 List of reagents used in Stimulation and Expansion, flow cytometry, and

Lymphoblastic cell lines (LCL) generation assays. Flow cytometry antibodies are

included on Table 2.4, tetramer reagents included in Tables 2.5-2.7 and ELISpot

associated reagents and materials are included in Table 2.8 ................................... 37

Table 2.4: Flow cytometry Fluorochrome conjugated antibodies. ............................. 41

Table 2.5: Tetramer folding mix. ................................................................................ 43

Table 2.6: Protease inhibitor mix. .............................................................................. 43

Table 2.7: Addition of streptavidin-PE labelled to peptide/HLA class-I complex. ...... 44

Table 2.8 List of reagents and materials used in ELISpot assay. ............................. 45

Table 2.9 List of reagents and materials used in molecular assays. Specific PCR

associated reagents and concentrations are listed in the relevant tables. ................ 49

Table 2.10: complementary DNA (cDNA) conversion reagents and Master mix. ...... 50

Table 2.11: Cycling conditions for complementary DNA (cDNA) conversion ............ 50

Table 2.12: M13F and M13R primers used to amplify the reference sequence to use

the amplicons to as a positive control. ...................................................................... 51

Table 2.13: Reference sequence amplification reagent Mastermix to amplify the

reference sequence to use the amplicons to as a positive control. ........................... 52

Table 2.14: Reference sequence amplification PCR cycling conditions for the

amplification of the reference sequence to use the amplicons to as a positive control.

.................................................................................................................................. 52

xvi

Table 2.15: Reagent mastermix for EvaGreen ddPCR. ............................................ 54

Table 2.16: Digital droplet PCR (ddPCR) cycling conditions for the amplification of

Variable beta (vβ)-25-1*01 and -12-3/4*01 CDR3 specific sequences, constant beta

(Cβ), and ribonuclease P RNA component H1 (RPPH1) reference gene sequences.

.................................................................................................................................. 55

Table 2.17: vβ 25 and 12 CDR3 specific, and reference gene primers. .................... 55

Table 3.1 HLA cross-reactive group (CREG) supertypes and MHC Cluster peptide-

binding (pb-) groups .................................................................................................. 60

Table 3.2 Distribution of the work within Chapter 3 and accreditation to those who

contributed ................................................................................................................ 63

Table 3.3: NVP-HSR with hepatitis and statistically significant HLA CREG supertype

and pb-groups ........................................................................................................... 64

Table 3.4: HLA-B, B-pocket key amino acid residues YYAVAENIYQY at positions 7,

9, 24, 34, 41, 45, 63, 66, 67, 70, and 99 respectively for the CREG B*07 and pb-B*78

alleles HLA-B*55:(01/02), B*56:(01/03/04), compared to the amino acid sequence of

other HLA-B alleles. .................................................................................................. 67

Table 3.5: NVP-SCAR statistically significant HLA-C alleles. .................................... 68

Table 3.6: HLA-C F-pocket key amino acid residues, DNKLYLRNFYWTKW at

positions 74, 77, 80, 81, 84, 95, 97, 114, 116, 123, 133, 143, 146, and 147,

respectively for the HLA-C*04:(01/03/06/07), HLA-C*05:(01/09) and HLA-C*18:01

alleles, compared to the amino acid sequence of other HLA-C allele. ...................... 71

Table 3.7: HLA-C, B-pocket key amino acid residues YSAVGEKYQF at the B-pocket

positions 7, 9, 24, 34, 41, 45, 63, 66, 67, and 70 respectively for the HLA-

C*04:(01/03/06/07), HLA-C*05:(01/09) and HLA-C*18:01 alleles, compared to the

amino acid sequence of other HLA-C alleles. ........................................................... 72

Table 3.8: Multivariate model of predictive and protective NVP-rash associated

SCARs HLA alleles, adjusted for ethnicity. ............................................................... 73

Table 3.9: Positive and negative predictive values for carriage of the identified

predicted and/or protective NVP-SCAR associated HLA class-I alleles, adjusted for

5% prevalence. ......................................................................................................... 74

Table 4.1: In-silico predicted HSV-1 and HSV-2, HLA-B*15:02-restricted peptides

which were synthesised and analysed ex-vivo (169, 170, 188, 189). ........................ 89

xvii

Table 4.2: In-silico predicted HSV-1 and HSV-2 shared, HLA-B*15:02-restricted

peptides and variants of the HSV-1 and HSV-2 shared peptide, YAHGHSLRLPY

which were synthesised and analysed, ex-vivo (169, 170, 188, 189). ....................... 90

Table 4.3: Distribution of the work within Chapter 4 and accreditation to those who

contributed. ............................................................................................................... 91

Table 5.1: Distribution of the work within this chapter and accreditation to those who

contributed .............................................................................................................. 123

Table 5.2: Summary of the ddPCR amplification and analysis of target Vβ CDR3

sequences as a percentage (%) of expression in Donor 1 (CBZ-TEN), Donor 2 (CBZ-

SJS), and Donor 3 (CBZ-SJS) pre-stimulation cells, CBZ, control drug stimulated and

expanded cells, and the unstimulated expanded cells............................................. 129

Table 6.1: Distribution of the work within this chapter and accreditation to those who

contributed .............................................................................................................. 136

List of Figures Figure 1.1: Human chromosome 6, the location of Major histocompatiblity complex

(MHC 6p21.1-21.3) and the Human leukocyte antigen (HLA) class-I, class-II, and

class-III genes. Adapted from Urayama KY et al. 2013 (13) ....................................... 4

Figure 1.2: Side view of (A) HLA class-I and (B) HLA class-II molecules. Also

represented are the top view of the (C) HLA class-I peptide-binding cleft pockets A-F

and (D) HLA class-II peptide-binding groove pockets 1-9. Adapted from Cole DK,

2013 (17). .................................................................................................................... 6

Figure 1.3: Human leukocyte antigen (HLA) and the processing and presentation of

antigens. (A) Endogenous antigen processing and HLA class-I presentation. (B)

Exogenous antigen processing and HLA class-II presentation. (C) HLA class-I cross-

presentation of exogenous antigens derived via endocytosis. β2M, Beta 2-

Microglobulin; TCR, T- cell Receptor; ER, endoplasmic reticulum; TAP 1 and TAP 2,

transporter associated with antigen processing 1 and 2; CLIP, Class-II-associated

invariant chain peptide. Adapted from, Guermonprez P, et al. 2002, Heath WR and

Carbone FR, 2001, Braciale et al. 1987, Boesteanu A et al. 1998, and Kobayashi KS

et al. 2012 (18, 20-22, 25) ........................................................................................... 6

xviii

Figure 1.4: Germline DNA T-cell receptor (TCR) rearrangement, translation, splicing

and transcription. Variable alpha (Vα), Variable beta (Vβ), Joining alpha (Jα), Joining

beta (Jβ), diverse beta (Dβ), Constant alpha (Cα) and Constant beta (Cβ) DNA

chains, adapted from Janeway CA et al. 2001 (26) ..................................................... 8

Figure 1.5: Peptide-loaded human leukocyte antigen (HLA) molecule (A) class-I and

(B) class-II complexes and T-cell receptor (TCR) complementarity-determining region

(CDR) loop interactions. CDR3 (Orange α3 and β3), CDR1 and CDR2 β-chains

(Green β1 and β2) CDR1 and CDR2 α-chains (Blue α1 and α2). Figure adapted from

Cole DK, 2013 (17). .................................................................................................... 8

Figure 1.6: Naive CD8+ T-cell activation, proliferation and clonal expansion. Tcm,

central memory T cells; Tem, effector memory T cells; Trm, resident memory T cells;

DC, dendritic cells. Adapted from Kaech S and Weiguo C, 2012 (51) ...................... 10

Figure 1.7: Abacavir (ABC) translational roadmap Showing how the discovery of the

association between HLA-B*57:01 carriage and ABC- HSS lead to the incorporation

of HLA-B*57:01 genetic screening into clinical practice. Figure adapted from Rive C

et al. 2013 and Phillips EJ et al. 2011(2, 78). ............................................................ 15

Figure 1.8: The frequency of HLA-B*15:02 allele carriage most prevalent in those of

South and South-east Asian ethnic groups. (2, 67, 100) ........................................... 17

Figure 1.9: Graphical representation published by White et al., 2015 of the

heterologous immunity model with the three neo-antigen generating, HLA-drug

binding models (A) hapten/prohapten model, the (B) P-I model, and the (C) altered

peptide repertoire model.(167) .................................................................................. 24

Figure 2.1: Timeline of the stimulation and cellular expansion culture protocol for

analysis of CBZ-specific polyclonal T cells. ............................................................... 40

Figure 2.2: General overview of digital droplet PCR (ddPCR) assay workflow from

PCR mastermix generation through to droplet generation, amplification, QX100/200

plate reading and analysis. ....................................................................................... 54

Figure 3.1: MHCcluster tree created for the functional clustering of HLA-B alleles.

Depicted are the peptide-binding groups, pb-B*78 (red), pb-B*35 (orange), pb-B*46,

and pb-B*18. (blue), with representative predicted peptide-binding motifs (168). ..... 65

xix

Figure 3.2: MHCcluster heatmap representation for the functional clustering of HLA-B

alleles, Depicted are the peptide-binding groups, pb-B*78 (red), pb-B*35 (orange),

pb-B*46, and pb-B*18 using the MHCcluster algorithm (168). .................................. 65

Figure 3.3: MHCcluster tree created for the functional clustering of HLA-C alleles.

Depicted are the various predicted peptide-binding groups, pb-C*05 and pb-C*18

(blue), with representative peptide-binding motifs (168). ........................................... 69

Figure 3.4: MHCcluster heatmap representation for the functional clustering of HLA-C

alleles. Depicted are the various predicted peptide-binding groups, including pb-C*05

and pb-C*18 (blue), using the MHCcluster algorithm (168). ...................................... 69

Figure 3.5: Virtual modelling of NVP and HLA-C*04:01 potential interaction, which

indicates the preferential docking or binding of NVP near the B-pocket of the HLA-

C*04:01 molecule. ..................................................................................................... 72

Figure 3.6 Timeline of diminishing IFN-γ responses SFU/million cells in Donor 1

(NVP-HSR), Donor 2 (NVP-HSR), and Donor 3 (NVP-HSR) Figure from Keane et al.

2014 (138) ................................................................................................................. 75

Figure 3.7: Regulatory T cells expressed at a % of CD4 T cells are shown in the plot

where clear symbols indicates samples from Donor 1 (NVP-HSR) and the time after

NVP cessation. The filled in symbols indicate samples from controls (black, blue) and

HIV patient not on medication (red symbol). Figure from Keane et al. 2014 (138). ... 75

Figure 3.8: Cytokine plasma profile of Donor 1 (NVP-HSR), plasma obtained pre-

NVP-HSR to day 102 post NVP-HSR. Concentrations were determined by using the

CBA kit internal standard controls and normalised against the NVP-tolerant and NVP-

naive healthy controls. Statistical significance was determined using an analysis of

variance (ANOVA) statistical analysis (p < 0.0001) and multiple comparison tests

(***p < 0.0001) to compare the concentration of each cytokine in Donor 1 (NVP-HSR)

to the concentration of the responding cytokine in the NVP-tolerant and NVP-naive

healthy controls. ........................................................................................................ 76

xx

Figure 3.9: Cytokine plasma profile of Donor 2 (NVP-HSR) plasma obtained pre-

NVP-DRESS to day 330 post NVP-DRESS. Concentrations were determined by

using the CBA kit internal standard controls and normalised against the NVP-tolerant

and NVP-naive healthy controls. Statistical significance was determined using an

analysis of variance (ANOVA) statistical analysis (p < 0.0001) and multiple

comparison tests (***p < 0.0001) to compare the concentration of each cytokine in

Donor 2 (NVP-HSR) to the concentration of the responding cytokine in the NVP-

tolerant and NVP-naive healthy controls. .................................................................. 77

Figure 3.10: Cytokine plasma profile of Donor 3 (NVP-HSR) plasma obtained day 7

to day 727 post NVP-HSR. Concentrations were determined by using the CBA kit

internal standard controls and normalised against the NVP-tolerant and NVP-naive

healthy controls. ........................................................................................................ 78

Figure 3.11: Plasma IL-6 cytokine profile of Donor 1 (NVP-HSR) and Donor 2 (NVP-

HSR) (Donor 3 (NVP-HSR) not included), highlighting the potential of IL-6 having a

role in NVP-HSR. Concentrations were determined by using the CBA kit internal

standard controls and normalised against the NVP-tolerant and NVP-naive healthy

controls. Statistical significance was determined using an analysis of variance

(ANOVA) statistical analysis (p < 0.0001) and multiple comparison tests to compare

the concentration of each cytokine to their respective pre-NVP plasma cytokine

concentration (***p < 0.0001). .................................................................................. 80

Figure 4.1: Example IFN-γ ELISpot raw data. The unstimulated background and

solvent controls (DMSO/DMF at 1μL/mL), positive controls anti-CD3 antibody (wells

B1 and B2 0.1μg/mL) and SEB (wells B3 and B4 1μg/mL) responses at 200,000 cells

per well. The test positive and test negative responses highlight the typical positive

and negative response to the test stimulant (drug or peptide 10μg/mL) with an input

of 50,000 and 200,000 cell per well. .......................................................................... 91

Figure 4.2 The Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 1 (CBZ-

TEN) were determined using ELISpot, described in Chapter 2, Enzyme-Linked

ImmunoSpot 2.2.3. Cut-off for positive responses, 50 SFU/million cells (red line).

Error bars represent intra-sample variation (n=3). .................................................... 93

Figure 4.3: The Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 2 (CBZ-

SJS) were determined using ELISpot, described in Chapter 2, Enzyme-Linked



xxi

Figure 4.4: The Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 3 (CBZ-

SJS) were determined using ELISpot, described in Chapter 2, Enzyme-Linked



Figure 4.5: Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 4 (CBZ-SJS

non- B*15:02) were determined using ELISpot, described in Chapter 2, Enzyme-

Linked ImmunoSpot 2.2.3. Cut-off for positive responses, 50 SFU/million cells (red

line). Error bars represent intra-sample variation (n=3). ............................................ 96

Figure 4.6: Mean (+/- SD) IFN-γ responses SFU/million cells in Control 1 (CBZ-naive)

were determined using ELISpot, described in Chapter 2, Enzyme-Linked



Figure 4.7: Mean (+/- SD) IFN-γ responses SFU/million cells in Control 2 (CBZ-naive

non-B*15:02) were determined using ELISpot, described in Chapter 2, Enzyme-

Linked ImmunoSpot 2.2.3. Cut-off for positive responses, 50 SFU/million cells (red

line). Error bars represent intra-sample variation (n=3). ............................................ 98

Figure 4.8: Mean (+/- SD) IFN-γ responses SFU/million cells to CBZ and predicted

HLA-B*15:02-restricted peptides identified in Figure 4.2 to Figure 4.7, determined

using ELISpot, described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3. Cut-off for

positive responses, 50 SFU/million cells (red line). Error bars represent intra-sample

variation (n=3). .......................................................................................................... 99

Figure 4.9: Mean (+/- SD) IFN-γ responses SFU/million cells to the increasing

concentrations of the predicted HLA-B*15:02-restricted peptide, YAHGHSLRLPY in

the three HLA-B*15:02 positive Donors were determined using ELISpot, described in

Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3 with a concentrations of 0.1 μg/ml, 1

μg/ml, and 10 μg/ml. Error bars represent intra-sample variation (n=3) and the p

values were determined using the Pearson correlation test (*p < 0.05). ................. 100


concentrations of the predicted HLA-B*15:02-restricted peptide, LAFVLDSPSAY in

the three HLA-B*15:02 positive Donors were determined using ELISpot, described in



values were determined using the Pearson correlation test (*p < 0.05). ................. 101

xxii


concentrations of the predicted HLA-B*15:02-restricted peptide, NIRQAGVQY in the

three HLA-B*15:02 positive Donors were determined using ELISpot, described in



values were determined using the Pearson correlation test (*p < 0.05) .................. 102

Figure 4.12 Mean (+/- SD) IFN-γ responses SFU/million cells to the increasing

concentrations of the predicted HLA-B*15:02-restricted peptide, FAFTINTAAY in the

three HLA-B*15:02 positive Donors were determined using ELISpot, described in


μg/ml, and 10 μg/ml. Error bars represent intra-sample variation (n=3). ................. 103

Figure 4.13 MS/MS analysis of YAHGHSLRLPY, original responsive peptide (left)

and pure form for tetramer folding (right). A Spectrum was obtained in de novo

peptide analysis via the MALDI TOF MS (4800 Proteomics analyser). ................... 104

Figure 4.14 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 1 (CBZ-TEN)

to the YAHGHSLRLPY peptide and the missing histidine variants, YAHG_SLRLPY,

YA_GHSLRLPY, and YA_G_SLRLPY determined using ELISpot, described in

Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3. Cut-off for positive responses, 50

SFU/million cells (red line). Error bars represent intra-sample variation (n=3). ....... 106

Figure 4.15 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 2 (CBZ-SJS)





Figure 4.16 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 3 (CBZ-SJS)





xxiii

Figure 4.17 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 4 (CBZ-SJS

non- B*15:02) to the YAHGHSLRLPY peptide and the missing histidine variants,

YAHG_SLRLPY, YA_GHSLRLPY, and YA_G_SLRLPY determined using ELISpot,

described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3. Cut-off for positive

responses, 50 SFU/million cells (red line). Error bars represent intra-sample variation

(n=3). ....................................................................................................................... 109

Figure 4.18 Mean (+/- SD) IFN-γ responses SFU/million cells in Control 1 (CBZ-

naive) to the YAHGHSLRLPY peptide and the missing histidine variants,




(n=3). ....................................................................................................................... 110

Figure 4.19 Mean (+/- SD) IFN-γ responses SFU/million cells in Control 2 (CBZ-naive

non- B*15:02) to the YAHGHSLRLPY peptide and the missing histidine variants,




(n=3). ....................................................................................................................... 111

Figure 4.20 Mean (+/- SD) IFN-γ responses SFU/million cells to the increasing

concentrations of the predicted HLA-B*15:02-restricted variant peptide,

YAHG_SLRLPY in the three HLA-B*15:02 positive CBZ-SJS/TEN Donors and Donor

4 (CBZ-SJS non-B*15:02) were determined using ELISpot, described in Chapter 2,

Enzyme-Linked ImmunoSpot 2.2.3 with a concentrations of 0.1 μg/ml, 1 μg/ml, and

10 μg/ml. Error bars represent intra-sample variation (n=3) and the p values were

determined using the Pearson correlation test (*p < 0.05) and r2 values non-linear

regression analysis (n=4) ........................................................................................ 112

Figure 4.21: Donor 1 CBZ-TEN, B*15:02-YAHG_SLRLPY, LAFVLDSPSAY, and

NIRQAGVQY tetramer stain, performed as outlined in Chapter 2, Tetramers 2.2.2.3.

Samples were compared to negative control which consisted of unstained sample

spiked with PE and washed out as per the Tetramer staining protocol. Error bars

represent intra-sample variation (n=5). ................................................................... 113

xxiv

Figure 4.22: Donor 2 (CBZ-SJS), B*15:02-YAHG_SLRLPY, LAFVLDSPSAY, and

NIRQAGVQY tetramer stain performed as outlined in Chapter 2, Tetramers 2.2.2.3.




Figure 4.23: Donor 3 (CBZ-SJS), B*15:02-YAHG_SLRLPY, LAFVLDSPSAY, and

NIRQAGVQY tetramer stain performed as outlined in Chapter 2, Tetramers 2.2.2.3.




Figure 4.24: Control 2 (CBZ-naive non-B*15:02), B*15:02-YAHG_SLRLPY,

LAFVLDSPSAY, and NIRQAGVQY tetramer stain performed as outlined in Chapter

2, Tetramers 2.2.2.3. Samples were compared to negative control which consisted of

unstained sample spiked with PE and washed out as per the Tetramer staining

protocol. Error bars represent intra-sample variation (n=5). .................................... 115

Figure 5.1: Left: Example IFN-γ ELISpot raw data. The unstimulated background and

solvent controls (DMSO/DMF at 1μL/mL), positive controls anti-CD3 antibody (wells

B1 and B2, 0.1μg/mL) and SEB (wells B3 and B4, 1μg/mL) responses at 200,000

cells per well. The test positive and test negative responses highlight the typical

positive and negative response to the test stimulant (drug or peptide 10μg/mL) with

an input of 50,000 and 200,000 cell per well. Right: Mean (+/- SD) IFN-γ responses

SFU/million cells in CBZ-SJS/TEN Donors and Controls, both pre- and post-CBZ

stimulation and culture, were determined using ELISpot, described in Chapter 2,

Enzyme-Linked ImmunoSpot 2.2.3. Error bars represent intra-sample variation (n=3)

................................................................................................................................ 124

Figure 5.2: Left: Example ICS raw data plots showing the CBZ-associated cellular

IFN-γ responses both pre- and post-stimulation (CBZ 10μg/mL) in Donor 1 (CBZ-

TEN). Right: Mean (+/- SD) IFN-γ ICS flow cytometry analysis for CBZ-SJS/TEN

Donor and controls cells both pre- and post- stimulation and culture were determined

using the ICS method outlined in Chapter 2, Intracellular cytokine staining 2.2.2.1.

Error bars represent intra-sample variation (n=3). ................................................. 124

Figure 5.3: TCR Vβ CDR3 sequence profile of Donor 1 CBZ-TEN. Raw data

analysed using Bio-Rad QuantaSoft software, results normalised against RPHH1

expression, and calculated as a percentage (%) of normalised Cβ expression.

xxv

Statistical significance was determined using analysis of variance (ANOVA) statistical

analysis (p value < 0.0001) and multiple comparison tests showed significant

differences between CBZ-stimulated cells to the other 3 conditions (***p value <

0.0001, *p value < 0.05). ......................................................................................... 126

Figure 5.4: Vβ TCR CDR3 sequence profile of Donor 2 (CBZ-SJS). Raw data






0.0001). ................................................................................................................... 127

Figure 5.5: Vβ TCR CDR3 sequence profile of Donor 3 (CBZ-SJS). Raw data






0.0001). ................................................................................................................... 128

Figure 6.1: Left: Example immunophenotyping and ICS raw data plots showing the

gating procedure used to isolate the cellular subpopulations and determine their

cytokine expression under the relevant conditions. Top Right: Example of negative to

low Granulysin expression. Bottom Right: Example of positive granulysin expression.

................................................................................................................................ 137

Figure 6.2: Mean (+/- SD) Granulysin expression in CD8+ cytotoxic T cells from

Donor and Control PBMC’s incubated ex-vivo with CBZ, control drug, or left

unstimulated. Statistical significance was determined using ANOVA analysis (p <

0.0001). Multiple comparison tests showed significance between unstimulated and

CBZ-stimulated donor PBMC’s (* p value < 0.05), CBZ and control drug-stimulated

donor PBMC’s (*** p value < 0.0001), CBZ and the unstimulated, CBZ and control

drug-stimulated donor PBMC’s (*** p value < 0.0001). The error bars represent inter-

sample variation (n=4). ............................................................................................ 138

xxvi

Figure 6.3: Mean (+/- SD) Granulysin expression in CD8+ cytotoxic T cells in Donor

and Control PBMC’s incubated ex-vivo with CBZ versus unstimulated PBMC’s.

Statistical significance between unstimulated and CBZ-stimulated for each set of

donors (***p value of < 0.0001), using the paired two-tailed T-test, compared to the

controls (not significant (ns)). The error bars represent intra-sample variation (n=4).

................................................................................................................................ 139

Figure 6.4: Mean (+/- SD) INF-γ expression in CD8+ cytotoxic T cells from Donor and

Control PBMC’s incubated ex-vivo with CBZ, control drug, or left unstimulated.

Statistical significance was determined using ANOVA analysis (p < 0.0001). Multiple

comparison tests showed significance between, unstimulated and CBZ-stimulated

donor PBMC’s (*** p value < 0.0001), CBZ and Control drug-stimulated donor

PBMC’s (*** p value < 0.0001), CBZ and the unstimulated, CBZ and Control drug-

stimulated donor PBMC’s (*** p value < 0.0001). The error bars represent inter-

sample variation (n=4). ............................................................................................ 140

Figure 6.5: Mean (+/- SD) IFN-γ expression in CD8+ cytotoxic T cells in Donor and

Control PBMC’s incubated ex-vivo with CBZ versus unstimulated PBMC’s. Statistical

significance between unstimulated and CBZ-stimulated Donor 1 (CBZ-TEN) (***p

value of < 0.0001), Donor 2 (CBZ- SJS) (*p value of < 0.05) and Donor 4 (CBZ-SJS

non- B*15:02) (*p value of < 0.05) using the paired two-tailed T-test. Donor 3 (CBZ-

SJS) and Controls CBZ-naive (both HLA-B*15:02 positive and negative) were not

significant (ns)). The error bars represent intra-sample variation (n=4)................... 141

Figure 6.6: Mean (+/- SD) Granulysin expression in CD3+ CD4+ CD56- T cells in


unstimulated. No statistical significance was determined between the means of the

various conditions. (ANOVA analysis and multiple comparison tests).The error bars

represent inter-sample variation (n=4). ................................................................... 143

Figure 6.7: Mean (+/- SD) Granulysin expression in CD4+ T cells in Donor and

Control PBMC’s incubated ex-vivo with CBZ versus unstimulated PBMC’s. Statistical

significance was found between Donor 4 (CBZ-SJS non- B*15:02) unstimulated and

CBZ-stimulated PBMC’s (*** p value < 0.0001). HLA-B*15:02 positive Donors 1

(CBZ-TEN), Donor 2 (CBZ-SJS), and Donor 3 (CBZ-SJS) unstimulated versus CBZ-

stimulated PBMC’s, and the Controls were found to be not significant using a paired

two-tailed T-test. The error bars represent intra-sample variation (n=4). ................ 144

xxvii

Figure 6.8: Mean (+/- SD) IFN-γ expression in CD4+ T cells in Donor and Control

PBMC’s incubated ex-vivo with CBZ versus unstimulated PBMC’s. Statistical

significance was found between Donor 4 (CBZ-SJS non- B*15:02) unstimulated and

CBZ-stimulated PBMC’s (*** p value < 0.0001). HLA-B*15:02 positive Donors 1

(CBZ-TEN), Donor 2 (CBZ-SJS), and Donor 3 (CBZ-SJS) unstimulated versus CBZ-

stimulated PBMC’s, and the Controls were found to be not significant using a paired

two-tailed T-test. The error bars represent intra-sample variation (n=4). ................ 145

Figure 6.9: Mean (+/- SD) Granulysin expression in CD56+ NK cells in Donor and

Control PBMC’s incubated ex-vivo with CBZ, control drug, or left unstimulated. The

error bars represent variation within the same sample on multiple test (N tests per

sample in each condition = 4). ................................................................................ 146

Figure 6.10: Mean (+/- SD) Granulysin expression in CD3+ CD8+ CD56+ NK-T cells in


unstimulated. The error bars represent intra-sample variation (n=4). ..................... 147

Figure 6.11: Mean (+/- SD) Granulysin expression in CD3+ CD4+ CD56+ NK-T cells in


unstimulated. The error bars represent inter-sample variation (n=4). ..................... 148

Figure 9.1: Analysis of specific Vβ CDR3 primers using serial dilutions of the

reference sequence amplicons, with the annealing temperature set at 58◦C .......... 182

Figure 9.2: Analysis of specific Vβ CDR3 primers using serial dilutions of the

reference sequence amplicons, with the annealing temperature set at and 60◦C. .. 183

Figure 9.3: the QuantaSoft raw digital droplet PCR plots. Reference sequence

amplicons, serially diluted to test the sensitivity of the digital droplet PCR and

qualifying the assay parameters. ............................................................................. 184

Figure 9.4: the QuantaSoft digital droplet PCR output of positive and total events

recorded for the raw plots outlined in Figure 9.10 Reference sequence amplicons,

serially diluted to test the sensitivity of the digital droplet PCR and qualifying the

assay parameters. ................................................................................................... 185

Figure 9.5: Digital droplet PCR Reference sequence amplicons serial dilution result

and QuantaSoft output (copies/μL). Reference sequence amplicons, serially diluted

to test the sensitivity of the digital droplet PCR and qualifying the assay parameters.

................................................................................................................................ 185

xxviii

Figure 9.6: Mean (+/- SD) Granulysin expression by CD8+ T cells, stimulated by the

epitopes discovered in Chapter 4 Predicted HLA-Restricted Peptide Screening

(10μg/mL) and CBZ (10μg/mL). .............................................................................. 186

Figure 9.7: TCR Vβ-12 CDR3 LAGELF and Vβ-11 CDR3 ISGSY sequence profile of

Donor 1 (CBZ-TEN) when stimulated with stimulated by the epitopes discovered in

Chapter 4 Predicted HLA-Restricted Peptide Screening (10μg/mL) and CBZ

(10μg/mL). Raw data analysed using Bio-Rad QuantaSoft software, results

normalised against RPHH1 expression, and calculated as a percentage (%) of

normalised Cβ expression. ...................................................................................... 187

Figure 9.8: Abstract of HLA Class-I restricted CD8 (+) and Class-II restricted CD4 (+)

T cells are implicated in the Pathogenesis of Nevirapine Hypersensitivity paper.

Published in AIDS, London, 24 August 2014 - Volume 28 - Issue 13 - p 1891–1901.

DOI:10.1097/QAD.0000000000000345 (138). ........................................................ 188

Figure 9.9: Abstract of Testing for Drug Hypersensitivity Syndromes. Published in

The Clinical biochemist. Reviews / Australian Association of Clinical Biochemists

02/2013; 34(1):15-38. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3626363/ (2).

................................................................................................................................ 189

Figure 9.10: Poster presented at Science on the Swan, April 2015.

https://www.researchgate.net/publication/281661878_HLA_Class_I-drug-T-

cell_receptor_interaction_and_cytokine_expression_in_SJSTEN DOI:

10.13140/RG.2.1.4937.5207 .................................................................................. 190

Figure 9.11: Poster presented at Conference on Retroviruses and Opportunistic

Infections (CROI), Boston, Massachusetts; 2014.

https://www.researchgate.net/publication/275332026_Specific_Binding_

Characterstics_of_HLA_Alleles_Associate_with_Nevirapine_Hypersensitivity DOI:

10.13140/RG.2.1.1206.1287. (241) ......................................................................... 191

Figure 9.12: Abstract for talk presented at Drug hypersensitivity meeting 6 (DHM6),

Bern Switzerland April, 2014. .................................................................................. 192

Figure 9.12: Abstract for talk presented at Drug hypersensitivity meeting 6 (DHM6),

Bern Switzerland April, 2014. .................................................................................. 193

1

Chapter 1

Introduction: Immunopathogenesis of Drug Hypersensitivity

2

3

1 1.1 Adverse Drug Reactions

Adverse drug reactions (ADR) represent a major cause of iatrogenic and preventable

disease and death (1, 2). A meta-analysis covering 39 different studies from 1966 to

1996 estimates that 2.2 million serious ADR occur every year in the US alone, with

over 100 thousand (4.5%) resulting in death (3-6). ADR are a common cause of

mortality in the developed world (2). ADR place high demands on hospitals, healthcare

facilities, and auxiliary healthcare services. The enormous burden and cost of ADR are

incompletely captured and underestimated (1, 5, 7).

Even with the stringency and oversight governing drug patents, development, and

licensing an estimated 1500 patients are likely to be exposed to a newly developed

drug during pre-clinical, pre-marketing testing phases (8-10). Only those ADR which

show a frequency of 1 in 500 or greater are identified during the pre-marketing phases

of drug development (8-10). The burden of ADR are likely to increase over time due to

new medication development, the discovery of new uses for already marketed

medications, and an ageing population in many developed countries. There are no

effective strategies to determine the propensity for a drug to cause severe

immunologically mediated-adverse drug reactions (IM-ADR) during development

(1, 3).

The different ADR classifications have been covered in the literature including the

paper by Rive et al. 2013 entitled Testing for Drug Hypersensitivity Syndromes

(Appendix 9, Journal Publications 9.4.1) (2). This paper reviewed IM-ADR or

hypersensitivity reactions (HSR), in particular, the well-known immediate

immunoglobulin (Ig) E and IgE-like HSR and the T-cell mediated, delayed IM-ADR.

Here we focus on the latter, looking at the severe cutaneous adverse drug reactions

(SCARs) that are mediated through adaptive, T cell-mediated immune responses

(2, 11).

1.2 Adaptive, Cell-Mediated Immune Response

Delayed IM-ADR are driven through adaptive, cell-mediated immune responses.

Adaptive, cell-mediated immunity involves T-cell activation through the T-cell receptor

(TCR) recognition of antigenic molecules, presented by the major histocompatiblity

complex (MHC) or human leukocyte antigen (HLA) cell-surface molecules.

4

1.2.1 Human Leukocyte Antigen

The antigen-presenting cell-surface HLA proteins are encoded by HLA genes in the

MHC chromosomal region, a 3.6 megabase pair section of deoxyribonucleic acid

(DNA) on the short arm of chromosome number 6 in humans (6q21.3). HLA class-I

proteins are encoded by three major loci: HLA-A, HLA-B, and HLA-C and three minor

loci: HLA-E, HLA-F, and HLA-G while the HLA class-II proteins are encoded by three

major loci: HLA-DR, HLA-DP, and HLA-DQ and two minor loci; HLA-DM and HLA-DO

(Figure 1.1) (2, 12). The HLA class-III region includes both immune and non-immune

associated genes like the tumour necrosis factor gene and heat shock proteins' genes

(Figure 1.1) (13, 14).

HLA class-I and class-II genes are highly variable, with polymorphic changes resulting

in amino acid changes within the HLA peptide-binding cleft/groove of the specific HLA

protein it encodes, influencing the peptide-binding capabilities of each HLA allotype

and therefore altering the repertoire of peptides that each HLA allotype can present

(15, 16).

Figure 1.1: Human chromosome 6, the location of Major histocompatiblity complex (MHC 6p21.1-21.3)

and the Human leukocyte antigen (HLA) class-I, class-II, and class-III genes. Adapted from Urayama

KY et al. 2013 (13)

1.2.2 Antigen Processing and Presentation

1.2.2.1 HLA class-I endogenous antigen processing

HLA class-I molecules are expressed by all nucleated cells and are composed of a

conserved Beta 2-Microglobulin (β2M) and a polymorphic alpha (α) chain, made of

5

three domains, α1 to α3 (Figure 1.2a) (17, 18). The peptide-binding cleft consists of

two antiparallel α-helices, formed by the α1 and α2 domains of the α chain and eight

antiparallel beta (β) sheets, which provides a binding cleft with specific peptide-binding

cleft pockets, identified as pockets A to F, at its base (Figure 1.2c) (17-19). The

chemical characteristics and size of each binding cleft pocket define the repertoire of

peptides that can be accommodated by the different HLA allotypes.

As highlighted in Figure 1.3a antigenic peptide presentation by HLA class-I

cell-surface protein to the TCR of T cells co-expressing the cluster of differentiation

(CD) 8 molecule, requires the endogenous processing of foreign or abnormal antigens

present in the cytoplasm (18, 20-22). HLA class-I bind and express antigenic peptides

8-9 amino acids in length, however, longer antigenic peptides of ten or more amino

acids can be accommodated by HLA class-I (17, 18).

1.2.2.2 HLA Class-II exogenous antigen processing

HLA class-II molecules differ from HLA class-I molecules in that, under normal

circumstances, HLA class-II are only expressed by specialised professional antigen-

presenting cells (APCs), including dendritic cells (DC), macrophages,

B cells, and certain activated epithelial cells. HLA class-II molecules are consists of

two α chain and two β chain domains (Figure 1.2b) (17, 18). The peptide-binding

groove is made by the interaction between the α1 and β1 domains which provides a

binding groove with specific peptide-binding groove pockets, P1 to P9 (Figure 1.2d)

(17-19). The binding groove pockets accommodate the amino acids residues of a

bound peptide.

As highlighted in Figure 1.3b, antigenic peptide presentation by HLA class-II

molecules, to the TCR of CD4 expressing (CD4+) T cells, requires the exogenous

processing of phagocytised antigens (18, 20-22). Unlike HLA class-I, HLA class-II

molecules present peptides 12 to 25 amino acids in length (17, 18).

1.2.2.3 Cross-presentation

Cross-presentation of antigenic peptides differs from the classical antigen presentation

models as it involves exogenous derived antigenic peptides being presented on HLA

class-I molecules (Figure 1.2c) (20, 23-25). Cross-presentation plays an essential role

in the defence against several viruses, bacteria; and tumours which have shown an

ability to use immune evasion methods, like antigen processing suppression. (20, 23-

25)

6

Figure 1.2: Side view of (A) HLA class-I and (B) HLA class-II molecules. Also represented are the top

view of the (C) HLA class-I peptide-binding cleft pockets A-F and (D) HLA class-II peptide-binding

groove pockets 1-9. Adapted from Cole DK, 2013 (17).

Figure 1.3: Human leukocyte antigen (HLA) and the processing and presentation of antigens. (A)

Endogenous antigen processing and HLA class-I presentation. (B) Exogenous antigen processing and

HLA class-II presentation. (C) HLA class-I cross-presentation of exogenous antigens derived via

endocytosis. β2M, Beta 2-Microglobulin; TCR, T- cell Receptor; ER, endoplasmic reticulum; TAP 1 and

TAP 2, transporter associated with antigen processing 1 and 2; CLIP, Class-II-associated invariant

chain peptide. Adapted from, Guermonprez P, et al. 2002, Heath WR and Carbone FR, 2001, Braciale

et al. 1987, Boesteanu A et al. 1998, and Kobayashi KS et al. 2012 (18, 20-22, 25)

7

1.2.3 T Cells

The generation and regulation of T-cell immune responses is directed by the HLA-

associated antigen processing and presentation, and recognition by the TCR. The

TCR plays an essential role in the development and selection of T cells with enough

TCR diversity while eliminating any TCR which may interact with self-peptides leading

to autoimmune and unwanted responses. Here we will concentrate on the α- and

β-chain derived TCR (αβTCR) development.

1.2.3.1 Thymic development

The lymphoid precursor cells move from the bone marrow to the thymus, where the

TCR is formed by the rearrangement of variable (V) segments with the joining (J)

and/or diverse (D) TCR gene segments (Figure 1.4) (26-28). This process of

recombination, transcription, splicing and translation, protein folding, and the formation

of the TCR heterodimer produces three unique variable regions within the TCR

molecule, the complementarity determining region (CDR) loops (26).

The three hypervariable CDR loops of both the α- and β-chains within the αβTCR are

essential in the interaction with the peptide-HLA complex. The CDR3 loop interacts

with the peptide in the peptide-HLA complex and its variability come from the D and/or

J gene segments rearrangement during recombination (Figure 1.4) (17, 28, 29). As

highlighted in Figure 1.5, the TCR CDR3 α-chain (Orange α3) interacts with the N-

terminus and the TCR CDR3 β-chain (Orange β3) with the C-terminus of a bound

peptide (17). The other variable regions within the TCR, the CDR1 and CDR2 loops,

are encoded within the V gene segments and interact with the HLA molecule in the

peptide-HLA complex, next to the peptide interacting CDR3 loops, with the CDR1 and

CDR2 loops on the TCR α-chain (blue α1 and α2) interacting with the HLA class-I α2

domain and the HLA class-II β1 domain and the CDR1 and CDR2 loops on the TCR β-

chain (Green β1 and β2) interacting with the HLA class-I and HLA class-II α1 domain

(Figure 1.5) (17, 30, 31). This rearrangement of TCR DNA and the variability in the

CDR regions allows for 1 x 1018 or greater different combinations for the αβTCR (17,

30, 31).

8

Figure 1.4: Germline DNA T-cell receptor (TCR) rearrangement, translation, splicing and transcription.

Variable alpha (Vα), Variable beta (Vβ), Joining alpha (Jα), Joining beta (Jβ), diverse beta (Dβ),

Constant alpha (Cα) and Constant beta (Cβ) DNA chains, adapted from Janeway CA et al. 2001 (26)

Figure 1.5: Peptide-loaded human leukocyte antigen (HLA) molecule (A) class-I and (B) class-II

complexes and T-cell receptor (TCR) complementarity-determining region (CDR) loop interactions.

CDR3 (Orange α3 and β3), CDR1 and CDR2 β-chains (Green β1 and β2) CDR1 and CDR2 α-chains

(Blue α1 and α2). Figure adapted from Cole DK, 2013 (17).

9

In the thymus, T-cell development leads to CD8+ / CD4+ double positive T cells which

express a mature αβTCR that interacts with the self-peptides presented by HLA class-I

or HLA class-II being expressed on the surface of the thymus cortical epithelial cells

(32-35). A double positive T cell which expresses a TCR that has either a high or low

affinity for the self-peptide being presented in a peptide-HLA complex undergoes

apoptosis. A TCR exhibiting an appropriate affinity for self-peptides presented on

either the HLA class-I or HLA class-II molecules, become single positive, naive CD8+

or CD4+ T cells, respectively. These single positive naive T cells then migrate out of

the medulla of the thymus to secondary lymphoid tissues such as the lymph nodes,

tonsils, spleen, Peyer’s patches and mucosa-associated lymphoid tissue (32-35).

1.2.3.2 Activation

Once in the secondary lymphoid tissue, a naive T cell becomes activated once it

interacts with the appropriate HLA-antigen presented by an APC, which has moved to

the secondary lymphoid tissue after encountering an antigen (Figure 1.6) (36-38).

Following activation, T cells undergo clonal expansion which produces numerous

activated effector T cells, expressing cell-surface molecules that allow them to leave

secondary lymphoid tissues and move to the site of inflammation (Figure 1.6). The

differentiation and effector functions of both, CD8+ T cells and CD4+ T cells, differ (39,

40). Effector CD4+ T cell or helper T (Th) cells direct immune responses by regulating

macrophages and B-cell antibody production, maintaining CD8+ T-cell responses, and

the recruitment and direction of innate response cells like neutrophils, eosinophils and

basophils to sites of infection and inflammation (41-43). The cytokine environment and

the initial antigen interaction dictates how the naive CD4+ T cell will differentiate into

one of several CD4+ Th cells, including Th1, Th2, Th17, regulatory T cells (Tregs),

Th22 or Follicular Th cells (41, 42, 44).

Effector CD8+ cytotoxic T cells migrate out of the secondary lymphoid organs and

move to the site of infection, interacting with cells expressing HLA class-I molecules

presenting specific antigen-peptides (41-43). Once an infected cell(s) is identified,

apoptosis is induced eliminating the infected cell. Effector T cells then undergo

apoptosis-dependent programmed contraction, an important phase in immune

homeostasis (41-43, 45). A small proportion of pathogen-specific T cells not

programmed for apoptosis will populate a pool of memory T cells, whose numbers are

sustained by low-level homeostatic proliferation (46, 47).

10

1.2.3.3 Memory

Memory T cells continue to act in a surveying manner, reacting when exposed to the

same specific antigen. Understanding into CD4+ memory T cell remains elusive when

compared to CD8+ memory T cells. CD8+ memory T cells are composed of central

memory T cells which remain in the secondary lymphoid organs, effector memory

T cells which circulate through non-lymphoid tissue, and the resident memory T cells

which remain in the tissue of the primary infection (Figure 1.6) (34, 48-50). Memory

T cells require a lowered activation threshold and do not require the same co-

stimulator signals governing naive T-cell activation (34).

Figure 1.6: Naive CD8+ T-cell activation, proliferation and clonal expansion. Tcm, central memory T

cells; Tem, effector memory T cells; Trm, resident memory T cells; DC, dendritic cells. Adapted from

Kaech S and Weiguo C, 2012 (51)

1.2.4 Heterologous Immunity

Heterologous immunity describes the immune response by antigen-specific memory

T cells from a previously encountered pathogen, cross-recognising, and influence the

immune response to a new or neo-antigen (52). Heterologous immunity uses a

restricted TCR repertoire directed at cross-reactive derived, subdominant epitopes

compared to homologous immunity in which a diverse TCR repertoire is directed at

specific epitopes maintained from primary infection (52). Heterologous immunity is

considered not as efficient and effective as homologous immunity, yet the severity and

course of infection may be reduced because of heterologous immunity (13, 14).

Diverging from normal immunity to a partial and non-protective immune response can

also result in altered immunopathology and even increased morbidity and mortality

(52, 53).

11

A considerable amount of evidence supports the heterologous immune model as a

mediator of organ transplant rejection (9, 10, 12, 16-19). Organ transplant offers the

unique circumstance where an abundance of neo-antigens, for which the host has no

tolerance to, are presented to the immune system and cross-recognised by pre-

existing HLA-restricted memory T cells, derived in response to and maintained by

prevalent viral infections, like those of the human herpesvirus (HHV) family, including

Herpes Simplex Virus-1 and Herpes Simplex Virus-2 (HSV-1 and HSV-2,

respectively), Epstein-Barr virus (EBV), and cytomegalovirus (CMV) (54-60).

1.3 Immunopathogenesis of Delayed

Immunologically Mediated-Adverse Drug

Reactions

Delayed IM-ADR involves a range of phenotypes of differing severity that are defined

by their clinical presentation. T cell-mediated reactions can often be defined by their

immunopathogenesis at a tissue level and by the effector cell involved. Stevens -

Johnson syndrome (SJS) / Toxic Epidermal Necrolysis (TEN) and drug reaction with

eosinophilia and systemic symptoms (DRESS) are among the severest of IM-ADR

with the former having up to 50% mortality (Table 1.1) (1, 2, 61, 62).

In SJS/TEN patients develop a widespread erythematous rash with flat or macular

atypical targets and bullae blisters with sections or sheets of the epidermis separating

from the dermis (2, 62-64). The spectrum of severity includes the percentage of body

surface epidermal detachment associated with bullae blister development. SJS

involves extensive small blisters with less than 10% epidermal detachment, SJS-TEN

overlap 10–30% epidermal detachment, and TEN involves greater than 30%

epidermal detachment (Table 1.1) (2, 62-65). In SJS/TEN short-term complications

include organ involvement and sepsis and longer-term complications involve the eyes

(adhesions or blindness), loss of nails, adhesions and structures throughout the

respiratory, gastrointestinal and urogenital tract and neuropsychiatric complications

including post-traumatic stress disorder and depression (Table 1.1) While only 0.4-6

incidences of SJS/TEN occur per million persons per year, SJS/TEN is a severe, life-

threatening ADR, with mortality rates ranging from 5 to 12.5% for SJS and 30 to 50%

for TEN and is associated with over 200 different drugs including several antiepileptic

drugs like carbamazepine (CBZ) (Table 1.1) (2, 63, 65, 66).

12

DRESS has a longer latency period than SJS/TEN with the first manifestation being

fever, followed by a rash of varying severity and involvement of organs such as the

liver. Besides fever, rash and organ involvement, other symptoms include

lymphadenopathy and haematological abnormalities including eosinophilia and/or

lymphocytosis (Table 1.1). Up to 80% of DRESS cases involve the liver. However,

DRESS can also include kidney damage (40%), and/or the lungs, heart, and the

pancreas (67). Symptoms appear within a two- to eight-week window after starting the

culprit drug and can persist after drug discontinuation or withdrawal (2, 68-70). Like

SJS/TEN, DRESS is associated with numerous drugs including the antiretroviral

(ARV) medication, nevirapine (NVP) (Table 1.1)

Despite not being a SCAR, delayed rash or maculopapular eruption (MPE) deserves a

brief mention (Table 1.1). Delayed rash often presents as macules or papules

erythematous, without other systemic features, appearing on the face and torso,

before becoming generalised (71, 72). Delayed rash has been associated with many

drugs, including NVP and CBZ, and symptoms appear at the end of the first week or

beginning of the second week of treatment with the culprit drug (Table 1.1) (71, 72).

Delayed drug-rash reactions have been associated with drug-viral interactions, for

example, the interaction between aminopenicillins and EBV (73).

In the last decade, the association between HLA alleles and severe delayed

IM-ADR has led to a greater understanding into the immunopathogenesis of these

diseases and the ability to use these specific HLA alleles as biomarkers. This began

with the association between HLA-B*57:01 and abacavir (ABC) associated

hypersensitivity syndrome (HSS) (Table 1.1).

13

Table 1.1: Delayed IM-ADR and their significant features. Adapted from Rive C et al. 2013, Phillips EJ

et al. 2010, and Pavlos R et al. 2012 (2, 67, 74)

HSR / IM-ADR

Time onset

Symptoms/features Associated drugs

Delayed Rash / MPE

end of first week or beginning of second week of treatment

Maculopapular rash appears on the face / torso, becoming generalised. No associated fever or systemic symptoms.

Antiepileptic drugs (including CBZ), aminopenicillins. antibacterial sulphonamides, ARV (NVP), Allopurinol

DRESS

2-8 weeks

Fever, rash, lymphocytosis (eosinophilia), lymphadenopathy, Liver / organ involvement. Viral reactivation of HHV6/7, EBV or CMV, 2 weeks or more. Possible delayed autoimmune disease.

Antiepileptic drugs (including CBZ),beta-lactam antibiotics, vancomycin, antibacterial sulphonamides, NSAIDs, and ARV’s (NVP)

ABC-HSS median 8 days

Fever, malaise, Gastrointestinal symptoms, 70% develop late mild to moderate rash. ABC rechallenge results in Hypotension, shock and even death.

ABC patients now screened for HLA-B*57:01

SJS/TEN 4-28 days (majority)

Skin detachment: SJS 1–10% body surface area SJS/TEN overlap 10–30% body surface area TEN >30% body surface area Fever and mucosal tissue involvement (precede rash), Hepatitis, jaundice, increased liver transaminases Eye involvement (post reaction complications), Gastrointestinal and respiratory manifestations, anaemia, lymphocytosis (neutropenia).

> 200 drugs Antiepileptic drugs (including CBZ, antibacterial sulphonamides, NSAID’s, ARV’s (NVP), and Allopurinol.

HSR, hypersensitivity reaction; MPE, maculopapular eruption; DRESS, drug reaction with eosinophilia and systemic symptoms; HSS, hypersensitivity syndrome; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis, HHV, human herpesvirus; EBV, Epstein-Barr virus; CMV, cytomegalovirus;, CBZ, carbamazepine; ARV, antiretroviral; NVP, nevirapine; NSAID, non-steroidal anti-inflammatory drugs.

1.3.1 Abacavir

ABC is a guanosine analogue ARV agent used in combination treatment of human

immunodeficiency virus (HIV). ABC is a nucleoside that inhibits the reverse

transcriptase enzyme, resulting in chain termination. Since the approval of ABC in

1998, 4-8% of patients receiving ABC would also experience the delayed IM-ADR,

now identified as ABC-HSS (67, 74, 75). The clinical presentation of ABC-HSS

includes symptoms such as fever, malaise, as well as respiratory and gastrointestinal

14

symptoms. (Table 1.1) (67, 68, 74, 76). A mild to moderate rash develops in 70% of

cases, which appears late in disease (Table 1.1) (67, 74, 76). White cell count

abnormalities like eosinophilia and hepatitis are uncommon (67, 68, 74). ABC-HSS

symptoms can appear within 36 hours to 6 weeks of ABC initiation, although most

cases occur within 2 weeks of first ABC exposure (74) The majority of ABC-HSS

symptoms are resolved within 72 hours of ABC discontinuation, when rechallenged

with ABC, patients experience extreme hypotension, shock and, even death (Table

1.1) (67, 68, 74, 76).

A genetic link to ABC-HSS was hypothesised with reports of a lowered frequency of

ABC-HSS in Asian and African populations. The report of ABC-HSS occurring in two

members of the same family added to this idea of a genetic association (67, 74, 76,

77). The association between HLA-B*57:01 carriage and ABC-HSS was reported by

two separate research groups. False positive clinical diagnosis minimised “true ABC-

HSS” in non-Caucasian patients where the carriage of HLA-B*57:01 is low (74, 76-78).

The specificity of ABC skin patch testing increased the negative predictive value of

HLA-B*57:01 in ABC-HSS to 100% (74, 79, 80). Further validation of the 100%

negative predictive value was seen in a case-controlled study called the Study of

Hypersensitivity to ABC and Pharmacogenetic Evaluation (SHAPE) (81). SHAPE

showed that both African (African-American) and Caucasian patients who tested

positive in patch tests all carried HLA-B*57:01 (74, 76-78). In a randomised clinical

trial, the Prospective Randomised Evaluation of DNA Screening in a Clinical Trail

(PREDICT-1), showed that screening for HLA-B*57:01 eliminated patch test positive

ABC-HSS patients and HLA-B*57:01 carriage has a 55% positive predictive value for

ABC-HSS (74, 76-79).

Figure 1.7 denotes how the genetic association of the HLA-B*57:01 allele with

ABC-HSS, lead to HLA-B*57:01 genetic screening in clinical practice and improved

treatment guidelines (Figure 1.7) (2, 78). The 100% negative predictive value

associated with HLA-B*57:01 in ABC-HSS showed that the underlying

immunopathogenesis of ABC-HSS is HLA-driven. The incomplete positive predictive

value of 55% associated with HLA-B*57:01 in ABC-HSS, suggests that

immunopathogenesis is driven by other factors, besides HLA-B*57:01 carriage.

A large body of experimental work supports the model that ABC-HSS is driven by

HLA-B*57:01-restricted, CD8+ T-cell response (2, 67, 82). CD8+ T-cell responses can

be seen in the peripheral blood mononuclear cells (PBMC’s) of HLA-B*57:01 positive,

15

ABC-naive Donors by culturing the cells and stimulating them with ABC-pulsed

HLA-B*57:01 transfected APCs (67, 82, 83). Changing a single amino acid in the

peptide-binding domain, like changing the serine at position 116 of the HLA-B*57:01

molecule, to a tyrosine, abolishes ABC-CD8+ T-cell recognition (67, 82, 83).

Virtual modelling data have provided insight into the underlying pathogenesis of HLA-

B*57:01-restricted ABC-HSS, by providing evidence that ABC binds non-covalently to

the peptide-binding cleft of HLA-B*57:01, altering its chemistry and shape (2, 68, 83-

85). The alteration of the peptide-binding cleft changes the repertoire of self-peptides

presented by HLA-B*57:01, from peptides with a tryptophan (W) or a phenylalanine (F)

at the C-terminus, to small aliphatic residue such as valine (V), alanine (A) or

isoleucine (I) at the carboxyl terminus (2, 68, 83-85). These altered peptides presented

when ABC is non-covalently bound to the peptide-binding cleft of HLA-B*57:01, were

not presented during T-cell thymic development and therefore a proportion of these

self-peptides are recognised by T cells of hypersensitive patients (2, 21, 87-89).

Figure 1.7: Abacavir (ABC) translational roadmap Showing how the discovery of the association

between HLA-B*57:01 carriage and ABC- HSS lead to the incorporation of HLA-B*57:01 genetic

screening into clinical practice. Figure adapted from Rive C et al. 2013 and Phillips EJ et al. 2011(2, 78).

16

1.3.2 Carbamazepine

An iminodibenzyl derivative, CBZ was developed in 1960 to treat trigeminal neuralgia

before the discovery of its antiepileptic properties (86). Besides its antiepileptic

properties and its ability to ease neuropathic pain accompanying diseases such as

Herpes Zoster (HZ), CBZ is used to treat a variety of “off-label” disease and conditions

including attention deficit hyperactive disorder, schizophrenia, bipolar disorder,

borderline personality disorder, phantom limb syndrome, and post-traumatic stress

disorder (87-91). The World Health Organization has placed CBZ on the essential

medication list due to its relative costs and multiple uses (92). CBZ-HSR include

maculopapular eruption and DRESS restricted to HLA-A*31:01 in Caucasian and other

ethnicities, however this body of research will concentrate on HLA-B*15:02-restricted

CBZ-SJS/TEN.

1.3.2.1 CBZ-SJS/TEN HLA associations

SJS/TEN is a rare disease with less than 10 in 10 thousand people experiencing the

disease. CBZ-specific SJS/TEN was found to have a strong association with the HLA-

B*15:02 allele in Han Chinese populations (19% of the world’s total population), where

the prevalence of the allele is approximately 10.2% (Figure 1.8) (2, 67). Similar to

ABC-HSS, studies have shown a negative predictive value of 99-100% for HLA-

B*15:02-restricted CBZ-SJS/TEN in Han Chinese populations (93-96). Unlike HLA-

B*57:01 and ABC-HSS, the positive predictive value of HLA-B*15:02 in CBZ-SJS/TEN

is lower (3.0-7.7%) (2). Given that 10.2% of Han Chinese carry the HLA-B*15:02 and

that 19% of the total world’s population (7.125 billion as of 2013) is of Han Chinese

ancestry, this places upwards of 4 to 9 million people worldwide at risk of experiencing

SJS/TEN reaction if given CBZ (97-99).

The CBZ-SJS/TEN-HLA-B*15:02 association has is significant in South and South-

east Asian populations where the HLA-B*15:02 allele frequency is high, with highest

incidence in Han Chinese (10.2%) and Taiwanese (10%) populations, while the

frequency of HLA-B*15:02 carriage in Hong Kong, Thailand, Malaysia, Vietnam,

Philippines, India, and Indonesia is 5% (2, 67, 100). The frequency of HLA-B*15:02 is

low in other populations, including African Americans (0.1-1%) and Caucasians (0.1%)

(2, 67, 100). There is an identified association with other derivatives of CBZ like

oxcarbazepine, and a weaker association between HLA-B*15:02-SJS/TEN and other

Antiepileptic drugs like phenytoin, and lamotrigine.

17

Figure 1.8: The frequency of HLA-B*15:02 allele carriage most prevalent in those of South and South-

east Asian ethnic groups. (2, 67, 100)

Based on the strong negative predictive value associated with CBZ-SJS/TEN and the

carriage of HLA-B*15:02 in South and Southeast Asians, the Food and Drug

Administration (USA) and similar regulatory agencies of different countries have

recommended that genetic screening for HLA-B*15:02 in those of South and South-

east Asian ancestry before CBZ therapy and that all HLA-B*15:02 carriers, should not

be administered CBZ, unless the benefits outweigh the risks (101).

Besides the HLA-B*15:02 allele association, other B75 serotype HLA alleles, including

HLA-B*15:08, -B*15:11, and -B*15:21, have been associated with CBZ-SJS/TEN, in

different populations, with the rationale being that the B75 serotype alleles share high

amino acid sequence homology that may resemble structural features of HLA-B∗15:02

and therefore may trigger a similar cutaneous adverse reaction to CBZ (102). Other

associations also include HLA-B*15:18, -B*59:01 and -C*07:04 (103-106).

Studies involving European and Japanese populations have suggested an association

between the HLA-A*31:01 allele and CBZ-SJS/TEN (107, 108). However, subsequent

studies involving both Taiwanese and European populations have shown a stronger

association between the HLA-A*31:01 allele and CBZ-DRESS / CBZ-MPE reactions

and only a weak to insignificant association with CBZ-SJS/TEN (107-110). This

inconsistency with HLA-A*31:01 might be due to several factors and stresses how

important it is to define the clinical presentation of delayed IM-ADR phenotypes and

eliminating false positive clinical diagnosis, which was clear in the association between

HLA-B*57:01 and “true” ABC-HSS (75).

18

The high negative predictive and incomplete/low positive predictive values associated

with carrying an at-risk HLA allele, like HLA-B*15:02 shows that while specific HLA

allele carriage is a necessary factor, it alone does not explain drug-SJS/TEN

immunopathogenesis.

1.3.2.2 Other factors driving CBZ-SJS/TEN

immunopathogenesis

Immune cells

Multiple studies have emphasised the prominence of T cells found in the blister fluid of

SJS/TEN patients (111-114). In early stages, CD8+ cytotoxic T cells are found in blister

fluid and epidermal layers of the skin, while CD4+ T cells are found in the dermal

layers of the skin (115). This contradicts the common finding that CD4+ T cells are

commonly found early in allergic reactions involving the epidermis, and not CD8+ T

cells (115). An increase in the interleukin (IL)-2 associated, soluble IL-2 receptors

within patient blister fluid and serum further support the role of CD8+ T cells in

SJS/TEN, as they are a marker for activated T cells and levels correlate to disease

activity (111, 113). Activated T cells expressing the cutaneous lymphocyte antigen skin

homing receptor, are elevated in the peripheral blood of patients with SJS/TEN, which

also correlates to disease activity and returns to normal with disease resolution (111,

113, 114).

Other immune cells found in the epidermis of SJS/TEN patients include natural killer

(NK) cells and natural killer T (NK-T) cells which can be identified in the blister fluid of

SJS/TEN patients (111, 116). Several pathogenic models highlighting the role of CD8+

cytotoxic T cells, NK, and NK-T cells in inducing the keratinocyte death seen in

SJS/TEN have been published (63, 111, 117). These models have theorised that the

death of keratinocytes is triggered through the various cellular apoptotic mechanisms.

More recent evidence suggests the cytotoxic effector molecule granulysin is involved

in the disseminated keratinocyte death in SJS/TEN (117).

Granulysin

Granulysin is a novel effector molecule that exhibits cytotoxic activity against several

pathogens and tumours, while also possessing chemoattractant and proinflammatory

abilities. (118-125). The cytotoxic effector molecule granulysin can be expressed by

several cells, including CD8+ cytotoxic T cells, CD4+ cytotoxic-like T cells, NK cells and

NK-T cells (2, 119, 125-127). Granulysin is a useful transplantation biomarker as

granulysin expression in-situ is a marker for acute rejection and steroid resistance

19

(128). Granulysin is also an important cytotoxicity mediator in several skin diseases

besides SJS/TEN including acne, psoriasis and folliculitis (119, 125, 129-132). CD4+

memory T cells that express granulysin are candidate marker in Mycobacterium

tuberculosis infection (119, 133).

Studies have shown that the 15kDa form of granulysin, found within the epidermal

blister fluid of SJS/TEN patients, show a similar cytotoxicity to the 9kDa form secreted

by both NK cells and CD8+ T cells (93, 103, 117). Exposing a mouse model to the

15kDa form of granulysin induces considerable cutaneous blistering and necrosis in

the mouse. These findings show that higher levels of extracellular secretory 15kDa

granulysin can lead to a rapid development of epidermal necrosis. It has also been

shown that the lower granulysin concentrations found in SJS lesions is due to an

increase in mononuclear cell infiltration when compared to the granulysin

concentrations of TEN lesions (93, 103, 117). This observation would explain the

observations in SJS/TEN histopathology, showing that infiltration of sparse dermal

mononuclear cells results in extensive epidermal necrosis (93, 103, 117).

In 2014, a case study of one patient with lamotrigine-induced TEN, HLA types

unknown, showed an increase CD56+ NK cell granulysin expression, when the

patients PBMC’s were exposed to lamotrigine ex-vivo over a four-day period (134).

Another finding by Porebski et al. 2013 showed an increase in granulysin expression

by CD4+ memory T cells from patients with lamotrigine-induced SJS/TEN when

exposed to lamotrigine ex-vivo (135). Chung et al. 2015 produced evidence that

besides HLA-B*58:01 carriage, oxypurinol-induced granulysin expression in T cells

contribute to the immunopathogenesis in allopurinol-SJS/TEN (116)

Granulysin is an important biomarker in the diagnosis and progression of SJS/TEN. A

robust granulysin-specific assay would offer further insight into drug-SJS/TEN

immunopathogenesis, which could be translated into clinical practice helping to better

predict and reduce incidences of drug-SJS/TEN.

1.3.3 Nevirapine

The ARV drug NVP is a non-competitive reverse transcriptase inhibitor that has been

effective in HIV treatment. The use of NVP as a prophylactic has proven to be

successful in reducing mother to child vertical transmission rates of HIV-1 in lower

socio-economical populations (136, 137). NVP is also effective in salvage regimens

after virological failure, especially in those who have not taken non-nucleoside reverse

transcriptase inhibitors. Approved for use in combination with dual nucleoside reverse

20

transcriptase inhibitors, NVP is listed on the List of Essential Medications by the World

Health Organization (92).

1.3.3.1 NVP-SCAR and HLA allele associations

NVP has been associated with inducing several IM-ADR, including delayed rash, the

more severe cutaneous reactions DRESS and SJS/TEN. NVP-SCAR

immunopathogenesis is driven by both CD8+ T-cell / HLA class-I responses and CD4+

T-cell /HLA class-II responses. Studies have shown that the response to NVP, ex-vivo

can be abrogated by removing CD8+ or CD4+ T cells (138). The ex-vivo NVP-

stimulated responses diminish over time after first in-vivo reaction/exposure. The

primary focus of this work will be on the diversity and complexity of HLA alleles

associated with NVP-SCAR phenotypes while exploring this diminished ex-vivo

responses associated with NVP-DRESS.

Up to 5% of patients who start NVP will experience a DRESS reaction (138-141). An

increased risk of NVP-DRESS has been associated with the HLA-DRB1*01:01

coupled with CD4+ T cells greater than 25% (CD4+ T-cells counts > 250 cells/μL in

women and > 400 cells/μL in men) (Table 1.2) (74, 142). While not a SCAR, NVP and

efavirenz delayed rash was observed in French patients who carried the HLA class-II

allele, HLA-DRB1*01 (Table 1.2) (143, 144). A study involving a South African cohort,

found that carriage of the HLA class-II allele HLA-DRB1*01:02 and the HLA class-I

allele HLA-B*58:01 was associated NVP-hepatotoxicity (Table 1.2) (145).

HLA class-I allele associations to NVP-DRESS include HLA-Cw*8 in an Italian cohort

and the HLA-Cw*8 / B*14 haplotype among the Japanese (Table 1.2) (146, 147). The

HLA-C*04 allele was shown to increase the risk of NVP-DRESS in the Han Chinese.

The HLA-B*35:01 and HLA-B*35:05 alleles were a significant risk factor for NVP-

DRESS in Caucasian and Asian populations, respectively (Table 1.2) (138, 148-150).

Again, the NVP-delayed rash, while not a SCAR reaction has also been linked to the

carriage of, HLA-C*04 and HLA-B*35:05, in Asian populations (Table 1.2). (144, 151-

153) NVP-SJS/TEN accounts for less than 1.5% of NVP-HSR (67, 76, 154). In terms

of SJS/TEN, a recent study has shown that carriage of the HLA-C*04:01 allele

increased the risk of NVP-SJS/TEN in a Malawian cohort (Table 1.2) (155).

Table 1.2: Genetic (HLA alleles) and tissue specificity variation seen in different ethnic groups

associated with the NVP-HSRs. Table adapted from Rive C et al. 2013 and Phillips EJ et al. 2011(2,

150)

21

IM-ADR Ethnic association Genetic association (HLA)

DRESS

Rash associated hepatitis CD4+ > 25% in Caucasians

DRB1*01:01 and DRB1*01:02

Italian (Sardinian) Cw8-B*14 haplotype

Japanese Cw8

Han Chinese Cw*4 and DRB1*15

Asian B*35:05

Australian (Caucasian)

B*35:01, B*15-DRB1*15

Delayed rash

French population DRB1*01

African, Asian, European and Thai Cw*04

Thai B*35:05, rs1756*G CCHCR1 (GWAS)

DILI South African B*58:01

SJS/TEN African HLA-C*04:01(needs confirmation)

IM-ADR; Immunologically mediated-adverse drug reaction, DRESS, drug reaction with eosinophilia and systemic symptoms; DILI; drug-induced liver injury, SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis, GWAS; genome-wide association study.

1.4 Established Models of HLA-driven delayed

Immunologically Mediated-Adverse Drug

Reactions

As highlighted earlier, specific HLA alleles, like HLA-B*15:02 in

CBZ-SJS/TEN in Southeast Asians and HLA-B*57:01 in ABC-HSS have a negative

predictive value of 100%. How these specific HLA alleles and drugs interact to elicit

T cell-mediated immune responses, leading to the associated IM-ADR phenotypes

described above is still an area of debate. Three hypotheses or models have been

proposed to explain HLA-driven delayed IM-ADR and possible that no single

hypothesised model is mutually exclusive. These three HLA-drug-binding models

proposed include the hapten/prohapten model, the pharmacological interaction (P-I)

model, and the altered peptide repertoire model (67, 74, 76, 156).

22

1.4.1 The Hapten / Prohapten Model

The hapten/prohapten model involves drug (or a metabolite) modifying host proteins to

creating a neo-antigen (Figure 1.9). The hapten model features chemically reactive

drugs covalently binding to high molecular weight proteins, forming a hapten-protein

complex (67, 157, 158). This modification can affect any autologous protein, including

extracellular, intracellular, and membrane-bound proteins. The hapten-protein

complexes are then processed and presented by the HLA molecule on the surface of

the cell forming a stable HLA-hapten-peptide complex (67, 157, 158). The location,

nature, and amount of the proteins modified influence the IM-ADR, leading to several

immune responses. The prohapten extension to the hapten model was used to explain

how chemically inert drugs can still elicit a delayed IM-ADR. The prohapten model

involves the metabolism of the inert parent drug into a reactive metabolite which then

becomes antigenic through the process mentioned.

1.4.2 The Pharmacological Interaction Model

The Pharmacological Interaction (P-I) model proposes that a non-covalent interaction

between the culprit drug and TCR / HLA molecules can activate certain T cells (Figure

1.9) (67, 74, 76, 156). This reversible reaction is comparable to the interaction of a

ligand with its receptor. Evidence of this P-I model exists in observations of aldehyde-

fixed APCs which cannot process or present any peptides but still have the ability to

activate specific T-cell clones (159-161).

1.4.3 The Altered Peptide Repertoire Model

The altered peptide repertoire model can be considered a refinement of the P-I model.

In-silico modelling has shown how certain drugs can to fit into an “empty” HLA

molecule, through a non-covalent bond to key residues and the occupation of anchor

sights within the HLA peptide-binding groove (2, 83-85, 162). This non-covalently

binding and occupation of anchor sites within an “empty” HLA protein, alters the

repertoire of peptides capable of binding within the drug-modified HLA (2, 83-85, 162).

Memory CD8+ T-cell responses derived from HLA-B*57:01-restricted EBV epitopes,

can also recognise endogenous ABC-HLA-B*57:01-restricted endotopes (83).

Evidence supporting this model comes from three independent studies that put

23

forward that the altered peptide repertoire model is the underlying mechanism in

ABC-HSS (83-85).

1.4.4 The Heterologous Immunity Model

The interactions of drugs with specific HLA alleles can be explained by neo-antigen

generation through the hapten/prohapten, the P-I and/or the altered peptide repertoire

models (Figure 1.9). These off-target interactions with specific HLA molecules or HLA-

presented peptides are not enough to explain several factors identified with various

delayed IM-ADR. For example, these models fail to explain how reactions can

sometimes manifest within days after the first drug exposure or the long-lasting

memory T-cell responses in drug-induced SJS/TEN and ABC-HSS reactions, for

example. The models fail to explain why different drug-HLA allele combinations can

produce different IM-ADR phenotypes, for example, why ABC-HLA-B57:01-HSS

differs from CBZ-HLA-B*15:02-SJS/TEN which differs again from the varied IM-ADR

phenotypes associated with NVP.(68)

The hapten/prohapten, the P-I and/or the altered peptide repertoire models fail to

account for the complete negative predictive value and low or incomplete positive

predictive values associated with carrying an at-risk HLA allele. This suggests that

while carrying an at-risk HLA allele is necessary, it alone does not explain the

immunopathogenesis. The heterologous immune model attempts to explain the

shortcomings.

Similar to organ transplantation, IM-ADR presents a novel situation in which neo-

antigens are created through one of the three HLA-drug-binding models, as described

earlier. Organ rejection is mediated by pre-existing HLA class-I–restricted effector

memory T-cell responses to prevalent viral infections (24, 54, 56, 59, 163-166). In this

setting memory T cells that interact with persistent viruses, for example, HHV

epitopes, may cross-react with neo-antigens are created through one of the three

HLA-drug-binding models (Figure 1.9) (68). Therefore, a pre-existing heterologous

immune response could decide whether an individual with a high-risk HLA allele

develops an IM-ADR. Heterologous immunity could also explain why IM-ADR can

manifest within days after the first drug exposure, and the variation in tissue specificity

and the long-lasting memory T-cell responses in drug-induced SJS/TEN and ABC-

HSS reactions, for example (68).

24

Figure 1.9: Graphical representation published by White et al., 2015 of the heterologous immunity

model with the three neo-antigen generating, HLA-drug binding models (A) hapten/prohapten model,

the (B) P-I model, and the (C) altered peptide repertoire model.(167)

1.5 Scope of this Thesis

A goal of IM-ADR research includes the ability to identify patients or specific

population(s) who is at risk of experiencing a drug-specific IM-ADR, before

administering the drug to prevent the resultant morbidity or mortality that could ensue.

To achieve this, an understanding of the immunopathogenesis underlying these

reactions is required. Both CBZ-SJS/TEN and NVP-SCAR are distinct clinical and

immunological syndromes, but they share are several significant similarities which

suggest that the underlying immunopathogenesis driving SCARs may share

commonalities. Both CBZ-SJS/TEN and NVP-SCAR offer important examples for the

immunopathogenesis of severe T cell-mediated drug reactions.

Chapter 1 has summarised ADR and delayed IM-ADR, a smaller subset of ADR,

driven by T cell-mediated immune responses. This chapter further explained the

reactions and interactions of T cell-mediated immunity before moving into what is now

known about immunopathogenesis of delayed IM-ADR. This chapter highlighted the

ABC-HSS example, before leading into the two drugs of interest, CBZ and NVP.

25

Following this, Chapter 2 will focus on the Materials and Methods utilised to discuss

the limitations to the established models of HLA-driven delayed drug hypersensitivity

and answer the hypotheses proposed in the following Chapters.

1.5.1 NVP-Induced HSR

NVP-HSR has been associated with a variety of HLA alleles. To understand the

immunopathogenesis driving NVP-HSRs, the question of varied tissue specificity and

HLA alleles are associated with NVP-HSR, needs answering. Chapter 3 will look at

examining two aspects of NVP-SCAR immunopathogenesis. HLA class-I alleles and

their binding cleft characteristics associated with NVP-SCAR and the diminishing

Gamma Interferon (IFN-γ) responses in NVP-stimulated, PBMC’s from NVP-DRESS

patients.

1.5.1.1 NVP-HSR and the shared HLA binding cleft specificities

The tissue specificity and varied HLA allele associations related to NVP-HSR might

be explained by the similarities in the HLA peptide-binding cleft structure and predicted

peptide/NVP-binding capabilities. Chapter 3 will analyse HLA class-I alleles identified

as statistically significant in NVP-SCAR and NVP-SCAR with hepatitis, within a cohort

of ethnically diverse, defined NVP-HSR cases and matched controls. HLA alleles will

be analysed to see if they share similarities in peptide-binding cleft structures by

grouping HLA alleles according to; shared key amino acid residues which make up the

HLA class-I binding cleft pockets (A-F), validated cross-reactive group (CREG) HLA

supertypes which groups HLA alleles based on similarities in cross-reactivity of public

epitopes among HLA class-I alleles and binding cleft pocket chemistry, and the

MHCcluster algorithm, which seeks to group HLA molecules based on their peptide-

binding (pb-) specificity as predicted by NetMHCpan 2.8 (168-172). Confirmation of the

HLA shared binding cleft specificities would explain many outstanding observations

regarding T cell-mediated delayed NVP-HSR including the low or incomplete

negative/positive predictive values for the various HLA alleles already associated. This

approach could apply to other delayed IM-ADR which exhibits an association to a

varied number of HLA alleles.

26

1.5.1.2 Diminishing ex-vivo PBMC’s responses and NVP-HSR

Immunopathogenesis

Unlike several well characterised delayed IM-ADR, ex-vivo NVP-stimulated, PBMC’s

IFN-γ responses diminish over time after first in-vivo reaction/exposure. Keane et al.

2014 explored the hypothesis that in cases of the NVP-SCAR DRESS phenotype, the

diminishing NVP-stimulated, PBMC IFN-γ responses was due to an increase in CD4+

Tregs levels, during the recovery period of NVP- DRESS. This increase in CD4+ Tregs

levels has been reported in other drug-induced IM-ADR (173). In the paper by Keane

et al. 2014, one patient exhibited an increase in Tregs that corresponded with the

waning IFN-γ, Enzyme-Linked ImmunoSpot (ELISpot) responses (138). In this section

of the chapter, we will expand on this hypothesis. However, as most of the NVP-

DRESS patient PBMC’s samples had been used in the initial study by Keane et al.

2014, cytokine plasma profiles of the same NVP-DRESS patients, used in the Keane

et al. 2014, were analysed. It is expected that the cytokine environment within the

donor’s plasma, over the time course of a NVP-DRESS reaction and recovery period,

alongside the Keane et al. 2014 data, might show changes in specific T-cell

subpopulations further explaining the diminishing NVP-stimulated, PBMC IFN-γ

responses

1.5.2 Carbamazepine-Induced SJS/TEN

Toxic Epidermal Necrolysis unlike NVP-SCAR and their varied HLA allele

associations, a 100% negative predictive value has been linked with HLA-B*15:02

carriage among those of Han Chinese and South-east Asian ancestry and CBZ-

SJS/TEN (174). Chapter 4, Chapter 5, and Chapter 6 will look at CBZ-SJS/TEN

immunopathogenesis. Chapter 4 and 5 will explore the hypothesis which states that

HLA-B*15:02-restricted CBZ-SJS/TEN is driven by memory T cells which expresses

the specific cross-reactive TCR, in a heterologous immune response, explaining the

tissue distribution of the clinical symptoms and the low or incomplete positive

predictive value associated with HLA-B*15:02-restricted CBZ-SJS/TEN.

1.5.2.1 CBZ-SJS/TEN and Herpes simplex virus

Chapter 4 involves assessing several in-silico derived HSV-1 and/or HSV-2, HLA-

B*15:02-restricted peptides, ex-vivo (186-190). Identifying HLA-B*15:02-restricted

peptides, shared and restricted to CBZ-SJS/TEN patients, is an essential step in

assessing the heterologous immune model as it relates to HLA-B*15:02-restricted

CBZ- SJS/TEN.

27

If a HHV-specific memory T cell is cross-recognising a neo-antigen created by one of

the three proposed HLA-drug-binding models, then it would stand to reason that the

immune response to the persistent pathogen should share similarities to the clinical

presentation of CBZ-SJS/TEN. Reactivation of HSV-1 and HSV-2 involve skin and

mucosal tissues. The skin and mucous membrane phenotype associated with HLA-

restricted CBZ-SJS/TEN might result from cross-reactive memory T cells primed for a

specific HSV epitope(s) and activated by the neo-antigen created by one of the HLA-

drug-binding models; the subsequent drug-related immune response becomes widely

distributed, resulting in expansive tissue involvement as seen in SJS/TEN. Besides the

persistent pathogen immune response sharing similarities to the clinical presentation

of SJS/TEN, the seroprevalence of the persistent pathogen within HLA-B*15:02

positive CBZ-SJS/TEN cases should differ from HLA-B*15:02 positive CBZ-tolerant

and naive populations.

1.5.2.2 CBZ-SJS/TEN and T-cell receptor clonotype

The highly variable TCR repertoire is defined through TCR gene rearrangement and

selection in the thymus also governed by HLA molecules, self-peptide repertoires, and

earlier exposure to pathogenic derived epitopes. Potentially dangerous TCR repertoire

may be selected in those with at-risk HLA alleles (175).

Ko et al. 2011 examined the CD8+ TCR repertoire in patients with CBZ-SJS/TEN

(176). The authors showed a T-cell repertoire containing the public sequences Vβ 25-

1*01 (Vβ 11s1) CDR3 CASSISGSY and CASSGLADVDN within the PBMC’s and

blister fluid of CBZ-SJS and CBZ-SJS-TEN overlap patients. Cellular based assays

have shown anti-Vβ-11 antibodies can block CBZ-associated cytotoxicity and both a T-

cell clonotype and transfectants, expressing a TCR with the Vβ 25-1*01 CDR3

CASSISGSY sequence and Vβ 25-1*01 CDR3 CASSISGSY, displayed cytotoxicity

against HLA-B*15:02 expressing APCs in the presence of CBZ (139). The authors

concluded that besides HLA-B*15:02 carriage, Vβ 25-1*01 CDR3 CASSISGSY and

CASSGLADVDN TCR expression could be another factor in HLA-B*15:02-specific

CBZ-SJS/TEN development (176).

Chapter 5 will involve developing methods and assays to assess before identified

specific Vβ CDR3 TCR sequence expression (161). The primary aim is to expand

CBZ-specific T cells and identify the expression of specific vβ CDR3 sequences using

the new and sensitive droplet digital (dd) polymerase chain reaction assay (PCR) (Bio-

Rad QX100/200 ddPCR system) (176).

28

1.5.2.3 Granulysin expression

Chapter 6 will look granulysin expression, another established factor in CBZ-SJS/TEN

immunopathogenesis and analysis granulysin expression in PBMC’s from patient

donors that had experienced a SJS/TEN reaction to CBZ. Cell-based assays have

shown that CBZ-SJS/TEN patient PBMC’s responses can be detected in PBMC’s

acquired years after CBZ withdrawal, further supporting the responses seen in-vivo

when a patient is rechallenged with CBZ. The overall aim is to characterise the cell

type that is expressing granulysin when re-exposed to pharmacological relevant doses

of CBZ ex-vivo to better understand CBZ-SJS/TEN immunopathogenesis.

1.5.3 Final Summary and Conclusion

Chapter 7 will summarise the findings within this body of work, highlighting the main

findings and offering recommendations and considerations for future research into the

underlying mechanisms governing both CBZ-SJS/TEN and NVP-HSR

immunopathogenesis. Chapter 8 includes the bibliography highlighting the associated

sources and references and the appendix is attached as Chapter 9.

29

Chapter 2

Methods and Materials

30

31

2 2.1 Samples, Ethics and Limitations

2.1.1 Nevirapine-Induced HSR

2.1.1.1 NVP-SCAR associated HLA class-I binding cleft

characteristics

Samples

Consenting participants from 11 countries (Argentina, Australia, Canada, France,

Germany, Netherlands, Spain, Taiwan, Thailand, United Kingdom and The United

States of America) were recruited for this study in collaboration with colleagues at

Vanderbilt University, School of Medicine and Boehringer Ingelheim. (ClinicalTrials.gov

NCT00310843). Cases were selected based on donors experiencing cutaneous and/or

hepatic symptoms 8 weeks after beginning NVP. Symptoms include:

1. NVP-induced severe cutaneous toxicity, characterised as a grade III or IV rash

(National Institute of Allergy and Infectious Disease, Division of acquired

immunodeficiency syndrome criteria) (149).

2. Acute hepatic failure or symptomatic grade 3 or higher hepatic transaminase

elevation (including aspartate transaminase and alanine transaminase) 5 times

or more upper limit of normal (149).

Cases and controls were excluded if donors had experience:

3. A CD4+ T-cell count of 150 cells/μL or lower 6 months prior to beginning NVP.

An acute viral hepatitis and/or no record of hepatic transaminase data 6 months

leading up to NVP initiation.

4. Received immunomodulatory drugs within the first 8 weeks of starting NVP

and/or was part of the two nucleoside analogue reverse transcriptase inhibitor

long-term study; Participants were given stavudine and lamivudine along with

either NVP and/or Efavirenz depending on test parameters (149, 177).

32

Ethical considerations and approvals

Project Title: The immunopathogenetic basis of nevirapine

hypersensitivity

Project No.: 2012/163 Murdoch University Human Research

Ethics Committee

Dates: 28 August, 2012 renewed until August 2015

Listed as: Student researcher

Limitations

Limitations included the incomplete nature of the initial demographic, phenotypic, and

sample information. Due to the study design, it was not possible to clarify any

ambiguous or unclear information. Prior to any analysis, the demographic, phenotypic,

and sample information for each donor was examined, organised, and confirmed

before being included. Several samples were excluded because; no sample was

available for HLA typing or issues with sample identity were identified (n=101) and

either no clinical data was reported (n=22) or ambiguous clinical data was reported

(n=62). The final analysis included 210 NVP-HSR cases (152 cutaneous adverse

events, 58 hepatic adverse events) with 468 controls.

2.1.1.2 Diminishing ex-vivo PBMC’s responses and NVP-HSR

Immunopathogenesis

Samples

Donor 1 (NVP-HSR), Donor 2 (NVP-HSR), and Donor 3 (NVP-HSR) were kindly

provided by The Institute for Immunology and Infectious Diseases, Murdoch

University, Perth, Australia. Patient 2, Patient 3 and Patient 4 used in HLA Class-I

Restricted CD8+ and Class-II Restricted CD4+ T Cells Implicated in the Pathogenesis

of Nevirapine Hypersensitivity, by Keane et al. 2014 (Appendix 9, Section 9.2

Publications Highlights), were renamed Donor 1 (NVP-HSR), Donor 2 (NVP-HSR),

and Donor 3 (NVP-HSR), respectively for this body of work (138).

Donor 1 (NVP-HSR), a Caucasian male (HLA-A*29:02, -A*31, -B*14:01, -B*44:03,

-C*08:02, -C*16:01, -DR*07:01:01) carried the HLA-B*14 / HLA-C*08 haplotype and

presented with rash, fever and eosinophilia two weeks after beginning a highly active

antiretroviral therapy (HAART) regime that included NVP, ABC, and lamivudine. Donor

1 (NVP-HSR) was positive for IgG to Varicella zoster virus (VZV), EBV, CMV, and

HHV-6 but showed no evidence of viral reactivation (no IgM).

Donor 2 (NVP-HSR), a 5-month-old infant with Asian ancestry (HLA-A*02:06,

-A*34:01, -B*1521, -B*5601, -C*04:03, -C*07:01, -DR*04:06, -DR*15:02), carried the

33

HLA-B*56:01 and HLA-C*04:03 alleles and experienced fever, rash, eosinophilia and

hepatitis on day 7 of a HAART regime containing NVP, ABC, and lamivudine. Donor 2

(NVP-HSR) at the time of NVP-HSR was only 2 months of age and was positive for

IgG to EBV, CMV and HHV-6 however, at 14 months of age only remained positive for

IgG to HHV-6, consistent with waning of passive immunity and lack of infection with

EBV and CMV. Donor 2 (NVP-HSR) showed tested negative for IgM and showed no

signs of an activated virus.

Donor 3 (NVP-HSR), Female of Asian ancestry (HLA-A*11, -A*24, -B*13:01,

-B*35, -C*03, -C*04, -DR*11, -DR*16), carried the HLA-B*35 and HLA-C*04 allele, and

presented with fever, rash, altered liver function and eosinophilia 14-15 days after

starting NVP and was immediately taken off NVP. Donor 3 (NVP-HSR) was positive

for IgG to EBV, CMV and HHV-6 but showed no evidence of viral reactivation (no

IgM). Donor 3 (NVP-HSR) also experienced a trimethoprim-sulfmethoxazole related

HSR, 2-3 months prior to the NVP HSR and showed evidence of elevated and

activated CD4+ and CD8+ T cells in a sample tested 1 month post NVP HSR compared

to a sample tested 4 months later.

Controls included several NVP-tolerant and NVP-naive donors.


Project Title: Pharmacogenomics and mechanistic basis of

drug hypersensitivity

Project No.: 2014/20 Murdoch University Human Research

Ethics Committee

Dates: 22 January, 2014 until December 2016


Limitations

Limitations included sample size, an issue in much of research involving delayed IM-

ADR. Another limitation includes obtaining enough patient PBMC’s/plasma samples

from various time points pre, during and post NVP-HSR. The initial study by Keane et

al. 2014, required the majority of the NVP-DRESS patient PBMC’s, therefore we

decided to assess the cytokine plasma profiles of the same NVP-DRESS patients,

used in the Keane et al. 2014, to see if they show changes in specific T-cell

subpopulations further explaining the diminishing NVP-stimulated, PBMC IFN-γ

responses.

34

Table 2.1: List of samples used in Diminishing ex-vivo PBMC’s responses and NVP-HSR

Immunopathogenesis

Samples Drug HSR phenotype HHV serology HLA

Donor 1 (NVP-HSR)

NVP DRESS

VZV IgG+ EBV IgG+ CMV IgG+ HHV-6 IgG+

A*29:02, -A*31, -B*14:01, -B*44:03, -C*08:02, -C*16:01, -DR*07:01:01

Donor 2 (NVP-HSR)

NVP DRESS

At two months old EBV IgG+ CMV IgG+ HHV-6 IgG+ After 14 months of age HHV-6 IgG+ only

A*02:06, -A*34:01, -B*1521, -B*5601, -C*04:03, -C*07:01, -DR*04:06, -DR*15:02

Donor 3 (NVP-HSR)

NVP DRESS

EBV IgG+ CMV IgG+ HHV-6 IgG+

A*11, -A*24, -B*13:01, -B*35, -C*03, -C*04, -DR*11, -DR*16

Naive NVP controls

NVP-naive

- - -

NVP tolerant Controls NVP - - -

HSR, hypersensitivity reaction; NVP, nevirapine; DRESS, Drug rash with eosinophilia and systemic

symptoms; HHV, Human herpesvirus; VZV, Varicella zoster virus; EBV, Epstein-Barr virus; CMV, ; HLA,

Human leukocyte antigen.

2.1.2 Carbamazepine - Induced SJS/TEN

Carbamazepine SJS/TEN samples

Patients, who had experienced a CBZ-SJS/TEN reaction, gave informed consent to

use their blood and products within in this study. The Controls and Donors are listed in

Table 2.2, briefly: Donor 1 (CBZ-TEN), a female with South-east Asian ancestry (HLA-

A*11:01, -B*15:02, -B*58:01, -C*08:01, -C*03:01, -DRB1*12:02, -DRB1*03:01), carried

the HLA-B*15:02 allele and experienced a TEN reaction to CBZ.

Donor 1 (CBZ-TEN), diagnosed with HIV-1 infection, had also experienced an acute

HZ with post-herpetic neuralgia, which was managed with CBZ along with other

regular pain medications and NSAIDs. Within 9-14 days of CBZ initiation, Donor 1

(CBZ-TEN) developed symptoms including fever, elevated transaminase levels, and

cutaneous symptoms consistent with TEN. Donor 1 (CBZ-TEN) tested positive for IgG

to HSV-2 and negative to HSV-1 while produced a positive CBZ cutaneous patch test

35

response 9.5 years after the original CBZ-TEN reaction, and produces positive CBZ-

stimulated, ex-vivo responses 17 years after the original CBZ-TEN reaction.

Donor 2 (CBZ-SJS), a female of South-east Asian ancestry (HLA-A*11:01 -B*15:02, -

C*08:01, -DRB1*12:02 common haplotype with 11.6-19.7% frequency in Han Chinese

populations), carried the HLA-B*15:02 allele, tested positive for IgG to HSV-1 and

negative to HSV-2 and experienced a SJS reaction to CBZ (144).

Donor 3 (CBZ-SJS), a female of Caucasian ancestry (HLA-A*26:12, -A*02:05,

-B*15:02, -B*41:01 -C*08:01, -C*07:01, -DRB1*12:02, -DRB1*01:01), carried the HLA-

B*15:02 allele, tested positive for IgG to HSV-1 and HSV-2, and experienced a SJS

reaction to CBZ.

Donor 4 (CBZ-SJS non-B*15:02), a female of Caucasian ancestry (HLA-A*01:01

-A*32:01/24, -B*35:03, -B*51:01 -C*01:02, -C*04:01, -DRB1*01:01, -DRB1*13:02),

tested positive for IgG to HSV-1 and was retroactively diagnosed with an earlier SJS

reaction to CBZ. Donor 4 (CBZ-SJS non-B*15:02) did not carry the HLA-B*15:02

allele, but carried several delayed IM-ADR alleles including the NVP-SCAR associated

HLA-C*04:01 allele and other IM-ADR associated alleles.

Whole blood samples from healthy donors were obtained through the Australian Bone

Marrow Donor Registry. The donors had given informed consent. CBZ-naive donors

included;

Control 1 (CBZ-naive) carried the HLA-B*15:02 allele (HLA-A*02:03:01, -A*24:07,

-B*15:02, -B*56:01, -DRB1*04, -DRB1*09), was serologically positive for HSV-1 and

negative for HSV-2, and was CBZ-naive.

Control 2 (CBZ-naive non-B*15:02) did not carry the HLA-B*15:02 allele (HLA-A*02,

-B*15:11, -B*38, -DRB1*11, -DRB1*16:02:01), was serologically positive for HSV-1

and negative for HSV-2, and was CBZ-naive.

Control 3 (CBZ-naive) carried the HLA-B*15:02 allele (HLA-A*11, -A*33:03:01,

-B*15:02, -B*58, -DRB1*03, -DRB1*15) and was CBZ-naive. Control 3 (CBZ-naive)

was also serologically negative for HSV-1 and HSV-2.

Control 4 (CBZ-naive non-B*15:02) did not carry the HLA-B*15:02 allele (HLA-A*02 -

A*322012, -B*35:01, -B*51, -C*04:01, -DRB1*01:01, -DRB1*03:02) and was CBZ-

naive. The HSV serological status for Control 4 (CBZ-naive non-B*15:02), was

unknown.

36


Project

Title:

Pharmacogenomics and mechanistic basis of drug

hypersensitivity

Project No.: 2014/20 Murdoch University Human Research Ethics

Committee

Dates: 22 January, 2014 until December 2016


Limitations

The limitations associated with sample size listed previously (Materials and Methods,

Section 2.1.1.2) also apply to CBZ-associated studies. A unique limitation with CBZ-

HSR samples includes the inability to get follow-up samples. Once a patient has

recovered from an SJS/TEN reaction, they are not required to return for ongoing

medical attention and as long as they avoid the culprit drug. This makes retaining

multiple Donor samples problematic. CBZ-tolerant, HLA-B*15:02 positive patients

have also been hard to obtain.

Table 2.2: List of samples used in CBZ-HSR immunopathogenesis associated studies.

Samples Drug HSR

phenotype HHV serology Other

Donor 1 (CBZ-TEN) CBZ TEN HSV-2 IgG +

Acute HZ and PHN

Patch test +

HLA-B*15:02 +

Donor 2 (CBZ-SJS) CBZ SJS HSV-1 IgG + HLA-B*15:02 +

Donor 3 (CBZ-SJS) CBZ SJS HSV-1 IgG + HSV-2 IgG +

HLA-B*15:02+

Donor 4 (CBZ-SJS

non-B*15:02) CBZ SJS HSV-1 IgG + Non-HLA-B*15:02

Control 1 (CBZ-naive)

CBZ-naive - HSV-1 IgG + HLA-B*15:02 +

Control 2 (CBZ-

naive non-B*15:02) CBZ-naive - HSV-1 IgG + Non-HLA-B*15:02

Control 3 (CBZ-naive)

CBZ-naive - Negative HLA-B*15:02+

Control 4 (CBZ-

naive non-B*15:02) CBZ-naive - Unknown Non-HLA-B*15:02

CBZ, carbamazepine; SJS, Stevens-Jonson syndrome; TEN, toxic epidermal necrolysis; HSR, hypersensitivity reaction; HHV, human herpesvirus; HZ, Herpes Zoster, PHN, post-herpetic neuralgia; HSV-1, Herpes simplex virus-1; HSV-2, Herpes simplex virus-2;

37

2.2 Cellular Assays

Table 2.3 List of reagents used in Stimulation and Expansion, flow cytometry, and Lymphoblastic cell

lines (LCL) generation assays. Flow cytometry antibodies are included on Table 2.4, tetramer reagents

included in Tables 2.5-2.7 and ELISpot associated reagents and materials are included in Table 2.8

Reagent Manufacture Catalogue Number

RPMI-1640 with phenol red Gibco 21-870-076

Foetal calf serum Invitrogen

Serana

C8056

S-FBS-AU-015

DNase Promega RQ1

L-glutamine Gibco 25030-081

Penicillin / Streptomycin Gibco 15-140-148

Sodium pyruvate Gibco 25030-081

Invitrogen Countess Invitrogen C10227

RNAlater stabilisation solution Life technologies AM2070

Recombinant human (rh) IL-2 R&D systems

Recombinant human (rh) IL-7 R&D systems 207-IL

Recombinant human (rh) IL-15 R&D systems 247-IL

Carbamazepine (CBZ) Sigma C4024

NVP Sigma

Boehringer Ingelheim

SML0097

Extracted from Viramune

Phosphate-buffered saline

(1 x PBS) Gibco 70013032

6-well flat-bottom culture plates BD Falcon 353046

12-well flat-bottom culture plates BD Falcon 353043

96-well U-bottom plate BD Falcon 353077

Brefelden A Sigma B7651

IsoFLOW Sheath fluid Beckman Coulter 8546859

IntraPrep Reagent kit Beckman Coulter A07802

Safe-Lock 1.5mL Eppendorf Tubes Crown Scientific 0030.120.086

Cryotubes Nunc Bank-It 374078

PE-streptavidin BD Biosciences 554061

BD Cytokine Bead Array (CBA) BD Scientific 560484

T-25 and T-75 culture flasks BD Falcon 353108 and 353136

0.45μm sterile filter with plunger Sigma CLS431225

Mycoplasma plus PCR primer kit Agilent 302008

Lectin Phytohemagglutinin (PHA-P) Sigma L8754

38

2.2.1 Drug Stimulation and Cellular Expansion

Day 0 - Cryopreserved PBMC’s were thawed and washed with RPMI-1640 media

(Gibco; Cat. No. 21-870-076) medium containing 10% heat-inactivated foetal calf

serum (FCS) (Invitrogen and Serana; Cat. No. C8056 and S-FBS-AU-015,

respectively) and 10U/mL DNase (Promega; Cat. No. RQ1) being added between the

first and second washes, as highlighted by Garcia-Piñeres et al. 2006) (178, 179).

PBMC’s were then resuspended in R10 media, which contained RPMI-1640 media,

10% heat-inactivated FCS, 1μM L-glutamine (Gibco; Cat. No. 25030-081), 50 U/mL of

penicillin, 50 μg/mL of streptomycin (1% v/v Gibco; Cat. No. 15-140-148), and 1μM

sodium pyruvate (Gibco; Cat. No. 25030-081) and incubated at 37°C overnight (Figure

2.1) (83).

Day 1 – recovery and viability of PBMC’s was determined by staining a 10μL aliquot of

PBMC’s with 0.4% Trypan Blue solution and counting cells via the Invitrogen Countess

method (Invitrogen; Cat. No. C10227). An appropriate aliquot of cells were removed

for pre-stimulation IFN-γ ELISpot, intracellular cytokine staining (ICS) and

immunophenotyping flow cytometry assays or stored in RNAlater stabilisation solution

(Life technologies; Cat. No. AM2070) until the Ribonucleic acid (RNA) extraction could

be performed.

10 million cells per mL (1 x 107 cells/mL) were resuspended in R10 media (Section

2.2.1, Drug Stimulation and Cellular Expansion) containing 50 μg/mL of recombinant

human (rh) IL-7 (R&D systems; Cat. No. 207-IL) (Figure 2.1). The cells were split into

three aliquots for control and test assays. R10 media (Section 2.2.1, Drug Stimulation

and Cellular Expansion), supplemented with 50 μg/mL of rhIL-7 containing and either

the control drug stimulant, or test stimulant (CBZ [Sigma; Cat. No. C4024]) was added

to the aliquots, giving a final stimulant concentration of 10 μg/mL, (unless otherwise

stated). For the unstimulated control, additional R10 media (Section 2.2.1, Drug

Stimulation and Cellular Expansion), supplemented with 50 μg/mL of rhIL-7 containing

was added to give the same final volume across all three aliquots (Figure 2.1). Cells

were than incubated at 37°C for 1-2 hours.

After stimulation, the cells were washed with a volume of cold RPMI-1640 media and

resuspended in R10 media (Section 2.2.1, Drug Stimulation and Cellular Expansion)

supplemented with 50 μg/mL of rhIL-7, giving a final concentration of 5 million cells per

mL (5 x 106 cells/mL). Cells were seeded into labelled wells of a 6- or 12-well flat-

39

bottom culture plates (BD Falcon; Cat. No. 353046 and 353043, respectively) and

incubated at 37°C for up to 48 hours.

Day 3 - 50% of the culture supernatant was removed and replaced with new R10

media (Section 2.2.1, Drug Stimulation and Cellular Expansion) supplemented with 50

μg/mL of rhIL-7 and with 50 μg/mL rhIL-15 (R&D systems; Cat. No. 247-IL). The

cultures were incubated at 37°C for another 48 hours.

Day 5 - Cultures were split, and cultures made up to the first volume by new R10

media (Section 2.2.1, Drug Stimulation and Cellular Expansion) containing rhIL-7 and

rhIL-15 as previously stated (Figure 2.1). The cultures were incubated at 37°C for

another 5 days.

Day 10 - cells were transferred to labelled 15 mL falcon tubes, washed and

resuspended in R10 media (Section 2.2.1, Drug Stimulation and Cellular Expansion).

The cultures were incubated at 37°C for another 4 days so they could stabilise.

Day 14 - cell viability and counts were performed and cells washed in RPMI-1640

media, centrifuged, and resuspended in the appropriate media for IFN-γ ELISpot, ICS

and immunophenotyping flow cytometry assays or stored in RNAlater stabilisation

solution until the RNA extraction could be performed (Figure 2.1).

40

Drug stimulation and cellular expansion protocol

Figure 2.1: Timeline of the stimulation and cellular expansion culture protocol for analysis of CBZ-

specific polyclonal T cells.

2.2.2 Flow Cytometry

The mutlicolour flow assays and reagent panel combinations were carefully selected

using spectralviewer anaylsis provided online by BD Biosciences (180). Instrament set

up controls inculded BD CompBeads (BD Biosciences, Cat. No. 644204). The

fluorescence minus one, single colour stained controls and unstained control methods

were also employed for compensation and adjusting photomultiplier tube voltages. All

flow cytometry experiments were analysed with specific biological controls like

unstimulated samples.

Removed cells from Liquid N2. Cells were thawed, washed, and resuspend in F10 media. Cells were rested overnight in the cellular incubator (37˚C with 5% CO2)

Cells were counted, washed, and aliquoted appropriately for -ICS, -ELISpot, -Storage in RNAlater stabilising solution for later RNA extraction and TCR analysis Remaining cells were aliquoted appropriately, washed and resuspended in rhIL-7-F10 media with the 10μg/mL of CBZ, control drug, or left unstimulated. After 1-2 hours of test/control stimulation, cells were washed, resuspended in rhIL-7-F10 media and incubated for 48 hours.

50% of the culture media was removed and replaced with rhIL-7-F10 media further supplemented with rhIL-15.

Cultures were split into 2 and the culture media replenished with rhIL-7-F10 media further supplemented with rhIL-15.

1 aliquot was stored in RNAlater stabilising solution. Remaining cells were washed, resuspended in F10 media, and rested in the cellular incubator for 48 hours. On day 12 this was repeated.

Cells were washed, counted, and resuspended in the appropriate media for ELISpot and ICS analysis.

41

Table 2.4: Flow cytometry Fluorochrome conjugated antibodies.

Immunogen Fluorochrome Exmax /Emmax

(nm)

Laser (nm)

Filter (nm) Cat. No

Anti – Human CD3

V450 405/448 405 450 BP *50 #560373

Anti – Human CD4

PE 496/578

488

575 BP *30 #340419

Anti – Human CD4

FITC 494/520

488

525 BP #340422

Anti – Human CD8

APC – H7 640/785

638

775 LP #641400

Anti – Human CD19

PE 496/578

488

575 BP *30 #555413

Anti – Human CD56

APC 650/660

638

660 BP *20 #555518

Anti – Human TCR α/β

APC 650/660

638

660 BP *20 +306718

Anti – Human IFN-γ

Alexa – 488 495/519

488

525 BP #557718

Anti – Human Granulysin (15- and 9 kDa)

Alexa – 488 495/519

488

525 BP #558254

BD Biosciences

CompBeads Multiple fluorochrome - Ig

κ anti-human Ab - - - #644204

*The number after the filter represents band width in nanometres (nm). #BD Bioscience and +Bio Legend Cat. No., catalogue numbers. BP, band pass filter; LP, long pass filter; EXmax, excitation maximum; Emmax, emission maximum.

2.2.2.1 Intracellular cytokine staining

Cryopreserved PBMC’s were thawed, washed, and resuspended in R10 media

(Section 2.2.1, Drug Stimulation and Cellular Expansion). Cells were incubated

overnight and viability and cell number determined. Cells were resuspended in R10

media (Section 2.2.1, Drug Stimulation and Cellular Expansion) giving a final

concentration of 0.5-1.0 x 107 cells/mL. Unlike the ELISpot and the Drug Stimulation

and Cellular Expansion protocols, R10 media (Section 2.2.1, Drug Stimulation and

Cellular Expansion) was not supplemented with cytokines as rhIL-2 and rhIL-15 have

been associated with the activation of granulysin expression within CD8+ T cells (181).

100 μL of cells were seeded into labelled wells of a sterile 96-well U-bottom plate (BD

Falcon; Cat. No. 353077). A further 80 μL of R10 media (Section 2.2.1, Drug

Stimulation and Cellular Expansion), either containing the appropriate test drug (CBZ)

or control drug was added to the 100 μL of cells, to obtain a final stimulant

concentration higher than 10 μg/mL, which would be rectified with the addition of

42

Brefelden A (Sigma; Cat. No. B7651). For the unstimulated control cells, 80 μL of R10

media (Section 2.2.1, Drug Stimulation and Cellular Expansion) was added. Cells were

incubated at 37°C for 2 hours, before adding 20 μL of Brefelden A giving a final

concentration 10 μg/mL. The cells were incubated at 37°C for 6 hours (unless

otherwise stated)

After incubation, the cells were washed with 200 μL of IsoFLOW Sheath fluid

(Beckman Coulter; Cat. No. 8546859) and the supernatant removed. The appropriate

cell-surface fluorochrome-conjugated antibodies, listed in Table 2.3 was added to the

cells and incubated in the dark (plate wrapped in aluminium foil) at 4˚C for 15 minutes.

Cells were then fixed and permeabilized using the IntraPrep Reagent kit, following the

protocol outlined by the manufacture (Beckman Coulter; Cat. No. A07802) (182).The

appropriate fluorochrome labelled cytokine antibody, was added and the cells

incubated at 24○C, in the dark (plate wrapped in aluminium foil) for 15 minutes (Table

2.3). Cells were washed and resuspended in 200 μL of IsoFLOW. Flow cytometry was

performed on the stained cells using the Gallios Flow Cytometer and the raw data

analysed using the Kaluza flow cytometry software.

2.2.2.2 Surface Marker staining and Immunophenotyping

Similar to the ICS flow cytometry assay (Section 2.2.2.1) The multicolour flow assays

and reagent panel combinations were carefully selected using spectralviewer anaylsis

provided online by BD Biosciences. Instrament set up controls inculded BD

CompBeads (BD Biosciences, Cat. No. 644204), fluorescence minus one and

unstained controls, and specific biological controls like unstimulated samples. Cells

were seeded into labelled wells, stained with cell-surface marker fluorochrome-

conjugated antibodies, listed in Table 2.3 and incubated on ice in the dark for 25-30

minutes. After incubation, the cells were washed and resuspended in 200 μL of

IsoFLOW reagent. Flow cytometry was performed on the stained cells using the

Gallios Flow Cytometer and the raw data analysed using the Kaluza flow cytometry

software.

2.2.2.3 Tetramers

Tetramer folding and construction

HLA class-I complex folding mixture was made as outlined in Table 2.4 and by Leisner

et al. and Svitek et al. (183, 184). The β2M and B*1502 reagents were supplied by our

colleagues and made as outlined in Leisner et al. and Svitek et al. (183, 184). After

incubating the HLA class-I complex folding mixture for 48 hours at 18-24 C, it was

43

filtered through a 0.5 mL 100kD membrane spin filter by centrifuging at 15,000 relative

centrifugal force (RCF) or 14245 revolutions per minute (RPM) (FX241.5P rotor in

Microfuge 16) for 2 minutes at 4-6◦C.

The folded peptide/HLA class-I complex was taken through tetramer production

process. 100 μL of the peptide/HLA class-I complex was added to separate labelled

sterile Safe-Lock 1.5mL Eppendorf Tubes (Crown Scientific; Cat. No. 0030.120.086)

and streptavidin-PE (1.85 μM BD Bioscience; Cat. No. 554061) was added in

sequential aliquots as highlighted Table 2.6 with a final concentration of streptavidin-

PE being 4.10% of the starting peptide/HLA class-I complex volume. The PE-

tetramers were aliquoted into labelled cryotubes (1 mL, Nunc Bank-It; Cat. No.

374078) at a smaller working volume of 10-20 μL to limit the amount of freeze/thawing

and stored at -80’C until required.

Table 2.5: Tetramer folding mix.

Reagents Stock Conc. (μM) Final Conc. (μM) Volume (μL) x1 reaction

Sterile deionised H2O - - 294

Tris/maleate (pH 6.6)

6 1 67

Lutrol (F68) 1 0.03 12

*Protease inhibitors 100 1 4

Peptide 500 6 5

β2M (FN345-346) 80.6 3 15

B1502 (FN286) 32.7 0.3 37

Total - - 400

*Outlined in Table 2.4 μM, micromolar; μL, microliter; β2M, Beta 2-2-Microglobulin; B1502, light chain B*15:02 sequence

Table 2.6: Protease inhibitor mix.

Protease inhibitor mix (100x concentration) Volume (μL)

Sterile deionised H2O 37.50

EDTA 14.80

Pepstatin A 4.80

1,10- phenantroline 4.80 NEM 7.40 TLCK 1.30

TPCK 1.30

PMSF 2.10

Total volume 74.00

EDTA, Ethylenediaminetetraacetic acid; NEM, N-Ethylmaleimide; TLCK, Tosyl-L-lysine chloromethyl ketone; TPCK, Tosyl-L-phenylalanine chloromethyl ketone; PMSF, Phenylmethanesulfonylfluoride; μL, microliter.

44

Table 2.7: Addition of streptavidin-PE labelled to peptide/HLA class-I complex.

Time (minutes) Volume of streptavidin-PE (1.85 μM) added to 100 μL of peptide/HLA class-I complex

0 1.2 μL

15 0.60 μL

25 0.60 μL

35 0.60 μL

45 0.60 μL

55 0.60 μL

Total streptavidin-PE 4.10 μL

μM, micromolar; μL, microliter; PE, phycoerythrin; HLA, human leukocyte antigen

Tetramer staining

Tetramer staining followed the same basic flow cytometry protocol seen in Section

2.2.2.2 Surface Marker staining and Immunophenotyping, with the exception that 5 μL

of the PE-tetramer was added and the cells incubated in the dark (plate wrapped in

aluminium foil) at 24 C for 20 minutes before the cells were stained with appropriate

cell-surface fluorochrome-conjugated antibodies, listed in Table 2.3 to the appropriate

samples. Once stained, cells were washed and resuspended in 200 μL of IsoFLOW

reagent. Flow cytometry was performed on the stained cells using the Gallios Flow

Cytometer and the raw data analysed using the Kaluza flow cytometry software.

2.2.2.4 Cytometric Bead Array cytokine assay

The Cytometric Bead Array (CBA) Th1/Th2/Th17 kit (BD Scientific; Cat No. 560484)

included capture beads for the cytokines; IL-2, IL-4, IL-6, IL-10, TNF, IFN-γ, and IL-

17A, each of which has a unique fluorescence intensity so that beads can be mixed

and the assay performed in a single tube. This assay was performed on plasma

samples, as outlined by the manufacture. Plasma samples were serial diluted with the

supplied assay diluent, to find the optimal plasma to assay diluent ratio. The optimal

dilution factor was 1:4; also the manufactures recommended dilution factor for serum

based samples.

45

2.2.3 Enzyme-Linked ImmunoSpot

Table 2.8 List of reagents and materials used in ELISpot assay.


MutliScreen-IP Hydrophobic PVDF

sterile Filter ELISpot Plate(s) Mabtech

MAIPS4510

anti–IFN-γ antibody Mabtech 1-D1k

Dimethyl sulfoxide (DMSO) Sigma 472301

N-N-Dimethylformamide (DMF) Sigma 140732

Anti-human CD3 monoclonal

antibodies Mabtech 3605-1-50

CEF peptide pool Mabtech 3615-1

Staphylococcal enterotoxin B (SEB) Sigma S4881-1MG

anti–IFN-γ biotinylated antibody Mabtech 7-B6-1

streptavidin-HRP Mabtech 3310-9

3’, 3’, 5’, 5’-tetramethylbenzidine (TMB) Mabtech 3652-R10

AID automated microplate ELISpot

reader and associated software AID V 5.0 - B7337

2.2.3.1 ELISpot assay

IFN-γ ELISpot assay was performed following the manufacturer’s instructions (R&D

Systems). Cells were resuspended in R10 media (Section 2.2.1, Drug Stimulation and

Cellular Expansion) supplemented with 10U/mL of rhIL-2 at 2x106 cells/mL. The 2x105

cells (100 μL), in duplicate/triplicate, were seeded into the appropriate wells of a

MultiScreen-IP, Hydrophobic PVDF Sterile Filter ELISpot Plate(s) (Mabtech; Cat No.

MAIPS4510), coated with 2 μg/mL of anti–IFN-γ antibody (Mabtech; Cat. No. 1-D1k),

then incubated overnight or for a minimum of 18 hours, at 4°C and blocked with R10

media (Section 2.2.1, Drug Stimulation and Cellular Expansion).

Negative controls included the unstimulated controls in which 100 μL of R10 media

(Section 2.2.1, Drug Stimulation and Cellular Expansion) supplemented with 10U/mL

of rhIL-2 was added to the seeded cells. The solvent controls included adding 100 μL

of R10 media (Section 2.2.1, Drug Stimulation and Cellular Expansion) supplemented

46

with 10U/mL of rhIL-2 media containing Dimethyl sulfoxide (DMSO) /

Dimethylformamide (DMF) (Sigma; Cat No. 472301 and 140732, respectively), giving

a final concentration 0.1 μL/200 μL. The negative controls were performed in triplicate

(unless otherwise stated).

The positive controls (duplicate) included, adding 100 μL R10 media (Section 2.2.1,

Drug Stimulation and Cellular Expansion) supplemented with 10U/mL of rhIL-2 with

Anti-human CD3 monoclonal antibodies (Mabtech; Cat. No. 3605-1-50), giving a final

concentration of 0.1 μg/mL per well. Secondary positive controls included 100 μL R10

media (Section 2.2.1, Drug Stimulation and Cellular Expansion) supplemented with

10U/mL of rhIL-2 with 1 μg/mL of either CEF peptide pool (Mabtech; Cat. No. 3615-1)

or Staphylococcal enterotoxin B (SEB) (Sigma; Cat No. S4881-1MG).

Drug-stimulated IFN-γ ELISpot assays involved control drugs (acyclovir) and the drugs

being analysed, CBZ or NVP (Sigma; Cat. No. SML0097 or Boehringer Ingelheim;

VIRAMUNE 200mg immediate release tablets (Appendix 9, Section 9.1 Nevirapine

isolation from Viramune tablets)). Drugs and synthesised peptides were stored in

DMSO or DMF at a stock concentration of 10 mg/mL were diluted with R10 media

(Section 2.2.1, Drug Stimulation and Cellular Expansion) supplemented with 10U/mL

of rhIL-2 and added (in triplicate) to seeded cells, to give a final concentration of 10

μg/mL (unless otherwise stated).

After incubation, ELISpot plates were washed with sterile Phosphate-buffered saline (1

x PBS) solution (Gibco, Cat. No. 70013032) and incubated with anti–IFN-γ biotinylated

antibody (Mabtech, Cat. No. 7-B6-1) for 2 hours. After incubation with the anti–IFN-γ

biotinylated antibody, ELISpot plates were washed, and incubated for 1 hour with

streptavidin-HRP (Mabtech, Cat. No. 3310-9), and then washed again. Plates were

substrated for 8-12 minutes using the 3’, 3’, 5, 5’-Tetramethylbenzidine substrate

(Mabtech, Cat. No.3652-R10). Once the plates were air-dried, counts were evaluated

with an AID automated microplate ELISpot reader and associated software (AID

ELISpot Version 5.0, B7337). The AID reader was staged, calibrated, and set to the

manufactures default settings. Files were saved as raw images and a PDF file was

created using the Nitro PDF created with the details of the time, date, cytokine being

analysed, media used, and the sample(s) identifying number.

47

2.2.4 Lymphoblastic Cell Lines

2.2.4.1 EBV supernatant

Lymphoblastic cell lines (LCL) were generated from patient peripheral B cells using

EBV B95-8 strain. EBV supernatant was generated from cryopreserved EBV B95-8

cells. Each vial was thawed and resuspended in RPMI-1640 media, which was added

in a drop wise fashion to make a final volume of 10 mL. EBV cells were centrifuged,

the supernatant removed, and cells resuspended in 10 mL of RPMI-1640 media

supplemented with sterile filtered, non-heat inactivated FCS (v/v 20%), sodium

pyruvate (1μM), and L-glutamine (1μM) (185).

The EBV cells were transferred to a labelled, sterile, T-25 culture flask (BD Falcon;

Cat. No. 353108) and incubated at 37°C, until an appropriate amount of cells were

acquired and/or a change in the phenol red pH indicator from red to yellow, indicating

the change of phenol to phenolic acid, (7-10 days). The supernatant was collected,

and filtered through a 0.45μm sterile filter (Sigma; Cat. No. CLS431225). The sterile

supernatant was then aliquoted into sterile 1.8 mL cryotubes and frozen to -80◦C

before transfer into liquid nitrogen storage. The EBV cells were also resuspended in

cryopreservation media (heat-inactivated FCS with 10% DMSO) and stored in liquid

nitrogen. Up to 5 million (5x106) cells were used to for Mycoplasma testing using the

Mycoplasma plus PCR primer kit (Agilent; Cat. No. 302008). (186, 187)

2.2.4.2 Lymphoblastic cell line generation

PBMC’s were thawed and washed with non-heat inactivated FCS based F20 media

(Section 2.2.4.1, EBV supernatant) and incubated at 37°C overnight (185). The

following day cell viability and counts were performed using 0.4% Trypan Blue solution

and the Invitrogen Countess method (Chapter 2, Section 2.2.1 Drug Stimulation and

Cellular Expansion). Cells were resuspended in non-HI serum based F20 media at 0.5

to 1 million (0.5x106 to 1x106) cell per mL

2.5 mL of PBMC’s / F20 were aliquoted into labelled wells of a 6-well flat-bottom

culture plate and a further 0.5 mL of EBV supernatant was added to make a final

volume of 3 mL15 μL of the Lectin Phytohemagglutinin from Phaseolus vulgaris

or PHA-P (Sigma; Cat No. L8754) at a concentration of 2.5 mg/mL was then added to

the 3 mL of PBMC’s / EBV supernatant sample. The cells were incubated at 37°C, for

up to 14 days, with an extra 1 mL of in non-heat inactivated FCS F20 media being

added at day 4 and 12 (185).

48

After 14 days the cells were counted, centrifuged, and resuspended in F20 media

made with components as mentioned in Section 2.2.4 Lymphoblastic Cell Lines

Lymphoblastic Cell Lines, except for the addition of sterile filtered, heat-inactivated

FCS (20% v/v) and Penicillin-Streptomycin (1% v/v). Cells were transferred to a T-25

culture flask and monitored as they grew, with cell counts being performed and the

appropriate amount of F20 (with heat-inactivated FCS and Penicillin-Streptomycin)

being added to keep a concentration of 1 million (1 x 106) cells per mL. Cultures were

transferred to sterile T-75 culture flasks (BD Falcon; Cat. No. 353136) when required

and once the suitable numbers of LCLs were generated, they were cryopreserved until

required. A 5 million (5x106) cells aliquot was used for Mycoplasma testing and a

further 1-2 million (1-2 x 106) cells were used to determine the B-cell and T-cell

frequencies in the flow cytometry LCL cultures at various stages (186, 187).

2.3 Virtual Modelling and in-silico Peptide Design

The virtual modelling of HLA molecules and drug/peptide interactions was performed

by Dr David Ostrov and colleagues, University of Florida College of Medicine,

Gainsville, Florida, USA Several 9-11mer, HLA-B*15:02-restricted HSV-1 (strain 17)

and HSV-2 (strain HG52) peptides were synthesised based on the in-silico epitope

mapping predictions performed by colleagues at The Institute for Immunology and

Infectious Diseases, Murdoch University, Perth, Australia using Net-MHC3.2 based

algorithms, combined with a novel measure of “targeting efficiency” developed in

collaboration with Microsoft research. (169, 170, 188-190)

49

2.4 Molecular Assays

Table 2.9 List of reagents and materials used in molecular assays. Specific PCR associated reagents

and concentrations are listed in the relevant tables.


Maxwell 16 Total RNA purification kit Promega 1050

RNeasy mini-kit QIAGEN 74104

Tris-EDTA (TE) buffer Sigma 93283

4% E-gel Invitrogen G700892

QIAGEN MinElute PCR purification kit QIAGEN 28604

NanoDrop DNA quantification Thermo Fisher ND-1000 with v3.2.1

software

Axygen 96-well PCR Half Skirt plate VWR PCR-96-HS-C

DG8 Cartridge and gaskets Bio-Rad 186-4007

DG8 Cartridge holder Bio-Rad 186-3051

QX100/200 ddPCR Droplet Generator Bio-Rad 186-4002

VIAFLOW Electronic pipette

(50-1250 μL) and tips (1250 μL)

INTEGRA

Biosciences

4623

(replacement tips 4445)

Hard-Shell 96-Well PCR plate Bio-Rad HSP 9601

Peirceable foil heat seal Bio-Rad 181-4040

ddPCR Droplet reader oil Bio-Rad 186-3004

C1000 Thermal Cycler 96-Deep Well Bio-Rad 185-1197

QuantaSoft, QX100/200 ddPCR reader Bio-Rad 186-4003

QuantaSoft, QX100/200 ddPCR

analysing software

Bio-Rad Version 1.5.38.1118

2.4.1 RNA isolation and complementary DNA conversion

Cells isolated from cell culturing assay’s (Section 2.2.1 Drug Stimulation and Cellular

Expansion) were stored in RNAlater stabilisation solution, following the manufactures

instructions. When ready for RNA extraction, the cells were removed from RNAlater

suspension and resuspended in the appropriate amount of 1 x PBS. RNA was

50

extracted from the cells using either the Maxwell 16 Total RNA purification kit

(Promega; Cat. No. 1050) or the RNeasy mini-kit (QIAGEN; Cat. No. 74104) and

following the appropriate protocol outlined by the manufacture. For the ddPCR, a two-

step reserve transcriptase PCR approach was used to assess RNA expression. RNA

conversion to complementary DNA (cDNA) was performed by preparing the reagent

mix and following the protocol outlined in Table 2.10 while following the thermocycling

conditions outlined in Table 2.11

Table 2.10: complementary DNA (cDNA) conversion reagents and Master mix.

Reagent Final concentration

x1 Reaction volumes (µL)

Oligo d(T) 20mer (50 µM) (Invitrogen; Cat. No. N8080128)

2.5 µM 1

RNA - 10

Samples incubated at 65◦C for 5 minutes

5x first strand buffer (Invitrogen; Cat. No. 18080-093)

1x 4

dNTPs (40mM) (Invitrogen; Cat. No. 10297-117)

4 mM 2

dithiothreitol (0.1mM) (Invitrogen; Cat. No.15508-013)

5 μM 1

Sterile deionised Water (Sigma; Cat. No. W3500-1L)

- 0.5

RNase out (40 U / μL) (Invitrogen, Cat. No.10777-019)

20 U 0.5

Samples then incubated at 37◦C for 5 minutes

SSIII Reverse Transcriptase (200 U / μL) (Invitrogen; Cat. No. 18080-093)

200 U 1

dNTPs, Deoxyribonucleotide triphosphates (dATP, dCTP, dGTP, and dTTP); μM, micromolar; μL, microliter.

Table 2.11: Cycling conditions for complementary DNA (cDNA) conversion

Cycle Temperature (°C) Time

Cycle 1 x 42 5 minutes

Post cycle 10 infinite

51

2.4.2 Digital Droplet PCR

2.4.2.1 Reference sequence Amplification, Purification, and

Quantitation

As TCRs are highly variable and it is difficult to have a positive control for all the TCR

combinations. A synthetic construct was designed as a positive control and for the

assay workup. The g-Blocks gene fragment reference sequence incorporated the

various primer sequences to be used in the amplification and identification Vβ TCR

clonotype mRNA using the ddPCR method.

Reference sequence.

5’-ATCGACA..(M13Fw)…TGTAAAACGACGGCCAGT..ATGGGCTCCAGGCTGCTCTGTTG

GGTAGCAGGTGAGTCCCTGCAGACAGGATAGCGTGACATCTACCAGACCCCAAGATACC

TTGTTTTATAGGGACAGGAAAGAAGATCACTCTGGAAATGTGAGTTCTCAGCTAAGATGC

CTAATGCAGGCTTT...(Vβ12Fw1)…TGAAGATCCAGCCCTC..GGGTC AAG...(CβFw1)…

TCCAGTTCTACGGGCTCTCG...GAGA...(Cβ-Fw2)…ATGACGAGTGGACCCAGGATAGGG

...TATATGCCAGGCCCTCACATACCTCTCAGTC...(Vβ25ISGSY-CDR3-Fw1)…TTGTAGGA

ACCTGAGATGGAA...AGTC...(Vβ25-GLAGVDN-CDR3)…TTGTTGTCAACTCCCGCTAACC

CACT...GGCCCG...(Vβ12Re1)…TTTAGCCGGGGAGCTGTTT..CCCCAA...(CβRe2)…ACCC

GTCACCCAGATCGTC...AGCGTGTTCTCAAACCATGGGCCAACGC…(Vβ25-Re)…GAAGC

AGAGATCTCCCACACC…ATCCCGCAGTAAAGGCTGGAGTCACTCAAACTAC...

(M13-Rw)…GGTCATAGCTGTTTCCTG..CAGAA TCCG-3’

Table 2.12: M13F and M13R primers used to amplify the reference sequence to use the amplicons to

as a positive control.

Primer Primer Sequence (5’-3’) Length GC% Tm (°C)

M13-Fw

(Forward)

TGTAAAACGACGGCCAGT 18 50 63.59

M13-Re

(Reverse)

CAGGAAACAGCTATGACC 18 50 63.59

GC%, percentage of cytosine and guanine within primer; Tm (°C), calculated melt temperature; and bp, base pairs.

The reference cDNA sequence was resuspended in Tris-EDTA (TE) buffer (Sigma;

Cat no. 93283) giving at 10 ng / μL. The M13 primers listed in Table 2.12 and the

mastermix reagents and concentrations highlighted in Table 2.13 were used, along

with the PCR thermocycling conditions, highlighted in Table 2.14, to amplify the

reference gene. Amplified products were then analysed via electrophoresis on either a

4% E-gel run for 20 minutes (Invitrogen; Cat No.G700892) and purified using the

52

QIAGEN MinElute PCR purification kit (QIAGEN; Cat. No. 28604) following the

manufacture protocols.

Table 2.13: Reference sequence amplification reagent Mastermix to amplify the reference sequence to

use the amplicons to as a positive control.

Reagent Final

concentration

x1 reaction volumes

(µL)

*PCR buffer (10x)

(Invitrogen; Cat. No. 18067-017)

*1x *2

MgCl2 (50 mM)

(Invitrogen; Cat. No. 18067-017

1.5 mM 0.6

dNTP (40 mM)


2 mM 1

Forward primer (25 µM) 0.625 nM 0.5

Reverse primer (25 µM) 0.625 nM 0.5

*Platinum Taq (5U / µL)


*2.5 Units *0.5

Sterile deionised H2O - 13.2

Ref sequence cDNA (10 ng / µL) 10 ng / μL 1 *1-2.5 Units recommended by Invitrogen for use of Platinum Taq DNA polymerase #Reference sequence was resuspended in TE buffer at 10ng/μL. IDT, the manufacture recommends using 0.1-10 ng for amplification. mM, millimolar; µM, micromolar; nM, nanomolar; U, units; cDNA, complementary DNA; PCR, polymerase chain reaction.

Table 2.14: Reference sequence amplification PCR cycling conditions for the amplification of the

reference sequence to use the amplicons to as a positive control.

Cycle Temperature (°C) Time

Pre-cycle 94 6 minutes

Cycle (40x)

95

60

72

30 seconds 30 seconds 45 seconds

Extension 72 5 minutes

Post cycle 15 Infinite

Quantification of the reference sequence amplicons was done using NanoDrop DNA

quantification (Thermo Fisher; Cat. No. ND-1000 with v3.2.1 software) and the DNA

copy number calculated using the following equation; The amplicons were diluted with

TE buffer to obtain a range of copy number concentrations ranging from 7.48 x 1011

copies per μL to 7.48 copies per μL.

53

Copy No. = Amount of DNA (ng) x 6.022x1023

((length (bp) x 1x109(ng/g) x 650 (g/mole of bp))

2.4.2.2 Digital droplet PCR protocol

The forward and reverse primers listed in Table 2.17 were validated using the

reference cDNA sequence (Appendix 9, Section 9.2). The validated primers were

added to into labelled sterile Safe-Lock 1.5mL Eppendorf Tubes, to give a final

concentration of 100 nM per primer, along with the 2x EvaGreen supermix for ddPCR

and sterile deionised water (Table 2.15). To automate the addition of mastermix and

samples, to an Axygen 96-well PCR Half Skirt plate (VWR: Cat. No. PCR-96-HS-C),

the Beckman Coulter Biomek FXP automated workstation requires the relevant

position, size, and volumes data, be saved into a comma-separated values (CSV) file,

which the robot can read and perform accordingly. This CSV file is simple file format

used to store tabular data which can be converted and saved in plain text.

For each sample, 22.5 μL of the mastermix was added to each well of an along with

2.5 μL of sample cDNA, or the appropriate control. The positive control reference

sequence amplicons were added manually in the designated Post-PCR Laboratory or

the laboratory designated for all samples that have already undergone PCR

amplification in order to prevent cross contamination of amplified PCR products in

samples being prepared in the designated Pre-PCR Laboratory or the laboratory

designated for all samples which have not undergone PCR amplification. Once

complete, plates were sealed using adhesive PCR film samples centrifuged

(Eppendorf Centrifuge) and droplet generation was performed in the Post-PCR

Laboratory.

The DG8 cartridge was placed into the ddPCR cartridge holder and 20 μL of the 25 μL

PCR was loaded into the middle sample well section of the DG8 Cartridge, in the

appropriate order. Then 70 μL of the ddPCR Droplet Reader Oil was loaded into the

first, reader oil well and the DG8 gasket added to the ddPCR cartridge holder with the

loaded ddPCR cartridge (Figure 2.2). The cartridge was then placed into the

QX100/200 ddPCR Droplet Generator and droplet generated (Figure 2.2).

After droplet formation, 40 μL of the sample droplets were transferred to the Hard-

Shell 96-Well PCR plate (Bio-Rad, Cat. No. HSP9601), in the same order as

highlighted in the CSV file generated for the automated process. The plates were

sealed with a peirceable foil heat seal, then plates placed into C1000 Thermal Cycler

54

96-Deep Well and exposed to the thermocycling conditions outlined in Table 2.16.

After thermocycling the plate was placed into the ddPCR droplet reader and the raw

data analysed using the QuantaSoft, QX100/200 ddPCR analysing software (Figure

2.2).

Figure 2.2: General overview of digital droplet PCR (ddPCR) assay workflow from PCR mastermix

generation through to droplet generation, amplification, QX100/200 plate reading and analysis.

Table 2.15: Reagent mastermix for EvaGreen ddPCR.

Reagent Final concentration x1 reaction volumes (µL)

ddPCR EvaGreen supermix (2x) (Bio-Rad Cat No. 186-4033)

*1x *12.5

Forward primer (25 µM) 100 nM 0.1

Reverse primer (25 µM) 100 nM 0.1

Sterile deionised H2O - 13.2

Ref sequence cDNA - 2.5

µM, micromolar; nM, nanomolar; µL microliters; cDNA, complementary DNA; ddPCR, digital droplet polymerase chain reaction.

55

Table 2.16: Digital droplet PCR (ddPCR) cycling conditions for the amplification of Variable beta (vβ)-

25-1*01 and -12-3/4*01 CDR3 specific sequences, constant beta (Cβ), and ribonuclease P RNA

component H1 (RPPH1) reference gene sequences.

Cycle Temperature (°C) Time Pre-cycle 95

5 minutes

95 30 seconds Cycle (40x) 58 30 seconds

60 30 seconds

Signal Stabilisation

4 5 minutes

90 5 minutes

Post cycle

10

Infinite

Table 2.17: vβ 25 and 12 CDR3 specific, and reference gene primers.

Primer Primer Sequence (5’-3’) Length

(bp)

GC% Tm (°C)

Vβ25– ISGSY fwd. CDR3 Specific primer

TTGTAGGAACCTGAGATGGAA 21 57.1 63.65

Vβ25 – GLAGVDN fwd. CDR3 Specific primer

TTGTTGTCAACTCCCGCTAACCCACT 26 50.0 63.87

Vβ25 – ISGSY and GLAGVDN rev

GGTGTGGGAGATCTCTGCTTC 21 57.1 63.79

Vβ12 Fwd. Primer TGAAGATCCAGCCCTC 16 56.3 63.53

Vβ12 CDR3 specific rev AAACAGCTCCCCGGCTAAA 19 52.6 63.66

Cβ fwd TCCAGTTCTACGGGCTCTCG 20 60.0 63.78

Cβ rev GACGATCTGGGTGACGGGT 19 63.2 63.77

RPPH1 F AGTGAGTTCAATGGCTGAGG 20 50.0 63.68

RPPH1 R GGCGGAGGAGAGTAGTCT 18 61.1 63.70

Vβ, Variable beta; Cβ, constant beta; and RPPH1, ribonuclease P RNA component H1 gene primers. GC%,

percentage of cytosine and guanine within primer; Tm (°C), calculated melt temperature; and bp,

2.4.2.3 QX100/200 ddPCR raw data analysis

The analysis of the QX100/200 raw data output and statistical parameters of the

ddPCR system is dependent on the partitioning of the template DNA/cDNA into

uniform droplets. The QuantaSoft software (Version 1.5.38.1118) measures the

amount of positive and negative droplets based upon the fluorophore in each sample.

56

The fraction of positive droplets is then fitted into a Poisson distribution algorithm to

determine the starting concentration of the Target DNA/cDNA template as copies per

μL and copies per reaction (copies per 20 μL). For accurate quantitation with a 95%

confidence level, a minimum of 14,000-15,000 total droplets are required when using

the EvaGreen ddPCR method. For the ddPCR specific Vβ CDR3 TCR mRNA analysis,

TCR mRNA expression levels were normalised using the expression of the

ribonuclease P RNA component H1 (RPPH1) gene and reporting Vβ CDR3 TCR

mRNA expression levels as a percentage of normalised Cβ mRNA expression levels.

57

Chapter 3

Nevirapine-Induced Hypersensitivity

58

59

3 3.1 Introduction

Based on established drug-HSR models, which illustrates the importance of the HLA

peptide-binding cleft in specific drug interactions, this chapter will explore the notion

that NVP-HSR tissue specificity and varied HLA allele associations explained in

Chapter 1, Section 1.3.3 Nevirapine, Table 1.2, can be explained by similarities in HLA

peptide-binding cleft structures and predicted peptide specificity and chemistry. We

sought to compare significant NVP-HSR HLA alleles across ethnic groups by grouping

NVP-HSR significant HLA alleles according to shared key amino acid residues which

make up the HLA class-I binding cleft pockets (A-F), based on similarities in cross-

reactivity of public epitopes among HLA class-I alleles and binding cleft pocket

chemistry (CREG HLA supertypes), and based on MHCcluster algorithm grouping

(pb), which groups HLA molecules on their peptide-binding specificity as predicted by

NetMHCpan 2.8, (Table 3.1) (168-172, 191).

This chapter will also explore the characteristic unique to NVP-SCARs, the diminishing

ex-vivo NVP-stimulated, PBMC’s IFN-γ responses in patients as they recover from

DRESS. As identified by Keane et al. 2014, the NVP-stimulated IFN-γ responses in

NVP-DRESS patient PBMC’s, diminish over time, becoming undetectable after

recovery. Keane et al. 2014 explored the hypothesis that these diminishing responses

were due to an increase in CD4+ Tregs levels after NVP withdrawal and during the

recovery period. In the paper by Keane et al. 2014, one patient exhibited an increase

in Tregs that corresponded with the waning IFN-γ ELISpot responses (138).

In this section of the chapter, this hypothesis will be investigated further. However, as

most of the NVP-DRESS patient PBMC’s samples had been used in the first study by

Keane et al. 2014, we analysed the cytokine plasma profiles of the same NVP-DRESS

patients, used in the Keane et al. 2014. We expected that the cytokine environment

within the donor’s plasma, over the time course of a NVP-DRESS reaction and

recovery period, alongside the Keane et al. 2014 data, might show changes in specific

T-cell subpopulations further explaining the diminishing NVP-stimulated, PBMC IFN-γ

responses.

60

Table 3.1 HLA cross-reactive group (CREG) supertypes and MHC Cluster peptide-binding (pb-) groups

HLA grouping HLA alleles (HLA-) Grouping characteristics

CREG B*07 supertype

B*07, B*08, B*13, B*54, B*55, B*56,

B*27, B*60, B*61, B*41, B*42, B*47,

B*48, B*59, B*67, B*81, and B*82 Grouped based on similarities in

cross-reactivity of public epitopes

among HLA class-I alleles and

binding cleft pocket chemistry

(171, 172, 191) CREG B*05

supertype

B*18, B*35, B*46, B*49, B*50, B*51,

B*52, B*53, B*57, B*58, B*62,-B*63,

B*71, B*72, B*73,B*75, B*76, B*77

and B*78

pb-B*35 B*35:(01/05/08), -B*53:01

Grouped based on their peptide-

binding specificity as predicted by

NetMHCpan 2.8 (160-162)

pb-B*78 B*78:01, B*55:01, B*56:01

pb-C*05 C*04:(03/06) C*05:(01/09)

C*08:(01/02) C*17:01

pb-C*18 C*04:(01/05/07), C*18:01

3.2 Hypothesis and Aims

Hypothesis 1

The predicted peptide-binding properties, the structure of the HLA class-I molecule

peptide-binding cleft, in particular, the B- and F-pockets, and predicted NVP-binding

properties may be shared between the various HLA alleles associated with

NVP-SCAR with hepatitis and NVP-SCAR without hepatitis.

Aim 1

Retroactively analysing the patient and demographic data from a NVP-HSR

associated cohort with controls (ClinicalTrials.gov NCT00310843 Chapter 2,

Nevirapine-Induced HSR 2.1.1) and the NVP-SCAR and NVP-SCAR with hepatitis

DNA samples will be sequenced and HLA typed. The data will then be analysed to

identify significant HLA alleles and grouping associations, their predicted peptide-

binding properties, the structure of the HLA class-I molecule peptide-binding cleft, in

particular, the B- and F-pockets, and predicted NVP-binding properties.

Hypothesis 2

The cytokine plasma profiles of NVP-SCAR patients over the time course of the

reaction represents the changes in T-cell subpopulations, explaining the diminishing

NVP-stimulated, PBMC IFN-γ responses.

61

Aim 2

The CBA Th1/Th2/Th17 kit (BD Scientific; Cat No. 560484) and the method outlined in

Chapter 2 Materials and Methods, Section 2.2.2.4 Cytometric Bead Array cytokine

assay will be used to analyse the cytokine profiles of the Plasma from NVP-DRESS

patents obtained over the time course of their NVP-DRESS reaction.

3.3 Methods

3.3.1 NVP-SCAR Associated HLA Class-I Binding Cleft

Characteristics

3.3.1.1 HLA four digit typing and Samples

On the samples highlighted in Chapter 2 Materials and Methods Section 2.1.1.1 NVP-

SCAR associated HLA class-I binding cleft characteristics, HLA-A, HLA-B,

HLA-C, and HLA-DR PCR amplification was performed by myself in conjunction with

staff at The Institute for Immunology and Infectious Diseases, Murdoch University,

Perth, Australia. Using the next generation sequencing 454 FLX+ sequencer platform,

the amplified products were sequenced and the raw data analysed using accredited

HLA allele caller software that uses the latest IMGT HLA allele database as the allele

reference library by colleagues at The Institute for Immunology and Infectious

Diseases, Murdoch University, Perth, Australia (Table 3.2).

3.3.1.2 Statistical analysis and virtual modelling

Both univariate and multivariate analyses were performed and stratified for race

according to HLA class-I alleles, HLA CREG supertypes, and peptide-binding groups

assigned from MHC Cluster binding specificities (Table 3.1) (168, 169, 171) In this

study, clinical notes were re-evaluated by and myself, Dr Rebecca Pavlos, and

Professor Elizabeth Phillips (Table 3.2). Samples were excluded from this analysis

because no sample was available for HLA typing or issues with sample identity were

identified (n=101) and either no clinical data was reported (n=22) or ambiguous clinical

data was reported (n=62). The final analysis included 210 clinically defined NVP-HSR

cases (152 SCAR, 58 HSR with hepatitis) with 468 controls. The statistical analysis

was performed in conjunction with colleagues at The Institute for Immunology and

Infectious Diseases, Murdoch University, Perth, Australia and Vanderbilt University,

Nashville Tennessee, USA in particular Dr Elizabeth McKinnon, Dr Rebecca Pavlos

and Professor Elizabeth Phillips (Table 3.2). Virtual modelling of NVP-binding in the

62

HLA-C*04:01 allele, was performed by Dr David Ostrov and colleagues, University of

Florida College of Medicine, Gainsville, Florida, USA (Table 3.2).

3.3.2 Diminishing IFN-γ Responses in NVP-stimulated,

PBMC’s from NVP-DRESS Patients

A complete description of the Controls and Donors is given in Chapter 2 Materials and

Methods, Section 2.1.1.2 Diminishing ex-vivo PBMC’s responses and NVP-HSR

Immunopathogenesis. The CBA flow cytometry methods, outlined in Chapter 2,

Materials and Methods, Section 2.2.2.4 Cytokine Bead Array, were used to assess the

cytokine plasma profiles of NVP-DRESS patients. This study was part of a larger

retrospective cohort study that involved the analysis of HLA class-I and class-II alleles

association and examined NVP specific CD4+ and CD8+ T-cell responses by IFN-γ

ELISpot assay and flow cytometry. The results of this larger study were published in

AIDS: HLA Class-I Restricted CD8+ and Class-II Restricted CD4+ T Cells Implicated in

the Pathogenesis of Nevirapine Hypersensitivity by Keane et al. 2014 (Appendix 9,

Section 9.2 Publications Highlights) (138). Statistical analysis and graphical

representations were calculated using GraphPad Prism software Version 5.02

(GraphPad Software Inc., California USA).

63

Table 3.2 Distribution of the work within Chapter 3 and accreditation to those who contributed

Work performed Accreditation

PCR amplification of HLA target sequences

from DNA

Craig Rive and staff at IIID

(NATA and ASHI accredited)

Sequencing of HLA amplicons Staff at IIID


Analysis of sequencing data and assigning of

HLA 4-digt types

Staff at IIID


Evaluation of clinical data Craig Rive

Rebecca Pavlos

Elizabeth Philips

NVP-SCAR analysis Rebecca Pavlos

Elizabeth McKinnon

NVP-HSR with Hepatitis analysis Craig Rive

Rebecca Pavlos

Elizabeth McKinnon

Virtual modelling of HLA-C*04:01 and NVP-

binding

David Ostrov, University of

Florida College of Medicine,

Gainsville, Florida, USA

NVP associated ELISpot, Treg flow cytometry,

and clinical data associated with

Donor 1 (NVP-HSR)

Niamh Keane and contributors

to the publication by

Keane et. al 2014 (138)

Cytokine analysis of NVP-DRESS patient

plasma

Craig Rive

PCR, Polymerase Chain Reaction; HLA, Human Leukocyte Antigen; DNA, deoxyribonucleic acid; NVP, nevirapine; SCAR, Severe cutaneous adverse reaction, IIID, Institute for Immunology and Infectious Diseases; NATA, National Association of Testing Authorities; ASHI, American Society for Histocompatibility and Immunogenetics.

3.4 Results

3.4.1 NVP-HSR Associated HLA Class-I Binding Cleft

Characteristics

3.4.1.1 NVP-HSR with hepatitis

The CREG B*07 and pb-B*78 groupings of HLA alleles as highlighted in Table 3.3

were both found to be significant in predicting NVP-HSR with hepatitis in Caucasians

(OR 2.18, p value 0.040 and OR 2.15, p value 0.016, respectively (Table 3.3). (168-

172, 191).

64

Table 3.3: NVP-HSR with hepatitis and statistically significant HLA CREG supertype and pb-groups

Phenotype Race Group P Value OR

Hepatitis Caucasian CREG B*07 0.016 2.150

pb-B*78 0.040 2.180

CREG, cross-reactive group; pb-, peptide-binding.

Associated HLA-B peptide-binding properties and MHC Cluster analysis

The MHCcluster analysis of the pb-B*78 group was compared to other common HLA

alleles identified within this cohort. Several HLA alleles belonged to the MHC Cluster

assigned pb-B*46 and pb-B*18 groups, as well as the pb-B*35 group (Table 3.1).The

MHCcluster analysis shows the pb-B*78 group alleles HLA-B*55:01,

-B*56:01, and -B*78:01 share close predicted peptide-binding capabilities (HLA-

B*55:01 to -B*56:01 distance = 0.093, HLA-B*55:01 to -B*78:01 distance = 0.111).

The pb-B*35 group alleles are represented by the HLA-B*35:01, -B*35:05, and -

B*53:01 alleles also share close predicted peptide-binding capabilities with (HLA-

B*35:01 to -B*35:05 distance = 0.039, HLA-B*35:01 and HLA-B*35:05 to HLA-B*53:01

distance = 0.254 and 0.273, respectively) (Figure 3.1 and Figure 3.2).

The MHCcluster tree and heatmap both show the pb-B*78 and pb-B*35 groups are

distinctly separate in terms of their predicted peptide-binding capabilities (distances

ranging from 0.361 to 0.539) (Figure 3.1 and Figure 3.2). The peptide-binding motifs in

Figure 3.1 shows how the pb-B*78 and pb-B*35 grouped alleles prefer to bind 9mer

peptides with a proline (P) or an alanine (A) at position 2 (P2 from the NH2 terminus of

the peptide) (168). We hypothesised that the HLA alleles within the pb-B*78 and pb-

B*35 groups would share structurally similar B-pockets.

65

Figure 3.1: MHCcluster tree created for the functional clustering of HLA-B alleles. Depicted are the

peptide-binding groups, pb-B*78 (red), pb-B*35 (orange), pb-B*46, and pb-B*18. (blue), with

representative predicted peptide-binding motifs (168).

Figure 3.2: MHCcluster heatmap representation for the functional clustering of HLA-B alleles, Depicted

are the peptide-binding groups, pb-B*78 (red), pb-B*35 (orange), pb-B*46, and pb-B*18 using the

MHCcluster algorithm (168).

66

Associated HLA-B binding cleft B-pocket.

While the MHCcluster analysis shows the pb-B*78 and pb-B*35 allele grouping are

distinctly different in their predicted peptide-binding capabilities, they both show a

tendency to bind a 9mer with a proline (P) or an alanine (A) at P2. The amino acid at

P2 of a HLA class-I bound peptide interacts with the HLA class-I binding cleft

B-pocket. An analysis of the binding cleft B-pocket structure was performed by

analysing the particular amino acids present in key positions that make up the binding

cleft B-pocket and therefore interact with the P2 amino acid of a bound peptide.

Multiple virtual modelling studies have shown that the amino acids at positions 7, 9,

24, 34, 41, 45, 63, 66, 67, 70, and 99 in the binding cleft B-pocket of the HLA-B

molecule are essential in the interaction with the P2 amino acid of a bound peptide

(171, 192-195). The HLA-B*55:(01/02), -B*56:(01/03/04) alleles of the MHCcluster

assigned pb-B*78 group, share the same amino acids sequence YYAVAENIYQY at

the binding cleft B-pocket positions 7, 9, 24, 34, 41, 45, 63, 66, 67, 70, and 99,

respectively. However, HLA-B*78:01 also grouped into pb-B*78 by MHCcluster, does

not share the same B-pocket amino acid sequence as the other pb-B*78 grouped

alleles but shares the same amino acid sequence as the MHCcluster pb-B*35 grouped

alleles (Table 3.4). The pb-B*78 alleles, HLA-B*55:01 and HLA-B*56:01 (minus HLA-

B*78:01) also belong to the validated HLA supertype CREG B*07 group, which was

also identified as significant in Table 3.1. This analysis has highlighted the potential for

the association of both HLA-B*55:01 and HLA-B*56:01 with NVP-HSR with hepatitis.

67

Table 3.4: HLA-B, B-pocket key amino acid residues YYAVAENIYQY at positions 7, 9, 24, 34, 41, 45,

63, 66, 67, 70, and 99 respectively for the CREG B*07 and pb-B*78 alleles HLA-B*55:(01/02),

B*56:(01/03/04), compared to the amino acid sequence of other HLA-B alleles.

HLA-B* Amino acid position B-pocket

7 9 24 34 41 45 63 66 67 70 99

55:(01/02),56:(01/03/04) YY Y A V A E N I Y Q Y

78:01,53:01,51:(01/02/04/08), 35:(01/02/03/04/05/08/43)

- - - - - T - - F - -

18:(01/02) - H S - - T - - S N -

44:(02/03/05) - - T - T K E I S N -

57:(01/02/03/04) - - - - - M E N M S -

58:(01/02) - - - - - T E N M S -

15:(01/12/24/25/35) - - - - - M E I S N -

15:02 - - - - - M - I S N -

15:03 - - S - - E - - - - -

46:(01/02) - - - - - M E K - - -

3.4.1.2 NVP-SCAR

The HLA-C*04:01 allele was significant in predicting NVP-SCAR, across all the broad

ethnic groups within this study, Asian (OR 5.73, p value < 0.0001), Caucasian (OR

1.92, p value 0.048) and, African (OR 4.00, p value 0.02) ancestry (Table 3.5). Ethnic-

specific analysis showed several HLA-C alleles to be significant in predicting NVP-

SCAR. These included HLA-C*05:01 in Caucasians/Hispanics when compared to non-

HLA-C*04:01 / non-HLA-C*05:01 carriers (OR 2.55, p value 0.004) and HLA-C*18:01

in African descendants, compared to non-HLA-C*04:01 / non-HLA-C*18:01 carriers

(OR 5.70, p value 0.007) (Table 3.5).

68

Table 3.5: NVP-SCAR statistically significant HLA-C alleles.

Phenotype Race/ethnicity Allele (HLA-) P Value OR

Rash Caucasian C*04:01 0.048 1.92

C*05:01 0.004 2.55

Rash African C*04:01 0.020 4.00

C*18:01 0.007 5.70

Rash Asian C*04:01 <0.0001 5.73

HLA-C peptide-binding properties and MHC Cluster analysis

The HLA-C*04:01, -C*05:01, and -C*18:01 alleles, and other common HLA-C alleles

identified within the cohort were analysed and assigned to specific peptide-binding

groups by the MHC Cluster program based on the NetMHCpan 2.8 (176, 177, 181). As

shown in Figure 3.3, the HLA-C*04:(01/05/07) alleles shared predicted peptide-binding

capabilities (distance < 0.0001). (168) The HLA-C*04:(01/05/07) alleles also shared

closely related predicted peptide-binding capabilities to the HLA-C*18:01 allele

(distance = 0.277). For this study, the HLA-C*04:(01/05/07) and HLA-C*18:01 alleles

were assigned to one peptide-binding group pb-C*18 (Figure 3.3 and Figure 3.4)

(168). The peptide-binding group pb-C*05 included the HLA-C*04:03, -C*05:(01/09)

alleles (HLA-C*04:03 to C*05:(01/09) distance = 0.0046). (168) Also identified in the

pb-C*05 group was HLA-C*08:(01/02) (in relation to HLA- C*05:(01/09) distance =

0.129), HLA-C*04:06 (in relation to HLA-C*05:(01/09) distance = 0.276), and HLA-

C*17:01 (in relation to HLA-C*05:(01/09) distance = 0.246) (Figure 3.3 and Figure 3.4)

(168). The peptide-binding motifs in Figure 3.3 show the predicted peptide-binding

capabilities of the MHCcluster assigned pb-C*18 and pb-C*05 HLA alleles. Both

MHCcluster assigned pb-C*18 and pb-C*05 group HLA alleles prefer to bind 9mer

peptides with either a leucine (L), phenylalanine (F), valine (V), isoleucine (I), or a

methionine (M) at P9 of the peptide which interacts with the binding cleft F-pocket of

the HLA class-I molecule (168). Given that the MHCcluster assigned pb-C*18 and pb-

C*05 HLA alleles are predicted to bind a 9mer similar amino acids at P2, we

hypothesised that the HLA alleles within these MHCcluster groups share structurally

similar F-pockets.

69

Figure 3.3: MHCcluster tree created for the functional clustering of HLA-C alleles. Depicted are the

various predicted peptide-binding groups, pb-C*05 and pb-C*18 (blue), with representative peptide-

binding motifs (168).

Figure 3.4: MHCcluster heatmap representation for the functional clustering of HLA-C alleles. Depicted

are the various predicted peptide-binding groups, including pb-C*05 and pb-C*18 (blue), using the

MHCcluster algorithm (168).

70

Associated HLA-C binding cleft F- and B-pocket

While the MHCcluster analysis showed that the HLA alleles assigned to the

pb-C*05 and pb-C*18 allele groups are different in their predicted peptide-binding

capabilities, they have the tendency to bind 9mer peptides with either a leucine (L),

phenylalanine (F), valine (V), isoleucine (I), or a methionine (M) at P9 of the peptide

which interacts with the binding cleft F-pocket of the HLA class-I molecule. Therefore,

analysis of the binding groove F-pocket structure was performed by analysing the

particular the amino acids in key positions of the binding cleft F-pocket (Table 3.6).

Multiple virtual modelling studies have shown that the amino acids at positions 74, 77,

80, 81, 84, 95, 97, 114, 116, 123, 133, 143, 146, and 147 in the F-pocket of the HLA-C

molecule, are essential in the interaction with the amino acid at P9 of a HLA class-I

bound peptide (171, 192-195). Table 3.4 shows the HLA alleles within the MHCcluster

assigned pb-C*18 and pb-C*05 groups share the same F-pocket amino acid sequence

DNKLYLRNFYWTKW at positions 74, 77, 80, 81, 84, 95, 97, 114, 116, 123, 133, 143,

146, and 147, respectively (171, 192-195). This analysis also looked at the key amino

acid residues of the binding cleft B-pocket amino acid residues positions 7, 9, 24, 34,

41, 45, 63, 66, 67, and 70. The B-pocket residue sequences of the MHCcluster

defined pb-C*18 and pb-C*05 grouped HLA alleles shared no significant similarities

(Table 3.6). The B-pocket residue sequences of the MHCcluster defined pb-C*18 and

pb-C*05 grouped HLA alleles are shared with by other HLA alleles identified in this

cohort, but not significant in NVP-SCAR. For example, the pb-C*18 alleles HLA-

C*04:(01/07) share the same B-pocket residues as HLA-C*14:(02/04) and the pb-C*05

alleles HLA-C*05:(01/09), share the same B-pocket residues as HLA-C*03:(03/04) and

HLA-C*08:(01/02/03) (Table 3.7).

71

Table 3.6: HLA-C F-pocket key amino acid residues, DNKLYLRNFYWTKW at positions 74, 77, 80, 81,

84, 95, 97, 114, 116, 123, 133, 143, 146, and 147, respectively for the HLA-C*04:(01/03/06/07), HLA-

C*05:(01/09) and HLA-C*18:01 alleles, compared to the amino acid sequence of other HLA-C allele.

HLA-C*

Amino acid position F-pocket

74 77 80 81 84 95 97 114 116 123 133 143 146 147

04:(01/03/06/07), 05:(01/09),18:01

D N K L Y L R N F Y W T K W

02:(02/10) - - - - - - - D S - - - - -

06:02, 16:02 - - - - - - W D S - - - - -

17:01 - - - - - I - - - - - S - L

15:02 - - - - - I - D L - - - - -

15:05 - - - - - I - D - - - - - -

15:06 - - - - - I - D Y - - - - -

01:(02/08) - S N - - - W D Y - - - - -

3:05 - S N - - - S D Y - - - - -

12:03,14:02, 16:(01/04)

- S N - - - W D S - - - - -

14:04 - - N - - - W D S - - - - -

03:02, 12:02 - S N - - - - D S - - - - -

07:(01/02) - S N - - - - D S - - - - L

7:04 - S N - - F - D - - - - - L

7:12 - S N - - F - D - - - - - -

03:(04/02) - S N - - I - D Y - - - - -

1:03 - S N - - - W - - - - - - -

08:(01/02/03) - S N - - - - - - - - - - -

72

Table 3.7: HLA-C, B-pocket key amino acid residues YSAVGEKYQF at the B-pocket positions 7, 9, 24,

34, 41, 45, 63, 66, 67, and 70 respectively for the HLA-C*04:(01/03/06/07), HLA-C*05:(01/09) and HLA-

C*18:01 alleles, compared to the amino acid sequence of other HLA-C alleles.

HLA-C*

Amino acid position B-pocket

7 9 24 34 41 45 63 66 67 70

04:(01/07), 14:(02/04) Y S A V G E K Y Q F

05:(01/09), 03:(03/04),

08:( 01/02/03) - Y - - - - - - - Y

04:(03/06) - Y - - - - - - - -

18:01 - D S - - - - - - -

Virtual modelling

The MHCcluster and HLA class-I binding cleft F- and B- pocket structure analysis

showed that the NVP-SCAR associated HLA-C alleles share the same binding cleft F-

pocket, but different B-pockets. Virtual modelling analysis performed and analysed by

colleagues at the University of Florida, showed that NVP preferentially interacts near

the binding cleft B-pocket of the HLA-C*04:01 molecule and not the F-pocket (Figure

3.5).

Figure 3.5: Virtual modelling of NVP and HLA-C*04:01 potential interaction, which indicates the

preferential docking or binding of NVP near the B-pocket of the HLA-C*04:01 molecule.

73

Multivariate analysis and HLA predictive values

The HLA-C*04:01 allele was significant in NVP-SCAR, when analysing the whole

cohort and remaining significant in all populations (Table 3.8). The MHCcluster pb-

C*18 and pb-C*05 grouped HLA alleles, which share the same binding groove

F-pocket amino acid residues, were also significant in predicting NVP-SCAR, when

analysing the whole cohort (OR 2.73, p value 0.0003) but only remained significant in

Caucasians, when each race was analysed separately (OR 3.35, p value 0.002)

(Table 3.8) (171).

Table 3.8: Multivariate model of predictive and protective NVP-rash associated SCARs HLA alleles,

adjusted for ethnicity.

MULTIVARIATE (+ adjusted for race)

ALL CAUCASIAN ASIAN BLACK

OR P OR P OR P OR P

HLA-C*04:01 4.28 <0.0001 3.00 0.002 8.23 <0.0001 6.38 0.006

pb-C*18 and pb-C*05 group sharing the same F-pocket

2.73 0.0003 3.35 0.002 1.51 0.4 3.27 0.1

Table 3.9 shows the positive and negative predictive values calculated based on

sample prevalence of NVP-SCAR and the conserved estimates of 5% prevalence

cited in the literature (138-141). HLA-C*04:01 allele possessed a positive predictive

value ranging from 4.6% in Taiwanese to 26.7% in Southeast Asians, and a negative

predictive value ranging from 94.7% in Taiwanese to 96.4% Africans, adjusted for a

prevalence of 5% NVP-SCAR (Table 3.9) (138-141). Carriage of a HLA allele that

shares the same key binding cleft F-pocket residues as HLA-C*04:01, possessed a

positive predictive value between 2.7% and 11.0% and a negative predictive value

between 94.5% and 97.4% was determined, depending on ethnicity and adjusted for

5% prevalence of NVP-SCAR (Table 3.9) (138-141).

74

Table 3.9: Positive and negative predictive values for carriage of the identified predicted and/or

protective NVP-SCAR associated HLA class-I alleles, adjusted for 5% prevalence.

Controls NVP- rash

HSR Sample

prevalence 5 % prevalence

Ethnicity Neg (N)

Pos (N)

Neg (N)

Pos (N)

Sensitivity (%)

Specificity (%)

PPV (%)

NVP (%)

PPV (%)

NVP (%)

HLA-C*04:01

African 44 8 11 8 42.1 84.6 50.0 80.0 12.6 96.5 European 115 29 26 16 38.1 79.9 35.6 81.6 9.1 96.1 Hispanic 52 14 19 6 24.0 78.8 30.0 73.2 5.6 95.2 SEA 154 6 40 14 25.9 96.2 70.0 79.4 26.7 96.1 Taiwanese 36 4 10 1 9.1 90.0 20.0 78.3 4.6 95.0

Predictive HLA-C allele

African 38 14 7 12 63.2 73.1 46.2 84.4 11.0 97.4 European 95 49 15 27 64.3 66 35.5 86.4 9.0 97.2 Hispanic 43 23 12 13 52 65.2 36.1 78.2 7.3 96.3 SEA 132 28 34 20 37 82.5 41.7 79.5 10.0 96.1 Taiwanese 33 7 10 1 9.1 82.5 12.5 76.7 2.7 94.5

3.4.2 Diminishing Ex-vivo PBMC’s Responses and NVP-HSR

Immunopathogenesis

3.4.2.1 IFN-γ ELISpot and CBA flow cytometry analysis

Work performed by Niamh Keane et al. 2014 showed that NVP-restricted responses

were detected by IFN-γ ELISpot assay in PBMC’s from the three donors; but these

responses waned and diminished becoming undetectable in PBMC’s shortly after the

first presentation of NVP-SCAR, DRESS phenotype (Figure 3.6) (138). Looking at the

cytokine plasma profiles of the same NVP-DRESS patients, used in Keane et al. 2014,

we expected that the cytokine environment within the donor’s plasma, over the time

course of a NVP-DRESS reaction and recovery period, might show changes in specific

T-cell subpopulations, further explaining the diminishing NVP-stimulated, PBMC IFN-γ

responses. For example an increase in IL-10 during the recovery period might disrupt

the Treg/Th17 balance and add weight to the first hypothesis proposed by Keane et al.

2014, that the diminishing NVP-stimulated, PBMC IFN-γ responses were due to an

increase in CD4+ Tregs levels after NVP withdrawal and during the recovery period.

Keane et al. 2014 proposed this hypothesis, based on Figure 3.7 which shows the

percentage of Tregs 8 days after NVP cessation, which matched the non-HSR control

Treg percentages and NVP IFN-γ responses were still detectable in Donor 1 (NVP-

75

HSR). As the IFN-γ response diminished, the percentage of Tregs became elevated

(Figure 3.6 and Figure 3.7) (138).

Figure 3.6 Timeline of diminishing IFN-γ responses SFU/million cells in Donor 1 (NVP-HSR), Donor 2

(NVP-HSR), and Donor 3 (NVP-HSR) Figure from Keane et al. 2014 (138)

Figure 3.7: Regulatory T cells expressed at a % of CD4 T cells are shown in the plot where clear

symbols indicates samples from Donor 1 (NVP-HSR) and the time after NVP cessation. The filled in

symbols indicate samples from controls (black, blue) and HIV patient not on medication (red symbol).

Figure from Keane et al. 2014 (138).

76

Donor 1 (NVP-HSR) showed a reduced IFN-γ ELISpot response in NVP-exposed

PBMC’s acquired 18 days post NVP-DRESS. Donor 1 (NVP-HSR) showed an IFN-γ

ELISpot response to NVP in the order of 1600 SFU/million cells, in PBMC’s acquired 5

post NVP-DRESS which had reduced to 150 SFU/million cells in PBMC’s acquired day

8 post NVP-DRESS (138). PBMC’s from Donor 1 (NVP-HSR) acquired 18 days post

NVP-DRESS, showed no IFN-γ ELISpot response (138). The cytokine plasma profile

of Donor 1 (NVP-HSR) also showed detectable IFN-γ release at day 1 post NVP-

DRESS which then decreased. The cytokine IL-6 was detected at day 1 and day 5 in

the plasma of Donor 1 (NVP-HSR) which decreased to become undetectable by day 8

post NVP-DRESS (Figure 3.8) (138).

Figure 3.8: Cytokine plasma profile of Donor 1 (NVP-HSR), plasma obtained pre-NVP-HSR to day 102

post NVP-HSR. Concentrations were determined by using the CBA kit internal standard controls and

normalised against the NVP-tolerant and NVP-naive healthy controls. Statistical significance was

determined using an analysis of variance (ANOVA) statistical analysis (p < 0.0001) and multiple

comparison tests (***p < 0.0001) to compare the concentration of each cytokine in Donor 1 (NVP-HSR)

to the concentration of the responding cytokine in the NVP-tolerant and NVP-naive healthy controls.

77

Donor 2 (NVP-HSR) showed an IFN-γ ELISpot response in the order of 300

SFU/million cells in NVP-exposed PBMC’s acquired 2 days post NVP-DRESS.

PBMC’s acquired 62 days post NVP-DRESS showed no detectable IFN-γ ELISpot

response (138). The cytokine plasma profile of Donor 2 (NVP-HSR) also showed

significant IL-6 release at day 1 post NVP-HSR which increased day 2 post NVP-

DRESS (Figure 3.9).

Figure 3.9: Cytokine plasma profile of Donor 2 (NVP-HSR) plasma obtained pre-NVP-DRESS to day

330 post NVP-DRESS. Concentrations were determined by using the CBA kit internal standard controls

and normalised against the NVP-tolerant and NVP-naive healthy controls. Statistical significance was

determined using an analysis of variance (ANOVA) statistical analysis (p < 0.0001) and multiple

comparison tests (***p < 0.0001) to compare the concentration of each cytokine in Donor 2 (NVP-HSR)

to the concentration of the responding cytokine in the NVP-tolerant and NVP-naive healthy controls.

78

Donor 3 (NVP-HSR), showed an IFN-γ ELISpot response in the order of 400

SFU//million cells in PBMC’s collected 97 days post NVP-DRESS induced and by day

476 post NVP-DRESS, no IFN-γ ELISpot response was detected. The cytokine

plasma profile of Donor 3 (NVP-HSR) showed no significant release of the cytokines

being analysed (Figure 3.10).

Figure 3.10: Cytokine plasma profile of Donor 3 (NVP-HSR) plasma obtained day 7 to day 727 post

NVP-HSR. Concentrations were determined by using the CBA kit internal standard controls and

normalised against the NVP-tolerant and NVP-naive healthy controls.

79

Analysis of the IL-6 plasma profile over the time course of a NVP-DRESS is shown in

Figure 3.11. Detectable levels of IL-6 in the plasma of the Donors at day 1 decreased

to insignificant levels in the Donors' plasma by day 12, post NVP cessation.

Figure 3.11: Plasma IL-6 cytokine profile of Donor 1 (NVP-HSR) and Donor 2 (NVP-HSR) (Donor 3

(NVP-HSR) not included), highlighting the potential of IL-6 having a role in NVP-HSR. Concentrations

were determined by using the CBA kit internal standard controls and normalised against the NVP-

tolerant and NVP-naive healthy controls. Statistical significance was determined using an analysis of

variance (ANOVA) statistical analysis (p < 0.0001) and multiple comparison tests to compare the

concentration of each cytokine to their respective pre-NVP plasma cytokine concentration

(***p < 0.0001).

80

3.5 Discussion


Characteristics

3.5.1.1 NVP- SCAR with hepatitis

A larger proportion of NVP-HSR involve the liver, with 80% of DRESS cases

presenting with clinical hepatitis (67). The validated HLA CREG supertype B*07, which

consists of a group of HLA alleles that share similarities in cross-reactivity of public

epitopes and binding cleft pocket chemistry and the MHCcluster pb-B*78 grouped

alleles, which share similar peptide-binding specificity, as predicted by NetMHCpan

2.8, showed a significant predictive association with NVP-HSR with hepatitis. After the

MHCcluster and binding cleft B-pocket structural analysis of the pb-B*78 group, the

HLA-B*55:01 and HLA-B*56:01 alleles both belonging to the CREG B*07 and pb-B*78

groups were found to be significant

Donor 2 (NVP-HSR) from the cytokine profile analysis, also possessed the

HLA-B*56:01 along with the HLA-C*04:03 allele also significant in NVP-SCAR. Donor

2 (NVP-HSR), experienced a fever, rash, eosinophilia and hepatitis, typical of DRESS,

on day 7 of a HAART regime containing NVP. Further analysis of the HLA-B*55:01

and HLA-B*56:01 alleles is required and future studies should include virtual modelling

to highlight the NVP-binding characteristics of HLA-B*55:01 and HLA-B*56:01.

Following that, setting up a series of peptide elution experiments and analysing the

peptide sequence data to show the binding characteristics of peptides would further

our understand of any potential interaction between the drug and how it influences the

peptides being presented. This work will lead to designing a working model for how

NVP generates neo-antigens, which can then be tested ex-vivo.

3.5.1.2 NVP- SCAR

The HLA-C*04:01 was shown to have a significant predictive association to NVP-

SCAR across all the ethnic groups, despite the variations in allele frequency. The

unique F-pocket amino acid sequence possessed by HLA-C*04:01 is also shared by

HLA-C*05:01 and HLA-C*18:01 shown to be predictive in Caucasian and Africans,

respectively (Table 3.5). The HLA-C*04:01, -C*05:01 and -C*18:01 alleles, and other

common HLA-C alleles identified within the cohort were analysed and assigned to

specific peptide-binding groups, based on the binding specificities of the MHC Cluster

algorithm to see if they shared are similarities in their peptide-binding cleft structure

81

and/or peptide-binding capabilities which could then be translated back to increased

risk of NVP-rash associated SCAR.

The MHCcluster analysis the HLA-C*04:01, HLA-C*05:01, HLA-C*18:01, and other

related alleles prevalent in the cohort showed these alleles all preferred to bind 9mer

peptides with an Aspartic acid (D) at position 2 of the peptide (from the NH2 terminal

end) which interacts with binding cleft B-pocket (Figure 3.3 and Figure 3.4). Further

analysis of the MHCcluster assigned pb-C*18 and pb-C*05 groups, showed the alleles

within both groups shared the binding groove F-pocket residues (Table 3.6). A

multivariate analysis of the cohort showed the alleles which have a shared binding

cleft F-pocket were predictive in NVP-SCAR, but, when each individual race was

analysed, the F-pocket sequence remained significant only in Caucasians (Table 3.8).

While these pb-C*18 and pb-C*05 group alleles shared the same binding cleft

F-pocket, their binding cleft B-pocket varied. For example the pb-C*18 alleles HLA-

C*04:(01/07) share the same key B-pocket residues as HLA-C*14:(02/04) and the pb-

C*05 alleles HLA-C*05:(01/09) share the same key B-pocket residues as HLA-

C*03:(03/04) and HLA-C*08:(01/02/03) (Table 3.5). The shared binding cleft F-pocket

and variation in binding cleft B-pocket among these NVP-SCAR associated HLA-C

alleles would suggest the importance of the binding cleft F-pocket, in terms of NVP-

SCAR immunopathogenesis and the HLA-drug-TCR interaction. HLA-C*04:01 and the

other predictive HLA-B and HLA-C alleles should be used in future studies to explore

NVP-SCAR immunopathogenesis.

Virtual modelling analysis performed by colleagues at the University of Florida show a

preferential interaction between NVP and a region close to the B-pocket, and not the

F-pocket as one might have predicted based on shared binding cleft F-pocket (83-85).

Before proposing a working model for ex-vivo based analysis, more research into

peptide-binding within this HLA and how the drug may interact is needed. One method

includes setting up a series of peptide elution experiments and analysing the peptide

sequence data to find the binding characteristics of peptides. This work will lead to

designing a working model for how NVP generates neo-antigens, which can then be

tested ex-vivo.


Immunopathogenesis

This smaller study was part of a larger study involving NVP-SCAR

immunopathogenesis (138). While the first section of this chapter looked at the binding

82

characteristics of HLA class-I alleles, the second half looked to find the immunological

mechanism driving the NVP-DRESS by analysing the cytokine environment within the

donor’s plasma, over the time course of a NVP-DRESS reaction and recovery period.

We expected this might show changes in specific T-cell subpopulations, further

explaining the diminishing NVP-stimulated, PBMC IFN-γ responses supporting the

data presented in the paper by Keane et al. 2014 (138, 173).

As highlighted in Keane et al. 2014, Donor 1 (NVP-HSR) exhibited an increase in

Tregs cell numbers corresponded with diminishing IFN-γ ELISpot responses (138).

After this first analysis, PBMC’s samples were limiting. Therefore to discuss this

question and offer more data to complement the data published by Keane et al. 2014,

we assessed donor plasma over the time course of NVP-DRESS reaction to see if the

cytokine environment reflects the changes in specific T-cell subpopulations. If the

cytokine environment were to support the Tregs cell numbers corresponded with

diminishing IFN-γ ELISpot responses, we suspected to see an increase in cytokines

like IL-10, which can show changes in the IL-17/Treg balance. Several cytokines were

measured in patients over the time course of a NVP-DRESS reaction. Furthermore,

two of the Donors carried HLA allotypes that were identified as significant NVP-SCAR

predictors or potential predictors in Section 3.4 Results. Donor 2 (NVP-HSR) carried

both HLA-B*56:01 and HLA-C*04:03, and Donor 3 (NVP-HSR), HLA-C*04. Donor 1

(NVP-HSR) however, carried the previously identified HLA-B*14 / HLA-C*08 haplotype

alleles (146, 147)

Donor 1 (NVP-HSR) showed a detectable IFN-γ release at day 1 post

NVP-DRESS, which then decreased. Both Donor 1 (NVP-HSR) and Donor 2 (NVP-

HSR) both showed a significant IL-6 release within the first 8 days post NVP-DRESS.

Elevated IL-6 levels in plasma samples can be linked to numerous different

pathologies (196-198).

IL-6 is a proinflammatory and the elevated levels could result from the immune

reaction which is not particularly significant to NVP-DRESS and commonly

concentration of IL-6 are typically below 1pg/ml due to its short plasma half-life and

corticosteroids down regulate IL-6 expression (199). Elevated IL-6 plasma levels have

also been shown to proceed HHV-6 reactivation and has been associated with higher

HHV-6 viral load and progression to encephalitis and systemic inflammation in stem

cell transplantation patients who experience HHV-6 infection-related complications

(200, 201) This association to HHV-6 reactivation is significant as it has been

83

associated with some IM-ADR and has been linked to prolonged disease and severe

organ involvement, fever and, cutaneous symptoms plus hepatosplenomegaly, and

severe lymphocytopenia. The reactivation of HHV’s like HHV-6 and corresponding

virus-specific Tregs in patients with DRESS has been reported (202). The Donors in

this study showed no specific evidence of HSV-1, HSV-2, HHV-3 (VZV), CMV, EBV or

HHV-6 reactivation. However, all patients possessed IgG antibodies for HHV-6,

including Donor 2 (NVP-HSR), an infant at the time of NVP-DRESS.

The accelerated production of IL-6, especially by cytotoxic T cells in MS patients helps

these cytotoxic T cells to become more resistant to Tregs and is potent modifier of

early Tregs and cytotoxic T-cell interactions and communication (203, 204). Other

studies have also highlighted the importance of IL-6 and the balance between Tregs /

Th17 cells (208, 210). Given this increase in IL-6 and other studies that suggest

HHV-6 reactivation in DRESS, in particular before reactivation, and the importance of

IL-6 in Tregs/Th17 balance, IL-6 could be a cytokine important to NVP-DRESS

alongside other key factors like carriage specific HLA class-I alleles. One model we

propose is that the decrease in Tregs at the time of the reaction might be

driving NVP-DRESS and the return of Tregs during recovery, responsible for the NVP-

stimulated diminishing IFN-γ PBMC responses.

To further analyse this proposed model, future analysis should look at the expression

and quantification of inhibitory immune receptors; CTLA4, (cytotoxic T-lymphocyte-

associated protein 4 or CD152), PD-1 (Programmed Death 1), and/or PD-L1

(Programmed Death Ligand 1) (205-208). Ex-vivo experiments to assess this could

include monoclonal antibodies directed towards blocking these inhibitory immune

receptors, to see if the ex-vivo response can be recovered (205, 207, 208). While it

would have been preferential to perform this on the specific samples used in Keane et

al. 2014 and within this thesis, this could not be performed as samples were limited

and were used in the first study by Niamh Keane et al. 2014, hence the reason we

analysed the Donors' plasma cytokine environment.

84

85

Chapter 4

Carbamazepine-Induced Stevens - Johnson Syndrome

and Toxic Epidermal Necrolysis:

Predicted HLA-Restricted Peptide Screening

86

87

4 4.1 Introduction

Chapter 1 (Section 1.3.2 Carbamazepine) highlighted the strong association between

HLA-B*15:02 and other B75 serotypes, including the HLA-B*15:08, -B*15:11, and -

B*15:21 alleles and CBZ-SJS/TEN (2, 209, 210). Based on a strong negative

predictive values associated with HLA-B*15:02 carriage (100% in Han Chinese), the

Food and Drug Administration (USA) and similar regulatory agencies of different

countries have recommended that genetic screening for HLA-B*15:02 for those of

South-east Asian ancestry, before CBZ therapy and that all HLA-B*15:02 carriers,

should not be administered CBZ, unless the benefits outweigh the risks (101).

However, the low/incomplete positive predictive value of HLA-B*15:02 carriage (3.3-

7% in Han Chinese) illustrates how carriage is necessary but not enough to explain

CBZ-SJS/TEN pathogenesis and other factors must also play a role in driving CBZ-

SJS/TEN.

As explained in Chapter 1 (Section 1.4 Established Models of HLA-driven delayed

Immunologically Mediated-Adverse Drug Reactions), a considerable amount of

evidence supports the heterologous immune model as a mediator of organ transplant

rejection (9, 10, 12, 16-19). In organ transplantation, an overabundance of neo-

antigens are presented to the hosts immune system, which can be cross-recognised

by pre-existing class-I memory T cells, created in response to prevalent viruses, like

those of the Herpesviridae family. (54-60) Similar to organ transplantation, IM-ADR

presents a novel situation, in which neo-antigens can be created by the off-target drug-

HLA interaction, which are then cross-recognised by epitope-specific class-I memory T

cells. Epitope-specific class-I memory T cells might determine CBZ-SJS/TEN

development in individuals who carry the at-risk HLA allele. Increasing the ability to

predict HLA-B*15:02-CBZ-SJS/TEN and further our understanding of the

immunopathogenic mechanisms driving these reactions.

88


Hypothesis

A proportion of the HSV-1 and/or HSV-2 HLA-B*15:02-restricted peptides can induce a

specific HLA-B*15:02-epitope memory T-cell response shared in patients who

experience a CBZ-SJS/TEN reaction.

Aim

To identify several in-silico predicted HLA-B*15:02-restricted HSV-1/HSV-2 peptides

which may elicit responses in the CBZ-SJS/TEN patient’s ex-vivo. Once identified, the

aim is to isolate and expand the HSV-1/HSV-2 specific memory T cells in-vitro for

future analysis of the cross-reactivity model.

4.3 Methods

4.3.1 Samples and Controls

The Controls and Donors used in this chapter, with a complete description, are listed

in Table 2.2 Chapter 2 Materials and Methods, Section 2.1.2 Carbamazepine -

Induced Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis.

4.3.2 In-Silico HLA-B*15:02-restricted Peptide Predictions

Several HLA-B*15:02-restricted HSV-1 (strain 17) and HSV-2 (strain HG52) 9-11mer

peptides were predicted by Bioinformatics colleagues at The Institute for Immunology

and Infectious Diseases, Murdoch University, Perth, Australia using in-silico epitope

mapping methods which involved Net-MHC3.2 based algorithms, combined with a

novel measure of “targeting efficiency” developed in collaboration with Microsoft

research (Table 4.3) (169, 170, 188-190) Based on personal communications, 42

predicted HLA-B*15:02-restricted peptides were synthesised for ex-vivo analysis,

highlighted Table 4.1 and Table 4.2 (169, 170). Listed in Table 4.2, are variants of the

HSV-1/HSV-2 shared peptide, YAHGHSLRLPY and a control peptide, which were also

synthesised and analysed as the results dictated (211).

89

Table 4.1: In-silico predicted HSV-1 and HSV-2, HLA-B*15:02-restricted peptides which were

synthesised and analysed ex-vivo (169, 170, 188, 189).

HSV-1 PROTEIN SEQUENCE ID

POSITION (Amino acid)

FAFTINIAAY DNA packaging tegument protein UL17 ref|NP_044618.1 667-676

FTFGGGYVY envelope glycoprotein B UL27 ref|NP_044629.1 641-649

FVYTPSPY tegument protein UL21 ref|NP_044622.1 161-168

HMRTLSAF tegument protein UL37 ref|NP_044639.1 773-780

ISYSSGAIAAF capsid portal protein UL6 ref|NP_044607.1 555-565

LAFVLDSPSAY envelope glycoprotein H UL22 ref|NP_044623.1 476-486

LAYNGSAY envelope glycoprotein I US7 ref|NP_044669.1 172-179

MAIAAEGTLAY envelope glycoprotein I US7 ref|NP_044669.1 164-174

MLRRRGEAY large tegument protein UL36 ref|NP_044638.2 1926-1934

NMYAKLSGAF helicase-primase helicase subunit UL5 ref|NP_044606.1 128-137

SMVPTGQAF helicase-primase primase subunit UL52 ref|NP_044655.1 539-547

YANAASGAF envelope protein UL43 ref|NP_044645.2 59-67

YTYARVFAF helicase-primase helicase subunit UL5 ref|NP_044606.1 731-739

FPAAAGSVAY capsid scaffold protein UL26.5 / capsid maturation protease UL26

ref|NP_044628.1 ref|NP_044627.1

48-57 354-363

HMRARTSAPY capsid scaffold protein UL26.5 / capsid maturation protease UL26

ref|NP_044628.1 ref|NP_044627.1

188-197 494-503

HSV-2 PROTEIN SEQUENCE ID


AMFAGLCAF envelope glycoprotein B UL27 ref|NP_044497.1 727-735

FAFTINTAAY DNA packaging tegument protein UL17 ref|NP_044486.1 666-675

FTINTAAY DNA packaging tegument protein UL17 ref|NP_044486.1 668-675

FAREYTTLAF helicase-primase subunit UL8 ref|NP_044477.1 208-217

FAYTPSPY tegument protein UL21 ref|NP_044490.1 161-168

FLFYPDPTY nuclear egress lamina protein UL31 ref|NP_044501.1 201-209

FLVAACAAAY capsid triplex subunit 1 UL38 ref|NP_044508.1 189-198

FLWDRHAQRAY envelope glycoprotein L UL1 ref|NP_044470.1 82-92

FQHQYIPAY capsid portal protein UL6 ref|NP_044475.1 509-517

LAFVLDSPAAY Envelope glycoprotein H UL22 ref|NP_044491.1 476-486

FVLDSPAAY Envelope glycoprotein H UL22 ref|NP_044491.1 478-486

LLASSGFAF envelope glycoprotein H UL22 ref|NP_044491.1 385-393

MAFRASGPAY membrane protein UL45 ref|NP_044515.1 1-10

MARGAGLVF envelope glycoprotein E US8 ref|NP_044538.1 1-9

YAQYMSRAY envelope glycoprotein H UL22 ref|NP_044491.1 352-360

SMFAYVKPM DNA replication origin-binding helicase UL9 ref|NP_044478.1 380-388

DTAFAFNPY large tegument protein UL36 ref|NP_044506.1 1388-1396

TAFAFNPY large tegument protein UL36 ref|NP_044506.1 1389-1396

HMRARTHAPY capsid scaffold protein UL26.5 / capsid maturation protease UL26

ref|NP_044496.1 ref|NP_044495.1

188-197 496-505

HSV Herpes simplex virus; HLA-B, Human leukocyte antigen B.

90

Table 4.2: In-silico predicted HSV-1 and HSV-2 shared, HLA-B*15:02-restricted peptides and variants of

the HSV-1 and HSV-2 shared peptide, YAHGHSLRLPY which were synthesised and analysed, ex-vivo

(169, 170, 188, 189).

HSV-1/2 shared

PROTEIN SEQUENCE ID


ELFPGHPVY large tegument protein HSV 1 UL36 ref|NP_044638.2 2488-2496

large tegument protein HSV 2 UL36 ref|NP_044506.1 2486-2494

EQFSLSSY envelope protein HSV 1 UL20 ref|NP_044621.1 31-38

envelope protein HSV 2 UL20 ref|NP_044489.1 31-38

FAASFAAIAY ribonucleotide reductase subunit 2 HSV 1 UL40 ref|NP_044642.1 191-200

ribonucleotide reductase subunit 2 HSV 2 UL40 ref|NP_044510.1 188-197

FLMHAPAF envelope glycoprotein D HSV 1 US6 ref|NP_044668.1 176-183

envelope glycoprotein D HSV 2 US6 ref|NP_044536.1 176-183

YMAKLHAY helicase-primase helicase subunit HSV 1 UL5 ref|NP_044606.1 404-411

helicase-primase helicase subunit HSV 2 UL5 ref|NP_044474.1 403-410

YMALVSAM envelope glycoprotein B HSV 1 UL27 ref|NP_044629.1 849-856

envelope glycoprotein B HSV 2 UL27 ref|NP_044497.1 849-856

YTFKTYQVF DNA packaging protein HSV 1 UL32 ref|NP_044634.1 552-559

DNA packaging protein HSV 2 UL32 ref|NP_044502.1 555-562

YAHGHSLRLPY helicase-primase primase subunit HSV 1 UL52 ref|NP_044655.1 811-821

helicase-primase primase subunit HSV 2 UL52 ref|NP_044523.1 823-833

HSV-1/2 shared PROTEIN SEQUENCE ID


AHGHSLRLPY Variant of YAHGHSLRLPY UL52

HGHSLRLPY Variant of YAHGHSLRLPY UL52 ref|NP_044655.1 811-821

YA_G_SLRLPY Variant of YAHGHSLRLPY UL52 and

YAHG_SLRLPY Variant of YAHGHSLRLPY UL52 ref|NP_044523.1 823-833

YA_GHSLRLPY Variant of YAHGHSLRLPY UL52

VZV (HHV3)

YSHGHSLRLPF helicase-primase primase subunit ORF6 ref|NP_040129.1 855-865

M. tuberculosis

NIRQAGVQY ESAT-6-like protein EsxB (M. tuberculosis)

ref|WP_009937166.1 72-80

HSV Herpes simplex virus; HLA-B, Human leukocyte antigen B.

4.3.3 Ex-vivo analysis of in-silico derived

HLA-B*15:02-restricted peptides.

The IFN-γ ELISpot assays and tetramer based flow cytometry methods, outlined in

Chapter 2, Section 2.2. Cellular assays were used to assess the

in-silico predicted HLA-B*15:02-restricted HSV-1/HSV-2 peptides, ex-vivo. Figure 4.1

includes an example of the raw ELISpot results from which spot forming unit/per one

million cells (SFU/million cells) were derived. SFU/million cells were calculated based

on the original cell number input and the spots counted, minus the background

response determined by the negative or solvent control SFU/million result. The mean

SFU/million cells (+/- Standard Deviation (SD)) were calculated with a minimum of

91

three replications per peptide. De novo sequencing by mass spectrum analysis was

performed by Proteomics International (Perth, Western Australia), as the research

dictated. Statistical analysis and graphical representations were calculated using

GraphPad Prism software Version 5.02 (GraphPad Software Inc., California, USA).

Table 4.3: Distribution of the work within Chapter 4 and accreditation to those who contributed.


In-silico predictions of

HSV-1/HSV-2 HLA-B*15:02- restricted peptides

Bioinformatics staff at IIID

Patient clinical analysis Elizabeth Phillips

Peptide de novo Sequencing and Mass

spectrometry analysis

Proteomics International

Ex-vivo analysis of predicted HSV-1/HSV-2 HLA-

B*15:02- restricted peptides

Craig Rive

HSV-1, HSV-2; herpes simplex virus type 1 and 2, HLA, human leukocyte antigen

Figure 4.1: Example IFN-γ ELISpot raw data. The unstimulated background and solvent controls

(DMSO/DMF at 1μL/mL), positive controls anti-CD3 antibody (wells B1 and B2 0.1μg/mL) and SEB

(wells B3 and B4 1μg/mL) responses at 200,000 cells per well. The test positive and test negative

responses highlight the typical positive and negative response to the test stimulant (drug or peptide

10μg/mL) with an input of 50,000 and 200,000 cell per well.

92

4.4 Results

4.4.1 Ex-vivo analysis of in-silico derived

HLA-B*15:02-restricted peptides.

4.4.1.1 IFN-γ ELISpot.

Figure 4.2 to Figure 4.7 shows the predicted HLA-15:02-restricted peptides which

elicited immune responses in CBZ-SJS/TEN Donor, and CBZ-naive control PBMC

samples. Figure 4.8 emphasises the key predicted HLA-15:02-restricted peptides

which elicited immune responses in more than one of the CBZ-SJS/TEN Donors.

These include the YAHGHSLRLPY HSV-1/2 shared peptide (Table 4.2) and the

LAFVLDSPSAY HSV-1 peptide (Table 4.1), compared to previously identified HLA-

B*15:(01:02)-restricted M. tuberculosis peptide, NIRQAGVQY initially used as a

control and the FAFTINITY HSV-2 peptide which only elicited an immune response in

Donor 2 (SJS) (211). Given the HSV-1 positive serological status for Control 1 (CBZ-

naive) (Table 2.2), the HSV-1 and shared HSV-1/HSV-2 predicted HLA-15:02-

restricted peptides could also elicit immune responses as seen in the shared

ELFGHPVY HSV-1/2 epitope response. No responses were predicted to occur in both

Donor 4 (CBZ-SJS non-B*15:02) (Figure 4.5) and Control 2 (CBZ-naive non-B*15:02)

(Figure 4.7), unless the predicted HLA-B*15:02-restricted peptides exhibited HLA

class-I allele promiscuity (212, 213). Figure 4.5 shows a response to the already

mentioned LAFVLDSPSAY HSV-1 peptide in Donor 4 (CBZ-SJS non-B*15:02). The

cut-off for positive response was determined to be 50 SFU/million cells.

93

Figure 4.2 The Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 1 (CBZ-TEN) were

determined using ELISpot, described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3. Cut-off for

positive responses, 50 SFU/million cells (red line). Error bars represent intra-sample variation (n=3).

94

Figure 4.3: The Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 2 (CBZ-SJS) were



95

Figure 4.4: The Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 3 (CBZ-SJS) were



96

Figure 4.5: Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 4 (CBZ-SJS non- B*15:02) were



97

Figure 4.6: Mean (+/- SD) IFN-γ responses SFU/million cells in Control 1 (CBZ-naive) were determined

using ELISpot, described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3. Cut-off for positive

responses, 50 SFU/million cells (red line). Error bars represent intra-sample variation (n=3).

98

Figure 4.7: Mean (+/- SD) IFN-γ responses SFU/million cells in Control 2 (CBZ-naive non-B*15:02) were



99

Figure 4.8: Mean (+/- SD) IFN-γ responses SFU/million cells to CBZ and predicted HLA-B*15:02-

restricted peptides identified in Figure 4.2 to Figure 4.7, determined using ELISpot, described in

Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3. Cut-off for positive responses, 50 SFU/million cells (red

line). Error bars represent intra-sample variation (n=3).

4.4.1.2 IFN-γ ELISpot dose-dependent responses

The affinity of cellular responses to decreasing peptide concentrations (10, 1, and

0.1 μg/mL) was assessed through the IFN-γ ELISpot assay. The cellular immune

responses to the of the predicted HSV-1/2 peptide YAHGHSLRLPY and the predicted

HSV-1 peptide LAFVLDSPSAY showed a significant decrease in SFU/million cells in

relationship to the concentration of the peptide in all three CBZ-SJS/TEN, HLA-

B*15:02 positive Donors (Figure 4.9 and Figure 4.10). The TB peptide NIRQAGVQY

100

show a decrease in SFU/million cells in Donor 1 (CBZ-TEN) and Donor 2 (CBZ-SJS);

but in Donor 3 (CBZ-SJS) no statistical significance was determined (Figure 4.11). No

significant decrease was seen in SFU/million cells in relationship to the HSV-2 peptide

FAFTINTAAY concentration in the CBZ-SJS/TEN, HLA-B*15:02 positive Donors

(Figure 4.12).

Figure 4.9: Mean (+/- SD) IFN-γ responses SFU/million cells to the increasing concentrations of the

predicted HLA-B*15:02-restricted peptide, YAHGHSLRLPY in the three HLA-B*15:02 positive Donors

were determined using ELISpot, described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3 with a

concentrations of 0.1 μg/ml, 1 μg/ml, and 10 μg/ml. Error bars represent intra-sample variation (n=3)

and the p values were determined using the Pearson correlation test (*p < 0.05).

101


predicted HLA-B*15:02-restricted peptide, LAFVLDSPSAY in the three HLA-B*15:02 positive Donors



and the p values were determined using the Pearson correlation test (*p < 0.05).

102


predicted HLA-B*15:02-restricted peptide, NIRQAGVQY in the three HLA-B*15:02 positive Donors



and the p values were determined using the Pearson correlation test (*p < 0.05)

103

Figure 4.12 Mean (+/- SD) IFN-γ responses SFU/million cells to the increasing concentrations of the

predicted HLA-B*15:02-restricted peptide, FAFTINTAAY in the three HLA-B*15:02 positive Donors


concentrations of 0.1 μg/ml, 1 μg/ml, and 10 μg/ml. Error bars represent intra-sample variation (n=3).

104

4.4.1.3 Variant YAHG_SLRLPY peptide analysis

Purified versions of the responsive peptides were used to make B*15:02 tetramers.

Tetramer grade peptides were tested ex-vivo to confirm the newly synthesised peptide

before tetramer folding and staining. The tetramer grade HSV-1/2-restricted

YAHGHSLRLPY peptide, failed to replicate the responses seen in the original IFN-γ

ELISpot analysis. The original and tetramer folding YAHGHSLRLPY peptides were

analysed independently by Proteomics International and through personal

correspondence, it was determined that the original YAHGHSLRLPY peptide product

contained several potential variants which may be responsible for the original

response (Figure 4.15). The MS analysis highlighted the original YAHGHSLRLPY

product contained variants of the original peptide in which one or both the Histidine (H)

residues were missing, as highlighted by the peaks with a mass of approximately 1176

(>60% intensity) and 1040 (>40% intensity).

Figure 4.13 MS/MS analysis of YAHGHSLRLPY, original responsive peptide (left) and pure form for

tetramer folding (right). A Spectrum was obtained in de novo peptide analysis via the MALDI TOF MS

(4800 Proteomics analyser).

105

A series of variant were synthesised following the same IFN-γ ELISpot analysis used

above, it was determined the variant peptide YAHG_SLRLPY missing the 5th amino

acid, (Histidine) was responsible for the original responses.

The IFN-γ ELISpot screening of the predicted HLA-B*15:02-restricted HSV-1/2 shared

peptide, YAHGHSLRLPY and the missing histidine variants, YAHG_SLRLPY,

YA_GHSLRLPY, and YA_G_SLRLPY in the Donor and Control PBMCs (Figure 4.14)

highlighted the significance of the YAHG_SLRLPY variant. Figure 4.21 shows affinity

of T cell responses to decreasing ex-vivo concentrations the variant YAHG_SLRLPY

peptide. Following this, tetramers were made for the YAHG_SLRLPY variant,

LAFVLDSPSAY and NIRQAGVQY as outlined in Chapter 2, Materials and Methods,

Section 2.2.2.3 Tetramers and Flow cytometry flow cytometry assays were performed

as outlined in Chapter 2 Materials and Methods, Section 2.2.2 Flow Cytometry

106

Figure 4.14 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 1 (CBZ-TEN) to the

YAHGHSLRLPY peptide and the missing histidine variants, YAHG_SLRLPY, YA_GHSLRLPY, and

YA_G_SLRLPY determined using ELISpot, described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3.

Cut-off for positive responses, 50 SFU/million cells (red line). Error bars represent intra-sample variation

(n=3).

107

Figure 4.15 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 2 (CBZ-SJS) to the




(n=3).

108

Figure 4.16 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 3 (CBZ-SJS) to the




(n=3).

109

Figure 4.17 Mean (+/- SD) IFN-γ responses SFU/million cells in Donor 4 (CBZ-SJS

non- B*15:02) to the YAHGHSLRLPY peptide and the missing histidine variants, YAHG_SLRLPY,

YA_GHSLRLPY, and YA_G_SLRLPY determined using ELISpot, described in Chapter 2, Enzyme-

Linked ImmunoSpot 2.2.3. Cut-off for positive responses, 50 SFU/million cells (red line). Error bars

represent intra-sample variation (n=3).

110

Figure 4.18 Mean (+/- SD) IFN-γ responses SFU/million cells in Control 1 (CBZ-naive) to the




(n=3).

111

Figure 4.19 Mean (+/- SD) IFN-γ responses SFU/million cells in Control 2 (CBZ-naive

non- B*15:02) to the YAHGHSLRLPY peptide and the missing histidine variants, YAHG_SLRLPY,

YA_GHSLRLPY, and YA_G_SLRLPY determined using ELISpot, described in Chapter 2, Enzyme-

Linked ImmunoSpot 2.2.3. Cut-off for positive responses, 50 SFU/million cells (red line). Error bars

represent intra-sample variation (n=3).

112

4.4.1.4 YAHG_SLRLPY dose-dependent IFN-γ ELISpot

responses

The affinity of cellular responses to decreasing variant peptide YAHG_SLRLPY

concentrations (10, 1, and 0.1 μg/mL) was assessed through the IFN-γ ELISpot assay.

The cellular immune responses to the of the predicted HSV-1/2 variant peptide

YAHG_SLRLPY showed a significant decrease in SFU/million cells in relationship to

the concentration of the peptide in all three CBZ-SJS/TEN, HLA-B*15:02 positive

Donors (Figure 4.25), but not in Donor 4 (CBZ-SJS non-B*15:02).

Figure 4.20 Mean (+/- SD) IFN-γ responses SFU/million cells to the increasing concentrations of the

predicted HLA-B*15:02-restricted variant peptide, YAHG_SLRLPY in the three HLA-B*15:02 positive

CBZ-SJS/TEN Donors and Donor 4 (CBZ-SJS non-B*15:02) were determined using ELISpot,

described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3 with a concentrations of 0.1 μg/ml, 1 μg/ml,

and 10 μg/ml. Error bars represent intra-sample variation (n=3) and the p values were determined using

the Pearson correlation test (*p < 0.05) and r2 values non-linear regression analysis (n=4)

113

4.4.1.5 Predicted peptide tetramer staining

B*15:02 tetramers where constructed using the variant YAHG_SLRLPY

HSV-1/2 shared peptide, the LAFVLDSPSAY HSV-1 peptide, and the NIRQAGVQY

M. tuberculosis peptide, to confirm these identified peptides are

HLA-B*15:02-restricted and recognised by CD8+ T cells. The tetramer responses show

that all three peptides are HLA-B*15:02-restricted epitopes, that are recognised by

CD8+ T cells (Figure 4.21 to Figure 4.25).

Figure 4.21: Donor 1 CBZ-TEN, B*15:02-YAHG_SLRLPY, LAFVLDSPSAY, and NIRQAGVQY tetramer

stain, performed as outlined in Chapter 2, Tetramers 2.2.2.3. Samples were compared to negative

control which consisted of unstained sample spiked with PE and washed out as per the Tetramer

staining protocol. Error bars represent intra-sample variation (n=5).

114

Figure 4.22: Donor 2 (CBZ-SJS), B*15:02-YAHG_SLRLPY, LAFVLDSPSAY, and NIRQAGVQY

tetramer stain performed as outlined in Chapter 2, Tetramers 2.2.2.3. Samples were compared to

negative control which consisted of unstained sample spiked with PE and washed out as per the

Tetramer staining protocol. Error bars represent intra-sample variation (n=5).

Figure 4.23: Donor 3 (CBZ-SJS), B*15:02-YAHG_SLRLPY, LAFVLDSPSAY, and NIRQAGVQY

tetramer stain performed as outlined in Chapter 2, Tetramers 2.2.2.3. Samples were compared to

negative control which consisted of unstained sample spiked with PE and washed out as per the

Tetramer staining protocol. Error bars represent intra-sample variation (n=5).

115

Figure 4.24: Control 2 (CBZ-naive non-B*15:02), B*15:02-YAHG_SLRLPY, LAFVLDSPSAY, and

NIRQAGVQY tetramer stain performed as outlined in Chapter 2, Tetramers 2.2.2.3. Samples were

compared to negative control which consisted of unstained sample spiked with PE and washed out as

per the Tetramer staining protocol. Error bars represent intra-sample variation (n=5).

4.5 Discussion

This chapter examined methods to screen CBZ-SJS/TEN patients potential cross-

reactive memory T-cell responses to HLA-B*15:02-restricted HSV-1 and HSV-2

peptides. A series of predicted HLA-B*15:02-restricted HSV-1 and/or HSV-2 peptides

were identified through in-silico based predicted HLA-restricted epitope mapping. The

ex-vivo analysis identified several predicted peptides which matched the hypothesis

predictions, in that a proportion of the HSV-1 and/or HSV-2 HLA-B*15:02-restricted

peptides can induce a specific HLA-B*15:02-epitiope memory T-cell response shared

in patients who experience a CBZ-SJS/TEN reaction. These included the HSV-1,

envelope glycoprotein H derived peptide, LAFVLDSPSAY (UL22, AA 476-486), the

HSV-1/HSV-2 shared Helicase-primase-primase subunit (UL52, AA 811-821 and 823-

833, in HSV-1 and HSV-2 respectively) peptide, YAHGHSLRLPY and the previously

identified M. tuberculosis ESAT-6-like protein EsxB derived peptide, NIRQAGVQY (AA

72-80).

The pure YAHGHSLRLPY peptide failed to illicit the same response the original

peptide product in repeated IFN-γ ELISpot experiments. Both the original and pure

116

YAHGHSLRLPY peptide products were analysed via mass spectrometry. The mass

spectra analysis showed that the original YAHGHSLRLPY peptide product had at least

three possible variants of the YAHGHSLRLPY, with a loss of a histidine (H) at position

3 and/or 5. The YAHG_SLRLPY variant, missing the 5th Histidine (H), gave the

greatest IFN-γ ELISpot responses, ex-vivo.

A Basic Local Alignment Search Tool (BLAST) search of the National Centre for

Biotechnology Information (NCBI) protein database using the YAHG_SLRLPY atypical

epitope showed the specific-YAHG_SLRLPY amino acid sequence matched none of

the reference protein sequences in the NCBI protein database and one possibility is

that the cellular immune response generated by this atypical epitope is an artefact of

ex-vivo analysis (169, 188, 189, 214). The original YAHGHSLRLPY peptide originates

from a highly conserved region of the UL52 protein in all the various strains of HSV-1

and HSV-2 that have been identified and published. (169, 188, 189, 214). It should be

noted that not finding a match in a BLAST search does not suggest this specific

atypical epitope does not exist, it merely suggests it does not exist in any identified

and published sequence contained within the specific databases.

The identification of such an atypical epitope has a remarkable impact on our

understanding of immunological and/or virological processes. We need to better

understand the mechanisms behind ambiguous transcription, translation, protein

folding, degradation, and epitope synthesis before we can understand the efficiency

and frequency of atypical epitopes. Several debated mechanisms have been

identified, which could cause atypical epitope generation. (215-220) These include the

DRiP hypothesis which describes substandard polypeptides developing from

alternative and/or defective mRNAs, ribosomal frameshifts, downstream initiation

on genuine mRNAs, and other errors inherent in translation including tRNA-amino acid

misacylation. (221-230) Using the netChop 3.1 program to analysing YAHGHSLRLPY

and surrounding amino acids within the HSV-1 and HSV-2 Helicase-primase-primase

subunit, four cleavage sites were predicted (231). The four cleavage sites included the

5th histidine (H), and the 1st tyrosine (Y) and both leucine residues (L), but the netChop

program is rather underdeveloped (231).

The HLA-B*15:02-restricted peptides LAFVLDSPSAY and NIRQAGVQY produced

positive tetramer responses, in at least one of the three CBZ-SJS/TEN, HLA-B*15:02

positive donors. As we could not determine the M. tuberculosis serology of the Donors

and Controls, the positive NIRQAGVQY tetramer response in Donor 2 (SJS) is of

117

interest. The majority M. tuberculosis infections are latent and asymptomatic and this

specific epitope is not one of the 483 T-cell epitopes in 12 BCG vaccine strains,

according to Skjøt et al., 2002. The NIRQAGVQY epitope is one of four recognised by

CD8+ T cells that Li et al. 2014 identified as important for the development of M.

tuberculosis peptide-based vaccines and for improved diagnosis of TB in humans

(211, 232). The identification of NIRQAGVQY-specific CD8+ T cells either means the

patient has or had a TB infection or the epitope is non-specific.

The assessment of three HLA-B*15:02-restricted CBZ-SJS/TEN donors is not powerful

enough to exclude the HLA-B*15:02-restricted peptide LAFVLDSPSAY and while the

hypothesis originally stated that a specific HSV-1/HSV-2 epitope HLA-B*15:02-

restricted memory T-cell response shared in patients who experience a CBZ-SJS/TEN

reaction cross-reacts with neo-antigens derived by the off-target actions of drugs with

high-risk HLA alleles, it is also probable that several of HSV-1/HSV-2 epitope HLA-

B*15:02-restricted memory T-cell response are associated with CBZ-SJS/TEN. Ideally

more predicted HLA-B*15:02 restricted HSV-1 / HSV-2, peptides should be screened

before proceeding. Given the association between the CBZ-TEN reaction and the

acute HZ experienced by Donor 1 (CBZ-TEN) in-silico epitope mapping strategies

should also be extended to generating several HLA-restricted VZV peptides.

Further analysis is required, involving more CBZ-SJS/TEN donors and CBZ-naive and

CBZ-tolerant controls which have been hard to acquire. In the recent paper by Jing et

al., 2016 showed recognition of VZV-infected cells, proteins, and peptides by

polyclonal HSV-1–driven CD8 T cells (233). Along with the detection of cross-

reactivity, this paper also discovered several novel cross-reactive

T-cell epitopes, including one HLA-B*15:02-restricted HSV / VZV cross-reactive T-cell

epitope which was found using PBMC’s acquired from the same donor used in this

study, Donor 1 (233). Future experiments should include the assessment of this newly

identified VZV / HSV epitope, following the method and experimentation outlined in

this chapter. If the newly identified VZV / HSV epitope can fulfil the requirements as

the atypical YAHG_SRLPY HSV-1 / HSV-2 epitope has, both these epitopes can be

used in future studies into the immunopathogenesis of HLA-B*15:02-restricted CBZ-

SJS/TEN.

118

119

Chapter 5



Specific T-cell Clonotype and Detection by Digital Droplet PCR

120

121

5 5.1 Introduction

This chapter will continue on this exploration of the heterologous immune model by

assessing specific cross-reactive TCR associated with HLA-B*15:02-resticted CBZ-

SJS/TEN. A strong association between HLA-B*15:02 CBZ-SJS/TEN in those of

South-east Asian ancestry has been established (2, 209, 210). Based on a strong

negative predictive values associated with HLA-B*15:02 carriage (100% in Han

Chinese), the Food and Drug Administration (USA) and similar regulatory agencies of

different countries have recommended that genetic screening for HLA-B*15:02, in

those of South-east Asian ancestry, before CBZ therapy and that all HLA-B*15:02

carriers, should not be administered CBZ, unless the benefits outweigh the risks (101).

The low/incomplete positive predictive value of HLA-B*15:02 carriage illustrates how

carriage is necessary, but not enough to explain CBZ-SJS/TEN pathogenesis and

other factors must also play a role in driving CBZ-SJS/TEN.

The heterologous immunity model explains, not only the variation of tissue distribution

between different clinical delayed IM-ADR, but also the low positive predictive value of

HLA-B*15:02 carriage in CBZ-SJS/TEN by explaining what is driving

immunopathogenesis. This study looked at expanding CBZ-specific T cells (CD8+ T

cells) to see if the patients in this study carried the oligoclonal CBZ-specific public TCR

sequences previously identified in Ko et al. 2011 (176). Through the development of a

Bio-Rad QX100/200 ddPCR assay, the mRNA expression of the CBZ-specific public

TCR sequences, Vβ 25-1*01 (Vβ 11s1) CDR3 CASSISGSY and CASSGLADVDN

sequences were analysed (176, 234). The Vβ 12-3/4*01 CDR3 CASSLAGELF was

identified in the analysis of CBZ-SJS/TEN patients blister fluid by colleagues at the

National Yang-Ming University and Chang Gung Memorial Hospital in Taiwan and

through personal communications with them, was also assessed (Table 5.1).

122


Hypothesis

Patients that have experienced a CBZ-SJS/TEN share a subset of oligoclonal T cells,

previously identified CBZ-specific public TCR sequences Vβ 25-1*01 (Vβ 11s1) CDR3

CASSISGSY / CASSGLADVDN or the newly identified Vβ 12-3/4*01 CDR3

CASSLAGELF sequence.

Aims

To expand CBZ-specific T cells and identify specific CDR3 sequences in a novel

ddPCR assay (176).

5.3 Methods




Induced Stevens - Johnson syndrome and Toxic Epidermal Necrolysis.

5.3.2 Experimental Assays and Analysis

The stimulation and expansion culture outlined in Chapter 2, Materials and Methods

Section 2.2.1 Drug Stimulation and Cellular Expansion, was used to stimulate and

expand polyclonal T cells. The IFN-γ ELISpot assays and ICS flow cytometry outlined

in Chapter 2, Section 2.2 were used to assess the CBZ-associated immune

responses. Figure 5.1 includes an example of the raw ELISpot results from which spot

forming unit/one million cells (SFU/million cells) was derived and an example of the

ICS flow cytometry raw data plots from which the CD8+ T-cell IFN-γ response to CBZ

was detected both pre- and post-stimulation is shown in Figure 5.2. Statistical analysis

and graphical representations were calculated using GraphPad Prism software

Version 5.02 (GraphPad software Inc., California, USA).

RNA was extracted from T cells acquired from pre-stimulation and

post-stimulation and expansion, and converted to cDNA as highlighted in Chapter 2,

Section 2.4.1 RNA isolation and complementary DNA conversion. The cDNA was

used in the ddPCR assay to amplify and detect the specific TCR Vβ CDR3 sequences.

QX100/200 raw data output was analysed using the QuantaSoft software (Version

123

1.5.38.1118). As outlined in Chapter 2, Digital Droplet PCR 2.4.2 and in Appendix 9,

Section 9.2 the forward and reverse primers listed in Table 2.17 were validated using

the reference cDNA sequence. For the ddPCR specific Vβ CDR3 TCR analysis,

expression levels were normalised using RPPH1 gene expression and reported as a

percentage of normalised Cβ mRNA expression levels. Statistical analysis and

graphical representations were calculated using GraphPad Prism software Version

5.02 (GraphPad software Inc., California, USA).

Table 5.1: Distribution of the work within this chapter and accreditation to those who contributed


Information about the specific Vβ CDR3 TCR

sequences to target in ddPCR assays

Ko et al. 2011 (176)

and personal correspondence with

colleagues at Taipei Taiwan and

Chang Gung Memorial Hospital

Patient Phenotype and clinical analysis Elizabeth Philips

Ex-vivo analysis of CBZ-stimulated T cells Craig Rive

ddPCR assay development Craig Rive

Vβ, variable beta; CDR3, complementarity-determining region 3; TCR, T-cell receptor, ddPCR; digital droplet PCR.

5.4 Results

5.4.1 ELISpot and ICS Flow Cytometry responses in Donor

and Control cells pre- and post- CBZ stimulation and

expansion

The IFN-γ responses to both pre- and post- cellular stimulation and expansion cultures

were assessed using ELISpot and ICS, to show the T-cell expansion culture was

successful in expanding polyclonal T cells, some of which are CBZ-specific CD8+ T

cells. Statistically significant increases in T-cell IFN-γ responses were seen in the

three CBZ-SJS/TEN Donors CBZ-stimulated and expanded culture cells, when re-

exposed to CBZ compared to the controls (Donor 1 (CBZ-TEN) and Donor 2 (CBZ-

SJS) (***) p value < 0.0001, Donor 3 (CBZ-SJS) (*) p value < 0.05 Figure 5.1). ICS

IFN-γ flow cytometry analysis shows an increase in IFN-γ expression by CD8+ T cells

in the three CBZ-SJS/TEN Donors CBZ-stimulated and expanded cells when re-

exposed to CBZ compared to the controls (Figure 5.1).

124

Figure 5.1: Left: Example IFN-γ ELISpot raw data. The unstimulated background and solvent controls

(DMSO/DMF at 1μL/mL), positive controls anti-CD3 antibody (wells B1 and B2, 0.1μg/mL) and SEB

(wells B3 and B4, 1μg/mL) responses at 200,000 cells per well. The test positive and test negative

responses highlight the typical positive and negative response to the test stimulant (drug or peptide

10μg/mL) with an input of 50,000 and 200,000 cell per well. Right: Mean (+/- SD) IFN-γ responses

SFU/million cells in CBZ-SJS/TEN Donors and Controls, both pre- and post-CBZ stimulation and

culture, were determined using ELISpot, described in Chapter 2, Enzyme-Linked ImmunoSpot 2.2.3.

Error bars represent intra-sample variation (n=3)

Figure 5.2: Left: Example ICS raw data plots showing the CBZ-associated cellular IFN-γ responses both

pre- and post-stimulation (CBZ 10μg/mL) in Donor 1 (CBZ-TEN). Right: Mean (+/- SD) IFN-γ ICS flow

cytometry analysis for CBZ-SJS/TEN Donor and controls cells both pre- and post- stimulation and

culture were determined using the ICS method outlined in Chapter 2, Intracellular cytokine staining

2.2.2.1. Error bars represent intra-sample variation (n=3).

125

5.4.2 TCR analysis of Donor and Control cells pre- and post-

CBZ stimulation and expansion using digital droplet

PCR

The assay workup required the validation of the specific primers, the thermocycling

conditions, constructing a positive control and assessment of the assay sensitivity.

Using a synthetic construct which contained the relevant primer sequences, the

ddPCR assay was validated (Appendix 9, Section 9.2).

5.4.2.1 TCR Vβ CDR3 ddPCR RNA expression analysis

Figure 5.3 to Figure 5.5 highlight the TCR Vβ CDR3 profile of Donor 1 (CBZ-TEN),

Donor 2 (CBZ-SJS) and Donor 3 (CBZ-SJS), with Table 5.2 summarising the findings.

In Donor 1 (CBZ-TEN) a significant increase in Vβ 25-1*01 (Vβ 11s1) CDR3

CASSISGSY expression was detected in the cells stimulated with CBZ and expanded

in culture compared to pre-stimulation, unstimulated and control drug-stimulated and

expanded cells. Donor 2 (CBZ-SJS) showed a significant increase in Vβ 25-1*01

CDR3 CASSISGSY expression in the CBZ-stimulated and expanded cells compared

to cells pre-stimulation and unstimulated expanded cells. However, the increase Vβ

25-1*01 CDR3 CASSISGSY expression, seen in CBZ-stimulated and expanded cells

was also seen control drug-stimulated and expanded cells.

Donor 2 (CBZ-SJS) and Donor 3 (CBZ-SJS) exhibited a significant increase in Vβ 12-

3/4*01 CDR3 CASSLAGELF expression in the cells stimulated with CBZ and

expanded in culture compared to the three control conditions. Donor 1 (CBZ-TEN) also

exhibited a significant increase in Vβ 12-3/4*01 CDR3 CASSLAGELF expression cells

stimulated with CBZ and expanded in culture when compared to the control pre-

stimulated cells only.

Vβ 25-1*01 CDR3 CASSGLAGVDN expression was found significant in Donor 3

(CBZ-SJS) CBZ-stimulated and expanded cells when compared to cells pre-

stimulation and unstimulated expanded cells. The increase in Vβ 25-1*01 CDR3

CASSISGSY expression seen in CBZ-stimulated and expanded cells was also seen

control drug-stimulated and expanded cells.

126

Figure 5.3: TCR Vβ CDR3 sequence profile of Donor 1 CBZ-TEN. Raw data analysed using Bio-Rad

QuantaSoft software, results normalised against RPHH1 expression, and calculated as a percentage

(%) of normalised Cβ expression. Statistical significance was determined using analysis of variance

(ANOVA) statistical analysis (p value < 0.0001) and multiple comparison tests showed significant

differences between CBZ-stimulated cells to the other 3 conditions (***p value < 0.0001, *p value <

0.05).

127

Figure 5.4: Vβ TCR CDR3 sequence profile of Donor 2 (CBZ-SJS). Raw data analysed using Bio-Rad




differences between CBZ-stimulated cells to the other 3 conditions (***p value < 0.0001).

128

Figure 5.5: Vβ TCR CDR3 sequence profile of Donor 3 (CBZ-SJS). Raw data analysed using Bio-Rad




differences between CBZ-stimulated cells to the other 3 conditions (***p value < 0.0001).

129

Table 5.2 Table 5.2 Summary of the ddPCR amplification and analysis of target Vβ CDR3 sequences as

a percentage (%) of expression in Donor 1 (CBZ-TEN), Donor 2 (CBZ-SJS), and Donor 3 (CBZ-SJS)

pre-stimulation cells, CBZ, control drug stimulated and expanded cells, and the unstimulated expanded

cells.

Sample Vβ 25-1*01 (Vβ 11s1) CDR3 CASSISGSY

Pre-stimulation CBZ stimulated Control drug stimulated Unstimulated

Donor 1

(CBZ-TEN)

1.46% 44.23%

5.72% 8.50%

Donor 2

(CBZ-SJS)

1.57% 8.23% 7.72% 0.79%

Donor 3

(CBZ-SJS)

0.73% 0.72% 0.01% 0.04%

Vβ 25-1*01 CDR3 CASSGLAGVDN


Donor 1

(CBZ-TEN)

1.94% 3.34% 4.76% 9.07%

Donor 2

(CBZ-SJS)

0.44% 0.63% 1.35% 0.67%

Donor 3

(CBZ-SJS)

1.44% 14.02% 10.44% 3.66%

Vβ 12-3/4*01 CDR3 CASSLAGELF


Donor 1

(CBZ-TEN)

0.07% 18.23%

6.84% 3.40%

Donor 2

(CBZ-SJS)

2.95% 75.33%

14.78% 2.90%

Donor 3

(CBZ-SJS)

1.23% 17.89%

0.34% 2.94%

5.5 Discussion

Using PBMC’s obtained from patients who had experienced a

CBZ-SJS/TEN reaction, this study showed that CBZ induced responses can be

expanded using a short cellular stimulation and expansion culture. CBZ-specific public

TCR sequences, Vβ 25-1*01 (Vβ 11s1) CDR3 CASSISGSY / CASSGLADVDN or the

130

newly identified Vβ 12-3/4*01 CDR3 CASSLAGELF sequence expression can be

detected using a two-step rtPCR and using the cDNA as the template in a ddPCR

assay (176).

The Vβ 25-1*01 (Vβ 11s1) CDR3 CASSISGSY sequence was found to be expressed

in Donor 1 (CBZ-TEN) when cells were stimulated with CBZ and expanded in culture

compared to the unstimulated and control drug-stimulated and expanded cells. The Vβ

12-3/4*01 CDR3 CASSSLAGELF sequence was found to be expressed in Donor 1

(CBZ-TEN) cells stimulated with CBZ and expanded in culture, when compared to pre-

stimulated cells and while not significant, an increase was seen when compared to the

unstimulated and control drug-stimulated and expanded cells (Table 5.2).

Donor 2 (CBZ-SJS) exhibited a significant increase in the Vβ 12-3/4*01 CD3

CASSLAGELF sequence expression in cells stimulated with CBZ and expanded in

culture compared to pre-stimulation, unstimulated and control drug-stimulated and

expanded cells (Table 5.2). Donor 3 (CBZ-SJS) also exhibited an increase in the Vβ

12-3/4*01 CD3 CASSLAGELF sequence expression. Cells from Donor 2 (CBZ-SJS)

and Donor 3 (CBZ-SJS) showed a CBZ-associated Vβ 25-1*01 CDR3 CASSISGSY

and CDR3 CASSSGLAGVDN increased expression, respectively (Table 5.2).

However, in both cases, the increase seen in the CBZ-stimulated and expanded cells

was also seen in the control drug-stimulated and expanded cells. It is possible the

increase in CDR3 CASSISGSY and CDR3 CASSSGLAGVDN sequence expression

results from a distorted representation of the T-cell repertoire driven by ex-vivo

expansion.

Jing et al. 2016 in collaboration with other members of our group showed recognition

of VZV-infected cells, proteins, and peptides by polyclonal HSV-1–driven CD8+ T cells.

In this paper, they identified several potential VZV / HSV cross-reactive Vβ TCR CDR3

sequences specific to Donor 1 (CBZ-TEN). The expression can be assessed using the

same method described in this chapter. Several IM-ADR studies have highlighted the

involvement of more complex TCR repertoire (116, 235, 236). It is also possible that

the specific Vβ or Vα TCR clonotype is not as essential as the specific-CDR3

sequence, which interacts with the HLA-bound peptide. This could mean that the TCR

repertoire can be heterogeneous and polyclonal, yet share the same or structurally

similar specific-CDR3 sequences. The data presented here shows in at least Donor 1

(CBZ-TEN) shows the CBZ-associated expansion of a polyclonal T-cell pool, each

expressing different Vβ specific-CDR3 sequences. Specific of cross-reactive memory

131

T cells might determine tissue specificity and severity of the reaction. Recent studies

have also highlighted how HSV-specific CD8αα+ tissue-resident T cells are persistent

within the epidermis over a prolonged period and cause rapid cytotoxicity to cells

expressing low levels of HSV reactivation (237). Going forward it may be prudent to

look at epidermal tissue samples from patient blister fluid, to identify CBZ-SJS/TEN

restricted TCR clonotype repertoire.

What this chapter has provided is a method that is adaptable and sensitive to pick up

the increased expression of known Vβ TCR CDR3 sequence. The method,

experimental design and in particular the ddPCR assay performed here can be utilised

in any study involving the quantitative assessment and expression of any known Vβ or

Vα TCR CDR3 sequence. The ddPCR method provided here can be utilised to further

assess Vβ TCR CDR3 sequences identified by Ko et al, the expression of VZV / HSV

cross-reactive Vβ or Vα TCR CDR3 sequences identified by Jing et al., 2016, or

identifying the HSV-specific CD8αα+ tissue-resident T-cell clonotype (176, 233, 237).

The ddPCR method outlined in this chapter can also be modified and adapted to

assess TCR expression in other delayed IM-ADR and can be used in several

immunological research disciplines, including T-cell immunology.

132

133

Chapter 6



Granulysin Expression

134

135

6 6.1 Introduction

Granulysin is an effector molecule that exhibits cytotoxic activity against several

pathogens and tumours while also possessing chemoattractant and proinflammatory

abilities (118-125). Granulysin can be expressed by several cells, including CD8+

cytotoxic T cells, CD4+ cytotoxic-like T cells, NK cells and NK-T cells (2, 119, 125-

127). Granulysin has proven to be a useful biomarker and an important cytotoxicity

mediator in several skin diseases besides SJS/TEN including acne, psoriasis and

folliculitis (119, 125, 129-132). In terms of SJS/TEN, granulysin has been associated

with the disseminated keratinocyte death seen in SJS/TEN pathology. (111, 117)

This chapter will explore CBZ-SJS/TEN immunopathogenesis through the expression

of granulysin in the PBMC’s acquired from patient donors that had experienced a

SJS/TEN reaction to CBZ, compared to CBZ-naive control PBMC’s. This will be

achieved using ICS and Immunophenotyping flow cytometry assays to define the

cellular subtypes within donor PBMC’s which show an increase in granulysin

expression when incubated with pharmacological relevant doses of CBZ, ex-vivo.


Hypothesis

The analysis of CBZ-SJS/TEN patient donor PBMCs show an increase granulysin

expression within several cellular subtypes when stimulated with pharmacological

amounts of CBZ (10μg/mL) ex-vivo.

Aims

To detect granulysin expression in cellular subtypes within PBMC’s from

CBZ-SJS/TEN Donors and Controls stimulated with pharmacological amounts of CBZ

(10μg/mL) compared to unstimulated and control drug stimulated PBMC’s.

6.3 Methods




Induced Stevens - Johnson syndrome and Toxic Epidermal Necrolysis.

136

6.3.2 Experimental assays and analysis

ICS and immunophenotyping flow cytometry assays were performed as outlined in

Chapter 2 Materials and Methods, Section 2.2.2 Flow Cytometry. Figure 6.1 shows the

flow cytometry gating procedure used to isolate the relevant cellular subpopulations to

assess granulysin expression. From the forward and side scatter plots, the lymphocyte

population was identified and labelled gate A, which was then further separated into

CD3+ and CD3- cells (Figure 6.1B). The CD3- / CD56+ NK cell population was

assessed for granulysin expression (Figure 6.1C), the CD3+ cells were further sorted

into CD56+ and CD56- cells (Figure 6.1D). The CD3+ CD56- cells were sorted into

CD8+ and CD4+ (Figure 6.1E) and each assessed for granulysin expression (Figure

6.1E1 and E2). The CD56+ cells were sorted into CD8+ and CD4+ (Figure 6.1F) and

each assessed for granulysin expression (Figure 6.1F1 and F2). Statistical analysis

and graphical representations were calculated using GraphPad Prism software

Version 5.02 (GraphPad software Inc., California, USA). Figure 6.1 Top Right

highlights an example of negative to low Granulysin expression and Figure 6.1 bottom

right highlights an example of positive granulysin expression.

Table 6.1: Distribution of the work within this chapter and accreditation to those who contributed


Patient Phenotype and clinical analysis Elizabeth Philips

Ex-vivo analysis of CBZ-stimulated PBMC’s via

Flow Cytometry

Craig Rive

137

Figure 6.1: Left: Example immunophenotyping and ICS raw data plots showing the gating procedure

used to isolate the cellular subpopulations and determine their cytokine expression under the relevant

conditions. Top Right: Example of negative to low Granulysin expression. Bottom Right: Example of

positive granulysin expression.

6.4 Results

6.4.1 Evaluation of granulysin in CD8+ T cells obtained from

CBZ SJS/TEN donors and naive CBZ controls

Comparisons between unstimulated, CBZ-, and or control drug-stimulated CD8+

cytotoxic T cells in both CBZ-SJS/TEN Donors and CBZ-naive controls are highlighted

in Figure 6.2 to Figure 6.5. All data points in Figure 6.2 show the mean (SD+/-)

percentage of PBMC’s CD8+ cytotoxic T cells that express granulysin, for each sample

(n > 3) under the specific ex-vivo stimulation condition. Figure 6.3 shows the same

data but compares the unstimulated versus the CBZ-stimulated ex-vivo granulysin

responses for each sample.

138

Figure 6.2: Mean (+/- SD) Granulysin expression in CD8+ cytotoxic T cells from Donor and Control

PBMC’s incubated ex-vivo with CBZ, control drug, or left unstimulated. Statistical significance was

determined using ANOVA analysis (p < 0.0001). Multiple comparison tests showed significance

between unstimulated and CBZ-stimulated donor PBMC’s (* p value < 0.05), CBZ and control drug-

stimulated donor PBMC’s (*** p value < 0.0001), CBZ and the unstimulated, CBZ and control drug-

stimulated donor PBMC’s (*** p value < 0.0001). The error bars represent inter-sample variation (n=4).

139

Figure 6.3: Mean (+/- SD) Granulysin expression in CD8+ cytotoxic T cells in Donor and Control PBMC’s

incubated ex-vivo with CBZ versus unstimulated PBMC’s. Statistical significance between unstimulated

and CBZ-stimulated for each set of donors (***p value of < 0.0001), using the paired two-tailed T-test,

compared to the controls (not significant (ns)). The error bars represent intra-sample variation (n=4).

140

Figure 6.4 and Figure 6.5 highlight IFN-γ expression in CD8+ cytotoxic T cells. The

IFN-γ expression results were obtained by exposing PBMC’s from the Donors and

Controls to the same conditions in which the granulysin expression results in Figure

6.2 and Figure 6.3 were obtained. Comparing granulysin versus IFN-γ expression in

CD8+ cytotoxic T cells, differences were observed, including the percentage of CD3+

CD8+ T cells expressing IFN-γ when stimulated with CBZ (10μg/mL) being only 0.1-

0.2% of total PBMC’s, compared the percentage of CBZ-stimulated CD3+ CD8+ T cells

expressing granulysin (Figure 6.2).

Figure 6.4: Mean (+/- SD) INF-γ expression in CD8+ cytotoxic T cells from Donor and Control PBMC’s

incubated ex-vivo with CBZ, control drug, or left unstimulated. Statistical significance was determined

using ANOVA analysis (p < 0.0001). Multiple comparison tests showed significance between,

unstimulated and CBZ-stimulated donor PBMC’s (*** p value < 0.0001), CBZ and Control drug-

stimulated donor PBMC’s (*** p value < 0.0001), CBZ and the unstimulated, CBZ and Control drug-

stimulated donor PBMC’s (*** p value < 0.0001). The error bars represent inter-sample variation (n=4).

141

Figure 6.5: Mean (+/- SD) IFN-γ expression in CD8+ cytotoxic T cells in Donor and Control PBMC’s

incubated ex-vivo with CBZ versus unstimulated PBMC’s. Statistical significance between unstimulated

and CBZ-stimulated Donor 1 (CBZ-TEN) (***p value of < 0.0001), Donor 2 (CBZ- SJS) (*p value of <

0.05) and Donor 4 (CBZ-SJS non- B*15:02) (*p value of < 0.05) using the paired two-tailed T-test.

Donor 3 (CBZ-SJS) and Controls CBZ-naive (both HLA-B*15:02 positive and negative) were not

significant (ns)). The error bars represent intra-sample variation (n=4).

142

6.4.2 Evaluation of granulysin in CD4+ T cells obtained from

CBZ SJS/TEN donors and naive CBZ controls

Comparisons between unstimulated, CBZ-, and control drug-stimulated CD4+ T cells in

both CBZ-SJS/TEN Donors and CBZ-naive controls are highlighted in Figure 6.6. All

data points in Figure 6.6 represents the mean (SD+/-) percentage of PBMC’s that are

granulysin-expressing, CD4+ T cells for each sample (n > 3) under the specific ex-vivo

stimulation condition. No significant difference in granulysin expression was seen in

CD4+ T cells from the HLA-B*15:02 positive donors, Donor 1 (CBZ-TEN), Donor 2

(CBZ-SJS) and Donor 3 (CBZ-SJS) and the CBZ-naive controls when stimulated with

a pharmacological relevant dose of CBZ compared to control drug, or left

unstimulated. Figure 6.7 shows, Donor 4 (CBZ-SJS non-B*15:02) showed a significant

increase in granulysin expression by CD4+ T cells exposed to CBZ, compared to the

HLA-B*15:02 donors, Donor 1 (CBZ-TEN), Donor 2 (CBZ-SJS) and Donor 3 (CBZ-

SJS) and controls. Figure 6.8 further supports this with a significant increase in IFN-γ

expression by CD4+ T cells in Donor 4 (CBZ-SJS non-B*15:02) when exposed to CBZ.

143

Figure 6.6: Mean (+/- SD) Granulysin expression in CD3+ CD4

+ CD56

- T cells in Donor and Control

PBMC’s incubated ex-vivo with CBZ, control drug, or left unstimulated. No statistical significance was

determined between the means of the various conditions. (ANOVA analysis and multiple comparison

tests).The error bars represent inter-sample variation (n=4).

144

Figure 6.7: Mean (+/- SD) Granulysin expression in CD4+ T cells in Donor and Control PBMC’s

incubated ex-vivo with CBZ versus unstimulated PBMC’s. Statistical significance was found between

Donor 4 (CBZ-SJS non- B*15:02) unstimulated and CBZ-stimulated PBMC’s (*** p value < 0.0001).

HLA-B*15:02 positive Donors 1 (CBZ-TEN), Donor 2 (CBZ-SJS), and Donor 3 (CBZ-SJS) unstimulated

versus CBZ-stimulated PBMC’s, and the Controls were found to be not significant using a paired two-

tailed T-test. The error bars represent intra-sample variation (n=4).

145

Figure 6.8: Mean (+/- SD) IFN-γ expression in CD4+ T cells in Donor and Control PBMC’s incubated ex-

vivo with CBZ versus unstimulated PBMC’s. Statistical significance was found between Donor 4 (CBZ-

SJS non- B*15:02) unstimulated and CBZ-stimulated PBMC’s (*** p value < 0.0001). HLA-B*15:02

positive Donors 1 (CBZ-TEN), Donor 2 (CBZ-SJS), and Donor 3 (CBZ-SJS) unstimulated versus CBZ-

stimulated PBMC’s, and the Controls were found to be not significant using a paired two-tailed T-test.

The error bars represent intra-sample variation (n=4).

146

6.4.3 Evaluation of granulysin in CD56+ NK and NK-T cells

obtained from CBZ SJS/TEN donors and naive CBZ

controls

The final PBMC’s cellular subsets included CD56+ NK cells and CD3+ CD56+ CD8+ or

CD4+ cell populations. CD56+ on the CD3+ T cells could reflect T-cell activation and

not specifically NK-T cells themselves (238). No significant correlation was seen

between unstimulated, CBZ-, and control drug-stimulated granulysin expression in

CBZ-SJS/TEN Donors and CBZ-naive controls as highlighted in Figure 6.9, Figure

6.10, and Figure 6.11. All data points in the following graphs represents the mean

(SD+/-) percentage of total PBMC’s that are granulysin-expressing, CD56+ NK cells

(Figure 6.9), CD56+ CD8+ NK-T cells (Figure 6.10), or CD56+ CD4+ NK-T cells (Figure

6.11) for each sample (n > 3) under the specific ex-vivo stimulation condition.

Figure 6.9: Mean (+/- SD) Granulysin expression in CD56+ NK cells in Donor and Control PBMC’s

incubated ex-vivo with CBZ, control drug, or left unstimulated. The error bars represent variation within

the same sample on multiple test (N tests per sample in each condition = 4).

147


+ CD56

+ NK-T cells in Donor and

Control PBMC’s incubated ex-vivo with CBZ, control drug, or left unstimulated. The error bars represent

intra-sample variation (n=4).

148


+ CD56

+ NK-T cells in Donor and

Control PBMC’s incubated ex-vivo with CBZ, control drug, or left unstimulated. The error bars represent

inter-sample variation (n=4).

6.5 Discussion

Granulysin has been attributed to the disseminated keratinocyte death seen in

SJS/TEN pathology. (111, 117) Granulysin in the epidermal blister fluid of SJS/TEN

patients; predominately the 15-kDa form is secreted by both NK cells and CD8+ T cells

via a nongranular exocytotic pathway (103, 117, 210). Lamotrigine is another

antiepileptic drug associated with SJS/TEN. In 2014, a case study of one patient with

lamotrigine-induced TEN showed an increase in CD56+ NK cell granulysin expression,

when the patient PBMC’s were exposed to lamotrigine ex-vivo over a four-day period.

(134) Another finding by Porebski et al. 2013 showed an increase in granulysin

expression by CD4+ T cell from patients who experienced a lamotrigine-induced

SJS/TEN, when re-exposed to lamotrigine ex-vivo (135).

149

In this specific analysis granulysin expression is not representative of disease

progression or a prediction marker, but it does have the ability to be an important test

in the diagnosis SJS/TEN and allow clinicians to have more confidence in identifying

the culprit drug. The performed in this chapter differs from these and other studies

involving antiepileptic drug-induced SJS/TEN and granulysin expression in that Figure

6.2 and Figure 6.3 show an increase in granulysin expression in CD8+ cytotoxic T cells

within PBMC’s exposed to CBZ, ex-vivo. These PBMC’s obtained from CBZ-SJS/TEN

donors up to 17 years after the initial CBZ-SJS/TEN reaction. Figure 6.2 shows an

increased expression of granulysin in unstimulated CD8+ cytotoxic T cells within the

donor PBMC’s when compared to the controls. This suggests the Donors have an

increased number CD8+ cytotoxic T cells already expressing granulysin, which

supports studies that show that granulysin is produced by CD8+ T cells, unlike other

apoptotic molecules like perforin and granzyme B (239). This increased frequency of

CD8+ cytotoxic T cells already expressing granulysin, coupled with the further increase

in granulysin expression by CD8+ cytotoxic T cells when PBMC’s are incubated with

CBZ ex-vivo, supports previous data that suggests CBZ-SJS/TEN is a memory based

immune response, with several CD8+ T cells “pre-loaded” with granulysin. This also

explains why some patients experience a whole skin necrolysis in as little as 24-36

hours.

No significant increase in granulysin expression was detected in CBZ-SJS/TEN Donor

and CBZ-naive control CD56+ cell subtypes when exposed to the various conditions

(unstimulated, CBZ-, and control drug-stimulated) (Figure 6.9, Figure 6.10, and Figure

6.11). This coupled with the increase in granulysin-expressing, CD8+ cytotoxic T cells

within PBMC’s exposed to CBZ, ex-vivo, we propose that CD8+ cytotoxic T cells are

the initial leukocyte subtype. The evidence of granulysin producing NK cells within

patient blister fluid at the time of a reaction could result from the initial CD8+ cytotoxic

T-cell reaction, leading to granulysin release which not only has a cytotoxic role but a

chemotaxic role, attracting NK cells to the site of inflammation exacerbating the

situation (239, 240). Further studies are required to verify this proposal.

Experimentation should involve immunophenotyping skin-infiltrating lymphocytes at

the acute phase of SJS/TEN. A recent presentation at the European Academy of

Allergy and Clinical Immunology’s 2016 Drug Hypersensitivity Meeting in Malaga,

Spain Villani A.P. et al. 2016 presented data which shows that CD8+ T cells and to a

lesser extent CD4+ T cells, were the main leukocyte subsets found in acute TEN

blisters (240).

150

Donor 4 (CBZ-SJS non-B*15:02) also showed a higher level of granulysin expression

in CD4+ T cells, that also increased with CBZ exposure, ex-vivo (Figure 6.7). This is

supported by the increase in IFN-γ under the same conditions (Figure 6.8). While only

one patient, it raises interesting questions and supports the results of Porebski et al.

2013 in lamotrigine-induced SJS/TEN (135). However, it was also noted that Donor 4

(CBZ-SJS non-15:02) also expressed higher levels of granulysin in the CD3+ CD4+

CD56+ cell subpopulation, irrespective of CBZ or control drug stimulation or left

unstimulated (Figure 6.11). Further analysis using samples acquired from

non-HLA-B*15:02 SJS/TEN patients is required before any potential models can be

hypothesised.

151

Chapter 7

Summary and Conclusion

152

153

7 7.1 Introduction

Studies have shown that approximately 2.2 million serious ADR occur every year in

the United States of America, with over 100 thousand (4.5%) resulting in death (3-6).

ADR are the fourth to sixth most common cause of mortality in the developed world.

This places high demands on hospitals, healthcare institutions and facilities, and

auxiliary health care services. The enormous burden and cost of ADR are incompletely

captured and underestimated (1, 5, 7).

The low frequency of these deadly, delayed IM-ADR means they are rarely seen in the

pre-clinical and pre-marketing phases of testing. Delayed IM-ADR were considered

unpredictable until researchers identified the association between carriage of the HLA-

B*57:01 allele and ABC-HSS. This pharmacogenetic association leads to HLA-

B*57:01 genetic screening been implemented globally in primary HIV clinical practice,

allowing ABC to stay on the market, while reducing the numbers of ABC-HSS case.

This HLA-drug association has also led to a greater understanding of the

immunopathogenesis underlying this T cell-driven delayed IM-ADR

7.2 Nevirapine Induced Hypersensitivity


Characteristics

NVP has been associated with a diverse number of HSR clinical phenotypes and

numerous HLA alleles. We explored the notion that tissue specificity of NVP-SCAR

and the various HLA alleles associations can be explained by similarities in HLA

peptide-binding cleft structures, by grouping HLA alleles according to shared key

amino acid residues which make up the HLA class-I binding cleft pockets (A-F),

MHCcluster algorithm, which seeks to group HLA molecules based on their peptide-

binding specificity as predicted by NetMHCpan 2.8, and looking at the validated CREG

HLA supertypes, which groups HLA alleles based on similarities in cross-reactivity of

public epitopes among HLA class-I alleles and binding cleft pocket chemistry. (168-

172)

154

7.2.1.1 NVP-HSR with hepatitis

A vast majority of NVP-HSR involve hepatic symptoms, with 80% NVP-DRESS cases

presenting with liver complications. An increased risk of NVP-HSR with hepatitis was

associated with the validated HLA supertype CREG B*07 and the MHCcluster

assigned pb-B*78 group. The MHCcluster assigned pb-B*78 group and binding cleft B-

pocket residue analysis of the showed the HLA-B*55:01 and HLA-B*56:01 alleles

shared the same binding cleft B-pocket residues and were the two alleles were shared

between the CREG B*07 and pb-B*78 groups. Further, analysis of these alleles is

required, but Donor 2 (NVP-HSR), who experienced a NVP-DRESS reaction, 7 days

after NVP initiation, carried the HLA-B*56:01 and HLA-C*04:03 alleles.

7.2.1.2 NVP-SCAR

The HLA-C*04:01 alleles was shown to be a significant predictive HLA allele in NVP-

SCAR (Table 3.8). Other significant HLA-C alleles associated with NVP-SCAR

included the HLA-C*05:01 alleles among Caucasian and the HLA-C*18:01 among

Africans. The HLA-C*04:01, -C*05:01, -C*18:01, and other related alleles prevalent in

the cohort were analysed using MHCcluster to see if they shared similar predicted

peptide-binding capabilities. MHCcluster analysis showed these alleles all preferred to

bind 9mer peptides with an Aspartic acid (D) at position 2 of the peptide (from the NH2

terminal end) which interacts with binding cleft B-pocket.

HLA-C*04:01, HLA-C*05:01, and HLA-C*18:01 also shared the same binding cleft F-

pocket residue. This F-pocket structure also proved to a significant predictor for NVP-

rash associated SCAR, when the cohort was analysed together, but only remained

significant in Caucasians, when the data was adjusted for race (Table 3.8). This group

of alleles which share the same binding cleft F-pocket possesses varied binding cleft

B-pocket sequences suggesting the potential importance of the binding cleft F-pocket.

The F-pocket sequence significance and the virtual modelling analysis performed by

colleagues at the University of Florida showed the interaction between NVP within the

B-pocket. More research is required before a model can be formulated to further

analyse NVP-SCAR and the significance of the shared F-pocket sequences and how

the preferred binding of NVP near the B-pocket interacts with peptide presentation (83-

85).

155


Immunopathogenesis

This smaller study was part of a larger study involving NVP-SCAR

immunopathogenesis (138). As highlighted in Keane et al. 2014, one patient exhibited

an increase in Tregs cell numbers corresponded with diminishing IFN-γ ELISpot

responses (138). After this first analysis, PBMC's samples were limiting, therefore to

discuss this question and provide more data to complement the data published by

Keane et al. 2014, donor plasma was analysed over the time course of NVP-DRESS

to see if the cytokine environment reflects the changes in specific T-cell

subpopulations. Several cytokines were measured in patients over the time course of

a NVP-DRESS reaction. Donor 1 (NVP-HSR) showed a detectable IFN-γ release at

day 1 post NVP- DRESS, which then decreased, potentially due to the withdrawal of

NVP. Both Donor 1 (NVP-HSR) and Donor 2 (NVP-HSR) both showed significant IL-6

levels within the first 8 days post NVP- DRESS.

Elevated IL-6 levels in plasma samples can be linked to numerous different

pathologies but taken in context could be significant in the immunopathogenesis of, or

recovery from, a NVP-DRESS reaction (195-197). Given this increase in IL-6 and the

role IL-6 plays in Tregs/Th17 balance, this cytokine alongside other key factors like

carriage specific HLA class-I alleles could be important in NVP-DRESS

immunopathogenesis. One model we propose is that the decrease in Tregs at the time

of the reaction, due to IL-6 altering the Th17/Treg balance might be partially driving

NVP-DRESS and the return of Tregs during recovery is responsible for the NVP-

stimulated diminishing IFN-γ PBMC's responses.

While this increase in IL-6 and the fact that Keane et al. 2014 showed the removal of

CD25+ T cells had no effect on restoring diminishing NVP-stimulated, IFN-γ ELISpot

responses, this data combined is only from a few patients (138). To further analyse

this proposed model, future analysis should look at the expression and quantification

of inhibitory immune receptors; CTLA4, PD-1, and/or PD-L1 (203-206). Ex-vivo

experiments including monoclonal antibodies directed towards blocking these

inhibitory immune receptors to see if the ex-vivo response can be recovered or further

rule out the role of Tregs in diminishing NVP-stimulated IFN-γ PBMC responses (203,

205, 206). While it would have been preferential to perform this on the specific

samples used in Keane et al. 2014 and within this thesis, this could not be performed

156

as samples were limited and used in the first study by Niamh Keane et al. 2014, hence

the reason we analysed the Donors' plasma cytokine environment.

7.3 Carbamazepine-Induced Stevens - Johnson

syndrome and Toxic Epidermal Necrolysis

A strong association between HLA-B*15:02 and other B75 serotypes, including the

HLA-B*15:08, -B*15:11, and -B*15:21 alleles and CBZ-SJS/TEN has already been

established (2, 207, 208). Furthermore, the HLA-B*15:02 and SJS/TEN association

has been extended to several other Antiepileptic drugs, including phenytoin,

lamotrigine, and oxcarbazepine. While screening for HLA-B*15:02 carriage based on

the 100% negative predictive value is a workable option, the low positive predictive

value of HLA-B*15:02 in CBZ-SJS/TEN (3.0-7.7% in Han Chinese) means screening

for HLA-B*15:02 prevents large number of patients who carry an associated HLA allele

from taking these CBZ. Limits are placed on the number of alternative drugs that can

be prescribed. Understanding this disparity between negative and positive predictive

values will help in our understanding of the immunopathogenesis driving HLA-

restricted drug-SJS/TEN and allow for more effective screening methods to be used.

The heterologous immune responses have been suggested as a model to explain not

only the low positive predictive value associated with at-risk HLA alleles and CBZ-

SJS/TEN but also explains the varied tissue distribution of IM-ADR.

One aspect this body of work has explored involves a heterologous immune response

explained in Chapter 1 (Section 1.4 Established Models of HLA-driven delayed

Immunologically Mediated-Adverse Drug Reactions) which might explain the low and

incomplete positive predictive value associated with HLA-restricted CBZ-SJS/TEN.

The existence or nonexistence of epitope-specific class-I memory T cells which

express a specific cross-reactive TCR, might determine CBZ-SJS/TEN development in

individuals who carry the at-risk HLA allele. This would also go further to explain the

low positive predictive value of HLA-B*15:02 in CBZ-SJS/TEN and further our

understanding of the immunopathogenic mechanisms driving drug-SJS/TEN.

7.3.1 Predicted HLA-restricted Peptide Screening

The identification of HLA-B*15:02-restricted peptides, shared and restricted to

CBZ-SJS/TEN patients, is an essential step assessing the heterologous immune

model as it relates to HLA-B*15:02-restricted CBZ-SJS/TEN. This initial screening of

HSV-1 and/or HSV-2 predicted HLA-B*15:02-restricted peptides generated by in-silico

157

epitope mapping techniques, designed and performed by colleagues at The Institute

for Immunology and Infectious Diseases, Murdoch University, Perth, Australia, elicited

genuine peptide IFN-γ ELISpot responses. The atypical YAHG_SLRLPY epitope, a

variant of the HSV-1/HSV-2 shared Helicase-primase-primase subunit (UL52 amino

acid positions 811-821 and 823-833, in HSV-1 and HSV-2 respectively) peptide

YAHGHSLRLPY, was the only peptide that fitted the parameters outlined by the

hypothesis.

A protein BLAST search of this atypical epitope did not match anything within the

NCBI databases (169, 186, 187, 212). This peptide originates from a highly conserved

region of the UL52 protein in all the various strains of HSV-1 and HSV-2 that have

been identified and published Specific-YAHG_SLRLPY amino acid sequence matched

none of the reference protein sequence in the NCBI protein database (169, 186, 187,

212).

The identification of such an atypical epitope has a remarkable impact on our

understanding of immunological and/or virological processes. A better understand of

the mechanisms behind ambiguous transcription, translation, protein folding and

degradation, and epitope synthesis is required before the efficiency and frequency of

atypical epitopes can be determined.

Several debated mechanisms have been identified, which could cause atypical epitope

generation (213-218). These include the DRiP hypothesis which describes

substandard polypeptides developing from alternative and/or defective mRNAs,

ribosomal frameshifts, downstream initiation on genuine mRNAs, and other errors

inherent in translation including tRNA-amino acid misacylation (219-228). Using the

netChop 3.1 program to analysing YAHGHSLRLPY and surrounding amino acids

within the HSV-1 and HSV-2 Helicase-primase-primase subunit, four cleavage sites

were predicted (229). The four cleavage sites included the 5th histidine (H), and the

1st tyrosine (Y) and both leucine residues (L), but the netChop program is rather

underdeveloped (229).

Continued analysis using more CBZ - SJS/TEN patients and CBZ-tolerant controls, is

required before an atypical epitope, like YAHG_SLRLY can be considered as a

potential epitope responsible in the primary immune response and the subsequent

maintenance of the memory T cell which express the specific cross-reactive TCR, in a

heterologous immune response. Future analysis should also not rule out other HHV or

persistent pathogens to screen for cross-reactivity. Given the association between the

158

CBZ-TEN reaction and the acute HZ experienced by Donor 1 (CBZ-TEN), in-silico

epitope mapping strategies should also be extended to generating several HLA-

restricted VZV peptides.

As highlighted in Chapter 4, a recent paper by Jing et al., 2016 discovered several

novel cross-reactive T-cell epitopes, including a HLA-B*15:02-restricted HSV / VZV

cross-reactive T-cell epitope found using PBMC’s acquired from Donor 1 (CBZ-TEN)

(231). This epitope should also be screened using the other Donors in this chapter

which have shown responses to the atypical YAHG_SRLPY HSV-1 / HSV-2 epitope. If

this newly identified VZV / HSV epitope meets the requirements outlined in the

hypothesis and aims of this chapter, as the atypical YAHG_SRLPY HSV-1 / HSV-2

epitope has, both these epitopes can be used in future studies into the

immunopathogenesis of HLA-B*15:02-restricted CBZ-SJS/TEN.

7.3.2 Specific T-cell Clonotype and Detection by ddPCR

This study also showed that CBZ induced responses can be expanded from PBMC’s

obtained from patients up to 17 years after the initial reaction. Expression of the

previously identified CBZ-specific public TCR sequences, Vβ 25-1*01 (Vβ 11s1) CDR3

CASSISGSY / CASSGLADVDN or the newly identified Vβ 12-3/4*01 CDR3

CASSLAGELF sequence can be detected using ddPCR (176).

The Vβ 25-1*01 (Vβ 11s1) CDR3 CASSISGSY sequence was found to be expressed

in Donor 1 (CBZ-TEN), when cells were stimulated with CBZ and expanded in culture

when compared to the unstimulated and control drug-stimulated and expanded cells.

The Vβ 12-3/4*01 CDR3 CASSSLAGELF sequence was found to be expressed in

Donor 1 (CBZ-TEN) cells stimulated with CBZ and expanded in culture when

compared to pre-stimulated cells and while not significant, an increase was seen when

compared to the unstimulated and control drug-stimulated and expanded cells.

Donor 2 (CBZ-SJS) exhibited a significant increase in the Vβ 12-3/4*01 CD3

CASSLAGELF sequence expression in cells stimulated with CBZ and expanded in

culture compared to pre-stimulation, unstimulated and control drug-stimulated and

expanded cells. Donor 3 (CBZ-SJS) also exhibited an increase in the Vβ 12-3/4*01

CD3 CASSLAGELF sequence expression. Cells from Donor 2 (CBZ-SJS) and Donor 3

(CBZ-SJS) showed a CBZ-associated increase in Vβ 25-1*01 CDR3 CASSISGSY and

CDR3 CASSSGLAGVDN expression, respectively. However, in both cases, the

increase in CBZ-stimulated and expanded cells was also seen in the control drug-

stimulated, expanded cells. It is possible the increase in CDR3 CASSISGSY and

159

CDR3 CASSSGLAGVDN sequence expression results from a distorted representation

of the T-cell repertoire driven by ex-vivo expansion.

This study did not confirm the findings by Ko et al, 2011, but an increase in Vβ 25-1*01

(Vβ 11s1) CDR3 CASSISGSY expression in CBZ-stimulated cells from Donor 1 (CBZ-

TEN) was seen. In contrast to the proposal by Ko et al, 2011, several IM-ADR studies

have highlighted the involvement of a more complex TCR repertoire (116, 233, 234). It

is also possible that the specific Vβ or Vα TCR clonotype is not as essential as the

specific-CDR3 sequence, which interacts with the HLA-bound peptide. This could

mean that the TCR repertoire can be heterogeneous and polyclonal, yet share the

same or structurally similar specific-CDR3 sequences.

Recent studies have also highlighted how HHV can drive the production of tissue-

resident memory T cells, possessing high cytolytic potential and a low trigger threshold

causing local tissue damage (235). One such example includes HSV-specific CD8αα+

tissue-resident T cells, persistent within the epidermis over a prolonged period and

causes rapid cytotoxicity to cells expressing low levels of HSV reactivation (235).

Going forward it will be prudent to look at epidermal tissue samples to identify both

HLA-restricted pathogenic epitopes and to screen for potential TCR clonotypes

involved in CBZ-SJS/TEN, to prove this heterologous immune model.

The recent paper by Jing et al., 2016 showed the recognition of VZV-infected cells,

proteins, and peptides by polyclonal HSV-1–driven CD8+ T cells. In this paper, they

identified several potential VZV / HSV cross-reactive Vβ TCR CDR3 sequences

specific the same Donor 1 (CBZ-TEN) also analysed here (231). The expression of

these VZV / HSV cross-reactive Vβ TCR CDR3 sequences can be assessed using the

same method described in this chapter.

What this chapter has provided is an adaptable and sensitive experimental method to

assess TCR expression. In particular, the ddPCR assay highlighted can be utilised in

any study involving the quantitative assessment and expression of any known Vβ or

Vα TCR CDR3 sequence. The ddPCR method used here can be utilised to further

assess Vβ TCR CDR3 sequences identified by Ko et al, the expression of VZV / HSV

cross-reactive Vβ or Vα TCR CDR3 sequences identified by Jing et al., 2016, or

identifying the HSV-specific CD8αα+ tissue-resident T-cell clonotype. This specific

ddPCR assay can be modified and adapted to not only assess TCR expression in

other delayed IM-ADR, this assay can be used across several immunological research

disciplines, including tumour T-cell immunology.

160

7.3.3 Granulysin Expression

Chapter 4 and 5 approached the study of CBZ-SJS/TEN immunopathogenesis by

evaluating the heterologous immunity model. Chapter 6 looked at another aspect of

CBZ-SJS/TEN immunopathogenesis, granulysin expression. The examination of

granulysin expression in patient PBMC’s obtained up to 17 years after the initial CBZ-

SJS/TEN reaction, involved the ex-vivo CBZ stimulation to show the relevant PBMC's

subtypes involved.

The granulysin analysis in this body of research differs from other bodies of work

because it shows an increase in granulysin-expressing by CD8+ cytotoxic T cells within

PBMC’s exposed to CBZ, ex-vivo (103, 116, 117, 134, 135). This study also shows an

increased expression of granulysin in unstimulated CD8+ cytotoxic T cells within the

donor PBMC’s when compared to the controls. The findings here suggest that CD8+

cytotoxic T cells are a primary cellular subtype within the PBMC’s that interacts with

CBZ. This would support the current model for immunopathogenesis, in that CBZ-

SJS/TEN is driven by CD8+ T cells through a HLA-restricted TCR interaction and their

activation recruits other granulysin producing cells into this reaction. This would also

explain why it is common to find not only activated CD8+ cytotoxic T cells, but other

activated granulysin excreting cells like NK and NK T cells, within the epidermal

blisters of patients during the reaction (113, 114, 117).

It has been shown that granulysin can cause a 2 to 7 fold increase in chemotaxis of

monocytes, CD4+, and CD8+ memory T cells, NK cells, and mature, monocyte-derived

dendritic cells (239, 240). While further studies are required to verify this,

experimentation should involve immunophenotyping skin-infiltrating lymphocytes at the

acute phase of SJS/TEN. A recent presentation at the European Academy of Allergy

and Clinical Immunology's 2016 Drug Hypersensitivity Meeting in Malaga, Spain Villani

A.P. et al. 2016 presented data which shows that CD8+ T cells and to a lesser extent

CD4+ T cells, were the main leukocyte subsets found in acute TEN blisters (238).

This study also showed an increase in granulysin expression by CBZ-stimulated CD4+

T cells in Donor 4 (CBZ-SJS non-B*15:02). Donor 4 (CBZ-SJS non-B*15:02), also

exhibited an increased expression of granulysin in unstimulated CD4+ T cells,

supported by the increase in IFN-γ under the same conditions. This is only one patient,

but it supports what Porebski et al. 2013 showed with lamotrigine- SJS/TEN (135).

161

7.3.4 Future Experimentation for HLA-15:02-restricted CBZ-

SJS/TEN

Future work should combine the experiments and results of Chapter 4, to see if the

identified HLA-B*15:02-restricted epitopes elicit granulysin responses in the memory T

cells which express the TCR sequences (Chapter 6) identified in Chapter 5. This

would be strong evidence of the heterologous model. The preliminary work on this has

so far failed to show the epitopes, identified in Chapter 4, elicit the same granulysin

response as seen in CBZ stimulation highlighted in Chapter 5, in memory T cells which

express the TCR sequences shown in Chapter 6 (Appendix 9, Section 9.3 Future

experimentation for HLA-15:02-restricted CBZ-SJS/TEN, Figure 9.16 and Figure 9.17).

Despite this, Chapters 4, 5, and 6 provide a way forward to continue the search for

potential HLA-B*15:02-restricted epitope responses to prove or disprove the

heterologous model.

7.4 Final Conclusions

Not only do ADRs place a great deal of demand and strain on hospitals, healthcare

facilities, auxiliary healthcare services, they are a major cause of iatrogenic and

preventable disease and death (1, 2). The ABC translational roadmap is a successful

example of how the research discoveries can be translated into clinical practice

leading to improved patient care. The ability to identify patients at risk of experiencing

a drug-specific IM-ADR, before administering the drug is an advantageous concept. To

achieve a greater understanding of IM-ADR immunopathogenesis is required. Both

CBZ-SJS/TEN and NVP-SCARs are distinct clinical and immunological syndromes,

but they share are several significant similarities which suggest that the underlying

immunopathogenesis driving SCARs may share commonalities. Both CBZ-SJS/TEN

and NVP-SCARs provide important examples for the immunopathogenesis of severe

T cell-mediated drug reactions.

As with ABC and HLA-B*57:01 genetic screening, the Food and Drug Administration

(USA) recommends genetic screening for HLA-B*15:02 before CBZ therapy and that

all HLA-B*15:02 carriers, should not be administered CBZ unless the benefits

outweigh the risks (101). The low positive predictive value associated with HLA-

B*15:02 carriage alone means numerous HLA-B*15:02 carriers who would otherwise

benefit from CBZ therapy, are required to undertake alternative treatments or risk

experiencing an SJS/TEN reaction. Further insight into the immunopathogenesis is

162

required to improve screening methods which would increase our ability to predict and

avoid potential CBZ-SJS/TEN reactions.

In terms of CBZ-SJS/TEN, we have shown that that increased expression of the

SJS/TEN associated cytotoxic molecule, granulysin in CD8+ T cells can be detected in

an ex-vivo assay in which PBMC’s isolated from CBZ-SJS/TEN patients are exposed

to CBZ. Future experimentation should include CBZ-naive and CBZ-tolerant controls

to determine if the flow cytometry ICS assay described in this body of work is useful in

improving CBZ-SJS/TEN prediction in conjunction with HLA-B*15:02 carriage. The

ddPCR method developed in this body of research should be used in further analysing

the TCR sequences identified in Jing et al. 2016 (231). A greater understanding of

cross-reactive memory T-cell responses would help us understand why not all patients

who carry at-risk alleles develop an IM-ADR. Continued research into the specificities

of the T cells involved in CBZ-SJS/TEN is required, however, the ddPCR method

outlined here can be translated into clinical practice, in conjunction with genetic

screening for HLA-B*15:02 to improve the predictability of IM-ADR.

Future research into the CBZ-SJS/TEN associated T cells should focus on advances

in the ability to characterise individual patient TCR repertoires by means of deep

sequencing, targeting the Vα and Vβ TCR genes and gene expression, and

sequencing based assays which can show paired TCR α- and β-chain sequences.

To add to our understanding of CBZ-SJS/TEN specific cross-reactive memory T-cell

responses we looked at potential epitope-specific T-cell responses. We have also

identified specific HSV-1/HSV-2 responses, with one atypical epitope being shared

among the HLA-B*15:02-CBZ-SJS/TEN patients. We propose that further

experimentation to answer the question of cross-reactivity between drug-derived

neo-antigen and viral-specific memory T cells should include the atypical epitope

identified in this body of work and the HSV/VZV cross-reactive epitope found by

Jing et al. 2016 (231). Continued work should include improving computational and

experimental methods to better identify HLA-restricted epitopes from candidate

pathogens like HHV potentially responsible for the initial priming and maintaining the

cross-reactive pathogenic T cell involved in heterologous driven immunopathogenesis

models.

Combining the experiments of Chapter 4, Chapter 5, and Chapter 6 to identify HLA-

B*15:02-restricted epitopes that elicit granulysin responses in the memory T cells

which express specific TCR sequences would provide strong evidence of the

163

heterologous model. The preliminary work on this has so far failed to show the

epitopes, identified in Chapter 4, elicit the same granulysin response as seen in CBZ

stimulation highlighted in Chapter 5, in memory T cells which express the TCR

sequences shown in Chapter 6 (Appendix 9, Section 9.6 Future experimentation for

HLA-15:02-restricted CBZ-SJS/TEN). The method and experimentation in these

Chapters have provided a way forward to continue the search for potential HLA-

B*15:02-restricted epitope responses to prove or disprove the heterologous model.

In contrast to ABC-HSS and CBZ-SJS/TEN, genetic screening for NVP-HSR

associated HLA alleles has been difficult, due to the number of varied HLA allele and

the associated incomplete predictive values. We proposed that the variation in NVP-

HSR associated HLA alleles was due to several HLA alleles similar HLA binding

cleft/groove properties and/or predicted peptide/NVP-binding characteristics. The

shared association between the F-pocket of the NVP-SCAR predictive allele HLA-

C*04:01 and other ethnic-specific predictive HLA alleles like HLA-C*05:01 and HLA-

C*18:01 was a predictive factor in NVP-SCAR.

164

165

8 Bibliography 1. Lazarou J, Pomeranz BH, Corey PN. Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies. Jama. 1998 Apr 15;279(15):1200-5. 2. Rive C, Bourke J, Phillips EJ. Testing for Drug Hypersensitivity Syndromes. The Clinical Biochemist Reviews. 2013;34(1):15-38. 3. Medicine. Io. Committee on Identifying and Preventing Medication Errors. Preventing Medication Errors. Washington, DC. : The National Academies Press; 2006. 4. Gandhi TK, Weingart SN, Seger AC, Borus J, Burdick E, Poon EG, et al. Outpatient prescribing errors and the impact of computerized prescribing. Journal of general internal medicine. 2005 Sep;20(9):837-41. 5. Bourgeois FT, Shannon MW, Valim C, Mandl KD. Adverse Drug Events in the Outpatient Setting: An 11-Year National Analysis. Pharmacoepidemiology and drug safety. 2010;19(9):901-10. 6. McCarthy R. The price you pay for the drug not taken. Business and health. 1998 Oct;16(10):27-8, 30, 2-3. 7. Krueger KP, Berger BA, Felkey B. Medication adherence and persistence: a comprehensive review. Advances in therapy. 2005 Jul-Aug;22(4):313-56. 8. Alomar MJ. Factors affecting the development of adverse drug reactions (Review article). Saudi Pharmaceutical Journal. 2014 4//;22(2):83-94. 9. Park BK, Pirmohamed M, Kitteringham NR. Role of drug disposition in drug hypersensitivity: a chemical, molecular, and clinical perspective. Chemical research in toxicology. 1998 Sep;11(9):969-88. 10. Lucena MI, Garcia-Cortes M, Cueto R, Lopez-Duran J, Andrade RJ. Assessment of drug-induced liver injury in clinical practice. Fundamental & clinical pharmacology. 2008 Apr;22(2):141-58. 11. Sasidharanpillai S, Riyaz N, Khader A, Rajan U, Binitha MP, Sureshan DN. Severe Cutaneous Adverse Drug Reactions: A Clinicoepidemiological Study. Indian Journal of Dermatology. 2015 Jan-Feb;60(1):102-. 12. Shiina T, Ando A, Suto Y, Kasai F, Shigenari A, Takishima N, et al. Genomic anatomy of a premier major histocompatibility complex paralogous region on chromosome 1q21-q22. Genome Res. 2001 May;11(5):789-802. 13. Urayama KY, Thompson PD, Taylor GM, Trachtenberg EA, Chokkalingam AP. Genetic variation in the extended major histocompatibility complex and susceptibility to childhood acute lymphoblastic leukemia: a review of the evidence. Frontiers in Oncology. [Review]. 2013 2013-December-12;3. 14. Parkin J, Cohen B. An overview of the immune system. Lancet. 2001 Jun 2;357(9270):1777-89. 15. Prugnolle F, Manica A, Charpentier M, Guegan JF, Guernier V, Balloux F. Pathogen-driven selection and worldwide HLA class I diversity. Current biology : CB. 2005 Jun 7;15(11):1022-7. 16. Qutob N, Balloux F, Raj T, Liu H, Marion de Proce S, Trowsdale J, et al. Signatures of historical demography and pathogen richness on MHC class I genes. Immunogenetics. 2012 Mar;64(3):165-75. 17. Cole DK. Re-directing CD4+ T cell responses with the flanking residues of MHC class II-bound peptides: the core is not enough. Frontiers in Immunology. [Review]. 2013 2013-July-1;4. 18. Braciale TJ, Morrison LA, Sweetser MT, Sambrook J, Gething MJ, Braciale VL. Antigen presentation pathways to class I and class II MHC-restricted T lymphocytes. Immunological reviews. 1987 Aug;98:95-114.

166

19. Greenberg S, Silverstein S, Paul W. Fundamental immunology. Fundamental Immunology. 1993:509. 20. Heath WR, Carbone FR. Cross-presentation in viral immunity and self-tolerance. Nat Rev Immunol. [10.1038/35100512]. 2001 11//print;1(2):126-34. 21. Boesteanu A, Brehm M, Mylin LM, Christianson GJ, Tevethia SS, Roopenian DC, et al. A molecular basis for how a single TCR interfaces multiple ligands. Journal of immunology (Baltimore, Md : 1950). 1998 Nov 1;161(9):4719-27. 22. Guermonprez P, Valladeau J, Zitvogel L, Théry C, Amigorena S. Antigen presentation and T cell stimulation by dendritic cells. Annual review of immunology. 2002;20(1):621-67. 23. Donohue KB. Cross-priming utilizes antigen not available to the direct presentation pathway. Immunology. [10.1111/j.1365-2567.2006.02406.x]. 2006 //;119:63-73. 24. Gamadia LE, Remmerswaal EB, Surachno S, Lardy NM, Wertheim-van Dillen PM, van Lier RA, et al. Cross-reactivity of cytomegalovirus-specific CD8+ T cells to allo-major histocompatibility complex class I molecules. Transplantation. 2004 Jun 27;77(12):1879-85. 25. Kobayashi KS, van den Elsen PJ. NLRC5: a key regulator of MHC class I-dependent immune responses. Nat Rev Immunol. [10.1038/nri3339]. 2012 12//print;12(12):813-20. 26. Janeway CA Jr, Travers P, Walport M, al. e. Immunobiology: The Immune System in Health and Disease. . New York: Garland Science; 2001. 27. Chitgopeker P, Sahni D. T-Cell Receptor Gene Rearrangement Detection in Suspected Cases of Cutaneous T-Cell Lymphoma. The Journal of investigative dermatology. [Research Techniques Made Simple]. 2014 04//print;134(4):e19. 28. Joachims ML, Chain JL, Hooker SW, Knott-Craig CJ, Thompson LF. Human αβ and γδ thymocyte development: TCR gene rearrangements, intracellular TCRβ expression, and γδ developmental potential—differences between men and mice. The Journal of Immunology. 2006;176(3):1543-52. 29. Janeway CA Jr, Travers P, Walport M, al. e. Immunobiology: The Immune System in Health and Disease. . New York: Garland Science; 2001. Available from: http://www.ncbi.nlm.nih.gov/books/NBK27136/. 30. Dik WA, Pike-Overzet K, Weerkamp F, de Ridder D, de Haas EF, Baert MR, et al. New insights on human T cell development by quantitative T cell receptor gene rearrangement studies and gene expression profiling. The Journal of experimental medicine. 2005;201(11):1715-23. 31. Janeway CA Jr, Travers P, Walport M, al. e. Immunobiology: The Immune System in Health and Disease. . New York: Garland Science; 2001. Available from: http://www.ncbi.nlm.nih.gov/books/NBK27145/. 32. Germain RN. T-cell development and the CD4-CD8 lineage decision. Nat Rev Immunol. [10.1038/nri798]. 2002 05//print;2(5):309-22. 33. Singer A. New perspectives on a developmental dilemma: the kinetic signaling model and the importance of signal duration for the CD4/CD8 lineage decision. Current opinion in immunology. 2002;14(2):207-15. 34. Liu X, Bosselut R. Duration of TCR signaling controls CD4-CD8 lineage differentiation in vivo. Nature immunology. 2004;5(3):280-8. 35. Branch EI. CD4/CD8 coreceptors in thymocyte development, selection, and lineage commitment: analysis of the CD4/CD8 lineage decision. T Cell Subsets: Cellular Selection, Commitment and Identity. 2004;83:91. 36. Takada K, Jameson SC. Naive T cell homeostasis: from awareness of space to a sense of place. Nature Reviews Immunology. 2009;9(12):823-32.

167

37. Surh CD, Sprent J. Homeostasis of naive and memory T cells. Immunity. 2008;29(6):848-62. 38. Sprent J, Surh CD. Normal T cell homeostasis: the conversion of naive cells into memory-phenotype cells. Nature immunology. 2011;12(6):478-84. 39. Wells A, Gudmundsdottir H, Turka L. Following the fate of individual T cells throughout activation and clonal expansion. Signals from T cell receptor and CD28 differentially regulate the induction and duration of a proliferative response. Journal of Clinical Investigation. 1997;100(12):3173. 40. Gudmundsdottir H, Wells AD, Turka LA. Dynamics and requirements of T cell clonal expansion in vivo at the single-cell level: effector function is linked to proliferative capacity. The Journal of Immunology. 1999;162(9):5212-23. 41. Zhu J, Paul WE. CD4 T cells: fates, functions, and faults. Blood. [10.1182/blood-2008-05-078154]. 2008;112(5):1557-69. 42. Zhu J, Yamane H, Paul WE. Differentiation of Effector CD4 T Cell Populations. Annual review of immunology. 2010;28:445-89. 43. Zhang N, Bevan Michael J. CD8+ T Cells: Foot Soldiers of the Immune System. Immunity. 2011;35(2):161-8. 44. Romagnani S. T-cell subsets (Th1 versus Th2). Annals of allergy, asthma & immunology : official publication of the American College of Allergy, Asthma, & Immunology. 2000 Jul;85(1):9-18; quiz , 21. 45. Garrod Kym R, Moreau Hélène D, Garcia Z, Lemaître F, Bouvier I, Albert Matthew L, et al. Dissecting T Cell Contraction In Vivo Using a Genetically Encoded Reporter of Apoptosis. Cell Reports. 2012 11/29/;2(5):1438-47. 46. Krammer PH, Arnold R, Lavrik IN. Life and death in peripheral T cells. Nat Rev Immunol. 2007 Jul;7(7):532-42. 47. Farber DL, Yudanin NA, Restifo NP. Human memory T cells: generation, compartmentalization and homeostasis. Nat Rev Immunol. [Review]. 2014 01//print;14(1):24-35. 48. Harty JT, Badovinac VP. Shaping and reshaping CD8+ T-cell memory. Nature Reviews Immunology. 2008;8(2):107-19. 49. Joshi NS, Kaech SM. Effector CD8 T cell development: a balancing act between memory cell potential and terminal differentiation. The Journal of Immunology. 2008;180(3):1309-15. 50. Williams MA, Bevan MJ. Effector and memory CTL differentiation. Annu Rev Immunol. 2007;25:171-92. 51. Kaech SM, Cui W. Transcriptional control of effector and memory CD8+ T cell differentiation. Nat Rev Immunol. [10.1038/nri3307]. 2012;12(11):749-61. 52. Rehermann B, Shin EC. Private aspects of heterologous immunity. The Journal of experimental medicine. 2005 Mar 7;201(5):667-70. 53. Welsh RM, Selin LK. No one is naive: the significance of heterologous T-cell immunity. Nat Rev Immunol. 2002 Jun;2(6):417-26. 54. D'Orsogna LJA, Roelen DL, Doxiadis IIN, Claas FHJ. Alloreactivity from human viral specific memory T-cells. Transplant immunology. 2010 8//;23(4):149-55. 55. Landais E, Morice A, Long HM, Haigh TA, Charreau B, Bonneville M, et al. EBV-specific CD4+ T cell clones exhibit vigorous allogeneic responses. Journal of immunology (Baltimore, Md : 1950). 2006 Aug 1;177(3):1427-33. 56. D'Orsogna LJ, Roelen DL, Doxiadis, II, Claas FH. Screening of viral specific T-cell lines for HLA alloreactivity prior to adoptive immunotherapy may prevent GvHD. Transplant immunology. 2011 Apr 15;24(3):141. 57. Koelle DM, Chen HB, McClurkan CM, Petersdorf EW. Herpes simplex virus type 2-specific CD8 cytotoxic T lymphocyte cross-reactivity against prevalent HLA class I alleles. Blood. 2002 May 15;99(10):3844-7.

168

58. Moss P, Khan N. CD8(+) T-cell immunity to cytomegalovirus. Hum Immunol. 2004 May;65(5):456-64. 59. D'Orsogna LJ, Roelen DL, van der Meer-Prins EM, van der Pol P, Franke-van Dijk ME, Eikmans M, et al. Tissue specificity of cross-reactive allogeneic responses by EBV EBNA3A-specific memory T cells. Transplantation. 2011 Mar 15;91(5):494-500. 60. Brook MO, Wood KJ, Jones ND. The impact of memory T cells on rejection and the induction of tolerance. Transplantation. 2006 Jul 15;82(1):1-9. 61. Farcas A, Bojita M. Adverse drug reactions in clinical practice: a causality assessment of a case of drug-induced pancreatitis. J Gastrointestin Liver Dis. 2009;18(3):353-8. 62. Mawson AR, Eriator I, Karre S. Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis (SJS/TEN): Could Retinoids Play a Causative Role? Medical Science Monitor : International Medical Journal of Experimental and Clinical Research. 2015 01/12;21:133-43. 63. Harr T, French LE. Toxic epidermal necrolysis and Stevens-Johnson syndrome. Orphanet journal of rare diseases. 2010;5:39. 64. R S, DA K. Drug allergy: an updated practice parameter. Annals of allergy, asthma & immunology : official publication of the American College of Allergy, Asthma, & Immunology. 2010 Oct;105(4):259-73. 65. Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau JC. Clinical classification of cases of toxic epidermal necrolysis, Stevens-Johnson syndrome, and erythema multiforme. Archives of dermatology. 1993 Jan;129(1):92-6. 66. Solensky R KD, Bernstein IL, Bloomberg GR, Castells MC, Mendelson LM, Weiss ME, Bernstein DI, Blessing-Moore J, Cox L, Lang DM, Nicklas RA, Oppenheinmer J, Portnoy JM, Randolph C, Schuller DE, Spector S, Tilles S, Wallace D, Dowling PJ, Dykewicz M, Greenberger PA, Macy EM, May KR, Nguygen MT, Schwartz L, Solensky R, Khan DA, Bernstein IL, Bloomberg GR, Castells MC, Mendelson L, Weiss ME, Bernstein DI, Blessing-Moore J, Cox L, Lang DM, Nicklas R, Oppenheimer J, Portnoy JM, Randolph C, Schuller DE, Spector SL, Tilles SA, Wallace D, Dowling PJ, Dykewicz MS, Greenberger PA, Macy EM, May KR, Nguyen MT, Schwartz LB. Drug allergy: an updated practice parameter. Annals of allergy, asthma & immunology : official publication of the American College of Allergy, Asthma, & Immunology. 2010 Oct;105(4):259-73. 67. Pavlos R, Mallal S, Phillips E. HLA and pharmacogenetics of drug hypersensitivity. Pharmacogenomics. 2012 Aug;13(11):1285-306. 68. Pavlos R, Mallal S, Ostrov D, Buus S, Metushi I, Peters B, et al. T Cell–Mediated Hypersensitivity Reactions to Drugs. Annual review of medicine. 2015 10/27;66:439-54. 69. Choudhary S, McLeod M, Torchia D, Romanelli P. Drug Reaction with Eosinophilia and Systemic Symptoms (DRESS) Syndrome. The Journal of Clinical and Aesthetic Dermatology. 2013;6(6):31-7. 70. Bocquet H, Bagot M, Roujeau JC. Drug-induced pseudolymphoma and drug hypersensitivity syndrome (Drug Rash with Eosinophilia and Systemic Symptoms: DRESS). Seminars in cutaneous medicine and surgery. 1996 Dec;15(4):250-7. 71. Roujeau JC. Clinical heterogeneity of drug hypersensitivity. Toxicology. 2005;209(2):123-9. 72. Hunziker T, Künzi UP, Braunschweig S, Zehnder D, Hoigné R. Comprehensive hospital drug monitoring (CHDM): adverse skin reactions, a 20-year survey. Allergy. 1997;52(4):388-93. 73. Lee A, Thomson J. Drug-induced skin reactions. Pharmaceut J 1999; 262: 357. 2006;362.

169

74. Phillips E, Mallal SA. HLA and Drug Reactions. In: Mehra NK, Mehra NK, editors. The HLA Complex in Biology and Medicine : A Resource Book. New Delhi, India: Jitendar P Vij, Jaypee Brothers Medical Publishers (P) Ltd; 2010. p. 332-49. 75. Phillips E, Mallal S. Successful translation of pharmacogenetics into the clinic: the abacavir example. Molecular diagnosis & therapy. 2009;13(1):1-9. 76. Phillips EJ, Mallal SA. Pharmacogenetics of drug hypersensitivity. Pharmacogenomics. 2010 Jul;11(7):973-87. 77. Symonds W, Cutrell A, Edwards M, Steel H, Spreen B, Powell G, et al. Risk factor analysis of hypersensitivity reactions to abacavir. Clin Ther. 2002 Apr;24(4):565-73. 78. Phillips EJ, Chung W-H, Mockenhaupt M, Roujeau J-C, Mallal SA. Drug hypersensitivity: pharmacogenetics and clinical syndromes. Journal of Allergy and Clinical Immunology. 2011;127(3):S60-S6. 79. Mallal S, Phillips E, Carosi G, Molina J-M, Workman C, Tomažič J, et al. HLA-B*5701 Screening for Hypersensitivity to Abacavir. New England Journal of Medicine. 2008;358(6):568-79. 80. Phillips EJ, Wong GA, Kaul R, Shahabi K, Nolan DA, Knowles SR, et al. Clinical and immunogenetic correlates of abacavir hypersensitivity. Aids. 2005 Jun 10;19(9):979-81. 81. Saag M, Balu R, Phillips E, Brachman P, Martorell C, Burman W, et al. High sensitivity of human leukocyte antigen-b*5701 as a marker for immunologically confirmed abacavir hypersensitivity in white and black patients. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2008 Apr 1;46(7):1111-8. 82. Lucas A, Lucas M, Strhyn A, Keane NM, McKinnon E, Pavlos R, et al. Abacavir-Reactive Memory T Cells Are Present in Drug Naïve Individuals. PLoS ONE. 2015 02/12;10(2):e0117160. 83. Ostrov DA, Grant BJ, Pompeu YA, Sidney J, Harndahl M, Southwood S, et al. Drug hypersensitivity caused by alteration of the MHC-presented self-peptide repertoire. Proceedings of the National Academy of Sciences. 2012 June 19, 2012;109(25):9959-64. 84. Illing PT, Vivian JP, Dudek NL, Kostenko L, Chen Z, Bharadwaj M, et al. Immune self-reactivity triggered by drug-modified HLA-peptide repertoire. Nature. [10.1038/nature11147]. 2012 06/28/print;486(7404):554-8. 85. Norcross MA, Luo S, Lu L, Boyne MT, Gomarteli M, Rennels AD, et al. Abacavir induces loading of novel self-peptides into HLA-B*57: 01: an autoimmune model for HLA-associated drug hypersensitivity. Aids. 2012 Jul 17;26(11):F21-9. 86. Medicine USNLo. Carbamazepine; Compound summary for CID2554. 2015 [updated 23/02/2015]. 87. Patterson JF. Carbamazepine in the treatment of phantom limb pain. South Med J. 1988 1988/09//;81(9):1100-2. 88. Ambrósio A, Soares-da-Silva P, Carvalho C, Carvalho A. Mechanisms of Action of Carbamazepine and Its Derivatives, Oxcarbazepine, BIA 2-093, and BIA 2-024. Neurochem Res. 2002 2002/02/01;27(1-2):121-30. 89. Davies JA. Mechanisms of action of antiepileptic drugs. Seizure. 1995 12//;4(4):267-71. 90. Bialer M. Why are antiepileptic drugs used for nonepileptic conditions? Epilepsia. 2012 Dec;53 Suppl 7:26-33. 91. Thomas JR, Hobson A, Gray RN, Roy YM, P W, C E. Cochrane Pain, Palliative and Supportive Care Group. The Cochrane Collaboration 2013;1. 92. Health topics/Essential medicines, (2013).

170

93. Chung WH, Hung SI, Hong HS, Hsih MS, Yang LC, Ho HC, et al. Medical genetics: a marker for Stevens-Johnson syndrome. Nature. 2004 Apr 1;428(6982):486. 94. Alfirevic A, Jorgensen AL, Williamson PR, Chadwick DW, Park BK, Pirmohamed M. HLA-B locus in Caucasian patients with carbamazepine hypersensitivity. Pharmacogenomics. 2006 Sep;7(6):813-8. 95. Lonjou C, Borot N, Sekula P, Ledger N, Thomas L, Halevy S, et al. A European study of HLA-B in Stevens–Johnson syndrome and toxic epidermal necrolysis related to five high-risk drugs. Pharmacogenet Genomics. 2008;18:99–107. 96. Man CB, Kwan P, Baum L, Yu E, Lau KM, Cheng AS, et al. Association between HLA-B*1502 allele and antiepileptic drug-induced cutaneous reactions in Han Chinese. Epilepsia. 2007 May;48(5):1015-8. 97. Bureau USC. International Data Base. Washington, D.C.2015 [cited 2016]; Available from: https://www.census.gov/data/developers/data-sets/international-database.html 98. Andreev K, Kantorova V, Bongaarts J. Demographic components of future population growth. Technical Paper. 2013;3. 99. United Nations DoEaSA, Population Division. Demographic Components of Future Population Growth: 2015 Revision. New York: United Nations.; 2015; Available from: unpopulation.org. 100. Khor AH, Lim KS, Tan CT, Wong SM, Ng CC. HLA-B*15:02 association with carbamazepine-induced Stevens-Johnson syndrome and toxic epidermal necrolysis in an Indian population: a pooled-data analysis and meta-analysis. Epilepsia. 2014 Nov;55(11):e120-4. 101. Ferrell PB, McLeod HL. Carbamazepine, HLA-B*1502 and risk of Stevens–Johnson syndrome and toxic epidermal necrolysis: US FDA recommendations. Pharmacogenomics. 2008;9(10):1543-6. 102. Wei C-Y, Chung W-H, Huang H-W, Chen Y-T, Hung S-I. Direct interaction between HLA-B and carbamazepine activates T cells in patients with Stevens-Johnson syndrome. Journal of Allergy and Clinical Immunology. 2012;129(6):1562-9. e5. 103. Chung W-H, Hung S-I. Recent advances in the genetics and immunology of Stevens-Johnson syndrome and toxic epidermal necrosis. Journal of dermatological science. 2012;66(3):190-6. 104. Kaniwa N, Saito Y, Aihara M, Matsunaga K, Tohkin M, Kurose K, et al. HLA-B*1511 is a risk factor for carbamazepine-induced Stevens-Johnson syndrome and toxic epidermal necrolysis in Japanese patients. Epilepsia. 2010 Dec;51(12):2461-5. 105. Kim SH, Lee KW, Song WJ, Kim SH, Jee YK, Lee SM, et al. Carbamazepine-induced severe cutaneous adverse reactions and HLA genotypes in Koreans. Epilepsy Res. 2011 Nov;97(1-2):190-7. 106. Ikeda H, Takahashi Y, Yamazaki E, Fujiwara T, Kaniwa N, Saito Y, et al. HLA class I markers in Japanese patients with carbamazepine-induced cutaneous adverse reactions. Epilepsia. 2010 Feb;51(2):297-300. 107. Ozeki T, Mushiroda T, Yowang A, Takahashi A, Kubo M, Shirakata Y, et al. Genome-wide association study identifies HLA-A*3101 allele as a genetic risk factor for carbamazepine-induced cutaneous adverse drug reactions in Japanese population. Human molecular genetics. 2011 Mar 1;20(5):1034-41. 108. McCormack M, Alfirevic A, Bourgeois S, Farrell JJ, Kasperavičiūtė D, Carrington M, et al. HLA-A* 3101 and carbamazepine-induced hypersensitivity reactions in Europeans. New England Journal of Medicine. 2011;364(12):1134-43. 109. Genin E, Chen DP, Hung SI, Sekula P, Schumacher M, Chang PY, et al. HLA-A*31:01 and different types of carbamazepine-induced severe cutaneous adverse

171

reactions: an international study and meta-analysis. The pharmacogenomics journal. 2014 Jun;14(3):281-8. 110. Hung SI, Chung WH, Jee SH, Chen WC, Chang YT, Lee WR, et al. Genetic susceptibility to carbamazepine-induced cutaneous adverse drug reactions. Pharmacogenetics and genomics. 2006 Apr;16(4):297-306. 111. Nickoloff BJ. Saving the skin from drug-induced detachment. Nat Med. [10.1038/nm1208-1311]. 2008 12//print;14(12):1311-3. 112. Viard I, Wehrli P, Bullani R, Schneider P, Holler N, Salomon D, et al. Inhibition of toxic epidermal necrolysis by blockade of CD95 with human intravenous immunoglobulin. Science (New York, NY). 1998 Oct 16;282(5388):490-3. 113. Nassif A, Bensussan A, Boumsell L, Deniaud A, Moslehi H, Wolkenstein P, et al. Toxic epidermal necrolysis: effector cells are drug-specific cytotoxic T cells. The Journal of allergy and clinical immunology. 2004 Nov;114(5):1209-15. 114. Nassif A, Bensussan A, Dorothee G, Mami-Chouaib F, Bachot N, Bagot M, et al. Drug Specific Cytotoxic T-Cells in the Skin Lesions of a Patient with Toxic Epidermal Necrolysis. 2002 04//print;118(4):728-33. 115. Downey A, Jackson C, Harun N, Cooper A. Toxic epidermal necrolysis: Review of pathogenesis and management. Journal of the American Academy of Dermatology. 2012;66(6):995-1003. 116. Chung W-H, Pan R-Y, Chu M-T, Chin S-W, Huang Y-L, Wang W-C, et al. Oxypurinol-Specific T Cells Possess Preferential TCR Clonotypes and Express Granulysin in Allopurinol-Induced Severe Cutaneous Adverse Reactions. [Original Article]. 2015 06/04/online. 117. Chung WH, Hung SI, Yang JY, Su SC, Huang SP, Wei CY, et al. Granulysin is a key mediator for disseminated keratinocyte death in Stevens-Johnson syndrome and toxic epidermal necrolysis. Nat Med. 2008 Dec;14(12):1343-50. 118. Stenger S, Rosat JP, Bloom BR, Krensky AM, Modlin RL. Granulysin: a lethal weapon of cytolytic T cells. Immunol Today. 1999 Sep;20(9):390-4. 119. Krensky AM, Clayberger C. Granulysin: a novel host defense molecule. American journal of transplantation : official journal of the American Society of Transplantation and the American Society of Transplant Surgeons. 2005 Aug;5(8):1789-92. 120. Sahiratmadja E, Alisjahbana B, Buccheri S, Di Liberto D, de Boer T, Adnan I, et al. Plasma granulysin levels and cellular interferon-gamma production correlate with curative host responses in tuberculosis, while plasma interferon-gamma levels correlate with tuberculosis disease activity in adults. Tuberculosis (Edinburgh, Scotland). 2007 Jul;87(4):312-21. 121. Di Liberto D, Buccheri S, Caccamo N, Meraviglia S, Romano A, Di Carlo P, et al. Decreased serum granulysin levels in childhood tuberculosis which reverse after therapy. Tuberculosis (Edinburgh, Scotland). 2007 Jul;87(4):322-8. 122. Farouk SE, Mincheva-Nilsson L, Krensky AM, Dieli F, Troye-Blomberg M. Gamma delta T cells inhibit in vitro growth of the asexual blood stages of Plasmodium falciparum by a granule exocytosis-dependent cytotoxic pathway that requires granulysin. European journal of immunology. 2004 Aug;34(8):2248-56. 123. Kaspar AA, Okada S, Kumar J, Poulain FR, Drouvalakis KA, Kelekar A, et al. A distinct pathway of cell-mediated apoptosis initiated by granulysin. Journal of immunology (Baltimore, Md : 1950). 2001 Jul 1;167(1):350-6. 124. Pages F, Berger A, Camus M, Sanchez-Cabo F, Costes A, Molidor R, et al. Effector memory T cells, early metastasis, and survival in colorectal cancer. The New England journal of medicine. 2005 Dec 22;353(25):2654-66. 125. Clayberger C, Krensky AM. Granulysin. Current Opinion in Immunology. 2003 10//;15(5):560-5.

172

126. Chung W-H, Hung S-I, Yang J-Y, Su S-C, Huang S-P, Wei C-Y, et al. Granulysin is a key mediator for disseminated keratinocyte death in Stevens-Johnson syndrome and toxic epidermal necrolysis. Nat Med. [10.1038/nm.1884]. 2008 12//print;14(12):1343-50. 127. Appay V. The physiological role of cytotoxic CD4+ T-cells: the holy grail? Clinical and Experimental Immunology. 2004 07/23/accepted;138(1):10-3. 128. Sarwal MM, Jani A, Chang S, Huie P, Wang Z, Salvatierra Jr O, et al. Granulysin expression is a marker for acute rejection and steroid resistance in human renal transplantation. Human Immunology. 2001 1//;62(1):21-31. 129. Raychaudhuri SP, Jiang WY, Raychaudhuri SK, Krensky AM. Lesional T cells and dermal dendrocytes in psoriasis plaque express increased levels of granulysin. Journal of the American Academy of Dermatology. 2004 Dec;51(6):1006-8. 130. Oono T, Morizane S, Yamasaki O, Shirafuji Y, Huh WK, Akiyama H, et al. Involvement of granulysin-producing T cells in the development of superficial microbial folliculitis. Br J Dermatol. 2004 May;150(5):904-9. 131. McInturff JE, Wang SJ, Machleidt T, Lin TR, Oren A, Hertz CJ, et al. Granulysin-derived peptides demonstrate antimicrobial and anti-inflammatory effects against Propionibacterium acnes. The Journal of investigative dermatology. 2005 Aug;125(2):256-63. 132. Kotsch K, Mashreghi MF, Bold G, Tretow P, Beyer J, Matz M, et al. Enhanced granulysin mRNA expression in urinary sediment in early and delayed acute renal allograft rejection. Transplantation. 2004 Jun 27;77(12):1866-75. 133. Mueller H, Fae KC, Magdorf K, Ganoza CA, Wahn U, Guhlich U, et al. Granulysin-expressing CD4+ T cells as candidate immune marker for tuberculosis during childhood and adolescence. PLoS One. 2011;6(12):e29367. 134. Won HK, Lee JW, Song WJ, Klaewsongkram J, Kang MG, Park HK, et al. Lamotrigine-induced toxic epidermal necrolysis confirmed by in vitro granulysin and cytokine assays. Asia Pacific allergy. 2014 Oct;4(4):253-6. 135. Porebski G, Pecaric-Petkovic T, Groux-Keller M, Bosak M, Kawabata TT, Pichler WJ. In vitro drug causality assessment in Stevens–Johnson syndrome – alternatives for lymphocyte transformation test. Clinical & Experimental Allergy. 2013;43(9):1027-37. 136. Bisio F, Masini G, Blasi Vacca E, Calzi A, Cardinale F, Bruzzone B, et al. Effectiveness of a project to prevent HIV vertical transmission in the Republic of Congo. Journal of Antimicrobial Chemotherapy. 2013 April 14, 2013. 137. Coovadia HM, Brown ER, Fowler MG, Chipato T, Moodley D, Manji K, et al. Efficacy and safety of an extended nevirapine regimen in infant children of breastfeeding mothers with HIV-1 infection for prevention of postnatal HIV-1 transmission (HPTN 046): a randomised, double-blind, placebo-controlled trial. The Lancet. 2012;379(9812):221-8. 138. Keane NM, Pavlos RK, McKinnon E, Lucas A, Rive C, Blyth CC, et al. HLA class I restricted CD8+ and class II restricted CD4+ T cells are implicated in the pathogenesis of nevirapine hypersensitivity. AIDS. 2014;28(13):1891-901. 139. Pollard RB, Robinson P, Dransfield K. Safety profile of nevirapine, a nonnucleoside reverse transcriptase inhibitor for the treatment of human immunodeficiency virus infection. Clinical Therapeutics. 1998;20(6):1071-92. 140. Gangar M, Arias G, O'Brien JG, Kemper CA. Frequency of cutaneous reactions on rechallenge with nevirapine and delavirdine. Annals of Pharmacotherapy. 2000;34(7-8):839-42. 141. Keane N. Nevirapine. Reactions Weekly. [journal article]. 2015;1559(1):122-.

173

142. Martin AM, Nolan D, James I, Cameron P, Keller J, Moore C, et al. Predisposition to nevirapine hypersensitivity associated with HLA-DRB1*0101 and abrogated by low CD4 T-cell counts. Aids. 2005 Jan 3;19(1):97-9. 143. Vitezica ZG, Milpied B, Lonjou C, Borot N, Ledger TN, Lefebvre A, et al. HLA-DRB1*01 associated with cutaneous hypersensitivity induced by nevirapine and efavirenz. Aids. 2008 Feb 19;22(4):540-1. 144. Gonzalez-Galarza FF, Christmas S, Middleton D, Jones AR. Allele frequency net: a database and online repository for immune gene frequencies in worldwide populations. Nucleic Acids Research. 2011 January 1, 2011;39(suppl 1):D913-D9. 145. Phillips E BJ, Sanne I, Lederman M, Hinkle J, Rousseau F, I J, et al. Associations between HLA-DRBA*0102, HLA-B*5801 and hepatotoxicity in patients who initiated nevirapine containing regimens in South Africa. 18th Conference on retroviruses and opportunistic infections [Conference proceeding]; Boston, MA, USA. Feb 27-March 1; 2011. 146. Littera R, Carcassi C, Masala A, Piano P, Serra P, Ortu F, et al. HLA-dependent hypersensitivity to nevirapine in Sardinian HIV patients. Aids. 2006 Aug 1;20(12):1621-6. 147. Gatanaga H, Yazaki H, Tanuma J, Honda M, Genka I, Teruya K, et al. HLA-Cw8 primarily associated with hypersensitivity to nevirapine. Aids. 2007 Jan 11;21(2):264-5. 148. Gao S, Gui XE, Liang K, Liu Z, Hu J, Dong B. HLA-Dependent Hypersensitivity Reaction to Nevirapine in Chinese Han HIV-Infected Patients. AIDS Res Hum Retroviruses. 2011:540-43. 149. Yuan J, Guo S, Hall D, Cammett AM, Jayadev S, Distel M, et al. Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African, Asian, and European descent. AIDS. 2011;25(Nevirapine Toxicogenomics Study Team). 150. Phillips EJ, Keane N, Blyth C. Both HLA Class I CD8+restricted and HLA Class II CD4+ restriced T Cells are implicated in the pathogenises of nevirapine hypersensitivity 51st ICAAC; Chicago IL, USA2011. 151. Puangpetch A, Koomdee N, Chamnanphol M, Jantararoungtong T, Santon S, Prommas S, et al. HLA-B allele and haplotype diversity among Thai patients identified by PCR-SSOP: evidence for high risk of drug-induced hypersensitivity. Frontiers in Genetics. 2014;5. 152. Chantarangsu S, Mushiroda T, Mahasirimongkol S, Kiertiburanakul S, Sungkanuparph S, Manosuthi W, et al. HLA-B*3505 allele is a strong predictor for nevirapine-induced skin adverse drug reactions in HIV-infected Thai patients. Pharmacogenetics and genomics. 2009 Feb;19(2):139-46. 153. Likanonsakul S, Rattanatham T, Feangvad S, Uttayamakul S, Prasithsirikul W, Tunthanathip P, et al. HLA-Cw*04 allele associated with nevirapine-induced rash in HIV-infected Thai patients. AIDS research and therapy. 2009;6(1):22. 154. Bonjoch A, Paredes R, Galvez J, Miralles C, Videla S, Martínez E, et al. Antiretroviral Treatment Simplification With 3 NRTIs or 2 NRTIs Plus Nevirapine in HIV-1-Infected Patients Treated With Successful First-Line HAART. JAIDS Journal of Acquired Immune Deficiency Syndromes. 2005;39(3):313-6. 155. Carr DF, Chaponda M, Jorgensen AL, Castro EC, van Oosterhout JJ, Khoo SH, et al. Association of human leukocyte antigen alleles and nevirapine hypersensitivity in a Malawian HIV-infected population. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2013 May;56(9):1330-9. 156. Pichler WJ. The pi concept: pharmacological interaction of drugs with immune receptors. The World Allergy Organization journal. 2008;1(6):96.

174

157. Adam J, Pichler WJ, Yerly D. Delayed drug hypersensitivity: models of T‐cell stimulation. British journal of clinical pharmacology. 2011;71(5):701-7. 158. Gerber BO, Pichler WJ. Cellular mechanisms of T cell mediated drug hypersensitivity. Current opinion in immunology. 2004;16(6):732-7. 159. Zanni MP, von Greyerz S, Schnyder B, Brander KA, Frutig K, Hari Y, et al. HLA-restricted, processing- and metabolism-independent pathway of drug recognition by human alpha beta T lymphocytes. J Clin Invest. 1998 Oct 15;102(8):1591-8. 160. Zanni MP, von Greyerz S, Schnyder B, Wendland T, Pichler WJ. Allele-unrestricted presentation of lidocaine by HLA-DR molecules to specific alphabeta+ T cell clones. International immunology. 1998;10(4):507-15. 161. Schnyder B, Mauri-Hellweg D, Zanni M, Bettens F, Pichler WJ. Direct, MHC-dependent presentation of the drug sulfamethoxazole to human alphabeta T cell clones. Journal of Clinical Investigation. 1997;100(1):136. 162. Pavlos R, Mallal S, Ostrov D, Pompeu Y, Phillips E. Fever, rash and systemic symptoms: understanding the role of virus and HLA in severe cutaneous drug allergy. The journal of allergy and clinical immunology In practice. 2014 Jan-Feb;2(1):21-33. 163. Amir AL, D'Orsogna LJ, Roelen DL, van Loenen MM, Hagedoorn RS, de Boer R, et al. Allo-HLA reactivity of virus-specific memory T cells is common. Blood. 2010 Apr 15;115(15):3146-57. 164. Amir AL, van der Steen DM, Hagedoorn RS, Kester MG, van Bergen CA, Drijfhout JW, et al. Allo-HLA-reactive T cells inducing graft-versus-host disease are single peptide specific. Blood. 2011 Dec 22;118(26):6733-42. 165. Deckers JG, Boonstra JG, Van der Kooij SW, Daha MR, Van der Woude FJ. Tissue-specific characteristics of cytotoxic graft-infiltrating T cells during renal allograft rejection. Transplantation. 1997 Jul 15;64(1):178-81. 166. Koelle DM, Chen HB, McClurkan CM, Petersdorf EW. Herpes simplex virus type 2–specific CD8 cytotoxic T lymphocyte cross-reactivity against prevalent HLA class I alleles. Presented in part at the 25th International Herpesvirus Workshop, Portland, OR, July 2000 [abstract 516]. 2002 2002-05-15 00:00:00;99(10):3844-7. 167. White KD, Chung WH, Hung SI, Mallal S, Phillips EJ. Evolving models of the immunopathogenesis of T cell-mediated drug allergy: The role of host, pathogens, and drug response. The Journal of allergy and clinical immunology. 2015 Aug;136(2):219-34. 168. Thomsen M, Lundegaard C, Buus S, Lund O, Nielsen M. MHCcluster, a method for functional clustering of MHC molecules. Immunogenetics. 2013 06/18;65(9):655-65. 169. Hoof I, Peters B, Sidney J, Pedersen LE, Sette A, Lund O, et al. NetMHCpan, a method for MHC class I binding prediction beyond humans. Immunogenetics. 2009 Jan;61(1):1-13. 170. Nielsen M, Lundegaard C, Blicher T, Lamberth K, Harndahl M, Justesen S, et al. NetMHCpan, a Method for Quantitative Predictions of Peptide Binding to Any HLA-A and -B Locus Protein of Known Sequence. PLoS ONE. 2007;2(8):e796. 171. Sidney J, Peters B, Frahm N, Brander C, Sette A. HLA class I supertypes: a revised and updated classification. BMC immunology. 2008;9:1. 172. Lund O, Nielsen M, Kesmir C, Petersen AG, Lundegaard C, Worning P, et al. Definition of supertypes for HLA molecules using clustering of specificity matrices. Immunogenetics. 2004 Mar;55(12):797-810. 173. Hanafusa T, Azukizawa H, Matsumura S, Katayama I. The predominant drug-specific T-cell population may switch from cytotoxic T cells to regulatory T cells during the course of anticonvulsant-induced hypersensitivity. J Dermatol Sci. 2012 Mar;65(3):213-9.

175

174. Chang CC, Too CL, Murad S, Hussein SH. Association of HLA-B*1502 allele with carbamazepine-induced toxic epidermal necrolysis and Stevens-Johnson syndrome in the multi-ethnic Malaysian population. International journal of dermatology. 2011 Feb;50(2):221-4. 175. Warren R.L., Freeman JD, Zeng T, Choe G, Munro S, Moore R, et al. Exhaustive T-cell repertoire sequencing of human peripheral blood samples reveals signatures of antigen selection and a directly measured repertoire size of at least 1 million clonotypes. Genome Res. 2011; 21:790-7. 176. Ko TM, Chung WH, Wei CY, Shih HY, Chen JK, Lin CH, et al. Shared and restricted T-cell receptor use is crucial for carbamazepine-induced Stevens-Johnson syndrome. The Journal of allergy and clinical immunology. 2011 Dec;128(6):1266-76.e11. 177. Kappelhoff BS, van Leth F, MacGregor TR, Lange J, Beijnen JH, Huitema AD. Nevirapine and efavirenz pharmacokinetics and covariate analysis in the 2NN study. Antiviral therapy. 2005;10(1):145-55. 178. García-Piñeres AJ, Hildesheim A, Williams M, Trivett M, Strobl S, Pinto LA. DNAse treatment following thawing of cryopreserved PBMC is a procedure suitable for lymphocyte functional studies. Journal of immunological methods. 2006;313(1):209-13. 179. Costantini A, Mancini S, Giuliodoro S, Butini L, Regnery CM, Silvestri G, et al. Effects of cryopreservation on lymphocyte immunophenotype and function. Journal of Immunological Methods. 2003 7//;278(1–2):145-55. 180. BD. BD FLUORESCENCE SPECTRUM VIEWER A MULTICOLOR TOOL. San Jose: BD Biosciences; 2016 [cited 2013-2015]; Available from: http://www.bdbiosciences.com/anz/research/multicolor/spectrum_viewer. 181. Hogg AE, Bowick GC, Herzog NK, Cloyd MW, Endsley JJ. Induction of granulysin in CD8+ T cells by IL-21 and IL-15 is suppressed by human immunodeficiency virus-1. Journal of Leukocyte Biology. 2009 November 1, 2009;86(5):1191-203. 182. Fischer AH, Jacobson KA, Rose J, Zeller R. Fixation and permeabilization of cells and tissues. CSH protocols. 2008;2008:pdb.top36. 183. Leisner C, Loeth N, Lamberth K, Justesen S, Sylvester-Hvid C, Schmidt EG, et al. One-pot, mix-and-read peptide-MHC tetramers. PLoS One. 2008;3(2):e1678. 184. Svitek N, Hansen AM, Steinaa L, Saya R, Awino E, Nielsen M, et al. Use of “one-pot, mix-and-read” peptide-MHC class I tetramers and predictive algorithms to improve detection of cytotoxic T lymphocyte responses in cattle. Veterinary Research. 2014 04/28;45(1):50. 185. Menezes J, Jondal M, Leibold W, Dorval G. Epstein-Barr virus interactions with human lymphocyte subpopulations: virus adsorption, kinetics of expression of Epstein-Barr virus-associated nuclear antigen, and lymphocyte transformation. Infection and Immunity. 1976;13(2):303-10. 186. Uphoff CC, Drexler HG. Detecting mycoplasma contamination in cell cultures by polymerase chain reaction. Methods Mol Biol. 2011;731:93-103. 187. Boslett B, Nag S, Resnick A. Detection and Antibiotic Treatment of Mycoplasma arginini Contamination in a Mouse Epithelial Cell Line Restore Normal Cell Physiology. BioMed Research International. 2014;2014:7. 188. Vita R ZL, Greenbaum JA, Emami H, Hoof I, Salimi N, Damle R, Sette A, Peters B. T. The immune epitope database 2.0. Nucleic Acids Res 2010;JAN 38:(Database issue):D854-62. 189. Tatusova T, Ciufo S, Fedorov B, O'Neill K, Tolstoy I. RefSeq microbial genomes database: new representation and annotation strategy. Nucleic Acids Res. 2014 Jan;42(Database issue):D553-9.

176

190. Hertz T, Nolan D, James I, John M, Gaudieri S, Phillips E, et al. Mapping the landscape of host-pathogen coevolution: HLA class I binding and its relationship with evolutionary conservation in human and viral proteins. Journal of virology. 2011;85(3):1310-21. 191. Wade JA, Katovich Hurley C, Takemoto SK, Thompson J, Davies SM, Fuller TC, et al. HLA mismatching within or outside of cross-reactive groups (CREGs) is associated with similar outcomes after unrelated hematopoietic stem cell transplantation. Blood. 2007 2007-05-01 00:00:00;109(9):4064-70. 192. Madden DR. The three-dimensional structure of peptide-MHC complexes. Annu Rev Immunol. 1995;13:587-622. 193. Guo HC, Jardetzky TS, Garrett TP, Lane WS, Strominger JL, Wiley DC. Different length peptides bind to HLA-Aw68 similarly at their ends but bulge out in the middle. Nature. 1992 Nov 26;360(6402):364-6. 194. Saper MA, Bjorkman PJ, Wiley DC. Refined structure of the human histocompatibility antigen HLA-A2 at 2.6 A resolution. Journal of molecular biology. 1991 May 20;219(2):277-319. 195. Madden DR, Gorga JC, Strominger JL, Wiley DC. The three-dimensional structure of HLA-B27 at 2.1 A resolution suggests a general mechanism for tight peptide binding to MHC. Cell. 1992 Sep 18;70(6):1035-48. 196. Ridker PM, Rifai N, Stampfer MJ, Hennekens CH. Plasma concentration of interleukin-6 and the risk of future myocardial infarction among apparently healthy men. Circulation. 2000 Apr 18;101(15):1767-72. 197. Schönrock LM, Gawlowski G, Brück W. Interleukin-6 expression in human multiple sclerosis lesions. Neuroscience Letters. 2000 11/10/;294(1):45-8. 198. Tackey E, Lipsky PE, Illei GG. Rationale for interleukin-6 blockade in systemic lupus erythematosus. Lupus. 2004 2004;13(5):339-43. 199. Sculco PK, McLawhorn AS, Desai N, Su EP, Padgett DE, Jules-Elysee K. The Effect of Perioperative Corticosteroids in Total Hip Arthroplasty: A Prospective Double-Blind Placebo Controlled Pilot Study. The Journal of Arthroplasty.31(6):1208-12. 200. Ogata M, Satou T, Kawano R, Takakura S, Goto K, Ikewaki J, et al. Correlations of HHV-6 viral load and plasma IL-6 concentration with HHV-6 encephalitis in allogeneic stem cell transplant recipients. Bone marrow transplantation. 2010 Jan;45(1):129-36. 201. Pritchett JC, Nanau RM, Neuman MG. The Link between Hypersensitivity Syndrome Reaction Development and Human Herpes Virus-6 Reactivation. International journal of hepatology. 2012;2012:723062. 202. Criado PR, Criado RFJ, Avancini JdM, Santi CG. Drug reaction with Eosinophilia and Systemic Symptoms (DRESS) / Drug-induced Hypersensitivity Syndrome (DIHS): a review of current concepts. Anais Brasileiros de Dermatologia. 2012;87:435-49. 203. Kimura A, Naka T, Nohara K, Fujii-Kuriyama Y, Kishimoto T. Aryl hydrocarbon receptor regulates Stat1 activation and participates in the development of Th17 cells. Proceedings of the National Academy of Sciences. 2008;105(28):9721-6. 204. Luckheeram RV, Zhou R, Verma AD, Xia B. CD4+T Cells: Differentiation and Functions. Clinical and Developmental Immunology. 2012;2012:12. 205. Ott PA, Hodi FS, Robert C. CTLA-4 and PD-1/PD-L1 blockade: new immunotherapeutic modalities with durable clinical benefit in melanoma patients. Clinical Cancer Research. 2013;19(19):5300-9. 206. Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, et al. Safety, activity, and immune correlates of anti–PD-1 antibody in cancer. New England Journal of Medicine. 2012;366(26):2443-54.

177

207. Brahmer JR, Tykodi SS, Chow LQ, Hwu W-J, Topalian SL, Hwu P, et al. Safety and activity of anti–PD-L1 antibody in patients with advanced cancer. New England Journal of Medicine. 2012;366(26):2455-65. 208. Thompson CB, Allison JP. The Emerging Role of CTLA-4 as an Immune Attenuator. Immunity. 1997;7(4):445-50. 209. Wang Q, Zhou JQ, Zhou LM CZ, Fang ZY, Chen SD, Yang LB, et al. Association between HLA-B*1502 allele and carbamazepine-induced severe cutaneous adverse reactions in Han people of southern China mainland. Seizure. 2011;20:446–8. 210. Chung WH, Hung SI, Hong HS, Hsih MS, Yang LC, Ho HC, et al. Medical genetics: a marker for Stevens–Johnson syndrome. Nature. 2004;428 486. 211. Li L, Yang B, Yu S, Zhang X, Lao S, Wu C. Human CD8+ T cells from TB pleurisy respond to four immunodominant epitopes in Mtb CFP10 restricted by HLA-B alleles. PLoS One. 2013;8(12):e82196. 212. Rao X, Hoof I Fau - Costa AICAF, Costa Ai Fau - van Baarle D, van Baarle D Fau - Kesmir C, Kesmir C. HLA class I allele promiscuity revisited. 2011(1432-1211 (Electronic)). 213. Frahm N, Yusim K, Suscovich TJ, Adams S, Sidney J, Hraber P, et al. Extensive HLA class I allele promiscuity among viral CTL epitopes. European journal of immunology. 2007;37(9):2419-33. 214. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997 Sep 1;25(17):3389-402. 215. Golovina TN, Morrison SE, Eisenlohr LC. The Impact of Misfolding versus Targeted Degradation on the Efficiency of the MHC Class I-Restricted Antigen Processing. The Journal of Immunology. 2005 March 1, 2005;174(5):2763-9. 216. Khan S, de Giuli R, Schmidtke G, Bruns M, Buchmeier M, van den Broek M, et al. Cutting Edge: Neosynthesis Is Required for the Presentation of a T Cell Epitope from a Long-Lived Viral Protein. The Journal of Immunology. 2001 November 1, 2001;167(9):4801-4. 217. Princiotta MF, Finzi D, Qian S-B, Gibbs J, Schuchmann S, Buttgereit F, et al. Quantitating Protein Synthesis, Degradation, and Endogenous Antigen Processing. Immunity. 2003 3//;18(3):343-54. 218. Schubert U, Antón LC, Gibbs J, Norbury CC, Yewdell JW, Bennink JR. Rapid degradation of a large fraction of newly synthesized proteins by proteasomes. Nature. [10.1038/35008096]. 2000 04/13/print;404(6779):770-4. 219. Reits EAJ, Vos JC, Gromme M, Neefjes J. The major substrates for TAP invivo are derived from newly synthesized proteins. Nature. [10.1038/35008103]. 2000 04/13/print;404(6779):774-8. 220. Mayrand S-M, Green WR. Non-traditionally derived CTL epitopes: exceptions that prove the rules? Immunology Today. 1998;19(12):551-6. 221. Yewdell JW, Anton LC, Bennink JR. Defective ribosomal products (DRiPs): a major source of antigenic peptides for MHC class I molecules? Journal of immunology (Baltimore, Md : 1950). 1996 Sep 1;157(5):1823-6. 222. Wang RF, Parkhurst MR, Kawakami Y, Robbins PF, Rosenberg SA. Utilization of an alternative open reading frame of a normal gene in generating a novel human cancer antigen. The Journal of experimental medicine. 1996 Mar 1;183(3):1131-40. 223. Shastri N, Nguyen V, Gonzalez F. Major histocompatibility class I molecules can present cryptic translation products to T-cells. The Journal of biological chemistry. 1995 Jan 20;270(3):1088-91. 224. Bullock TN, Patterson AE, Franlin LL, Notidis E, Eisenlohr LC. Initiation codon scanthrough versus termination codon readthrough demonstrates strong potential for

178

major histocompatibility complex class I-restricted cryptic epitope expression. The Journal of experimental medicine. 1997 Oct 6;186(7):1051-8. 225. Bullock TN, Eisenlohr LC. Ribosomal scanning past the primary initiation codon as a mechanism for expression of CTL epitopes encoded in alternative reading frames. The Journal of experimental medicine. 1996 Oct 1;184(4):1319-29. 226. Fetten JV, Roy N, Gilboa E. A frameshift mutation at the NH2 terminus of the nucleoprotein gene does not affect generation of cytotoxic T lymphocyte epitopes. Journal of immunology (Baltimore, Md : 1950). 1991 Oct 15;147(8):2697-705. 227. Berglund P, Finzi D, Bennink JR, Yewdell JW. Viral alteration of cellular translational machinery increases defective ribosomal products. Journal of virology. 2007 Jul;81(13):7220-9. 228. Netzer N, Goodenbour JM, David A, Dittmar KA, Jones RB, Schneider JR, et al. Innate immune and chemically triggered oxidative stress modifies translational fidelity. Nature. 2009 Nov 26;462(7272):522-6. 229. Yewdell JW, Nicchitta CV. The DRiP hypothesis decennial: support, controversy, refinement and extension. Trends in immunology. 2006 Aug;27(8):368-73. 230. Yewdell JW. DRiPs Solidify: Progress in Understanding Endogenous MHC Class I Antigen Processing. Trends in immunology. 2011 09/29;32(11):548-58. 231. Nielsen M, Lundegaard C, Lund O, Kesmir C. The role of the proteasome in generating cytotoxic T-cell epitopes: insights obtained from improved predictions of proteasomal cleavage. Immunogenetics. 2005 Apr;57(1-2):33-41. 232. Skjøt RLV, Brock I, Arend SM, Munk ME, Theisen M, Ottenhoff THM, et al. Epitope Mapping of the Immunodominant Antigen TB10.4 and the Two Homologous Proteins TB10.3 and TB12.9, Which Constitute a Subfamily of the esat-6 Gene Family. Infection and Immunity. 2002;70(10):5446-53. 233. Jing L, Laing KJ, Dong L, Russell RM, Barlow RS, Haas JG, et al. Extensive CD4 and CD8 T Cell Cross-Reactivity between Alphaherpesviruses. The Journal of Immunology. 2016 March 1, 2016;196(5):2205-18. 234. Folch G, Lefranc MP. The human T cell receptor beta variable (TRBV) genes. Experimental and clinical immunogenetics. 2000;17(1):42-54. 235. Koning D, Costa AI, Hasrat R, Grady BP, Spijkers S, Nanlohy N, et al. In vitro expansion of antigen-specific CD8+ T cells distorts the T-cell repertoire. Journal of immunological methods. 2014;405:199-203. 236. Adam J, Eriksson KK, Schnyder B, Fontana S, Pichler WJ, Yerly D. Avidity determines T-cell reactivity in abacavir hypersensitivity. European journal of immunology. 2012 Jul;42(7):1706-16. 237. Zhu J, Peng T, Johnston C, Phasouk K, Kask AS, Klock A, et al. Immune surveillance by CD8alphaalpha+ skin-resident T cells in human herpes virus infection. Nature. 2013 May 23;497(7450):494-7. 238. Pittet MJ, Speiser DE, Valmori D, Cerottini J-C, Romero P. Cutting edge: cytolytic effector function in human circulating CD8+ T cells closely correlates with CD56 surface expression. The Journal of Immunology. 2000;164(3):1148-52. 239. Ogawa K, Takamori Y, Suzuki K, Nagasawa M, Takano S, Kasahara Y, et al. Granulysin in human serum as a marker of cell-mediated immunity. European journal of immunology. 2003;33(7):1925-33. 240. Villani A.P., Rozières A., Bensaïd B., Tardieu M., Albert F., Mutez V., et al. Massive Expansion Of Clonotypic And Polycytotoxic CD8+ T Cells In Toxic Epidermal Necrolysis. 7th Drug Hypersensitivity Meeting (DHM) 21st - 23rd April, 2016; Malaga, Spain Clinical and Translational Allergy; 2016. 241. Rebecca Pavlos CR, Elizabeth McKinnon, David Ostrov, Pablo C. Plascencia, Bjoern Peters, Jing Yuan, Silvana Gaudieri, David Haas, Simon Mallal, Elizabeth J.

179

Phillips, editor. Specific Binding Characteristics of HLA Alleles Associate with Nevirapine Hypersensitivity. Conference on Retroviruses and Opportunistic Infections (CROI), At Boston, Massachusetts; 2014. 242. Rive C, Pavlos R, Ostrov D, Plasencia P, Hung S-I, Chung W, et al. HLA class I-drug-T-cell receptor interactions in SJS/TEN. Clinical and Translational Allergy. 2014;4(3).

180

181

9 Appendix 9.1 Nevirapine isolation from Viramune tablets

One tablet of Viramune contains 200mg of NVP; one tablet of Viramune extended-

release (XR) contains 400mg of NVP. NVP extraction from its tablet form requires the

separation of NVP from the drug excipients. In a sterile environment, like that of a

class-II biological hood:

A 200mg NVP tablet was crushed into a fine powder, transferred into a sterile 50

mL Schott bottle (autoclaved), and diluted in 40 mL of DMSO (5mg of NVP per mL

of DMSO).

Tablet/DMSO solution was then transferred to the appropriate tubes, centrifuged at

7224 RCF (10,000 RPM) for 5 minutes.

Supernatant, containing NVP dissolved in DMSO, was then transferred to a sterile

50 mL Schott bottle (autoclaved).

NVP/DMSO solution was diluted with sterile 1x PBS, giving a final concentration of

1 mg/mL (1:5 dilution).

NVP solution was then filtered using a 0.22 μm filter and transferred to a correctly

labelled sterile 50 mL Schott bottle (autoclaved) wrapped in aluminium foil, and

stored in a dark environment at room temperature. As a means of quality control

any NVP made and stored in this manner would be disposed of after 2 weeks.

Check on Nano drop for absorbance at 260nm, should read at 6.0 (For NanoDrop,

make up three Eppendorf tubes).

o Tube 1 = 750uls of PBS alone

o Tube 2 = 750uls of PBS plus 250 of DMSO

o Tube 3 = 750uls of PBS plus 250 of NVP 1mg/mL

182

9.2 Chapter 5 Specific T-cell clonotype and

detection by ddPCR

9.2.1 Reference Sequence Amplicon Analysis QPCR Primer

Validation

The first step in setting up a TCR CDR3 sequence specific PCR assay involves

designing the oligonucleotide primer sets and optimising various aspects of the assay,

including the optimal thermocycling conditions. This was done using a real time,

quantitative PCR assay and the EvaGreen master mix. Firstly multiple variations of the

primer sets were designed, but failed and were excluded. The primer sets and master

mix highlighted Chapter 2 Methods and Materials, Section 2.4.2 Digital Droplet PCR

Table 2.15, Table 2.17, and Table 2.17 were used, and thermocycling conditions

optimised. This included analysing various annealing temperatures. As shown in

Figure 9.2 and Figure 9.2, it was determined that an annealing temperature of 58°C for

30 sec was better for all this specific ddPCR assay, in preference to the manufactures

recommended protocol of having a 60 sec annealing and denaturing combined step at

60°C.

Figure 9.1: Analysis of specific Vβ CDR3 primers using serial dilutions of the reference sequence

amplicons, with the annealing temperature set at 58◦C

183

Figure 9.2: Analysis of specific Vβ CDR3 primers using serial dilutions of the reference sequence

amplicons, with the annealing temperature set at and 60◦C.

9.2.2 The ddPCR Sensitivity Analysis

The sensitivity of the ddPCR assay was assessed using the serially diluted synthetic

construct amplicons. This analysis confirms the manufactures claims that the ddPCR

method is sensitive enough to detect single copies of target DNA. The range of the

ddPCR copy number detection, using the EvaGreen ddPCR assay is sensitive enough

to detect DNA from 105 copies of DNA, down to the single copy level. This was shown

in the analysis of serially diluted synthetic construct amplicons (Figure 9.3 to Figure

9.5).

184

Figure 9.3: the QuantaSoft raw digital droplet PCR plots. Reference sequence amplicons, serially

diluted to test the sensitivity of the digital droplet PCR and qualifying the assay parameters.

185

Figure 9.4: the QuantaSoft digital droplet PCR output of positive and total events recorded for the raw

plots outlined in Figure 9.10 Reference sequence amplicons, serially diluted to test the sensitivity of the

digital droplet PCR and qualifying the assay parameters.

Figure 9.5: Digital droplet PCR Reference sequence amplicons serial dilution result and QuantaSoft

output (copies/μL). Reference sequence amplicons, serially diluted to test the sensitivity of the digital

droplet PCR and qualifying the assay parameters.

186

9.3 Future experimentation for HLA-15:02-restricted

CBZ-SJS/TEN

9.3.1 Preliminary granulysin expression data

Stimulation of Donor PBMC’s with CBZ and with the HLA-B*15:02-restricted peptides

identified in Chapter 4 Predicted HLA-Restricted Peptide Screening, failed to show

increased granulysin expression in peptide stimulation CD8+ T cells significantly above

that of the unstimulated response, matching that of the CBZ-stimulated CD8+ T cells

(Figure 9.16).

Figure 9.6: Mean (+/- SD) Granulysin expression by CD8+ T cells, stimulated by the epitopes discovered

in Chapter 4 Predicted HLA-Restricted Peptide Screening (10μg/mL) and CBZ (10μg/mL).

187

9.3.2 Preliminary TCR expression data

Stimulation of Donor PBMC’s with CBZ and with the HLA-B*15:02-restricted peptides

identified in Chapter 4 Predicted HLA-Restricted Peptide Screening, failed to show an

expansion of TCR Vβ-12 CDR3 LAGELF and Vβ-11 CDR3 ISGSY in Donor 1 (CBZ-

TEN) matching that of the CBZ-stimulated response (Figure 9.17).

Figure 9.7: TCR Vβ-12 CDR3 LAGELF and Vβ-11 CDR3 ISGSY sequence profile of Donor 1 (CBZ-

TEN) when stimulated with stimulated by the epitopes discovered in Chapter 4 Predicted HLA-

Restricted Peptide Screening (10μg/mL) and CBZ (10μg/mL). Raw data analysed using Bio-Rad


(%) of normalised Cβ expression.

188

9.4 Publication and Conference Abstracts

9.4.1 Journal Publications

Title: HLA Class-I restricted CD8 (+) and Class-II restricted CD4 (+) T cells are

implicated in the Pathogenesis of Nevirapine Hypersensitivity.

DOI:10.1097/QAD.0000000000000345 AIDS, London, 24 August 2014 - Volume 28 -

Issue 13 - p 1891–1901.

Niamh M Keanea*, Rebecca K Pavlosa*, Elizabeth McKinnona*, Andrew Lucasa, Craig

Rivea, Christopher C Blythb,c, David Dunna, Michaela Lucasd, Simon Mallala,e,

Elizabeth Phillipsa,e: aInstitute for Immunology and Infectious Diseases, Murdoch

University, bDepartment of Paediatrics and Adolescent Medicine, Princess Margaret

Hospital for Children, cSchool of Paediatrics & Child Health, dSchool of Medicine &

Pharmacology and School of Pathology & Laboratory Medicine, The University of

Western Australia, Perth, Western Australia, Australia, and eVanderbilt University

School of Medicine, Nashville, Tennessee, USA

Figure 9.8: Abstract of HLA Class-I restricted CD8 (+) and Class-II restricted CD4 (+) T cells are

implicated in the Pathogenesis of Nevirapine Hypersensitivity paper.

Published in AIDS, London, 24 August 2014 - Volume 28 - Issue 13 - p 1891–1901.

DOI:10.1097/QAD.0000000000000345 (138).

Title: Testing for Drug Hypersensitivity Syndromes.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3626363/ The Clinical biochemist.

Reviews / Australian Association of Clinical Biochemists 02/2013; 34(1):15-38.

189

Craig M Rive1, Jack Bourke2, Elizabeth J Phillips1-4: 1Institute for Immunology and

Infectious Diseases, Murdoch University, 2Department of Clinical Immunology &

Immunogenetics, Royal Perth Hospital, Western Australia; 3Department of Clinical

Immunology & Infectious Diseases, Sir Charles Gairdner Hospital, Western Australia; 4Pathology & Laboratory Medicine, University of Western Australia, Australia

Figure 9.9: Abstract of Testing for Drug Hypersensitivity Syndromes.

Published in The Clinical biochemist. Reviews / Australian Association of Clinical Biochemists 02/2013;

34(1):15-38. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3626363/ (2).

9.4.2 Publications in development

1. Title: Peptide-Binding Characteristics of HLA Alleles Associate with Cutaneous

Nevirapine Hypersensitivity – Co-author based on Chapter 3: Nevirapine.

2. Title: The use of droplet digital PCR to explore shared and restricted T-cell

receptor use in HLA-B*15:02 driven, carbamazepine-induced Steven-Johnson

syndrome / Toxic epidermal necrolysis – Primary author based on Chapter 5 -

Carbamazepine induced Stevens-Johnson syndrome and toxic epidermal

necrolysis Immunopathogenesis: Specific T-cell clonotypes.

190

9.4.3 Conference Proceedings

9.4.3.1 Posters

Title: HLA Class-I-drug-T-cell receptor interaction and cytokine expression in SJS/TEN.

DOI: 10.13140/RG.2.1.4937.5207 Science on the Swan, Western Australia, Perth.

April, 2015

Authors: Craig Rive1, David Ostrov2, Abha Chopra1, Rebecca Pavlos1 Simon Mallal1,3,

Elizabeth J. Phillips1,3:1Institute for Immunology and Infectious Diseases, Murdoch,

Australia,2University of Florida College of Medicine, Gainesville, FL,3Vanderbilt

University Medical Centre, Nashville, TN.

Figure 9.10: Poster presented at Science on the Swan, April 2015.

https://www.researchgate.net/publication/281661878_HLA_Class_I-drug-T-

cell_receptor_interaction_and_cytokine_expression_in_SJSTEN

DOI: 10.13140/RG.2.1.4937.5207

191

Title: Specific Binding Characteristics of HLA Alleles Associate with Nevirapine Hypersensitivity.

DOI: 10.13140/RG.2.1.1206.1287 Conference on Retroviruses and Opportunistic

Infections (CROI), Boston, Massachusetts; 2014

Authors: Rebecca Pavlos1, Craig Rive1, Elizabeth McKinnon1, David Ostrov2, Bjoern

Peters3, Jing Yuan4, Silvana Gaudieri1,5, David Haas6, Simon Mallal1,6, and Elizabeth

J. Phillips1,6: Institute for Immunology and Infectious Diseases, Murdoch, AUSTRALIA, 2University of Florida College of Medicine, Gainesville, FL, 3La Jolla Institute for Allergy

and Immunology, La Jolla, CA, 4Boehringer Ingelheim Pharmaceuticals Inc.,

Ridgefield, CT, 5University of Western Australia, Crawley, AUSTRALIA, 6Vanderbilt

University Medical Centre, Nashville, TN

Figure 9.11: Poster presented at Conference on Retroviruses and Opportunistic Infections (CROI),

Boston, Massachusetts; 2014. https://www.researchgate.net/publication/275332026_Specific_Binding_

Characterstics_of_HLA_Alleles_Associate_with_Nevirapine_Hypersensitivity

DOI: 10.13140/RG.2.1.1206.1287. (241)

192

9.4.3.2 Presentations

Title: HLA class-I-drug-T-cell receptor interactions in SJS/TEN.

DOI:10.1186/2045-7022-4-S3-P2 at Drug hypersensitivity meeting 6 (DHM6), Bern

Switzerland April, 2014

Authors: Craig Rive1, Rebecca Pavlos1, David Ostrov2, Pablo Plasencia2, Shien-Iu

Hung3 WenHung Chung4, Bjoern Peters5, Soren Buus6, Simon Mallal1,7, Elizabeth J.

Phillips1,7, 1Institute for Immunology and Infectious Diseases, Murdoch, Australia, 2University of Florida College of Medicine, Gainsville, FL, 3Institute of Pharmacology,

School of Medicine, Genome Research Centre, National Yang-Ming University, Taipei,

Taiwan., 4Department of Dermatology, Chang Gung Memorial Hospital, College of

Medicine, Chang Gung University, Keelung,5La Jolla Institute for Allergy and

Immunology, La Jolla, CA, 6Laboratory of Experimental Immunology, University of

Copenhagen, Copenhagen, Denmark. 7Vanderbilt University Medical Centre,

Nashville, TN.

Figure 9.12: Abstract for talk presented at Drug hypersensitivity meeting 6 (DHM6), Bern Switzerland

April, 2014.

http://www.ctajournal.com/content/4/S3/P2 DOI:10.1186/2045-7022-4-S3-P2 or

https://www.researchgate.net/publication /274893067_

HLA_class_I-drug-T-cell_receptor_interactions_in_SJSTEN. (242)

193

Title: HLA Class-I-drug-T-cell receptor and Carbamazepine induced Stevens-Johnson syndrome / Toxic epidermal necrolysis

DOI: 10.13140/RG.2.1.2709.2962. The Australian Society for Medical Research

(ASMR), Perth Western Australia June, 2014.

Authors: Craig Rive1, Rebecca Pavlos1, David Ostrov2, Pablo Plasencia2, Shien-Iu

Hung3 WenHung Chung4, Bjoern Peters5, Soren Buus6, Simon Mallal1,7, Elizabeth J.

Phillips1,7, 1Institute for Immunology and Infectious Diseases, Murdoch, Australia, 2University of Florida College of Medicine, Gainsville, FL, 3Institute of Pharmacology,

School of Medicine, Genome Research Centre, National Yang-Ming University, Taipei,

Taiwan., 4Department of Dermatology, Chang Gung Memorial Hospital, College of

Medicine, Chang Gung University, Keelung,5La Jolla Institute for Allergy and

Immunology, La Jolla, CA, 6Laboratory of Experimental Immunology, University of

Copenhagen, Copenhagen, Denmark. 7Vanderbilt University Medical Centre,

Nashville, TN.

Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN). CBZ-SJS/TEN has been associated with HLA-B*15:02 carriage and specific T-cell Background: Carbamazepine (CBZ) is associated with the severe cutaneous drug reaction clonotypes. We aimed to characterise the interactions between specific T-cell receptor (TCR) clonotypes, HLA-B*15:02 and CBZ.

Methods: Peripheral blood mononuclear cells (PBMC’s) were isolated from patients with CBZ-SJS/TEN and healthy controls, stimulated with 10ug/mL CBZ (Sigma, cat. no. C4024-5G) and cultured over a 9-14 day period. RNA was isolated at various time points and TCR Vβ subtypes were assessed using digital droplet PCR (Bio-Rad QX100 droplet digital PCR system). Drug-specific T-cell IFN-γ and granulysin responses were assessed by ELISpot and ICS. In-silico modelling was used to examine the interaction between the known specific TCR Vβ and Vα chains, CBZ and HLA-B*15:02.

Results: CBZ-specific T-cell IFN-γ and granulysin responses were detected many years following the original SJS/TEN reaction. Expansion of specific TCR CDR3 sequences was confirmed in CBZ-SJS/TEN patient cultures. In-silico modelling of the CBZ-HLA-B*15:02-TcR interaction suggests that CBZ binds non-covalently in the P4 binding pocket of the HLA-B*15:02 antigen binding cleft in a site that is typically occupied with bound peptide.

Conclusions: These findings raise two non-mutually exclusive possibilities: 1) CBZ may block peptide-binding and be presented by HLA-B*15:02 in a solvent exposed manner available for direct recognition by the TCR, 2) CBZ binding the central portion of the HLA-B*15:02 antigen binding cleft may permit long peptides to bind conventionally at the peptide termini (in the A- and F-pockets) and bulge over the drug in the central residues to permit indirect recognition of CBZ (peptide mediated TCR contact).

Documents

The Immunopathogenesis of drug hypersensitivity