The Peptide Binding Specificity of the Inhibitory and Activating KIR2D Receptors
Sorcha Cassidy
Imperial College London, Division of Medicine
Department of Hepatology
Thesis submitted for the Degree of Doctor of Philosophy
2014
Declaration of Originality
I declare that this body of work is all my own work which was carried out between 2010 and 2014. All the work, data analysis, and illustrations were performed by me, unless otherwise specified in the text. Where figures contain collaborative data, this is specified in the figure panel.
Both I and Thet Mon Myint genotyped the donors obtained from NHS Collindale for KIR2DL2 and KIR2DL3. Dr Helen North kindly provided access to the donors for typing.
In Chapter 3 the HLA-C typing and further KIR genotyping was performed by Franco Tavarozzi at The Anthony Nolan Institute London and Dr James Traherne at Cambridge University respectively.
In Chapter 3 Fig 3.16, Professor Jayajit Das and Dr Sayak Mukherjee constructed a mathematical model to depict my experimental data. All of the peptide mixing experiments were performed by myself and Professor Das and Dr Mukherjee used this experimental data to construct the model that is shown in Fig 3.16 and detailed in section 3.6.
In Chapter 4, Dr Liam Fanning provided the 31 isolate sequences from the Hepatitis C subtype 1b Irish cohort. All of the bioinformatic analysis was carried out by myself.
In Chapter 5, Dr Connie Liu provided the Huh7 replicon cells transfected with HLA-Cw*01 and the HCV replicon for use in this set of experiments.
Publications
Cassidy S, Mukherjee S, Myint TM, Mbribindi B, North H, Traherne J, Mulder A, Claas FH, Purbhoo MA, Das J, Khakoo SI. Peptide selectivity discriminates NK cells from KIR2DL2- and KIR2DL3-positive individuals. Eur J Immunol. 2015;45(2):492-500.
Cassidy SA, Cheent KS, Khakoo SI. Effects of Peptide on NK cell-mediated MHC I recognition. Front Immunol. 2014 Mar 31;5:133. doi: 10.3389/fimmu.2014.00133.
Cheent KS, Jamil KM, Cassidy S, Liu M, Mbiribindi B, Mulder A, Claas FH, Purbhoo MA, Khakoo SI. Synergistic inhibition of natural killer cells by the nonsignaling molecule CD94. Proc Natl Acad Sci USA. 2013;15;110(42):16981-6. doi: 10.1073/pnas.1304366110.
Knapp S, Meghjee N, Cassidy S, Jamil K, Thursz M. Detection of allele specific differences in IFNL3 (IL28B) mRNA expression. BMC Med Genet. 2014;5;15(1):104.
2 Copyright Declaration
The copyright of this thesis rests with the author and is made available under a Creative Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy, distribute or transmit the thesis on the condition that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it. For any reuse or redistribution, researchers must make clear to others the licence terms of this work.
3 Abstract
Human natural killer (NK) cells play an invaluable role in the first line of innate immune defence against viral infection and cancer. They are the main sub-group of lymphocytes with predominant expression of polymorphic KIR (Killer-immunoglobulin-like-receptors), which allows them to engage with their MHC Class I ligands. Past immunogenetic analysis has shown that possession of KIR2DL3 may confer an advantage in eliminating acute Hepatitis C virus (HCV) infection. This contrasts with those possessing KIR2DL2 who were found not to spontaneously resolve infection.
Previous work has shown that different variants of an endogenous peptide VAPWNSLSL (VAP) can modulate the KIR2DL2/3-HLA-C1 interaction. Here we have analysed the peptide repertoire in the context of the inhibitory KIR2DL3, KIR2DL2 receptors and the activating KIR2DS2 receptor. We firstly investigated any differences in KIR2DL3+ or KIR2DL2+ NK cell reactivity in response to mixes containing VAP peptide derivatives with the aim that these peptide mixes would be representative of the diverse peptide repertoire on a cell. Overall, we found that NK cells from KIR2DL3 homozygous donors were more sensitive to changes in the peptide content of MHC class I than those from KIR2DL2 homozygous donors.
As HCV is an infection with a well-recognised association of outcome with specific KIR, we also sought to determine whether HCV peptides could modulate KIR2DL2/3+ NK cell reactivity. We investigated the ability of the panel of HCV peptides to influence NK cell reactivity firstly via KIR2DL2 or KIR2DL3. The majority of HCV peptides were weak binders for HLA-Cw*0102 and/or non-inhibitory. We found that a HCV peptide (HCV NS31254-1263 LNPSVAATL) was able to promote weak binding of the KIR2DS2 to the HLA-Cw*0102 allele and subsequently affect KIR2DS2+ NK cell reactivity. Here we report that LNP can effectively promote binding of KIR2DS2 tetramers and affect the functionality of KIR2DS2+ NK cell clones. Thus, we propose that the HCV-derived peptide LNP can act as an agonistic peptide on NK cell activation through KIR2DS2.
4 Acknowledgements
First and foremost I would like to sincerely thank my PhD supervisor Professor Salim Khakoo for being an excellent, supportive and inspirational supervisor throughout the duration of this PhD project. Your enthusiasm and guidance has pushed me forward and created great opportunities for me. I feel like I have made great progression under your supervision and I am privileged to have worked with you.
I would also like to sincerely thank my PhD co-supervisor Dr Marco Purbhoo. Your outstanding support and helpful advice has truly been appreciated over the past few years. Your excellent technical expertise and great ideas have always been incredibly helpful and insightful.
I would like to say a big thank you to all of the members of the Khakoo and Purbhoo lab, both past and present. Dr Gwenoline Borhis and Dr Kuldeep Cheent were a huge help in the early days of my PhD training and their useful support and advice aided me greatly along the way. Dr Khaleel Jamil, Dr Lemonica Koumbi and Dr Mina Aletrari were great mentors and colleagues that one could hope for and also my good friends and I will miss everyone dearly. If I have forgotten anyone my thank you extends to all lab members and I have thoroughly enjoyed working with you all.
I would like to thank all of my collaborators throughout the course of my PhD. There are many to name but in particular I would like to thank Dr Salah Mansour whom I worked closely with in Southampton and who provided excellent expertise. Thet Mon Myint and Dr Helen North were also superb in their assistance in obtaining much-needed donors for my experiments. Their flexibility and kindness was truly appreciated.
I would like to thank my fellow PhD students Katie Twigger and Anna Cauldwell who started their journey with me on their PhD studies and their support was much needed over the past few years. They were always there to provide a helping hand or a much needed friendly face throughout the tough days. Thank you for all those times.
I would finally like to say an overwhelming thank you to my family and friends for all of their unbelievable understanding, patience and support over the past few years. It has been a long road but I hope that I will make you proud.
5 Table of Contents
Section Page
Title Page 1
Declaration of Originality 2
Copyright Declaration 3
Abstract 4
Acknowledgements 5
Table of Contents 6-11
Figures 12-14
Tables 15
Abbreviations 16-17
Chapter 1. Introduction 18-61
1.1 A brief history of Natural killer cells 18
1.2 Defining human NK cell surface markers 18-19
1.3 NK cell functions (Innate immunity) 19
1.4 NK cell functions (Adaptive immunity) 20
1.5 NK cell cytotoxicity (Perforin/Granzyme-mediated) 20
1.6 NK cell cytotoxicity (Death receptor pathway) 21
1.7 Main NK cell targets for cytotoxicity 21
1.8 The Missing-Self Hypothesis 21-22
1.8.1 NK cell Self-Tolerance 22
1.9 The NK cell Licensing model 22-23
1.9.1 NK cell Arming and Disarming models 23
1.9.2 The Rheostat concept and the Tuning model 24
1.10 Human NK cell receptors summary 24-27
1.10.1 The KIR receptor family overview 24-25
1.10.2 The NKG2/CD94 receptor family overview 25-26
1.10.3 NCRs (NKp30, NKp44, NKp46) 26
1.10.4 LILR/ILT/MIR receptors 26-27
1.10.5 2B4 receptor 27
1.11 Ly49 receptors in mice 27-28
6 1.12 KIR nomenclature and structure 28
1.13 HLA Class I nomenclature and structure 29-30
1.14 KIR Haplotypes and genetic diversity on the LRC 30-31
1.15 KIR Lineages 31-32
1.16 Inhibitory KIR 32-35
1.16.1 KIR2DL2/3 32-33
1.16.2 KIR2DL1 33
1.16.3 KIR3DL1 33-34
1.16.4 KIR3DL2 34
1.16.5 KIR2DL4 34
1.16.6 KIR2DL5 34-35
1.17 Ligands for the inhibitory KIR: HLA Class I 35-37
1.17.1 HLA-C 35-36
1.17.2 HLA-B 36
1.17.3 HLA-A 37
1.17.4 HLA-G 37
1.18 Activating KIR and their HLA Class I ligands 37-38
1.18.1 KIR2DS1 and ligand HLA-C2 38
1.18.2 KIR2DS2 and ligand HLA-C1 38-39
1.18.3 KIR2DS4 39-40
1.18.4 KIR2DS3 and KIR2DS5 40
1.18.5 KIR3DS1 40
1.19 Co-crystallisation and structure of the inhibitory and activating KIR2D
receptors with their respective HLA-C ligands 40-42
1.20 Expression of KIR on T cells 42
1.21 The inhibitory KIR synapse and intracellular signalling 43
1.22 The activating KIR synapse and intracellular signalling 43
1.23 Coevolution of the KIR and HLA genes and outcome of viral infection 44
1.23.1 KIR and HIV infection 44
1.23.2 KIR and Hepatitis C infection 45
1.24 Hepatitis C virus and NK cells 45-46
1.25 HCV genotypes and prevalence of HCV in population 46
1.26 NK cells and cancer/ Immunotherapies 47
7 1.27 KIR/HLA and autoimmunity/inflammatory disease 47-48
1.28 NK cells (KIR/HLA) and pregnancy 48-49
1.29 NK cells (KIR/HLA) and transplantation 49
1.30 The importance of peptides in HLA:Inhibitory KIR interactions 50-53
1.31 Peptide antagonism of NK cell inhibition 53-54
1.32 The role of peptides in HLA: Activating KIR interactions 54
1.33 Peptide synergy and the NKG2/CD94 receptor family 55-56
1.34 Presentation of peptide on MHC Class I of healthy cells versus virally infected cells 56-59
1.34.1 Cleavage and processing of peptides for presentation 56-57
1.34.2 Presentation of self- and viral-epitopes during an infection 57-59
1.35 Background and Aims to Project 60-61
Chapter 2. Materials and Methods 62-73
2.1 Immortalised cell lines and culture 62
2.2 Human Peripheral Blood Mononuclear Cell isolation and preparation 62-63
2.3 KIR genotyping 63
2.4 HLA Class I typing 63
2.5 Peptides 63
2.6 Peptide stabilisation assay 64
2.7 CD107a NK Cell Degranulation Assay 64-65
2.7.1 Target cell preparation 64
2.7.2 Effector cells (PBMC) 64
2.8 CD107a NK Cell Degranulation Assay 64-65
2.9 Flow cytometry analysis 65
2.9.1 Flow cytometry Antibody panel 65
2.9.2 Flow cytometry compensation 65
2.10 KIR2DL2/3-Fc binding assay 67
2.11 Production of KIR2DS2 tetramers 67-70
2.11.1 PCR (Amplification of DNA encoding KIR2DS2 extracellular domain) 67
2.11.2 Cloning of KIR2DS2 extracellular domain 67-68
2.11.3 Expression of KIR2DS2 proteins 68
2.11.4 Purification of KIR2DS2 protein from inclusion bodies 68-69
8 2.11.5 Refolding of KIR2DS2 protein using a simple dilution method 69
2.11.6 Reducing SDS-polyacrylamide gel electrophoresis 69
2.11.7 Biotinylation of KIR2DS2 monomers 70
2.11.8 Tetramerisation of KIR2DS2 70
2.11.9 KIR2DS2 tetramer staining 70
2.12 Cloning of human KIR2DS2+ NK cells 70-71
2.13 KIR expression on NK clones using RTqPCR 71-72
2.13.1 RNA isolation and cDNA preparation from NK clones 71-72
2.13.2 Reverse-Transcriptase (RT) and Quantitative Real-Time (q) PCR 72
2.14 Cytotoxicity assay with NK clones using CTO (cell-tracker-orange) 72
2.15 Conjugate formation and staining 72-73
2.16 Bioinformatic analysis 73
2.17 Statistical analysis 73
Chapter 3. Results 74-112
Changes in the peptide repertoire can differentially influence KIR2DL2+ and KIR2DL3+ NK cell reactivity
3.1 Peptide stabilisation of HLA-Cw*0102 and HLA-E molecules 77-81
3.2 The binding specificity of KIR2DL2-Fc and KIR2DL3-Fc proteins to
HLA-Cw*0102:VAP-LS peptide and derivatives 82-84
3.3 Optimisation of CD107a assay with K562 and 721.174 cells 85-87
3.3.1 Gating strategy for Flow Cytometry 85-86
3.4 Optimisation of IL-15 concentration in CD107a assays 88-89
3.5 KIR2DL2/3 and HLA-C genotyping 90-92
3.6 Inhibition by one or two peptides is similar for NK cells from KIR2DL2 and KIR2DL3 donors 93-98
3.7 Triple peptide mixes with VAP-DA, VAP-RA and VAP-FA 99-103
3.8 Modelling the data from the triple peptide experiments 104-106
3.9 Discussion 107-112
Chapter 4. Results 113-156
KIR2DL2/3+ NK cell responses to peptides derived from Hepatitis C virus
4.1 Identification and synthesis of HCV peptides 113-156
4.2 Conservation of the predicted HLA-Cw*0102 binding epitopes across different HCV genotypes 120-122
9 4.3 Peptide stabilisation assays with HCV peptides 123-127
4.4 Peptide binding affinities with online prediction algorithms 128-129
4.5 The binding specificity of HCV peptides to KIR2DL2-Fc and KIR2DL3-Fc proteins 130-132
4.6 CD107a NK degranulation assay with HCV peptides 133-136
4.7 Effect of HCV peptides 5 & 6 on NK inhibition in KIR2DL2 and KIR2DL3 homozygotes 137-147
4.8 Specificity of effect with HCV peptide 5/HCV6 and KIR2DL2/KIR2DL3/KIR2DS2
population of NK cells 148-149
4.9 The involvement of NKG2/CD94 receptor family in the effects observed with
HCV peptide 5 and 6 150-151
4.10 Discussion 152-156
Chapter 5. Results 157-197
An investigation into the peptide binding specificity of the activating KIR2DS2 receptor: A viral peptide with potential specificity for KIR2DS2
5.1 KIR2DS2 tetramers 160-173
5.1.1 Cloning strategy for the generation of KIR2DS2 extracellular domain constructs 160-164
5.1.2 Expression of KIR2DS2 Monomers 165-166
5.1.3 Refolding and purification of KIR2DS2 monomers 165-167
5.1.4 Biotinylation and tetramerisation 168-169
5.1.5 Staining with KIR2DS2 tetramers 170-173
5.2 LNPSVAATL does not stabilise HLA-A2 molecules on .174 target cells 174-175
5.3 Confocal microscopy 176-178
5.4 Staining NK cells with 1F12 Antibody 179-181
5.5 NK cell cloning 182-193
5.5.1 NK clones show increased cytotoxicity in response to target cells loaded with LNP peptide 182-188
5.5.2 KIR2DS2+ NK clones can kill Huh7 replicon cells transfected with the HLA-Cw*0102 allele 189-191
5.5.3 qRT-PCR analysis to determine relative levels of KIR2DS2 and KIR2DL2 gene expression 192-193
5.6 Discussion 194-197
Chapter 6. Discussion 198-213
6.1 The peptide binding specificity of the inhibitory KIR receptors 198
6.2 The peptide binding specificity of activating KIR receptors 198
6.3 Three models to activate NK cells using peptides 199-207
10 6.3.1 Antagonistic peptides for inhibitory KIR 199-204
6.3.2 Synergy of inhibition via the NKG2A/CD94 receptor 204
6.3.3 Peptides specific for activating KIR receptors (agonist peptide(s) for activating KIR) 204-207
6.4 The immunopeptidome and self/viral peptides 207-210
6.5 HCV adaptation and KIR 210-211
6.6 Evolutionary and clinical relevance 211-213
Chapter 7. Final Conclusions and Future Direction 214-215
Chapter 8. References 216-241
Appendix I 242-244
Appendix II 245-247
11 Figures
Figure Page
1.1 Structure of the main inhibitory and activating KIR receptors 29
1.2 The KIR locus on chromosome 19q13.4 31
1.3 Interaction of KIR2DL1 and KIR2DL2/3 receptors with their respective
HLA-C2 and HLA-C1 ligands 36
1.4 The interaction of KIR and TCR with peptide:MHC 51
3.1 The three different classes of peptide that may be presented on MHC Class I
to CD158+ NK cells 75
3.2 HLA-Cw*0102 peptide stabilisation assay with endogenous VAPWNSLSL and derivatives 79
3.3 Scatchard Analysis: Affinity of VAPWNSLSL peptides for HLA-Cw*0102 80
3.4 HLA-E stabilisation with endogenous VAPWNSLSL and derivatives 81
3.5 KIR2DL2-Fc and KIR2DL3-Fc binding to .174 cells pulsed with VAP peptide and derivatives 83
3.6 KIR2DL3-Fc and KIR2DL3-Fc binding to .174 cells pulsed with VAP peptide and derivatives 84
3.7 Flow cytometry gating strategy for CD158b-/+ NK cells in CD107a degranulation assay 86
3.8 KIR2DL2/3+ NK cell degranulation in response to K562, 174 target cells or no target cells 87
3.9 Increasing IL-15 concentration has no effect on the overall inhibition
mediated by VAP-FA peptide 89
3.10 KIR2DL2/3 genotyping of donors by PCR-SSP 90
3.11 Effects of VAP-FA, VAP-RA and VAP-DA peptides on NK degranulation in
KIR2DL2 homozygous and KIR2DL3 homozygous donors 95
3.12 Frequency of degranulation of CD3- CD56+ CD158b+ NK cells in response to
three VAP peptide derivatives loaded on 721.174 target cells 96
3.13 All peptide mixes containing two peptides give similar levels of CD107a expression for both
KIR2DL2 and KIR2DL3 homozygous donors 98
3.14 Mixes containing three peptides result in significant differences in degranulation between
KIR2DL2 and KIR2DL3 homozygous donors 102
3.15 Analysis of degranulation of CD158b+ NK cells as determined by Equation (1) 106
4.1 Primary sequence analysis between HCV subtype 1b isolate sequences
from Irish cohort 116
4.2 An illustration of the peptide binding specificity of Class I HLA-Cw*0102 allotype 117
4.3 HCV peptides and their location within the HCV polyprotein 119
4.4-4.5 Multiple sequence alignment between each HCV genotype at potential
12 HLA-Cw*0102 binding-epitopes 121-122
4.6 The majority of HCV peptides have a low affinity for HLA-Cw*0102 and do not stabilise well 124
4.7 HLA-Cw*0102 stabilisation assay with intermediate affinity HCV peptides 125
4.8 VLPCSFTTL (HCV p6) can bind HLA-Cw*0102 to a high-affinity 126
4.9 Scatchard Analysis: Affinity of HCV peptide 5 & 6 for HLA-Cw*0102 127
4.10 Algorithm binding predictions of HCV peptides to HLA-Cw*0102 129
4.11 KIR2DL2-Fc binding to .174 cells pulsed with HCV peptides 131
4.12 KIR2DL3-Fc binding to .174 cells pulsed with HCV peptides 132
4.13 The majority of HCV peptides have a weak or no inhibitory effect on
NK cell degranulation in CD158b+ NK cells 134
4.14 HCV core peptide LLPRRGPRL has an inhibitory effect on NK cell degranulation
with a naturally occuring mutation at position 8 (R->K) relieving this inhibition 136
4.15 Initial screen with eight HCV peptides and VAP-FA peptide 138
4.16 Gating strategies used for CD56+ CD158b+ and CD56+ CD158b- NK cells 139
4.17 LNPSVAATL (HCV5) reduces CD158b+ NK cell inhibition in KIR2DL2 homozygous donors 140
4.18 LNPSVAATL (HCV5) does not have an effect on CD158b+ NK cell inhibition
in KIR2DL3 homozygous donors 141
4.19 LNPSVAATL (HCV5) reduces CD158b+ NK cell inhibition in KIR2DL2/3 heterozygotes 142
4.20 Normalised data for HCV5 from experiments with KIR2DL2 homozygous,
KIR2DL3 homozygous and KIR2DL2/3 heterozygous donors 143
4.21 VLPCSFTTL (HCV6) reduces CD158b+ NK cell inhibition in KIR2DL2 homozygous donors 144
4.22 VLPCSFTTL (HCV6) reduces CD158b+ NK cell inhibition in KIR2DL3 homozygous donors 145
4.23 VLPCSFTTL (HCV6) reduces CD158b+ NK cell inhibition in KIR2DL2/3 heterozygous donors 146
4.24 Normalised data from experiments with HCV6 (VLPCSFTTL) with
KIR2DL2 homozygous, KIR2DL3 homozygous and KIR2DL2/3 heterozygous donors 147
4.25 Effects observed with HCV5 (LNPSVAATL) in KIR2DL2 homozygous donors is
not found in the CD158a+ NK cell sub-population 149
4.26 HCV peptides LNPSVAATL (HCV 5) and VLPCSFTTL (HCV 6) do not stabilise
HLA-E molecules on .174 cells 151
5.1 Structure of the activating KIR2DS2 receptor 161
5.2 Cloning strategy with SnapGene software 162
5.3 Restriction digests of pET23d vector and DNA encoding the extracellular domain of KIR2DS2 163
5.4 PCR screen and restriction digest showing successful cloning of KIR2DS2
extracellular domain into vector pET23d 164
13 5.5 Expression of protein encoding the KIR2DS2 extracellular domain in two
colonies of Rosetta bacteria post-induction with IPTG 166
5.6 FPLC plot to show purity of the refolded KIR2DS2 extracellular domain protein 167
5.7 FPLC to assess refolded biotinylated KIR2DS2 proteins 169
5.8 KIR2DS2 tetramers can bind HLA-Cw*0102: LNPSVAATL complexes on
the surface of .174 target cells 171
5.9 KIR2DS2 tetramers can bind .174 target cells after pulsing with
HCV5 (LNPSVAATL) peptide only 172
5.10 KIR2DS2 tetramers can also mediate binding to VAP peptide derivatives
depending on key residues at P7 and P8 173
5.11 LNPSVAATL peptide does not stabilise HLA-A2 molecules on .174 target cells 175
5.12 LNPSVAATL (HCV p5) induces clustering of KIR2DS2 at the interface
between target and effector cells 177
5.13 LNP induces aggregation of KIR2DS2 at the interface of effector and target cells 178
5.14 Staining KIR2DL2 and KIR2DL3 homozygous donors with monocloncal 1F12 antibody 180
5.15 Staining untransfected and transfected Ba/F3 cells with 1F12 monoclonal antibody 181
5.16 The clones used in cytoxicity assays stained CD158b+ (positive for KIR2DL2/KIR2DS2) 184
5.17 Flow cytometry gating strategy for CTO assay 185
5.18 Flow cytometry gating strategies for CTO assay with clones 186
5.19 Screen of clones for cytotoxicity against peptide-loaded target cells 187
5.20 Clones 4, 6 and 9 show increased cytotoxicty in response .174
target cells loaded with LNPSVAATL peptide 188
5.21 CD107a assay with Clone 6 and Huh7-Cw*0102 target cells (+/- HCV replicon) 190
5.22 CD107a assay with Clone 9 and Clone 27 and Huh7-Cw*0102 target cells (+/- HCV replicon) 191
5.23 qRT-PCR analysis of KIR2DL2 and KIR2DS2 gene expression in 20 NK clones 193
6.1 The effects of VAP-FA peptide alone and antagonism of DA/FA mix on
KIR2DL3+ and KIR2DL2+ NK cells 200
6.2 The effects of the peptide mixes containing three peptides on
KIR2DL2+ and KIR2DL3+ NK cells 202
6.3 The agonistic effect of a viral peptide on KIR2DS2 206
14 Tables
Tables Page
1.1 An overview of the Inhibitory KIR members 33
1.2 An overview of the Activating KIR members 38
1.3 Prominent studies of KIR interaction with peptide(s) to date 55
2.1 Antibodies used for Flow Cytometry 66
3.1 qPCR results detailing KIR2DL2, KIR2DL3 and KIR2DS2 copy numbers for each donor group 91
3.2 HLA-C typing and KIR typing results 92
3.3 Final concentrations of peptide in mixes containing three VAPWNSLSL derivatives 101
3.4 Concentrations of VAPWNSLSL peptide derivatives alone or in each mix at %CD107a50 for
KIR2DL3 and KIR2DL2 donors 103
15 Abbreviations
ADCC Antibody-dependent Cellular Cytotoxicity ANOVA Analysis of Variance Arg Arginine Asn Asparagine CMV Cytomegalovirus CTL Cytotoxic T lymphocyte DC Dendritic cells DMEM Dulbecco’s Modified Eagle’s Medium EBV Epstein Barr Virus ERK Extracellular signal-regulated kinase FasL Fas Ligand FCS Foetal Calf Serum Gln Glutamine GCSF Granulocyte Colony-Stimulating Factor GM-CSF Granulocyte/macrophage colony stimulating factor HCV Hepatitis C virus HIV Human Immunodeficiency Virus HLA Human Leukocyte Antigen HLA-C1 Group 1 HLA-C allotypes HLA-C2 Group 2 HLA-C allotypes ICAM-1 Intercellular adhesion molecule-1 IFN-γ Interferon Gamma Ig Immunoglobulin IL Interleukin Ile Isoleucine IS Immunological Synapse ITAM Immuno-receptor tyrosine activating motif ITIM Immuno-receptor tyrosine- inhibitory motif KIR Killer Immunoglobulin-like Receptor LAMP Lysosomal associated membrane glycoprotein Leu Leucine LFA-1 Lymphocyte function-associated antigen-1 LILR Leukocyte Immunoglobulin-like Receptor LNP LNPSVAATL peptide Lys Lysine mAbs Monoclonal antibodies MAPK Mitogen-activated extracellular signalling regulated kinase MEK Mitogen-activated or extracellular signal-regulated protein kinase MHC Major Histocompatibility Complex MFI Mean Fluorescence Intensity NCAM Neural cell adhesion molecule NCR Natural Cytoxcity Receptor NK Natural Killer
16 NKAT NK associated transcripts PBMC Peripheral blood mononuclear cells Phe Phenylalanine RANTES Factor regulated on activation-normal T cells expressed and secreted SD Standard Deviation SEM Standard Error of the Mean Ser Serine SHP-1 Src homology (SH)2-containing protein tyrosine phosphotase-1 SPR Surface Plasmon Resonance TAP Transporter associated with antigen processing TGF-β Transforming growth factor-β Th1 T helper 1 Thr Threonine TM Transmembrane TNF-α Tumour necrosis factor- α TRAIL Tumour necrosis factor-related apoptosis-inducing ligand Tyr Tyrosine VAP-DA VAPWNSDAL peptide VAP-FA VAPWNSFAL peptide VAP-LS VAPWNSLSL peptide VAP-RA VAPWNSRAL peptide
Zn2+ Zinc β2-m β2-microglobulin
17 Chapter 1. Introduction
1.1 A brief history of Natural killer cells
Natural killer cells (NK cells) exist as crucial cells of the human innate immune system and constitute 5-15% of total lymphocytes in peripheral blood (Caligiuri, 2007; Cheent and Khakoo, 2009; Vivier et al, 2008). NK cells were originally identified by Rolf Kiessling in 1975 as “null lymphocytes”, elusive cytotoxic cells that lacked expression of the well-defined T and B lymphocyte markers (Kiessling et al, 1975; Heberman et al, 1975). Their discovery came about following observations made in studies using tumour cell lines (Kiessling et al, 1975; Heberman et al, 1975; Karre et al, 1986; Ljunggren et al, 1989). In order for T lymphocytes to kill specific tumour cells prior sensitisation with tumour antigen was required. However, it was found that a certain sub-group of lymphocytes could mediate spontaneous cytotoxicity of these tumour cells without any prior sensitisation. Furthermore this activity was thought not to be MHC (major histocompatibility complex) restricted. This non-MHC restricted killing was termed “natural cytotoxicity” and Kiessling and colleagues coined the ‘natural killer cell’ (Kiessling et al, 1975). There has been significant progression in recent years in terms of unravelling the complexity of NK cell interactions with other cells.
1.2 Defining human NK cell surface markers
To date, it has been difficult to ascribe a surface antigen as entirely NK cell specific. Mature human NK cells are distinguishable from the CD3+ T Lymphocyte population based on their cell surface phenotype as CD3- and do not express the T Cell Receptor (TCR) (Lanier et al, 1989). NK cells express the CD56molecule or neural cell adhesion molecule (NCAM). NK cells can be further divided into CD56bright or CD56dim subpopulations depending on their level of surface CD56 expression (Caligiuri, 2008; Jacobs et al, 2001). CD16 is also another prominent NK surface marker that assorts with CD56. There are two main subsets of NK cells depending on their surface CD56 expression. CD56bright CD16- NK cells are not usually highly cytotoxic but release large quantities of IFN-γ (Cooper et al, 2001; Maghazachi et al, 2005). CD56dim CD16+ NK cells express a low level of CD56 but these cells are highly cytotoxic. The CD56dim subset comprises around 90% of all NK cells. It is now thought that the CD56dim NK cells are more mature than the CD56bright NK cells (Jacobs et al, 2001). However, it has been shown that CD56dim NK cells can express large quantities of Interferon- γ (IFN-γ) in response to red blood cells infected with Malaria parasite (Korbel et al, 2005). Korbel et al found that IFN-γ+ CD56dim NK cells can outnumber IFN-γ+ CD56bright NK cells in the periphery (Korbel et al, 2009). Numerous studies have now proposed that CD56bright NK
18 cells are a functionally distinct lymphocyte population and most likely they are also the direct precursors to the more mature CD56dim NK cells (Freud et al, 2014). CD57 is also another recently described NK cell maturation marker and may discriminate NK cells based on their final stages of peripheral maturation (Lopez-Verges et al, 2010; Nielsen et al, 2013). On CD8+ T cells CD57 is a marker of terminal differentiation (Lopez-Verges et al, 2010). Interestingly, CD57+ NK cells have been found to be highly cytotoxic and the expression of the marker has been associated with some chronic infections to date (Nielsen et al, 2013). Another marker that is perhaps more specifically expressed on all NK cells is the activating Natural Cytotoxicity Receptor (NCR) NKp46 marker which is a pan-NK cell marker conserved on all NK cells across many different species (Tomasello et al, 2012; Freud and Caligiuri 2006;Freud et al, 2013; Freud et al, 2014).
1.3 NK cell functions (Innate immunity)
The functional role of NK cells was traditionally thought to be limited within the innate immune defences because of their lack of antigen-specific cell surface receptors. Whilst NK cells provide an essential source of IFN-γ to prime a subsequent adaptive immune response, their main function is to instigate non-MHC restricted killing of virally infected or abnormal tumour cells (O’Connor et al, 2005; Vivier et al, 2008). Importantly, unlike T and B ymphocytes, NK cells accomplish this targeted killing without prior sensitisation, which is why they are traditionally grouped with the innate system (Cheent and Khakoo, 2009). This allows NK cells to discriminate between normal MHC Class I positive cells and cells that have lost or down-regulated the expression of MHC class I molecules as a consequence of tumour transformation or viral infection. NK cells release IFN-γ at a very early stage of infection and this primes the macrophage response towards release of IL-12 (Interleukin-12) which subsequently encourages a CD4+ Th1 (T Helper 1) and CTL (cytotoxic-T- lymphocytes) response which is essential in viral infection (Malnati et al, 1993; Moretta et al, 2008; Sun et al, 2009). Some of the other cytokines produced by NK cells include tumour necrosis factor–α (TNF-α), IL-3 and growth factors such as GM-CSF (Granulocyte Macrophage Colony-Stimulating Factor), G-CSF (Granulocyte Colony-Stimulating Factor) (Vivier et al, 2011). NK cells also secrete numerous chemokines, including CCL2 (MCP-1), CCL3 (MIP1-α), CCL4 (MIP1-β), CCL5 (RANTES), XCL1 (lymphotactin), and CXCL8 (IL-8) (Vivier et al, 2011).
19 1.4 NK cell functions (Adaptive immunity)
It has been difficult to assign NK cells as strictly innate cells as their functions also bridge the adaptive immune system and they have recently been shown to have features traditionally associated with adaptive immunity, such as enhanced responses to second antigen exposure such as memory or ‘recall responses’ (Sun et al 2009; Marcus and Raulet, 2013). Through secretion of chemokines, NK cells can attract Dendritic Cells (DC) and macrophages to a site of infection and/or inflammation. In response, DC and macrophages produce IL-15, which further heightens NK cell cytotoxicity (Vivier et al, 2011). As noted above, NK cells have a profound effect on the driving of the T cell response through the secretion of IFN-γ. The crosstalk between NK cells and DC is vital as NK cells are able to kill immature DC or also stimulate them for better antigen presentation with T cells. In situations of chronic inflammation or infection it has been found that NK cells can also secrete the immunosuppressive cytokine IL-10 which influences DC, macrophage and T cell responses (Vivier et al, 2011). NK cells have also been found to possess similarities with cytotoxic T lymphocytes in that there may be a subset of memory NK cells that can rapidly produce cytokines upon reactivation (Sun et al, 2009).
1.5 NK cell Cytotoxicity- Perforin/Granzyme-mediated
NK cell mediated cytotoxicity is primarily carried out by release of granules containing perforin and granzymes into target cells which initiates the apoptotic cascade within that target cell (Caligiuri, 2007). Perforin/granzyme-mediated apoptosis is the primary way in which NK cells kill their targets (Catalfalmo et al, 2003). The cytotoxic granules that are released by the NK cell contain perforin, a calcium-dependent pore-forming protein and granzymes, which are related to the pro-apoptotic serine proteases that usually initiate apoptosis within a cell (Raja et al, 2002). Perforin is able to insert itself into the target cell phospholipid bilayer and generate a pore for the granzymes to enter (Uellner et al, 1997). In particular, Granzyme B is most important as it can activate the cellular caspases, which triggers the cascade of apoptosis (Barry et al, 2000). Importantly, the cytotoxic granules that are released by the NK cell contain certain components within their lipid bilayer that can be detected as markers of NK cell degranulation. The granules contain proteins called lysosomal associated membrane glycoproteins (LAMPs), which include LAMP1, LAMP2 and LAMP3 (Alter et al, 2004). Detecting the level of LAMP proteins on the surface of NK cells or CTLs can be a useful measure of NK cell degranulation or activation (Alter et al, 2004).
20 1.6 NK cell Cytotoxicity- Death receptor pathway
NK cells possess another alternative mechanism to kill target cells effectively. They can also use death-receptor mediated apoptosis (Montel et al, 1995). NK cells can express certain death ligands that belong to the Tumour necrosis factor (TNF) family of ligands. These ligands are usually the Fas ligand (FasL), TNF ligand and TRAIL (TNF-related-apoptosis- inducing-ligand) (Montel et al, 1995; Screpanti et al, 2001). Depending on whether these ligands bind to their corresponding receptor on a target cell the ligation can initiate the apoptosis cascade within that target cell. In particular, TRAIL has been found to be upregulated on infected hepatocytes of the liver during Hepatitis C virus (HCV) infection which induces cellular apoptosis (Bantel and Schulze-Osthoff, 2003; Lan et al, 2008).
1.7 Main NK cell targets for Cytotoxicity
The balance between activating and inhibitory signals regulates the overall cytotoxic function of NK cells and this maintains tolerance to healthy self. NK cells are able to unleash their cytotoxic action on virally infected cells, tumour cells and cells infected with intracellular bacteria (Fitzgerald et al, 1986; Blanchard et al, 1987; Horowitz et al, 2012). Moreover, they have also been found to kill immature dendritic cells (iDC) (Piccioli et al, 2002; Gerosa et al, 2002). NK cells can also effectively target and lyse erythrocytes infected with intracellular parasites such as Plasmodium falciparum (malaria parasite) (Horowitz et al, 2010; Horowitz and Riley, 2013). There are numerous receptor families on the surface of NK cells that regulate the fine balance between NK activation and inhibition which will be discussed in further detail in later sections.
1.8 The Missing Self-Hypothesis
The missing-self hypothesis was originally ascribed to NK cells by Klas Karre whilst conducting his PhD studies to explain how they were able to distinguish between healthy/uninfected self and unhealthy/infected self (Ljunngren and Karre, 1985; Karre et al, 1986). The down-regulation and absence of MHC class I molecules on unhealthy target cells induced NK cell cytotoxicity in early studies (Ljunngren and Karre, 1985). MHC down- regulation is a common occurrence throughout viral infections as the virus avoids recognition by CD8+ CTL (Ljunggren and Karre, 1990). One model proposed by Karre et al stated that NK cells may express inhibitory receptors, that once engaged with MHC Class I on a target cell, would transmit an inhibitory signal to the NK cell instructing it not to lyse that target cell. Thus, normal healthy cells would keep NK cells in an inhibitory resting state, always transmitting these inhibitory signals in the presence of normal amounts of MHC Class I. The presence of inhibitory receptors that specifically recognise self-MHC class I molecules that
21 are expressed on the surface of all nucleated cells partly explains how NK cell activation is prevented.
There was more to it when it came to explaining how an NK cell may be activated. MHC class I down-regulation or absence on a target cell could account for NK cell killing but in some cases NK cells were able to kill even without MHC class I down-regulation (Cerwenka et al, 2001). In the context of viral infection, Malnati et al demonstrated that NK cells could directly kill cells infected with Herpesvirus that had “normal” levels of HLA class I expression (Malnati et al, 1993). Therefore, it was proposed that unhealthy cells may express some activating ligands that can instruct NK cells to kill. This model therefore led to the proposal that NK cells must possess a plethora of receptors able to transduce both inhibitory and activating signals to mediate cytotoxicity (Anfossi et al, 2006).
1.8.1 NK cell Self-tolerance and the ‘At least one’ model
Human erythrocytes do not express any MHC class I molecules and they are an interesting example of how NK cells mediate tolerance to healthy self in the situation where there is no MHC class I expression. It has also been proposed that erythrocytes lack the activating ligands needed that may engage activating NK receptors and this spares them from an NK cell effector response (Lanier, 2005). In 1997, Valiante and colleagues proposed the ‘At least one’ receptor model that tried to explain NK cell self-tolerance. The ‘At least one’ model simply proposes that every NK cell within a given individual expresses ‘at least one’ inhibitory receptor with self-specificity for MHC Class I. If every NK cell expresses at least one inhibitory receptor for self-MHC Class I this would explain self-tolerance. Valiante et al conducted this research on NK clones derived from only two individuals and found that each of these clones did express at least one inhibitory receptor that was known at the time for MHC Class I (Valiante et al, 1997). However, it is now known that the education and self- tolerance of NK cells is far more complex than the ‘At least one’ model as a subset of NK cells do not possess an inhibitory receptor for self MHC Class I and these NK cells can be alternatively activated via cytokine stimulation (Raulet and Vance, 2006). Further models have now been proposed which are introduced next.
1.9 The NK cell Licensing model
The licensing model of NK cell regulation was proposed after numerous studies were carried out on the Ly49 receptors in mice. It was proposed that the NK cells in mice underwent a ‘licensing’ or ‘education’ process following the engagement of their inhibitory Ly49 receptors
22 with self-MHC Class I molecules. Simply, NK cells became ‘licensed to function’ after engagement with self-MHC (Kim et al, 2005). Anfossi and colleagues proposed that human NK cells also undergo an education process and this is dependent on the inhibitory NK receptors and the MHC Class I molecules they bind (Anfossi et al, 2006). Therefore, the inhibitory receptors that an NK cell possesses ultimately guide its education and tolerance to self. The licensing model describes that during the early development of an NK cell there is a key-point whereby the receptors on an NK cell can recognise self-MHC Class I on other cells and can calibrate themselves accordingly so from that point on they can recognise the down- regulation or absence of any MHC Class I. Simply put, an early NK cell calibrates itself to recognise a target cell expressing a normal level of MHC class I. The licensing model states that the engagement of inhibitory NK receptors with self-MHC class I in an early NK cell provides a go-ahead signal to license fully-effector NK cells (Kim et al, 2005). There may be an unlicensed subset of NK cells and these remain hyporesponsive and non-functional (Anfossi et al, 2006). Licensing has provided a good model for NK self-tolerance so far and has since been further divided into the ‘Arming’ and ‘Disarming’ models by Raulet and Vance (Raulet and Vance, 2006).
1.9.1 NK cell Arming and Disarming models
The NK ‘Arming’ model that was proposed by Raulet and Vance applies to immature NK cells only. The Arming model states that immature NK cells receive positive signals from inhibitory receptors on NK cells that are specific for self-MHC Class I. This induces functional maturation and ‘arming’ of the immature NK cell. In the absence of any inhibitory signals, there is no positive signal and therefore the immature NK cells stay in an unresponsive and ‘unarmed’ state (Raulet and Vance, 2006)
The NK ‘Disarming’ model that was also proposed can be applied to both immature and mature NK cell subsets. NK cells that do not possess any inhibitory receptors for self-MHC Class I can be ‘disarmed’ or switched off. This means that they would enter a hyporesponsive state and mediate no effector functions. It has been proposed that chronic stimulation of NK cells by self-cells expressing activating ligands could initiate ‘Disarming’ of that NK cell. It is a similar concept to T cell anergy and the disarming model could also account for loss of responsiveness in the mature NK cells subset (Raulet and Vance, 2006).
23 1.9.2 The Rheostat concept and the Tuning model
The Rheostat concept is an interesting new concept that describes NK responsiveness from a more quantitative point of view (Joncker and Raulet, 2008; Brodin et al, 2009). NK cells have been subject to much scrutiny about whether they fall under the ‘responsive’ or ‘hyporesponsive’ categories. The rheostat concept was proposed based on observations that NK cells may fall anywhere between the two states and that this responsiveness can vary quantitatively. The Tuning model uses the Arming/Disarming models combined with the Rheostat concept (Shifrin et al, 2014). The tuning model proposes that each NK cell, depending on the inhibitory and activating receptors expressed, and the MHC genotype of the individual, receives a varying degree of stimulation and inhibition when interacting with target cells. The NK cell processes all of the activating and inhibitory signals and depending on the overall net stimulation received, the NK cell assumes a responsive state that is quantitatively appropriate for the level of stimulation, i.e. if an NK cell receives strong net stimulation it can drive the NK cell to its lowest responsive state (hyporesponsive), whereas weak (or no) net stimulation can drive the NK cell to its highest responsive state (effector). Intermediate net stimulation results in intermediate responsiveness (Shifrin et al, 2014).
1.10 Human NK cell receptors summary
To date it has indeed been found that NK cells express a plethora of receptors. There are many NK cell receptors that bind HLA Class I that have been well documented to date (Biassoni et al, 2001; Parham 2005; Gardiner, 2007; Cheent and Khakoo 2009). There are numerous families of receptors; The polymorphic KIR receptor family (Killer Immunoglobulin- like Receptor) arising from the Immunoglobulin-like receptor superfamily, the conserved CD94/NKG2 heterodimeric receptors (Biassoni et al, 2001; Lanier, 2008), the LIR/ILT receptors (Leukocyte Ig-like receptor/Ig-like transcript) and the activating natural cytoxicity receptors (NCRs).
1.10.1 The KIR receptor family overview
In total, 15 KIR genes (KIR2DL1, KIR2DL2, KIR2DL3, KIR2DL4, KIR2DL5A, KIR2DL5B, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS4, KIR2DS5, KIR3DL1, KIR3DL2, KIR3DL3, KIR3DS1) and 2 pseudogenes (KIR2DP1, KIR3DP1) exist in humans, encoded on chromosome 19q13.4 (IPD KIR database, 2013; Hsu et al, 2002; Moretta, 2004; Gardiner, 2007). Taken together with the ILT/LIR gene family, this region is called the Leukocyte receptor complex (LRC) (Trowsdale et al, 2001). The LRC is composed of a centromeric and telomeric region, which is present on all KIR genotypes. The KIR locus is divided into centromeric and telomeric regions which have highly diverse and complex gene content.
24 This will be discussed further in a later section.
Like their HLA Class I ligand counterparts, KIRs constitute a highly polymorphic gene family, which are expressed on all NK cells and a subset of CD8+ T cells. Each individual possesses a different number of KIR genes of which there are many different alleles that are clonally expressed on NK cells (Moretta, 2004; Parham, 2005). There is also further variation with regard to both the frequency and number of specific KIR genes expressed on the total NK cell population within that individual (Gardiner, 2007; Parham, 2005). Without a doubt, expression of KIR receptors is variegated and stochastic. The variability in the gene content at the KIR locus arises because of gene duplication and non-allelic homologous recombination (Martin et al, 2003; Martin et al, 2004). A fundamental aspect about KIR receptors is that they can transduce either inhibitory or activating signals when bound to their respective HLA class I ligands on target cells. The inhibitory KIR have been well characterised to date and they bind the classical HLA Class Ia ligands (HLA-A, B and C) as well as a short peptide of 8-10 amino acids (aa) that is presented on all MHC Class I molecules (Vilches and Parham, 2002). Inhibitory signals must dominate over activating signals to prevent the unwarranted killing of healthy HLA-Class I-expressing cells. The fine balance of NK cell inhibition and activation resides predominantly on the engagement of these specific receptors.
1.10.2 The NKG2/CD94 receptor family overview
The other main family of NK cell receptors are the NKG2/CD94 heterodimeric C-type lectin receptors, which are very conserved when compared to the polymorphic KIR family. Like the KIR family, this family of receptors also consists of both inhibitory and activating members. There are four primary NKG2 genes, NKG2A, NKG2C, NKG2E and NKG2D/F, which are encoded on chromosome 12p12.3-p13.1 (Lanier, 1998). Perhaps the most-well studied inhibitory member is NKG2A and the key activating members are NKG2D and NKG2C. CD94 is an NK cell receptor that belongs to the C-type lectin family and usually forms heterodimeric complexes with the NKG2 receptors by disulphide bonds (Lazetic et al, 1996; Braud et al, 1998; Lee et al, 1998). The only member of the NKG2 family that does not form a heterodimeric complex with CD94 is NKG2D. Instead NKG2D forms homodimeric complexes (Li et al, 2001; Vivier et al, 2002).
The ligand for the NKG2/CD94 family of receptors was found to be the non-classical HLA Ib molecule, HLA-E (Braud et al, 1998; Lee et al, 1998). HLA-E is also non-polymorphic in comparison to the other HLA molecules. In order for HLA-E to engage with its NKG2/CD94
25 ligand it must also bind a leader peptide derived from other HLA molecules such as HLA-A, - B, -C and -G. HLA-E cannot bind its own leader peptide to inhibit NK cells (Braud et al, 1998). As HLA-E binds the leader peptides for other HLA Class I molecules it can sense any downregulation in MHC Class I expression that may signal a viral infection (Cassidy et al, 2014). The NKG2 receptors possess a cytoplasmic domain, which is responsible for transmitting the inhibitory or activating signal whereas the CD94 moiety is the binding region for HLA-E and other ligands (Brooks et al, 1997). The NKG2A inhibitory receptor has two ITIM (Immunoreceptor Tyrosine Inhibitory Motif) domains in its cytoplasmic tail responsible for inhibitory signalling (Brooks et al, 1997).
It has been found that the activating NKG2D homodimer can engage with stress associated molecules such as MICA/B (MHC Class I related antigen A/B), which can be upregulated during times of cell stress or infection and DNA damage (Bauer et al, 1999; Mistry and O’Callaghan, 2007). This provides an extra signal to NK cell activating receptors, initiating a cytotoxic response. Additionally, a signal peptide derived from heat-shock protein 60 (Hsp60) was found to be upregulated on the surface of stressed cells and this was able to abrogate the inhibitory CD94/NKG2A-HLA-E interaction (Michealsson et al, 2002). Thus, the C-type lectin-like receptors have at least two distinct mechanisms for recognition of cellular stress.
1.10.3 NCRs (NKp30, NKp44, NKp46)
The activating natural cytotoxicity receptors (NCRs) are specifically expressed on the NK cell population and contain three members NKp30, NKp44, NKp46 (Vitale et al 1998; Pessino et al, 1998; Moretta et al, 2001). NCRs are directly involved in the process of NK cell activation by initiating cytotoxicity (Moretta et al, 2001; Biassoni et al, 2001). The three NCR member names were coined after their respective molecular weights. The ligands for the NCRs had remained elusive until recently. In 2001 they were suspected to bind viral haemagluttinins (Mandelboim et al, 2001). More recently, it has been shown that NKp30 can bind to a tumour-cell surface molecule called B7-H6 (Brandt et al, 2009). NKp44 was also recently found to bind a leukemia transcript (Rajagopalan and Long, 2013; Baychelier et al, 2013).
1.10.4 LILR/ILT/MIR receptors
The leukocyte Immunoglobulin-like receptors (LILRs) are type I glycoproteins found on NK cells and are members of the Immunoglobulin superfamily (Moretta et al, 1993; Braud et al, 1998; Brooks et al, 1998). They have been found to bind a broad range of HLA Class I molecules such as HLA-A, HLA-B, HLA-C and HLA-G, HLA-F (Chapman et al, 1999; Vitale et al, 1999; Navarro et al, 1999; Lapin et al, 2000). They have also been found to recognise
26 the HCMV (human cytomegalovirus) MHC Class I homolog UL-18 with a much greater binding affinity than for HLA Class I (Cosman et al, 1997). In particular, LILR receptors bind a relatively conserved region of MHC Class I and UL-18 that is located in the α3 domain. This explains the broad specificity of these receptors for different HLA Class I molecules. Structurally, LILR-1 has four extracellular Ig-domains and has four ITIMs in its cytoplasmic tail. It can associate with SHP-1 to inhibit NK cell activation (Cosman et al, 1997; Colonna, 1997).
1.10.5 2B4 receptor
2B4 (CD244) is a predominantly activating NK cell receptor that is broadly expressed on all NK cells and can mediate non-MHC restricted killing (Mathew et al, 1993). 2B4 is a member of the CD2 subset of the IgG family of receptors and its cellular ligand is CD48a glycosylphosphatidylinositol-anchored molecule expressed on all hematopoietic cells. Usually, 2B4 serves as a coreceptor and participates in homophilic or heterophilic interactions (McNerney et al, 2005; Lee et al, 2006). Human 2B4 is capable of inducing NK cell cytotoxicity and the release of cytokines IFN-γ and TNF-α when it acts as a co-receptor and it can also synergise with the NCRs (NKp30, NKp44 and NKp46) and the activating NKG2D receptor (Sivori et al, 2000; Watzl and Claus, 2014). For the most part 2B4 plays an activating role in NK cell biology, however it has been found to mediate inhibitory signalling in other cases (Watzl and Claus, 2014).
1.11 Ly49 receptors in mice
The mouse lectin-related Ly49 receptor family are members of the type II glycoproteins, which belong to the larger C-type lectin superfamily (Anderson et al, 2001). They possess an entirely different structure to human KIR but have remarkably evolved to carry out extremely similar functions. Similar to the KIR receptors, the Ly49 receptors also encode both activating and inhibitory receptors. The inhibitory Ly49 receptors bind MHC class I and inhibit NK cell attack against healthy cells but allow a response against cells that have lost class I expression. Similar to their human counterparts, the function of the activating Ly49 receptors has remained elusive (Arase and Lanier, 2002). Despite their differences to human KIR in their extracellular domains the Ly49 receptors possess highly similar intracellular signalling mechanisms as KIR. The Ly49 family has provided a useful model to investigate NK cell education, function and anti-viral immune responses. Interestingly, Ly49 receptors can bind to MHC class I molecules in a different orientation to KIR. They can undergo both trans- interactions with MHC Class I molecules (cell-to-cell) and cis-interactions with MHC Class I
27 expressed in the plane of the same membrane. Therefore there may some important differences in their functions with those of the human KIR (Held and Mariuzza, 2008).
1.12 KIR nomenclature and structure
KIR were originally cloned by Marco Colonna in 1995 and termed NKAT (NK-associated transcripts) (Colonna and Samaridis, 1995). They were then named according to their respective molecular weights (Long et al, 1996). For example, the 70 kDa KIR3DL1 receptor was first termed ‘p70’ and 58 kDa KIR2DL1 was ‘p58.1’. Now they are referred to by their respective KIR name (Long et al, 1996; Selvakumar et al, 1997; Andre et al, 2001; Marsh et al, 2002). All KIR also possess individual Cluster of Differentiation (CD) numbers. KIR are named according to how many extracellular Ig-like domains they possess KIR2D (two domains D1 and D2) or KIR3D (three domains D0, D1 and D2) (Andre et al, 2001; Marsh et al, 2002). In KIR3D receptors, the D0 domain is the membrane distal domain joined to the other extracellular D1 and D2 domains. The extracellular Ig-like domains are the regions responsible for interacting with their respective HLA ligands (Vilches and Parham, 2002; Hsu et al, 2002).
KIR were then termed S (short) or L (Long) according to the length of their cytoplasmic domains (Andre et al, 2001). The inhibitory KIR have long cytoplasmic domains because they contain conserved signalling motifs called ITIMs (Immuno-tyrosine-signalling-motifs) and the activating KIR have short cytoplasmic domains without these signalling motifs (Vivier et al, 2004). The activating KIR do not possess any ITIMs and instead recruit positive signalling adaptor molecules, such as DAP12, to transmit their activating signals. Charged amino acids, such as Lysine, are commonly found in the transmembrane domain of most activating KIR to allow them to associate with the adaptor molecules (Vilches and Parham, 2002). The only exception to the above is the KIR2DL4 receptor which possesses a long cytoplasmic domain and a charged residue that can initiate an activating signal rather than an inhibitory one (Selvakumar et al, 1996; Faure and Long, 2002). More will be discussed in a later section on the intracellular signalling paths of KIR.
28
Fig 1.1 Structure of the main inhibitory and activating KIR receptors
Illustration depicting the extracellular and Intracellular structures of the KIR2DL/S, KIR3DL/S and KIR2DL4 receptors. ITIM (Immuno Tyrosine Inhibitory Motif). ITAM (Immuno Tyrosine Activatory Motif). Charged amino acids (K/D) are present in transmembrane region of the activating KIR for interaction with adaptor molecules.
1.13 HLA Class I nomenclature and structure
The MHC on chromosome 6 is a complex and highly polymorphic area of the human genome encoding greater than 220 genes (Parham, 1988; Parham, 2005; Robinson et al, 2013). In particular, the genes encoding MHC class I are located at the telomeric end and encode the polymorphic HLA-A, HLA-B and HLA-C (Parham, 2005). The first HLA alleles were named by the WHO (World Health Organisation) Nomenclature Committee for Factors of the HLA System in 1987 (Robinson et al, 2013). At that time in 1987, 12 class I alleles and 9 class II alleles were named. Around 25 years later in 2012, the WHO Nomenclature Committee was able to assign names to 1163 alleles (Robinson et al, 2013). HLA genes are named according to which HLA type they fall under (HLA- A, B, C, G, E, F) followed by the allele number (Marsh et al, 2002). Structurally, MHC Class I molecules are noncovalently linked heterodimers. The extracellular region is comprised of a heavy chain consisting of three α domains α1, α2 and α3 and a light chain which is a non-MHC protein called β2- microglobulin (β2m) (Bjorkman and Parham, 1990). The α1 and α2 domains of MHC Class I
29 molecules are the most variable parts and are responsible for interaction with both KIR and TCR (Parham et al, 1988). The α1 and α2 domains also contain the binding groove for the presentation of peptide to KIR and TCR (Bjorkman and Parham, 1990).
1.14 KIR Haplotypes and genetic diversity on the LRC
The KIR genes have been broadly defined into two main genotypes to date. The two genotypes were originally separated according to the different numbers of inhibitory and activating KIR genes they possessed (Uhrberg et al, 1997). KIR haplotype A was classified as individuals who possess up to eight main KIR genes which demonstrate allelic variation (Trowsdale et al, 2001; Valiante et al, 1997) and KIR haplotype B was classified as those who possess more KIR genes than just the eight main ones. In simple terms, KIR B haplotypes are defined by the possession of at least one of the following genes: KIR2DL5, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS5 and KIR3DS1, whereas, KIR A haplotypes do not possess any of those genes. Therefore, KIR B haplotype can also be defined by the fact that they possess one or more activating KIR genes whereas the KIR A haplotype group do not. KIR B haplotypes can contain any combination of KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS5, KIR3DS1 and KIR2DS4 genes (Carrington and Norman, 2003). The KIR B haplotype has been found to possess up to 12 KIR genes and has considerable variability. However, five framework KIR genes have been found in every individual tested to date regardless of haplotype and they are KIR2DL4, KIR3DL2, KIR3DL3 and KIR2DP1 and KIR3DP1.
The ratio of haplotypes found in each population usually varies between each ethnic group (Uhrberg et al, 2002). In most populations tested, the KIR A haplotype is found in about one- third of individuals. Variation on KIR B haplotype is generated by the gene content and whether or not certain KIR genes are present or absent. Variation on the KIR A haplotype relies more on allelic variation rather than variation of gene content. 26 KIR haplotypes have now been fully sequenced according to the IPD KIR database and for haplotype B there are at least 40 distinct haplotypes (IPD KIR database, 2013; Robinson et al, 2013). There is no official nomenclature system in use for the KIR haplotypes as of yet, according to the IPD KIR database.
The locus for the KIR genes is found on chromosome 19q13.4 9 (Wilson et al, 2000; Carrington and Norman, 2003). The KIR locus is divided into two gene clusters that are located at the centromeric and telomeric ends with a unique region separating them (Hsu et al, 2002; Pyo et al, 2010). The framework KIR genes KIR3DL3 and KIR3DL2 are located on the centromeric and telomeric ends respectively (Wilson et al, 2000). The other framework genes KIR3DP1 and KIR2DL4 are located in the central region (Wilson et al, 2000; Pyo et al,
30 2010). The rest of the KIR genes are dispersed around the centromeric and telomeric clusters. There are two types of gene-content motifs that are different and called Cen-A, Cen-B, Tel-A and Tel-B (Moesta and Parham, 2012). KIR Haplotype A is made up of the combination of Cen-A and Tel-A. Whereas, the group B haplotype can be made up of three other combinations of Cen-B with Tel-B, Cen-B with Tel-A, and Cen-A with Tel-B (Moesta and Parham, 2012). As mentioned previously, the variation in gene content at the KIR loci is mediated by asymmetric recombination. There is also homologous recombination at the center of the locus, which changes the different Cen and Tel motifs (Moesta and Parham, 2012; Wilson et al, 2000; Pyo et al, 2010). As shown in Fig 1.2, some KIR genes are found only in the centromeric or telomeric region. The main example of this is found with the KIR2DL5 gene. KIR2DL5 is encoded by alleles of two different loci KIR2DL5A and KIR2DL5B and map to different regions of the KIR-gene cluster (Gomez-Lozano et al, 2002). It has been proposed that KIR2DL5A and KIR2DL5B are products of a gene duplication event and after subsequent recombination are now separated on KIR haplotypes (Gomez- Lozano et al, 2002). A very interesting point to note is that the KIR genes located within the same cluster are in much stronger linkage disequilibrium than KIR genes that are located in different clusters (Uhrberg, 2005). Linkage disequilibrium has been defined as “the nonrandom association of alleles at two or more neighboring loci” (Gourraud et al, 2010). An important example of this is the linkage disequilibrium found between the KIR2DL2 and KIR2DS2 genes (Gourraud et al, 2010).
Figure 1.2 The KIR locus on chromosome 19q13.4
1.15 KIR Lineages
The evolution of the KIR genes in humans has been both rapid and convergent to keep up with the polymorphism of the MHC Class I genes (Khakoo et al, 2000; Vilches and Parham, 2002; Parham and Moffett, 2013). There are five phylogenetic lineages of KIR (I, II, III, IV
31 and V). The oldest are lineage I KIR which include the KIR2DL4, KIR2DL5A and KIR2DL5B genes that recognise HLA-G ligands (Parham and Moffett, 2013). Lineage II KIR consist of KIR3DL1, KIR2DL2, KIR3DS1 genes that recognise HLA-A and HLA-B molecules. Lineage III KIR are the largest group consisting of KIR2DL1, KIR2DL2/3, KIR2DS1, KIR2DS2, KIR2DS4, KIR2DS3/5, KIR2DP1 and KIR3DP1 that recognise HLA-C1 and HLA-C2 alleles. Lineage V KIR consists of KIR3DL3 and its ligand is not known at present (Parham and Moffett, 2013). The most recently evolved KIR are lineage III and lineage V. Lineage III KIR are also found in both orangutans and chimpanzees. In terms of evolution there was also an increase in the number of lineage III KIR in humans. Interestingly, the activating receptor KIR2DS4 is the only lineage III KIR found in both humans and chimpanzees (Parham and Moffett, 2013).
1.16 Inhibitory KIR Receptors
1.16.1 KIR2DL2/3
The KIR2DL receptors are some of the most best-studied inhibitory KIR found in humans. They are the most recently evolved KIR receptors and belong to the Lineage III KIR (Parham and Moffett, 2013). There are two main types of KIR2DL in humans: type 1 has extracellular Ig-like domains homologous to D1 and D2, whereas type 2 has extracellular Ig-like domains homologous to D0 and D2 (in humans, type 2 is represented by KIR2DL4 and KIR2DL5) (Moretta and Moretta, 2004).
In 1990, Moretta and colleagues identified two 58 kDa proteins that were found primarily on NK cells and were able to engage MHC Class I to regulate NK cell killing functions (Moretta et al, 1990). A monoclonal antibody called GL183 was screened for its ability to induce NK cell killing of target cells and it was found to essentially block KIR2DL2/3 interactions. Therefore, this led to the identification of both the KIR2DL2 and KIR2DL3 receptors (Morettaet al, 1990; Moretta etal, 1993). KIR2DL2 and KIR2DL3 were found to segregate as alleles of the same locus (Keaney et al, 2004). However, they differ with respect to binding their HLA-C1 ligand due to differences in the hinge angle between their extracellular D1 and D2 domains with KIR2DL2 being a stronger binder than KIR2DL3 (Moesta et al, 2008). KIR2DL2 and KIR2DL3 are highly similar in sequence and vary by no more than six amino acids in their extracellular domains between any two allelic products (VandenBussche et al, 2006). Interestingly, any amino acid differences that have been observed between KIR2DL2 and KIR2DL3 are not located within the region that interacts with HLA-C1 and the bound peptide. The differences in the hinge angle between KIR2DL2 and KIR2DL3 may allow greater flexibility for KIR2DL2 to engage more strongly with its ligand HLA-C1. As of 2013,
32 there are 29 known existing alleles of KIR2DL2 and 44 alleles of KIR2DL3 (IPD KIR database, 2013; Moesta and Parham, 2012). A key study by Winter and Long demonstrated that a Lysine residue at position 44 of KIR2DL2 and KIR2DL3 confers specificity to HLA- Cw*03 and other HLA-C1 alleles which will be detailed further in later sections (Winter and Long, 1997).
Table 1.1 Overview of the Inhibitory KIR members (Allele and protein numbers obtained from IPD KIR database, IPD database last updated 11th October 2013; Release 2.5.0). All novel KIR nucleotide sequences are reported by the IPD KIR database. However, Multiple KIR alleles shown here may code for the same KIR protein.
1.16.2 KIR2DL1
Moretta et al also identified KIR2DL1 in 1990 as a 58 kDa protein (Moretta et al, 1990). A different monoclonal antibody (EB6) identified KIR2DL1 (Moretta et al, 1993). There are now more than 44 alleles identified for the KIR2DL1 gene (IPD KIR database, 2013). KIR2DL1 is an inhibitory receptor that mediates a very strong interaction to its ligand HLA-C2 (Winter and Long, 1997; Winter et al, 1998). The co-crystal structure of KIR2DL1 complexed to one of its ligands HLA-Cw*04 was carried out by Fan and colleagues (Fan et al, 2001). Winter and Long also showed that methionine at position 44 (p44) of KIR2DL1 confers specificity to HLA-Cw*04 and other HLA-Cw*02 alleles, which will also be detailed further in later sections (Winter and Long, 1997).
1.16.3 KIR3DL1
KIR3DL1 is a 70kDa inhibitory receptor that contains three extracellular immunoglobulin (Ig) domains (3D) (D0, D1 and D2) and a long (L) cytoplasmic domain that contains two ITIMs (Vivian et al, 2011). There are more than 92 alleles now identified for the KIR3DL1 gene (IPD KIR database, 2013; Parham and Moesta, 2012). The specific ligand for KIR3DL1 has
33 been identified as HLA-B alleles with serological Bw4 and/or Bw6 epitopes. KIR3DL1 has also been shown to recognise some HLA-A alleles with the Bw4 motif such as HLA-A*2402 (Thananchai et al, 2007; Maloveste et al, 2012). KIR3DL1 has been implicated in numerous studies with HIV which will be discussed in further detail in a later section (Martin et al, 2002; Martin et al, 2007).
1.16.4 KIR3DL2
KIR3DL2 is structurally similar to KIR3DL1. It is one of the three framework KIR genes mentioned previously (along with KIR3DL3 and KIR2DL4) present on all haplotypes (Pende et al, 1996). The elusive ligand of KIR3DL2 has been under much debate and it is thought to bind to HLA-A3 and HLA-A11 although this interaction may be highly peptide-dependent (Dohring et al, 1996; Hansasuta et al, 2004). KIR3DL2 has been reported to bind HLA-A*11 and Lys44, Pro71 and Val72 are particularly important for this interaction (Graef et al, 2009). Human KIR3DL2 shares this Pro71/Val72 motif with the activating KIR2DS4 receptor (Graef et al, 2009). There are 86 alleles in existence for the KIR3DL2 gene (IPD KIR database, 2013). HLA-B27 dimers have also been recently shown to bind KIR3DL2 in a peptide- independent manner and this may be important in the development of ankylosing spondylitis (Kollnberger et al, 2007; Wong-Baeza et al, 2013).
1.16.5 KIR2DL4
KIR2DL4 is an interesting exception to all KIR receptors as it is expressed endosomally and has been classified as both an inhibitory receptor and an activating receptor (Faure and Long, 2002). It has been classified this manner because it contains a long cytoplasmic domain with one ITIM and it also possesses a charged amino acid in its transmembrane region (Selvakumar et al, 1996). The presence of this charged amino acid suggests it has an activating function (Faure and Long, 2002). Ponte et al have previously described inhibitory functions of KIR2DL4 and more recently other studies have described predominantly activating functions of KIR2DL4 (Ponte et al, 1999; Rajagopalan et al, 2001; Shajahan Miah et al, 2008; Rajagopalan and Long, 2012). The ligand for KIR2DL4 is soluble HLA-G that accumulates within the endosomes (Rajagopalan and Long, 2012). There are 51 known alleles identified for the KIR2DL4 gene (IPD KIR database, 2013).
1.16.6 KIR2DL5
The polymorphic KIR2DL5 gene was most recently identified in 2000 and is around 60 kDa in size when expressed as a monomer on the NK cell surface (Vilches et al, 2000; Estefania et al, 2007). Like KIR2DL4, KIR2DL5 is an ancestral KIR and has Ig-like domains of the D0-
34 D2 type (Cisneros et al, 2012). KIR2DL5 possesses alleles that originate from two different loci, KIR2DL5A Telomeric (T) and KIR2DL5B Centromeric (C), and these are commonly referred to as two separate KIR genes as they map to different regions of the KIR cluster (Fig 1.2) (Gomez-Lozano et al, 2002). Four haplotypes of KIR2DL5 exist; one haplotype has both KIR2DL5A and KIR2DL5B, the second and third haplotypes have either KIR2DL5A or KIR2DL5B, and finally the fourth haplotype does not possess KIR2DL5 (Gomez-Lozano et al, 2002). Like the other inhibitory KIR members KIR2DL5 transmits inhibitory signals intracellularly through the recruitment of SHP-2 phosphatase (Estefania et al, 2007). Relatively little is known to date regarding the ligand of KIR2DL5 and its protective or susceptible relationship with disease (Cisneros et al, 2012).
1.17 Ligands for the Inhibitory KIR: HLA Class I
The ligands for the inhibitory KIR are MHC Class I molecules, which are normally expressed on all healthy cells. The genes encoding MHC Class I molecules in humans are the HLA Class I genes found on chromosome 6p21.3, which are also highly polymorphic at the allelic level (Parham, 2005). MHC Class I molecules also engage with T cell receptors on CD8+ CTLs, which triggers adaptive immune responses (Single et al, 2007). The main ligands that have been defined for KIR to date are the HLA-C and HLA-B alleles.
1.17.1 HLA-C
HLA-C1 and HLA-C2 have been defined as the ligands for the inhibitory receptors KIR2DL2/3 and KIR2DL1 respectively. The predominant HLA-C1 alleles can be classified as HLA-Cw*01, HLA-Cw*03, HLA-Cw*07 and HLA-Cw*08 (Marsh et al, 2000; Parham and Moffett, 2013). The main HLA-C2 alleles can be classified as HLA-Cw*02, HLA-Cw*04, HLA- Cw*05 and HLA-Cw*06 (Marsh et al, 2000). Winter and Long showed that methionine at p44 of KIR2DL1 confers specificity to HLA-Cw*04 and other HLA-Cw*02 alleles (Winter and Long, 1997). Whereas, a lysine residue at p44 of KIR2DL2 and KIR2DL3 is essential to confer specificity to HLA-Cw*03 and the other HLA-C1 alleles (Winter and Long, 1997).
Later crystallisation studies of inhibitory KIR, such as KIR2DL1 and KIR2DL2, all revealed that residues 77 and 80 of the class I HLA-C α1 heavy chain were important for HLA-C binding to KIR2DL receptors (Boyington et al, 2000; Fan et al, 2001; Synder et al, 1999). The inhibitory KIR receptors KIR2DL2 and KIR2DL3 were found to specifically bind Group 1 HLA-C (HLA-C1) alleles that have a serine residue at P77 and an asparagine residue at P80 (Boyington et al, 2000). Whereas, the inhibitory KIR2DL1 receptor binds specifically to Group 2 HLA-C (HLA-C2) alleles that have an asparagine residue at position 77 and a lysine
35 residue at position 80 (Fan et al, 1997; Fan et al, 2001). Maenaka and colleagues successfully crystallised the KIR2DL3 receptor in 1999 (Maenaka et al, 1999).
It was shown that KIR2DL2/3 also demonstrated weak binding to HLA-C2 in-vitro (Moesta et al, 2008). From these results, it was proposed that the KIR2DL2/3-HLA-C1 pairing evolved first and is older than the KIR2DL1-HLA-C2 pairing which is a very strong inhibitory interaction and is highly specific for HLA-C2 only (Moesta et al, 2008; Hilton et al, 2012). HLA-C1 also predominantly emerged in the orangutans which suggest that it evolved before HLA-C2 in humans. For example, the HLA-Cw*01 epitope evolved first and functioned as a KIR ligand before the HLA-C2 epitope.
Fig 1.3 Interaction of KIR2DL1 and KIR2DL2/3 receptors with their respective HLA-C2 and HLA-C1 ligands
1.17.2 HLA-B
The only KIR found to bind HLA-B alleles is the inhibitory receptor KIR3DL1. KIR3DL1 binds the serological Bw4 motif (position 77-83) that is shared amongst HLA-B alleles and there has been some weak binding reported with Bw6 motifs (Cella et al, 1994; Gumperz et al, 1995; Carr et al, 2005). It appears that Bw4 alleles with Isoleucine at position 80 (Bw4–80I) are the best group of ligands to mediate binding to most of the KIR3DL1 alleles identified (Cella et al, 1994; Carr et al, 2005). Other evidence has suggested that Bw4 alleles with threonine at position 80 are also good ligands for other KIR3DL1 alleles. The main example of this is allele HLA-Bw*2705 (Luque et al, 1996; Martin et al, 2007). However, Barber et al have previously reported one HLA-Bw*4601 allele that is a recombinant between HLA-B and HLA-C and can mediate inhibition of NK cells in the same manner as HLA-C does (Barber et al, 1996).
36 1.17.3 HLA-A
KIR3DL2 may have specificity for HLA-A3 and HLA-A1 molecules although this is still under some ongoing debate at present as it may be highly peptide-dependant (Dohring et al, 1996; Hansasuta et al, 2004).
1.17.4 HLA-G
Other less well-studied ligands for KIR receptors are the conserved HLA-G molecules, which are usually expressed on fetal trophoblasts, thymic endothelial cells and the cornea (Apps et al, 2008). Soluble HLA-G has been found to be a ligand for KIR2DL4 (Rajagopalan and Long, 1998; Rajagopalan and Long, 2012). Interestingly, when a peptide derived from the leader sequence of HLA-G is presented on HLA-E molecules it has been found to mediate a strong inhibitory interaction between the inhibitory CD94/NKG2A receptor and HLA-E (Braud et al, 1998; Llano et al, 1998; Petrie et al, 2008).
1.18 Activating KIR and their HLA Class I ligands
Overall, the activating KIR members possess a large amount of sequence similarities to their inhibitory counterparts in their extracellular domains. There is a considerable degree of linkage disequilibrium between inhibitory KIR genes and the activating KIR genes. In particular, this is found between KIR2DL2 and KIR2DS2 and between KIR3DL1 and KIR3DS1 (Uhrberg et al, 1997; Uhrberg, 2005; Gourraud et al, 2010). Unlike the inhibitory KIR, there is a lack of specific monoclonal antibodies available for the activating KIR, which has hampered significant research into these important members. However, there has been some progress made recently by David et al into the development of monoclonal antibodies that can recognise KIR2DS2 and KIR2DS1 (David et al, 2009). The downside into research on the activating KIR is also that the ligands for the majority of activating KIR continue to remain most ambiguous. As we have only mentioned the ligands for the inhibitory KIR we must address the issue of HLA ligands for the activating KIR. Most of the activating KIR were thought to share the same HLA class I ligands as the inhibitory KIR due to their overall sequence similarities. The activating KIR were discovered in 1995 when Moretta and colleagues identified 50 kDa receptor proteins on NK cells that resembled the previous 58 kDa proteins that were now known to be KIR2DL1 or KIR2DL2/3 (Moretta et al, 1995). Moreover, the engagement of these new 50 kDa proteins was able to stimulate the NK cell rather than inhibit its effector functions. The main difference that was found between the p58 and p50 receptors was with regard to the intracellular domain (Biassoni et al, 1996). The p50
37 receptors had shorter cytoplasmic domains and a transmembrane region containing a charged lysine residue.
Table 1.2 An overview of the Activating KIR members (Allele and protein numbers obtained from IPD KIR database, last updated 11th October 2013; Release 2.5.0). All novel KIR nucleotide sequences are reported by the IPD KIR database. However, Multiple KIR alleles shown here may code for the same KIR protein.
1.18.1 KIR2DS1 and ligand HLA-C2
KIR2DS1 is probably the most well-studied activating KIR. 15 alleles of KIR2DS1 have been identified to date (IPD KIR database, 2013; Parham and Moesta, 2012).To date, the best- known ligand for an activating KIR is HLA-C2 for KIR2DS1. Stewart et al showed that KIR2DS1 tetramers can weakly bind to some HLA-C2 alleles (HLA-Cw*04:01 and HLA- Cw*06:02) in the presence of specific peptides. However, the binding affinity of KIR2DS1 to these particular alleles was not as strong as the binding of its inhibitory counterpart KIR2DL1 (Biassoni et al, 1997; Stewart et al, 2005). Further binding studies by Moesta et al demonstrated binding of KIR2DS1-Fc to bead bound HLA Class I (Moesta et al, 2010). In terms of functional studies on KIR2DS1, Chewning et al showed there was induced killing and IFN-γ release against HLA-C2-bearing target cells (Chewning et al, 2007). Foley et al also showed KIR2DS1-mediated killing against HLA-C2-bearing PHA blast cells (Foley et al, 2008). Finally Pende et al, demonstrated binding of KIR2DS1-Fc proteins to HLA-Cw*06 on transfected 721.221 cells and then demonstrated killing of HLA-C2/C2 leukemic blast cells (Pende et al, 2009).
1.18.2 KIR2DS2 and HLA-C1
The activating receptor KIR2DS2 possesses 99% sequence similarities with the inhibitory receptor KIR2DL2 (Vales-Gomez et al, 1998). However, the binding affinities of KIR2DS2 and KIR2DL2 are remarkably different. Vales-Gomez et al used surface plasmon resonance
38 (SPR) analysis to demonstrate that the affinity of KIR2DS2 for HLA-C1 was much lower than its inhibitory counterpart KIR2DL2 (Vales-Gomez et al, 1998). There are 22 known alleles of KIR2DS2 that have been identified to date (IPD KIR database, 2013). The ligand for KIR2DS2 still remains largely unknown and it was thought to be HLA-C1, the same ligand as its inhibitory counterpart KIR2DL2/3. Winter et al had failed to show binding of KIR2DS2-Fc to 721.221 cells transfected with HLA-Cw*0102, HLA-Cw*0304 and HLA-Cw*0702 alleles (Winter et al, 1998). However, a recently published structural study has shown co- crystallisation of KIR2DS2 with HLA-A*11:01 molecules and a vaccinia-derived peptide (Liu et al, 2014). The crystal structure of the KIR2DS2 receptor had been documented previous to this in 2003 by Saulquin et al and it failed to bind HLA-Cw*0304 (Saulquin et al, 2003). Saulquin et al showed that activating KIR2DS2 does not bind HLA-Cw*03 molecules that are recognised by inhibitory KIR2DL2, despite 99% extracellular amino acid identity between KIR2DL2 and KIR2DS2. Saulquin et al also demonstrated that two mutations of residues (Tyr45 and Gln71) were involved in determining the interaction of KIR2DL2 with HLA-Cw*03 (Saulquin et al, 2003). KIR2DS2 tetramers failed to bind both HLA-Cw*03- and HLA-Cw*04- transfected 721.221 cells (Saulquin et al, 2003). Importantly, Stewart et al also investigated binding of KIR2DS2 to HLA-Cw*0302 and reported very weak binding only in the presence of some endogenous peptide derivatives (Stewart et al, 2005). A very recent functional study carried out by David et al has now shown that KIR2DS2+ NK clones stained with a monoclonal antibody specific for KIR2DS2 can recognize and kill HLA-Cw3+ target cells (David et al, 2013).
1.18.3 KIR2DS4
In 2009 the crystal structure of KIR2DS4 was also documented (Graef et al, 2009). It was reported that KIR2DS4 is found in both functional (full-length) and non-functional (deleted) forms (Graef et al, 2009). In the same year it was shown that KIR2DS4 can interact with numerous alleles, HLA-A*11:02, HLA-Cw*05:01 and HLA-Cw*16:01 (Graef et al, 2009). To date, both KIR2DS2 and KIR3DL2 have been reported to bind HLA-A*11 and Lys44, Pro71, and Val72 are important for this interaction (Liu et al, 2014; Jones et al, 2006; Winter et al, 1998). KIR2DS4 was previously shown to exhibit HLA-A*11:01 and HLA-A*11:02 binding specificity by means of a Pro71/Val72 motif at the D1 domain (Graef et al, 2004). Human KIR2DS4 shares this Pro71/Val72 motif with KIR3DL2 and it is believed to have arisen through a gene conversion event between the ancestral lineages of these two genes (Graef et al, 2009). Katz et al also reported that KIR2DS4 may be also able to bind other ligands independently of MHC Class I (Katz et al, 2004). Interestingly, one of these non-MHC Class I ligands for KIR2DS4 may be a protein expressed on melanoma-derived tumour cells (Katz et
39 al, 2004).
1.18.4 KIR2DS3 and KIR2DS5
The ligands for the other two remaining activating KIR members KIR2DS3 and KIR2DS5 have not been as well described to date. There are 15 known alleles for the KIR2DS3 gene and 18 for the KIR2DS5 gene (Moesta and Parham, 2012; IPD KIR database, 2013).
1.18.5 KIR3DS1
Alongside KIR2DS1, the KIR3DS1 gene is one of the much-discussed activating KIR genes, particularly with regard to its role in HIV infection which will be discussed in further detail later. KIR3DS1 is an activating KIR receptor, an allele of KIR3DL1, which has been implicated in the slower progression of HIV infection. There are 16 known alleles for the KIR3DS1 gene (Moesta and Parham, 2012; IPD KIR database, 2013). It was thought originally that KIR3DS1 was not expressed on NK cells. It has been since found that KIR3DS1 is indeed expressed (O’Connor et al, 2007; Trundley et al, 2007; Carr et al, 2007; Pascal et al, 2007). However, it has still not been shown whether there is actually any direct interaction of KIR3DS1 with HLA-Bw4 alleles despite intense investigation (Carr et al, 2007; Single et al, 2007).
1.19 Co-crystallisation and structure of the Inhibitory and Activating KIR2DL receptors with their respective HLA-C ligands
The KIR2D molecules have two N-terminal Ig-like extracellular domains (D1 and D2) joined by a linker sequence of around 3-5 amino acids. There are two anti-parallel β sheets that make up each domain. There is a hinge angle between each of the domains that varies in each receptor. For the free-form KIR2DL1 receptor the hinge angle is 55° and for the free- form KIR2DL2 receptor the hinge angle is 84° (Fan et al, 1997; Fan et al, 2001; Synder et al, 1999). The different hinge angles between the two domains are of the utmost importance as they help regulate binding. A conserved salt-bridge that is found in all KIR families restricts this hinge angle (Boyington and Sun, 2002). Interestingly, when both KIR2DL2 and KIR2DL1 are engaged with their respective HLA-C ligand their hinge angle becomes quite similar (Fan et al, 2001; Boyington et al, 2001). The hinge angle between the D1 and D2 domains of KIR2DL1 is 66° when KIR2DL1 is complexed with HLA-Cw*04 (Fan et al, 2001) and the hinge angle between the D1 and D2 domains of KIR2DL2 is 81° when complexed with HLA- Cw*03 (Boyington et al, 2000).
Maenaka et al reported that the hinge angle of the free-form KIR2DL3 receptor is 78°
40 (Maenaka et al, 1999). In 2008, Moesta and colleaguesdemonstrated that there was a stronger binding affinity of KIR2DL2 to HLA-C1 compared with KIR2DL3. This difference in binding affinities was not related to differences in the ligand-binding site or peptide-binding site of the receptors but is instead related to residues in the hinge angle between the extracellular Ig-like D1 and D2 domains (Moesta et al, 2008).
Three important loops from the D1 and D2 domains of the KIR2D receptors are responsible for binding to the α1 and α2 helices of HLA-C respectively (Boyington and Sun, 2002). Importantly, the D1 and D2 domains of the KIR2D receptors also make direct contact with the peptide that is presented on all HLA-C molecules. KIR2DL2 makes direct contact with p7 and p8 of peptide (Synder et al, 1999; Boyington et al, 2000; Boyington et al, 2001). Boyington et al found that KIR2DL2 contact with HLA-Cw*03 requires position 8 of the peptide to be a residue smaller than valine (Boyington et al, 2000). KIR2DL1 only interacts with p8 of peptide and also requires this to be a small amino acid residue such as alanine (Fan et al, 2001).
Other structural and kinetic studies have shown that KIR binds to HLA-C in a 1:1 stoichiometry (Fan et al, 1996; Vales-Gomez et al, 1998; Maenaka et al, 1999). Overall, the KIR/HLA interface is highly charged and has salt-bridges that are essential for the interaction. In further detail, the KIR2DL2-HLA-Cw*03 interface contains an overall 12 residues contributed from HLA-Cw*03 (Boyington et al, 2000; Boyington et al, 2001). The only variable residue out of these 12 is at position 80 (Boyington et al, 2000). Lysine 44 (Lys44) in KIR2DL2 has been found to make a critical hydrogen bond with Asparagine at position 80 of HLA-Cw*03 (Boyington et al, 2001). In comparison with KIR2DL1, it has a methionine at position 44 and this forms a bond with Lysine 80 of HLA-Cw*04 (Fan et al, 1997; Fan et al, 2001). This Met44 to Lys80 interaction is mediated by both polar and hydrophobic interactions (Fan et al, 2001).
With regard to crystallisation studies performed on the activating KIR2DS members, Saulquin and colleagues performed a key crystallisation study on KIR2DS2 in 2003 (Saulquin et al, 2003). The hinge angle between the D1 and D2 domains of KIR2DS2 was reported by Saulquin et al to be 73° (Saulquin et al, 2003). The KIR2DS2 receptor has 99% sequence similarity to KIR2DL2 with the exception of one residue located at position 45. KIR2DS2 possesses tyrosine at position 45 (tyr45) whereas KIR2DL2 has a phenylalanine (phe45) (Vales-Gomez et al, 1998; Saulquin et al, 2003). This single amino acid change is enough to abrogate all binding to HLA-C1 molecules and it has been proposed that KIR2DS2 evolved away from binding the HLA-C1 epitope at once point using this single residue substitution (Winter et al, 1998; Older Aguilar et al, 2011). Vales-Gomez et al
41 showed that KIR2DL2 and KIR2DS2 bound to HLA-Cw*07 with remarkably different binding affinities with KIR2DS2 having weak or no affinity for HLA-Cw*07 (Vales-Gomez et al, 1998). Restoration of some degree of KIR2DS2 binding to HLA-C1 has been reported when Phe45 replaces Tyr45 (Saulquin et al, 2003). As reported by Boyington et al, Gln71 is also involved in the KIR2DL2–HLA-Cw*03 interface via two hydrogen bonds; one water-mediated and one with the Alanine at p8 of the GAVDPLLAL (GAV) peptide (Boyington et al, 2000). However, in the KIR2DS2 structure, the fully solvent accessible Gln71 side chain is positioned in a way that both hydrogen bonds are not possible, which ultimately restricts KIR2DS2 successully binding to HLA-C1. The only other activating KIR members that have been crystallised to date are KIR2DS4 by Graef et al in 2009. The hinge angle for KIR2DS4 was reported to be 69° (Graef et al, 2009).
1.20 Expression of KIR on T cells
KIR receptors are expressed on a small subset of T cells and have been found to modulate TCR-mediated T cell responses. The majority of KIR expression on T cells has been on the cytotoxic CD8+ T cell population rather than the CD4+ T helper cell population (Mingari et al, 1996). Interestingly, all of the known KIR can be expressed on T cells (Mingari et al, 1998). Most KIR+ T cells are found to be from the activated memory T cell subset that have already undergone clonal expansion. Expression of inhibitory KIR receptors on CTLs was of initial concern as it was hypothesised to lead to inhibition of the cytotoxic T cell response against viral infection or tumourigenesis. However, expression of inhibitory KIR on CTLs may actually help to control or regulate the CTL response in situations where the CTL may crossreact with self-antigens. High frequencies of KIR+ T cells have been detected in HIV- infected patients (De Maria et al, 1977). It was found that KIR expression on CTLs can also inhibit cytokine production such as IFN-γ (Hengartner, 1996). Seich Al-Basatena et al recently showed that expression of KIR2DL2 on the CD8+ T cell population enhanced HLA- class I mediated immunity in Hepatitis C virus infection (HCV) and HTLV-I (Human T cell Leukemic virus) infection. They proposed that, in the face of chronic antigen stimulation, such as with HCV, protective T cells may survive longer if they possess the KIR2DL2 gene. They propose that when a CD8+ T cell is activated by engagement of its TCR by the cognate HLA:peptide complex, the CD8+ T cell will be less likely to undergo Activation- induced cell death (AICD) if it co-expresses KIR2DL2 which binds HLA-C1 on an infected cell. These KIR2DL2+ T cells may be able to provide stronger regulatory protection in chronic infection (Seich Al-Basatena et al, 2012).
42 1.21 The Inhibitory KIR synapse and intracellular signalling
One of the main structural features of the inhibitory KIR is the presence of ITIMs in their intracellular cytoplasmic tails (Bolland and Ravetch, 1999). Each ITIM consists of the consensus sequence I/V/LxYxxL/V (Burshtyn and Long, 1997). Most inhibitory KIR have two ITIMs within their cytoplasmic domain (Vivier and Daeron, 1997). Upon engagement of their HLA Class I ligands, the inhibitory receptors become phosphorylated on the tyrosine residues of their ITIMs and recruit specific Src homology (SH) 2-containing phosphatases (SHP-1, SHP-2). SHP1/SHP2 are crucial for transmitting the inhibitory signal to the rest of the NK cell as they block activation signals (Campbell et al, 1996). The tyrosine phosphorylation of the ITIMs is critical for the recruitment of SHP1/SHP2. Further downstream, SHP1/SHP2 dephosphorylates Vav1, a key substrate, which results in blocking the actin cytoskeletal rearrangements of the NK cell. Overall, SHP1/SHP2 play a crucial role in inhibiting conjugate formation, Ca2+ influx and NK cell degranulation.
1.22 The Activating KIR synapse and intracellular signalling
In general, the activating KIR receptors have a high level of sequence similarities to their corresponding inhibitory KIR receptors. The main difference is that the activating KIR have a short cytoplasmic tail lacking any ITIMs. Alongside this, they also have a slightly different transmembrane region in that they usually have a positively charged amino acid such as lysine in this domain (Lanier, 2005; Lanier, 2008). This charged lysine residue allows the association of the activating receptors with DAP12/KARAP, CD3ζ and FcεR1γ adaptor molecules and allows them to transmit their activating signals (Lanier, 2008). The adaptor molecules contain an ITAM (Immuno Tyrosine Activating Motif), which consists of a negatively charged amino acid, such as aspartic acid, and this allows them to interact with the charged lysine residue in the activating KIR transmembrane domain (Lanier, 2008). DAP12 signalling and ITAM signalling all eventually result in the activation of the mitogen- activated extracellular signalling regulated kinase (ERK) (Vivier et al, 2004). Usually, Src- kinase dependent phosphorylation of DAP12 occurs which recruits Syk and allows the initiation of LAT-dependent signaling pathways (PI3K-AKT, NKFAT and MAPK) (Davis, 2002; Vivier et al, 2004). This signalling cascade eventually leads to release of the cytotoxic granules by NK cells. The immunological synapse that forms between an NK cell and its respective target cell has been referred to as the activating synapse (Davis, 2002). The formation of the activating synapse resides heavily on rearrangements of the actin cytoskeleton and microtubule organisation in order to transport molecules such as granzymes into the target cell (Davis et al, 1999; Davis, 2002; Orange et al, 2003).
43 1.23 Co evolution of the KIR and HLA genes and outcome of viral infection
It has been through the study of different diseases that the relevance of certain KIR-HLA pairings has come into focus. Although they are inherited physically independent of each other on separate chromosomes, certain KIR-HLA ligand pairings were found to confer protection in numerous diseases or grant susceptibility to certain infections or autoimmune diseases (Kulkarni et al, 2008).
1.23.1 KIR and HIV infection
Some of the most informative studies on KIR:HLA interactions in relation to disease have perhaps been with HIV (human immunodeficiency virus) infection and measurement of its progression to AIDS (acquired immunodeficiency syndrome). It has been found that there is a positive correlation between slower progression to AIDS infection and possession of KIR3DL1 and HLA-B*57 genes (Martin et al, 2002; Migueles et al, 2000; Flores-Villanueva et al, 2001; Lopez-Vazquez et al, 2005; Martin et al, 2007). In a study carried out on African sex workers in the Cote d’Ivoire it was found that possession of inhibitory KIR receptors but absence of their cognate HLA ligands was associated with increased protection against transmission of HIV (Jennes et al, 2006). For example, some of the sex workers possessed KIR3DL1 in the absence of HLA-Bw4 and some possessed KIR2DL2/3 in the absence of HLA-C1. This study proposed that in the absence of the ligands for the inhibitory KIR, they wouldn’t become engaged thus leading to less overall NK inhibition (Jennes et al, 2006).
A positive correlation between the activating receptor gene KIR3DS1 and delayed progression to AIDS has also been found (Martin et al, 2002). Martin et al published a finding that KIR3DS1-positive individuals infected with HIV progressed slower to AIDS but only if they also possessed HLA-Bw*04 with isoleucine at position 80 (Martin et al, 2002). Homozygousity for HLA-Bw*04 was found to be very important in the slower progression to AIDS as it may provide increased ligand to engage with KIR3DS1. However, it has still not been shown whether there is any direct interaction between KIR3DS1 and HLA-Bw*04 molecules (Single et al, 2007).
Tiemessen et al demonstrated that possession of at least one HLA-C1 allele is associated with higher NK cell responses to a HIV-1 Env peptide pool and also a decreased viral load (Tiemessen et al, 2011). There was also an increased possession of KIR2DL3+HLA-C1+ in responders compared with non-responders. Overall, Tiemessen et al found that the responder sub-group had possession of more activating KIR genes (KIR2DS2, KIR2DS1, KIR2DS5 and KIR3DS1) and also at least one HLA-C1 allele (Tiemessen et al, 2011).
44 1.23.2 KIR and Hepatitis C infection A significant study using a large HCV cohort (352 individuals with spontaneous clearance of HCV infection and 685 individuals with persistent infection) by Khakoo et al has shown that individuals homozygous for the inhibitory NK receptor gene KIR2DL3 and its human leukocyte antigen C group 1 (HLA-C1) ligand have a higher frequency of resolution of acute Hepatitis C virus (HCV) infection (Khakoo et al, 2004; Knapp et al, 2010). This may be explained in part by the weaker binding affinity of inhibitory KIR2DL3 to HLA-C1, thus allowing stronger NK activating signals to dominate (Khakoo et al, 2004; Moesta et al, 2008). KIR2DL2 is an allele that segregates from the same locus as KIR2DL3 and interestingly it was found that it did not mediate the same frequency of resolution of HCV infection (Khakoo et al, 2004). Interestingly, the protective effect of KIR2DL3/HLA-C1 against HCV was only observed in those individuals that had been infected via injected drug use or accidental needlestick. It was not observed in individuals that had been infected via blood transfusion. Khakoo et al explain that this was most likely due to the innate immune responses being overloaded with a large amount of virus that would undoubtedly be transmitted via transfusion (Khakoo et al, 2004).
HLA-C1 and KIR3DS1 were also found to be protective against the development of hepatocellular carcinoma in individuals with chronic HCV infection (López-Vázquez et al, 2005). Another study was conducted on a small liver transplant cohort (44 individuals) by Askar et al. In patients who had recurrent HCV infection after liver transplantation, the absence of KIR2DS2 and/or KIR2DL2 genes was associated with failure of Peg/RBV (Peginterferon and Ribavirin) therapy (Askar etal, 2009). This cohort was tested for sustained virological response (SVR) post-transplantation and only 2 of 18 (11%) patients who lacked the KIR2DS2/KIR2DL2genes achieved SVR compared to 13 of 26 (50%) who possessed these two genes (Askar et al, 2009).
In another significant study, certain HLA Class I alleles were also found to have an association with clearance or persistence of HCV infection. Thio et al found that HLA- A*1101, HLA-B*57 and HLA-Cw*0102 were associated with viral clearance, whilst HLA- A*2301 and HLA-Cw*04 were associated with viral persistence (Thio et al, 2002).
1.24 Hepatitis C virus and NK cells Hepatitis C is the only member of the hepacivirus genus within the flaviviridae family of viruses and an estimated 180 million individuals are infected worldwide (Rehermann and Nascimbeni, 2005). Of those that are infected with the virus, 60-80% develop chronic infection which can lead to liver cirrhosis and hepatocellular carcinoma (Rehermann and
45 Nascimbeni, 2005). NK cells have a pivotal role in early defence mechanisms against HCV (Amadei et al, 2010; Hiroishi et al, 2008). In acute stages of HCV infection, increased NK cell activation has been reported to occur (Amadei et al, 2010). Most of the clinical studies on HCV infection have been on the chronic phase of infection rather than the acute phase of infection due to the difficulties faced in obtaining patient samples during the acute phase. Long-term persistence of HCV may have a significant impact on the balance of NK cell responses in chronically infected patients. The innate immune system is particularly important in the early stages of HCV infection as it takes 1-2 months for the cellular and humoral adaptive system to ‘kick-in’. It is perceived that by this stage the virus has unfortunately had a chance to “outpace” the adaptive immune response (Thimme et al, 2001; Thimme et al, 2002). Early in infection HCV can induce changes in the expression of Type I Interferon genes (IFN α/β). It is partly through manipulation of the interferon response at the start that HCV can lead to a prolonged chronic infection. To date, HCV envelope (E2) proteins have been reported to have a negative effect on NK cell function by cross-linking the inhibitory surface molecule CD81 (Tseng and Klimpel, 2002). In order to combat later stages of HCV infection, a potent CD4+ T cell response is also required along with generation of virus-specific CTLs (Rehermann and Nascimbeni, 2005). It is known that approximately 20% of infected individuals can spontaneously clear HCV infection (Khakoo et al, 2004).
Studies of HCV-infected patients have also shown that HCV may also affect an NK cell's ability to effectively cross-talk with DC which would hinder the subsequent adaptive response. NK cells may overexpress the inhibitory CD94/NKG2A receptor and produce immunosuppressive cytokines such as IL-10 and TGF-β (Jinushi et al, 2004).
1.25 HCV genotypes and prevalence of HCV in population There are six main genotypes of HCV in the world with HCV genotype 1 being the most prevalent (Simmonds, 2004; Simmonds et al, 2005). Each genotype can be further divided into subtypes and quasispecies. Genotypes of HCV differ from one another by approximately 30-35% in their nucleotide sequence and subtypes differ from each other by approximately 20-25% (Simmonds, 2004). Genotypes 1 and 4 are the most difficult to tackle with the current treatments when compared with the other genotypes (Horner and Gale, 2013). HCV subtype 1b is particularly prevalent in Europe whereas subtype 1a is more prevalent in North America. Wide distribution of Genotypes 1a, 1b and 3a has occurred in Western countries due to IDU (Intravenous drug use) and contaminated blood transfusions (Simmonds, 2004).
46 1.26 NK cells and cancer/ Immunotherapies NK cells play a vital role in the early surveillance and elimination of abnormal or tumour cells in the body. NK cells are mainly triggered by the downregulation of self-MHC Class I that occurs on many tumour cells (Ljunggren and Karre, 1985; Karre et al, 1986). Studies conducted on NK-cell deficient mice showed that NK cells played a critical role against tumour metastasis (Kim et al, 2000). NK cells have been found to kill different mouse and human tumour cell lines in vitro, as well as freshly isolated human tumour cells. The activating receptor NKG2D has been implicated in the NK cell mediated anti-tumour response as it was found that certain tumour cell-lines can upregulate NKG2D expression in- vitro (Nausch and Cerwenka, 2008).
With regard to immunotherapies using NK cells against human cancer, IL-2 has been used to treat individuals and this has shown positive correlation between the increased NK cell function and anti-tumour responses (Ljunggren and Malmberg, 2007). However, administering cytokines such as IL-2, IL-12 and IL-21 has been tested and generated mixed results depending on the type of cancer and the toxic side effects (Ljunggren and Malmberg, 2007). Another possibility of NK cell immunotherapy may be to use monoclonal antibodies or siRNA to block expression of NK cell inhibitory receptors thus aiding the NK cell activating responses. Adoptively transferred NK cells may be another future option in cancer immunotherapy with the aid of certain cytokines and growth factors. Rosenberg and colleagues attempted this first in the 1980’s and adoptive transfers of autologous ‘lymphokine activated killer’ cells were performed together with high-dose IL-2 to patients with advanced metastatic renal cancer and melanoma (Rosenberg et al, 1985).
1.27 KIR/HLA and autoimmunity/inflammatory disease There have been numerous correlations reported of activating KIR genes with development or increased risk of inflammatory or autoimmune diseases. In psoriatic arthritis there has been an association of KIR2DS1 with disease (Williams et al, 2005). Two studies also reported associations of KIR2DS2 with scleroderma, either presence of KIR2DS2 predisposes individuals to disease (Momot et al, 2004) or presence of KIR2DS2 and/or KIR2DS1 was increased in diseased individuals (Pellett et al, 2007). There have also been some associations of activating KIR genes with Type I diabetes. Type I diabetes is an autoimmune disease caused by destruction of the insulin-secreting β cells in the pancreas. Nikitina-Zake et al showed that presence of KIR2DS2 and KIR2DL2 had positive correlation with Type I diabetes (Nikitina-Zake et al, 2004). In Rheumatoid arthritis, there was increased expression of KIR2DS2 on both NK cells and T cells in patients with the disease and who
47 developed vasculitis (Yen et al, 2001). Finally, the frequency or presence of KIR2DS1 has been implicated in systemic lupus erythematosus (Hov et al, 2010; Pellett et al, 2007).
With respect to the KIR:HLA associations and autoimmune disease, certain KIR:HLA pairings have also been associated. In Crohn’s disease, KIR2DL2/3 heterozygosity and HLA-C2 homozygosity is protective whereas KIR2DL2/3 heterozygosity and presence of HLA-C1 confers susceptibility (Hollenbach et al, 2009). In Ulcerative colitis, a disease related to Crohn’s disease, one study found that KIR2DL2/KIR2DS2 are overrepresented in patients and KIR2DL3 in the presence of HLA-C1 is protective (Jones et al, 2006). Two studies show contradicting results with Type I diabetes and KIR associations. One study shows that KIR2DS2 and HLA-C1 predisposes to disease and that KIR2DL1 and HLA-C2 is protective (van der Slik et al, 2003). Another study showed that the combination of KIR2DL2 and HLA- C2 confers susceptibility and the absence of KIR2DL2 and HLA-C2 is protective. Either effect is stronger in the absence of KIR2DS1 and KIR2DS2 (Shastry et al, 2008). Multiple sclerosis (MS) is a degenerative disease that results in degradation of the myelin sheath surrounding neuronal axons by the immune system. Fusco et al found that the presence of the KIR2DS1 gene is protective in MS and the effect is stronger in the presence of HLA-C2 (Fusco et al, 2010).
1.28 NK cells (KIR/HLA) and pregnancy A subset of NK cells are particularly important during early pregnancy with regard to maintenance and tolerance to the developing embryo and placenta. In the uterus, NK cells are referred to as uterine NK cells (uNK) and have unique properties. uNK cells help with the formation of the placenta and control trophoblast invasion by the fetal extravillous trophoblast cells (EVTs) (Redman and Sargent, 2005; Moffett-King, 2002). uNK cells also ensure that there is a sufficient blood and nutrient flow from placenta to the embryo (Moffett- King, 2002). Certain pregnancy-related disorders can occur because there has been insufficient trophoblast invasion. Examples of these disorders are pre-eclampsia, recurrent miscarriage, and fetal growth restriction (Moffett-King, 2002). The KIR/HLA-C interaction is essential in regulating this whole process because the EVTs express only HLA-C and HLA-E molecules and not HLA-A/HLA-B (Hiby et al, 2004). Therefore, the fetal HLA-C molecules are some of the main HLA that will be presented to the maternal uNK cells (Apps et al, 2009). KIR expression on the uNK cells has been previously analysed and is predictably biased towards expression of KIR that interact with HLA-C (Sharkey et al, 2008). Importantly, EVT’s express both maternal and paternal HLA-C molecules so this can lead to mismatching with the maternal KIR+ NK cells (Hiby et al, 2010).
It has been shown that there is an increased risk of pre-eclampsia, recurrent miscarriage,
48 and fetal growth restriction for mothers who are homozygous for the KIR A haplotype (Hiby et al, 2004; Parham and Moffett, 2013). If the foetus is homozygous for the HLA-C2 epitope, this risk is increased, particularly if it is inherited from the father (Hiby et al, 2004, Hiby et al, 2010). The strong inhibitory receptor KIR2DL1 expressed on maternal uNK cells can engage with HLA-C2 on EVTs and this has been associated with increasing the risk of pregnancy disorders (Parham and Moffett, 2013; Trowsdale and Moffett, 2008). Expression of KIR2DS1 is associated with a protective effect and it is most likely related to the interaction of KIR2DS1 with HLA-C2. Hiby et al have suggested that the increased uNK cell activation triggered by activating KIR would aid greater trophoblast invasion (Hiby et al, 2010). More recently, another study showed that expression of the activating KIR2DS1 receptor on maternal NK cells has been associated with increased trophoblast invasion and therefore may be beneficial for placentation during pregnancy (Xiong et al, 2013). The HLA-C1 specific KIR have not been implicated in susceptibility to pregnancy disorders (Hiby et al, 2004).
1.29 NK cells (KIR/HLA) and transplantation With regard to haematopoietic cell transplants, HLA mismatching can lead to NK cell alloreactivity from the donor to the recipient of a transplant. Donor NK cells that are transplanted in the graft possess KIR specific for the donor HLA Class I molecules only and can react against the recipient cells that may or may not express the same HLA molecules (Ruggeri et al, 2002; Velardi, 2012). This alloreactivity was found to confer some benefit to the recipient by means of a graft-versus-leukemia (GVL) effect in a “missing-self” model.
49 1.30 The importance of peptides in mediating HLA: Inhibitory KIR interactions In order for KIR and HLA to successfully interact, there must be a short peptide presented in the MHC Class I peptide binding groove. The peptide binding groove of MHC class I binds peptides of 8-10 amino acid residues in length (Parham et al, 2005; Marsh et al, 2000). Peptides provide crucial stability to the newly folded MHC Class I molecules and their presence can also dictate whether a cell will be targeted for NK cell lysis (Wagtmann et al, 1995; Boyington et al, 2001). Key studies that originally determined the importance of peptides in protection against NK cell lysis were carried out by Peruzzi and colleagues in 1996. Peruzzi et al discovered that certain residues at P7 and P8 of the peptide,when presented on HLA-B27 molecules, could protect against NK cell lysis. It was also found that the presence of charged residues at P7 and P8 such as, Lysine, Arginine, Glutamine and Asparagine left the cells more vunerable to NK cell lysis (Peruzzi et al, 1996a, Peruzzi et al, 1996b). Much recent focus has been placed on the potential role that self or viral-derived peptides may have on the NK cell response when presented on HLA class I molecules. It is well established that each TCR on CD8+ T lymphocytes has a high specificity for each individual MHC:peptide complex whereas KIR receptors on NK cells exhibit a broader peptide specificity range. In particular, residues 7 and 8 of presented peptide were found to be essential for peptide-KIR interaction (Peruzzi et al, 1996; Boyington et al, 2000; Boyington et al, 2001; Parham, 2005). In comparsion, the TCR binds the central region of the MHC- peptide complex usually at positions 4 and 6 of peptide (Garcia et al, 1996). It is worth noting that both peptide dependency and peptide selectivity between HLA class I and NK cell receptors has been previously documented by numerous structural and functional investigations since the early 1990s.
50
Fig 1.4 The interaction of KIR and TCR with peptide:MHC
Key initial studies demonstrated the peptide selectivity of KIR:MHC interactions.One of the first functional studies that demonstrated the effect of peptides on inhibitory KIR:MHC interactions was using a specific self-peptide loaded onto HLA-B*2705 molecules. This peptide was found to have a protective effect on NK-cell mediated lysis via KIR3DL1: HLA- B*2705 interaction (Malnati et al, 1995). Therefore, Malnati et al demonstrated that this interaction of KIR3DL1 with HLA-B*2705 was peptide-specific. Later, a co-crystallisation study of KIR3DL1 interacting with HLA-B*5701 demonstrated that P8 of the LSSPVTKSF peptide, presented on HLA-B*5701, was found to make direct contact with residue 166 of KIR3DL1 (Vivian et al, 2011).
The peptide binding specificity of KIR2DL1 was also investigated in a functional study by Rajagopalan and Long in 1997 and they demonstrated that binding of KIR2DL1 to the HLA- C2 allotype HLA-Cw*04 occured only in the presence of specific peptides. Mutation of the residues at positions 7 and 8 of the nonamer peptide QYDDAVYKL was found to disrupt the HLA-Cw*04 interaction with KIR2DL1 (Rajagopalan and Long, 1997).
Another early functional study demonstrated that KIR2DL2+ NK cells could only recognise HLA-Cw*0304 when peptides were bound (Zappacosta et al, 1997). Zappacosta et al showed that the presence of exogenous peptides was essential to prevent NK cell-mediated lysis of cells expressing HLA-Cw*0304.The endogenous GAVDPLLAL (GAV) peptide derived from the human importin α1 subunit and its derivatives were tested in this study
51 (Zappacosta et al, 1997).
In 1999, Maenaka et al reported the crystal structure of KIR2DL3 and later in the same year they used SPR analysis to study the binding of soluble forms of KIR2DL3 and KIR2DL1 to several peptide-HLA-Cw*07 complexes (Maenaka et al, 1999a; Maenaka et al, 1999b). They found that KIR2DL3 bound to HLA-Cw*07 presenting the peptide RYRPGTVAL (Maenaka et al, 1999). This binding occurred with fast kinetics and favourable binding entropy. Some weak binding of KIR2DL1 to HLA-Cw*07:RVRPGTVAL was also reported although HLA- Cw*07 is not a reported ligand for KIR2DL1 (Maenaka et al, 1999).
A pivotal structural investigation by Boyington and colleagues undertook the co- crystallisation of KIR2DL2 and HLA-Cw*03 complexed with the GAV peptide which provided further confirmation that it was residues P7 and P8 of the GAV peptide that made direct contact with KIR2DL2 (Boyington et al, 2000; Boyington et al, 2001). Boyington et al showed that a Glutamine residue at position 71 (Gln71) of KIR2DL2 forms an essential hydrogen bond to the main-chain nitrogen atom of alanine at p8 of the presented GAV peptide (Boyington et al, 2000). The leucine residue at P7 of the GAV peptide also made important hydrophobic interactions with leucine 104 and Tyrosine 105 of KIR2DL2. It was found that this hydrophobic packing was “loose” and could accommodate other residues at P7 of GAV (Boyington et al, 2001). Interestingly, P8 of the GAV peptide mediated the most important interaction with KIR2DL2 as the hydrogen bond formed from P8 brings three residues of KIR2DL2 Gln71, Ser184 and Glu187 into such close contact with the peptide that it restricts the actual size of the peptide that can be accommodated. Therefore, Boyington et al demonstrated that as the size of the residue at P8 increases, the affinity to KIR2DL2 decreases (Boyington et al, 2000; Boyington et al, 2001).
A decade ago, in 2004, it was reported by Hansasuta and colleagues that the interaction between HLA-A3/HLA-A11 and KIR3DL2 was also peptide-dependant (Hansasuta et al, 2004). Hansasuta et al reported that the residue at P8 was key in mediating interaction with KIR3DL2. Hansasuta et al used various nonamer and decamer peptides derived from self- proteins or HIV-1, EBV (Epstein Barr virus) and Influenza virus. They also showed that other HLA alleles did not interact with KIR3DL2, such as, HLA-A2, HLA-B7, HLA-B8, HLA-B27, HLA-B58, HLA-E and HLA–G (Hansasuta et al, 2004). Some of the peptides studied were also CTL epitopes and only one out of eight CTL epitopes tested permitted KIR binding. The one CTL epitope that did permit KIR binding was derived from EBV protein EBNA3A. Substitution of the residue at p8 of the EBV EBNA3A peptide was found to abolish KIR binding (Hansasuta et al, 2004). Overall, although peptide binding has been clearly
52 demonstrated for many inhibitory KIR receptors to date its relevance to NK cell functions in- vivo still remains unclear.
1.31 Peptide antagonism of NK cell inhibition
In recent work carried out in our research group, it was demonstrated that different variants of an endogenous self-peptide can induce different NK cell responses when exogenously loaded on target cells (Fadda et al, 2010). The endogenous nonamer peptide VAPWNSLSL is derived from the TIMP1 protein (tissue inhibitor of metalloproteinase) and was naturally eluted from cells expressing the HLA-Cw*0102 allele (Barber et al, 1996; Andersen et al, 1999). Fadda et al screened derivatives of this peptide for binding of inhibitory KIR2DL2/3 to HLA-Cw*0102 on a B lymphoblastoid cell-line deficient in TAP (transporter associated with antigen processing) (Fadda et al, 2010). The panel of peptide analogs that were created contained the key anchor residues required for MHC Class I binding. However, each peptide derivative differed at residue 7 and 8 which are important for KIR binding. Overall, three key classes of peptide were reported; strong KIR-binding peptides that could induce NK cell inhibition (e.g VAPWNSFAL), those that led to weak NK cell inhibition (intermediate-binding peptides) (e.g VAPWNSRAL) and those that did not result in any NK cell inhibition (e.g VAPWNSDAL) (weak-binding peptide) (Fadda et al, 2010). Interestingly, using different ratios of peptide, Fadda et al showed that VAP-DA could significantly antagonise any low- level inhibition mediated by VAP-RA. Similarily, inhibition of NK cells by the strong-binding peptide VAP-FA was abrogated in the presence of the 'antagonistic' VAP-DA peptide. Antagonistic peptide derivatives did not affect KIR clustering at the interface between NK cells and target cells (Fadda et al, 2010). Hence, it was proposed antagonistic peptides may have a negative effect on the downstream NK inhibitory signalling pathway.
In 2012, further analysis with the VAPWNSLSL derivatives was carried out in our group by Borhis et al, whereby confocal imaging and western blotting was used to determine the effects of these peptide derivatives on KIR clustering and intracellular signalling (Borhis et al, 2012). Borhis et al focused on the KIR2DL3 receptor in particular and found that although VAP-DA does not bind KIR2DL2/3-Fc proteins it can somehow induce clustering of KIR2DL3 at an NK cell: target cell interface. VAP-DA can also enable recruitment of SHP-1 molecules to KIR2DL3 but by doing so it prevents the tight KIR clustering induced by the VAP-FA peptide, a strong binder for KIR2DL2/3. Borhis et al have proposed that somehow the DA:MHC complex is able to sequester SHP-1 molecules and thus uncouple SHP-1 recruitment from the downstream inhibitory signalling (Fadda et al, 2010; Borhis et al, 2012).
53 Borhis et al also showed that the presence of ITIMs in the intracellular domain was crucial for the antagonism with VAP-DA to occur (Borhis et al, 2012).
1.32 The role of peptides in HLA: Activating KIR interactions There has been significant work carried out on the role of peptide(s) in modulating inhibitory KIR/HLA interactions. In comparison, the number of investigations into the role of peptide(s) in activating KIR/HLA interactions has remained quite sparse. However, some recent studies have shown that the binding specificity of the activating KIR receptors to their mostly elusive HLA Class I ligands can also be influenced by the presence of peptide(s). KIR2DS1 exhibited some peptide selectivity in binding HLA-Cw*04 with some peptide derivatives tested leading to better binding affinities (Stewart et al, 2005). Stewart and colleagues also investigated KIR2DS2 tetramers binding to HLA-Cw*0304. Biacore analysis carried out by Stewart et al, showed that mutants of the GAVDPLLAL peptide at p8 disrupted KIR/HLA binding whereas mutants at p7 contributed to affinity. HLA-Cw*03 was complexed with the peptide mutants of GAVDPLLAL at p8 (GAVDPLLSL and GAVDPLLYL, GAVDPLLVL, GAVDPLLKL). The affinity of KIR2DS2 tetramers for the GAVDPLLAL peptide mutants were as follows; GAVDPLLSL > GAVDPLLYL > GAVDPLLVL > GAVDPLLAL > GAVDPLLKL (in decreasing order of affinities) (Stewart et al, 2005).
In 2014, a most recent study was published which demonstrated the interaction of KIR2DS2 with HLA-A*11:01, a rather unexpected ligand for KIR2DS2. This study also showed that variants of a vaccinia viral peptide were able to modulate KIR2DS2 interaction with HLA- A*11:01 (Liu et al, 2014). Liu et al used a vaccinia virion membrane protein A14-derived peptide MLIYSMWGK.Liu et al tested different derivatives of this peptide all varying at P8 and found that the affinity for binding to KIR2DS2 tetramers was MLIYSMWAK> MLIYSMWWK> MLIYSMWSK> MLIYSMWGK> MLIYSMWVK> MLIYSMWTK in decreasing order of affinities (Liu et al, 2014).
54 KIR$binding$ $ Pep-des$ Origin$of$ $ Deriva-ves$ HLA$Allele$ studied$to$ pep-de$ KIR$receptor$ date$ $ GAVDPLLAL>SL>VL> KIR2DL2/ Human&impor+n& $GAVDPLLAL$ KL& KIR2DS2/& HLA=Cw*0304& α1&subunit& KIR2DS1& KIR2DL1/& QYDDAVYKL$ N/A& QYDDAVYKL& HLA=Cw*0401& KIR2DS1& AAADAAAAL$ N/A& N/A& KIR2DL2& HLA=Cw*0304& Human&histone& N/A& TAMDVVYAL$ KIR2DL2& HLA=Cw*0304& H4&
Human&TIMP1& VAPWNSFAL>VAP= VAPWNSLSL$ KIR2DL2/3& HLA=Cw*0102& (endogenous)& LS>VAP=RA>VAP=DA&
MLIYSMWAK>& Vaccina&virus& WWK>&WSK>& MLIYSMWGK$ membrane& KIR2DS2& HLA=A*11:01& WGK>WVK>WTK& protein&A14& &
Table 1.3 Prominent studies of KIR interaction with peptide(s) to date. GAVDPLLAL peptide (Stewart et al 2005); QYDDAVYKL peptide (Rajagopalan and Long, 1997); AAADAAAL and TAMDVVYAL peptides (Stewart et al, 2005); VAPWNSLSL peptide (Fadda et al, 2010); MLIYSMWGK peptide (Liu et al, 2014).
1.33 Peptide Synergy and the NKG2/CD94 receptor family
The non-classical HLA Class Ib molecule HLA-E is also able to present peptides to the NKG2/CD94 family of receptors. Unlike the other HLA Class I molecules, HLA-E presents peptides derived from the leader sequences of other MHC Class I molecules (HLA-A, B, C, G) (Braud et al, 1998). HLA-E molecules usually bind peptides of 8-10 residues in length. Unlike the KIR receptors, residues P5, P6 and P8 of peptide are most important in mediating binding to CD94 (Llano et al, 1998; Petrie et al, 2008; Kaiser et al, 2008). The CD94 receptor is the part that directly interacts with the peptide (Kaiser et al, 2008). Co-crystallisation studies of the inhibitory receptor CD94/NKG2A complexed with HLA-E and VMAPRTLFL peptide demonstrated that this interaction was also peptide-dependent (Petrie et al, 2003; Kaiser et al, 2008). A significant study by Natterman et al in 2005 showed that a decamer peptide derived from HCV (core peptide YLLPRRGRPL) was able to stabilise HLA-E molecules but doesn’t inhibit NK cells via NKG2A (Natterman et al 2005). More recently, work published from our research group has shown that the peptides presented on the CD94/NKG2 family of receptors may actually work together in synergy to promote NK cell
55 inhibition via the CD94/NKG2A receptor. Cheent et al found that small amounts of leader peptides alone could weakly inhibit a small-fraction of NKG2A+ NK cells. However, this inhibition appeared to be significantly enhanced in the presence of exogenous viral peptide. This was therefore coined ‘synergistic inhibition’. Peptides from HCV, HIV and EBV were tested by Cheent et al and synergy was observed for each and NK cell inhibition was enhanced. Importantly, it was also found that the HCV core peptide (YLLPRRGPRL) could induce clustering of CD94 but not NKG2A suggesting that CD94 may potentially form homodimers (Cheent et al, 2013).
1.34 Presentation of peptide on MHC Class I of healthy cells versus virally infected cells
1.34.1 Cleavage and processing of peptides for presentation The processing of peptides for MHC Class I presentation is a highly a complex process as it involves presentation to not only KIR receptors on NK and T cells but the TCR on CD8+ CTLs. The first step of MHC Class I antigen presentation involves degradation of proteins or polypeptides by the proteasome. Proteasomes (20S) are made up of 14 subunits (seven α and seven β) arranged in four symmetrical rings (α7β7-β7α7) to form a barrel-like structure which can be opened to allow the entry of substrates for cleavage (Ferrington and Gregerson, 2012). During the early stages of a viral infection, the release of proinflammatory cytokines IFN-γ or TNF-α, induces the expression of immunoproteasomes that take over the normal proteasome activity (Aki et al, 1994). Three catalytic subunits of the proteasome are replaced by homologous subunits (β1i (LMP2), β2i (MECL1), and β5i (LMP7)) to form the immunoproteasome (Basler et al, 2013). Tanaka and colleagues first used the term ‘immunoproteasome’, referring to its increased ability to produce peptides with a proper motif for efficient MHC binding (Aki et al, 1994).
Immunoproteasomes are generally only expressed in immune cells and ultimately result in a greater variety of antigenic peptides being available for presentation on MHC Class I molecules to CD8+ T cells. Interestingly, within a week of a virus infection in the liver, almost all proteasomes are replaced by immunoproteasomes (Yewdell et al, 2005). A well- described function of immunoproteasomes is to generate peptides with a hydrophobic C terminus that can be processed to fit in the groove of MHC class I molecules (Ferrington and Gregerson, 2012; Basler et al, 2013; Seifert et al, 2010).
56 Following cleavage by the proteasome or immunoproteaome, peptides are transported to the endoplasmic reticulum (ER) where the peptides are loaded onto MHC Class I molceules and then sent to the cell surface (Purcell and Elliott, 2008). There are numerous chaperone molecules in the ER which play an important role in the peptide loading process such as Tapasin, ERP57 and TAP (Purcell and Elliot, 2008). TAP and tapasin are the main chaperones involved. Tapasin is a glycoprotein that connects peptide receptive class I heterodimers to the transporter associated with antigen processing (TAP) heterodimer (Ortmann et al, 1997; Lehner et al, 1998; Williams et al, 2002). Tapasin co-localises these complexes to TAP and this allows for the loading of peptides into the antigen binding cleft of the class I molecules (Williams et al, 2002). Tapasin may also have a stabilising role of the peptide receptive state of the class I complex (Purcell and Elliot, 2008). Interestingly CD8+ T cell responses are generated by only a small number of antigenic epitopes. This has been termed ‘immunodominance’ (Yewdell, 2006).
1.34.2 Presentation of self- and viral- epitopes during an infection The difficulties in investigating antigen presentation to T cells and NK cells lie around the exact quantification and timing of antigen that is presented (i.e how much is presented, where is it presented and for how long). We also currently lack a valuable tool which is monoclonal antibodies that can detect specific MHC:peptide complexes. A recent investigation by Croft et al nicely summarises these difficulties (Croft et al, 2013). Croft et al state that most kinetic studies of peptide binding to date have only focused on a single peptide over a range of time-points and the broader studies have been limited to one time- point.Perhaps one of the most fundamental questions still rests upon the fact that we still do not know the intimate details of how exactly viral infection or tumour transformation fully affects the MHC Class I peptide repertoire of a cell. Many previous studies have focused on collecting quantitative data and analysed the peptide repertoire of healthy cells versus virally-infected cells.
Significant work by Hickman et al in 2003, showed that there is altered expression of host- derived peptides early in infection in HIV-infected T cells (Hickman et al, 2003). Hickman et al studied the peptide repertoire that was eluted from HLA-B*0702 molecules found in uninfected T cells versus HIV-infected T cells. Hickman et al reported a difference in expression of 15 self-peptides in the HIV-infected T cells. Hickman et al observed that self- peptides were particularly prevalent in the initial days post-infection (day 1- day 3). However, after day 4 of infection the HIV viral proteins gp120 (envelope) and p24 were also detected. The unique presentation of self-peptides early in HIV infection is likely to do with early
57 protein fluctuations and an upregulation of MHC Class I molecules before viral epitopes are available for presentation. Another similar study conducted on Measles virus also reported upregulation of expression of host peptides (Herberts et al, 2003). A recent significant study carried out by Croft et al simultaneously quantified the presence of eight vaccinia virus (VACV) epitopes on infected cells at multiple time-points post-infection. Croft et al showed that the abundance of each epitope varied greatly from cell-to-cell, from as low as 10 copies per cell up to greater than 30,000 copies per cell. Importantly, the largest changes in the peptide repertoire with the eight VACV epitopes occurred early on in infection (3.5-6.5 hours post-infection) (Croft et al, 2013). Interestingly, Croft et al state that the infection had a very rapid and substantial effect on the immunopeptidome and in this case it contrasts with HIV infection in which altered self-peptides were expressed early on rather than viral epitopes. Van Els et al reported that a peptide from Measles virus (KLWESPQEI), an HLA-A*0201 epitope, may be as abundant as 5 x 104 copies per cell (Van Els et al, 2000). Herberts et al discuss the fact that certain measles virus epitopes can be as abundant as 5 x 105 copies per cell if they are from highly abundant C-protein derived peptides. Whereas, other measles virus epitopes are less abundant (200 copies per cell) if they are derived from haemagglutinin derived peptide (Herberts et al, 2001). Indeed, a lot of viral CTL epitopes have been identified and quantified in the past and most range between 1-1000 copies per cell (Crotzer et al, 2000; Levitsky et al, 1996; Lucchiari-Hartz et al, 2000; Shi et al, 1997; Tsai et al, 1996; Tsomides et al, 1994).
Therefore, an altered NK cell response may be in part caused by subtle changes in the sequence of self-peptide presented by HLA class I. This may allow certain peptides presented on virally-infected or stressed cells to tip the balance over an NK cell inhibitory threshold leading to activation. Conversely, viral-derived peptides favouring certain mutations may bind the inhibitory KIR and tip the balance in favour of NK cell inhibition.
The source of MHC Class I peptides can be quite diverse. Both DRiPs (Defective Ribosome initiation Products) and mature proteins have been implicated as important peptide sources (Dolan et al, 2002; Dolan et al, 2011; Farfan-Arribas DJ et al, 2012; Croft et al, 2013). Professor Jonathan Yewdell originally proposed that most self and viral peptides are derived from DRiPs. DRiPs are rapidly degraded and consist of prematurely terminated polypeptides and misfolded polypeptides produced from translation of mRNAs in the proper reading frame (Yewdell et al, 1996; Yewdell, 2011). DRiPs are now known to be just one of many potential sources of peptides, including other forms of defective proteins, as well as normal protein turnover (“retirees”) (Yewdell, 2001). DRiPs are now thought to include defective
58 polypeptides arising from alternative/defective mRNAs, ribosomal frame shifting, downstream initiation on bona fide mRNAs and all other errors that occur in converting genetic information into proteins (including tRNA-amino acid misacylation) (Yewdell, 2011).
It is well known that viral infection can subvert and alter host cell protein synthesis and MHC Class I peptide loading. Most viruses can promote MHC Class I downregulation on the cell surface and interfere with the recycling process of MHC Class I. Khan et al showed that viral epitopes for CTLs are from recently synthesised proteins rather than mature peptides which shows that a virus may alter the host peptide repertoire (Khan et al, 2001). Overall, we can appreciate the fact that the generation of the immunopeptidome for presentation on MHC Class I remains a highly complex process and one that requires more in-depth investigations. The relevance of the immunopeptidome with regard to viral infection and KIR is of high significance.
59 1.35 Background and Aims of the Project
The original aim of this project was to explore KIR2DL2/3+ NK cell recognition of target cells with an emphasis on how viral peptides may affect NK cell responses. As Khakoo et al previously showed a protective association of KIR2DL3 and not KIR2DL2 in acute HCV infection, we sought to investigate this further and whether this may be attributed to how KIR+ NK cells respond to the fluctuating peptide content of a cell. In terms of Hepatitis C infection, we hypothesise that certain self or viral peptides may play an important role in promoting or dampening the KIR+ NK cell response. However, we firstly wanted to address the potential differences in activation of KIR2DL3+ NK cells versus KIR2DL2+ NK cells using an already established repertoire of endogenous peptides.
Therefore, the specific aims of this project can be outlined and addressed by asking the following questions:
1) Defining an optimal peptide repertoire for NK cell activation: Can KIR2DL2+ and KIR2DL3+ NK cells respond differently to shifts in the peptide repertoire using endogenous peptides?
The first objective of this project was to further investigate mixes of the three VAPWNSLSL peptide derivatives generated by Fadda et al in 2010 (VAPWNSDAL, VAPWNSRAL and VAPWNSFAL) in order to determine optimal peptide repertoires in-vitro that would selectively define an activating threshold of KIR2DL2+ and KIR2DL3+ NK cells derived from healthy human donors. Through the use of different set ratios of each peptide, it was hoped that this would provide more insight into the fine balance that exists between NK cell activating and inhibitory responses and how they may be subtly modulated by changes in peptide. Using the three VAP peptide derivatives already created in our research group, we sought to undertake more complex mixes using these peptides and we hoped they would be representative of the peptide repertoire on a cell. Each mix consisted of different ratios of the three peptides and were presented on TAP-deficient 721.174 cells to KIR2DL2+ or KIR2DL3+ NK cells. The .174 cell-line is deficient in MHC Class II expression and expresses only the HLA-Cw*0102, HLA-A*0201 and HLA-Bw*5101 Class I alleles (Young et al, 1998). A previous study had also shown an association between HCV clearance and possession of the HLA-Cw*0102 allele (Thio et al, 2002). We therefore questioned if different mixes of these peptides, presented on HLA-Cw*0102 molecules, would differentially influence KIR2DL2+ and KIR2DL3+ NK cell reactivity. The aim of this was to obtain a more
60 quantitative foundation into how KIR2DL2+ and KIR2DL3+ NK cells respond to shifts in the peptide content of a cell.
2) Can viral (HCV) peptides mediate any effect on KIR2DL2/3+ NK cell reactivity?
Our next question was regarding how the peptide repertoire of a cell is relevant in the context of infection. We specifically chose HCV infection as there has been previous genetic evidence that there is a protective effect against HCV when individuals are homozygous for the inhibitory KIR2DL3 receptor (Khakoo et al, 2004). No functional evidence to support this has yet been shown. We hypothesise that NK cells derived from KIR2DL3+ donors may be able to respond to HCV peptides in a more efficient manner than KIR2DL2+ NK cells. We also hypothesise that some HCV peptides may be able to promote inhibition of NK cell activation in order to promote virus survival.
3) Are there any significant differences with regard to how KIR2DL2+ NK cells and KIR2DL3+ NK cells respond to viral peptides?
Can KIR2DL2+ NK cells and KIR2DL3+ NK cells respond differently to HCV peptides? The next aim was to screen viral peptides and investigate the potential role of “antagonistic” and inhibitory peptides in the context of HCV infection. HCV peptides were synthesised as part of a future effort to essentially determine whether they could mediate an effect on the NK cell response from specific donors and whether this correlates with KIR genotype. Our aim in this case was to observe if any peptides behave in an inhibitory manner for KIR2DL2 and non- inhibitory for KIR2DL3 and vice-versa. We also questioned whether ‘peptide antagonism’ of NK cell inhibition would be mediated with HCV peptides.
4) What is the role (if any) for the activating KIR2DS2 receptor in mediating HLA- Cw*0102: peptide interactions?
Towards the end of the project our aim became more focused on the role of the activating KIR2DS2 receptor which has almost 99% sequence similarity with the KIR2DL2 receptor. We questioned whether a HCV peptide could potentially act agonistically through an activating KIR receptor and whether this could be deemed as an alternative route for NK cell activation.
61 Chapter 2. Materials and Methods
2.1 Immortalised cell lines and culture
The following cell lines were used for NK cell assays: The human B lymphoblastoid cell line 721.174 (.174) expressing the HLA-Cw*0102 allele and the human K562 erythroleukemic cell line. K562 is an erythroleukemic human cell line derived from a chronic myeloid leukemia patient that lacks MHC class I, and is a highly sensitive NK cell target (Koeffler et al, 1980). The B lymphoblastoid cell line .174, does not express normal amounts of MHC class I at the cell surface due to the lack of a functional transporter associated with antigen processing (TAP) heterodimer. However, it synthesises and expresses HLA-A*0201, HLA- Cw*0102 and HLA-Bw*5101 alleles which can be stabilised at the cell surface by exogenous peptides (Andersen et al, 1999). All cell lines were cultured and maintained in R10 growth medium consisting of RPMI 1640 (Gibco® RPMI 1640, Life technologies U.K.) supplemented with 1% penicillin/streptomycin (InvitrogenTM, Life technologies U.K.) and 10% heat inactivated foetal calf serum (FCS, Lonza, U.K.) at 37°C in 5% CO2. Each cell line was passaged twice per week at equally spaced intervals.
The Ba/F3 cell lines were a gift from Professor Lewis Lanier and are a murine Interleukin-3 dependent pro-B cell line transfected with human KIR2DL2, KIR2DL3, KIR2DS2 receptors. Ba/F3 cells were cultured in R10 medium containing recombinant murine IL-3 (10 ng/mL) (Peprotech, NJ, USA). The Ba/F3 cell line transfected with KIR2DS2 required puromycin (1 µg/mL) (Life technologies, U.K) after every cell passage. All of the transfected Ba/F3 cell- lines required Geneticin G418 (1 mg/mL) (Life technologies, U.K) after every cell passage. The Huh7 hepatoma cell lines transfected with HLA-Cw*0102 and the Huh7 HCV replicon JFH1ΔE1E2-luc were kindly obtained from Dr Connie Liu in our research group and were maintained in R10 medium containing Geneticin G418 (200 µg/mL) and Blasticidin (5 µg/mL) (Life technologies, U.K). The Huh7 cell-lines were passaged every week using Trypsin- EDTA (InvitrogenTM, Life technologies, U.K).
2.2 Human Peripheral Blood Mononuclear Cell isolation and preparation
Fresh peripheral blood mononuclear cells (PBMC) were isolated from the blood of healthy screened donors using Hypaque-Ficoll (Amersham Biosciences, GE Healthcare, U.K.) density centrifugation. PBMC were frozen (90% FCS/10% DMSO) and stored in liquid nitrogen. For this investigation, donors were screened as negative for HIV, Hepatitis C and Hepatitis B. PBMC isolated in this study were dervied from haemochromatosis patients with
62 ethical committee approval. Healthy PBMC were also obtained from the NHS Bloodbank (NHS Collindale, London, U.K.).
2.3 KIR genotyping
DNA extraction for PCR typing was prepared from PBMC using the QIAamp Blood Kit according to manufacturer’s instructions (QIAGEN, U.K.). KIR genotyping was performed by sequence specific primer PCR (PCR-SSP) using the GeneAmp PCR system 9700 (Applied Biosystems®, Life Technologies, U.K.) as a thermocycler. For KIR2DL2 the primers were, sense (forward): 5’- CCA TGA TGG GGT CTC CAA A -3’ and antisense (reverse): 3’- GCC CTG CAG AGA ACC TAC A -5’, amplifying a 1800 bp fragment, and for KIR2DL3 the primers were, sense (forward) 5’- CCT TCA TCG CTG GTG CTG -3’ and antisense (reverse): 3’- CAG GAG ACA ACT TTG GAT CA -5’, amplifying a 798 bp fragment (Uhrberg et al, 1997). Amplification was performed in a 25 µL reaction, using the primers at a concentration of 0.5 µM (specific) or 0.05 µM (internal control), each dNTP (Promega, U.K.) at a concentration of 2 mM, 0.625 U AmpliTaq polymerase (Applied Biosystems®, Life Technologies, U.K.) and 200 ng DNA under the following conditions: initial denaturation, 95°C 2 min; followed by 5 cycles at 94°C for 60 s, 62°C for 60 s and 72°C for 45 s; then 30 cycles at 94°C for 60 s, 60°C for 45 s and 72°C for 45 s; and then a final extension 72°C for 5 min. Each experiment included negative and positive controls from other known genotyped donors. Amplification products were analysed on 1.5% agarose gels using ethidium bromide.
Dr James Traherne at the University of Cambridge performed further in-depth KIR genotyping using a qPCR Ct method.
2.4 HLA Class I typing
Donors were typed for HLA-B and HLA-C by Franco Tavarozzi at the Anthony Nolan Institute (London, UK) using Luminex xMAP® technology.
2.5 Peptides
Peptides identified by bioinformatic analysis were synthesised by three companies; GL Biochem Ltd (Shanghai, China), Mimotopes (U.K.) and Peptide Protein Research (Fareham, U.K.). All peptides were purified to a level of greater than 95% purity.
63 2.6 Peptide stabilisation assay
721.174 cells (2 x 105) were incubated with or without synthesised peptides for 16 hours at
26°C in 5% CO2. Cells were washed with FACs wash buffer (1% BSA and 0.1% NaN3 prepared in PBS) the following day and stained with a primary HLA-Cw*0102-specific antibody VP6G3 (a gift from Arend Mulder; Leiden, Netherlands) for 1h at room temperature. Cells were washed twice with FACs wash buffer and stained with a secondary FITC- conjugated goat anti-human IgM antibody (AbD Serotec, U.K.) for 30 min at room temperature. Cells were then washed twice in FACs wash buffer and fixed in 1% (w/v) PFA (paraformaldehyde) prepared in PBS. Cells were analysed on C6 flow cytometer (BD Biosciences, U.K.) or aBD LSR II cytometer (BD Biosciences, U.K.) using FACS Diva software (BD Biosciences, U.K) and Flowlogic software (Inivai Technologies, Australia). The mean fluorescence intensity (MFI) of at least 10,000 cells was analysed.
2.7 CD107a NK Cell Degranulation Assay
2.7.1 Target cell preparation
Target 721.174 cells (6 x 104) were incubated with single peptide or peptide mixes (10 µM final peptide concentration) for 16 hours at 26°C in 5% CO2. For the peptide mixes, target cells were pulsed with different concentrations of peptide simultaneously.
2.7.2 Effector cells (PBMC)
Human PBMC (3 x 105) were stimulated overnight with 1 ng/mL recombinant human IL-15 (R&D systems, U.K.) prepared in R10 media. PBMC were incubated overnight at 37°C in 5%
CO2..
2.8 CD107a NK Cell Degranulation Assay
After 16 hours, target cells were re-suspended with PBMC at an effector-to-target (E: T) ratio of 5:1 in fresh R10 medium containing peptide and an anti-CD107a Alexa fluor-647 antibody prepared at an optimised 1/600 dilution (see FACs antibodies table: Table 2.1). K562 target cells were also re-suspended with PBMC at an E:T ratio of 5:1 in R10 media without peptide.
Cells were incubated for 1h at 26°C in 5% CO2. A protein transport inhibitor containing monensin (Golgi-StopTM BD Biosciences, U.K.) (final concentration of Golgi-Stop 6 µg/mL) was added after 1h. Cells were incubated for a further 3h at 26°C in 5% CO2. Cells were washed in FACs wash buffer and blocked with blocking buffer (10% filtered human serum prepared in FACs wash buffer) for 30 min at 4°C. Cells were firstly stained with a Live/Dead
64 fixable Aqua dead cell stain for 30 min at 4°C(InvitrogenTM, Life Technologies, U.K.). This was followed by washing and staining the cells with antibodies for cell surface markers (seeantibodies table 2.1) for 30 min at 4°C in the dark. Cells were washed twice and fixed in 1% (w/v) PFA prepared in PBS. Cells were analysed on either a C6 flow cytometer (BD Biosciences, U.K.) or aBD LSR II cytometer (BD Biosciences, U.K.) using FACS Diva software (BD Biosciences, U.K) and FlowLogic software (Inivai Technologies, Australia).
2.9 Flow Cytometry analysis
2.9.1 Antibody panel (see Table 2.1)
The monoclonal antibodies that were used in flow cytometry are outlined in Table 2.1. All antibodies for flow cytometry were prepared in blocking buffer with the exception of VP6G3 which was prepared in RPMI 1640 (Gibco® RPMI 1640 (Life technologies U.K.). and FITC- conjugated goat anti-human IgM secondary antibody which was prepared in Dulbecco’s Modified Eagle medium (DMEM) (Gibco® DMEM (Life technologies U.K.).
2.9.2 Compensation
The BD™ CompBeads Set Anti-Mouse Ig, κ (BD Biosciences, U.K.) was used to optimise fluorescence compensation settings. Compensation beads were stained with each individual antibody for 30 min at 4°C in the dark prior to analysis. This method accurately establishes the compensation corrections for spectral overlap for any combination of fluorochrome- labelled antibodies.
65 Table 2.1 Antibodies used for Flow Cytometry experiments
Cell Surface Fluorochr Species Clone Isotype Company marker ome
Biolegend, San Mouse, CD3 Human UCHT1 PerCP Diego, CA, IgG1, κ USA Biolegend,San Mouse, CD3 Human UCHT1 Pacific Blue Diego, CA, IgG1, κ USA BD Mouse, CD56 (N-CAM) Human B159 PE Biosciences, IgG1, κ U.K. Biolegend,San Mouse, CD56 (N-CAM) Human HCD56 PE Diego, CA, IgG1, κ USA Biolegend,San Mouse, CD56 (N-CAM) Human HCD56 PE/Cy7 Diego, CA, IgG1, κ USA CD158b BD Mouse (KIR2DL2/3/S2 Human CH-L FITC Biosciences, IgG2b, κ NKAT2) U.K. Biolegend, CD158b (KIR2DL2/3/S2 Mouse Human DX27 PE San Diego, CA, NKAT2) IgG2a, κ USA
CD158b Mouse, BeckmanCoulte Human GL183 PE (KIR2DL2/3/S2NKAT2) IgG1, κ r, U.K.
Biolegend,San CD158b Mouse Human DX27 Purified Diego, CA, (KIR2DL2/3/S2NKAT2) IgG2a, κ USA
Mouse, eBioscience, CD158a (KIR2DL1/S1) Human HP-MA4 efluor-450 IgG2b U.K.
eBioH4A Mouse, Alexafluor- eBioscience, CD107a (LAMP1) Human 3 IgG1, κ 647 U.K.
Mouse, Alexafluor- Biolegend, San CD16 Human 3G8 IgG1, κ 700 Diego USA
66 2.10 Peptide binding to KIR2DL2 and KIR2DL3 receptors
2.10.1 KIR2DL2-Fc and KIR2DL3-Fc Binding Assay
KIR2D2L2-IgG and KIR2DL3-IgG fusion constructs (KIR2DL2-Fc & KIR2DL3-Fc Chimera; R&D Systems Inc, U.K.) were conjugated with protein A Alexa Fluor-488 (InvitrogenTM, Life Technologies, U.K.) at a molar ratio of 6:1. 721.174 cells (1 x 106) were incubated with 200
µM peptide for 16 hours at 26°C in 5% CO2 as described previous and then stained with KIR-Fc the following day. After fixation in 4% (w/v) PFA, cells were analysed by flow cytometry.
2.11 Production of KIR2DS2 tetramers
2.11.1 PCR for amplification of DNA encoding KIR2DS2 extracellular domain
DNA encoding the extracellular domain of KIR2DS2 was amplified by PCR using Taq DNA polymerase (New England Biolabs Inc, U.K.) and the following primers (InvitrogenTM, Life Technologies, U.K.); KIR2DS2 forward primer sequence 5’- GCAGCACCATGGGACATGAAGGAGTTCACAGAAAACCTTCCCT-3’. KIR2DS2 reverse primer sequence 5’-TACTGGATCCCGGACCGGTTTTGGAGCTTGGTTC-3’. Appropriate restriction sites were incorporated into the primer design along with an engineered Biotinylation tag at the C-terminus. A 25-cycle PCR run was used to generate the PCR products. PCR products were analysed on 1.5% agarose gels. DNA bands were visualised using ethidium bromide. PCR products were excised from the gels using a sterile blade and extracted with QIAquick gel extraction kit (QIAGEN, U.K.).
2.11.2 Cloning of KIR2DS2 extracellular domain pET23d vector DNA was kindly provided by Professor Stephan Gadola (University of Southampton, U.K.) PCR products were cloned into the pET23d vector using a ligation kit (InvitrogenTM, Life Technologies, U.K.) according to the manufacturer’s protocol. Ligation was performed for one hour at room temperature. The vector DNA was then transformed into competent Escherichia coli cells. Cells were incubated with DNA for 30 min on ice and then spread onto agarose plates containing Lucia Bertoni medium and 100 mg/mL ampicillin (Sigma-Aldrich, U.K.) (LB/Amp). Colonies were selected and grown in 5 mL of LB/AMP broth overnight at 37°C with shaking in preparation for small-scale purification of plasmid DNA. DNA was purified from the colonies using a QIAamp DNA Mini kit (QIAGEN, U.K.).
Restriction digests were performed using DNA cut with the restriction enzymes NcoI and BamHI (New England Biolabs Inc, U.K.). To confirm the presence of the insert in the vector,
67 restriction enzyme digestion using Nco I and BamHI was used to cut at sites flanking the insert. Restriction digests were performed for two hours at 37°C. The presence of an insert was confirmed after running the DNA on 1.5% agarose gels. The fidelity of the construct was confirmed by DNA sequencing (MRC Clinical Sciences centre, London).
2.11.3 Expression of KIR2DS2 proteins
KIR2DS2 DNA was transformed overnight into RosettaTM2(DE3) bacteria (MerckMillipore, U.K.).Two out of the five colonies were picked and grown overnight in 10 mL of LB/Amp (100 mg/mL). After overnight incubation, 10 mL of the bacterial culture was added to one litre of LB/AMP. Cultures were grown for a further 3-4 hours. The optical density (O.D) values were monitored every hour at 600nm on a spectrophotometer for every hour pre-induction. After three hours growth, bacterial cultures were at ideal O.D (0.3-0.6) for protein expression. IPTG (InvitrogenTM, Life Technologies, U.K.) was added to the culture (600 µL of IPTG per litre of bacterial culture). Cultures were placed in the shaking incubator at 37°C. Bacteria were induced with IPTG for 3-4 hours. After this period, the bacteria were centrifuged at 5000 rpm for 5 min per each spin to obtain bacterial pellets. The O.D values were monitored post-induction with IPTG.
2.11.4 Purification of KIR2DS2 protein from inclusion bodies
The bacterial pellets were resuspended in PBS and transferred to fresh glass beakers surrounded by ice. PMSF protease inhibitors (Abcam, U.K.) were added to the pellets before sonication. Bacteria were sonicated for 20 min in total, which consisted of 20 seconds sonication with 20 seconds rest. The beaker containing the bacterial pellets was surrounded by ice-cold water during sonication. After sonication, the bacterial pellets were centrifuged at 5000 rpm for 10 min. The supernatant was disposed of and the resulting white pellet contained the inclusion bodies. The inclusion bodies were pooled from the bacterial lysate and centrifuged at 10,000 rpm for 30 min. Inclusion bodies were then resuspended in inclusion body wash buffer (50mM Tris-base pH 7.5; 100mM NaCl; 1mM EDTA; 1mM DTT;
0.5% TritonX; 0.1% NaN3). The resuspended pellets were transferred to a glass homogeniser with clean spatula and homogenised for a further 5-10 min. The inclusion bodies were centrifuged again (10,000 rpm for 5-10 min). The washing steps were repeated twice more with the homogeniser until the final supernatant was clear (three washes in total). The washing steps were then repeated twice more except replacing the wash buffer with inclusion body resuspension buffer (50mM Tris-base pH 7.5; 100mM NaCl; 1mM EDTA;
68 1mM DTT), which removes all the detergent that was contained in the wash buffer. The inclusion bodies were finally resuspended in 10 mL of resuspension buffer. The inclusion bodies were centrifuged at 10,000 rpm for 20 min and the final pellet was solubilised in Denaturing buffer (6M Guanidine-HCL, 50 mM Tris-HCL pH 8.1, 100 mM NaCl, 10 mM EDTA) (Sigma-Aldrich, U.K.) using the glass homogeniser and separated into 10-12 mg/mL aliquots for refolding. The protein content of the inclusion body lysates was checked on a Thermo Scientific NanodropTM 1000. The inclusion body lysates were kept at 4°C where they remained stable until refolding could be performed.
2.11.5 Refolding of KIR2DS2 protein using a simple dilution method
Refolding of KIR2DS2 proteins was performed using a simple dilution method. A glass homogeniser was firstly used to solubilise the KIR2DS2 protein properly. The proteins were then left in denaturing buffer for one hour at 4°C. The proteins were then centrifuged at 10,000 rpm for 45 min (4°C) and the supernatant containing the protein was retained. Fresh DTT (20 mM) (Sigma-Aldrich, U.K.) was added to the denatured solubilised protein. The protein was incubated in a water bath for 1hr at 37°C. 10 mg of the protein was used for the refolding method. 10mg (~2 mL) of solubilised KIR2DS2 protein was added to 200 mL of refolding buffer(0.8M Urea; 3mM Reduced Glutathione; 0.3mM Oxidised Glutathione; 1 M Tris-HCL pH 8.1; 0.4M L-Arginine (Avocado Research Chemicals Ltd)) and left stirring overnight at room temperature. The following day the KIR2DS2 protein was concentrated using Vivaspin20 concentrators (Vivascience, U.K.). The KIR2DS2 protein was spun for 10- 20 min at 4,000 rpm until it was concentrated to 10 mL. The KIR2DS2 protein was then purified with gel-filtration chromatography using FPLC (Pharmacia, Sweden) with Superdex GL 10/300 gel filtration columns (GE Healthcare, U.K). After purification the KIR2DS2 protein eluate was again concentrated to 1 mL using the Vivaspin20 concentrators. The protein concentration was measured using a Thermo Scientific NanodropTM1000 and prepared for biotinylation at ~1 mg/mL.
2.11.6 Reducing SDS-polyacrylamide gel electrophoresis (SDS-PAGE)
Bacterial cultures were centrifuged at 3000 rpm for 10 min to obtain pellets. The pellets were resuspended in lysis buffer containing 5% β-mercaptoethanol and boiled for 5-10 min prior to SDS-PAGE analysis. 12% SDS-PAGE gels were prepared and 5 µg of each KIR protein was loaded on each gel and run for approx 1 hour at 80V. Gels were then stained by Coomasie blue stain.
69 2.11.7 Biotinylation of KIR2DS2 monomers
Specific biotinylation of the KIR2DS2 protein at the engineered C-terminus BirA-tag site was performed overnight in the dark using a BirA biotinylation kit (Avidity, USA). The following day excess biotin was removed from the refolded biotinylated KIR2DS2 proteins by further gel filtration chromatography using FPLC (Pharmacia, Sweden).
2.11.8 Tetramerisation of KIR2DS2
For the production of fluorescent-conjugated KIR2DS2 tetramers, biotinylated KIR2DS2 complexes were bound to PE-streptavidin overnight (Molecular Probes, USA) at a 3:1 molar ratio respectively.
2.11.9 KIR2DS2 tetramer staining
721.174 cells (2 x 105) were incubated with or without 200 µM of endogenous or HCV peptides overnight at 26°C in 5% CO2. The following day the .174 cells were washed once with FACS washing buffer, centrifuged at 1500 rpm and the supernatant discarded. .174 cells were then stained with 3 µL of KIR2DS2 tetramer which was first diluted (1/10) in sterile PBS. Cells were stained with tetramer for one hour at 4°C. Cells were then subject to two washes with FACS washing buffer and resuspended in PBS. Cells were analysed straight away post-staining using aBD LSR II Flow Cytometer (BD Biosciences U.K.) using FACS Diva software (BD Biosciences, U.K.) and for later analysis Flowlogic software was used (Inivai Technologies, Australia).
2.12 Cloning of KIR2DS2+ human NK cells
NK cells (CD3-, CD56+, CD158b+) were single cell sorted from PBMC (from a KIR2DL2/S2 homozygous donor) using the BD FACSAria II cell sorter (BD Biosciences, U.K.). Approximately 20,000 NK cells were sorted into GMP Serum-free Stem Cell Growth Medium (SCGM) media (CellGenix, Freiburg, Germany) containing 5% heat-inactivated human serum. SCGM has been used successfully for cultivation of hematopoietic stem and progenitor cells and expansion of NK cells in a previous study (Morris et al, 2005). A cocktail of cytokines was added to the starting clone media ((IL-12 (10 ng/mL) (Peprotech, NJ, USA), IL-15 (20 ng/mL) (R&D systems, Oxon, UK) and IL-18 (100 ng/mL) (R&D Systems, Oxon, UK)). Feeder cells, PBMC (1 x 106/mL) from three other allogeneic donors were irradiated and added to the NK cells along with 2.5 µg/mL PHA (Fischer Scientific, U.K.). IL-2 (250 iU/mL) (R&D systems, Oxon, UK) was added to the cells 24 hours post-sorting. Morris et al had previously described an effective NK cell cloning method using SCGM media and higher concentrations of IL-2 (Morris et al, 2005). Clones were plated according to the limiting dilution method of Cella and Colonna (Cella and Colonna, 2000) with minor modifications.
70 Briefly, human NK cells were resuspended at 3 cells per mL in SCGM media containing PHA (2.5 µg/mL) and irradiated PBMC from mixed donors. Cells were thoroughly resuspended and plated in 96-well round bottom plates at 100 µL per well, which gave an average of 0.3 NK cells per well. Six 96-well plates were prepared for the clones. After one week, an additional 100 µL of fresh clone media containing IL-2 (100 iU/mL) was added to each well. After two weeks, wells in which the media had turned yellow were resuspended and transferred to a new 96-well round-bottom plate with fresh clone SCGM media containing IL- 2 (100 iU/mL) (no PHA or irradiated feeder PBMC added). Clones were eventually transferred to 48-well plates when they were confluent and used for NK cytotoxicity assays. Clones were stimulated at the end of each week with new irradiated feeder PBMC cells and IL-2 (100 iU/mL).
2.13 KIR expression on NK clones using Reverse Transcriptase (RT) and Quantitative Real-Time (q) PCR
2.13.1 RNA isolation and cDNA preparation from NK clones
RNA was extracted from the NK clones using the TRIzol method of RNA extraction. Clones in 48-well plates were washed with ice-cold sterile PBS. 200 µL of TRIzol (TRIzol® Reagent, Ambion Inc, Life Technologies, U.K.) was added to each individual well containing the clones. Clones were pipetted up and down in the TRIzol until none remained adhered to the bottom of the wells. Clones were then transferred to RNAse-free eppendorfs and homogenised in TRIzol by vortexing for 15sec. The eppendorfs were then incubated at room temperature for 10 min. 80 µL of chloroform (Ambion, Inc, Life Technologies, U.K.) was then added to each of the eppendorf tubes. Each eppendorf was shaken by hand for 15 sec after the addition of chloroform followed by a 5 min incubation on ice. Eppendorfs were then centrifuged at 11,000rpm for 15 min at 4°C. New eppendorfs were prepared containing 200 µL of Isopropanol. The aqueous phase (upper layer) after centrifugation was carefully transferred into the new eppendorfs containing isopropanol and mixed by gentle inversion. They were then incubated for an hour at -20°C. Eppendorfs were then centrifuged at 13,000rpm for 15 min at 4°C. Supernatants were removed and the pellets were resuspended in 200 µL of 75% Ethanol. All eppendorfs were vortexed and incubated at -80°C for a further hour. After the hour incubation, the eppendorfs were centrifuged for 15 min at 11,000rpm at 4°C. The ethanol was carefully removed by pipette and any of the excess ethanol was left to evaporate before resuspending the pellets in 20 µL of DNase and Rnase free water. RNA was quantified spectrophotometrically using a Thermo Scientific NanodropTM 1000. For cDNA synthesis, 85ng of mRNA was reverse transcibed using a high capacity cDNA reverse
71 transcriptase kit (Applied Biosystems®, Life Technologies, U.K.) following the manufacturers protocol. A GeneAmp 9700 Thermocycler (Applied Biosystems®, Life Technologies, U.K.) was used to obtain the cDNA. Approximately 1 µg of cDNA was prepared from mRNA samples and used for real-time PCR analysis.
2.13.2 Reverse-Transcriptase (RT) and Quantitative Real-Time (q) PCR mRNA abundance was quantified by RTqPCR analysis on a StepOne Plus real-time PCR system (Applied Biosystems®, Life Technologies, U.K.) using Taqman fast universal PCR master mix (2X) (Applied Biosystems®, Life Technologies, U.K.) A Taqman (MGB/FAM) gene expression assay for KIR2DS2, KIR2DL2, KIR2DL3 was designed using primers and probes as described by Cooley et al (2007) (Applied Biosystems®, Life Technologies, U.K.). Eukaryotic 18S rRNA was used for an endogenous housekeeping control (reference gene) for normalisation of expression (Applied Biosystems®, Life Technologies, U.K.). Donors who were genotyped as negative for KIR2DS2 and KIR2DL2 genes were included as a control in the real time qRT-PCR of that gene. Each sample was read in triplicate on a 96-well plate under recommended cycling conditions.
2.14 Cytotoxicity assay with NK clones using Cell Tracker Orange (CTO)
4 174 cells (3 × 10 ) were incubated in R10 medium alone or R10 + peptide at 26 °C/ 5% CO2 o for 16 hours. The following day, 174 cells were incubated for 1 hour at 37 C in 5% CO2 with the cell marker CellTracker™Orange CMTMR (InvitrogenTM, Life Technologies, U.K.). They were then washed and resuspended with the NK cell clones at an E:T ratio of 10:1 in fresh R10 medium with or without peptide. .174 cells incubated in medium alone were used as controls. Cells were incubated at 26 °C for 4h, stained with a Live/Dead fixable Aqua dead cell stain (InvitrogenTM, Life Technologies, U.K.), and fixed in 1% (w/v) PFA in PBS. Analysis by Flow Cytometry allowed for the calculation of specific cytotoxicity as follows:
% Specific Cytotoxicity = % Dead .174 cells in test wells - % Dead .174 cells in control well
2.15 Conjugate formation and staining
.174 cells were incubated for 16 hours at 26°C in 5% CO2 in the absence or presence of 100 µM peptide. The following day, 1 x 105 Ba/F3 KIR2DS2 or Ba/F3 KIR2DL2 cells were co- incubated with the .174 cells at an E:T ratio of 2:1 for 10 min at 37°C in 5% CO2 . Cells were centrifuged gently and resuspended in 100 μL PBS before being fixed in 100 μL pre-warmed
4% (w/v) PFA in PBS for 30 min at 37°C in 5% CO2. Cells were centrifuged gently and resuspended in 0.01% Triton X-100 for 1 min at room temperature. Cells were stained with
72 20 μg/mL CD158b-PE (GL183) antibody (Beckmann Coulter, U.K.) for 30 min. Cells were then washed twice and resuspended in 1% (w/v) BSA in PBS. Cells were imaged using a Leica SP5 resonance scanning microscope (Leica Microsystems, U.K.). Transmission images and PE emission were collected in different channels. Data was processed using Leica Imaging software (Leica Microsystems, U.K.) and ImageJ software.
2.16 Bioinformatic analysis
The online database immune epitope database and analysis resource (IEDB) was used to search for potential HLA-Cw*0102 binding epitopes. Search input was restricted to source organism: (Hepatitis C virus genotype 1b); MHC binding; HLAClass I restricted; host organism: human. See appendix I for a list of accession numbers of each HCV isolate. Other Bioinformatic programs and tools used were the online ExPASy translate tool, ExPASy ProtParam and ClustalW2 Multiple Sequence Alignment tool which are detailed further in Results Chapter 4 and 5.
2.17 Statistical analysis
All statistical analyses were performed using GraphPad Prism version 6.0 software (GraphPad Software, San Diego, CA, USA). Linear regression analysis was performed on data using GraphPad Prism software and an ANCOVA (Analysis of Covariance) test was used to determine whether the slopes were significantly different. A p (probability) value of less than 0.05 (p<0.05) was significant. An unpaired, two tailed t test (non-parametric) was used for comparisons between two groups. A one-way or two-way ANOVA (Analysis of variance) was used for comparisons between two or more groups with the recommended multiple comparison test chosen to obtain p values. All error bars shown are SD or SEM values.
73 Chapter 3. Results
Changes in the peptide repertoire can differentially influence KIR2DL2+ and KIR2DL3+ NK cell reactivity
In 2010, Fadda et al screened 58 derivatives of the endogenous VAPWNSLSL (VAP-LS) peptide with each amino acid varying at the KIR-binding position 7 and 8. Out of this screen, only one-third of the peptide derivatives could bind KIR2DL2/3 fusion constructs and the rest gave weak or no detectable binding. Three representative peptide derivatives that gave the most differing results in KIR-Fc binding were chosen and incorporated into functional assays. From this data, we hypothesised that three peptides represent the major classes of peptide presented on MHC class I to NK cells. Each class of peptide could stabilise HLA-C1 to the same extent. The first class of peptides can engage inhibitory KIR receptors strongly and induce strong inhibition of NK cells, represented by VAPWNSFAL (VAP-FA) peptide. The second class are weak inhibitory peptides that do not bind as strongly to KIR and thus give a weaker inhibition, represented by VAP-LS and VAPWNSRAL (VAP-RA) peptides. Finally, the third class of peptide (VAPWNSDAL (VAP-DA)) give weak or undetectable binding to KIR but are at low peptide concentration able to reduce NK cell inhibition by strongly inhibitory peptides (Fig 3.1). This was coined 'peptide antagonism'. One mechanism to explain peptide antagonism was recently documented and it was found that the strong inhibitory peptide VAP-FA could induce tight micro clustering of inhibitory KIR2DL3 receptors whereas in the presence of the VAP-DA peptide such tight clustering is disrupted (Borhis et al, 2012). It was found that for this clustering to take place, inhibitory signalling via SHP-1 recruitment must occur (Borhis et al, 2012). There is also a fourth class of peptide termed 'null' peptides that do not give any detectable binding to KIR2DL2/3, do not mediate an inhibitory effect and do not antagonise NK cell inhibition. Null peptides therefore do not contribute to NK cell inhibitory signalling. However, null peptides will not be investigated as part of this project.
74
Fig 3.1 The three main classes of peptide that may be presented on MHC Class I to CD158b+ NK cells.
Due to the finding that KIR2DL3 homozygousity may provide some advantage in the resolution of Hepatitis C in a cohort of acutely infected individuals we sought to further investigate whether this is to do with how KIR2DL3+ and KIR2DL2+ NK cells respond to shifts in the peptide repertoire of a cell (Khakoo et al, 2004; Romero et al, 2008). The possession of KIR2DL2 did not contribute to any protective effect in the study by Khakoo et al. As highlighted in Chapter 1, both KIR2DL2 and KIR2DL3 receptors are almost identical in structure and both segregate as alleles of the same locus (Uhrberg et al, 2002). The differences between the extracellular domains of KIR2DL2 and KIR2DL3 are not located within the binding region that interacts with HLA-C1 and bound peptide. However, it has been shown that KIR2DL2 binds with a stronger affinity to HLA-C1 than KIR2DL3 (Moesta et al, 2008). The stronger binding affinity of KIR2DL2 to HLA-C1 is not related to differences in the ligand-binding site of the receptors but it is related to the hinge angle between the extracellular Ig-like D1 and D2 domains which may allow KIR2DL2 to engage with HLA-C1 more strongly (Moesta et al, 2008). Certain HLA Class I alleles have also been found to have association with HCV clearance. In particular, the alleles associated with clearance were HLA-A*1101, HLA-B*57 and HLA-Cw*0102 (Thio et al, 2002).
In this chapter we propose that due to the lower affinity of the inhibitory KIR2DL3 receptor to HLA-C1, KIR2DL3 may be able to sample more of the subtle shifts in peptide on the
75 repertoire of an infected cell than KIR2DL2, thus allowing an alternative system of NK activation. A previous study by Barber et al had performed peptide elution studies with cells expressing the HLA-Cw*0102 allele and found that the VAP-LS peptide derived from TIMP1 (tissue inhibitor of metalloproteinase 1) was naturally processed and bound by HLA- Cw*0102 (Barber et al, 1996). This led us to therefore also hypothesise that the majority of peptides represented on the cell surface are from the second class of peptide (VAP-RA and VAP-LS). They may engage inhibitory KIR receptors weakly and are able to maintain NK cells in a low or weak state of inhibition. By maintaining NK cells in a constant yet weaker state of inhibition it may be advantageous to the host during a viral infection or cancerous transformation when the peptide repertoire of an infected cell can subtly change and therefore the activating threshold easier breached leading to NK activation.
The difference between KIR2DL2 and KIR2DL3 receptors and how they may respond to shifts in the peptide repertoire has been previously touched upon. In functional assays both seem to produce similar results in response to each class of peptide alone and mixes containing two peptides. Therefore, the first chapter of this project involves a detailed in- vitro investigation into optimal peptide repertoires that may help to define an activating threshold of KIR2DL3+ and KIR2LD2+ NK cells derived from different donors. It is hoped that this will help to define a ‘threshold’ or ‘set-point’ for NK cell activation based on combinations of these peptides. Stabilisation assays were firstly carried out in order to investigate binding of the VAPWNSLSL peptide and its derivatives to HLA-Cw*0102, an allele associated with HCV clearance. I then investigated binding of the VAP peptides to KIR2DL2 and KIR2DL3 receptors using KIR2DL2-Fc and KIR2DL3-Fc proteins. Finally, NK cell degranulation assays were carried out using mixes representative of each of the three main classes of peptide (VAP-FA, VAP-RA and VAP-DA) to investigate how NK cells may respond to changes in the peptide content of MHC class I. With the data obtained from each group of donors we have investigated how the peptide repertoire can differentially influence activation of KIR2DL2+ and KIR2DL3+ NK cells.
76 3.1 Peptide stabilisation of HLA-Cw*0102 and HLA-E molecules with VAPWNSLSL peptide and its derivatives
The human B lymphoblastoid cell-line 721.174 (.174) are MHC Class II-deficient but express only three Class I alleles; HLA-Cw*0102, HLA-A*0201 and HLA-B*5101 (Young et al, 1998). The 721.174 cell-line are also deficient in two genes coding for TAP (transporter associated with antigen processing) and can therefore be exogenously loaded with peptide (Andersen et al, 1999). 721.174 were the primary target cells used for all assays carried out. During resting state, there is low-level expression of HLA Class I molecules on the surface of 721.174 cells and some are unstably loaded with low-affinity hydrophobic peptides (Andersen et al, 1999). When a peptide is added exogenously that is able to bind HLA- Cw*0102, the number of stable peptide-HLA-Cw*0102 complexes increases on the cell surface (Andersen et al, 1999). This results in increased stabilisation of HLA-Cw*0102. The presence of exogenous peptide leads to increased stabilisation and hence increased expression of HLA-Cw*0102. We detected this increased expression of HLA-Cw*0102 using a primary antibody specific for this allele (VP6G3) which was a gift from Dr Arend Mulder (Netherlands).
721.174 cells were pulsed with a titration of peptide concentrations (0.5 µM to 200 µM). The peptides used were; VAPWNSLSL (VAP-LS), VAPWNSFAL (VAP-FA), VAPWNSRAL (VAP- RA), VAPWNSDAL (VAP-DA). The .174 cells were pulsed for 16 hours at 26°C. Incubation time and pulsing temperature had been optimised in the lab prior to this (Fadda et al, 2010; Borhis et al, 2012). The following day, cells were stained with antibodies specific for HLA- Cw*0102 (VP6G3), and HLA-E and analysed by flow cytometry. A FITC-conjugated secondary antibody was used to detect expression of HLA-Cw*0102 molecules. As a negative control, an antibody was also used to detect HLA-E expression on cells (3D12-PE).
Results show that increasing the concentration of peptide leads to stabilisation of HLA- Cw*0102 on the cell surface (Fig 3.2C). Each VAP peptide stabilises HLA-Cw*0102 to the same extent with concentrations >10 µM approaching saturation. Even at lower concentrations of peptide (0.5 µM - 1 µM) there was stabilisation of HLA-Cw*0102 (Fig. 3.2C). With regard to HLA-E stabilisation, none of the VAP peptides were able to reach stabilisation (Fig 3.4). As a control peptide for HLA-E stabilisation, VMAPRTLFL (VMA) leader peptide was used as a negative and positive control for both HLA-Cw*0102 and HLA- E stabilisation assays respectively. VMA is a HLA-G leader peptide that can stabilise HLA-E molecules. VMA results in stabilisation of HLA-E on the cell surface (Fig. 3.4A-C). However, VMA does not stabilise HLA-Cw*0102 (Fig 3.2). Therefore, VMA was also used as a negative control peptide for VP6G3 binding to HLA-Cw*0102. Scatchard Analysis was also
77 carried out on VAPWNSLSL and each of the peptide derivatives from the stabilisation data.
Kd and Bmax values obtained showed that all of the peptides have similar affinities for HLA-
Cw*0102 (0.14 - 0.19 µM Kd) (Fig 3.3).
78 A! No peptide! VAPWNSLSL! VAPWNSFAL! VAPWNSRAL! VAPWNSDAL! VMAPRTLFL!
10587.8! 90586.7! 90362.0! 89331.1! 90094.7! 9874.5!
1986.6! 1986.6! 1986.6! 1986.6! 1986.6! 1986.6!
HLA-Cw*0102 (VP6G3)! B!
VAPWNSLSL! VAPWNSFAL! VAPWNSRAL! VAPWNSDAL! VMAPRTLFL!
107411.8! 99952.3! 102249.1! 105284.4! 63400.2!
1986.6! 1986.6! 1986.6! 1986.6! 1986.6!
HLA-Cw*0102 (VP6G3)! C!
150000 150000 VAPWNSLSL VAPWNSLSL VAPWNSRAL VAPWNSRAL 100000 VAPWNSFAL 100000 VAPWNSFAL VAPWNSDAL VAPWNSDAL VMAPRTLFL VMAPRTLFL (MFI) (MFI) 50000 50000 HLA-Cw*0102 stabilisation 0 HLA-Cw*0102 stabilisation 0 0 50 100 150 200 0.1 1 10 100 1000 Peptide concentration (µM) Peptide concentration (µM) (Log10 scale)
Fig 3.2 HLA-Cw*0102 peptide stabilisation assay with endogenous VAPWNSLSL and derivatives
.174 cells were pulsed for 16 hours at 26°C with a titration of 0.5 µM to 200 µM of each VAP peptide. Cells were then stained with a HLA-Cw*0102 specific antibody (VP6G3). Overlay plots depicting the expression of HLA-Cw*0102 in the presence (black filled-in overlays) or absence of exogenous VAPWNSLSL peptide and its derivatives (VAP-RA, VAP-FA and VAP-DA). Overlay shown with no fill is the secondary antibody only. (A) Peptide concentration of 10 µM and (B) at a peptide concentration of 200 µM. VMAPRTLFL (HLA-G leader peptide) was used as a negative control for VP6G3 binding. (C) Line graphs of MFI versus peptide concentration for each peptide. Results are from two independent experiments (n=2) and show the means ±SD.
79
Fig 3.3 Scatchard Analysis: Affinity of VAPWNSLSL peptides for HLA-Cw*0102
.174 cells were pulsed for 16 hours at 26°C with a titration of 0.5 µM to 200 µM of each VAP peptide. Cells were stained with a HLA-Cw*0102 specific antibody (VP6G3) and MFI values analysed. Scatchard Plots were made from stabilisation experiments of the HLA-Cw*0102 allele and analysis by flow cytometry with VP6G3 antibody. (A) VAPWNSLSL (B) VAPWNSFAL (C) VAPWNSRAL and (D)
VAPWNSDAL. The regression line was derived using the formula Y= Bmax*X/(Kd + X) for each peptide using GraphPad Prism6 software. Kd is the dissociation constant and Bmax is the maximal binding.
80
A! No peptide! VAPWNSLSL! VAPWNSFAL! VAPWNSRAL! VAPWNSDAL! VMAPRTLFL!
754.97 1112.3 754.9 924.3! 754.9 !! 931.0 !! 974.2 !! 754.9! 907.4 !! 754.9! 964.4 !! 754.9 !! !! !! !
HLA-E (3D12-PE)! B! VAPWNSLSL! VAPWNSFAL! VAPWNSRAL! VAPWNSDAL! VMAPRTLFL!
754.9! 955.0! 754.9 !! 932.9 !! 754.9! 964.9 !! 754.9! 912.5 !! 754.9 !! 1811.2 !
HLA-E (3D12-PE)! C!
2000 2000 VMAPRTLFL VMAPRTLFL VAPWNSFAL VAPWNSFAL 1500 1500 VAPWNSRAL VAPWNSRAL VAPWNSDAL VAPWNSDAL 1000 VAPWNSLSL 1000 (MFI) VAPWNSLSL (MFI)
500 500 HLA-E Stabilisation HLA-E Stabilisation 0 0 50 100 150 200 0 1 10 100 1000 Peptide concentration (µM) Peptide concentration ( M) Log10 scale µ
Fig 3.4 HLA-E stabilisation with endogenous VAPWNSLSL and derivatives
.174 cells were pulsed for 16 hours at 26°C with a titration of 0.5 µM to 200 µM of each VAP peptide. Cells were then stained with a HLA-E specific antibody (3D12-PE). Overlay histogram plots depicting the expression of HLA-E in the presence (black filled-in overlays) or absenceof exogenous VAPWNSLSL peptide and derivatives (VAP-RA, VAP-FA and VAP-DA). Black lines with no-fill are the HLA-E 3D12 antibody only. (A) Peptides shown on histogram plots are at a concentration of 10 µM (B) Peptides are at a concentration of 200 µM. VMAPRTLFL (HLA-G leader peptide) was used as a positive control for HLA-E binding. Expression of HLA-E was determined by an HLA-E PE-conjugated antibody. (C) Line graph of MFI versus peptide concentration for each peptide. Results are from two independent experiments (n=2) and show the means ±SD.
81 3.2 The binding specificity of KIR2DL2-Fc and KIR2DL3-Fc proteins to HLA- Cw*0102:VAP-LS peptide and derivatives
As in previous experiments 721.174 cells were incubated with each peptide for 16 hours at 26°C and the following day stained with KIR2DL2-Fc and KIR2DL3-Fc proteins. 721.174 cells were loaded with 10 µM of each VAP peptide (VAP-LS, VAP-FA, VAP-RA and VAP- DA) to ensure maximum stabilisation of HLA-Cw*0102 on the cell surface. 721.174 cells with no peptide loaded (‘no peptide’) were used as a control. Detection of KIR2DL2-Fc or KIR2DL3-Fc binding was with Protein-A conjugated to Alexafluor-488. The optimal molar ratios of KIR-Fc and Protein-A Alexafluor-488 were used for the assay (6:1; KIR-Fc: Protein- A).
As outlined in Fig 3.5 and Fig 3.6, a clear shift in KIR2DL2-Fc and KIR2DL3-Fc binding can be seen in the histograms for VAP-FA peptide (10 µM) compared with the ‘no peptide’ control. A shift can also be seen in the histograms for peptides VAP-RA (10 µM) and VAP- LS (10 µM). However, with VAP-DA there is no shift and it is similar to the ‘no peptide’ control. Results are also displayed as MFIs in bar charts (Fig 3.5B & Fig 3.6B). KIR2DL2-Fc binds significantly to .174 cells pulsed with VAP-FA peptide (p<0.001) and the same is true for KIR2DL3-Fc binding to .174 cells pulsed with VAP-FA (p<0.01) (Fig 3.5B & Fig 3.6B). There is also significant binding of KIR2DL2-Fc and KIR2DL3-Fc to .174 cells pulsed with VAP-RA (p<0.01) (p< 0.05) and VAP-LS peptides (p<0.01). KIR2DL2-Fc and KIR2DL3-Fc do not bind significantly to unpulsed .174 cells or .174 cells pulsed with VAP-DA peptide.
Overall, the MFI values obtained indicate that there is slightly stronger binding of VAP-FA peptide to KIR2DL2-Fc compared with KIR2DL3-Fc at 10 µM (Fig 3.6). However, the rest of the peptides behave in a similar manner for binding both KIR2DL3-Fc and KIR2DL2-Fc proteins with VAP-DA giving no detectable binding. Therefore, the peptide hierarchies for binding KIR2DL2-Fc and KIR2DL3-Fc are similar in that they each bind the peptides in this order of decreasing affinities VAP-FA> VAP-LS=VAP-RA>VAP-DA.
82
A No peptide! VAPWNSLSL! VAPWNSFALL! VAPWNSRAL! VAPWNSDAL!
1308.3! 1211.1! 1211.1! 9032.7! 1211.1! 14105! 1211.1! 8877.9! 1211.1! 1500.0!
KIR2DL2-Fc binding! B No peptide! VAPWNSLSL! VAPWNSFAL! VAPWNSRAL! VAPWNSDAL!
13779.7! 9236.0! 1867.7! 1652.7! 1836.7! 1652.7! 10541.9! 1652.7! 1652.7! 1652.7!
KIR2DL3-Fc binding !
Fig 3.5 KIR2DL2-Fc and KIR2DL3-Fc binding to .174 cells pulsed with VAP peptide and derivatives
Target 721.174 cells were pulsed with 10 µM of each peptide (VAP-LS, VAP-FA, VAP-RA and VAP- DA) for 16 hours at 26°C. The following day the .174 cells were stained with KIR2DL2-Fc or KIR2DL3- Fc proteins followed by Protein-A alexafluor488 (Molar ratio of 6:1 respectively). (A) KIR2DL2-Fc binding and (B) KIR2DL3-Fc binding. Histograms depicting the shift in MFIs for .174 cells pulsed with each individual peptide or .174 cells with ‘no peptide’ (Black filled-in overlays). Protein-A alexafluor488 alone is shown by no-fill overlays only in the first histogram plot. In the rest of histograms the no-fill overlay is the 'no peptide' control.
83
A KIR2DL2-Fc binding! B KIR2DL3-Fc binding!
ns ns ** * 20000 20000 ** *** ** ** 15000 15000
10000 10000
5000 5000 KIR2DL3-Fc binding (MFI) KIR2DL2-Fc binding (MFI) 0 0
No peptide No peptide VAPWNSLSLVAPWNSFALVAPWNSRALVAPWNSDAL VAPWNSLSLVAPWNSFALVAPWNSRALVAPWNSDAL
Fig 3.6 KIR2DL2-Fc and KIR2DL3-Fc binding to .174 cells pulsed with VAP peptide and derivatives
Target 721.174 cells were pulsed with 10 µM of each peptide (VAP-LS, VAP-FA, VAP-RA and VAP- DA) for 16 hours at 26°C. 174 cells were stained with KIR2DL2-Fc or KIR2DL3-Fc proteins followed by Protein-A alexafluor488 (Molar ratio of 6:1 respectively). (A) KIR2DL2-Fc binding results (B) KIR2DL3-Fc binding results. Bar charts with the displayed mean MFI values for each peptide. Each bar chart represents three independent experiments (n=3) and show means ±SD. One-way ANOVA was used with Tukeys multiple comparison test to determine significance. (**, p<0.01; ***, p<0.001).
84 3.3 Optimisation of CD107a assay with K562 and 721.174 cells
In order to optimise the CD107a degranulation assay for peptide pulsing experiments, K562 and 721.174 target cell lines were both used. K562 is an erythroleukemic human cell-line derived from a chronic myeloid leukemia patient and is deficient in MHC class I expression (Klein et al, 1976; Koeffler et al, 1980). K562 cells served as a positive control in each assay for NK cell activation as they are considered an ideal NK cell target. Previous to this, the assay has been optimised on 721.221 cells and T2 cells.
PBMC from healthy screened donors were used and KIR genotyped. CD107a or LAMP1 (Lysosomal associated membrane protein 1) is a marker of NK degranulation. LAMP1 proteins are the most abundant proteins found in the lysosomal membrane and comprise the main protein component of the lipid bilayer of NK cell cytotoxic granules that are released upon degranulation as an NK cell interacts with its target (Alter et al, 2004; Aktas et al, 2009). Thus, CD107a expression can be used as a marker of NK cell degranulation.
3.3.1 Gating strategy for Flow cytometry
Outlined in Fig 3.7 is the flow cytometry gating strategy for NK CD107a (LAMP1) degranulation assays. PBMC were firstly incubated overnight with IL-15 (1 ng/mL). The following day, PBMC were co-incubated with target cells for a period of 4 hours at 26°C and then subject to multi-parameter Flow Cytometry. Before this 4-hour co-incubation, an anti- human CD107a antibody was added followed by a protein transport inhibitor containing monensin one hour later to prevent endocytosis of any CD107a-antibody complexes. After the 4-hour co-incubation, cells were then stained for 30 min with a Live-dead staining kit followed by 30 min staining with other antibodies.
The lymphocyte population was separated from the remaining PBMC population using side- scatter and forward-scatter characteristics. Singlet cells were separated from doublets using cell width characteristics. As depicted in the gating strategy (Fig 3.7), live cells (70-80%) were those cells that did not take up Live-dead stain and this allowed dead or dying cells to be gated out. From the live lymphocyte population, CD3-, CD56+ NK cells were gated on. The CD3- CD56+ NK cells were further gated into sub-populations of NK cells, CD158b+ or CD158b- NK cells. The CD158b monoclonal antibody binds to both inhibitory KIR2DL2 and KIR2DL3 receptors. It also binds the activating KIR2DS2 receptor. Three human clones of CD158b were used throughout these assays. DX27 was the primary CD158b clone that was used until further optimisation. A monoclonal antibody CD158a, specific for the inhibitory
85 KIR2DL1 and activating KIR2DS1 receptors was also used in later assays. Within the gated CD158b+ and CD158b- sub-populations of NK cells, CD107a (LAMP1) expression was analysed. All of the donors that were analysed had between 2-15% CD3- CD56+ NK cells in their peripheral blood and this varied per individual donor within that percentage range.
Fig 3.7 Flow cytometry gating strategy for CD158b-/+ NK cells in CD107a degranulation assay
Dotplots demonstrating the simplified gating strategy used to distinguish the CD3- CD56+ CD158b+ NK cell population as part of CD107a assays. (A) Fresh PBMC were stimulated overnight with IL-15 (1 ng/mL) and separated using a lympohcyte forward scatter (FSC) and side scatter (SSC) (P1). (B) Live cells were separated by staining with a fixable livedead stain (Invitrogen) that is taken up by dead cells only. (C) The live lymphocyte population was analysed in terms of surface CD3 and CD56 expression and the CD3- CD56+ NK subpopulation was defined (R3). (D) CD3- CD56+ lymphocytes were separated and analysed in terms of CD158b (KIR2DL2/3/KIR2DS2) expression.
The level of NK cell degranulation (% CD107a expression) in response K562 and 721.174 target cells or no target cells was consistent for all donors regardless of the different KIR2DL2/3 genotypes of the cells (Fig 3.8). In every donor, the basal level of NK cell degranulation in response to IL-15 alone was ~3-4%. %CD107a expression in response to K562 cells was between 40-60% in all donors tested. In comparison, %CD107a expression in response to 721.174 cells was much lower in the range of 20-40% (Fig 3.8A&B).
86
Fig 3.8 KIR2DL2/3+ NK cell degranulation in response to K562, 721.174 target cells or no target cells
CD107a assay conducted on PBMC from three different donors incubated with IL-15 alone (1 ng/mL) or in presence of two target cell-lines K562 and 721.174 at an E:T ratio of 5:1. (A) Representative flow cytometry histograms show the percentage of CD56+ CD158b+ NK cells expressing CD107a following incubation with an anti-CD107a antibody in the presence or absence of target cells. Numbers above the black gated lines indicate the percentage of NK cells expressing CD107a within that gate. (B) Bar-chart displaying the overall percentage of CD158b+ CD107a+ NK cells in the presence or absence of target cells. Data is from three donors (n=3) and shows the means ±SD. One way ANOVA was used to compare the means of each group using Tukeys multiple comparison test (**, p<0.01; ***, p<0.001).
87 3.4 Optimisation of IL-15 concentration in CD107a assays
Interleukin 15 (IL-15) is a key cytokine that augments NK cell cytotoxicity (Bezbradica and Medzhitov, 2009). Optimisation of IL-15 concentration for the CD107a assays was also carried out. PBMC were stimulated overnight with different concentrations of IL-15 (1 ng/mL, 2 ng/mL, 5 ng/mL and 10 ng/mL). The following day, PBMC were co-incubated with 721.174 cells or no target cells. It was necessary to ensure that increasing the concentration of IL-15 would maintain the fine balance between NK cell activation and inhibition. At a higher concentration of IL-15, NK cells may become de-sensitised to small changes of peptide as part of future assays. The strong inhibitory peptide VAP-FA was used in the IL-15 optimisation experiments to observe if increasing the concentration of IL-15 would result in a decrease in NK cell inhibition observed with this peptide.
As shown in Fig 3.9B, titration of IL-15 to 10 ng/mL increased the level of CD107a expression in the presence of 721.174 target cells up to ~49%. In the presence of 721.174 cells pulsed with 10 µM VAP-FA peptide, the frequency of degranulation also increased from ~9% at 1 ng/mL IL-15 to 20% at 10 ng/mL IL-15 (Fig 3.9B). The inhibitory effect of VAP-FA was clearly observed at all concentrations of IL-15. Overall, 1 ng/mL IL-15 was chosen as the optimal concentration as it allowed for good discrimination between NK cell activation in response to 721.174 cells alone (absence of VAP-FA) and maximal NK cell inhibition (presence of 10 µM VAP-FA peptide). To serve as a control, PBMC were stimulated with either 1 ng/mL or 10 ng/mL IL-15 in the absence of target cells (Fig 3.9A). There was no difference in the frequency of CD158b+ CD107a+ NK cells at 1 ng/mL or 10 ng/mL of IL-15 in the absence of target cells.
88
Fig 3.9 Increasing the IL-15 concentration has no effect on the overall inhibition mediated by VAP-FA peptide
PBMC from one KIR2DL3 homozygous donor were incubated overnight with a titration of IL-15 concentrations ranging from 1 ng/mL to 10 ng/mL. The following day the PBMC were co-incubated with (A) No target cells (R10 media alone) or (B) Target .174 cells alone or .174 cells pulsed with 10 µM of VAP-FA peptide. CD107a expression on CD158b+ NK cells was analysed after a 4-hour co- incubation.
89 3.5 KIR2DL2/3 and HLA-C genotyping
DNA was extracted from whole blood or PBMC from 11 healthy donors (See Chapter 2: Materials and Methods) and analysed by PCR-SSP for KIR2DL2/3 genotype using the method and primers documented by Uhrberg et al (2007). Donors with confirmed KIR2DL2 homozygous or KIR2DL3 homozygous genotypes were used in all further CD107a assays. As shown in Fig 3.10, individuals homozygous for KIR2DL3 possess only one band migrating at ~798 bp in size whereas donors homozygous for KIR2DL2 possess only one band of ~1800 bp in size. Donors heterozygous for KIR2DL2/3 have both bands present. The rest of the donors used for our further assays were processed and typed using PCR- SSP by Thet Mon Myint (NHS Blood transfusion service, Collindale) and also typed by Dr James Traherne (University of Cambridge) using quantitative real-time PCR to obtain copy numbers of each KIR gene that was present in each donor. All of the genotyping carried out on each donor by PCR-SSP was confirmed by real-time PCR analysis as shown in Table 3.1. All of the donors were also HLA-C genotyped by Franco Tavarozzi at the Anthony Nolan Institute in London (Table 3.2). There was originally some difficulty obtaining KIR2DL2 homozygous donors as the allele is of low frequency in certain populations.
Fig 3.10 KIR2DL2/3 genotyping of donors by PCR-SSP
KIR2DL2/3 genotypes were determined for 11 donors using PCR-SSP and primers specific for KIR2DL2 and KIR2DL3 (Chapter 2. Materials and Methods). A DNA ladder of (100-1000 bp) (DNA HyperLadder IV (Bioline)) was used to identify the bands present and their corresponding sizes. Lanes labelled ‘Bl’ indicate Blank lanes. Lanes 1, 4, 10, 11 identify four KIR2DL3 homozygous donors. Lanes 3, 5, 6, 9 identify four KIR2DL2 homozygous donors. Lanes 2, 7, 8 identify KIR2DL2/3 heterozygotes.
90 A! Donors! KIR2DL3 KIR2DL2! KIR2DS2! (L3)! (copy no)! (copy no)! (copy no)! Donor 1! 2! 0! 0! Donor 2! 2! 0! 0! Donor 3! 2! 0! 0! Donor 4! 2! 0! 0! Donor 5! 2! 0! 0! Donor 6! 2! 0! 0! Donor 7! 2! 0! 0! Donor 8! 2! 0! 0! B! Donors! KIR2DL3 KIR2DL2! KIR2DS2! (L2)! (copy no)! (copy no)! (copy no)! Donor 1! 0! 2! 2! Donor 2! 0! 2! 2! Donor 3! 0! 3! 2! Donor 4! 0! 2! 2! Donor 5! 0! 2! 2!
Donor 6! 0! 2! 2!
Table 3.1 qPCR results (provided by Dr. James Traherne, University of Cambridge, U.K.) detailing KIR2DL2, KIR2DL3 and KIR2DS2 copy numbers for each donor group
(A) Eight KIR2DL3 homozygous donors (L3 group) (B) Six KIR2DL2 homozygous donors (L2 group).
91 A! Donors! HLA-C typing KIR typing results! (L3)! results! (Cambridge)! (" (Anthony Nolan)! L3 Donor 2! Cw*03/ Cw*07 (C1/ KIR2DL3 homozygous! ! C1)! ! ! ! L3 Donor 3! Cw*15/ Cw*03 KIR2DL3 homozygous! (C1/C1! L3 Donor 5! Cw*16/ Cw*02 (C1/ KIR2DL3 homozygous! C2) ! ! L3 Donor 6! Cw*07/ Cw*08 (C1/ KIR2DL3 homozygous! C1)! ! L3 Donor 8! Cw*06/Cw*16 (C2/ KIR2DL3 homozygous! C1)!
B! Donors HLA-C typing KIR typing results! (L2)! results! (Cambridge)! (Anthony Nolan)! L2 Donor 1! Cw*07/ Cw*07 (C1/ KIR2DL2/S2 homozygous! C1) ! L2 Donor 2! Cw*07/ Cw*07 (C1/ KIR2DL2/S2 homozygous! C1) ! L2 Donor 3! Cw*03/ Cw*07 (C1/ KIR2DL2/S2 homozygous! C1) ! !
L2 Donor 4! Cw*07/ Cw*07 (C1/ KIR2DL2/S2 homozygous! C1)! L2 Donor 5! Cw*03/Cw*07 (C1/ KIR2DL2/S2 homozygous! C1)!!
Table 3.2 HLA-C typing results (Franco Tavarozzi, Anthony Nolan Institute London) and KIR typing (James Traherne, University of Cambridge)
(A) Five KIR2DL3 homozygous donors (L3 group) (B) Five KIR2DL2 homozygous donors (L2 group).
92 3.6 Inhibition by one or two peptides is similar for NK cells from KIR2DL2 and KIR2DL3 homozygous donors
Following optimisation of the CD107a assay with 721.174 target cells, peptide titration assays were carried out with three different endogenous peptide derivatives originally screened by Fadda et al for HLA-Cw*0102 binding to KIR2DL2 or KIR2DL3. A peptide concentration of 10 µM was chosen as the optimal concentration for each assay as Fadda et al had previously demonstrated >90% stabilisation of HLA-Cw*0102 molecules on the cell surface at this concentration for each peptide. Saturation of HLA-Cw*0102 at concentrations above 10 µM of each peptide had also been documented previously (Fadda et al, 2010).
To confirm that there was no difference between the effects that each individual peptide derivative had on NK cells expressing KIR2DL2 or NK cells expressing KIR2DL3, healthy donors were selected for their KIR2DL2 or KIR2DL3 homozygousity. All future assays were based upon this group of individuals. It was essential to firstly observe the effects that each individual peptide alone, loaded exogenously on 721.174 target cells, had on the NK cell response before pursuing further peptide mixes. Previous assays had been carried out using T2 cells so it was crucial to carry out peptide titrations using the 721.174 cell line and determine if results correlated with previous findings. CD107a degranulation assays were carried out as outlined before.
721.174 target cells were incubated with different concentrations of each peptide (10 µM, 8 µM, 6 µM, 4 µM and 2 µM) for 16 hours at 26°C. On the following day, target cells pulsed with each respective peptide were co-incubated with PBMC from KIR2DL2 homozygous or KIR2DL3 homozygous donors stimulated overnight with 1 ng/mL IL-15 and an anti-human CD107a antibody. PBMC and target cells were co-incubated for 4hours at 26°C at an optimal Effector: Target cell (E:T) ratio of 5:1. Cells were stained and analysed by flow cytometry. Fig 3.11 shows representative histograms of the titrations of 721.174 target cells pulsed with each peptide and the frequency of CD158b+ NK cell degranulation in response to each. It can be observed in Fig 3.11A that the frequency of NK cell degranulation in response to 721.174 target cells pulsed with the lowest 2 µM and highest 10 µM VAP-DA concentrations was similar as for unpulsed 721.174 target cells. Expression of CD107a remained between ~30-35% at all concentrations of VAP-DA for both KIR2DL2 and KIR2DL3 homozygous donors (Fig 3.11A). Unlike VAP-DA, the other two peptides VAP-FA and VAP-RA gave significant inhibition of CD107a expression in comparison to unpulsed 721.174 cells (Fig 3.11B&C). This inhibition was weaker with VAP-RA and CD107a expression was reduced to (20-22%) (Fig 3.11B). The only significant difference observed
93 between the KIR2DL2 and KIR2DL3 donor groups was at a higher concentration of VAP-RA 8µM (**, p<0.01) and 10 µM (*, p<0.05).
The strong KIR-binding VAP-FA peptide mediated a stronger inhibitory effect on CD158b+ NK cell degranulation in both groups of donors. As shown in Fig 3.11C, the frequency of CD158b+ NK cells expressing CD107a decreased to ~10-15% at a concentration of 10 µM VAP-FA. Inhibition of NK degranulation with a low concentration of VAP-FA (~2 µM) was consistently observed with both KIR2DL2 and KIR2DL3 donors.
VAP-DA did not induce any inhibitory effect on NK cell degranulation in both donor groups. VAP-DA gave comparable levels of NK degranulation in response to unpulsed 721.174 target cells (‘no peptide’ control). Therefore, in further assays the percentage of NK cell degranulation observed in response to all peptides was normalised to the degranulation in response to unpulsed 721.174 cells (‘no peptide’ control) to minimise any experimental variation. The effect that each VAP-peptide (10 µM) alone had on CD158b+ NK cell degranulation is summarised in Fig 3.12.
94
Fig 3.11 Effects of VAP-FA, VAP-RA and VAP-DA peptides on NK degranulation in KIR2DL2 homozygous and KIR2DL3 homozygous donors
CD107a assay conducted on PBMC incubated with 721.174 cells loaded with each VAP peptide derivatives at an E:T ratio of 5:1. (A) No peptide and VAP-DA peptide (B) VAP-RA peptide and (C) VAP-FA peptide. Data shown is the mean values from three independent experiments with six KIR2DL2 homozygous donors (n=6) and eight KIR2DL3 homozygous donors (n=8). PBMC from each donor were co-incubated with 721.174 cells pulsed with a range of peptide concentrations from 2 µM to 10 µM. Line-graphs show the data from each donor group normalised to the ‘No peptide’ sample. ANOVA was used to determine significance (Bonferroni’s Multiple comparison test) (*, p<0.05), (**, p<0.01).
95 (n=4)
** * 50 ns
+ ** 40 * 30 CD107a + 20 NK cells NK
10 % CD158b 0
µ M) µ M) µ M)
174+ DA (10 174 + RA (10 174+ FA (10
174 cells (no peptide)
Fig 3.12 Frequency of degranulation of CD3- CD56+ CD158b+ NK cells in response to three VAP peptide derivatives loaded on 721.174 target cells
Graph highlighting the percentage of CD56+ CD158b+ CD107a+ NK cells post-incubation with 721.174 target cells or 721.174 target cells loaded with 10 µM of each VAP-peptide derivative. Results are mean values ± SD and are representative of four independent experiments from one KIR2DL3 homozygous donor (n=1) and one KIR2DL2 homozygous donor (n=1). **, p<0.01 174 cells alone vs 174 + VAP-FA (10 µM), **, p<0.01 174 + VAP-DA (10 µM) vs 174 + VAP-FA (10 µM), *, p<0.05 174 cells alone vs 174 + VAP-RA (10 µM), *, p<0.05 174 + VAP-RA (10 µM) vs 174 + VAP-FA (10 µM), Unpaired T Test.
96 Following the single peptide titrations, it was further investigated what effect peptide mixes consisting of different ratios of VAP-FA, VAP-DA and VAP-RA would have on NK cell activation in KIR2DL2 and KIR2DL3 homozygous donors. All subsequent assays using different ratios of peptide in each mix were prepared to an overall final peptide concentration of 10 µM and each peptide varied by 20% increments within each mix. Peptides were simultaneously loaded onto 721.174 cells and incubated for 16 hours at 26°C with the target cells. CD107a assays were carried out as outlined previous.
In both KIR2DL2 and KIR2DL3 homozygous donors the peptide mixes consisting of VAP-DA and VAP-FA resulted in increased CD107a expression above that observed for VAP-FA alone (Fig 3.13A). However, the mix containing VAP-FA (8 µM) and VAP-DA (2 µM) looked relatively similar to what was observed with VAP-FA alone (%CD107a at ~40%). Therefore, when present ata concentration higher than 2 µM, VAP-DA could antagonise the strong inhibitory effect of VAP-FA in both groups of donors. This correlateswith previous findings observed whereby Fadda et al showed that this change in inhibition is not caused by VAP- DA outcompeting VAP-FA but it is dependent on the ratio of peptide in each mix (Fadda et al, 2010). Overall, the effect of VAP-DA peptide on inhibition mediated by VAP-FA looks similar for both donor groups (Fig 3.13A).
As shown in Fig 3.13B, the intermediate KIR-binding peptide VAP-RA does not have asmuch effect on the inhibition mediated by VAP-FA until a peptide mix ratio of 6 µM VAP- RA: 4 µM VAP-FA is reached (~50:50 mix point). The frequency of NK cell degranulation gradually increases past the 50%:50% point indicating that the weaker inhibitory peptide VAP-RA is not as efficient as VAP-FA in promoting NK cell inhibition in both donor groups.
In the mix containing the weak inhibitory peptide VAP-RA and antagonistic peptide VAP-DA some antagonism was also observed but because the VAP-RA peptide yields a much weaker inhibitory effect than VAP-FA peptide the effect of the VAP-DA peptide was not as apparent (Fig 3.13C). Upon reaching the 6 µM VAP-RA: 4 µM VAP-DA mix, the maximal level of degranulation was reached. However, there was some significant difference (p<0.03) between the KIR2DL2 homozygous and KIR2DL3 homozygous donors when VAP-RA was at a high concentration (~8 µM) in this peptide mix.
97
Fig 3.13 All peptide mixes containing two peptides give similar levels of CD107a expression for both KIR2DL2 and KIR2DL3 homozygous donors
Three peptide mixes containing two peptides per mix (A) VAP-DA/VAP-RA; (B) VAP-DA/VAP-FA and (C) VAP-FA/VAP-RA mixes were used. A titration of each of these mixes was performed ensuring a final peptide concentration of 10µM. Data plotted is the mean values from three independent experiments with eight KIR2DL3 homozygous donors (n=8) (squares) and six KIR2DL2 homozygous donors (n=6) (circles). Data is normalised to a ‘No peptide’ control sample. ANOVA was used to determine significance (Bonferroni’s Multiple comparison test) (*, p<0.05).
98 3.7 Differences in reactivity of NK cells from KIR2DL2 and KIR2DL3 homozygous donors are predominantly at low levels of inhibition: Triple peptide mixes with VAP- DA, VAP-RA and VAP-FA
In order to explore the effects of changing the peptide repertoire in more detail we devised experiments using all three peptides. These experiments were conducted using fixed ratios of VAP-RA: VAP-FA and then gradually increasing the amount of VAP-DA in the mix (Table 3.3). All experiments were performed at a final peptide concentration of 10μM. 721.174 target cells were pulsed simultaneously with VAP-RA: VAP-FA combinations in the presence of a titration of VAP-DA peptide. 721.174 target cells were incubated for 16 hours with the peptide mixes at 26°C. CD107a assays were carried out as done previous and analysed by flow cytometry.
The most significant differences between the KIR2DL2 and KIR2DL3 homozygotes were observed when VAP-RA peptide was present at a high ratio to the VAP-FA peptide (80% VAP-RA: 20% VAP-FA) (Fig 3.14A). The significant differences in this peptide mix were observed at a low concentration of VAP-DA 2 µM (p<0.001) and VAP-DA 4 µM (p<0.01) (Fig 3.14A). As the concentration of VAP-DA was increased in this mix (greater than 6 µM) this difference was not as obvious between the two groups of donors.
However, in the other two mixes (40% VAP-RA: 60% VAP-FA) and (60% VAP-RA and 40% VAP-FA) the differences between the KIR2DL2 homozygotes and KIR2DL3 homozygotes were only more evident at a higher concentration of VAP-DA peptide (Fig 3.14B&C). At a ratio of (20% VAP-RA: 80% VAP-FA) all of the differences were smaller and less significant between the two donor groups (Fig 3.14D).
The concentration of peptide(s) required to reach %CD107a50 (50% of maximal NK cell degranulation) were also manually analysed for each peptide and for each peptide mix (Table 3.4A-C). As the concentrations of VAP-RA:VAP-FA were fixed in each mix for each group of donors the only variable we could measure was the concentration of VAP-DA present at 50% CD107a. From manual analysis, the only differences observed were observed in the mixes containing the three peptides (Table 3.4C). Overall, a higher concentration of VAP-DA was required to reach the 50% degranulation threshold in the majority of the mixes for the KIR2DL2 donors when compared to the KIR2DL3 donors. In the 80% VAP-RA: 20% VAP-FA mix for the KIR2DL3 donors, 0 µM of VAP-DA was present at the 50% degranulation threshold in comparison to 1.2 µM of VAP-DA in the KIR2DL2 donors. In the 60% VAP-RA: 40% VAP-FA mix, the concentration of VAP-DA was equal (0 µM) in both groups of donors. However, in the remaining two mixes 40% VAP-RA: 60% VAP-FA and 20% VAP-RA: 80% VAP-FA the concentration of VAP-DA at the 50%
99 degranulation point was again higher in the KIR2DL2 donor group (1.2 µM and 2.5 µM) (Table 3.4C).
100 DA (μM) RA (μM) FA (μM) RA:FA (%)
0 2 8 20:80
0 4 6 40:60
0 6 4 60:40
0 8 2 80:20
2 1.6 6.4 20:80
2 3.2 4.8 40:60
2 4.8 3.2 60:40
2 6.4 1.6 80:20
4 1.2 4.8 20:80
4 2.4 3.6 40:60
4 3.6 2.4 60:40
4 4.8 1.2 80:20
6 0.8 3.2 20:80
6 1.6 2.4 40:60
6 2.4 1.6 60:40
6 3.2 0.8 80:20
8 0.4 1.6 20:80
8 0.8 1.2 40:60
8 1.2 0.8 60:40
8 1.6 0.4 80:20
Table 3.3 Final concentrations of peptide in mixes containing three VAPWNSLSL derivatives (VAP-DA, VAP-RA and VAP-FA).
101
Fig 3.14 Mixes containing three peptides result in significant differences in degranulation between KIR2DL2 and KIR2DL3 homozygous donors.
Mixes containing the three-peptide variants (VAP-FA, VAP-DA and VAP-RA) were performed. Four mixes of different VAP-RA: VAP-FA were used; (A) VAP-RA 80%: VAP-FA 20% (B) VAP-RA 60%: VAP-FA 40% (C) VAP-RA 40%: VAP-FA 60% (D) VAP-RA 20%: VAP-FA 80%. Each mix containing three peptides was prepared to a final concentration of 10 µM and loaded on 721.174 cells. PBMC from six KIR2DL2 homozygotes (n=6) (circles) and eight KIR2DL3 homozygotes (n=8) (squares) were co-incubated with each of the peptide mixes and %CD107a analysed. Results were normalised to a ‘No peptide’ control sample. Results shown are mean values and are representative of three independent experiments per individual donor. ANOVA was used to determine significance (Bonferroni’s Multiple comparison test) (*, p<0.05, **, p<0.01, ***, p<0.001).
102 A! % CD107a50 values!
Peptide! KIR2DL3 donors! KIR2DL2 donors! Concentration!
of peptide at %CD107a50!
VAP-RA only! [RA] = ~2 µM! [RA] = ~2 µM!
VAP-FA only! [FA] = >10 µM! [FA] = >10 µM!
VAP-DA only! [DA] = >2 µM! [DA] = >2 µM!
B! Peptide mix! KIR2DL3 donors! KIR2DL2 donors! Concentration !
of peptide at %CD107a50!
FA + DA mix! [FA] = 8 µM [DA] = 2 µM! [FA] = 8 µM [DA] = 2 µM!
FA + RA mix! [FA] = 8 µM [RA] = 2 µM! [FA] = 8 µM [RA] = 2 µM!
DA + RA mix! [DA] = 2 µM [RA] = 8 µM! [DA] = 2 µM [RA] = 8 µM!
C!
Peptide mix! KIR2DL3 donors! KIR2DL2 donors! ! ! Concentration of peptide at %CD107a50! ! ! 80% RA/ 20% FA: [RA] = 6.4 µM [FA] = 1.6 µM [RA] = 6.4 µM [FA] = 1.6 DA mix! [DA] = <0 µM! µM [DA] = ~1 µM! ! ! ! 60% RA/ 40% FA: [RA] = 4.8 µM [FA] = 3.2 µM [RA] = 4.8 µM [FA] = 3.2 µM DA mix! [DA]= <0 µM! [DA] = 0 µM! ! ! [RA] = 3.2 µM [FA] = 4.8 µM [RA] = 3.2 µM [FA] = 4.8 µM 40% RA/ 60% FA: [DA] = 0.7 µM! [DA] = 1.2 µM! DA mix! ! ! [RA] = 1.6 µM [FA] = 6.4 µM [RA] = 1.6 µM [FA] = 6.4 µM 20% RA/ 80% FA : [DA] = 1.2 µM! [DA] = 2.5 µM! DA mix! ! !
Table 3.4 Concentrations of VAPWNSLSL peptide derivatives alone or in each mix at
%CD107a50 for KIR2DL3 and KIR2DL2 donors. (Displays each peptide concentration at 50% of maximum %CD107a expression). All data is normalised to ‘no peptide’ which is set at 100% CD107a degranulation.
103 3.8 Modelling the data from the triple peptide experiments from KIR2DL2 homozygous and KIR2DL3 homozygous donors (Model constructed by Professor. Jayajit Das and Dr. Sayak Mukherjee)
In order to understand the data using a more mathematical approach we sent it to Professor Jayajit Das (Battelle Center for Mathematical Medicine, The Research Institute at the Nationwide Children’s Hopsital, Columbus, Ohio, US). He and Dr Sayak Mukherjee modelled the data using regression modelling on MATLAB software. They derived an equation from the raw data that I sent (shown below as Equation 1). In simple terms, the equation they derived shows that the level of degranulation (A) changes as you vary the concentration of each peptide in each peptide mix.
Equation 1:
A= 100– kF[FA] – kD[DA] – kR[RA] + kDF[DA][FA] + kFR[FA][RA] + kRD[RA][DA]+ kDFR[DA] [FA] [RA]
In equation 1 above, [x] indicates the concentration of peptide. The k values are model parameters that have been evaluated by fitting the above equation to the data charactersing NK cell activation stimulated by mixes of the three peptides. NK cell activation is assumed to be 100% in the absence of any of the three peptides.
Overall, this produced the results shown below:
A2DL2 = 100–6.4[FA]– 5.27[RA] + 0.003[DA][FA] + 0.018[FA][RA] + 0.125[RA][DA]
A2DL3 = 100 - 6.05[FA] – 2.36[RA] + 0.0007[DA][FA] + 0.0016[FA][RA] + 0.54[RA][DA]
(Both kD and kDFR[DA] [FA] [RA] being are set to zero for both KIR2DL2 and KIR2DL3)
Overall, the model created by Professor Das and Dr Mukherjee (Fig 3.15) strongly reinforces what we have already observed with the primary data. We observed that KIR2DL3 donors seem to have a lower set-point for activation than the KIR2DL2 donors and this seems to be most dependent on the concentration of VAP-RA present in the peptide mixes. The result was that Equation 1 derived by Professor Das andDrMukherjee agrees marginally better with the data from the KIR2DL3 donors. The x2 values obtained from the equation quantify the difference between the model and the experimental data. The smaller x2 value of 8 for KIR2DL3 donors shown in Fig 3.15A correlates with how well the model fits the data (goodness of fit measurement). An x2 of 8 for the KIR2DL3 donors would indicate a better goodness of fit of the data compared with the larger x2 value of 12.6 for the KIR2DL2 donors. We already proposed that the [VAP-RA] in the triple mix and also in the mix of [VAP- RA]/[VAP-DA] are the most important peptide/peptide combinations in influencing the NK cell
104 activation. This was reinforced by the kR and kRD values obtained by Professor Das and Dr Mukherjee for the KIR2DL2 and KIR2DL3 donors which are quite different when compared to the k values for the other peptides. As shown above in underlined text, the kR value for the
KIR2DL2 donors was 5.27 and the kR value for the KIR2DL3 donors was 2.36. The kRD value for the KIR2DL2 donors was 0.125 and the kRD value for the KIR2DL3 donors was 0.54.
In Fig 3.15B Professor Das and Dr Mukherjee also show that more peptide combinations giving greater than 65% CD107a degranulation (red) are found for KIR2DL3+ NK cells compared with KIR2DL2+ cells. Green indicates the number of peptide combinations giving less than 65% CD107a degranulation and this is observed in a greater number for the KIR2DL2+ cells.
105
Fig 3.15 Analysis of CD107a degranulation of CD158b+ NK cells from the KIR2DL2 and KIR2DL3 homozygous donors as determined by equation (1) (Courtesy of Dr Sayak Mukherjee and Professor Jayajit Das)
(A) Comparison between the model and the experiments. The activation determined by the model (gray surface) in Eq. (1). The left panel shows the comparison for KIR2DL2+ donors and the right panel for KIR2DL3+ donors. (B) NK cells from the KIR2DL3 donors show a lower “set point” for activation. The peptide combinations determined by Eq. 1 that produce greater than 65% degranulation as compared to no peptide for KIR2DL2 donors (left panel) and KIR2DL3 donors (right panel). Degranulation of more than 65% is shown in red, and less than 65% is shown in green. Each ‘dot’ on Fig 3.15A shows the individual experimental data-points.
106 3.9 Discussion
The aim of this chapter was to investigate endogenous peptide repertoires that would be representative of MHC:peptide complexes presented on a target cell surface to KIR2DL2+ or KIR2DL3+ NK cells. Furthermore, we questioned what effect changes in these specific peptide repertoires would have on KIR2DL2+ and KIR2DL3+ NK cells.
VAPWNSLSL is an endogenous peptide that was originally eluted from 721.221 cells expressing HLA-Cw*0102 (Barber et al, 1996). We tested variants of this peptide that had residue differences at position 7 and 8 to explore the effects that they would have on NK cell reactivity via KIR2DL2 and KIR2DL3 inhibitory receptors. Results demonstrated that the variants of the VAPWNSLSL peptide used singly had similar effects on the NK cell response in both KIR2DL2 homozygous and KIR2DL3 homozygous donor groups, which correlated with those previously reported by Fadda et al in 2010. The weak-affinity peptide VAPWNSDAL (VAP-DA) did not induce any KIR2DL2/3-mediated inhibition when presented on HLA-Cw*0102 to CD158b+ NK cells. Therefore, the effects that VAP-DA had when loaded on target cells were comparable to the response of CD158b+ NK cells to the target cells without peptide. It was demonstrated that the high-affinity KIR-binding peptide VAP-FA induced a much lower level of CD158b+ NK cell degranulation when loaded on target cells, which resulted in NK inhibition in both donor groups. In comparison, the intermediate affinity KIR-binding peptide VAP-RA exhibited a much lower inhibitory effect than VAP-FA. Therefore, a higher concentration of VAP-RA peptide was required to obtain the same level of inhibition as observed with VAP-FA in both KIR2DL2 and KIR2DL3 donor groups.
Peptide mixes allowed for further discrimination between the effects that each VAP peptide derivative had on NK cell reactivity. A significant change in the level of NK cell degranulation was observed with a peptide mix containing both VAP-FA and VAP-DA. As the peptide repertoire shifted within this mix to a small extent a distinct increase in NK cell degranulation was observed. Therefore, a very small change in the repertoire in favour of a peptide that has weak binding affinity (VAP-DA) versus a peptide that elicits strong binding affinity (VAP- FA) was able to antagonise the inhibition mediated by VAP-FA, leading to substantial CD158b+ NK cell degranulation. At a low concentration, VAP-DA was found to be an effective antagonist of NK cell inhibition mediated by VAP-FA in both KIR2DL3 homozygous and KIR2DL2 homozygous donor groups.
The precise molecular mechanism to explain peptide antagonism is unknown. Borhis et al previously observed that VAP-DA can induce diffuse clustering of KIR2DL3 receptors in confocal experiments. However, this differs from the KIR-Fc binding data reported here where VAP-DA does not mediate any detectable binding to either KIR2DL3-Fc or KIR2DL2-
107 Fc. This demonstrates that the KIR: MHC-peptide affinity in solution versus in-situ is quite different and KIR-Fc proteins may not be the most reliable tools to use for direct affinity measurements. Borhis et al proposed that the VAP-DA peptide is able to somehow disrupt the early inhibitory signal needed for tight KIR micro clustering to take place. Disruption of this inhibitory signal by VAP-DA appears to be ITIM-dependent and may involve recruitment of SHP-1 molecules to VAP-DA:MHC complexes (Borhis et al, 2012). Borhis et al proposed that VAP-DA:MHC complexes may sequester the SHP-1 molecules needed to transduce an inhibitory signal. Alternatively, these antagonistic peptide:MHC complexes may deliver another signal that counteracts the traditional ITIM-based inhibitory signal.
Within a mix containing the strong inhibitory peptide VAP-FA and weak inhibitory peptide VAP-RA, the level of NK degranulation only progressively rose when there was a much larger shift in the peptide repertoire. It must be highlighted again that the antagonistic effect of VAP-DA on NK cell inhibition mediated by VAP-FA was distinct from the weaker inhibitory effect of VAP-RA. The only significant difference between the two donor groups was observed in a peptide mix containing the weak inhibitory peptide VAP-RA and antagonist peptide VAP-DA. At a higher concentration of VAP-RA (~8 µM) there was increased degranulation in the KIR2DL3 donor group, which was significantly different to that observed in the KIR2DL2 group.
Degranulation assays incorporating the three peptides were undertaken to represent a cell surface repertoire of the different classes of peptide that can influence KIR+ NK cell functionality. As part of these complex degranulation assays, it was observed that in the KIR2DL3 homozygous donor group, NK degranulation only significantly changed when the proportion of VAP-RA was highest in the mix relative to the proportion of VAP-FA present. Essentially, the subtle differences we observed between KIR2DL2 homozygous donors and KIR2DL3 homozygous donors were only evident at high concentrations of the weak inhibitory peptide VAP-RA. This also correlated with the results we observed in the peptide mix containing VAP-RA and VAP-DA only. Again, the differences were only apparent at high VAP-RA concentrations and also after a small proportion (~2 µM) of the antagonist peptide VAP-DA was titrated into the peptide mix. When a concentration greater than 4-5 µM of VAP-DA was titrated into the VAP-RA/VAP-FA mixes the subtle difference observed between donor groups was not as apparent from this point. This is best highlighted when we analysed the concentrations of each peptide at %CD107a50 (50% of maximal NK cell degranulation). It was observed that in three out of the four mixes containing the three peptides, the concentration of VAP-DA needed to reach 50% of max NK degranulation was consistently lower in the KIR2DL3 donor group when compared to the KIR2DL2 donors. This
108 would suggest that less 'antagonistic' VAP-DA peptide was required in the KIR2DL3 donor group to reach maximal NK cell degranulation. Whereas, the KIR2DL2 donor group required a higher concentration of VAP-DA to surpass this 50% degranulation threshold. This data indicates to us that KIR2DL3+ NK cells may therefore respond more rapidly to small shifts in the peptide repertoire compared with KIR2DL2+ NK cells.
It must be noted that VAP-RA is a weakly inhibitory peptide and its effects are distinct from the peptide antagonism previously reported with the VAP-DA peptide (Fadda et al, 2010). VAP-RA may not induce as strong an inhibitory signal nor may it induce the tight clustering of inhibitory KIR as observed with the VAP-FA peptide (Borhis et al, 2012). The effects of the VAP-RA peptide on inhibitory KIR signalling and KIR clustering are unknown at present. We hypothesise that the VAP-RA peptide may also affect the clustering of inhibitory KIR and that this differs slightly between KIR2DL2 and KIR2DL3 receptors. KIR2DL2 and KIR2DL3 are alleles of the same gene and differ only by four amino acids in their extracellular hinge domain. These residue differences are not located in the region that binds to MHC Class I and peptide. Thus, the polymorphic differences between KIR2DL2 and KIR2DL3 may not affect MHC:peptide binding directly but may indirectly affect KIR clustering or KIR expression on the cell surface. As both the VAP-DA and VAP-FA peptides can mediate an effect on KIR clustering, we can hypothesise that the VAP-RA peptide does also.
Results obtained from both the primary data and later the regression model depict a more gradual increase in degranulation for KIR2DL2 homozygous donors at high VAP-RA concentrations whereas a more abrupt increase in degranulation occurs for KIR2DL3 homozygous donors. Our model also suggests that the role of the VAP-RA/VAP-DA pair is more important than the other two pairings as it has a greater effect on KIR2DL3 homozygous donors than the KIR2DL2 homozygous donors. We observe this when we compare the KRD values, whereby KDA equalling zero, indicates minimal effect on NK cell activation. The model also agrees better with the data from the KIR2DL3 homozygous donors rather than the KIR2DL2 homozygous donors. Using this model we can show that NK cells from the KIR2DL3 donors show a lower “set point” or “threshold” for activation.
There are numerous factors coming into play with regard how to quantify and explain the differences observed between the two groups of donors. However, the differences seem to be only important at a high concentration of the weak inhibitory peptide VAP-RA. Thus, the only differences between the donor groups were always observed at lower levels of inhibition even in the experiments with one or two peptides only. The model derived in Equation (1) fitted slightly better with the data for KIR2DL3 compared to KIR2DL2, which could indicate
109 the presence of additional factors (e.g., other activating and inhibitory NK receptors) not included in the model that may influence activation for KIR2DL2 donors.
Given these results, we propose that KIR2DL3+ NK cells may be more sensitive to fluctuations in the surface peptide repertoire on a target cell. Inhibition of KIR2DL3+ NK cells may be more easily ‘antagonised’ by these small fluctuations than KIR2DL2+ NK cells. KIR2DL2 homozygous donors also possess a plethora of other inhibitory and activating KIR receptors that could influence their activation. In particular, the possession of more activating KIR receptors is associated with a KIR Haplotype B whereas KIR2DL3 homozygous donors possess less activating KIR that is associated with KIR Haplotype A (Wilson et al, 2000; Khakoo and Carrington, 2006). It is known that the activating KIR2DS2 gene is in strong linkage disequilibrium with the KIR2DL2 gene (Moesta et al, 2008; Moesta et al 2010). The monoclonal CD158b antibody that was used in this study is able to bind KIR2DL2, KIR2DL3 and KIR2DS2. Therefore, the effect of KIR2DS2 in this study would have to be considered when analysing the results that were gated on a CD158b+ NK population. Given the differences we observed between the KIR2DL2 homozygous donors and KIR2DL3 homozygous donors we propose that any activating effect of KIR2DS2 was likely to be very small with the endogenous VAP peptides.
Previous studies have shown that KIR2DL2 is a stronger inhibitory receptor than KIR2DL3 and can therefore engage its HLA-C1 ligand strongly (Winter et al, 1998; Moesta et al, 2008). The importance of KIR2DL3/HLA-C1 homozygousity has been highlighted as it shows correlation with spontaneous resolution in the acute stage of Hepatitis C infection (Khakoo et al, 2004). The explanation behind this protection conferred by KIR2DL3 homozygousity was that as KIR2DL3 doesn’t engage or bind HLA-C1 as strongly as KIR2DL2 it may be more easily overcome by other stronger activating signals. Defining these activating signals remains a challenge to date and they may come from other activating NK receptors. We therefore now propose that another simple alternative route to NK cell activation may be determined by shifts in the peptide repertoire on a target cell. If KIR2DL3 can act as a more flexible sensor to detect these subtle shifts that may be triggered by viral infection then it could have the upper hand over KIR2DL2. KIR2DL3 homozygousity has also been found important for fulminant Malaria (Hirayasu et al, 2012). Interestingly in HIV-infected individuals who possess the stronger inhibitory receptor KIR2DL2, it appears that HIV- positive CD4+ T cells can activate KIR2DL2+ NK cells. It has been proposed that altered or viral HIV peptides can manipulate the stronger HLA-C:KIR2DL2 interaction and promote NK cell inhibition instead of NK activation (Alter et al, 2011).
110 Fluctuations in the peptide repertoire of a cell are commonly triggered by a viral infection or tumour transformation. Thus, if changes in the peptide repertoire of a cell can differentially activate KIR2DL2+ or KIR2LD3+ NK cells as shown here with the VAPWNSLSL derivatives, this provides another rationale for how abnormal or infected cells may be recognised by different KIR receptors. The peptides that are ultimately presented on MHC Class I molecules in an infected cell come from complex processing and may be altered self- peptides or viral peptides. These peptides may originate from the degradation of mature proteins or alternatively DRiPS (defective ribosomal initiation products) (Yewdell et al, 1996; Yewdell, 2002; Yewdell and Hill, 2002). Quantitation of the exact ratio of viral peptide to endogenous peptide that may be processed and presented on MHC Class I has remained difficult. However, previous significant work by Van Els and Herberts et al found that a HLA- B*2705-restricted peptide derived from Measles virus had abundant expression in cells infected with the virus. Its level of expression in infected cells was up to ~1500 copies (MHC- peptide complexes) per cell (Herberts et al, 2001; Van Els et al, 2000). Other significant work by Hickman et al in 2003 demonstrated that in HIV-infected cells there is a larger presence of altered self-peptides compared with what was found in normal cells. Therefore the plethora of peptides that may be presented by MHC Class I to our NK cells may be of viral or altered self-origin. Another study by Hickman et al in 2004 sampled the peptide repertoire of the HLA-B*1801 allele expressed on the MHC Class I-deficient 721.221 B cell line. Over 200 endogenously loaded HLA-B*1801 peptides were found and were considerably varied. Most were from intracellular proteins and others originated from different proteins such as binding proteins, catalytic proteins, and proteins involved in cell metabolism and growth, (Hickman et al, 2004). Overall, it was found that the range of peptides presented by HLA-B*1801 was relatively unbiased although there was a slight preference found for peptides derived from RNA- and nucleic acid-binding proteins (Hickman et al, 2004).
The relevance of this comes into play with our findings, which shows that these subtle shifts of peptide can differentially activate different subgroups of KIR+ NK cells. Thus, by merely changing the peptide content of MHC Class I the reactivity of KIR2DL2+ and KIR2DL3+ NK cells can be altered significantly. We have also shown that the ‘threshold’ or ‘set-point’ of activation for KIR2DL3+ NK cells is lower than that of KIR2DL2+ NK cells, which may imply that they are easier activated by shifts in the peptide repertoire. Future work would have to determine whether this is the case for viral peptides and potentially use MHC-Class I- restricted viral peptides or altered self-peptides that have already been eluted in binding studies. The use of altered self-peptides that may occur naturally during tumour transformation would also be ideal to determine if they can affect NK cell reactivity via KIR2DL2 and KIR2DL3. A peptide of interest could potentially be derived from heat-shock
111 proteins (Hsp) or other stress proteins that are upregulated during times of cellular stress. As mentioned previous, the panel of over 200 endogenous peptides sequenced by Hickman et al in 2004 featured many peptides derived from heat shock proteins Hsp27 (27 kDa), Hsp60 (60 kDa) and Hsp90 (90 kDa).
112 Chapter 4. Results
The KIR2DL2/3+ NK cell response to peptides derived from Hepatitis C virus
The second part of this project has focused on the identification of peptides derived from Hepatitis C virus (HCV) to determine what effects they may have on the KIR2DL2+ and KIR2DL3+ NK cell response when presented on HLA-C1 molecules. The original goal was that the isolation of such peptides could eventually lead to their incorporation in NK functional assays using a defined optimal peptide repertoire.
As outlined in the previous chapter, peptide antagonism of the inhibitory KIR2DL2/3 receptors was explored with derivatives of an endogenous peptide VAPWNSLSL in KIR2DL2+ and KIR2DL3+ donors but this has not been tested using viral peptides and specifically peptides derived from HCV. HCV was deemed a good candidate for investigation due to the possibility that possessing KIR2DL3 may confer an advantage in eliminating the virus in the early acute stage of infection (Khakoo et al, 2004).
Some past studies have investigated presentation of viral peptides on MHC Class I and the corresponding NK cell reactivity. Mandelboim et al investigated peptides derived from cocksackie virus which has been associated with autoimmune diabetes mellitus. This study found that certain peptides from cocksackie virus could abrogate the inhibitory recognition via HLA-Cw*07 (Mandelboim et al, 1997). Zappacosta et al also investigated the ability of a panel of peptides eluted from HLA-Cw*0304 to protect target cells from NK cell-mediated lysis and they observed different NK cell responses with each of the presented peptides (Zappacosta et al, 1997). A significant study by Hickman et al reported that during HIV infection, altered self-peptides are presented uniquely on HIV-infected cells that are not normally presented on uninfected cells (Hickman et al, 2003). The presentation of “altered” or “modified” self-peptides on MHC class I may trigger NK cell lysis and this would be of importance during a cellular stress response to viral infection or tumour transformation. Although these altered peptides that were presented on infected cells early on in infection were not from HIV, the study by Hickman et al is particularly relevant in showing that presentation of altered self-peptides could accompany that of viral peptides in the peptide repertoire of an infected cell.
More recently, similar work published by Fadda et al investigated the effects that peptide variants of p24 GAG from HIV-1 had on the NK cell response via the KIR2DL2 receptor (Fadda et al, 2012). The peptide variants of p24 Gag that were investigated were HLA- Cw*0102 restricted. Out of this large screen of 217 HIV peptides, only 11 were found that
113 could stabilise HLA-Cw*0102 (~5% of screened peptides) and only one allowed binding of KIR2DL2 (Fadda et al, 2012).
Given our previous work, in this chapter we aimed to generate a small screen of peptides derived from Hepatitis C virus using database searching and bioinformatic analysis. We obtained HCV isolate sequences from a cohort of 31 individuals that had been infected with the same HCV subtype (subtype 1b). Using algorithms and manual analysis, we predicted which HCV epitopes from this cohort were most likely to bind HLA-Cw*0102, an allele previously associated with HCV clearance (Barber et al, 1996; Thio et al, 2002). Following this we then tested whether the HCV peptides could stabilise HLA-Cw*0102. We questioned whether these peptides would have an effect on NK cell inhibition in KIR2DL2 homozygous or KIR2DL3 homozygous donors. Using KIR2DL2-Fc and KIR2DL3-Fc proteins we investigated binding of HCV peptides to these receptors. Following this, the HCV peptides were incorporated into functional NK degranulation assays with KIR2DL2 homozygous, KIR2DL3 homozygous donors and KIR2DL2/3 heterozygous donors.
114 4.1 Identification and synthesis of HCV peptides
I firstly set out to try identify potential HCV viral HLA-Cw*0102 binding epitopes in a selected cohort of patients through the use of database searching. In the late 1970s over 500 women in Ireland were infected with the same isolate of Hepatitis C virus (subtype 1b) upon receiving single source contaminated anti-Immunoglobulin D (Itakura et al, 2005). Dr Liam Fanning (University College Cork, Ireland) kindly provided us with the nucleotide sequences of 31 HCV subtype 1b isolates (including the primary source HCV isolate) from this cohort. The sequences were firstly translated using the ExPASy proteomics translate tool (Gasteiger et al, 2003; ExPASy, 2014). The translated, full-length sequence of the primary isolate can be found in Appendix I. Following the translation of all 31 sequences, a ClustalW2 multiple sequence alignment (ClustalW2, EMBL-EBI, 2014) was performed (Fig 4.1). Each HCV isolate is approximately 9000 nucleotides (~3100 amino acid residues) in length and displayed high conservation overall.
Firstly, in order to pinpoint potential HLA-Cw*0102 binding epitopes within the HCV subtype 1b genome, a number of online databases were used. Limited information was available on both Syfpeithi and BIMAS databases for HLA-Cw*0102 binding epitopes for HCV. Both Syfpeithi and BIMAS are databases that provide HLA peptide binding predictions (Hans- Georg Rammenseeet al, 1999; Parker et al, 1994). An online database termed immune epitope database (IEDB) was also used to search for potential HLA-Cw*0102 binding epitopes within the HCV subtype 1b genome (IEDB; Vita et al, 2009).
The peptide binding specificity of the HLA-Cw*0102 allotype is illustrated in Fig 4.2 (Marsh et al, 2000). As shown in Fig 4.2, HLA-Cw*0102 preferably binds nonamer peptides and the dominant anchor residues for binding are a proline residue (P) at position 3 (P3) and a leucine residue (L) at position 9 (P9) (motif xxPxxxxxL). Other residues are also found enriched at certain positions, such as alanine or leucine at position 2 (P2). Using these key anchor residues as a guide we manually selected for potential HCV peptides that could bind HLA-Cw*0102.
Potential HLA-Cw*0102 binding epitopes within the HCV subtype 1b primary sequence were also screened by searching for these dominant anchor residues in the search output obtained from IEDB.
115
Fig 4.1 Primary sequence analysis between HCV subtype 1b isolate sequences obtained from Irish cohort
Multiple sequence alignments were carried out using ClustalW2 (Pearson format). Each HCV isolate sequence is approximately 3100 amino acid residues in length. The first part of the alignment is depicted (up to 60 amino acid residues). Amino acid residues highlighted in blue show variation in sequence that differs from the consensus. The nonamer sequence (LLPRRGPRL) highlighted in the black box represents a potential HLA-Cw*0102 binding epitope. Residues outlined in the red box highlight the sequence variation present within this nonamer peptide in isolate number 12 and 13 (LLPRRGPKL). Residues denoted * = identical. Residues denoted : = conserved. Residues denoted . = semi-conserved.
116
Fig 4.2 An illustration of the peptide binding specificity of Class I HLA-Cw*0102 allotype
HLA-Cw*0102 preferably binds nonamer peptides with dominant amino acid residues proline at P3 and Leucine at P9 that anchor the peptide within the HLA class I peptide binding groove. Alanine or leucine are commonly enriched at P2. Figure is adapted from (Marsh et al, 2000, The HLA FactsBook).
The structure of the HCV polyprotein is outlined in Fig 4.3A. After translation of the HCV genome into the polyprotein it is cleaved into three structural proteins (Core protein, envelope protein 1 (E1), envelope protein 2 (E2)) and seven non-structural proteins (P7, NS2, NS3, NS4A, NS4B, NS5A, NS5B) (Takamizawa et al, 1991; Reed and Rice, 2000).
In total, eight nonamer peptide sequences from HCV were identified that may potentially bind HLA-Cw*0102 (Fig 4.3B). Potential HLA-Cw*0102 binding epitopes identified were analysed further on the ClustalW2 alignment carried out between the 31 HCV isolate sequences. Any sequence variation between each isolate sequence was recorded. Overall, there was a high degree of conservation between each HCV isolate. As shown in Fig 4.1, one peptide that was identified from the core region of HCV exhibited some sequence variation within one isolate. The nonamer sequence LLPRRGPRL contained an arginine (R) to lysine (K) mutation LLPRRGPKL at position 8 (P8) in these two out of 31 isolates (Fig 4.1). As position 8 of peptide modulates KIR binding, the variant peptide containing the mutation at P8 (LLPRRGPKL) was also synthesised. Previous studies have identified the decamer peptide YLLPRRGPRL (HCV core35-44) as an HLA-A*0201 epitope that can lead to viral escape from the HLA-A*0201-restricted CTL response (Battegay et al, 1995) and that
117 can also upregulate HLA-E expression (Nattermann et al, 2005). However, for HLA-Cw*0102 binding we synthesised the nonamer peptide (LLPRRGRPL) without tyrosine at P1.
Seven other potential HLA-Cw*0102 binding epitopes were also identified within different regions of the HCV polyprotein (Fig 4.3B). All identified peptides were searched using a protein BLASTP search (Altschul et al, 1990) in order to confirm their location within the HCV subtype 1b polyprotein. The other peptides that were identified did not exhibit any major sequence variation between isolates. In particular, the HCV NS31253-1262 epitope (LNPSVAATL) was identical in all 31 isolate sequences obtained from the Irish cohort (Appendix I).
118
Fig 4.3 HCV peptides synthesised and their location within the HCV polyprotein
(A) Structure of HCV genome and polyprotein (post-cleavage) into three structural proteins (Core, E1 and E2) and seven non-structural proteins (p7, NS2, NS3, NS4A, NS4B, NS5A, NS5B). (B) Table depicting the nine peptides synthesised and their location within the translated HCV polyprotein. Amino acid residues highlighted in red indicate the positions of the peptide that are important for anchoring to HLA-Cw*0102.
119 4.2 Conservation of the predicted HLA-Cw*0102 binding epitopes across different HCV genotypes
There are six main genotypes of HCV in the world with HCV genotype 1 being the most prevalent (Simmonds, 2004; Simmonds et al, 2005). Each genotype can be further divided into subtypes and quasispecies. Each genotype of HCV differs from another by approximately 30-35% in nucleotide sequence and subtypes differ from each other by approximately 20-25% (Simmonds, 2004). Genotypes 1 and 4 are the most diificult to tackle with the current treatments when compared with the other genotypes (Horner and Gale, 2013). The cohort of Irish patients which we received data for were infected with HCV subtype 1b. HCV subtype 1b is particulary prevalent in Europe whereas subtype 1a is more prevalent in North America. Wide distribution of Genotypes 1a, 1b and 3a has occured in Western countries due to IDU (Intravenous drug use) and contaminated blood transfusions (Simmonds, 2004).
A ClustalW2 alignment was performed on the protein sequences encoding each main HCV genotype (1a, 1b, 2, 3, 4, 5 and 6) to observe if there was any variation or conservation between the peptide residues that we had synthesised (Fig 4.4 & 4.5). Overall, HCV peptide 1 (LLPRRGPRL) and HCV peptide 5 (LNPSVAATL) were highly conserved between genotypes with seven and eight residues out of nine being identical respectively (Fig 4.4 & 4.5). LLPRRGPRL is derived from the Core protein of HCV and LNPSVAATL is derived from the NS3 protein.
HCV p3 (LSPRPVSYL), HCV p2 (AQPGYPWPL), HCV p7 (ARPDYNPPL) and HCV p8 (RAYLNTPGL) exhibited some conservation between the different genotypes with at least six residues out of nine being identical in each genotype (Fig 4.4 & 4.5). LSPRPVSYL is derived from NS2 protease. AQPGYPWPL is from the core protein. ARPDYNPPL is from the NS5A protein. RAYLNTPGL is derived from the NS3 protein.
The least conserved of the peptides across the different HCV genotypes were HCV p4 (LSPHYKVFL) and HCV p6 (VLPCSFTTL). These two peptides only had four and two out of nine identical residues respectively (Fig 4.4 & 4.5). LSPHYKVFL is derived from the NS2 protein and VLPCSFTTL from the envelope (E2) protein.
120
HCV ! Peptides! Genotypes!
HCV p1! LLPRRGPRL! Core region!
HCV p2! AQPGYPWPL! Core region!
HCV p6! E2 region! VLPCSFTTL!
Fig 4.4 Multiple sequence alignment between each HCV genotype at potential HLA-Cw*0102 binding-epitopes
Multiple sequence alignments were carried out using ClustalW2. Each HCV genotype sequence (genotypes 1a, 1b, 2, 3, 4, 5 and 6) was approximately 3100 amino acid residues in length. . Amino acid residues highlighted in red show the peptides synthesised and the variation or conservation in the potential epitopes. Residues denoted * = identical. Residues denoted : = conserved. Residues denoted . = semi-conserved.
121 HCV ! Peptides! Genotypes!
HCV p4! NS2 region! LSPHYKVFL!
HCV p3! NS3 region! LSPRPVSYL!
HCV p5! NS3 region! LNPSVAATL!
HCV p8! RAYLNTPGL! NS3 region!
HCV p7! ARPDYNPPL! NS5A region!
Fig 4.5 Multiple sequence alignment between each HCV genotype at potential HLA-Cw*0102 binding-epitopes
Multiple sequence alignments were carried out using ClustalW2. Each HCV genotype sequence (genotypes 1a, 1b, 2, 3, 4, 5 and 6) was approximately 3100 amino acid residues in length. Amino acid residues highlighted in red show the peptides synthesised and the variation or conservation in the potential epitopes. Residues denoted * = identical. Residues denoted : = conserved. Residues denoted . = semi-conserved.
122 4.3 Peptide stabilisation assays with HCV peptides
To determine whether HCV peptides could stabilise HLA-Cw*0102, cell surface stabilisation assays were carried out as shown previously in Chapter 3 using different concentrations of each HCV peptide. The human B lymphoblastoid 721.174 cell-line expressing the class I HLA-Cw*0102 allele was used for all stabilisation assays carried out (Young et al, 1998; Andersen et al, 1999). The 721.174 cell-line are deficient in TAP (transporter associated with antigen processing) and can be exogenously pulsed with peptide.
721.174 target cells were incubated with different concentrations of each HCV peptide ranging from 0.5 µM to 200 µM for 16 hours at 26°C. Following the incubation, cells were stained with the HLA-Cw*0102 specific primary antibody VP6G3 (a kind gift from Dr Arend Mulder). A FITC-conjugated secondary antibody was used to detect expression of HLA- Cw*0102 molecules. Total fluorescence events were collected and analysed within the live cell population.
As shown in Fig 4.6, HLA-Cw*0102 was not stabilised to a maximal level on 721.174 cells by 200 µM of HCV core-derived peptide LLPRRGPRL or the variant peptide LLPRRGPKL in comparison with the endogenous VAPWNSLSL peptide. Overall, five of the HCV peptides (LLPRRGPRL (HCV 1a), LLPRRGPKL (HLA 1b), AQPGYPWPL (HCV 2), ARPDYNPPL (HCV 7) and RAYLNTPGL (HCV 8)) were deemed as low affinity for HLA-Cw*0102 and did not reach saturation at concentrations >200 µM peptide (Fig 4.6). Three of the HCV peptides stabilised HLA-Cw*0102 with an intermediate affinity and approached saturation at >100 µM of peptide (LSPRPVSYL (HCV 3), LSPHYKVFL (HCV 4), LNPSVAATL (HCV5)) (Fig 4.7). Only one of the HCV peptides (VLPCSFTTL (HCV 6)) was able to stabilise HLA-Cw*0102 at low concentrations similar to that of the VAP endogenous peptide (Fig 4.8). This HCV peptide was deemed as high affinity for HLA-Cw*0102 (Fig 4.8).
Scatchard analysis was also carried out on HCV 5 (LNPSVAATL) and HCV 6 (VLPCSFTTL)
(Fig 4.9). Kd values showed that HCV 6 had a similar affinity to HLA-Cw*0102 as the VAP peptide derivatives in Chapter 3 (Kd ~0.15 µM) (Fig 4.10B). The Kd of LNP was 55.5 µM and of a much lower affinity for HLA-Cw*0102 than the VAP peptide derivatives (Fig 4.9A).
123
Fig 4.6 The majority of HCV peptides have a low affinity for HLA-Cw*0102
Peptides were incubated with 721.174 target cells for 16 hours at 26°C. Expression of HLA-Cw*0102 was determined by an HLA-Cw*0102-specific antibody (VPG63) (A) Dotplot showing gating on live .174 cell population (B) Overlay histogram plots depicting the expression of HLA-Cw*0102 in the presence or absence of four HCV peptides (HCV 1 LLPRRGPRL; HCV 2 AQPGYPWPL; HCV 7 ARPDYNPPL; HCV 8 RAYLNTPGL). All peptides shown on the histogram plots are at a concentration of 10 µM. Black line with no fill is the secondary antibody only and black filled-in histograms are no peptide (.174 cells alone) or .174 cells with peptides. (C) Line graph of MFI versus peptide concentration for each peptide. The endogenous VAPWNSLSL peptide was used as a positive control. Results are from two independent experiments (n=2) and show means ±SD.
124
Fig 4.7 HLA-Cw*0102 stabilisation assay with intermediate affinity HCV peptides
Peptides were incubated with 721.174 target cells for 16 hours at 26°C. Expression of HLA-Cw*0102 was determined by an HLA-Cw*0102-specific antibody (VPG63). (A) Overlay histogram plots depicting the expression of HLA-Cw*0102 in the presence or absence of three intermediate-affinity HCV peptides (HCV 3 LSPRPVSYL; HCV 4 LSPHYKVFL; HCV 5 LNPSVAATL). All peptides shown on the histogram plots are at a concentration of 10 µM. Black line with no fill is the secondary antibody only and black filled-in histograms are no peptide (.174 cells alone) or .174 cells with peptides (B) Line graphs of MFI versus peptide concentration for each peptide. The endogenous VAPWNSLSL peptide was used as a positive control. Results are from two independent experiments (n=2) and show means ±SD.
125
Fig 4.8 VLPCSFTTL (HCV p6) can bind HLA-Cw*0102 to a high-affinity
Peptides were incubated with 721.174 target cells for 16 hours at 26°C. Expression of HLA-Cw*0102 was determined by an HLA-Cw*0102-specific antibody (VPG63). (A) Overlay histogram plot depicting the expression of HLA-Cw*0102 in the presence or absence of one high-affinity HCV peptide (HCV peptide 6; VLPCSFTTL). Peptide shown on the histogram plot is at a concentration of 10 µM. Black line with no fill is the secondary antibody only and black filled-in histogram is .174 cells with VLP peptide. (B) Line graph of MFI versus peptide concentration for each peptide. The endogenous VAPWNSLSL peptide was used as a positive control. Results shown are from two independent experiments (n=2) and show means ±SD.
126
Fig 4.9 Scatchard Analysis: Affinity of HCV peptide 5 & 6 for HLA-Cw*0102
Scatchard Plots made from stabilisation experiments of the HLA-Cw*0102 allele and analysis by Flow Cytometry with HLA-Cw*0102-specific antibody (VP6G3). (A) and (B) show HCV peptides LNPSVAATL (HCV 5) and VLPCSFTTL (HCV 6) respectively. The regression line was derived using the formula Y= Bmax*X/ (Kd + X) for each peptide using GraphPad Prism6 software. Kd is the dissociation constant and Bmax is the maximal binding.
127 4.3 Peptide binding affinities with online prediction algorithms
Online algorithms were also used to compare the predicted binding affinities of each HCV peptide for HLA-Cw*01 with observed experimental data. The three online prediction algorithms used were: NetMHCPan (Nielsen et al, 2009), KISS (Kernel-based Inter-allele peptide binding prediction SyStem) (Jacob and Vert, 2008) and ADT (Adaptive Double Threading) (Schueler Furman et al, 2006). All algorithm prediction-binding scores are shown in Fig 4.10.
Most prediction algorithms have been well trained for both HLA-A and HLA-B alleles but not for HLA-C. The endogenous VAPWNSLSL peptide and its derivatives were also entered into each programme to obtain predicted binding affinities for HLA-Cw*01. This would allow for a direct comparison with output obtained for each HCV peptide.
NetMHCPan was the only algorithm that gave results which correlated with the observed experimental stabilisation data (stabilisation data Fig 4.6 - 4.9). The HCV subtype 1b consensus sequence was also entered into each algorithm to obtain a list of scores for nonamer peptides that may bind HLA-Cw*01. One nonamer peptide was also chosen that gave high binding scores in each algorithm (RAYLNTPGL). This peptide also contained key anchor residues at positions for binding HLA-Cw*01 and interaction with KIR so was therefore deemed a good candidate. However, this peptide (RAYLNTPGL) did not stabilise HLA-Cw*0102 on 721.174 cells (Fig 4.6)
128 Actual$Binding$ Predicted$Binding$Affini'es$ Affini'es$$
Peptides! Protein/ Net MHC Pan! KISS! ADT! Approximate Region! Peptide Score! (Binding Affinity (µM) Affinity (nM)! Energies)! at 50% stabilisation!
VAPWNSLSL! TIMP1 22.70! 1! 5.919544! ~1 µM! (Human)! VLPCSFTTL! E2 ! 199.73! 1! 10.485140! ~1 µM!
LNPSVAATL! NS3! 1306.03! 0! 7.113784! ~ 50 µM! LSPRPVSYL! NS3! 276.65! 0! 7.147722! > 100 µM! Core! 296.22! 1! 8.483209! > 100 M! HCV$$ LLPRRGPRL! µ pep'des$ LLPRRGPKL! Core! 461.97! 1! 8.422783! > 100 µM! RAYLNTPGL! NS3! 250.21! 1! 5.517011! > 100 µM! AQPGYPWPL! Core! 1899.79! 1! 7.790346! > 100 µM! ARPDYNPPL! NS5A! 4401.92! 1! 7.753846! > 100 µM! 1! 7.521453! > 100 µM! LSPHYKVFL! NS2! 5333.43!
Fig 4.10 Algorithm binding predictions of HCV peptides to HLA-Cw*0102
Table summarising the algorithm binding predictions by NetMHCPan, KISS and ADT for each HCV peptide. The endogenous VAPWNSLSL peptide is listed at the top of the table. Approximate affinities of each peptide for HLA-Cw*0102 from experimental stabilisation data are also plotted in the table. ADT (lower binding energies corresponds to better binders). A KISS score of 0 represents no binding and a KISS score of 1 represents binding. NetMHCPan peptide binding output is shown in terms of predicted affinity (nM).
129 4.5 The binding specificity of HCV peptides to KIR2DL2-Fc and KIR2DL3-Fc proteins
As in Chapter 3, 721.174 cells were incubated with each 10 µM HCV peptide for 16 hours at 26°C and the following day were stained with KIR2DL2-Fc and KIR2DL3-Fc proteins. Detection of KIR2DL2-Fc or KIR2DL3-Fc binding was with Protein-A conjugated to Alexafluor-488. The optimal molar ratios of KIR-Fc and Protein-A Alexafluor 488 were used for the assay (6:1) (R&D Systems). 721.174 cells without peptide (‘No peptide’) were used as a control. The endogenous VAP-LS peptide and inhibitory VAP-FA peptide were also used as controls.
As shown in Fig 4.11 & 4.12, the population of live 721.174 cells was gated. Non-specific staining with secondary Protein-A Alexafluor488 alone was accounted for. A ‘no peptide’ sample was used as a negative control for KIR2DL2-Fc or KIR2DL3-Fc binding. The only clear shift in KIR2DL2-Fc binding was with the LLPRRGPRL (HCV 1a) peptide with a much lesser shift in binding with LLPRRGPKL (HCV 1b) (Fig 4.11). LLPRRGPRL peptide also generated a shift in KIR2DL3-Fc binding (Fig 4.12). The rest of the HCV peptides presented on .174 cells did not bind significantly to KIR2DL2-Fc or KIR2DL3-Fc proteins.
130
Fig 4.11 KIR2DL2-Fc binding to .174 cells pulsed with HCV peptides
(A) Scatter dotplot showing gating on live .174 cell population. (B) Histograms depicting the shift in MFIs for .174 cells pulsed with each individual HCV peptide (200 µM) or .174 cells with ‘no peptide’ (Black filled-in overlays). Protein-A alexafluor488 alone is shown by black (no-fill) lines only. (C) Bar- chart with the displayed mean MFI values for each peptide. Bar chart represents three independent experiments (n=3) and means are ±SD. Dashed black line represents the MFI for the control ‘no peptide’ sample as the baseline negative control.
131
Fig 4.12 KIR2DL3-Fc binding to .174 cells pulsed with HCV peptides
(A) Scatter dotplot showing gating on live .174 cell population. (B) Histograms depicting the shift in MFIs for .174 cells pulsed with each individual HCV peptide (200 µM) or .174 cells with ‘no peptide’ (Black filled-in overlays). Protein-A alexafluor488 alone is shown by black (no-fill) lines only. (C) Bar- chart with the displayed mean MFI values for each peptide. Bar chart represents three independent experiments (n=3) and means are ±SD. Dashed black line represents the MFI for the control ‘no peptide’ sample as the baseline negative control.
132 4.6 CD107a NK degranulation assay with HCV peptides
All nine HCV peptides were subsequently tested in functional NK CD107a degranulation assays to investigate what effect they may have on NK degranulation in KIR2DL2+ or KIR2DL3+ donors. 721.174 target cells were pulsed with 100 µM of each HCV peptide to ensure maximum stabilisation of HLA-Cw*0102 for 16 hours at 26°C. On the following day, target cells pulsed with each respective peptide were co-incubated with PBMC from KIR2DL2 homozygous or KIR2DL3 homozygous donors stimulated overnight with 1 ng/mL IL-15 and an anti-human CD107a antibody. PBMC and target cells were co-incubated for 4h at 26°C at an optimised E:T ratio of 5:1. Cells were stained and analysed by flow cytometry.
Results showed that only one of the invariant HCV peptides LLPRRGPRL (HCV 1a) derived from the core protein had a significant inhibitory effect and on NK cell degranulation in one KIR2DL2 homozygous donor (p<0.001) and one KIR2DL3 homozygous donor (p<0.01). %CD107a expression decreased to levels comparable with the inhibitory VAP-FA peptide in these two donors tested (15-20 %CD107a expression) (Fig 4.13C). The inhibitory effect of HCV 1a could be observed in the KIR2DL2/3 heterozygous donor tested, however it was not statistically significant compared to the 'no peptide' sample. The remaining HCV peptides (HCV 1b-8) had a weak or no inhibitory effect on KIR2DL2/3+ NK cell de-granulation in all donors tested (Fig 4.13B&C). These differences were not significant. Fig 4.13C summarises results of CD107a degranulation assays with three donors possessing different KIR2DL2/3 genotypes; one KIR2DL3 homozygous, one KIR2DL2 homozygous and one KIR2DL2/3 heterozygous.
133 A!
! R3 15%! R4 34%! !
P2 42%! ! CD3-PerCP SSC-A CD56-PE
FSC-A! CD56-PE! !CD158b-FITC! B! ! ! ! ! VAP-DA! No peptide! VAP-FA ! HCV 1a! HCV 1b! HCV 2 ! 32%! 34%! 28%! 48%! 3%! 5%!
! ! Count! HCV 3! HCV 4 ! HCV 5! HCV 6! HCV 7! HCV 8!
36%! 44%! 33%! 27%! 16%! 32%!
CD107a- AF647!
C! Combined donors! **** 60 **** *** 40
20 % CD107a expression 0
HCV HCVp2 HCVp3 HCVp4 HCVp5 HCVp6 HCVp7 p8 VAP-DAVAP-FAHCV HCVp1a p1b No peptide Peptide (100 µM)
KIR2DL2 ! KIR2DL3 ! KIR2DL2/3 ! homozygous donor! homozygous donor! heterozygous donor! *** 60 ns *** * 60 ns 60 ** ** ns 40 * 40 40
20 20 20 % CD107a expression % CD107a expression % CD107a expression 0 0 0
VAP-FA HCV HCVp2 HCVp3 HCVp4 HCVp5 HCVp6 HCVp7 p8 VAP-FA HCV HCVp2 HCVp3 HCVp4 HCVp5 HCVp6 HCVp7 p8 HCV HCVp2 HCVp3 HCVp4 HCVp5 HCVp6 HCVp7 p8 VAP-DA HCV HCVp1a p1b VAP-DA HCV HCVp1a p1b VAP-DAVAP-FAHCV HCVp1a p1b No peptide No peptide No peptide Peptide (100 µM) Peptide (100 µM) Peptide (100 µM)
Fig 4.13 The majority of HCV peptides have a weak or no inhibitory effect on NK cell degranulation in CD158b+ NK cells from three different donors
(A) Gating strategy for CD3- CD56+ CD158b+ NK cells. (B) Representative histogram plots for each endogenous or HCV peptide depicting %CD107a expression. (C) Bar-charts show overall percentage of CD56+ CD158b+ NK cells from three donors expressing CD107a post-incubation with .174 target cells alone or .174 target cells pulsed with 100 µM of each peptide VAP-FA, VAP-DA or HCV peptide 1-8. HCV 1a (LLPRRGPRL); HCV1b (LLPRRGPKL); HCV 2 (AQPGYPWPL); HCV 3 (LSPRPVSYL) HCV 4 (LSPHYKVFL); HCV 5 (LNPSVAATL); HCV 6 (VLPCSFTTL); HCV 7 (ARPDYNPPL); HCV 8 (RAYLNTPGL). Bar-charts represent two independent experiments for each of the three donors (n=3) and combined bar chart shows all the replicate values ±SD. One- way ANOVA was used (Newman Keuls post-test) (*, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001).
134 LLPRRGPRL core peptide (HCV 1a) was significantly inhibitory in a KIR2DL2 homozygous donor and a KIR2DL3 homozygous donor. However, the variant LLPRRGPKL (HCV 1b) peptide did not induce any inhibition of NK degranulation highlighting the importance of the residue at position 8 of peptide for mediating the KIR2DL2/3:HLA-C1 interaction (Fig 4.14). There was a significant difference in degranulation between HCV 1a and HCV 1b in both KIR2DL2 and KIR2DL3 homozygous donors at a lower concentration of the HCV peptides (10 µM) (Fig 4.14). As shown in Fig 4.14, HCV peptide 1a gave approximately the same level of inhibition as the strongly inhibitory peptide VAP-FA and more inhibition than the weakly inhibitory peptide VAP-RA.
Each HCV peptide was used in the assays at a concentration of 100 µM. However, inhibition of NK degranulation by LLPRRGPRL peptide was dose-dependent as inhibition was only achieved at concentrations greater than or equal to 10 µM of this peptide in comparison with inhibition by VAP-FA peptide which was achieved in previous assays by Fadda et al at 0.5-1 µM (Fig 4.13 & 4.14).
135
A! KIR2DL2 homozygous donors B! KIR2DL3 homozygous donors
60 60 *** **** *** **** *** **** 40 40
20 20 % CD107a expression % CD107a expression 0 0
VAP-FA VAP-FA VAP-RA VAP-DA HCV p1a HCV p1b VAP-RA VAP-DA HCV p1a HCV p1b No peptide No peptide
Peptide (10 M) Peptide (10 µM) µ
Fig 4.14 HCV core peptide LLPRRGPRL has an inhibitory effect on NK cell degranulation with a naturally occuring mutation at position 8 (R->K) relieving this inhibition
Bar charts displaying the mean %CD107a expression from PBMC from (A) Two KIR2DL2 homozygous donors (n=2) and (B) Two KIR2DL3 homozygous donors (n=2) after co-incubation with .174 cells pulsed with endogenous peptide derivatives (VAP-FA, VAP-RA and VAP-DA) and two HCV peptides (LLPRRGPRL (HCV p1a) and LLPRRGPKL (HCV p1b)). All peptides were used at a final concentration of 10 µM. CD107a expression was analysed on the CD158b+ NK cell population. Data are from one independent experiment per donor. Bar-charts show means from each donor group ±SD. One-way ANOVA was used (Newman Keuls post-test) (***, p<0.001; ****, p<0.0001).
136 4.7 Effect of HCV peptides 5 & 6 on NK inhibition in KIR2DL2 and KIR2DL3 homozygotes
In peptide mix assays conducted by Fadda et al and as part of this project, low levels of the weakbinding VAP-DA peptide were found to antagonise KIR2DL2/3+ NK cell inhibition induced by VAP-FA. We therefore aimed to screen each of the HCV peptides for any potential antagonistic effect on inhibition by performing peptide mixes with the inhibitory VAP-FA peptide.
Each HCV peptide was screened in KIR2DL2 homozygous, KIR2DL3 homozygous and KIR2DL2/3 heterozygous donors using peptide mixes consisting of 5 µM VAP-FA and 5 µM HCV peptide. 10 µM was chosen as the final concentration of peptide in the mixes as this was deemed to be the lowest concentration at when HLA-Cw*0102 was approaching saturation with peptide. In an initial screen, two of the HCV peptides, LNPSVAATL (HCV peptide 5 NS31254-1263) and VLPCSFTTL (HCV peptide 6 E2673-682) appeared to relieve inhibition by VAP-FA to some extent in some donors (Fig 4.15). These two HCV peptides were also chosen for further assays as they gave good stabilisation of HLA-Cw*0102 molecules on .174 cells (Fig 4.8 & 4.9). The rest of the HCV peptides did not have any effect on the inhibition mediated by VAP-FA.
Further assays with HCV peptide 5 and HCV peptide 6 were carried out on five KIR2DL2 homozygous donors (n=5), five KIR2DL3 homozygous donors (n=5) and two KIR2DL2/3 heterozygous donors (n=2). All peptide mixes were carried out simultaneously using 5 µM of each peptide as done previously. A peptide mix containing VAP-DA (5 µM) and VAP-FA (5 µM) was used as a control peptide mix. In peptide mixes containing 5 µM VAP-FA and 5 µM of HCV5, there was a reduction in NK cell inhibition in KIR2DL2 homozygous donors and in KIR2DL2/3 heterozygous donors (Fig 4.17 A&B & Fig 4.19 A&B respectively). This decrease in inhibition was significant in the KIR2DL2 homozygous group and comparable to that observed with the VAP-DA and VAP-FA peptide mix (Fig 4.17 A&B). In the KIR2DL3 homozygous group, there was no effect of HCV5 on NK cell inhibition and there was no significant difference in degranulation (Fig 4.18 A&B). The normalised data for HCV5 in each group of donors is shown in Fig 4.20. All data in Fig 4.20 was normalised to the ‘No peptide’ control sample, whereby ‘no peptide’ is equal to 100% CD107a degranulation.
Similar peptide mixes with HCV peptide 6 (5 µM) and VAP-FA (5 µM) were carried out and showed that HCV peptide 6 was able to reduce inhibition levels mediated by VAP-FA in all groups of donors (Fig 4.21, Fig 4.22 and Fig 4.23). The normalised data for HCV6 in each
137 group of donors is shown in Fig 4.24. Again, all data for HCV peptide 6 was normalised to the ‘no peptide’ control sample (Fig 4.24).
All donors combined!
50
40
30
20 % CD107a 10
0 VAP-FA FA + DA No peptide FA + HCV p1 FA + HCV p2 FA + HCV p3 FA + HCV p4 FA + HCV p5 FA + HCV p6 FA + HCV p7 FA + HCV p8
KIR2DL2 ! KIR2DL2/3 ! KIR2DL3 !
homozygous donor! heterozygousKIR2DL2/3 heterozygous donor donor! KIR2DL3homozygous homozygous donor donor ! KIR2DL2 homozygous donor 50 50 50 40 40 40 30 30 30
20 20 20 % CD107a % CD107a % CD107a 10 10 10
0 0 0 VAP-FA VAP-FA VAP-FA FA + DA FA + DA FA + DA No peptide No peptide No peptide FA + HCV p1 FA + HCV p2 FA + HCV p3 FA + HCV p4 FA + HCV p5 FA + HCV p6 FA + HCV p7 FA + HCV p8 FA + HCV p1 FA + HCV p2 FA + HCV p3 FA + HCV p4 FA + HCV p5 FA + HCV p6 FA + HCV p7 FA + HCV p8 FA + HCV p1 FA + HCV p2 FA + HCV p3 FA + HCV p4 FA + HCV p5 FA + HCV p6 FA + HCV p7 FA + HCV p8
Fig 4.15 Initial screen with eight HCV peptides and VAP-FA peptide
CD107a assay was carried out on PBMC from three donors (KIR2DL2 homozygous, KIR2DL3 homozygous and KIR2DL2/3 heterozygous). PBMC from each donor were co-incubated with .174 cells that had been simultaneously pulsed with equal concentrations of each HCV peptide (1-8) (5 µM) and VAP-FA peptide (5 µM) or VAP-FA alone (5 µM). All peptide mixes used were prepared to a final concentration of 10 µM. CD107a expression was analysed on the CD56+ CD158b+ NK cell population. Dashed black line represents the level of degranulation for the negative control ‘no peptide’ sample. Dashed red line represents the lower level of degranulation for the VAP-FA sample.
138
4.16 Gating strategies used for CD56+ CD158b+ and CD56+ CD158b- NK cells
Scatter dotplots demonstrating the simplified gating strategy used to distinguish the(A) CD3- CD56+ CD158b+ and the (B) CD3- CD56+ CD158b- NK cell populations as part of CD107a assays. Fresh PBMC stimulated overnight with IL-15 (1 ng/mL) were separated on lymphoid forward scatter (FSC-A) and side scatter (SSC-A) (gate P2). Live cells were separated by staining with a fixable Livedead stain (Invitrogen) that is taken up by dead cells only (gate R1). The live lymphocyte population was analysed in terms of surface CD3 and CD56 expression and the CD3- CD56+ NK sub-population was defined (gate R2). CD3- CD56+ lymphocytes were separated and analysed in terms of CD158b (KIR2DL2/KIR2DL3/KIR2DS2) expression (gate R6).
139
Fig 4.17 LNPSVAATL (HCV5) reduces CD158b+ NK cell inhibition in KIR2DL2 homozygous donors .174 cells were pulsed with 5 µM of each peptide (HCV5 (LNPSVAATL), endogenous derivatives (VAP-FA, VAP-DA) or peptide mixes containing equal concentrations of each peptide (5 µM: 5 µM). All mixes were prepared to a final concentration of 10 µM. %CD107a expression was analysed in the gated CD158b+ or CD158b- NK sub- populations. (A) and (B) show representative histogram plots and a bar chart respectively of CD158b+ data from five KIR2DL2 homozygous donors. (C) and (D) show representative histogram plots and a bar chart respectively of CD158b- data from four of the KIR2DL2 homozygous donors. Each experiment was repeated twice per donor. Data in (B) shows all replicates from the donors ±SEM. Data in (D) shows means ±SEM. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (*, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001).
140
Fig 4.18 LNPSVAATL (HCV5) does not have an effect on CD158b+ NK cell inhibition in KIR2DL3 homozygous donors .174 cells were pulsed with 5 µM of each peptide (HCV5 (LNPSVAATL), endogenous derivatives (VAP-FA, VAP-DA) or peptide mixes containing equal concentrations of each peptide (5 µM: 5 µM). All mixes were prepared to a final concentration of 10 µM. %CD107a expression was analysed in the gated CD158b+ or CD158b- NK sub-population. (A) and (B) show representative histogram plots and a bar chart respectively of CD158b+ data from five KIR2DL3 homozygous donors. (C) and (D) show representative histogram plots and a bar chart respectively of CD158b- data from five KIR2DL3 homozygous donors. Each experiment was repeated twice per donor. Data in (B) shows all replicates from the donors ±SEM. Data in (D) shows means ±SEM. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (*, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001).
141
Fig 4.19 LNPSVAATL (HCV5) reduces CD158b+ NK cell inhibition in KIR2DL2/3 heterozygous donors .174 cells were pulsed with 5 µM of each peptide (HCV5 (LNPSVAATL), endogenous derivatives (VAP-FA, VAP- DA) or peptide mixes containing equal concentrations of each peptide (5 µM: 5 µM). All mixes were prepared to a final concentration of 10 µM. %CD107a expression was analysed in the gated CD158b+ or CD158b- NK sub- populations. (A) and (B) show representative histogram plots and a bar chart respectively of CD158b+ data from two KIR2DL2/3 heterozygous donors. (C) and (D) show representative histogram plots and a bar chart respectively of CD158b- data from two KIR2DL2/3 heterozygous donors. Each experiment was repeated twice per donor. Data in (B) and (D) is means ±SEM. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (All non-significant).
142
Fig 4.20 Normalised data for HCV5 from experiments with KIR2DL2 homozygous, KIR2DL3 homozygous and KIR2DL2/3 heterozygous donors
Data from Fig 4.17, 4.18 and 4.19 were normalised using the ‘no peptide’ control sample. %CD107a expression with ‘no peptide’ was set at 100%. Fig 4.20 depicts %CD107a expression data from after normalisation. Data show means ±SEM from five KIR2DL2 donors, five KIR2DL3 donors and two KIR2DL2/3 heterozygous donors. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (*, p<0.05; **, p<0.01).
143
Fig 4.21 VLPCSFTTL (HCV6) reduces CD158b+ NK cell inhibition in KIR2DL2 homozygous donors .174 cells were pulsed with 5 µM of each peptide (HCV6 (VLPCSFTTL), endogenous derivatives (VAP-FA, VAP-DA) or peptide mixes containing equal concentrations of each peptide (5 µM: 5 µM). All mixes were prepared to a final concentration of 10 µM. %CD107a expression was analysed in the gated CD158b+ or CD158b- NK sub- populations. (A) and (B) show representative histogram plots and a bar chart respectively of CD158b+ data from five KIR2DL2 homozygous donors. (C) and (D) show representative histogram plots and a bar chart respectively of CD158b- data from four KIR2DL2 homozygous donors. Each experiment was repeated twice per donor. Data in (B) shows all replicates ±SEM. Data in (D) shows means ±SEM. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (***, p<0.001; ****, p<0.0001).
144
Fig 4.22 VLPCSFTTL (HCV6) reduces CD158b+ NK cell inhibition in KIR2DL3 homozygous donors
.174 cells were pulsed with 5 µM of each peptide (HCV6 (VLPCSFTTL), endogenous derivatives (VAP-FA, VAP- DA) or peptide mixes containing equal concentrations of each peptide (5 µM: 5 µM). All mixes were prepared to a final concentration of 10 µM. %CD107a expression was analysed in the gated CD158b+ or CD158b- NK sub- populations. (A) and (B) show representative histogram plots and a bar chart respectively of CD158b+ data from five KIR2DL3 homozygous donors. (C) and (D) show representative histogram plots and a bar chart respectively of CD158b- data from five KIR2DL3 homozygous donors. Each experiment was repeated twice per donor. Data in (B) shows all replicates ±SEM. Data in (D) shows means ±SEM. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (*, p<0.05; **, p<0.01).
145
Fig 4.23 VLPCSFTTL (HCV6) reduces CD158b+ NK cell inhibition in KIR2DL2/3 heterozygous donors
.174 cells were pulsed with 5 µM of each peptide (HCV6 (VLPCSFTTL), endogenous derivatives (VAP-FA, VAP- DA) or peptide mixes containing equal concentrations of each peptide (5 µM: 5 µM). All mixes were prepared to a final concentration of 10 µM. %CD107a expression was analysed in the gated CD158b+ or CD158b- NK sub- populations. (A) and (B) show representative histogram plots and a bar chart respectively of CD158b+ data from two KIR2DL2/3 heterozygous donors. (C) and (D) show representative histogram plots and a bar chart respectively of CD158b- data from two KIR2DL2/3 heterozygous donors. Each experiment was repeated twice per donor. Data shown in (B) and (D) is means ±SEM. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (All non-significant).
146
Fig 4.24 Normalised data from experiments with HCV6 (VLPCSFTTL) with KIR2DL2 homozygous, KIR2DL3 homozygous and KIR2DL2/3 heterozygous donors
Data from Fig 4.21, 4.22 and 4.23 was normalised using the ‘no peptide’ control sample. %CD107a expression with ‘no peptide’ was set at 100%. Fig 4.24 depicts %CD107a expression data from after normalisation. Data is means ±SEM from five KIR2DL2 donors, five KIR2DL3 donors and two KIR2DL2/3 heterozygous donors. One-way ANOVA (Tukey's Multiple comparison test) was used to determine significance. (*, p<0.05; **, p<0.01; ***, p<0.001).
147 4.8 Specificity of effect with HCV peptide 5 and HCV peptide 6 and KIR2DL2/KIR2DL3/KIR2DS2 population of NK cells
As stated previously, human anti-CD158b is a monoclonal antibody specific for the KIR2DL2, KIR2DL3 and KIR2DS2 receptors. To check that the effect on inhibition observed with HCV peptide 5 and 6 was only observed within the sub-population of CD158b+ NK cells and was not observed in the CD158b- sub-population results were also analysed on CD158b- NK cell sub-population.
Gating on the CD56+ CD158b- NK cell population in KIR2DL2 homozygous, KIR2DL3 homozygous and KIR2DL2/3 heterozygous donors showed that VAP-FA peptide had no inhibitory effect confirming it is specific for the inhibitory KIR2DL2/3-C1 interaction (Fig 4.17- Fig 4.24 C&D). No effect on inhibition (positive or negative) was observed in CD158b- NK cells with either VAP-DA or HCV peptides 5/6. Overall, as shown in all Figures 4.17-4.24 C&D, NK cell degranulation remained at the same level as the no peptide control within the CD158b- sub-population of NK cells. This was consistent for all groups of donors.
Human anti-CD158a is a monoclonal antibody that binds the inhibitory KIR2DL1 receptor and the activating KIR2DS1 receptor. The CD158a+ and CD158a-sub-populations of NK cells were also analysed in only KIR2DL2 homozygous donors for HCV peptide 5 to confirm no involvement of KIR2DL1 and/or KIR2DS1 receptors. There was no difference in degranulation levels within the CD158a+ population of NK cells (Fig 4.25). As shown in Fig 4.25, the only significant differences in NK degranulation were observed when gating on the CD158b+CD158a+ and CD158b+CD158a- populations. Thus, the effects mediated with these peptides was only relevant for CD158b+ NK cells.
148
Fig 4.25 Effects observed with HCV5 (LNPSVAATL) in KIR2DL2 homozygous donors is not found in the CD158b- NK cell sub-population
.174 cells were pulsed with 5 µM of each peptide HCV p5 (LNPSVAATL), endogenous derivatives (VAP-FA, VAP-DA) or peptide mixes containing equal concentrations of each peptide (5 µM: 5 µM). All peptide mixes were prepared to a final concentration of 10 µM. The following day PBMC from KIR2DL2 homozygous donors were co- incubated with the .174 cells with or without peptides. %CD107a expression was analysed in four gated NK cell sub-populations (CD158b+ CD158a+ double-positive NK cells, CD158b+ CD158a- , CD158b- CD158a+ single- positive NK cells and CD158b- CD158a- NK cells). Data displayed on bar-charts were normalised to the ‘no peptide’ control samples. Normalised data is from three KIR2DL2 homozygous donors. Each experiment was repeated twice per donor. Data is means ±SEM. One-way ANOVA (Tukey’s multiple comparison test) was used to determine significance. (*, p<0.05).
149 4.9 The involvement of NKG2/CD94 receptor family in the effects observed with HCV peptide 5 and 6
To rule out involvement of the NKG2/CD94 receptor family in the results we observed with these HCV peptides 5 & 6 we proposed that gating on the CD158b- sub-population of NK cells ought to aid this. We expected that CD94/NKG2 receptors should be expressed on both CD158b+ and CD158b- gated sub-populations so if the peptides were acting through this family of receptors we would expect to see the inhibitory or activating effects of the peptide(s) present in the CD158b- NK sub-population.
To further confirm the likelihood that there was no involvement of the CD94/NKG2 receptor family, peptide stabilisation assays were carried out on its ligand HLA-E. As HLA-E molecules are constitutively expressed on the surface of .174 cells, stabilsation assays for HLA-E were carried out simultaneously. Neither HCV 5 (LNPSVAATL) nor HCV 6 (VLPCSFTTL) were able to stabilise HLA-E molecules on the cell surface in comparison with the positive control HLA-G leader peptide that stabilises HLA-E (VMAPRTLFL) (Fig 4.26). VAPWNSLSL peptide was used as a negative control for HLA-E stabilisation.
150 A!
No peptide! VMAPRTLFL! VAPWNSLSL! LNPSVAATL! VLPCSFTTL!
! 1652.4! 2105.0! 1550.7! 1581.5! 1608.5! ! Count
3D12-PE! B! HLA-E stabilisation!
4000 VAPWNSLSL VMAPRTLFL 3000 LNPSVAATL VLPCSFTTL 2000 (3D12 MFI) (3D12 1000 HLA-E Stabilisation
0 0 20 40 60 80 100 Peptide (µM)
Fig 4.26 HCV peptides LNPSVAATL (HCV 5) and VLPCSFTTL (HCV 6) do not stabilise HLA-E molecules on .174 cells
(A) Overlay histogram plots depicting the expression of HLA-E in the presence or absence of endogenous VAPWNSLSL peptide, HCV peptide 5 (LNPSVAATL) and HCV peptide 6 (VLPCSFTTL). All peptides shown on histogram plots are at a concentration of 10 µM. VMAPRTLFL (HLA-G leader peptide) was used as a positive control for HLA-E binding. Expression of HLA-E was determined by an HLA-E PE-conjugated antibody (3D12-PE). (B) Line graph of MFI versus peptide concentration for each peptide. Results are from two independent experiments (n=2) and show means ±SD.
151 4.10 Discussion
Overall, our data suggest that HCV has not adapted substantially to polymorphic Inhibitory KIR:HLA-C interactions. This is implied by the observation that eight out of the nine invariant HCV peptides tested were either weakly binding to HLA-C and/or non-inhibitory to KIR2DL2/3-positive NK cells when presented by HLA-Cw*0102. Thus, either they are less likely to be loaded onto HLA-C molecules, or if loaded do not inhibit KIR2DL2/3+ NK cells. For the one peptide that did inhibit NK cells in this study (LLPRRGPRL), HLA-C stabilisation was not saturable, even at 200 μM of peptide. Furthermore, a naturally occurring mutation of this peptide at P8 (LLPRRGPKL) rendered it non-inhibitory. The non-inhibitory variant of the HCV core peptide LLPRRGPRL containing a lysine at P8 would have no selective advantage for the virus implying that selection at this locus is not KIR mediated. This mutation was found in two isolate sequences out of 31 within this cohort. The three algorithms that were used as part of this study gave outputs with poor correlation to the experimental data that was observed. The best example of this was observed with HCV peptide 8 (RAYLNTGPL). Each of the three algorithms predicted a high binding affinity of this peptide to HLA- Cw*0102. However, experimental data demonstrated that the RAYLNTPGL peptide was not able to stabilise HLA-Cw*0102 on .174 cells and did not have any inhibitory effect on NK cell degranulation. Therefore, although these algorithms have been well-trained for HLA-A/HLA- B binding, careful interpretation of results should be made with regard to HLA-C binding until experimental data can be also obtained for analysis.
Itakura et al had previously reported that no specific point mutation was directly associated with a change in viral load in the Irish HCV cohort (Itakura et al, 2005). Although the mutation of Arginine to Lysine in LLPRRGPRL has been found in other HCV genotypes (2, 4, 5a and 6), it was not frequent. Interestingly, the HCV core peptide also binds HLA-A*0201 and HLA-E, and the mutation has been previously suggested to lead to viral escape from the HLA-A*0201-restricted CTL response (Battegay et al, 1995; Nattermann et al, 2005). Therefore, as this is also a CTL epitope, there may be other selection pressures acting on it. Overall, there appears to be little selection pressure on these epitopes to induce inhibition of NK cells but there may be other selection pressures acting on these epitopes such as antibody-driven selection on the Envelope (E2) protein.
With regard to the conservation of these epitopes amongst different HCV genotypes we observed that most were relatively conserved. In particular, LLPRRGPRL and LNPSVAATL were the most conserved epitopes amongst genotypes 1a, 1b, 2, 3, 4, 5 and 6. Both epitopes were also conserved within the 31 isolate sequences of the HCV subtype 1b Irish cohort. The mutation from Arginine to Lysine at P8 in LLPRRGPRL was also present in the
152 consensus sequence obtained from HCV genotype 5. LNPSVAATL was also identical in each of the patients from the Irish HCV subtype 1b cohort. HCV peptide 6 (VLPCSFTTL) is derived from the E2 region of HCV, a usually hypervariable region, and showed the least conservation of all the HCV peptides amongst the different genotypes.
In order to bind KIR there is preference for large hydrophobic amino acids at P7 and small amino acids at P8 (Boyington et al, 2000). Mandelboim et al had previously reported that a lysine residue at P8 of peptides derived from both self and viral proteins (cocksackie virus) abolishes recognition of HLA-Cw*07-peptide complexes by KIR2DL2/3 receptors on NK cells (Mandelboim et al, 1997). However, substitution of P8 lysine with an arginine or histidine residue resulted in the generation of peptides that had a full inhibitory effect. This poses a question regarding adaptation of the virus to host KIR:HLA interactions as mutating this residue at P8 can result in increased or decreased levels of NK cell degranulation.
If HCV is not as well adapted to polymorphic KIR2DL2/3 receptors this would provide a clear selective evolutionary advantage for the host. It has been questioned whether HCV may have adapted better to the conserved NKG2/CD94 family of NK receptors and may have an inhibitory effect on NK cells via NKG2A/CD94- HLA-E interaction. In a study carried out by Nattermann in 2005, ~30 HCV core-derived peptides were analysed and the decamer YLLPRRGPRL core peptide was found to bind and stabilise HLA-E (Nattermann et al, 2005).It was found that this peptide had an inhibitory effect on NK cell cytotoxicity and this was mediated through interactions of HLA-E and the inhibitory receptor CD94/NKG2A (Nattermann et al, 2005). Work published from our group recently has also shown that the decamer core peptide YLLPRRGPRL does not inhibit degranulation by itself via NGK2A/CD94 and requires other peptides to mediate inhibition, a 'synergistic inhibitory' effect (Cheent et al, 2013).
In the context of viral infection, it is known that loss or down-regulation of surface HLA Class I does not account for NK cell killing of virally infected cells in a number of cases (Ljunggren and Karre, 1990; Malnati et al 1993). Malnati et al demonstrated that NK cells can directly kill cells infected with Herpesvirus that have “normal” levels of HLA class I expression (Malnati et al, 1993). It can be hypothesised that viral peptides presented on HLA class I molecules may act in a manner analogous to the endogenous VAP-DA peptide derivative, binding with very weak affinity to KIR receptors but yet significantly able to antagonise NK cell inhibition in the presence of a range of “normal” self- inhibitory peptide. This would prove advantageous to the protective host immune response early on in infection.
If it is assumed that the normal peptide repertoire of healthy cells is not 100% inhibitory then
153 a small fluctuation in peptide presented may be a good trigger of NK cell activation. As part of the “DRiP hypothesis” proposed by Professor Jonathan Yewdell, it has been suggested that the majority of antigenic peptides presented by MHC class I are defective ribosomal products (DRiPs), generated as by-products of protein synthesis (Yewdell, 1996). This would favour rapid presentation of viral peptides in an infected cell as the concentration of viral mRNAs surpasses the concentration of cellular mRNAs early in infection (Yewdell, 1996; Yewdell and Nicchitta, 2006). The other major contributor to the peptide content of our cells is from aged cellular proteins or ‘retirees’ (Yewdell et al, 2001; Yewdell, 2011; Yewdell, 2013; Croft et al, 2013). When investigating these invariant HCV peptides for future studies it is important to consider whether they naturally exist in infection and are cleaved and processed in the immunoproteasome for antigen presentation.
The majority of the peptides that we tested were only weakly binding to HLA-Cw*01 and we did not cover all HCV peptides with this approach. However, our approach was biased towards the selection of peptides with a high probability of binding to HLA-C1. Epitope prediction methods do not work as well for HLA-C as they do for HLA-A/HLA-B. Therefore, further non-intuitive screening of HCV peptides is unlikely to substantially affect the conclusion that most HCV peptides do not influence NK cell reactivity via the HLA- C1/KIR2DL2/3 interaction. However, as the output given to us by the binding algorithms correlated poorly with HLA-Cw*01 binding there may be some HCV epitopes that were missed during the initial screen. It would be essential to perform future elution studies on HCV infected cells at different time-points post-infection and analyse the epitopes for HLA- C1 binding.
Interestingly, our results are similar to the findings for HIV infection, in which a KIR-binding screen of 217 peptides revealed only 11 peptides that stabilised HLA-Cw*0102, but only one of these inhibited KIR2DL2-expressing NK cells (Fadda et al, 2012). This study also correlates with the work from our group in 2010 whereby two-thirds of 58 endogenous VAPWNSLSL peptide derivatives originally tested did not bind KIR2DL2/3-fusion constructs (Fadda et al, 2010). Therefore, less than one third of the 58 variants tested bound KIR2DL2 or KIR2DL3. Results here show that the inhibitory LLP-R peptide (HCV p1a) did not mediate any significant binding to KIR2DL2-Fc or KIR2DL3-Fc proteins in-vitro but was yet able to induce CD158b+ NK cell inhibition in the degranulation assays. This contrasts with the results reported in Chapter 3 for the VAP-DA peptide as VAP-DA gave no detectable binding to KIR-Fc but yet had the opposite activating (antagonistic) effect on CD158b+ NK cells. We propose that the LLP-R (HCV p1a) peptide can mediate a strong inhibitory effect via KIR2DL2/3 receptors, however the KIR-Fc reagents may not give the most reliable affinity
154 measurement. This correlates with the previous results using the VAP-DA peptide as VAP- DA does not mediate any binding to KIR2DL3-Fc or KIR2DL2-Fc in-vitro but can induce diffuse clustering of the KIR2DL3 receptor in-situ (Borhis et al, 2012). Interestingly, the strongly inhibitory VAP-FA peptide does elicit strong binding with KIR2DL2/3-Fc. We therefore hypothesise that the LLP-R (HCV p1a) peptide may bind KIR2DL2/3 receptors with a lower affinity than that of VAP-FA but overall may be able to induce a very strong inhibitiory signal comparable to that of VAP-FA.
The essential differences between NK cell and T cell recognition of peptides lie in the fact that NK cells can respond to a broader range of MHC-peptide complexes whereas T cells respond only to a specific MHC-peptide complex (Malnati et al, 1995; Lanier and Phillips, 1996; Long, 1996). In our study using HCV peptides there appears to be relatively limited selection of specific HCV epitopes for inhibition of KIR2DL2/3+ NK cell responses. However, Chapter 3 of this project has highlighted that NK cells may respond to changes in the overall peptide content of a cell. Therefore, NK cells may be evolutionarily better adapted to sensing larger shifts in the self-peptide repertoire of a cell that could occur during viral infection or tumour transformation.
HCV has an exceptionally high mutation rate similar to HIV (Purcell et al, 1991; Gomez et al, 1992; Coffin, 1995; Belshaw et al, 2010). However, despite this high mutation rate, the peptides that we have tested suggest that HCV peptides do not inhibit KIR+ NK cells very efficiently. In order to make this assumption we are also basing it on the fact that the majority of HLA-C binding peptides did not inhibit KIR+ NK cells in the recently published HIV study by Fadda et al (Fadda et al, 2012). Although our HCV peptide screen was considerably smaller, the economics of the situation remain the same. We found one inhibitory HCV peptide out of nine peptides tested in our biased approach and the HIV study found one inhibitory HIV peptide out of 211 peptides screened. The finding of a single inhibitory peptide (LLPRRGPRL) amongst our HCV peptides is therefore not substantially different from a chance occurrence, consistent with our thesis that HCV peptide epitopes are not selected on the basis of their ability to inhibit NK cells.The total contribution of HCV to the peptidome still remains unclear at present.
As well as the observation that most of the HCV peptides were non-inhibitory, we observed that one peptide derived from the NS3 region of HCV, LNPSVAATL (LNP), was able to reduce NK inhibition mediated by another inhibitory peptide (VAP-FA) in KIR2DL2 homozygous donors. This reduction in inhibition led to enhanced NK cell reactivity (more NK cell degranulation) in these donors. The peptide screen using LNP in mixes with the VAP-FA peptide showed a significant difference in NK degranulation in five KIR2DL2 homozygous
155 donors compared with six KIR2DL3 homozygous donors. In a mix containing equal concentrations (5 µM) of VAP-FA and LNP, there was increased NK degranulation when compared with VAP-FA alone. Importantly, this subtle difference was not observed in the six KIR2DL3 homozygous donors.
We also observed increased degranulation with the LNP and VAP-FA mix in the KIR2DL2/3 heterozygous group. Previous genetic analysis of the donors led us to also propose that LNP may be binding an activating receptor that KIR2DL2 homozygous donors possess but KIR2DL3 homozygous donors do not, namely the activating receptor KIR2DS2. We now propose that KIR2DS2 may be involved as the inhibition or antagonism we observed with any of these peptides was always specific for the CD158b+ subset of NK cells. We did not see effect of any peptides on the CD158b- subset. Moreover, the LNPSVAATL epitope also appears to be highly conserved both in the Irish HCV cohort sequences and between the six main HCV genotypes, which makes it a good candidate to warrant further investigation.
We also observed a significant reduction in NK cell inhibition with HCV6 (VLPCSFTTL) in all groups of donors tested. We did not investigate this further in this study as we mainly attributed it to the fact that VLPCSFTTL had an exceptionally high-affinity for HLA-Cw*0102 in the stabilisation assays and may outcompete VAP-FA for binding HLA-Cw*0102. Our hypothesis regarding KIR2DS2 and the potential agonistic effects observed with LNPSVAATL peptide will be explored further in the next chapter of this study.
156 Chapter 5. Results
An investigation into the peptide binding specificity of the activating KIR2DS2 receptor:A viral peptide with potential specificity for KIR2DS2
In the previous chapter the results with one of the HCV peptides NS31253-1262(LNPSVAATL) led us to propose that there was a difference in inhibition of NK cell degranulation observed between KIR2DL2 homozygous and KIR2DL3 homozygous donor groups. KIR genotyping in Chapter 3 had previously confirmed that the KIR2DL2 homozygous donors were also homozygous for the activating KIR2DS2 gene, an unsurprising finding considering the strong linkage between these two genes. However, the KIR2DL3 homozygous donor group did not possess any copies of KIR2DS2. It is known that possession of both KIR2DL2 and KIR2DS2 genes is allelic to KIR2DL3 (Uhrberg et al 2002). Due to the differences in genetics and observed functional differences, we proposed that the difference in NK inhibition observed in the KIR2DL2 donor group with HCV peptide 5 (HCV 5) (LNPSVAATL) might be associated with the activating KIR2DS2 receptor.
In the recent study carried out by Lena Fadda et al, a number of different KIR and HLA genotypes were investigated and overall no difference in peptide antagonism using the VAP- FA and VAP-DA peptides was reported. Overall, the peptide hierarchy was similar for KIR2DL2 homozygote, KIR2DL3 homozygote and KIR2DL2/3 heterozygote donors (Fadda et al, 2010). However, more intermediate KIR- binding peptides were found to bind KIR2DL2 than KIR2DL3.Structurally, the KIR2DS2 receptor is most identical to the KIR2DL2 receptor. However, KIR2DS2 differs from KIR2DL2 by means of one key single amino acid substitution of phenylalanine (F) to tyrosine (Y) at position 45. This single residue difference is enough to abrogate binding of KIR2DS2 to HLA-C1 (Saulquin et al, 2003; Vales-Gomez et al, 2008). To date, most of the previous binding studies for KIR2DS2-HLA-C1 have proven very difficult and the ligand for this receptor remains elusive. Although functionally KIR2DS2 does remain capable of transducing activating signals. Saulquin et al reported the free crystal structure of KIR2DS2 in 2003 and failed to show binding of KIR2DS2 to HLA- Cw*0304 tetramers (Saulquin et al, 2003). KIR2DS2-Fc failed to bind 721.221 cells transfected with HLA-Cw*0102, HLA-Cw*0304 or HLA-Cw*0702 alleles (Winter et al, 1998). Surface plasmon resonance analysis also failed to show any binding of KIR2DS2 to HLA- Cw*07 (Vales-Gomez et al, 1998).
One study reported weak KIR2DS2 binding to HLA-C1 through the use of different peptide derivatives. This study found weak but detectable binding of KIR2DS2 to HLA-Cw*0302 with certain peptides (Stewart et al, 2005). The peptide used in this study was GAVDPLLAL. This particular peptide was used as it has been endogenously isolated and sequenced from HLA-
157 Cw*0304 and was found to confer protection against NK cell lysis by NK clones in a previous study (Zappacosta et al, 1997). Biacore analysis carried out by Stewart et al, showed that mutants of the GAVDPLLAL peptide at p8 disrupted KIR/HLA binding whereas mutants at p7 contributed to affinity. HLA-Cw*03 was complexed with the peptide mutants of GAVDPLLAL at p8 (GAVDPLLSL and GAVDPLLYL, GAVDPLLVL, GAVDPLLKL). Affinity of KIR2DS2 tetramers for the GAV peptide mutants was as follows; GAVDPLLSL>GAVDPLLYL>GAVDPLLVL>GAVDPLLAL>GAVDPLLKL (in decreasing order of affinities).
Stewart et al also reported binding of KIR2DS2 tetramers to EBV-infected cells from HLA-C1 homozygous donors. However, this binding was not reported with EBV-infected cells from HLA-C2 homozygous donors or with primary B cells. Moesta et al found weak binding of KIR2DS2-Fc to a single bead bound HLA-C1 allotype (HLA-Cw*1601) (Moesta et al, 2010).
Prior to this, the best binding results carried out previous to this with an activating KIR had been with KIR2DS1. KIR2DS1 has been shown to bind HLA-C2 using many different methods. KIR2DS1 binding to HLA-C1 has been shown to be significantly lower than binding of its inhibitory counterpart KIR2DL1. More recently, in 2010 Moesta et al tested all seven HLA-C2 alleles and detected binding of KIR2DS1-Fc to bead bound HLA-Class I (Moesta et al, 2010). Also in 2009 Pende et al demonstrated binding of KIR2DS1-Fc to HLA Class I transfected 721.221 cells. The HLA-C2 ligand investigated in this binding study was HLA- Cw*06. The same group showed recognition and killing of HLA-C2/C2 leukemic blast cells via HLA-C*04/05 and HLA-Cw*04/06 (Pende et al, 2009). As for KIR2DS2 tetramer binding carried out by Stewart et al, similar results were published of KIR2DS1 tetramers binding to HLA-C2 alleles (HLA-Cw*04:01 and HLA-Cw*06:02) pulsed with specific peptides. Chewning et al found that there was induced cytotoxicity and IFN-γ release against HLA-C2- bearing target cells. However, the HLA Class I ligand was not identified (Chewning et al, 2007). Foley et al showed KIR2DS1-mediated killing of HLA-C2-bearing PHA blast cells. Killing only occurred with HLA-C1/C2 or HLA-C2/C2 PHA blasts (Foley et al, 2008).
In this body of work I investigated the potential binding of KIR2DS2 to HLA-Cw*0102 pulsed with certain viral peptides, specifically LNPSVAATL (HCV 5). As this thesis was being written, a structural study emerged that tested binding of KIR2DS2 to HLA-A*011. Binding was found between KIR2DS2 and HLA-A complexed with a vaccinia peptide (Liu et al, 2014). Although this somewhat takes away from the novelty of my findings we continue to propose that KIR2DS2 can also weakly bind to HLA-C1 alleles when complexed with certain peptides as originally reported by Stewart et al in 2005. Importantly, I have tried to obtain some functional data to support the hypothesis that a viral peptide could influence NK
158 reactivity via the KIR2DS2 receptor. To investigate this further in this chapter I firstly wished to examine the binding specificity of this HCV peptide (LNPSVAATL) to KIR2DS2 by generating KIR2DS2 tetramers followed by confocal microscopy to observe if there was any KIR2DS2 clustering in the presence of HLA-Cw*0102:LNPSVAATL complexes. Finally, I wanted to test the functionality of the KIR2DS2 receptor with the LNP peptide by generating KIR2DS2+ NK clones for use in further cell-based functional assays.
159 5.1 Generation of KIR2DS2 tetramers
5.1.1 Cloning strategy for the generation of KIR2DS2 extracellular domain constructs
The gene for the activating KIR2DS2 receptor is approximately 1,381bp in length (Homo sapiens natural killer cell inhibitory receptor (KIR2DS2) mRNA, KIR2DS2*00102 allele, complete cds (>gi|38305393 |gb|AY366243.1|) (See Appendix II). The translated protein is 304 amino acids (aa) in length. The KIR2DS2 receptor is comprised of two extracellular Immunoglobulin-like domains (nucleotide positions 74-725), which are responsible for the potential binding to HLA Class I molecules. We translated the extracellular domain of KIR2DS2 only, which gave a total aa sequence of 217 residues. We used specific primers to clone the extracellular domain of KIR2DS2 using template DNA obtained from a KIR2DS2 fusion construct (a kind gift from Professor Eric Long and Professor Christine Winter) (Fig 5.1 & 5.2). The extracellular domain of KIR2DS2 was linked by Glycopeptide optimised for protein expression to a Bir-A tag in the vector pET23d. The purpose of this tag is to allow enzymatic biotinylation of the protein in later stages by the biotin ligase enzyme (BirA). The forward primer contained a NcoI restriction site. The reverse primer used contained a BamHI restriction site, followed by the GP-linker to a BirA tag in pET23d (Fig 5.2). These primers were used to amplify the extracellular domain of KIR2DS2 using PCR, which was done in triplicate.
PCR products were double-digested with restriction enzymes NcoI and BamHI. The three digested PCR products were then visualised on 1.5% agarose gels at approximately 660 bp in size (Fig 5.3B). pET23d vector was also double digested with restriction enzymes NcoI and BamHI. pET23d was run on an 1.5% agarose gel post-digestion, visualised with ethidium bromide(Fig 5.3A). Following the double-digests of pET23d vector and KIR2DS2 DNA insert, ligation of the KIR2DS2 extracellular domain into the vector was performed at a ratio of 3:1 (insert:vector). Escherchia coli competent cells were transformed and any colonies resulting from the ligation and transformation were chosen for a final PCR screen (Fig 5.4A). To ensure the presence of KIR2DS2 DNA insert in the vector, a restriction digest was also performed on the colony DNA (Fig 5.4B). The KIR2DS2 construct in pET23d was confirmed by DNA sequencing.
160
Fig 5.1 Structure of the activating KIR2DS2 receptor and Cloning strategy
(A) The human KIR2DS2 receptor is approximately 1381 nucleotides in length (304 amino acid residues). It contains two extracellular Ig-like domains which comprise approximately 224 residues of the total protein. Illustration of KIR2DS2 receptor is not drawn to scale. (B) Cloning strategy for generation of KIR2DS2 construct in pET23d vector.
161
Fig 5.2 Cloning strategy with SnapGene software
(A) SnapGene software was used to check the cloning strategy for the KIR2DS2 construct into the pET23d vector. The final construct size resulting from ligation of the extracellular domain of KIR2DS2 (665 bp) into pET23d (3627 bp) was expected to be around 4292 bp in size. (B) Both restriction sites (BamHI and NcoI) positions are shown and the ligation resulting from the correct orientation of the insert into the vector.
162 A! Undigested pET23d ! Digested pET23d ! Undigested pET23d ! (~3663 bp)! (BamHI/NcoI) (~3627 bp)! (~3663 bp)!
Lane! 1! bl! 2! bl! 3! bl! 4!
bp! 10037% 4000% 3000% 1500% 1000% 800% 600% 400% 200%
B!
Lane! bl! 1! bl! 2! bl! 3!
bp! 10,037#
KIR2DS2 extracellular domain! 600# Digested with BamHI and NcoI! 400# 200# (~665bp)!
Fig 5.3 Restriction digests of pET23d vector and DNA encoding the extracellular domain of KIR2DS2
(A) 1.5% agarose gel displaying pET23d vector double-digested with restriction enzymesNcoI and BamHI. Digested vector is represented by bands migrating at approximately 3660 bp. Lane 2&3 = digested pET23d vector; Lane 1&4 = undigested pET23d vector. (B) 1.5% agarose gel displaying amplified PCR product of KIR2DS2 extracellular domain digested with NcoI and BamHI to allow for ligation into pET23d. Digested PCR products are represented by bands migrating at approximately 665bp. Lane 1-3 = Digested PCR products of KIR2DS2 extracellular domain. DNA was visualised using Ethidium Bromide.
163 A!
Lane! 1" 2" 3" 4" 5" 6" 7" 8" 9" 10" 11" 12" 13" 14" 15"
bp! 10037#
600# KIR2DS2 extracellular domain(~665bp)! 400# Post-ligation into pET23d! 200#
B!
Lane! bl! 1! bl! 2! bl! 4! bl! 5! bl! 7! bl! 13! bl! 14!
bp! 10,037! pET23d! 4000! vector (3627bp)!
800! 600! KIR2DS2 insert! 400! (665bp)! 200!
Fig 5.4 PCR screen and restriction digest showing successful cloning of KIR2DS2 extracellular domain into vector pET23d
(A) 1.5% agarose gel showing PCR screen of coloniespost-ligation of KIR2DS2 extracellular domain into pET23d vector. Colonies were picked and subjected to PCR analysis with specific primers for KIR2DS2. Amplified PCR products represented by bands migrating at approximately 665 bp. Lane 1, 2 4, 5, 7, 13 and 14 show presence of KIR2DS2 insert in pET23d vector. (B) 1.5% agarose gel showing the restriction digestion of plasmid DNA from colonies that grew. Restriction digest was performed with enzymes NcoI and BamHI which show presence of KIR2DS2 insert (band ~665 bp) and digested pET23d vector (band ~3660 bp). DNA was visualised using Ethidium Bromide.
164 5.1.2 Expression of KIR2DS2 Monomers
Construct DNA encoding the extracellular domain of KIR2DS2 in pET23d vector were transformed overnight into RosettaTM 2(DE3) competent cells. The following day five colonies were picked and were grown overnight in LB/Amp (100 mg/mL Ampicillin). The following day the bacteria were grown for a further three hours and protein expression was induced by 1 mM IPTG. Bacteria were harvested 3 hours post-induction with IPTG, centrifuged and sonicated for 20 min to obtain inclusion bodies. Inclusion bodies were homogenised and centrifuged. One of the colonies selected expressed ~100 mg of protein. Proteins were subject to to reducing SDS-PAGE gel analysis (12.5%) (Fig 5.5). The extracellular domain of KIR2DS2 can be seen on the SDS gel as a band migrating around ~25 kDa. Using an online tool called ProtParam, the expected molecular weight of the extracellular domain of KIR2DS2 was checked and it gave a result of ~24 kDa.
5.1.3 Refolding and purification of KIR2DS2 monomers
The inclusion bodies were solubilised in refolding buffer (Guanidine-HCL (6M)) to 10-12 mg/mL using a glass homogeniser. Inclusion bodies were left in refolding buffer for one hour at 4°C and then centrifuged for 45 min at 10,000 rpm. Supernatant containing solubilised protein was preserved after the spin and DTT added (20 mg/mL). The protein was left for one hour at 37°C. A simple overnight dilution method was used to re-fold 10 mg of KIR2DS2 protein (outlined in Chapter 2. Materials and Methods). After dilution the protein was concentrated and purified by FPLC (Fig 5.6). After purification the protein was concentrated further again to prepare for biotinylation.
165
Fig 5.5 Expression of protein encoding the KIR2DS2 extracellular domain in two colonies of Rosetta bacteria post-induction with IPTG
Reducing SDS PAGE analysis (12.5%) of induction experiment carried out with IPTG (1 mM) at 37°C. Lane 2 = 3-hour post-induction with one selected colony Lane 3 = 3-hour post-induction with another selected colony. Proteins were visualised with Coomassie blue stain. Expected molecular weight of the extracellular domain of KIR2DS2 was calculated using the online ProtParam tool which gave a result of ~24 kDa. Bands can be seen on the gel migrating at around ~25 kDa.
166 KIR2DS2 extracellular domain! ~24 kDa! !
167!
Elution volume (ml)!
Fig 5.6 FPLC plot to show purity of the refolded KIR2DS2 extracellular domain protein
The correctly refolded KIR2DS2 domain of the expected size was purified from by preparative grade and analytical size-exclusion chromatography using FPLC (Pharmacia, Sweden) with Hi Load 26/600 Superdex pg and Superdex GL 10/300 gel filtration columns (GE Healthcare, UK). Correctly refolded KIR2DS2 proteins were purified by gel filtration chromatography using fast protein liquid chromatography (FPLC).
167 5.1.4 Biotinylation and tetramerisation
The KIR2DS2 monomer was diluted to 1 mg/mL for optimal biotinylation. Biotinylation was performed overnight at 4°C (see Chapter 2. Materials and Methods). Biotinylated monomers were re-purified by FPLC and concentrated (Fig 5.7). A sandwich ELISA confirmed successful biotinylation of the KIR2DS2 monomers. Biotinylated monomers were resuspended at a final concentration of 1 mg/mL and PE-conjugated Streptavidin was added accordingly overnight at a molar ratio of 3:1 of to produce KIR2DS2 tetramers.
168 Biotin!
Biotinylated KIR2DS2! Monomers! !
Elution volume (ml)!
Fig 5.7 FPLC to assess refolded biotinylated KIR2DS2 proteins
Stability of the refolded biotinylated KIR2DS2 proteins was assessed during repurification by FPLC gel filtration following overnight biotinylation. Excess biotin was removed by further gel filtration chromatography.
169 5.1.5 Staining with KIR2DS2 tetramers
The KIR2DS2 tetramers were then tested for binding to 721.174 cells pulsed with 200 µM of each HCV peptide and endogenous peptides. .174 cells were pulsed with 200 µM of each HCV peptide and each endogenous VAP peptide for 16 hours at 26°C. The following day the cells were stained with KIR2DS2 tetramers that were pulsed with no peptide, endogenous peptide or HCV peptides. We observed increased staining with KIR2DS2 tetramers on cells pulsed with 200 µM of LNPSVAATL peptide and this was significant (*, p<0.05) when we compared the MFI to that of ‘No peptide’ control or any other endogenous or HCV peptide (Fig 5.8 & Fig 5.9).
We also synthesised further peptide derivatives of the VAPWNSLSL endogenous peptide containing mutations at positions 7 and 8 that were similar to the residues at position 7 and 8 of the LNPSVAATL HCV peptide. We synthesised three other derivatives VAPWNAAAL, VAPWNSATL, VAPWNAATL. .174 cells were pulsed with 200 µM of each of these derivatives for 16 hours at 26°C and were stained with the KIR2DS2 tetramer as outlined above. Fig 5.10 depicts KIR2DS2 tetramer staining with these derivatives. As can be seen in Fig 5.10, the strongest binding of tetramer was still observed in the presence of LNPSVAATL peptide. However, there was a subtle shift in tetramer staining to .174 cells pulsed with the VAPWNAATL and VAPWNSATL peptides in particular. In order of decreasing affinity KIR2DS2 tetramers bind LNPSVAATL> VAPWNAATL> VAPWNSATL> VAPWNAAAL> VAPWNSLSL peptides. Although the shifts observed with the VAP derivatives were not significant it highlights that the residues at P7 and P8 are key in mediating binding to KIR2DS2 tetramers.
170
Fig 5.8 KIR2DS2 tetramers can bind HLA-Cw*0102: LNPSVAATL complexes on the surface of .174 target cells
(A) Live .174 target cells were separated on their side-scatter (SSC-A) and forward-scatter characteristics (FSC-A). .174 cells were pulsed with (B) 200 µM of each endogenous peptide and derivatives and (C) 200 µM of each HCV peptide for 16 hours at 26°C. Staining was performed the following day with PE-conjugated human KIR2DS2 tetramers and cells were subject to Flow Cytometry analysis. MFIs displayed on representative overlay histograms indicate the MFI of the ‘no peptide’ negative control sample alongside the MFI of the samples with peptide present. No peptide MFI shown here is 230.67.
171 A B * * * 600 * * 500 *
400 400
300 MFI MFI 200 200 (KIR2DS2-PE) (KIR2DS2-PE) 100
0 0
No peptide VAPWNSLSLVAPWNSFAL LNPSVAATL VAPWNSDALVAPWNSRAL No peptide VLPCSFTTL LSPRPVSYL VAPWNSLSLLNPSVAATLAQPGYPWPLARPDYNPPLLSRHYKVFLRAYLNTPGLLLPRRGPRLLLPRRGPKL Peptides (200 µM) Peptides (200 µM)
Fig 5.9 KIR2DS2 tetramers can bind .174 target cells after pulsing with HCV5 (LNPSVAATL) peptide only
Staining of .174 cells pulsed with (A) 200 µM of endogenous peptides and (B) 200 µM of HCV peptides with PE-conjugated KIR2DS2 tetramers. Data shown in (A) are means ±SD and are representative of three independent experiments (n=3). Data in (B) are means ±SD and are representative of two independent experiments (n=2). (*, p<0.05) One-way ANOVA, Tukeys multiple comparison test.
172
Fig 5.10 KIR2DS2 tetramers can also mediate binding to VAP peptide derivatives depending on key residues at P7 and P8
(A) .174 target cells were separated on their side-scatter (SSC-A) and forward-scatter (FSC-A) characteristics and pulsed with (B) 200 µM of endogenous peptide derivatives and LNPSVAATL peptide (C) Bar chart displaying the MFI data. Staining was performed the following day with PE- conjugated human KIR2DS2 tetramers and cells were subject to Flow Cytometry analysis. MFIs displayed on histograms indicate the MFI of the ‘No peptide’ control sample alongside the MFI of the samples with peptide present. MFI of no peptide sample is 233.88. Data in (C) are means ±SD and are representative of four independent experiments (n=4). (*, p<0.05) One-way ANOVA, Tukeys multiple comparison test.
173 5.2 LNPSVAATL does not stabilise HLA-A2 molecules on .174 target cells
According to the work published in February 2014 by Liu et al which showed that KIR2DS2 is able to bind HLA-A*01:11 molecules in the presence of certain peptides we had already tested whether the LNPSVAATL peptide could potentially mediate binding to HLA-A2 molecules on the surface of target cells. The .174 cell-line expresses HLA-A*0201, HLA- Cw*0102 and HLA-B*5101 alleles (Andersen et al, 1999). Simply, .174 cells were pulsed with 100 µM of each peptide overnight and then stained with the HLA-A2-specific antibody BB7.2 and analysed by flow cytometry. The Influenza peptide GILGFVFTL (GILG) that mediates binding to HLA-A2 was used as a positive control for binding as shown in Fig 5.11, the MFI otbtained for GILG peptide demonstrates that it does stabilise HLA-A2 molecules. MFI data shown in Fig 5.11 indicate that LNPSVAATL does not stabilise HLA-A2 molecules on the surface of .174 cells.
174
Fig 5.11 LNPSVAATL peptide does not stabilise HLA-A2 molecules on .174 target cells
Peptides were incubated with 721.174 target cells for 16 hours at 26°C. Expression of HLA-A2 was determined by an HLA-A2-specific antibody conjugated to FITC (BB7.2) (A) Scatter Dotplot showing gating on live .174 cell population and histograms with MFIs for each condition. (B) Overlay histogram plots depicting the expression of HLA-A2 in the presence or absence of HCV 5 (LNPSVAATL) or a control peptide GILGFVFTL. All peptides were used at a concentration of 100 µM. Black line with no fill is the no peptide control. Black filled-in histograms on the overlays are with peptide LNP or GILG. One independent experiement was performed.
175 5.3 Confocal microscopy to detect clustering of KIR2DS2 receptor with LNPSVAATL: HLA-Cw*0102 peptide complexes
Confocal microscopy was used to examine the effect of peptides presented by HLA- Cw*0102 on the clustering of KIR2DS2 and KIR2DL2 receptors at the immunological synapse formed between effector and target cells. The transfected Ba/F3 cell lines were obtained as a kind gift from Professor Lewis Lanier and are a murine pre-B cell line. The Ba/F3 cell lines were already transfected with either human KIR2DS2 or human KIR2DL2 receptors. The transfected Ba/F3 cell line were used as effector cells and co-incubated with .174 cells that had been pulsed with 200 µM of peptide (VAPWNSFAL peptide or LNPSVAATL peptide). The target and effector cell complexes were stained by the monoclonal CD158b antibody conjugated to PE (clone GL183), which binds both KIR2DS2 and KIR2DL2 receptors. LNPSVAATL (LNP) peptide induced clustering on Ba/F3 KIR2DS2 transfectant cells stained with the GL183 antibody (****, p<0.0001) but did not induce significant clustering on KIR2DL2 Ba/F3 transfectant stained with GL183 antibody (Fig 5.12& 5.13). As a control peptide for KIR2DL2, VAPWNSFAL promoted significant clustering when co-incubated with KIR2DL2 transfectant cells (****, p<0.0001) but not with KIR2DS2 transfectant cells (Fig 5.12 & 5.13). For each condition at least 25-30 conjugates were analysed. The fold increase in fluorescence intensity at the interface between the Ba/F3 effector cells and .174 target cells was compared to a noncontact area of the Ba/F3 cell membrane. As can be seen in Fig 5.13B, the percentage of conjugates with aggregation of KIR2DS2 was higher in the target cells pulsed with LNP compared to target cell pulsed with VAP-FA. The percentage of conjugates with aggregation of KIR2DL2 was higher with the target cells that had been pulsed with VAP-FA peptide.
176
Fig 5.12 LNPSVAATL (HCV 5) induces clustering of KIR2DS2 at the interface between target and effector cells
Comparison of KIR2DS2 and KIR2DL2 clustering at the interface between Ba/F3 2DS2 and Ba/F3 2DL2 effector cells and .174 target cells, loaded with 200 µM of the indicated peptides. Unpulsed .174 cells (No peptide) were used as a control. After a 10-min co-incubation at 37°C the cells were fixed and analysed by confocal microscopy. White arrows indicate regions of observed clustering. All scale bars are 10 µm.
177 A ns 4 **** 3 **** ns
2
Fold change 1
0
2DS2 FA 2DL2 FA 2DS2 LNP 2DL2 LNP B No peptide
100
80
60
40
20 KIR2DL2 aggregation KIR2DL2 0 % conjugates with KIR2DS2 or
2DS2 FA 2DL2 FA 2DS2 LNP 2DL2 LNP No peptide
Fig 5.13 LNP induces aggregation of KIR2DS2 at the interface of effector and target cells
(A) The fold-increase in fluorescence intensity at the interface between Ba/F3 cells and .174 target cells was compared with a noncontact area of the Ba/F3 plasma membrane. The data was plotted as a fold-increase in fluorescence intensity. 25-30 conjugates were analysed per condition and data is means ±SD. One-way ANOVA was used to determine significance (Dunnetts multiple comparison test) ****, p<0.0001. (B) The percentage of conjugates with aggregation of KIR2DS2 or KIR2DL2 at the effector:target cell interface in each sample was analysed. Conjugates with a fold-increase greater than 1.2 were used to determine aggregation at the interface. 25-30 conjugates were analysed per condition.
178 5.4 Staining NK cells with 1F12 (KIR2DS2/KIR2DL3-specific) monoclonal antibody
Dr Christelle Retieres research group recently devloped a monoclonal antibody termed ‘1F12’ that was found to have specificity for the KIR2DS2 and KIR2DL3 receptors only (David et al, 2009). David et al found that 1F12 did not stain KIR2DL2 receptors. 1F12 was used to stain some of the KIR2DL2 and KIR2DL3 homozygous donors used in this study to observe if we got any differential staining. In order to separate the KIR2DS2+ cell population from the KIR2DL2+ cell population, it was proposed that staining KIR2DL2 homozygous donors with the CD158b monoclonal antibody and the 1F12 antibody would do so. CD158b binds all three receptors, KIR2DL2, KIR2DL3 and KIR2DS2. KIR genotyping had shown that all of our KIR2DL2 homozygous donors were also homozygous for KIR2DS2. KIR2DL3 homozygous donors were also stained with the CD158b antibody and the 1F12 antibody. Thus, by gating on CD158b+ and 1F12+ NK cells it could allow us to identify and separate the KIR2DS2+ cell population in KIR2DL2 homozyous donors.
The 1F12 antibody conjugted to FITC was obtained as a kind gift from Dr Christelle Retiere. A number of donors were screened using the 1F12 antibody and did not achieve the same results as reported by David et al in 2009. The main issue was with the results in the control group, the KIR2DL3 homozygous donors, which failed to give an expected double-positive 1F12+ CD158b+ population (Fig 5.14B). Instead, either single positive populations CD158b+ or 1F12+ population of NK cells were observed. A double-positive population with these donors was expected as both antibodies should stain positive for KIR2DL3. It was not known at the allelic level what KIR2DL3 alleles these donors expressed as that may have influenced this staining. It was also reported by Retiere et al that the 1F12 antibody has crossreactivity with some KIR2DL1 alleles. The KIR2DL3 donors that were stained did have copies of KIR2DL1 but we did not have the in-depth KIR allelic typing results so they may well possess alleles of KIR2DL1 that bind 1F12.
As we screened only a small number of the specific donors and did not obtain positive results, I did not pursue staining with the 1F12 antibody any further. However, a cell line transfected with the the KIR2DL2, KIR2DS2, KIR2DL3 receptors Ba/F3 (a gift from Professor Lewis Lanier) was also stained with the 1F12 antibody and we obtained positve results (Fig 5.15). The 1F12 antibody stained Ba/F3 cells transfected with KIR2DL3 and KIR2DS2 but did not stain Ba/F3 cells transfected with KIR2DL2 as reported (Fig 5.15).
179 A
B
Fig 5.14 Staining KIR2DL2 and KIR2DL3 homozygous donors with monoclonal 1F12 and CD158b antibodies
PBMC were separated on side-scatter (SSC-A) and forward-scatter (FSC-A) characterisitcs followed by staining with the following antibodies; CD3-PerCP, CD56-PeCy7, CD158b-PE and 1F12-FITC. (A) shows unstained and stained cells from a KIR2DL2 homozygous donor (B) shows unstained and stained cells from a KIR2DL3 homozygous donor.
180
Fig 5.15 Staining untransfected and transfected Ba/F3 cells with 1F12 monoclonal antibody
Ba/F3 cells (untranfected and transfected with either human KIR2DS2, KIR2DL3 or KIR2DL2) were separated according to their side-scatter (SSC-A) and forward-scatter (FSC- A) characteristics using flow cytometry analysis. The Ba/F3 cells were stained with 1F12- FITC antibody and gated. Both unstained and stained Ba/F3 cell populations are shown in the representative histograms.
181 5.5 NK cell cloning
5.5.1 NK cell clones show a trend towards increased cytotoxicity in response to target cells pulsed with LNP peptide
NK cells (CD3-, CD56+, CD158b+) were single cell sorted from PBMC from a KIR2DL2+/S2+/L3- donor. NK cells were sorted into NK clone media containing 5% heat- inactivated human serum. A cocktail of cytokines was also added to the starting clone media ((IL-12 (10 ng/mL), IL-15 (20 ng/mL) and IL-18 (100 ng/mL)). Feeder cells, PBMC (1 x106/mL) from three other allogeneic donors were irradiated and added to the NK cells with 2.5 µg/mL PHA. IL-2 was added to the cells 24 hours post-sorting. Clones were plated according to the limiting dilution method of Cella and Colonna (2000) (see Chapter 2. Materials and Methods). At first, 33 clones in total survived the sorting and plating and could be observed after one week as those wells in which the media had changed from a pink colour to yellow/orange and there was visible cell proliferation. All cytotoxicity and RTqPCR assays on the clones were always performed at least 5-7 days after the addition of the irradiated feeder PBMC.
The clones were stained with an antibody that is specific for both KIR2DL2 and KIR2DS2 receptors (CD158b antibody conjugated to PE (clone GL183)). CD158b staining for clone 4, clone 6, clone 9 is depicted in Fig 5.16. ~80-95% of each clone population stained positively with the CD158b antibody. Clones stained negative for both CD3 and CD158a (antibody that binds both KIR2DL1 and KIR2DS1). Clones also stained positive for the NK cell CD56 marker as expected.
For the cytotoxicity assay, .174 target cells were pulsed overnight with 200 µM of peptide and on the following day stained with a cell tracker orange dye (CTO) for 1 hour. They were then co-incubated with each of the clones (at an E:T ratio of 10:1) at 26°C for a further 4 hours. Cells were then stained with a Livedead fixable stain and analysed by Flow Cytometry. The % specific cytotoxicity for each of the clones was calculated by subtracting the % cytotoxicity (% of dead .174 cells) with no effector cells present (negative control) from the % cytotoxicity with effector cells present (test).
% Specific Cytotoxicity = % Dead .174 cells in test wells - % Dead .174 cells in control well
The gating strategy for the CTO assay is outlined in Fig 5.17. Most of the.174 target cells were identified with the CTO dye and firstly gated on. From the gated .174 CTO-stained population, cells were separated on their live-dead characteristics using a fixable live-dead stain. Most of the clones were able to kill the .174 cells alone apart from a few which didn’t
182 mediate any cytotoxicity in initial assays. However, only 4 out of the 33 clones gave significant changes in %specific cytotoxicity when co-incubated with the .174 target cells plus peptides (Fig 5.18 & Fig 5.19). There was a greater percentage of specific cytotoxicity observed with a number of clones in response to target cells pulsed with the LNPSVAATL peptide which was only observed in clone 4, clone 6 and clone 9 (Fig 5.19 & Fig 5.20). However, the % specific cytotoxicity was decreased in clone 6 and clone 27 when they were co-incubated with target cells that had been pulsed with the VAP-FA peptide. The overall % specific cytotoxicity for the clones is depicted in Fig 5.19.
183
Fig 5.16 Clones used in cytoxicity assays stained CD158b+ (positive for KIR2DL2/KIR2DS2)
NK cell clones (A) Clone 4 (B) Clone 6 and (C) Clone 9 were identified on their forward (FSC-A) and side scatter (SSC-A) characteristics and then separated by firstly staining with a fixable Livedead marker. The live population of clones were gated on and separated on their CD158b expression using a CD158b antibody (clone GL183) conjugated to PE. Clones were also stained with CD3, CD56 and CD158a antibodies and expression analysed. No fill overlays in histogram plots are unstained cells. Black overlays are stained cells.
184
Fig 5.17 Flow cytometry gating strategy for CTO assay
.174 target cells were separated on their side-scatter (SSC-A) and forward-scatter (FSC-A) characteristics. They were stained with CTO marker for one hour .174 cells that stained positively with CTO were gated on and then separated into live and dead populations using the fixable Livedead marker. Gate P2 represents the dead .174 target cells. Representative scatter dotplots show the % cytotoxicity (% of dead .174 cells) (P2) (19%) without any effector cells present. The dead .174 cells alone were the negative control for any non-specific cytotoxicity. The % of .174 cells that took up the Livedead stain were used as a control to calculate the %specific cytotoxicity.
185
Fig 5.18 Flow cytometry gating strategies for CTO assay with clones
.174 target cells were pulsed with 200 µM of LNP/VAP-FA peptides or no peptide and were stained with the CTO marker for 1 hour. The .174 cells that stained positively with CTO were gated on and then separated using the fixable Livedead stain. The % of cells that took up the Livedead stain were used to calculate %cytotoxicity and %specific cytotoxicity. .174 cells pulsed with peptide(s) or no peptide were co-incubated with effector cells (clone 4, clone 6, clone 9 and clone 27) (E:T ratio 10:1) for 4 hours at 26°C and subject to the same gating strategy. Scatter dot plots for each clone show the % cytotoxicity in Gate P2 (% of dead .174 cells).
186 No peptide A LNPSVAATL VAP-FA 60
40
20 % Specific cytotoxicity 0
Clone 1 Clone 2 Clone 3 Clone 4 Clone 5 Clone 6 Clone 7 Clone 8 Clone 9 Clone 10Clone 11Clone 12Clone 13Clone 14Clone 15Clone 16
No peptide B LNPSVAATL VAP-FA 60
40
20 % Specific cytotoxicity 0
Clone 17Clone 18Clone 19Clone 20Clone 21Clone 22Clone 23Clone 24Clone 25Clone 26Clone 27Clone 28Clone 29Clone 30Clone 31Clone 32Clone 33
Fig 5.19 Screen of clones for cytotoxicity against peptide-loaded target cells
.174 target cells were pulsed with 200 µM of LNP/VAP-FA peptides or no peptide. .174 cells were then incubated with clones 1-33 for 4 hours and analysed by Flow Cytometry. (A) Clones 1-16. (B) Clones 17-33. Bar-chart for the clones shows the % specific cytotoxicity for each clone at each condition. %Specific cytotoxicity was calculated by subtracting the %cytotoxicity with target cells alone from the %cytotoxcity with the target cells and clones (effector cells). One CTO screening experiment was performed with all the clones (n=1).
187
80 No peptide VAP-FA 60 * LNPSVAATL
40
20
% Specific cytotoxicity 0
Clone 4 Clone 6 Clone 9 Clone 27
Fig 5.20 Clones 4, 6 and 9 show a trend towards increased cytotoxicty in response .174 target cells pulsed with LNPSVAATL peptide
.174 target cells were pulsed with 200 µM of LNP/VAP-FA peptides or no peptide .174 cells were then incubated with clone 4, clone 6, clone 9 and clone 27 for 4 hours and analysed by Flow Cytometry. Bar-chart for the clones shows the % specific cytotoxicity for each clone at each condition. %Specific cytotoxicity was calculated by subtracting the %cytotoxicity with target cells alone from the %cytotoxcity with the target cells and clones. Means ±SEM are shown for two independent experiments (n=2). A two-way ANOVA was performed on the data using Tukeys multiple comparison test (*, p<0.05).
188 5.5.2 KIR2DS2+ NK clones can kill Huh7 replicon cells transfected with the HLA- Cw*0102 allele
Huh7 is a human hepatoma cell line that is permissive for HCV infection. Using a replicon system, a modified form of the HCV genome is able to replicate to high levels within this cell line. The stably transfected Huh7 replicon cell lines can be maintained after continuous selection with Geneticin (G418). The development of robust RNA replicon systems for HCV has significantly advanced the analysis of viral replication (Lohmann et al, 1999; Blight et al, 2000). TheHCV replicons replicate well in human Huh7 cells and contain the viral 5′ and 3′ untranslated regions.
The Huh7 cell line expresses HLA-A*11 on the cell surface but no HLA-C. I obtained some Huh7 target cells from Dr Connie Liu in our research group that had been transfected with the HLA-C alleleHLA-Cw*0102. In order to determine the cytotoxic effect of the NK cell clones on HCV replication, the HCV replicon JFH1ΔE1E2-luc had also been previously transfected into the Huh7 cell lines. I tested the NK clones for CD107a (LAMP1) expression using a CD107a degranulation assay with the Huh7 cells as target cells. The NK clones were tested only once with the Huh7 cells due to limitations of time and low clone cytotoxicity which occurred after 1-2 months of proliferation.
A blocking antibody (purified CD158b) was also used in the CD107a assay with the clones to observe if it could have any effect on degranulation after co-incubation with the target Huh7 cells transfected with HCV replicon. The blocking antibody to KIR2DL2/S2/L3 was added to the clones 1 hour prior to the addition of the target cells.
Both clone 6 (Fig 5.21) and clone 9 (Fig 5.22) showed increased CD107a expression in the presence of Huh7 target cells transfected with HLA-Cw*0102 compared to clone cells incubated without any target cells. Furthermore, the level of CD107a expression was increased in clone 6 and clone 9 when they were co-incubated with Huh7 cells transfected with both HLA-Cw*0102 and the HCV replicon. In the presence of a blocking antibody (purified CD158b), CD107a expression was slightly reduced in clone 6 in the presence of the Huh7 replicon target cells (Fig 5.21). CD107a expression was reduced in clone 9 in the presence of the blocking antibody and the Huh7 replicon target cells (Fig 5.22)
Conversely, clone 27 showed no reduction in CD107a expression in the presence of the blocking CD158b antibody. Clone 27 gave the same level of CD107a expression in response to Huh7 target cells transfected with HLA-Cw*0102 alone or transfected with both HLA- Cw*0102 and the HCV replicon.
189
Fig 5.21 CD107a assay on NK cell clone 6 (CD158b+) as effector cells and Huh7 cells transfected with HLA-Cw*0102 (with/without HCV replicon) as target cells.
An E:T ratio of 5:1 was used and cells were co-incubated for 4 hours at 37°C. A blocking unlabelled CD158b antibody was also added prior to the co-incubation. Gating strategy shows separation by forward scatter (FSC-A) and side scatter (SSC-A) characteristics followed by gating on live cells (stained with a fixable Livedead stain). The CD158b+ population was gated on and CD107a expression was analysed in histogram plots. Clones were also stained with anti-CD158a as a control.
190
Fig 5.22 CD107a assay with Clone 9 and Clone 27 and Huh7-Cw*0102 target cells (+/- HCV replicon)
CD107a assay using NK cell clone 9 and 27 (CD158b+) as effector cells and Huh7 cells transfected with HLA-Cw*0102 (with/without HCV replicon) as target cells. An E:T ratio of 5:1 was used and cells were co-incubated for 4 hours at 37°C. A blocking (unlabelled) CD158b antibody was also added prior to the co-incubation. Gating strategy shows separation by forward-scatter (FSC-A) and side-scatter characteristics (SSC-A) followed by gating on live cells (stained with the fixable livedead stain). The CD158b+ population was gated on and CD107a expression was analysed in histogram plots. Clones were also stained with anti-CD158a as a control.
191 5.5.3 qRT-PCR analysis to determine relative levels of KIR2DS2 and KIR2DL2 gene expression in NK clones
In order to quantitatively determine the levels of KIR2DS2 and KIR2DL2 gene expression in the NK cell clones reverse transcriptase (RT) quantitative real time (q) PCR was carried out. mRNA from each NK clone was obtained from homogenisation of the clone cells using the Trizol method of RNA extraction (see Chapter 2. Materials and Methods). Each clone was analysed by qRT-PCR and gene expression was normalised to that of the 18S ribosomal gene. Gene expression was determined using the ΔCt method. A donor genotyped by PCR- SSP as negative for both KIR2DL2 and KIR2DS2 genes was used as a negative control for KIR2DS2 and KIR2DL2 expression (Donor control). Levels of KIR2DS2 and KIR2DL2 gene expression were also measured in PBMC from an individual homozygous for both genes. The majority (20) out of the 33 clones were analysed at least once by qRT-PCR except for some clones which failed to give good quality RNA post-extraction.
Calculation for comparative method (ΔΔCT)
ΔΔCT= ((CT Clone KIR2DS2)- CT (Clone reference gene 18S))- ((CT (Donor control, KIR2DS2)- CT (Donor control reference gene 18S)).
Relative level of expression or fold change= 2-(ΔΔCT)
As shown in Fig 5.23, the KIR2DL2 and KIR2DS2 genes were readily detectable in most clones analysed. For the clones in which we originally observed differences in cytotoxicity of target cells pulsed with different peptides, namely clone 4, clone 6, clone 9 and clone 27, we also observed different levels of KIR2DS2 and KIR2DL2 mRNA expression. In particular, with clone 6 and clone 9 there were elevated levels of KIR2DS2 expression compared with KIR2DL2 expression (Fig 5.23). In clone 4 there were comparable levels of expression of both KIR2DS2 and KIR2DL2 genes. However, with clone 27 there was a higher level of expression of the KIR2DL2 gene (Fig 5.23). Taken together these results indicate there is indeed elevated expression of the KIR2DS2 gene in the clones that had mediated specific cytotoxicity against target .174 cells presenting the LNPSVAATL peptide (clone 4 and clone 6). With regard to clone 27 where there is an almost two-fold higher KIR2DL2 expression compared with KIR2DS2 expression we observed increased inhibition of cytotoxicity against target cells presenting the inhibitory VAP-FA peptide. As we have already documented that the VAP-FA peptide can strongly bind the KIR2DL2 receptor and induce inhibition of NK degranulation we can assume that this increased expression of KIR2DL2 is related to the increased inhibition when VAP-FA is present.
192
KIR2DL2 expression KIR2DS2 expression
32768
1024
32
1 Relative mRNA expression level
Clone Clone1 Clone2 Clone3 Clone4 Clone5 Clone6 9 Clone Clone10 Clone11 Clone12 Clone13 Clone15 Clone16 Clone17 Clone20 Clone23 Clone25 Clone27 Clone30 31 Clones
Fig 5.23 qRT-PCR analysis of KIR2DL2 and KIR2DS2 gene expression in 20 NK clones
Quantitative RT-PCR analysis of KIR2DL2 and KIR2DS2 expression in NK clones. Gene expression is normalised to the 18S ribosomal gene (ΔCt) and is presented as relative expression level rather than fold change. Values obtained for 2-(ΔΔCT) are depicted on a Log2 scale. Error bars shown are means ±SD values representative of two independent experiments. Clones used in the Cell Tracker Orange (CTO) cytotoxicity assays are highlighted by red boxes.
193 5.6 Discussion
The peptide binding specificity of the activating KIR2DS2 receptor has remained ambiguous to date. Here we report that a peptide derived from Hepatitis C virus is able to promote binding of KIR2DS2 to HLA-Cw*0102. We observed subtle differences in NK cell degranulation using this HCV peptide between KIR2DL2 homozygous donors all possessing KIR2DS2 and KIR2DL3 homozygous donors without KIR2DS2. As this difference was specific for the CD158b+ NK population in both groups one of the main differences between the two donor groups was presence or absence of KIR2DS2. If a peptide from HCV can facilitate binding to activating KIR2DS2 this may confer protection in the early stage of HCV infection. LNPSVAATL (HCV NS31254-1263) is derived from domain 1 of the NS3 helicase region in the HCV subtype 1b polyprotein. We found that the nonamer LNPSVAATL peptide is highly conserved within viruses from the Irish HCV cohort used in this project and also in all HCV genotypes (1b, 1a, 2, 3, 4, 5 and 6). There has been no reported binding or functional studies with the LNPSVAATL peptide.
Structurally, the extracellular region of KIR2DS2 is almost identical to the inhibitory KIR2DL2 receptor apart from one single amino acid substitution, which is enough to disrupt binding to MHC Class I. Past structural investigations have demonstrated that the key interface residues between KIR:MHC are the same for KIR2DS2 and KIR2DL2 (Leu104 and Tyr105). However, the side chain of Glutamic acid 71 (Gln71) in KIR2DS2 points away from MHC. In the presence of LNPSVAATL peptide, T8 may provide an essential hydrogen bond (H-bond) to Gln71 of KIR2DS2. As Threonine is a polar uncharged amino acid it may form a H-bond with Gln71 more readily. We proposed that LNPSVAATL is one of first viral peptides of known origin that can bind the activating KIR2DS2 receptor to date and subsequently affect the functionality of NK cells. At present, we do not know if only the residues at P7 and P8 contribute to KIR2DS2 binding although the potential hydrogen bond formed by threonine to Gln71 of KIR2DS2 could be an important interaction.
As this work was ongoing, a structural study was published in February 2014 by Liu et al that demonstrated binding of KIR2DS2 to HLA-A*01:11, a previously unknown ligand for KIR2DS2 (Liu et al, 2014). The study reports HLA-A*01:11 as the cognate ligand for KIR2DS2. The crystal structure of KIR2DS2 and HLA-A*01:11 was solved at 2.5A and showed that key KIR2DS2 residues Tyr45 and Asp72 were essential to mediate binding. However, Liu et al found that this binding could be modulated by changes in the peptide presented by MHC Class I, particularly at position 8 (P8) of peptide. Interestingly, Liu et al also reported something highly similar to what we have proposed in our study, Gln71 from KIR2DS2 appears to be particularly important when it comes to binding the peptide
194 presented by MHC. Liu et al show that the tip of KIR2DS2 (Gln 71 position) is able to interact with the P8 of peptide. They also showed that residues Threonine, Alanine, Serine, Valine and Tryptophan permitted binding of KIR2DS2 to HLA-A*01:11. However, replacing the residue at P8 with a charged residue such as aspartic acid, glutamic acid, lysine and arginine led to the loss of binding. We observed something similar with the LNPSVAATL peptide by creating mutants of the VAPWNSLSL peptide containing mainly alanine, serine or threonine residues at P7 and P8. These derivatives seemed to bind KIR2DS2 tetramers albeit to a lower affinity than the LNPSVAATL peptide.
Unlike the activating KIR receptors, the peptide binding specificity of the inhibitory KIR receptors has been well investigated to date by previous studies. Eric Long was the first to show that inhibitory KIR receptors are sensitive to the peptide bound by MHC Class I (Rajagopalan and Long, 1997). This was demonstrated first withthe KIR3DL1 receptor and its ligand HLA-B*2705. It was then found for KIR2DL1 (Malnati et al, 1995; Peruzzi et al, 1996; Rajagopalan et al 1997, Boyington et al, 2000). These findings were finally extended to KIR2DL2, KIR2DL3 and KIR3DL2 (Boyington et al, 2000; Maenaka et al, 1999; Hansasuta et al, 2004; Zappacosta et al, 1997; Mandelboim et al, 1997). The co-crystal structure of KIR2DL2 and HLA-Cw*03 with the endogenous peptide GAVDPLLAL showed that residues P7 and P8 are essential for KIR binding (Boyington, 2000). It was shown with the KIR3DL1 and HLA-B*5701 pair that P8 of peptide LSSPVTKSF makes contact with residue 166 of KIR3DL1 (Vivian et al, 2011). Recently, the crystal structure of KIR2DL1 with HLA-Cw*04 shows that direct contact between KIR and MHC:peptide was not observed. Fan et al previously reported that P8 is solvent accessible and changes to P8 lead to alterations in NK function (Fan et al, 2001).
If LNPSVAATL is a peptide that could be presented during HCV infection on hepatocytes to KIR+ NK cells we need to ask numerous important questions such as: How likely is this peptide to be presented on MHC Class I during HCV infection? How much would be expressed quantitatively? Would this imply that expression of KIR2DS2 on our NK cells gives us some advantage in infection if this peptide was indeed presented? We also need to question whether presentation of LNP would be limited to HLA-Cw*0102 molecules. Would it exhibit broader specificity to other HLA-C1 allotypes or perhaps even the newly described ligand HLA-A*1011? It is known for peptide binding to HLA-Cw*0102 that the key peptide binding residues are at position 3 and position 9. In this case it is Proline at P3 and Leucine at P9 (motif xxPxxxxxL). Therefore, this peptide may not bind HLA-A*1011 as well as HLA- Cw*0102. However, the peptide-binding motif of HLA-Cw*0102 is similar to that of HLA- Cw*0304 (Barber et al, 1996). HLA-Cw*0304 also has the same anchor residues at P2 and P9. Therefore, LNPSVAATL may bind to numerous HLA-C1 alleles and it may not be solely
195 limited to HLA-Cw*0102. A very recent study by David et al demonstrated that KIR2DS2+ KIR2DL2- NK cell clones were strongly activated (showed increased degranulation) in response to the 721.221 cell-line expressing the HLA-C1 allele HLA-Cw*0304 but not in response to the cell-line expressing the HLA-C2 allele HLA-Cw*04:01 (David et al, 2013). We may assume that if KIR2DS2+ NK clones were activated in response to cells expressing the HLA-Cw*0304 they should also be activated in response to cells expressing HLA- Cw*0102 which is what we have observed here with the .174 cell line presenting LNP peptide.
Confocal data shows the LNPSVAATL peptide inducing clustering of the KIR2DS2 receptor on the surface of transfected Ba/F3 cells. We expect that through these induced clusters there would be increased activating signals and signalling interactions via KIR2DS2 to activate the NK cell. Other data from our group suggests that this is the case. We could also hypothesise that these activating signals could outweigh the inhibitory signals mediated by the VAP-FA peptide. Importantly, LNPSVAATL acts in a different manner to the VAP-DA peptide in inducing activation as we know it can mediate binding to KIR2DS2. Therefore, we assume that the peptide antagonism already described with VAP-DA acts in a different manner yet the end-result is the same which is increased NK cell activation/degranulation (Fadda et al, 2010; Borhis et al, 2012).
As KIR receptors are indeed peptide selective, would changing the peptide repertoire alter NK cell reactivity in viral infection? We do not know if and how many copies of LNPSVAATL may be presented in HCV infection. In measles infection, the HLA-A*0201 epitope KLWESPQEI has been suggested to be as abundant as 5 x 104 copies per cell (Van Elset al, 2000). Quantitation of these peptides is absolutely essential to determine what proportion may be viral or self-derived in infection. As the HCV RNA-dependent RNA polymerase has a high mutation rate (lack of proof-reading capacity compared to human) during transcription the likelihood that there would be many mutations in viral proteins is high. It has been estimated that the error rates of the HCV RNA-dependent RNA polymerase may be as high as one per 1000 nucleotides (Powdrill, 2011). The HCV genome is 9.3 kb so there is substantial probability of mutation. Therefore, this would favour viral escape mutation and also lead to synthesis of DRIPs.
We also do not know at present the exact ratios of host peptide: viral peptide that may be presented on MHC Class I molecules (Croft et al, 2013). It is likely that viruses, such as HCV, can alter the host peptide repertoire or ‘peptidome’. This was highlighted to us in the study by Hickman et al where altered self peptides were found on HIV-infected cells and not found on uninfected cells (Hickman et al, 2003). Viral epitopes for CTLs have been shown to
196 be from recently synthesised proteins rather than mature peptides. This confirms potential for this mechanism for viruses altering host peptide repertoire (Khan, 2001). However, the generation of peptide repertoire is complex process that is not clear at present and is something that requires much future work.
Finally, in terms of clinical relevance, if the LNPSVAATL peptide derived from HCV was indeed processed and presented on infected hepatocytes or other antigen-presenting cells during Hepatitis C infection this may only confer advantage on individuals possessing a KIR genotype containing KIR2DS2. In particular, individuals with a KIR haplotype B. Group B Haplotypes are defined by the possession of at least one of the following genes: KIR2DL5, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS5 and KIR3DS1 (Martin et al, 2000; Trowsdale et al, 2001; Khakoo and Carrington, 2006). The Group A haplotypes do not possess as many activating KIR receptors. If the LNP peptide is indeed specific for engaging KIR2DS2 on NK cells from individuals with haplotype B it may aid activation of NK cells. We already know there has been a protective association shown between the possession of KIR2DL3 and HLA-C1 as this is a weaker inhibitory interaction (Khakoo et al 2004; Moesta et al, 2008). As mentioned previously, those individuals with the KIR2DS2 gene nearly always inherit the stronger inhibitory gene KIR2DL2 as there is strong linkage disequilibrium between these two genes. It is known that possession of both KIR2DL2 and KIR2DS2 is allelic to KIR2DL3 (Uhrberg et al, 2002). Perhaps those individuals possessing a haplotype B have evolutionarily adapted to an alternative mechanism of NK cell activation, which simply focuses more on engagement of their activating KIR receptors.
197 Chapter 6. Discussion
Peptides form a crucial part of the stability of KIR:MHC interactions and without them there would be no interaction between MHC and KIR and no subsequent NK cell inhibition or activation (Wagtmann et al, 1995; Peruzzi et al, 1996; Boyington et al, 2001). I would like to therefore discuss the impact of peptides on the inhibitory KIR2DL2/3 and activating KIR2DS2 interactions and the results that I have found with regard to how both endogenous and HCV-derived peptides can modulate these interactions.
6.1 The peptide binding specificity of the inhibitory KIR receptors
The peptide specificity of the inhibitory KIR has been been documented by many previous studies. Peptide specificity has been observed for KIR2DL1 (Peruzzi et al, 1996; Rajagopalan and Long, 1997), KIR2DL2 and KIR2DL3 (Zappacosta et al, 1997; Boyington et al, 2000), KIR3DL1 (Malnati et al, 1995; Peruzzi et al, 1996) and finally KIR3DL2 (Hansasuta et al, 2004). The co-crystal structure of KIR2DL2 and HLA-Cw*03 with the endogenous peptide GAVDPLLAL showed that residues P7 and P8 are essential for KIR binding (Boyington et al, 2001). Fan et al previously reported that P8 is solvent accessible and changes to P8 lead to alterations in NK function (Fan et al, 2001). Until now, Fadda et al have produced the most recent data regarding the effect of different endogenous peptide derivatives (VAPWNSLSL) on the inhibitory KIR2DL2/3 receptors (Fadda et al, 2010).
6.2 The peptide binding specificity of activating KIR receptors
The peptide binding specificity for activating KIR has been only documented by a handful of studies to date, mainly for KIR2DS1/KIR2DS2 (Stewart et al, 2005) and KIR2DS2 again (Liu et al, 2014). Vales-Gomez et al demonstrated that KIR2DL2 and KIR2DS2, although of high sequence similarity, bound HLA-Cw*07 with strikingly different affinities (Vales-Gomez et al, 1998). Stewart et al investigated the binding affinity of KIR2DS1 and KIR2DS2 to the endogenous GAVDPLLAL peptide and its derivatives. In 2014, Liu et al investigated the binding of KIR2DS2 to one vaccinia viral (VACV) peptide (MLYSMWGK) and derivatives. Liu et al reported HLA-A*01:11 as the cognate ligand for KIR2DS2 and found that this binding could be modulated by changes in the peptide presented by MHC particularly at P8 of peptide. Gln71 of KIR2DS2 was found to be essential for peptide-KIR interaction in this study. The affinity for the derivatives binding to KIR2DS2 tetramers was MLIYSMWAK> MLIYSMWWK> MLIYSMWSK> MLIYSMWGK> MLIYSMWVK> MLIYSMWTK in decreasing order of affinities (Liu et al, 2014). Any charged residues at P8 such as Aspartic acid (D), Glutamic acid (E), Arginine (R) or Lysine (K) were found to abrogate binding completely.
198 6.3 Three models to activate NK cells using peptides (two have been studied in this project)
6.3.1 Antagonistic peptides for inhibitory KIR
Fadda et al first proposed the phenomenon of peptide antagonism with inhibitory KIR in 2010 for both KIR2DL2 and KIR2DL3 inhibitory receptors. Our research group have since defined four broad classes of peptide that can bind inhibitory KIR: strongly inhibitory peptides, weak inhibitory peptides and antagonistic/null peptides. Null peptides do not mediate any effect on NK cell signalling. Borhis et al went on to show that the strongly inhibitory peptide (VAP-FA) can induce tight clusters of KIR2DL3 receptors whereas the antagonistic VAP-DA peptide induces diffuse clustering of KIR2DL3 and disrupts the inhibitory signal (Borhis et al, 2012). Borhis et al generated mutations in the ITIMs of KIR2DL3 and this abrogated the antagonism observed with VAP-DA. Therefore, it was proposed that the VAP-DA peptide affects SHP-1 recruitment and this uncouples early inhibitory signalling.
The mechanism behind the disruption of the inhibitory signal by VAP-DA is ITIM-dependent and may involve recruitment of SHP-1 molecules to VAP-DA:MHC complexes (Borhis et al, 2012). Borhis et al proposed that VAP-DA:MHC complexes may sequester the SHP-1 molecules needed to transduce the downstream inhibitory signal. Alternatively, these antagonistic peptide:MHC complexes may deliver another unknown signal that counteracts the traditional ITIM-based inhibitory signal. The precise molecular mechanism for peptide antagonism is still unknown but it operates at the early stages of inhibitory synapse formation.
Interestingly, Borhis et al also reported that the diffuse clustering observed in the DA/FA mix is comparable to the diffuse clustering observed with VAP-DA peptide alone (Borhis et al, 2013). Borhis et al did not investigate the KIR2DL2 receptor or the VAP-RA peptide in the scope of this work. Fig 6.1 illustrates the effect of the VAP-FA peptide and the antagonistic effect of the DA/FA mix on both KIR2DL3+ and KIR2DL2+ NK cells. As the effects of the VAP-FA peptide alone and the DA/FA mix were highly similar for both groups of donors in this study and in the study by Fadda et al, (2010) we hypothesise that there is no significant difference in the manner that the antagonism of the DA/FA mix behaves in both these donor groups. We can propose that VAP-FA can induce tight clusters of both KIR2DL3 and KIR2DL2 and the DA/FA mix induces more diffuse clustering (Fig 6.1).
199
A!
="VAP&FA"(strong"inhibitory"pep5de)" Induces"5ght"clustering"of"inhibitory"KIR"
Strong ! Strong ! Inhibition! Inhibition!
KIR2DL2! KIR2DL3!
SHP1! SHP1!
ITIMS! ITIMS!
KIR! KIR! Inhibitory synapse! Tight clustering of KIR2DL3! (Borhis et al, 2012)!
MHC:FA peptide! MHC:FA peptide!
B! ="VAP&FA"(Strong"inhibitory"pep5de)" Antagonism (VAP-DA and VAP-FA)! ="VAP&DA"(Antagonis5c"pep5de)"
No inhibitory signal! Other ITIM-mediated signal?!
SHP1! SHP-1 sequestered! No SHP-1 recruitment! by DA?! ITIMS!
KIR!
Inhibitory synapse! Diffuse clustering of KIR2DL3 (Borhis et al, 2012)!
MHC:Peptide!
Fig 6.1 The effects of VAP-FA peptide alone (A) and antagonism of DA/FA mix (B) on KIR2DL3+ and KIR2DL2+ NK cells
200 In this study we chose peptides representative of the three different classes of peptide and compared how KIR2DL3 homozygous or KIR2DL2 homozygous donors responded to more complex peptide mixes in an effort to represent the peptide repertoire of a cell. Results from chapter one have suggested that KIR2DL3 homozygous donors may respond better than KIR2DL2 homozygous donors to larger shifts in the self-peptide repertoire of a target cell. In particular the presence of the weak inhibitory peptide VAP-RA and the antagonistic peptide VAP-DA contributed the most discriminating differences between the two donor groups. Quantitatively speaking, in most of the complex mixes used, NK cells from KIR2DL3 donors degranulated more rapidly in response to an almost two-fold lower concentration of peptide (VAP-DA ~1 µM) than the NK cells from KIR2DL2 donors (~2 µM VAP-DA). We do not yet know how the weakly inhibitory VAP-RA peptide behaves in terms of KIR clustering and SHP-1 recruitment. As VAP-RA was the key peptide in determining the subtle differences between the KIR2DL2 and KIR2DL3 donor groups in the complex peptide mixes we can propose that high concentrations of VAP-RA may alter the peptide repertoire to a greater extent for KIR2DL3+ NK cells thus generating more diffuse clusters and weakening the downstream inhibitory signal (as depicted in illustration Fig 6.2). This alteration in the peptide repertoire may not affect KIR2DL2+ NK cells to the same extent thus giving slightly stronger inhibition and less activation (degranulation).
201 Repertoire of three peptides ((>80%) VAP-RA )! KIR2DL2! KIR2DL3!
Overall Stronger inhibitory signal! Overall Weaker inhibitory signal!
SHP1! SHP1!
ITIMS! ITIMS!
KIR! KIR!
Less Diffuse Clustering! More Diffuse Clustering! With KIR2DL2?! With KIR2DL3?!
MHC:Peptide!
="VAP&FA"(strong"inhibitory"pep5de)"
="VAP&DA"(Antagonis5c"pep5de)"
="VAP&RA"(weak"inhibitory"pep5de)"
Fig 6.2 The effects of the peptide mixes containing three peptides on KIR2DL2+ and KIR2DL3+ NK cells
KIR2DL2 has been previously shown to be a stronger inhibitory receptor than KIR2DL3 due to differences in the hinge angle between D1 and D2 extracellular domains and can therefore engage its HLA-C1 ligand strongly (Winter et al, 1998; Moesta et al, 2008). The importance of KIR2DL3/HLA-C1 homozygosity has been highlighted previous as it shows correlation with spontaneous resolution in Hepatitis C infection (Khakoo et al, 2004). The explanation behind the protection conferred by KIR2DL3 homozygosity was that as KIR2DL3 doesn’t engage or bind HLA-C1 as strongly as KIR2DL2 it may be more easily overcome by other stronger activating signals. This may have something to do with the KIR2DL3/C1 being a weaker interaction and activating signals may dominate. As the reported differences in KIR2DL2 and KIR2DL3 are not found in the region that binds HLA or peptide, it has been proposed that the allelic differences may not affect HLA:peptide binding directly and instead they may have an indirect impact on KIR clustering and/or KIR expression.
KIR2DL3 homozygosity has also been found important for fulminant malaria (Hirayasu et al, 2012). Interestingly in HIV-infected individuals who possess the stronger inhibitory receptor KIR2DL2, it appears that HIV-positive CD4+ T cells can enhance the binding of KIR2DL2+ NK cells. Alter and colleagues have also proposed that altered or viral HIV peptides can manipulate the stronger HLA-C:KIR2DL2 interaction and promote NK cell inhibition instead of NK activation (Alter et al, 2011). Under increased immune pressure, HIV can select
202 sequence polymorphisms that evade the immune responses by deliberately promoting interaction with KIR2DL2+ NK cells. We therefore now propose that another simple alternative route to NK cell activation may indeed be determined by shifts in the peptide repertoire on a target cell. If KIR2DL3 can act as a more flexible sensor to these subtle shifts that may be triggered by viral infection then it would have the upper hand over KIR2DL2.
The tuning/rheostat model of NK cell licensing may play a part with regard to how an NK cell is actually activated in this manner. Activation and inhibition can be quantitated on a large scale and peptides may have an additional role in this fine-tuning. It can be viewed from both angles, antagonistic peptides may be ‘better adapted’ for KIR2DL3/C1 interaction or the KIR2DL3/C1 system is ‘better adapted’ for the antagonism to occur. The presence of more weakly inhibitory peptides or antagonistic peptides can disrupt a strong inhibitory peptide. This mechanism may prove effective for KIR A haplotype individuals, which lack numerous activating KIR, but may have increased sensitivity to sensing shifts in the peptide repertoire. The shifts may be better sensed by the weaker inhibitory receptor such as KIR2DL3. The limitations of the antagonism phenomenon are that it has so far only been defined for KIR2DL2/3-HLA-Cw*0102 pairing so it is essential to investigate whether this is found in other KIR2DL2/3-HLAC1 allele pairings let alone for other KIR receptors. Testing the models in different KIR receptor systems and also in-vivo is key as part of future work.
Overall, we propose that KIR2DL3+ NK cells are more sensitive to fluctuations in the surface peptide repertoire on a target cell. Inhibition of KIR2DL3+ NK cells may be more easily ‘antagonised’ by these small fluctuations than KIR2DL2+ NK cells. The phenomenon of peptide antagonism may also be easier to accomplish in the presence of weak inhibitory peptides as we have observed with VAP-RA.
The effect of peptides on other inhibitory KIR:HLA (KIR2DL1/KIR2DS1 and HLA-C2 interaction)
Most of the donors in this study were found to possess at least one copy of the strong inhibitory receptor KIR2DL1. KIR2DL1 mediates a very strong interaction with its ligand HLA- C2 (Winter and Long, 1997; Winter et al, 1998). However, the effect of all of the peptides we used was only apparent in the CD158b+ population rather than the CD158b- population so we assume that there was no major effect of these peptides on the KIR2DL1+ NK cell population. We did not observe inhibition or any enhanced activation in the KIR2DL1+ (CD158a+) population using any of the peptides (endogenous or HCV). The majority of the donors we used were also homozygous for HLA-C1 and did not possess HLA-C2. However,
203 there may be peptides that do engage this receptor system that have not been described as of yet. We do not know if antagonism is strictly specific for the KIR2DL2/3-C1 interaction, which is something that is required for future work.
6.3.2 Synergy of inhibition via the NKG2A/CD94 receptor
Another mechanism by which to influence NK cell reactivity is using synergistic inhibitory peptides through the NKG2A/CD94 receptor family. Synergistic inhibition of NK cells has not been covered in detail within the scope of this study but recent results published in our group by Cheent et al support synergy with numerous viral peptides derived from HCV, HIV and EBV as a viable mechanism to promote NK cell inhibition using the conserved CD94/NKG2A receptor system (Cheent et al, 2013). Cheent et al discovered that small amounts of leader peptides presented on HLA-E molecules could weakly inhibit NKG2A+ NK cells. This inhibition was greatly enhanced in the presence of viral peptides derived from HIV, HCV and EBV hence the term ‘synergy of inhibition’ was employed. If HCV is not as well adapted to polymorphic KIR2DL2/3 receptors this would provide a clear selective evolutionary advantage for the host. We have presently questioned whether HCV may have adapted better to the conserved NKG2/CD94 family of NK receptors and may have an inhibitory effect on NK cells via NKG2A/CD94-HLA-E interaction (Cassidy et al, 2014). We have recently proposed that the HLA-C specific KIR are specialised to detect changes in the peptide repertoire of a cell and the conserved NKG2/CD94 system is better adapted to sensing any downregulation of MHC Class I (Cassidy et al, 2014). The reasoning behind this is that KIR2DL2/3+ and NKG2A+ cells can respond with different stoichiometries to changes in the levels of cell surface MHC Class I (Cheent et al, 2013). In particular, NKG2A+ NK cells are more sensitive to changes in MHC Class I surface levels at low levels of MHC Class I than KIR2DL2/3+ NK cells are.
6.3.3 Peptides specific for activating KIR receptors (agonist peptide(s) for activating KIR)
Here we report that a peptide derived from Hepatitis C virus is able to augment binding of KIR2DS2 to HLA-Cw*0102. In this present study we report that a viral peptide derived from HCV (LNPSVAATL) could weakly engage KIR2DS2 tetramers and activate KIR2DS2+ NK cell clones. We observed increased NK cell degranulation in KIR2DL2+ KIR2DS2+ homozygous donors with the LNP peptide that we did not observe in KIR2DL2- KIR2DS2- KIR2DL3+ donors. We also observed increased clustering of KIR2DS2 receptors in the presence of this peptide. Thus, we propose that LNP can act as an agonistic peptide on NK cell activation via KIR2DS2. If a peptide from HCV can facilitate binding to KIR2DS2 this may
204 confer protection in the acute stages of HCV infection. LNPSVAATL (aa residues 1253- 1262) is derived from the NS3 region in the HCV subtype 1b polyprotein. We found that the nonamer LNPSVAATL peptide is relatively conserved in all HCV genotypes 1b, 1a, 2, 3, 4, 5 and 6. As far as the literature goes, there has been no reported binding or functional studies with the LNPSVAATL peptide as of yet. We proposed that LNPSVAATL is one of first viral peptides of known origin that can bind the activating KIR2DS2 receptor to date and subsequently affect the functionality of NK cells on which it is expressed. At present we do not know if only the residues at P7 and P8 contribute to KIR2DS2 binding although we propose that the potential hydrogen bond formed by threonine of LNPSVAATL to Q71 of KIR2DS2 could be a most important interaction.
At present it is not known of any other viral peptides that can mediate an interaction between the activating KIR2DS2 receptor and HLA besides one vaccinia-derived viral (VACV) peptide (MLYSMWGK) that was reported a few months ago by Liu et al as this investigation was also ongoing (Liu et al, 2014). Interestingly, Liu et al also reported something highly similar as to what we have proposed in our study, Q71 from KIR2DS2 appears to be particularly important when it comes to binding the peptide presented by MHC. Liu et al show that the tip of KIR2DS2 (Q71 position) is able to interact with the P8 of peptide. They also show that residues Threonine, Alanine, Serine, Valine and Tryptophan permitted binding of KIR2DS2 to HLA-A*11:01. However, replacing the residue at P8 with a charged residue such as Aspartic acid, Glutamic acid, Lysine and Arginine led to the loss of binding. In the presence of LNPSVAATL peptide, T8 may provide an essential hydrogen-bond (H-bond) to Q71 of KIR2DS2. As Threonine is a polar uncharged amino acid it may form a H-bond with Q71 more readily. We observed something similar with the LNPSVAATL peptide by creating mutants of the VAPWNSLSL peptide containing mainly Alanine, Serine or Threonine residues around P7 and P8. These derivatives seemed to bind KIR2DS2 tetramers albeit to a lower affinity than the LNPSVAATL peptide.
Liu et al report HLA-A*11:01 as the cognate ligand for KIR2DS2 (Liu et al, 2014). This may be the case but we propose that KIR2DS2 can also engage with HLA-Cw*01 molecules and this is highly peptide-dependent as we have shown for LNP. A recent study by David et al indicates that KIR2DS2 may also interact with HLA-Cw*03 (David et al, 2013). Given the broad ligand binding capababilities of some other activating KIR family members such as KIR2DS4, it would not be surprising if we observe something similar with KIR2DS2. KIR2DS4 can interact with numerous alleles (HLA-A*11:02, HLA-Cw*05:01 and HLA- Cw*16:01) (Graef et al, 2009). The Pro71-Val72 motif in KIR2DS4 confers reactivity with HLA-A*11 whereas KIR2DS2 possesses a Pro71-Asp72 motif (Liu et al, 2014).
205 The potential to activate NK cells using specific peptides that bind activating KIR receptors is another suggested mechanism for modulating NK cell activation. It is perhaps tantalising to suggest that it could be another key potential mechanism of activation during viral infection or tumorigenesis as the activating KIR may engage with specific HLA:peptide complexes comprising of viral peptide and/or altered self peptides. These peptides would not usually be expressed on healthy or uninfected cells. Therefore, the presence of these peptides could be another form of activating ligand for the activating KIR. It would be essential as part of future work to determine if any other viral peptides can mediate binding to KIR2DS2 or other activating KIR. Now that we are aware that certain residues at P7 and P8 are key to mediating the stability of KIR:HLA interactions we could focus on the screening of prominent viral epitopes containing these residues for binding.
Fig 6.3 The agonistic effect of a viral peptide on KIR2DS2
In 2013, a key study by David et al demonstrated that KIR2DS2+ KIR2DL2- NK cell clones were strongly activated (showed increased degranulation) in response to the 721.221 cell- line expressing the HLA-C1 allele HLA-Cw*0304 but not in response to the cell-line expressing the HLA-C2 allele HLA-Cw*04:01 (David et al, 2013). However, David et al found that when KIR2DL2 and KIR2DS2 were co-expressed on the NK cells, the inhibition mediated by KIR2DL2 dominated. Interestingly, the peptide-binding specificity of HLA- Cw*0304 is similar to the allele tested in this study HLA-Cw*0102 (Barber et al, 1996). We may assume that if KIR2DS2+ NK clones were activated in response to HLA-Cw*0304 cells they should also be activated in response to HLA-Cw*0102 cells which is what we have observed here with the .174 cell line presenting LNP peptide. The recognition of this viral peptide relies on presentation on MHC Class I and the interaction of activating KIR2DS2 with
206 HLA:peptide complexes. However, activating KIR may be able to also engage with non-MHC ligands. Katz et al have previously shown that activating KIR2DS4 may be able to bind a non-MHC tumour ligand, a protein derived from melanoma (Katz et al, 2004).
KIR3DS1 is the most well studied activating KIR member to date, which is summarised in a nice review by Korner and Altfield (Korner and Altfield, 2012). It has been mostly studied within the context of HIV/AIDS infection. An important study conducted by Martin et al showed that individuals possessing the KIR3DS1 gene have slower progression to AIDS. In- vitro work by Alter et al, showed that KIR3DS1+ KIR3DL1- NK cells taken from HLA- Bw*0480I+ individuals were able to suppress HIV replication in-vitro better than KIR3DS1- KIR3DL1+ and KIR3DS1- KIR3DL1- NK cells (Alter et al, 2007). Copy numbers of KIR3DS1 have also found to be important with regard to viral load set points. In HLA-Bw*04+ individuals, more copies of KIR3DS1 are associated with lower viral loads (Pelak et al, 2011).
There has been some conflicting previous evidence that the KIR B haplotypes consisting of activating members KIR2DS2, KIR2DS3 and KIR2DS5 have been associated with faster progression to AIDS (Gaudieri et al, 2005) and lower CD4+ T cell counts in a west African cohort of HIV-infected women (Jennes et al, 2011). However, if the HIV-infected women were also homozygous for HLA-C1 this decreased the association between lower CD4+ T cell counts and the activating KIR (Jennes et al, 2011). Dring et al reported that the frequency of KIR2DS3+ individuals is higher in chronic HCV patients than with resolvers of HCV (Dring et al, 2011). Interestingly the negative effect of KIR2DS3 was only observed in HLA-C2+ individuals, which does not act as a known ligand for KIR2DS3. The overall impression of these studies could suggest the importance of the presence of HLA-C1 rather than HLA-C2 as a ligand in terms of better resolution and prognosis in HIV and HCV infection. It is known that HLA-C is not down regulated on HIV-infected cells (Cohen et al, 1999). There is still no direct evidence that the activating KIR can directly recognise HIV or HCV infected cells via HLA-C1 or HLA-C2 ligands.
6.4 The immunopeptidome and self/viral peptides
Fluctuations in the self-peptide repertoire of a cell are commonly triggered by a viral infection or tumour transformation. Thus, if changes in the peptide repertoire of a cell can differentially activate KIR2DL2+ or KIR2LD3+ NK cells as shown here with the VAPWNSLSL derivatives, this provides another rationale for how abnormal or infected cells may be recognised by different KIR receptors. The peptides that are ultimately presented on MHC Class I
207 molecules in an infected cell come from complex processing and may be altered self- peptides or viral peptides. In 2004, Hickman et al sequenced over 200 endogenously-loaded HLA-B*1801 peptides in a human B cell line and showed that these peptides can be presented from a majority of the proteomic content of the cell (Hickman et al, 2004). Hickman et al reported that many of these peptides originated from such as binding proteins, catalytic proteins, and proteins that are involved in cell metabolism, growth, and maintenance. Hickman et al report that the range of peptides presented by MHC Class I molecules is relatively unbiased. Interestingly however, an abudance of RNA binding- proteins was found in the peptide repertoire of HLA-B*1801 (Hickman et al, 2004).
These peptides may originate from the degradation of mature proteins or alternatively DRiPS (Yewdell et al, 1996; Yewdell, 2002; Yewdell and Hill, 2002). The other major contributor to the peptide content of our cells is from aged cellular proteins or ‘retirees’ (Yewdell et al, 2001; Yewdell, 2011; Yewdell, 2013; Croft et al, 2013). An important consideration to take into account when investigating viral peptides for future studies is whether they naturally exist in infection and are cleaved and processed in the proteasome for antigen presentation. As the HCV RNA dependent RNA polymerase enzyme has a high mutation rate (lack of proof-reading capacity compared to human) during transcription of new proteins the likelihood that there would be many mutations in viral proteins is high. It has been estimated that the error rates of the HCV RNA dependent RNA polymerase estimated may be as high as one per 1000 nucleotide site (Powdrill, 2011). The HCV genome is 9.3 kb long so there is substantial probability of mutation. Therefore, this would favour viral escape mutation and also lead to synthesis of DRIPs.
Quantitation of the exact ratio of viral peptide to endogenous peptide that may be processed and presented on MHC Class I has remained difficult. Previous significant work by Van Els and Herberts et al found that a HLA-B*2705-restricted peptide derived from Measles virus had abundant expression in cells infected with the virus. Its level of expression in infected cells was up to ~1500 copies (MHC-peptide complexes) per cell (Herberts et al, 2001; Van Els et al, 2000). As KIR receptors are indeed peptide selective would changing the peptide repertoire change our NK cell reactivity in viral infection? We do not know if and how many copies of viral peptides such as LNPSVAATL may be presented in HCV infection. It was found that in Measles infection, the HLA-A*0201 epitope KLWESPQEI has been suggested to be as abundant as 5 x 104 copies per cell (Van Els et al, 2000). Quantitation of these peptides is absolutely essential to determine what proportion may be viral or self-derived in infection. A key recent study by Croft et al has shown in vaccina virus infection that epitope/antigen presentation is an overall highly dynamic process. The abundance of virus
208 epitopes varied per cell, as low as 10 copies per cell to as high as 3 x 104 copies per cell. Croft et al were able to quantify 8 VACV epitopes simultaneously across numerous time- points of infection. They monitored epitopes from viral proteins such as KSYNYMLL from vaccinia core protein and TSYKFESV, which was derived from an immunomodulator protein. The TSYKFESV epitope was dominant all throughout infection. Croft et al demonstrated that 6/8 viral epitopes appeared on infected cells as early as 30 min post-infection and this peaked around 3.5 hours post-infection. This contrasts with the previous HIV study conducted by Hickman et al in 2003 where HIV peptides were only detected at later time- points post-infection. Hickman et al observed that altered host peptides were particularly prevalent in the initial days post-HIV infection (day 1-day 3). However, after day 4 of HIV infection the viral proteins gp120 (envelope) and p24 were then detected. Herberts et al also found that altered self-peptides were upregulated in Measles virus infection (Herberts et al, 2003).
The relevance of this comes into play with our findings, which shows that these subtle shifts of peptide can differentially activate different subgroups of KIR+ NK cells. Thus, by merely changing the peptide content of MHC Class I the reactivity of KIR2DL2+ and KIR2DL3+ NK cells can be altered significantly. We have also shown that the ‘threshold’ or ‘set-point’ of activation for KIR2DL3+ NK cells is lower than that of KIR2DL2+ NK cells, which may imply that they are easier activated by shifts in the peptide repertoire. Future work would have to determine whether this is the case for viral peptides and potentially use MHC-Class I- restricted viral peptides or altered self-peptides that have already been eluted in binding studies. The use of altered self-peptides that may occur naturally during tumour transformation would also be ideal to determine if they can affect NK cell reactivity via KIR2DL2 and KIR2DL3. A peptide of interest could potentially be derived from heat-shock proteins or other stress proteins that are upregulated during times of cellular stress.
If LNPSVAATL is a peptide that could be presented during HCV infection on infected hepatocytes to our KIR+ NK cells we need to ask numerous important questions such as: How likely is this peptide to be presented during HCV infection? How much would be expressed quantitatively? Would this mean that possessing KIR2DS2 on our NK cells gives us some advantage in infection if this peptide was indeed presented. We also need to question whether presentation of LNP would be limited to HLA-Cw*0102 molecules. Would it exhibit broader specificity to other HLA-C1 allotypes or perhaps even the newly described ligand HLA-A*1011? It is known for peptide binding to HLA-Cw*0102 that the key peptide binding residues are at position 3 and position 9. In this case it is Proline at P3 and Leucine
209 at P9. Therefore, this peptide may not bind HLA-A*1011 as well as HLA-Cw*0102. However, the peptide-binding motif of HLA-Cw*0102 is similar to that of HLA-Cw*0304 (Barber et al, 1996). HLA-Cw*0304 also has the same anchor residues at P2 and P9. KIR2DS2+ NK cell clones have been shown to be activated by HLA-Cw*0304 expressing cells (David et al, 2013). Therefore, the potential binding of LNPSVAATL to numerous HLA-C1 alleles may occur and it may not be limited to HLA-Cw*0102.
Viral epitopes for CTLs have been shown to be from recently synthesised proteins rather than mature peptides. This confirms potential for this mechanism for viruses altering host peptide repertoire (Khan, 2001). However, the generation of peptide repertoire is complex process that is not clear at present and is something that requires much future work.
6.5 HCV adaptation and KIR
Overall, our data with the HCV peptides suggest that HCV has not adapted substantially to polymorphic Inhibitory KIR:HLA-C interactions. This is implied by the observation that eight out of the nine invariant HCV peptides tested were either weakly binding to HLA-C and/or non-inhibitory to KIR2DL2/3-positive NK cells when presented by HLA-Cw*0102. The non- inhibitory variant of the HCV core peptide LLPRRGPRL containing a lysine at P8 (LLPRRGPKL) would have no selective advantage for the virus implying that selection at this locus is not important. This poses a question regarding adaptation of the virus to host KIR:HLA interactions as mutating this residue at P8 can result in increased or decreased levels of NK cell degranulation. LLPRRGPRL was one of the most conserved epitopes amongst genotypes 1a, 1b, 2, 3, 4, 5 and 6.
Itakura et al had previously reported that no specific point mutation was directly associated with a change in viral load in these patients (Itakura et al, 2005). Although this R->K mutation has been found in other HCV genotypes (2, 4, 5a and 6), it was not frequent. Interestingly, the HCV core peptide also binds HLA-A*0201 and HLA-E, and the mutation has been previously suggested to lead to viral escape from the HLA-A*0201-restricted CTL response (Battegay et al, 1995; Nattermann et al, 2005). Therefore, as this is also a CTL epitope, there may be other selection pressures acting on it. Overall, there appears to be little selection pressure on these epitopes to induce inhibition of NK cells. Neutralising antibodies have been shown to drive the sequence evolution of the HCV E2 protein (Dowd et al, 2009).
The essential differences between NK cell and T cell recognition of peptides lie in the fact that NK cells can respond to a broader range of MHC-peptide complexes whereas T cells respond only to a specific MHC-peptide complex (Malnati et al, 1995; Lanier and Phillips,
210 1996; Long, 1996). In our study using HCV peptides there appears to be relatively limited selection of specific HCV epitopes for inhibition of NK responses. However, chapter 3 of this study has highlighted that NK cells may respond to changes in the overall peptide content of a cell. Therefore, NK cells may be evolutionarily better adapted to sensing larger shifts in the peptide repertoire of a cell that could occur during viral infection or tumour transformation.
HCV has an exceptionally high mutation rate similar to HIV (Purcell et al, 1991; Gomez et al, 1992; Coffin, 1995; Belshaw et al, 2010). However, despite this high mutation rate, the peptides that we have tested suggest that HCV peptides do not inhibit KIR+ NK cells very efficiently and most are weakly binding to HLA-C. In order to make this assumption we are also basing it on the fact that the majority of HLA-C binding peptides did not inhibit KIR- positive NK cells in the recently published HIV study by Fadda et al (Fadda et al, 2012). Although our HCV peptide screen was considerably smaller, the economics of the situation remain the same. We found one inhibitory HCV peptide out of nine peptides tested in our biased approach and the HIV study found one inhibitory HIV peptide out of 211 peptides screened. The finding of a single inhibitory peptide (LLPRRGPRL) amongst our HCV peptides tested is therefore not substantially different from a chance occurrence, consistent with our thesis that HCV peptide epitopes are not selected on the basis of their ability to inhibit NK cells. The total contribution of HCV to the peptidome still remains unclear at present. All of the epitope prediction methods we used did not work as well for HLA-C as they do for HLA-A/HLA-B. Therefore, further non-intuitive screening of HCV peptides is unlikely to substantially affect the conclusion that most HCV peptides do not influence NK cell reactivity.
6.6 Evolutionary and clinical relevance
Finally, in terms of clinical relevance, if the LNPSVAATL peptide derived from HCV was indeed processed and presented on infected hepatocytes or other antigen-presenting cells during Hepatitis C infection this may only confer advantage on individuals possessing a KIR genotype containing KIR2DS2. In particular, individuals with a KIR haplotype B. Group B haplotypes are defined by the possession of at least one of the following genes: KIR2DL5, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS5 and KIR3DS1 (Martin et al, 2000; Trowsdale et al, 2001; Khakoo and Carrington, 2006). The Group A haplotypes do not possess as many activating KIR receptors. If the peptide is indeed specific for engaging KIR2DS2 on NK cells from individuals with haplotype B it may aid activation of NK cells.
In evolutionary terms, the KIR A haplotypes along with HLA-C1 homozygousity have been strongly associated with more effective immunity against infections such as HCV. This has
211 been best demonstrated in HCV infection where there has been a protective association shown between the possession of KIR2DL3 and HLA-C1 as this is a weaker inhibitory interaction (Khakoo et al 2004; Moesta et al, 2008). However, those with KIR A haplotypes have also been the most high-risk group for pregnancy syndromes such as miscarriage and pre-eclampsia (Hiby et al, 2004; Hiby et al, 2008; Hiby et al, 2010).
The opposite is the case for the KIR B haplotypes, which have been associated with poorer outcomes in most infectious disease studies and protective in pregnancy disorders. An interesting review by Parham and Moffett discusses the fact that during human evolution a compromise was made between reproductive success and immunological success against infections (Parham and Moffett, 2013). For this reason human populations have always maintained consistent frequencies of both KIR A and KIR B Haplotypes. Parham and Moffett have ascertained that the maintenance of both haplotypes has been essential for human survival and evolution as a whole but possession of either haplotype is not essential for survival on an individual basis (Parham and Moffett, 2013).
As mentioned previously, those individuals with the KIR2DS2 gene nearly always inherit the stronger inhibitory gene KIR2DL2 as there is strong linkage disequilibrium between these two genes. It is known that possession of both KIR2DL2 and KIR2DS2 is allelic to KIR2DL3 (Uhrberg et al, 2002). As KIR haplotype A fares better in terms of infectious disease immunity on many whole population analysis, studies perhaps on an individual level, those possessing a KIR B haplotype, have evolutionarily adapted to an alternative mechanism of NK cell activation, which has simply had to place the focus more on engagement of their activating KIR receptors such as engagement of KIR2DS2 with a HCV peptide.
Our results support the theory detailed by Parham and Moffett as we propose that those individuals possessing KIR2DL3 (KIR A haplotypes) can respond well to overall shifts in the peptide repertoire of a cell compared with KIR2DL2 (KIR B haplotypes). Therefore, the shifts in peptide, whether altered self-peptide or viral peptide that could occur early on in an infection would aid in the triggering NK cell activation in those possessing KIR2DL3. However, we also propose that upon finding that a HCV-derived peptide may engage an activating KIR receptor, KIR2DS2, this has provided those individuals possessing a KIR B haplotype containing KIR2DS2 with another mechanism by which to recognise virally- infected cells and subsequent NK cell activation. Therefore, possession of either KIR haplotype B or KIR haplotype A is both sufficient for survival on an individual basis. However, analysing larger cohorts or populations such as the HCV cohort analysed by Khakoo et al has demonstrated the slight advantage conferred by the KIR A haplotype in infectious disease studies. The HLA-C genotype of an individual is also of great importance
212 in these numerous disease studies. HLA-C1 homozygousity appears to confer the greatest protection in most infectious disease studies even if the individual possesses a KIR haplotype B. This may explain all the conflicting evidence obtained in disease studies with KIR3DS1 for example.
213 Chapter 7. Conclusions and Future Direction
One of the main challenges faced by researchers when performing functional experiments on human NK cells is the stochastic and variegated expression of their receptors within each individual. A real obstacle that has remained is the lack of monoclonal antibodies that uniquely identify each KIR receptor as each antibody may recognise the same epitope on more than one receptor. We investigated the inhibitory KIR2DL2 and KIR2DL3 receptors along with the activating receptor KIR2DS2. These three receptors possess an extraordinarily high degree of sequence similarity. In order to distinguish KIR2DL2/KIR2DL3 receptor expression we used PBMCs from numerous donors that were KIR genotyped. Using KIR-Fc proteins we confirmed results reported previously by Fadda et alin 2010. Peptide analogs of the endogenous VAPWNSLSL peptide, mutated at position 7 and 8, can mediate differing degrees of binding to KIR2DL2/3. The binding correlated well with the level of NK cell inhibition that was observed. All of these peptide derivatives could equally stabilise HLA-Cw*0102, an allele associated with viral clearance in HCV (Thio et al, 2002). Futhermore, CD107a functional assays were able to show that changing the peptide content of MHC Class I could result in different responses of KIR2DL3+ NK cells and KIR2DL2+ NK cells. Overall, our data shows that KIR2DL3+ NK cells are more easily activated by fluctuations in the peptide repertoire of a cell than KIR2DL2+ NK cells. This may be in part explained by the higher affinity of KIR2DL2 for its HLA-C1 ligand (Moesta et al, 2008). Whereas, KIR2DL3 has a weaker affinity for HLA-C1 but because of this may be able to sample more of the shifts in the peptide repertoire. In quantitative terms, KIR2DL3+ NK cells were 'activated' with two-fold less antagonistic peptide (~1 µM DA) than KIR2DL2+ NK cells (~2 µM DA). These differences were only observed at high concentrations of the weak inhibitory VAP-RA peptide. Therefore, KIR2DL3+ NK cells surpassed the 50% level of degranulation with fewer shifts in the peptide repertoire, i.e. they were more sensitive to these smaller shifts.
It is crucial to determine if we can find other peptide:MHC complexes that can mediate a differential effect on KIR2DL3 and KIR2DL2 reactivity. Of importance is to use viral peptides or self-peptides, which we initialised as part of this investigation. We used a predictive approach with three algorithms in order to try predict HCV peptides that could bind HLA- Cw*0102 molecules and be presented to KIR2DL2/3+ NK cells. The majority (8 out of 9) of the HCV peptides were non-inhibitory and were poor binders for HLA-Cw*0102. The algorithm outputs correlated poorly with the observed experimental data which highlights that we must be cautious in interpretation of predictive binding data for HLA-C until experimental data can be obtained. This led us to hypothesise that HCV may not be as well adapted to the
214 polymorphic KIR system as we would assume that peptides mediating inhibition would be advantageous to virus survival early in an infection. One of the members of our panel of HCV peptides (LNP) generated surprising data with KIR2DL2+ KIR2DS2+ donors as peptide inhibition appeared to be reversed in the presence of this HCV peptide. We hypothesised that this HCV peptide was able to bind the activating KIR2DS2 receptor and elicit an 'agonistic' effect on NK cell activation. This activation was only obvious in the presence of an inhibitory peptide and was otherwise masked in the presence of no peptide or the peptide alone. Further binding and functional experiments with LNP demonstrated that it can mediate a subtle binding interaction between KIR2DS2 and HLA-Cw*0102. KIR2DS2+ NK cell clones showed increased activation in the presence of LNP although the effect was subtle.
In order to further investigate LNP and KIR2DS2, the Huh7 HCV JFH1ΔE1E2-luc replicon line are a good avenue of investigation followed by the use of cells in further killing assays that have been transfected with HLA-Cw*0102 and other HLA alleles. It would be essential to mutate key KIR binding and/or HLA-C1 anchor residues (P3, P7, P8 or P9) in the LNP epitope within the HCV replicon to observe if this affects HCV replication. Further down the line, if in-vitro results obtained are positive, in-vivo studies may be performed on mice transfected with human KIR2DS2. Peptide elution experiments may be the best means of which to firstly identify other essential viral epitopes and compare the peptide repertoire of HCV-infected cells versus uninfected cells as carried out by Hickman et al for HIV in 2003. It would be essential to then use the eluted peptides in functional assays on both NK cells and possibly CTLs to determine any epitopes of importance. We also have other tools available such as KIR tetramers to detect binding of HCV peptides to these receptors.
Overall, the inhibition and activation of NK cells via KIR is a highly complex balancing process and the fact that peptides can influence the interaction adds only further complexity. However, the fact that different KIR receptors can respond differentially to the same peptide and possibly peptides of viral origin is of utmost importance as the regulation of NK cells may be manipulated with peptides binding to KIR early in an infection. It provides an alternative yet flexible mechanism of NK cell inhibition and activation alongside what has been described to date. Undoubtedly the role of peptides will have to be taken into consideration in all future studies on KIR and other NK cell receptor families.
215 Chapter 8. References
Ahlenstiel G, Martin MP, Gao X, Carrington M, Rehermann B. Distinct KIR/HLA compound genotypes affect the kinetics of human antiviral natural killer cell responses. J Clin Invest. 2008;118(3):1017-26.
Aki M, Shimbara N, Takashina M, Akiyama K, Kagawa S, Tamura T, Tanahashi N, Yoshimura T, Tanaka K, Ichihara A. Interferon-gamma induces different subunit organizations and functional diversity of proteasomes. J Biochem. 1994;115(2):257-69.
Aktas E, Kucuksezer UC, Bilgic S, Erten G, Deniz G. Relationship between CD107a expression and cytotoxic activity. Cell Immunol. 2009;254:149-54.
Alter G, Altfeld M. NK cell function in HIV-1 infection. Curr Mol Med. 2006;6(6):621-9.
Alter G, Malenfant JM, Altfeld M. CD107a as a functional marker for the identification of natural killer cell activity. J Immunol Methods. 2004;294:15-22.
Alter G, Martin MP, Teigen N, Carr WH, Suscovich TJ, Schneidewind A, Streeck H, Waring M, Meier A, Brander C, Lifson JD, Allen TM, Carrington M, Altfield M. Differential natural killer cell-mediated inhibition of HIV-1 replication based on distinct KIR/HLA subtypes. J Exp Med. 2007;204(12):3207-3036.
Alter G, Heckerman D, Schneidewind A, Fadda L, Kadie CM, Carlson JM, Oniangue- Ndza C, Martin M, Li B, Khakoo SI, Carrington M, Allen TM, Altfeld M. HIV-1 adaptation to NK-cell-mediated immune pressure. Nature. 2011;476(7358):96-100.
Altschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J. "Basic local alignment search tool”. J. Mol. Biol. 1990;215:403-410.
Altschul, S.F., Madden, T.L., Schäffer, AA., Zhang, J., Zhang, Z., Miller, W., and Lipman, DJ. "Gapped BLAST and PSI-BLAST: a new generation of protein database search programs". Nucleic Acids Res. 1997;25:3389-3402.
Amadei B, Urbani S, Cazaly A, Fisicaro P, Zerbini A, Ahmed P, Missale G, Ferrari C, Khakoo SI. Activation of natural killer cells during acute infection with hepatitis C virus. Gastroenterology. 2010;138:1536-45.
Andersen MH, Søndergaard I, Zeuthen J, Elliott T, Haurum JS. An assay for peptide binding to HLA-Cw*0102.Tissue Antigens. 1999;54:185-90.
Anderson SK, Ortaldo JR, McVicar DW. The ever-expanding Ly49 gene family: repertoire and signaling. Immunol Rev. 2001;181:79-89.
Andre P, Biassoni R, Colonna M, Cosman D, Lanier LL, Long EO, Lopez-Botet M, Moretta A, Moretta L, Parham P, Trowsdale J, Vivier E, Wagtmann N, Wilson MJ. New nomenclature for MHC receptors. Nat Immunol. 2001;2(8):661.
Anfossi N, André P, Guia S, Falk CS, Roetynck S, Stewart CA, Breso V, Frassati C, Reviron D, Middleton D, Romagné F, Ugolini S, Vivier E. Human NK cell education by inhibitory receptors for MHC class I. Immunity. 2006;25:331-342.
216 Apps R, Murphy SP, Fernando R, Gardner L, Ahad T, Moffett A. Human leucocyte antigen (HLA) expression of primary trophoblast cells and placental cell lines, determined using single antigen beads to characterize allotype specificities of anti-HLA antibodies.Immunology. 2009;127(1):26-39.
Arase H, Lanier LL. Virus-driven evolution of natural killer cell receptors. Microbes Infect. 2002;4(15):1505-12.
Askar M, Avery R, Corey R, Lopez R, Thomas D, Pidwell D, Eghtesad B, Miller C, Fung J, Zein NN. Lack of killer immunoglobulin-like receptor 2DS2 (KIR2DS2) and KIR2DL2 is associated with poor responses to therapy of recurrent hepatitis C virus in liver transplant recipients. Liver Transpl. 2009;15(11):1557-63. BantelH, Schulze-Osthoff K. Apoptosis in hepatitis C virus infection. Cell Death Differ.2003;10 Suppl 1:S48-58.
Barber LD, Percival L, Valiante NM, Chen L, Lee C, Gumperz JE, Phillips JH, Lanier LL, Bigge JC, Parekh RB, Parham P. The inter-locus recombinant HLA-B*4601 has high selectivity in peptide binding and functions characteristic of HLA-C. J Exp Med. 1996;184(2):735-40.
Basler M, Kirk CJ, Groettrup M. The immunoproteasome in antigen processing and other immunological functions.Curr Opin Immunol. 2013;25(1):74-80.
Battegay M, Fikes J, Di Bisceglie AM, Wentworth PA, Sette A, Celis E, Ching WM, Grakoui A, Rice CM, Kurokohchi K, et al. Patients with chronic hepatitis C have circulating cytotoxic T cells which recognize hepatitis C virus-encoded peptides binding to HLA-A2.1 molecules. J Virol. 1995;69(4):2462-70.
Bauer S, Groh V, Wu J, Steinle A, Phillips JH, Lanier LL, Spies T. Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science. 1999;285(5428):727- 9.
Baychelier F, Sennepin A, Ermonval M, Dorgham K, Debré P, Vieillard V. Identification of a cellular ligand for the natural cytotoxicity receptor NKp44. Blood. 2013;122(17):2935-42.
Bezbradica JS, Medzhitov R. Integration of cytokine and heterologous receptor signaling pathways. Nat Immunol. 2009;10:333-9.
Biassoni R, Cantoni C, Falco M, Verdiani S, Bottino C, Vitale M, Conte R, Poggi A, Moretta A, Moretta L. The human leukocyte antigen (HLA)-C-specific "activatory" or "inhibitory" natural killer cell receptors display highly homologous extracellular domains but differ in their transmembrane and intracytoplasmic portions. J Exp Med. 1996;183(2):645-50.
Biassoni R, Cantoni C, Pende D, Sivori S, Parolini S, Vitale M, Bottino C, Moretta A. Human natural killer cell receptors and co-receptors. Immunol Rev. 2001;181:203-14.
Biassoni R, Pessino A, Malaspina A, Cantoni C, Bottino C, Sivori S, Moretta L, Moretta A.Role of amino acid position 70 in the binding affinity of p50.1 and p58.1 receptors for HLA- Cw4 molecules. Eur J Immunol. 1997;27(12):3095-9.
217 Bjorkman PJ, Parham P. Structure, function, and diversity of class I major histocompatibility complex molecules. Annu Rev Biochem. 1990;59:253-88.
Blanchard DK, Stewart WE 2nd, Klein TW, Friedman H, Djeu JY. Cytolytic activity of human peripheral blood leukocytes against Legionella pneumophila-infected monocytes: characterization of the effector cell and augmentation by interleukin 2. J Immunol. 1987;139(2):551-6.
Blight KJ, Kolykhalov AA, Rice CM. Efficient initiation of HCV RNA replication in cell culture. Science. 2000;290(5498):1972-4.
Bolland S and Ravetch JV. Inhibitory pathways triggered by ITIM-containing receptors. Adv Immunol. 1999;72:149-77.
Borhis G, Ahmed PS, Mbiribindi B, Naiyer MM, Davis DM, Purbhoo MA, Khakoo SI. A Peptide Antagonist Disrupts NK Cell Inhibitory Synapse Formation. J Immunol. 2013.
Boyington JC, Motyka SA, Schuck P, Brooks AG, Sun PD. Crystal structure of an NK cell immunoglobulin-like receptor in complex with its class I MHC ligand. Nature. 2000;405(6786):537-43.
Boyington JC, Brooks AG, Sun PD. Structure of killer cell immunoglobulin-like receptors and their recognition of the class I MHC molecules. Immunol Rev. 2001;181:66-78.
Boyington JC, Sun PD. A structural perspective on MHC class I recognition by killer cell immunoglobulin-like receptors. Mol Immunol. 2002;38(14):1007-21.
Braud VM, Allan DS, O'Callaghan CA, Soderstrom K, D'Andrea A, Ogg GS, Lazetic S, Young NT, Bell JI, Phillips JH, Lanier LL, McMichael AJ. HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature. 1998;391(6669):795-9.
Brandt CS, Baratin M, Yi EC, Kennedy J, Gao Z, Fox B, Haldeman B, Ostrander CD, Kaifu T, Chabannon C, Moretta A, West R, Xu W, Vivier E, Levin SD. The B7 family member B7-H6 is a tumor cell ligand for the activating natural killer cell receptor NKp30 in humans. J Exp Med. 2009;206(7):1495-503.
Brodin P, Kärre K, Höglund P. NK cell education: not an on-off switch but a tunable rheostat. Trends Immunol. 2009;30(4):143-9.
Brooks AG, Posch PE, Scorzelli CJ, Borrego F, Coligan JE. NKG2A complexed with CD94 defines a novel inhibitory natural killer cell receptor. J Exp Med. 1997;185(4):795-800.
Burshtyn DN, Long EO. Regulation through inhibitory receptors: Lessons from natural killer cells. Trends Cell Biol. 1997;7(12):473-9.
Caligiuri MA. Human natural killer cells. Blood. 2008;112(3):461-9.
Campbell, K, Dessing M, Lopez-Botet M, Cella M, Colonna M. Tyrosine phosphorylation of a human killer inhibitory receptor recruits protein tyrosine phosphatase 1C. J Exp Med 1996;184:93.
218 Carr WH, Pando MJ, Parham P. KIR3DL1 polymorphisms that affect NK cell inhibition by HLA-Bw4 ligand. J Immunol. 2005;175(8):5222-9.
Carr WH, Rosen DB, Arase H, Nixon DF, Michaelsson J, Lanier LL. Cutting Edge: KIR3DS1, a gene implicated in resistance to progression to AIDS, encodes a DAP12- associated receptor expressed on NK cells that triggers NK cell activation. J Immunol. 2007;178(2):647-51.
Carretero M, Cantoni C, Bellon T, Bottino C, Biassoni R, Rodriguez A, Perez-Villar JJ, Moretta L, Moretta A, Lopez-Botet M. The CD94 and NKG2-A C-type lectins covalently assemble to form a natural killer cell inhibitory receptor for HLA class I molecules. Eur J Immunol. 1997;27(2):563-7.
Cassidy SA, Cheent KS, Khakoo SI. Effects of Peptide on NK cell-mediated MHC I recognition. Front Immunol. 2014;5:133.
Catalfamo M, and Henkart P. Perforin and the granule exocytosis cytotoxicity pathway. Curr Opin Immunol 2003;15: 522-527.
Cella M, Colonna M. Cloning human natural killer cells. Methods Mol Biol. 2000;121:1-4.
Cella M, Longo A, Ferrara GB, Strominger JL, Colonna M. NK3-specific natural killer cells are selectively inhibited by Bw4-positive HLA alleles with isoleucine 80. J Exp Med. 1994;180(4):1235-42.
Cerwenka, A, Lanier L. Natural killer cells, viruses and cancer. Nat Rev Immunol 2001; 1(1):41-9.
Chapman T, Heikeman AP, Bjorkman PJ. The inhibitory receptor LIR-1 uses a common binding interaction to recognise class I MHC molecules and the viral homolog UL18. Immunity. 1999;11: 603-613.
Cheent K, Khakoo SI. Natural killer cells: integrating diversity with function. Immunology. 2009;126:449-57.
Cheent KS, Jamil KM, Cassidy S, Liu M, Mbiribindi B, Mulder A, Claas FH, Purbhoo MA, Khakoo SI. Synergistic inhibition of natural killer cells by the nonsignaling molecule CD94. Proc Natl Acad Sci U S A. 2013;110(42):16981-6.
Chewning JH, Gudme CN, Hsu KC, Selvakumar A, Dupont B. KIR2DS1-positive NK cells mediate alloresponse against the C2 HLA-KIR ligand group in vitro. J Immunol. 2007;179(2):854-68.
Cisneros E, Moraru M, Gómez-Lozano N, López-Botet M, Vilches C. KIR2DL5: An Orphan Inhibitory Receptor Displaying Complex Patterns of Polymorphism and Expression. Front Immunol. 2012;3:289.
Colonna M. Specificity and function of immunoglobulin superfamily NK cell inhibitory and stimulatory receptors. Immunol Rev. 1997;155:127-33.
Colonna M and Samaridis J. Cloning of immunoglobulin-superfamily members associated
219 with HLA-C and HLA-B recognition by human natural killer cells. Science. 1995;268:405- 408.
Cooley S, Xiao F, Pitt M, Gleason M, McCullar V, Bergemann TL, McQueen KL, Guethlein LA, Parham P, Miller JS. A subpopulation of human peripheral blood NK cells that lacks inhibitory receptors for self-MHC is developmentally immature. Blood. 2007. 15;110(2):578-86.
Cooper M, Fehniger TA, Caligiuri MA. The biology of human natural killer-cell subsets. Trends Immunol. 2001;22(11):633-640.
Cosman D, Fanger N, Borges L, Kubin M, Chin W, Peterson L, Hsu ML. A novel immunoglobulin superfamily receptor for cellular and viral MHC class I molecules. Immunity 1997;7:273-282.
Croft NP, Smith SA, Wong YC, Tan CT, Dudek NL, Flesch IE, Lin LC, Tscharke DC, Purcell AW. Kinetics of antigen expression and epitope presentation during virus infection. PLoS Pathog. 2013;9(1):e1003129.
Crome SQ, Lang PA, Lang KS, Ohashi PS. Natural killer cells regulate diverse T cell responses. Trends Immunol. 2013;34(7):342-9.
Crotzer VL, Christian RE, Brooks JM, Shabanowitz J, Settlage RE, Marto JA, White FM, Rickinson AB, Hunt DF, Engelhard VH. Immunodominance among EBV-derived epitopes restricted by HLA-B27 does not correlate with epitope abundance in EBV- transformed B-lymphoblastoid cell lines. J Immunol. 2000;164(12):6120-9.
David G, Djaoud Z, Willem C, Legrand N, Rettman P, Gagne K, Cesbron A and Retiere C. Large spectrum of HLA-C recognition by killer Ig-like receptor (KIR)2DL2 and KIR2DL3 and restricted C1 specificity of KIR2DS2: dominant impact of KIR2DL2/KIR2DS2 on KIR2D NK cell repertoire formation. J Immunol. 2013;191(9):4778-88.
David G, Morvan M, Gagne K, Kerdudou N, Willem C, Devys A, Bonneville M, Folléa G, Bignon JD, Retière C. Discrimination between the main activating and inhibitory killer cell immunoglobulin-like receptor positive natural killer cell subsets using newly characterized monoclonal antibodies. Immunology. 2009;128(2):172-84.
Davis, D. Assembly of the immunological synapse for T cells and NK cells. Trends in Immunol. 2002;23:356-363.
Davis, Chiu I, Fassett M, Cohen GB, Mandelboim O, Strominger JL. The human natural killer cell immune synapse. PNAS. 1999;96:15062-15067.
Di Santo JP. Natural killer cells: diversity in search of a niche. Nat Immunol. 2008;9:473-5.
Döhring C, Scheidegger D, Samaridis J, Cella M, Colonna M. A human killer inhibitory receptor specific for HLA-A1,2. J Immunol. 1996;156(9):3098-101.
Dolan BP, Bennink JR, Yewdell JW. Translating DRiPs: progress in understanding viral and cellular sources of MHC class I peptide ligands. Cell Mol Life Sci. 2011;68(9):1481-9.
220 Estefanía E, Flores R, Gómez-Lozano N, Aguilar H, López-Botet M, Vilches C. Human KIR2DL5 is an inhibitory receptor expressed on the surface of NK and T lymphocyte subsets. J Immunol. 2007;178(7):4402-10.
Fadda L, Borhis G, Ahmed P, Cheent K, Pageon SV, Cazaly A, Stathopoulos S, Middleton D, Mulder A, Claas FH, Elliott T, Davis DM, Purbhoo MA, Khakoo SI. Peptide antagonism as a mechanism for NK cell activation. Proc Natl Acad Sci U S A. 2010;107(22):10160-5.
Fadda L, Körner C, Kumar S, van Teijlingen NH, Piechocka-Trocha A, Carrington M, Altfeld M. HLA-Cw*0102-restricted HIV-1 p24 epitope variants can modulate the binding of the inhibitory KIR2DL2 receptor and primary NK cell function. PLoS Pathog. 2012;8(7):e1002805.
Fan Q, Mosyak L, Winter CC, Wagtmann N, Long EO, Wiley DC. Structure of the inhibitory receptor for human natural killer cells resembles haematopoietic receptors. Nature. 1997;389(6646): 96-100.
Fan QR, Long EO, Wiley DC. Crystal structure of the human natural killer cell inhibitory receptor KIR2DL1-HLA-Cw4 complex. Nat Immunol. 2001;2(5):452-60.
Farfan-Arribas DJ, Stern LJ, Rock KL. Using intein catalysis to probe the origin of major histocompatibility complex class I-presented peptides. Proc Natl Acad Sci U S A. 2012;109(42):16998-7003.
Faure M, and Long E. Spontaneous clustering and tyrosine phosphorylation of NK cell inhibitory receptor induced by ligand binding. J Immunol. 2003;170: 6107-6114.
Ferrington DA, Gregerson DS. Immunoproteasomes: structure, function, and antigen presentation.Prog Mol Biol Transl Sci. 2012;109:75-112.
Flores-Villanueva PO, Yunis EJ, Delgado JC, Vittinghoff E, Buchbinder S, Leung JY, Uglialoro AM, Clavijo OP, Rosenberg ES, Kalams SA, Braun JD, Boswell SL, Walker BD, Goldfeld AE. Control of HIV-1 viremia and protection from AIDS are associated with HLA-Bw4 homozygosity. Proc Natl Acad Sci U S A. 2001;98(9):5140-5.
Foley BA, De Santis D, Van Beelen E, Lathbury LJ, Christiansen FT, Witt CS. The reactivity of Bw4+ HLA-B and HLA-A alleles with KIR3DL1: implications for patient and donor suitability for haploidentical stem cell transplantations. Blood. 2008;112(2):435-43.
Freud AG, Caligiuri MA. Human natural killer cell development. Immunol Rev. 2006;214:56-72.
Freud AG, Yu J, Caligiuri MA. Human natural killer cell development in secondary lymphoid tissues. Semin Immunol. 2014;26(2):132-7.
Freud AG, Zhao S, Wei S, Gitana GM, Molina-Kirsch HF, Atwater SK, Natkunam Y. Expression of the activating receptor, NKp46 (CD335), in human natural killer and T-cell neoplasia. Am J Clin Pathol. 2013;140(6):853-66.
221 Fusco C, Guerini FR, Nocera G, Ventrella G, Caputo D, Valentino MA, Agliardi C, Gallotti J, Morra VB, Florio C, Clerici M, Lombardi ML. KIRs and their HLA ligands in remitting-relapsing multiple sclerosis. J Neuroimmunol. 2010;229(1-2):232-7.
Garcia KC, Degano M, Stanfield RL, Brunmark A, Jackson MR, Peterson PA, Teyton L, Wilson IA. An alphabeta T cell receptor structure at 2.5 A and its orientation in the TCR- MHC complex. Science. 1996;274(5285):209-19.
Gardiner CM. Killer cell immunoglobulin-like receptors on NK cells: the how, where and why. Int J Immunogenet.2008;35:1-8.
Gasteiger E, Gattiker A, Hoogland C, Ivanyi I, Appel R.D and Bairoch A. ExPASy: the proteomics server for in-depth protein knowledge and analysis. Nucleic Acids Res. 2003;31:3784-3788.
Gerosa F, Baldani-Guerra B, Nisii C, Marchesini V, Carra G, Trinchieri G. Reciprocal activating interaction between natural killer cells and dendritic cells. J Exp Med. 2002;195(3):327-33.
Gómez-Lozano N, Gardiner CM, Parham P, Vilches C. Some human KIR haplotypes contain two KIR2DL5 genes: KIR2DL5A and KIR2DL5B. Immunogenetics. 2002;54(5):314-9.
Goujon M, McWilliam H, Li W, Valentin F, Squizzato S, Paern J, Lopez R.A new bioinformatics analysis tools framework at EMBL-EBI. Nucleic acids research. 2010;38Suppl: W695-9 doi:10.1093/nar/gkq313.
Gourraud PA, Meenagh A, Cambon-Thomsen A and Middleton D. Linkage disequilibrium organization of the human KIR superlocus: implications for KIR data analyses. Immunogenetics. 2010;62(11-12):729-40.
Graef T, Moesta AK, Norman PJ, Abi-Rached L, Vago L, Older Aguilar AM, Gleimer M, Hammond JA, Guethlein LA, Bushnell DA, Robinson PJ, Parham P. KIR2DS4 is a product of gene conversion with KIR3DL2 that introduced specificity for HLA-A*11 while diminishing avidity for HLA-C. J Exp Med. 2009;206(11):2557-72.
Gumperz JE, Litwin V, Phillips JH, Lanier LL, Parham P. The Bw4 public epitope of HLA- B molecules confers reactivity with natural killer cell clones that express NKB1, a putative HLA receptor. J Exp Med. 1995;181(3):1133-44.
Hansasuta P, Dong T, Thananchai H, Weekes M, Willberg C, Aldemir H, Rowland- Jones S, Braud VM. Recognition of HLA-A3 and HLA-A11 by KIR3DL2 is peptide-specific. Eur J Immunol. 2004;34(6):1673-9.
Herberman R, Nunn ME, Lavrin DH. Natural cytotoxic reactivity of mouse lymphoid cells against syngeneic acid allogeneic tumors I. Distribution of reactivity and specificity. Int. J. Cancer. 1975a;16; 216–229.
Herberts CA, Stittelaar KJ, van der Heeft E, van Gaans-Van den Brink J, Poelen MC, Roholl PJ, van Alphen LJ, Melief CJ, de Jong AP, van Els CA. A measles virus glycoprotein-derived human CTL epitope is abundantly presented via the proteasomal- dependent MHC class I processing pathway. J Gen Virol. 2001;82(Pt 9):2131-42.
222 Hiby S, Walker JJ, O'shaughnessy KM, Redman CW, Carrington M, Trowsdale J, Moffett A. Combination of maternal KIR and fetal HLA-C genes influence the risk of preeclampsia and reproductive success. J Exp Med. 2004;200:957-965.
Hiby SE, Apps R, Sharkey AM, Farrell LE, Gardner L, Mulder A, Claas FH, Walker JJ, Redman CW, Morgan L, Tower C, Regan L, Moore GE, Carrington M, Moffett A. Maternal activating KIRs protect against human reproductive failure mediated by fetal HLA- C2. J Clin Invest. 2010;120(11):4102-10.
Hickman H, Luis AD, Buchli R, Few SR, Sathiamurthy M, VanGundy RS, Giberson CF, Hildebrand WH. Toward a definition of self: proteomic evaluation of the class I peptide repertoire. J Immunol. 2004;172:2944-2952
Hickman HD, Luis AD, Bardet W, Buchli R, Battson CL, Shearer MH, Jackson KW, Kennedy RC, Hildebrand WH. Cutting edge: class I presentation of host peptides following HIV infection.J Immunol. 2003;171:22-6.
Held W, Mariuzza RA. Cis interactions of immunoreceptors with MHC and non-MHC ligands. Nat Rev Immunol. 2008;8(4):269-78.
Henderson RA, Michel H, Sakaguchi K, Shabanowitz J, Appella E, Hunt DF, Engelhard VH. HLA-A2.1-associated peptides from a mutant cell line: a second pathway of antigen presentation. Science. 1992;255(5049):1264-6.
Herberts CA, van Gaans-van den Brink J, van der Heeft E, van Wijk M, Hoekman J, Jaye A, Poelen MC, Boog CJ, Roholl PJ, Whittle H, de Jong AP, van Els CA. Autoreactivity against induced or upregulated abundant self-peptides in HLA-A*0201 following measles virus infection. Hum Immunol. 2003;64(1):44-55.
Herberts CA, Meiring HM, van Gaans-van den Brink JA, van der Heeft E, Poelen MC, Boog CJ, de Jong AP, van Els CA. Dynamics of measles virus protein expression are reflected in the MHC class I epitope display. Mol Immunol. 2003;39(10):567-75.
Hilton HG, Vago L, Older Aguilar AM, Moesta AK, Graef T, Abi-Rached L, Norman PJ, Guethlein LA, Fleischhauer K, Parham P. Mutation at positively selected positions in the binding site for HLA-C shows that KIR2DL1 is a more refined but less adaptable NK cell receptor than KIR2DL3. J Immunol. 2012;189(3):1418-30.
Hirayasu K, Ohashi J, Kashiwase K, Hananantachai H, Naka I, Ogawa A, Takanashi M, Satake M, Nakajima K, Parham P, Arase H, Tokunaga K, Patarapotikul J, Yabe T. Significant Association of KIR2DL3-HLA-C1 Combination with Cerebral Malaria and Implications for Co-evolution of KIR and HLA. PLoS Pathog. 2012;8(3):e1002565.
Hiroishi K, Ito T, Imawari M. Immune responses in hepatitis C virus infection and mechanisms of hepatitis C virus persistence. J Gastroenterol Hepatol. 2008;23:1473-82.
Hoare HL, Sullivan LC, Clements CS, Ely LK, Beddoe T, Henderson KN, Reid HH, Brooks AG, Rossjohn, J. Subtle changes in peptide conformation profoundly affect recognition of the non-classical MHC class I molecule HLA-E by the CD94-NKG2 natural killer cell receptors. J Mol Biol. 2008;377(5):1297-303.
223 Hollenbach JA, Ladner MB, Saeteurn K, Taylor KD, Mei L, Haritunians T, McGovern DP, Erlich HA, Rotter JI, Trachtenberg EA. Susceptibility to Crohn's disease is mediated by KIR2DL2/KIR2DL3 heterozygosity and the HLA-C ligand. Immunogenetics. 2009;61(10):663-71.
Hoof I, Peters B, Sidney J, Pedersen LE, Sette A, Lund O, Buus S, Nielsen M. NetMHCpan, a method for MHC class I binding prediction beyond humans. Immunogenetics. 2009;61(1):1-13.
Horner SM, Gale M Jr. Regulation of hepatic innate immunity by hepatitis C virus. Nat Med. 2013;19(7):879-88.
Horowitz A, Riley EM. Activation of human NK cells by malaria-infected red blood cells. Methods Mol Biol. 2010;612:429-46.
Horowitz A, Riley EM. Activation of human NK cells by Plasmodium-infected red blood cells. Methods Mol Biol. 2013;923:447-64.
Hov JR, Lleo A, Selmi C, Woldseth B, Fabris L, Strazzabosco M, Karlsen TH, Invernizzi P. Genetic associations in Italian primary sclerosing cholangitis: heterogeneity across Europe defines a critical role for HLA-C. J Hepatol. 2010;52(5):712-7.
Hsu KC, Chida S, Geraghty DE, Dupont B. The killer cell immunoglobulin-like receptor (KIR) genomic region: gene-order, haplotypes and allelic polymorphism. Immunol Rev. 2002;190:40-52.
Hsu KC, Chida S, Geraghty DE, Dupont B. The killer cell immunoglobulin-like receptor (KIR) genomic region: gene-order, haplotypes and allelic polymorphism. Immunol Rev. 2002;190:40-52.
Itakura J, Nagayama K, Enomoto N, Hamano K, Sakamoto N, Fanning LJ, Kenny- Walsh E, Shanahan F, Watanabe M. Viral load change and sequential evolution of entire hepatitis C virus genome in Irish recipients of single source-contaminated anti-D immunoglobulin. J Viral Hepat. 2005;12:594-603.
Jacob L and Vert JP. Efficient peptide-MHC-I binding prediction for alleles with few known binders. Bioinformatics. 2008;24(3):358-66.
Jacobs R, Hintzen G, Kemper A, Beul K, Kempf S, Behrens G, Sykora KW, Schmidt RE. CD56bright cells differ in their KIR repertoire and cytotoxic features from CD56dim NK cells. Eur J Immunol. 2001;(10):3121-7.
Jamil KM and Khakoo SI. KIR/HLA interactions and pathogen immunity. J Biomed Biotechnol. 2011.
Jennes W, Verheyden S, Demanet C, Adjé-Touré CA, Vuylsteke B, Nkengasong JN, Kestens L. Cutting edge: resistance to HIV-1 infection among African female sex workers is associated with inhibitory KIR in the absence of their HLA ligands. J Immunol. 2006;177(10):6588-92.
224 Jiang W, Johnson C, Jayaraman J, Simecek N, Noble J, Moffatt MF, Cookson WO, Trowsdale J, Traherne JA. Copy number variation leads to considerable diversity for B but not A haplotypes of the human KIR genes encoding NK cell receptors. Genome Res. 2012;22(10):1845-54.
Jinushi M, Takehara T, Tatsumi T, Kanto T, Miyagi T, Suzuki T, Kanazawa Y, Hiramatsu N, Hayashi N. Negative regulation of NK cell activities by inhibitory receptor CD94/NKG2A leads to altered NK cell-induced modulation of dendritic cell functions in chronic hepatitis C virus infection. J Immunol. 2004;173(10):6072-81
Jojic N, Reyes-Gomez M, Heckerman D, Kadie C, Schueler-Furman O. Learning MHC I-- peptide binding. Bioinformatics. 2006;22(14):e227-35.
Joncker NT and Raulet DH. Regulation of NK cell responsiveness to achieve self-tolerance and maximal responses to diseased target cells. Immunol Rev. 2008;224:85-97.
Jones DC, Edgar RS, Ahmad T, Cummings JR, Jewell DP, Trowsdale J, Young NT. Killer Ig-like receptor (KIR) genotype and HLA ligand combinations in ulcerative colitis susceptibility. Genes Immun. 2006;7(7):576-82.
Kaiser BK, Barahmand-Pour F, Paulsene W, Medley S, Geraghty DE, Strong RK., Interactions between NKG2x immunoreceptors and HLA-E ligands display overlapping affinities and thermodynamics. J Immunol. 2005;174(5):2878-84.
Kaiser BK, Pizarro JC, Kerns J, Strong RK. Structural basis for NKG2A/CD94 recognition of HLA-E. Proc Natl Acad Sci U S A. 2008;105(18):6696-701.
Karre K, Ljunggren HG, Piontek G and Kiessling R. Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. Nature. 1986;319(6055):675-8.
Katz, G, Markel G, Mizrahi S, Arnon TI, Mandelboim O. Recognition of HLA-Cw4 but not HLA-Cw6 by the NK cell inhibitory receptor two-domain short tail number 4. J Immunol .2001;166:7260- 7267
Keaney L, Williams F, Meenagh A, Sleator C, Middleton D. Investigation of killer cell immunoglobulin-like receptor gene diversity III. KIR2DL3. Tissue Antigens. 2004; 64:188-94.
Khakoo SI, Carrington M. KIR and disease: a model system or system of models? Immunol Rev. 2006;214:186-201.
Khakoo SI, Rajalingam R, Shum BP, Weidenbach K, Flodin L, Muir DG, Canavez F, Cooper SL, Valiante NM, Lanier LL, Parham P. Rapid evolution of NK cell receptor systems demonstrated by comparison of chimpanzees and humans. Immunity (2000) 12:687-98.
Khakoo SI, Thio CL, Martin MP, Brooks CR, Gao X, Astemborski J, Cheng J, Goedert JJ, Vlahov D, Hilgartner M, Cox S, Little AM, Alexander GJ, Cramp ME, O'Brien SJ, Rosenberg WM, Thomas DL, Carrington M. HLA and NK cell inhibitory receptor genes in resolving hepatitis C virus infection. Science. 2004;305(5685):872-4.
225 Khan S, de Giuli R, Schmidtke G, Bruns M, Buchmeier M, van den Broek M, Groettrup M. Cutting edge: neosynthesis is required for the presentation of a T cell epitope from a long-lived viral protein. J Immunol. 2001;167(9):4801-4.
Kiessling R, Klein E, Pross H, Wigzell H. "Natural" killer cells in the mouse. II. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Characteristics of the killer cell. Eur J Immunol. 1975;5:117-21.
Kim, S,Poursine-Laurent J, Truscott SM, Lybarger L, Song YJ, Yang L, French AR, Sunwoo JB, Lemieux S, Hansen TH, Yokoyama WM. Licensing of natural killer cells by host major histocompatability complex class I molecules. Nature. 2005; 436(4):709-713.
Kim, S, Iizuka K, Aguila HL, Weissman IL, Yokoyama WM. In vivo natural killer cell activities revealed by natural killer cell- deficient mice. Proc Natl Acad Sci U S A. 2000; 97(6):2731-2736.
Kollnberger S,Chan A, Sun MY, Chen LY, Wright C, di Gleria K, McMichael A, Bowness P. Interaction of HLA-B27 homodimers with KIR3DL1 and KIR3DL2, unlike HLA- B27 heterotrimers, is independent of the sequence of bound peptide. Eur J Immunol. 2007;37(5):1313-22.
Korbel DS, Newman KC, Almeida CR, Davis DM, Riley EM. Heterogeneous human NK cell responses to Plasmodium falciparum-infected erythrocytes. J Immunol. 2005;175(11):7466-73.
Korbel DS, Norman PJ, Newman KC, Horowitz A, Gendzekhadze K, Parham P, Riley EM. Killer Ig-like receptor (KIR) genotype predicts the capacity of human KIR-positive CD56dim NK cells to respond to pathogen-associated signals. J Immunol. 2009;182(10):6426-34.
Knapp S, Warshow U, Hegazy D, Brackenbury L, Guha IN, Fowell A, Little AM, Alexander GJ, Rosenberg WM, Cramp ME, Khakoo SI. Consistent beneficial effects of killer cell immunoglobulin-like receptor 2DL3 and group 1 human leukocyte antigen-C following exposure to hepatitis C virus. Hepatology. 2010;51:1168-75.
Kraft JR, Vance RE, Pohl J, Martin AM, Raulet DH, Jensen PE., Analysis of Qa-1(b) peptide binding specificity and the capacity of CD94/NKG2A to discriminate between Qa-1- peptide complexes. J Exp Med. 2000;192(5):613-24.
Kulkarni S, Martin MP, Carrington M. The Yin and Yang of HLA and KIR in human disease. Semin Immunol. 2008;20(6):343-52.
Lampen MH, Hassan C, Sluijter M, Geluk A, Dijkman K, Tjon JM, de Ru AH, van der Burg SH, van Veelan PA, van Hall T. Alternative peptide repertoire of HLA-E reveals a binding motif that is strikingly similar to HLA-A2. Mol Immunol. 2013;53(1-2):126-31.
Lan L, Gorke S, Rau SJ, Zeisel MB, Hildt E, Himmelsbach K, Carvajal-Yepes M, Huber R, Wakita T, Schmitt-Graeff A, Royer C, Blum HE, Fischer R, Baumert TF. Hepatitis C virus infection sensitizes human hepatocytes to TRAIL-induced apoptosis in a caspase 9- dependent manner. J Immunol. 2008;181(7):4926-35.
226 Lanier L,Testi R, Bindl J, Phillips JH. Identity of Leu-19 (CD56) leukocyte differentiation antigen and neural cell adhesion molecule. J Exp Med. 1989;169:2233-2238.
Lanier, L. NK cell receptors. Ann Rev Immunol. 1998;16: 359-393.
Lanier, L. NK cell recognition. Ann Rev Immunol. 2005;23: 225-274.
Lanier LL. Up on the tightrope: natural killer cell activation and inhibition. Nat Immunol. 2008;9(5):495-502.
Lanier LL, Phillips JH. Inhibitory MHC class I receptors on NK cells and T cells. Immunol Today. 1996;17(2):86-91.
Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, Thompson JD, Gibson TJ and Higgins DG. ClustalW and ClustalX version 2. Bioinformatics. 2007;23(21): 2947-2948.
Lazetic S, Chang C, Houchins JP, Lanier LL, Phillips JH. Human natural killer cell receptors involved in MHC class I recognition are disulfide-linked heterodimers of CD94 and NKG2 subunits. J Immunol. 1996;157(11):4741-5.
Lee KM, Forman JP, McNerney ME, Stepp S, Kuppireddi S, Guzior D, Latchman YE, Sayegh MH, Yagita H, Park CK, Oh SB, Wülfing C, Schatzle J, Mathew PA, Sharpe AH, Kumar V. Requirement of homotypic NK-cell interactions through 2B4(CD244)/CD48 in the generation of NK effector functions. Blood. 2006;107(8):3181-8.
Lee N, Llano M, Carretero M, Ishitani A, Navarro F, Lopez-Botet M, Geraghty DE. HLA- E is a major ligand for the natural killer inhibitory receptor CD94/NKG2A. Proc Natl Acad Sci U S A. 1998;95(9):5199-204.
Lehner PJ, Surman MJ, Cresswell P. Soluble tapasin restores MHC class I expression and function in the tapasin-negative cell line .220. Immunity. 1998;8(2):221-31.
Lepin EJ, Bastin JM, Allan DS, Roncador G, Braud VM, Mason DY, van der Merwe PA, McMichael AJ, Bell JI, Powis SH, O'Callaghan CA. Functional characterization of HLA-F and binding of HLA-F tetramers to ILT2 and ILT4 receptors. Eur J Immunol. 2000;30(12):3552-61.
Levitsky V, Zhang QJ, Levitskaya J, Masucci MG. The life span of major histocompatibility complex-peptide complexes influences the efficiency of presentation and immunogenicity of two class I-restricted cytotoxic T lymphocyte epitopes in the Epstein-Barr virus nuclear antigen 4. J Exp Med. 1996;183(3):915-26.
Li P, Morris DL, Willcox BE, Steinle A, Spies T, Strong RK. Complex structure of the activating immunoreceptor NKG2D and its MHC class I-like ligand MICA. Nat Immunol. 2001;2(5):443-51.
Liu J, Xiao Z, Ko HL, Shen M, Ren EC. Activating killer cell immunoglobulin-like receptor 2DS2 binds to HLA-A*11. Proc Natl Acad Sci U S A. 2014;111(7):2662-7.
227 Llano M, Lee N, Navarro F, Garcia P, Albar JP, Geraghty DE, Lopez-Botet M. HLA-E- bound peptides influence recognition by inhibitory and triggering CD94/NKG2 receptors: preferential response to an HLA-G-derived nonamer. Eur J Immunol. 1998;28(9):2854-63.
Ljunggren, H, Pääbo S, Cochet M, Kling G, Kourilsky P, Kärre K. Molecular analysis of H-2-deficient lymphoma lines. Distinct defects in biosynthesis and association of MHC class I heavy chains and beta 2- microglobulin observed in cells with increased sensitivity to NK cell lysis. J Immunol. 1989; 142:2911–17.
Ljunggren HG, Kärre K. Host resistance directed selectively against H-2-deficient lymphoma variants. Analysis of the mechanism. J Exp Med. 1985;162(6):1745-59.
Ljunggren HG, Kärre K. In search of the 'missing self': MHC molecules and NK cell recognition. Immunol Today. 1990;11(7):237-44.
Ljunggren HG, Malmberg KJ. Prospects for the use of NK cells in immunotherapy of human cancer. Nat Rev Immunol. 2007;7(5):329-39.
Lohmann V, Körner F, Koch J, Herian U, Theilmann L, Bartenschlager R. Replication of subgenomic hepatitis C virus RNAs in a hepatoma cell line. Science. 1999;285(5424):110-3.
Long EO., Negative signaling by inhibitory receptors: the NK cell paradigm. Immunol Rev. 2008;224:70-84.
Long EO, Colonna M, Lanier LL. Inhibitory MHC class I receptors on NK and T cells: a standard nomenclature. Immunol Today. 1996;17(2):100.
Long EO, Kim HS, Liu D, Peterson ME, Rajagopalan S. Controlling natural killer cell responses: integration of signals for activation and inhibition. Annu Rev Immunol. 2013;31:227-58.
López-Vázquez A, Rodrigo L, Martínez-Borra J, Pérez R, Rodríguez M, Fdez-Morera JL, Fuentes D, Rodríguez-Rodero S, Gonzáez S, López-Larrea C. Protective effect of the HLA-Bw4I80 epitope and the killer cell immunoglobulin-like receptor 3DS1 gene against the development of hepatocellular carcinoma in patients with hepatitis C virus infection. J Infect Dis. 2005;192(1):162-5.
López-Vázquez A, Miña-Blanco A, Martínez-Borra J, Njobvu PD, Suárez-Alvarez B, Blanco-Gelaz MA, González S, Rodrigo L, López-Larrea C. Interaction between KIR3DL1 and HLA-B*57 supertype alleles influences the progression of HIV-1 infection in a Zambian population. Hum Immunol. 2005;66(3):285-9.
Lopez-Vergès S, Milush JM, Pandey S, York VA, Arakawa-Hoyt J, Pircher H, Norris PJ, Nixon DF, Lanier LL. CD57 defines a functionally distinct population of mature NK cells in the human CD56dimCD16+ NK-cell subset. Blood. 2010;116(19):3865-74.
Lorenz U. SHP-1 and SHP-2 in T cells: two phosphatases functioning at many levels. Immunol Rev. 2009;228(1):342-59.
Lucchiari-Hartz M, van Endert PM, Lauvau G, Maier R, Meyerhans A, Mann D, Eichmann K, Niedermann G. Cytotoxic T lymphocyte epitopes of HIV-1 Nef: Generation of
228 multiple definitive major histocompatibility complex class I ligands by proteasomes. J Exp Med. 2000;191(2):239-52.
Luque I, Solana R, Galiani MD, González R, García F, López de Castro JA, Peña J. Threonine 80 on HLA-B27 confers protection against lysis by a group of natural killer clones. Eur J Immunol. 1996;26(8):1974-7.
Maenaka K, Juji T, Stuart DI, Jones EY. Crystal structure of the human p58 killer cell inhibitory (KIR2DL3) specific for HLA-Cw3-related MHC class I. Structure. 1999a; 7(4): 391- 398.
Maenaka K, Juji T, Nakayama T, Wyer JR, Gao GF, Maenaka T, Zaccai NR, Kikuchi A, Yabe T, Tokunaga K, Tadokoro K, Stuart DI, Jones EY, van der Merwe PA. Killer cell immunoglobulin receptors and T cell receptors bind peptide-major histocompatibility complex class I with distinct thermodynamic and kinetic properties. J Biol Chem. 1999;274(40):28329-34.
Maghazachi AA. Compartmentalization of human natural killer cells. Mol Immunol. 2005;42(4):523-9.
Malnati MS, Lusso P, Ciccone E, Moretta A, Moretta L, Long EO. Recognition of virus- infected cells by natural killer cell clones is controlled by polymorphic target cell elements. J Exp Med. 1993;178:961-9.
Malnati MS, Peruzzi M, Parker KC, Biddison WE, Ciccone E, Moretta A, Long EO. Peptide specificity in the recognition of MHC class I by natural killer cell clones. Science. 1995;267(5200):1016-8.
Maloveste SM,Chen D, Gostick E, Vivian JP, Plishka RJ, Iyengar R, Kruthers RL, Buckler-White A, Brooks AG, Rossjohn J, Price DA, Lafont BA. Degenerate recognition of MHC class I molecules with Bw4 and Bw6 motifs by a killer cell Ig-like receptor 3DL expressed by macaque NK cells. J Immunol. 2012;189(9):4338-48.
Mandelboim O, Wilson SB, Vales-Gomez M, Reyburn HT, Strominger JL. Self and viral peptides can initiate lysis by autologous natural killer cells. Proc Natl Acad Sci U S A. 1997;94(9):4604-9.
Mandelboim O. Recognition of haemagglutinins on virus-infected cells by NKp46 activates lysis by human NK cells. 2001;409(6823).
Marcus A, Raulet DH. Evidence for natural killer cell memory. Curr Biol. 2013;23(17):R817- 20.
Marsh SGE, Albert ED, Bodmer WF, Bontrop RE, Dupont B, Erlich HA, Hansen JA, Mach B, Mayr WR, Parham P, Petersdorf EW, Sasazuki T, Schreuder GMTh, Strominger JL, Svejgaard A, Terasaki PI. Nomenclature for factors of the HLA system, 2002. European Journal of Immunogenetics (2002) 29:463-517. Human Immunology (2002) 63:1213-1268. Tissue Antigens (2002) 60:407-464.
229 Marsh SGE, Parham P, Dupont B, Geraghty DE, Trowsdale J, Middleton D, Vilches C, Carrington M, Witt C, Guethlein LA, Shilling H, Garcia CA, Hsu KC, Wain H. Killer-cell Immunoglobulin-like Receptors (KIR) Nomenclature Report, 2002. Tissue Antigens (2003) 62:79-86. Immunogenetics (2003) 55:220-226. Human Immunology (2003) 64:648-654. European Journal of Immunogenetics (2003) 30:229-234. Journal of Immunological Methods (2003) 281:1-8.
Marsh SGE, Parham P and Barber L. The HLA FactsBook (2000). 1st Edition.
Martell M, Esteban JI, Quer J, Genescà J, Weiner A, Esteban R, Guardia J, Gómez J. Hepatitis C virus (HCV) circulates as a population of different but closely related genomes: quasispecies nature of HCV genome distribution. J Virol. 1992;66(5):3225-9.
Martin MP, Pascal V, Yeager M, Phair J, Kirk GD, Hoots K, O'Brien SJ, Anderson SK, Carrington M. A mutation in KIR3DS1 that results in truncation and lack of cell surface expression. Immunogenetics. 2007;59(10):823-9.
Martin MP, Nelson G, Lee JH, Pellett F, Gao X, Wade J, Wilson MJ, Trowsdale J, Gladman D, Carrington M. Cutting edge: susceptibility to psoriatic arthritis: influence of activating killer Ig-like receptor genes in the absence of specific HLA-C alleles. J Immunol. 2002;169(6):2818-22.
Martin MP, Bashirova A, Traherne J, Trowsdale J, Carrington M. Cutting edge: expansion of the KIR locus by unequal crossing over. J Immunol. 2003;171(5):2192-5.
Martin AM, Kulski JK, Gaudieri S, Witt CS, Freitas EM, Trowsdale J, Christiansen FT. Comparative genomic analysis, diversity and evolution of two KIR haplotypes A and B. Gene. 2004;335:121-31.
Martin MP, Qi Y, Gao X, Yamada E, Martin JN, Pereyra F, Colombo S, Brown EE, Shupert WL, Phair J, Goedert JJ, Buchbinder S, Kirk GD, Telenti A, Connors M, O'Brien SJ, Walker BD, Parham P, Deeks SG, McVicar DW, Carrington M. Innate partnership of HLA-B and KIR3DL1 subtypes against HIV-1. Nat Genet. 2007;39(6):733-40.
Mathew PA, Garni-Wagner BA, Land K, Takashima A, Stoneman E, Bennett M, Kumar V. Cloning and characterization of the 2B4 gene encoding a molecule associated with non- MHC-restricted killing mediated by activated natural killer cells and T cells. J Immunol. 1993;151(10):5328-37.
McNerney ME, Guzior D, Kumar V. 2B4 (CD244)-CD48 interactions provide a novel MHC class I-independent system for NK-cell self-tolerance in mice. Blood. 2005;106(4):1337-40.
McWilliam H, Li W, Uludag M, Squizzato S, Park YM, Buso N, Cowley AP, Lopez R. Analysis Tool Web Services from the EMBL-EBI. Nucleic acids research 2013;41(WebServer issue):W597-600 doi:10.1093/nar/gkt376.
Meier UC, Owen RE, Taylor E, Worth A, Naoumov N, Willberg C, Tang K, Newton P, Pellegrino P, Williams I, Klenerman P, Borrow P. Shared alterations in NK cell frequency, phenotype, and function in chronic human immunodeficiency virus and hepatitis C virus infections. J Virol. 2005; 79:12365-74.
230 Miah SM, Hughes TL, Campbell KS. KIR2DL4 differentially signals downstream functions in human NK cells through distinct structural modules. J Immunol. 2008;180(5):2922-32.
Michaëlsson J, Teixeira de Matos C, Achour A, Lanier LL, Kärre K, Söderström K. A signal peptide derived from hsp60 binds HLA-E and interferes with CD94/NKG2A recognition. J Exp Med. 2002;196:1403-14.
Migueles SA, Sabbaghian MS, Shupert WL, Bettinotti MP, Marincola FM, Martino L, Hallahan CW, Selig SM, Schwartz D, Sullivan J, Connors M. HLA B*5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors. Proc Natl Acad Sci U S A. 2000;97(6):2709-14.
Mingari MC, Moretta A, Moretta L. Regulation of KIR expression in human T cells: a safety mechanism that may impair protective T-cell responses. Immunol Today. 1998;19(4):153-7.
Mingari MC, Schiavetti F, Ponte M, Vitale C, Maggi E, Romagnani S, Demarest J, Pantaleo G, Fauci AS, Moretta L. Human CD8+ T lymphocyte subsets that express HLA class I-specific inhibitory receptors represent oligoclonally or monoclonally expanded cell populations. Proc Natl Acad Sci U S A. 1996;93(22):12433-8.
Mingari MC, Vitale C, Schiavetti F, Cambiaggi A, Bertone S, Zunino A, Ponte M. HLA- class I-specific inhibitory receptors of NK type on a subset of human T cells. Chem Immunol. 1996;64:135-45.
Mistry AR, O'Callaghan CA. Regulation of ligands for the activating receptor NKG2D. Immunology. 2007;121(4):439-47.
Moesta AK, Norman PJ, Yawata M, Yawata N, Gleimer M, Parham P. Synergistic polymorphism at two positions distal to the ligand-binding site makes KIR2DL2 a stronger receptor for HLA-C than KIR2DL3. J Immunol. 2008.180:3969-79.
Moesta AK, Graef T, Abi-Rached L, Older Aguilar AM, Guethlein LA, Parham P. Humans differ from other hominids in lacking an activating NK cell receptor that recognizes the C1 epitope of MHC class I. J Immunol. 2010;185(7):4233-7.
Moesta AK, Parham P. Diverse functionality among human NK cell receptors for the C1 epitope of HLA-C: KIR2DS2, KIR2DL2, and KIR2DL3. Frontiers in immunology. 2012;3:336.
Moffett-King A. Natural killer cells and pregnancy. Nat Rev Immunol. 2002;2(9):656-63.
Momot T, Koch S, Hunzelmann N, Krieg T, Ulbricht K, Schmidt RE, Witte T. Association of killer cell immunoglobulin-like receptors with scleroderma. Arthritis Rheum. 2004;50(5):1561-5.
Montel A, Bochan MR, Goebel WS, Brahmi Z. Fas-mediated cytotoxicity remains intact in perforin and granzyme B antisense transfectants of a human NK-like cell line. Cell Immunol 1995; 165:312-317.
Moretta A, Vitale M, Bottino C, Orengo AM, Morelli L, Augugliaro R, Barbaresi M, Ciccone E, Moretta L. P58 molecules as putative receptors for major histocompatibility complex (MHC) class I molecules in human natural killer (NK) cells. Anti-p58 antibodies
231 reconstitute lysis of MHC class I-protected cells in NK clones displaying different specificities. J Exp Med. 1993;178(2):597-604.
Moretta A, Tambussi G, Bottino C, Tripodi G, Merli A, Ciccone E, Pantaleo G, Moretta L. A novel surface antigen expressed by a subset of human CD3- CD16+ natural killer cells. Role in cell activation and regulation of cytolytic function. J Exp Med. 1990;171(3):695-714.
Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC, Biassoni R, Moretta L. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu Rev Immunol. 2001;19:197-223.
Moretta L, Moretta A. Killer immunoglobulin-like receptors. Curr Opin Immunol. 2004;16:626-33.
Moretta A., The dialogue between human natural killer cells and dendritic cells. Curr Opin Immunol. 2005;17(3):306-11.
Moretta A., NK cells at the interface between innate and adaptive immunity. Cell Death Differ. 2008;15(2):226-33.
Morris RJ, Chong LK, Wilkinson GW, Wang EC. A high-efficiency system of natural killer cell cloning. J Immunol Methods. 2005;307(1-2):24-33.
Nattermann J, Nischalke HD, Hofmeister V, Ahlenstiel G, Zimmermann H, Leifeld L, Weiss EH, Sauerbruch T, Spengler U. The HLA-A2 restricted T cell epitope HCV core 35- 44 stabilizes HLA-E expression and inhibits cytolysis mediated by natural killer cells. Am J Pathol. 2005;166:443-53.
Nausch N, Cerwenka A. NKG2D ligands in tumor immunity. Oncogene. 2008;27(45):5944- 58.
Navarro F, Llano M, García P, López-Botet M. NK cell mediated recognition of HLA class Ib molecules: role of CD94/NKG2 receptors. J Reprod Immunol. 1999;43(2):167-73.
Newman KC, Riley EM. Whatever turns you on: accessory-cell-dependent activation of NK cells by pathogens. Nat Rev Immunol. 2007;7(4):279-91.
Nielsen CM, White MJ, Goodier MR, Riley EM. Functional Significance of CD57 Expression on Human NK Cells and Relevance to Disease. Front Immunol. 2013;4:422.
Nikitina-Zake L, Rajalingham R, Rumba I, Sanjeevi CB. Killer cell immunoglobulin-like receptor genes in Latvian patients with type 1 diabetes mellitus and healthy controls. Ann NY Acad Sci. 2004;1037:161-9.
O'Connor GM, Hart OM, Gardiner CM. Putting the natural killer cell in its place. Immunology. 2006;117:1-10.
Older Aguilar AM, Guethlein LA, Abi-Rached L, Parham P. Natural variation at position 45 in the D1 domain of lineage III killer cell immunoglobulin-like receptors (KIR) has major effects on the avidity and specificity for MHC class I. Immunogenetics. 2011;63(8):543-7.
232 Older Aguilar AM, Guethlein LA, Hermes M, Walter L, Parham P. Rhesus macaque KIR bind human MHC class I with broad specificity and recognize HLA-C more effectively than HLA-A and HLA-B. Immunogenetics. 2011;63(9):577-85.
Oliveira CC, van Hall T. Importance of TAP-independent processing pathways. Mol Immunol. 2013;55(2):113-6.
Orange JS, Harris KE, Andzelm MM, Valter MM, Geha RS, Strominger JL. The mature activating natural killer cell immunologic synapse is formed in distinct stages. Proc Natl Acad Sci U S A. 2003;100(24):14151-6.
Ortmann B, Copeman J, Lehner PJ, Sadasivan B, Herberg JA, Grandea AG, Riddell SR, Tampé R, Spies T, Trowsdale J, Cresswell P. A critical role for tapasin in the assembly and function of multimeric MHC class I-TAP complexes. Science. 1997;277(5330):1306-9.
Parham P. Immunology. NK cells lose their inhibition. Science. 2004;305:786-7.
Parham P. MHC class I molecules and KIRs in human history, health and survival. Nat Rev Immunol. 2005;5:201-14.
Parham P, Lomen CE, Lawlor DA, Ways JP, Holmes N, Coppin HL, Salter RD, Wan AM, Ennis PD. Nature of polymorphism in HLA-A, -B, and -C molecules. Proc Natl Acad Sci U S A. 1988;85(11):4005-9.
Parham P, Moffett A., Variable NK cell receptors and their MHC class I ligands in immunity, reproduction and human evolution. Nat Rev Immunol. 2013;13(2):133-44.
Parker, K. C., M. A. Bednarek, and J. E. Coligan. Scheme for ranking potential HLA-A2 binding peptides based on independent binding of individual peptide side-chains. J. Immunol. 1994;152:163. http://www-bimas.cit.nih.gov/molbio/hla_bind/.
Pascal V, Yamada E, Martin MP, Alter G, Altfeld M, Metcalf JA, Baseler MW, Adelsberger JW, Carrington M, Anderson SK, McVicar DW. Detection of KIR3DS1 on the cell surface of peripheral blood NK cells facilitates identification of a novel null allele and assessment of KIR3DS1 expression during HIV-1 infection. J Immunol. 2007;179(3):1625- 33.
Pellett F, Siannis F, Vukin I, Lee P, Urowitz MB, Gladman DD. KIRs and autoimmune disease: studies in systemic lupus erythematosus and scleroderma. Tissue Antigens. 2007;69 Suppl 1:106-8.
Pende D, Biassoni R, Cantoni C, Verdiani S, Falco M, di Donato C, Accame L, Bottino C, Moretta A, Moretta L. The natural killer cell receptor specific for HLA-A allotypes: a novel member of the p58/p70 family of inhibitory receptors that is characterized by three immunoglobulin-like domains and is expressed as a 140-kD disulphide-linked dimer. J Exp Med. 1996;184(2):505-18.
Pende D, Marcenaro S, Falco M, Martini S, Bernardo ME, Montagna D, Romeo E, Cognet C, Martinetti M, Maccario R, Mingari MC, Vivier E, Moretta L, Locatelli F, Moretta A. Anti-leukemia activity of alloreactive NK cells in KIR ligand-mismatched
233 haploidentical HSCT for pediatric patients: evaluation of the functional role of activating KIR and redefinition of inhibitory KIR specificity. Blood. 2009;113(13):3119-29.
Perussia B, Chen Y, Loza MJ. Peripheral NK cell phenotypes: multiple changing of faces of an adapting, developing cell. Mol Immunol. 2005;42(4):385-95.
Peruzzi M, Parker KC, Long EO, Malnati MS. Peptide sequence requirements for the recognition of HLA-B*2705 by specific natural killer cells. J Immunol. 1996;157(8):3350-6.
Peruzzi M, Wagtmann N, Long EO. A p70 killer cell inhibitory receptor specific for several HLA-B allotypes discriminates among peptides bound to HLA-B*2705. J Exp Med. 1996;184(4):1585-90.
Pessino A, Sivori S, Bottino C, Malaspina A, Morelli L, Moretta L, Biassoni R, Moretta A. Molecular cloning of NKp46: a novel member of the immunoglobulin superfamily involved in triggering of natural cytotoxicity. J Exp Med. 1998;188(5):953-6.
Petrie EJ, Clements CS, Lin J, Sullivan LC, Johnson D, Huyton T, Heroux A, Hoare HL, Beddoe T, Reid HH, Wilce MC, Brooks AG, Rossjohn J. CD94-NKG2A recognition of human leukocyte antigen (HLA)-E bound to an HLA class I leader sequence. J Exp Med. 2008;205(3):725-35.
Piccioli D, Sbrana S, Melandri E, Valiante NM. Contact-dependent stimulation and inhibition of dendritic cells by natural killer cells. J Exp Med. 2002;195(3):335-41.
Ponte M, Cantoni C, Biassoni R, Tradori-Cappai A, Bentivoglio G, Vitale C, Bertone S, Moretta A, Moretta L, Mingari MC. Inhibitory receptors sensing HLA-G1 molecules in pregnancy: decidua-associated natural killer cells express LIR-1 and CD94/NKG2A and acquire p49, an HLA-G1-specific receptor. Proc Natl Acad Sci U S A. 1999;96(10):5674-9.
Purcell AW, Elliott T. Molecular machinations of the MHC-I peptide loading complex. Curr Opin Immunol. 2008;20(1):75-81.
Pyo CW, Guethlein LA, Vu Q, Wang R, Abi-Rached L, Norman PJ, Marsh SG, Miller JS, Parham P, Geraghty DE. Different patterns of evolution in the centromeric and telomeric regions of group A and B haplotypes of the human killer cell Ig-like receptor locus. PLoS One. 2010;5(12).
Rajagopalan S, Fu J, Long EO. Cutting edge: induction of IFN-gamma production but not cytotoxicity by the killer cell Ig-like receptor KIR2DL4 (CD158d) in resting NK cells. J Immunol. 2001;167(4):1877-81.
Rajagopalan S, Long EO. Antagonizing inhibition gets NK cells going. Proc Natl Acad Sci U S A. 2010;107(23):10333-4.
Rajagopalan S, Long EO. KIR2DL4 (CD158d): An activation receptor for HLA-G. Front Immunol. 2012;20;3:258
Rajagopalan S, Long EO. The direct binding of a p58 killer cell inhibitory receptor to human histocompatibility leukocyte antigen (HLA)-Cw4 exhibits peptide selectivity. J Exp Med. 1997;185(8):1523-8.
234 Rajagopalan S, Long EO. Found: a cellular activating ligand for NKp44. Blood. 2013; 122(17):2921-2.
Hans-Georg Rammensee, Jutta Bachmann, Niels Nikolaus Emmerich, Oskar Alexander Bachor, Stefan Stevanovic: SYFPEITHI: database for MHC ligands and peptide motifs. Immunogenetics. 1999;50:213-219 (access via :www.syfpeithi.de).
Raulet DH, Vance RE. Self-tolerance of natural killer cells. Nat Rev Immunol. 2006;6(7):520-31.
Redman CW, Sargent IL. Latest advances in understanding preeclampsia. Science. 2005;308(5728):1592-4.
Reed KE, Rice CM. Overview of hepatitis C virus genome structure, polyprotein processing, and protein properties. Curr Top Microbiol Immunol. 2000;242:55-84.
Rehermann B, Nascimbeni M. Immunology of hepatitis B virus and hepatitis C virus infection. Nat Rev Immunol. 2005;5:215-29.
Robinson J, Halliwell JA, Marsh SG. IMGT/HLA and the Immuno Polymorphism Database. Methods Mol Biol. 2014;1184:109-21.
Robinson J, Halliwell JA, McWilliam H, Lopez R, Marsh SG. IPD--the Immuno Polymorphism Database. Nucleic Acids Res. 2013;41 (Database issue):D1234-40.
Robinson J, Waller MJ, Parham P, de Groot N, Bontrop R, Kennedy LJ, Stoehr P, Marsh SGE. IMGT/HLA and IMGT/MHC: sequence databases for the study of the major histocompatibility complex. Nucleic Acids Research. 2003;31:311-314.
Romero V, Azocar J, Zuniga J, Clavijo OP, Terreros D, Gu X, Husain Z, Chung RT, Amos C, Yunis EJ. Interaction of NK inhibitory receptor genes with HLA-C and MHC class II alleles in Hepatitis C virus infection outcome. Mol Immunol. 2008;45(9):2429-36.
Rosenberg SA, Lotze MT, Muul LM, Leitman S, Chang AE, Ettinghausen SE, Matory YL, Skibber JM, Shiloni E, Vetto JT, et al. Observations on the systemic administration of autologous lymphokine-activated killer cells and recombinant interleukin-2 to patients with metastatic cancer. N Engl J Med. 1985;313(23):1485-92.
Ruggeri L, Capanni M, Urbani E, Perruccio K, Shlomchik WD, Tosti A, Posati S, Rogaia D, Frassoni F, Aversa F, Martelli MF, Velardi A. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science. 2002;295(5562):2097-100.
Sanjuán R, Nebot MR, Chirico N, Mansky LM, Belshaw R. Viral mutation rates. J Virol. 2010;84(19):9733-48.
Saulquin X, Gastinel LN, Vivier E. Crystal structure of the human natural killer cell activating receptor KIR2DS2 (CD158j). J Exp Med. 2003;197(7):933-8.
Schonberg K, Sribar M, Enczmann J, Fischer JC, Uhrberg M. Analyses of HLA-C- specific KIR repertoires in donors with group A and B haplotypes suggest a ligand-instructed model of NK cell receptor acquisition. Blood. 2011;117(1):98-107.
235 Schubert U, Anton LC, Gibbs J, Norbury CC, Yewdell JW, Bennink JR. Rapid degradation of a large fraction of newly synthesized proteins by proteasomes. Nature. 2000;404(6779):770-4.
Screpanti V, Wallin RP, Ljunggren HG, Grandien A. A central role for death receptor- mediated apoptosis in the rejection of tumors by NK cells. J Immunol. 2001;167(4):2068-73.
Seich Al Basatena NK,Macnamara A, Vine AM, Thio CL, Astemborski J, Usuku K, Osame M, Kirk GD, Donfield SM, Goedert JJ, Bangham CR, Carrington M, Khakoo SI, Asquith B. KIR2DL2 enhances protective and detrimental HLA class I-mediated immunity in chronic viral infection. PLoS Pathog. 2011;7(10).
Seifert U, Bialy LP, Ebstein F, Bech-Otschir D, Voigt A, Schröter F, Prozorovski T, Lange N, Steffen J, Rieger M, Kuckelkorn U, Aktas O, Kloetzel PM, Krüger E. Immunoproteasomes preserve protein homeostasis upon interferon-induced oxidative stress. Cell. 2010;142(4):613-24.
Selvakumar A, Steffens U, Dupont B. NK cell receptor gene of the KIR family with two IG domains but highest homology to KIR receptors with three IG domains. Tissue Antigens. 1996;48(4 Pt 1):285-94.
Selvakumar A, Steffens U, Dupont B. Polymorphism and domain variability of human killer cell inhibitory receptors. Immunol Rev. 1997;155:183-96.
Sharkey AM, Gardner L, Hiby S, Farrell L, Apps R, Masters L, Goodridge J, Lathbury L, Stewart CA, Verma S, Moffett A. Killer Ig-like receptor expression in uterine NK cells is biased toward recognition of HLA-C and alters with gestational age. J Immunol. 2008 1;181(1):39-46.
Shastry A, Sedimbi SK, Rajalingam R, Rumba I, Kanungo A, Sanjeevi CB. Different KIRs confer susceptibility and protection to adults with latent autoimmune diabetes in Latvian and Asian Indian populations. Ann N Y Acad Sci. 2008;1150:133-8.
Shastry A, Sedimbi SK, Rajalingam R, Nikitina-Zake L, Rumba I, Wigzell H, Sanjeevi CB. Combination of KIR 2DL2 and HLA-C1 (Asn 80) confers susceptibility to type 1 diabetes in Latvians. Int J Immunogenet. 2008;35(6):439-46.
Shi Y, Smith KD, Kurilla MG, Lutz CT. Cytotoxic CD8+ T cells recognize EBV antigen but poorly kill autologous EBV-infected B lymphoblasts: immunodominance is elicited by a peptide epitope that is presented at low levels in vitro. J Immunol. 1997;159(4):1844-52.
Shifrin N, Raulet DH, Ardolino M. NK cell self tolerance, responsiveness and missing self recognition. Semin Immunol. 2014;26(2):138-44.
Simmonds P. Genetic diversity and evolution of hepatitis C virus--15 years on. J Gen Virol. 2004;85(11):3173-88.
Simmonds P, Bukh J, Combet C, Deléage G, Enomoto N, Feinstone S, Halfon P, Inchauspé G, Kuiken C, Maertens G, Mizokami M, Murphy DG, Okamoto H, Pawlotsky JM, Penin F, Sablon E, Shin-I T, Stuyver LJ, Thiel HJ, Viazov S, Weiner AJ, Widell A.
236 Consensus proposals for a unified system of nomenclature of hepatitis C virus genotypes. Hepatology. 2005;42(4):962-73.
Single RM, Martin MP, Gao X, Meyer D, Yeager M, Kidd JR, Kidd KK, Carrington M. Global diversity and evidence for coevolution of KIR and HLA. Nat Genet. 2007;39(9):1114- 9.
Sivori S, Parolini S, Falco M, Marcenaro E, Biassoni R, Bottino C, Moretta L, Moretta A. 2B4 functions as a co-receptor in human NK cell activation. Eur J Immunol. 2000;30(3):787-93.
Snyder GA, Brooks AG, Sun PD. Crystal structure of the HLA-Cw3 allotype-specific killer cell inhibitory receptor KIR2DL2. Proc Natl Acad Sci U S A. 1999;96(7):3864-9.
Standeven LJ, Carlin LM, Borszcz P, Davis DM, Burshtyn DN. The actin cytoskeleton controls the efficiency of killer Ig-like receptor accumulation at inhibitory NK cell immune synapses. J Immunol. 2004;173(9):5617-25.
Stewart CA, Laugier-Anfossi F, Vély F, Saulquin X, Riedmuller J, Tisserant A, Gauthier L, Romagné F, Ferracci G, Arosa FA, Moretta A, Sun PD, Ugolini S, Vivier E. Recognition of peptide–MHC class I complexes by activating killer immunoglobulin-like receptors. Proc Natl Acad Sci U S A. 2005;102(37):13224-9.
Sun JC, Beilke JN, Lanier LL. Adaptive immune features of natural killer cells. Nature. 2009;457(7229):557-61.
Takamizawa A, Mori C, Fuke I, Manabe S, Murakami S, Fujita J, Onishi E, Andoh T, Yoshida I, Okayama H. Structure and organization of the hepatitis C virus genome isolated from human carriers. J Virol. 1991;65(3):1105-13.
Thananchai H, Gillespie G, Martin MP, Bashirova A, Yawata N, Yawata M, Easterbrook P, McVicar DW, Maenaka K, Parham P, Carrington M, Dong T, Rowland-Jones S. Cutting Edge: Allele-specific and peptide-dependent interactions between KIR3DL1 and HLA-A and HLA-B. J Immunol. 2007;178(1):33-7.
Thimme R, Bukh J, Spangenberg HC, Wieland S, Pemberton J, Steiger C, Govindarajan S, Purcell RH, Chisari FV. Viral and immunological determinants of hepatitis C virus clearance, persistence, and disease. Proc Natl Acad Sci U S A. 2002;99(24):15661- 8.
Thimme R, Oldach D, Chang KM, Steiger C, Ray SC, Chisari FV. Determinants of viral clearance and persistence during acute hepatitis C virus infection. J Exp Med. 2001;194(10):1395-406.
Thio CL, Gao X, Goedert JJ, Vlahov D, Nelson KE, Hilgartner MW, O'Brien SJ, Karacki P, Astemborski J, Carrington M, Thomas DL. HLA-Cw*04 and hepatitis C virus persistence. J Virol. 2002;76(10):4792-7.
Tiemessen CT, Paximadis M, Minevich G, Winchester R, Shalekoff S, Gray GE, Sherman GG, Coovadia AH, Kuhn L. Natural killer cell responses to HIV-1 peptides are
237 associated with more activating KIR genes and HLA-C genes of the C1 allotype. J Acquir Immune Defic Syndr. 2011;57(3):181-9.
Tomasello E, Yessaad N, Gregoire E, Hudspeth K, Luci C, Mavilio D, Hardwigsen J, Vivier E. Mapping of NKp46(+) Cells in Healthy Human Lymphoid and Non-Lymphoid Tissues. Front Immunol. 2012;3:344.
Trowsdale, J. Genetic and functional relationships between MHC and NK receptor genes. Immunity. 2001;15:363-374.
Trundley A, Frebel H, Jones D, Chang C, Trowsdale J. Allelic expression patterns of KIR3DS1 and 3DL1 using the Z27 and DX9 antibodies. Eur J Immunol. 2007;37(3):780-7.
Trowsdale J, Moffett A. NK receptor interactions with MHC class I molecules in pregnancy. Semin Immunol. 2008;20(6):317-20.
Tsai SL, Chen MH, Yeh CT, Chu CM, Lin AN, Chiou FH, Chang TH, Liaw YF. Purification and characterization of a naturally processed hepatitis B virus peptide recognized by CD8+ cytotoxic T lymphocytes. J Clin Invest. 1996;97(2):577-84.
Tseng CT, Klimpel GR. Binding of the hepatitis C virus envelope protein E2 to CD81 inhibits natural killer cell functions. J Exp Med. 2002;195:43-9.
Tsomides TJ, Aldovini A, Johnson RP, Walker BD, Young RA, Eisen HN. Naturally processed viral peptides recognized by cytotoxic T lymphocytes on cells chronically infected by human immunodeficiency virus type 1. J Exp Med. 1994;180(4):1283-93.
Uellner R, Zvelebil MJ, Hopkins J, Jones J, MacDougall LK, Morgan BP, Podack E, Waterfield MD, Griffiths GM. Perforin is activated by a proteolytic cleavage during biosynthesis which reveals a phospholipid-binding C2 domain. EMBO J. 1997;16(24):7287- 96.
Uhrberg M, Valiante NM, Shum BP, Shilling HG, Lienert-Weidenbach K, Corliss B, Tyan D, Lanier LL, Parham P. Human diversity in killer cell inhibitory receptor genes. Immunity. 1997;7(6):753-63.
Uhrberg M, Parham P, Wernet P. Definition of gene content for nine common group B haplotypes of the Caucasoid population: KIR haplotypes contain between seven and eleven KIR genes. Immunogenetics. 2002;54(4):221-9.
Uhrberg M. The KIR gene family: life in the fast lane of evolution. Eur J Immunol. 2005;35(1):10-5.
Vales-Gomez M, Reyburn HT, Erskine RA, Strominger J. Differential binding to HLA-C of p50-activating and p58-inhibitory natural killer cell receptors. Proc Natl Acad Sci U S A. 1998;95(24):14326-31.
Valiante NM, Uhrberg M, Shilling HG, Lienert-Weidenbach K, Arnett KL, D'Andrea A, Phillips JH, Lanier LL, Parham P. Functionally and structurally distinct NK cell receptor repertoires in the peripheral blood of two human donors. Immunity. 1997;7(6):739-51.
238 Vance RE, Kraft JR, Altman JD, Jensen PE, Raulet DH. Mouse CD94/NKG2A is a natural killer cell receptor for the nonclassical major histocompatibility complex (MHC) class I molecule Qa-1(b). J Exp Med. 1998;188(10):1841-8.
VandenBussche CJ, Dakshanamurthy S, Posch PE, Hurley CK. A single polymorphism disrupts the killer Ig-like receptor 2DL2/2DL3 D1 domain. J Immunol. 2006;177(8):5347-57. van Els CA, Herberts CA, van der Heeft E, Poelen MC, van Gaans-van den Brink JA, van der Kooi A, Hoogerhout P, Jan ten Hove G, Meiring HD, de Jong AP. A single naturally processed measles virus peptide fully dominates the HLA-A*0201-associated peptide display and is mutated at its anchor position in persistent viral strains. Eur J Immunol. 2000;30(4):1172-81. van der Slik AR, Koeleman BP, Verduijn W, Bruining GJ, Roep BO, Giphart MJ. KIR in type 1 diabetes: disparate distribution of activating and inhibitory natural killer cell receptors in patients versus HLA-matched control subjects. Diabetes. 2003;52(10):2639-42.
Velardi A. Natural killer cell alloreactivity 10 years later. Curr Opin Hematol. 2012;19(6):421- 6.
Vidal-Castineira JR, Lopez-Vazquez A, Diaz-Pena R, Alonso-Arias R, Martinez-Borra J, Perez R, Fernandez-Suarez J, Melon S, Prieto J, Rodrigo L, Lopez-Larrea C. Effect of killer immunoglobulin-like receptors in the response to combined treatment in patients with chronic hepatitis C virus infection. J Virol. 2010;84(1):475-81.
Vilches C, Gardiner CM, Parham P. Gene structure and promoter variation of expressed and nonexpressed variants of the KIR2DL5 gene. J Immunol. 2000;165(11):6416-21.
Vilches C, Parham P. KIR: diverse, rapidly evolving receptors of innate and adaptive immunity. Annu Rev Immunol. 2002;20:217-51.
Vitale M, Bottino C, Sivori S, Sanseverino L, Castriconi R, Marcenaro E, Augugliaro R, Moretta L, Moretta A. NKp44, a novel triggering surface molecule specifically expressed by activated natural killer cells, is involved in non-major histocompatibility complex-restricted tumor cell lysis. J Exp Med. 1998;187(12):2065-72.
Vitale M, Castriconi R, Parolini S, Pende D, Hsu ML, Moretta L, Cosman D, Moretta A. The leukocyte Ig-like receptor (LIR)-1 for the cytomegalovirus UL18 protein displays a broad specificity for different HLA class I alleles: analysis of LIR-1 + NK cell clones. Int Immunol. 1999;11(1):29-35.
Vita R, Zarebski L, Greenbaum JA, Emami H, Hoof I, Salimi N, Damle R, Sette A, Peters B. The immune epitope database 2.0. Nucleic Acids Res. 2010;38(Database issue):D854-62. Epub 2009 Nov 11.
Vivian JP, Duncan RC, Berry R, O'Connor GM, Reid HH, Beddoe T, Gras S, Saunders PM, Olshina. Killer cell immunoglobulin-like receptor 3DL1-mediated recognition of human leukocyte antigen B. Nature. 2011;479(7373):401-5.
Vivier E, Daëron M. Immunoreceptor tyrosine-based inhibition motifs. Immunol Today. 1997;18(6):286-91.
239 Vivier E, Nunès JA, Vély F. Natural killer cell signaling pathways. Science. 2004;306:1517- 9.
Vivier E,Tomasello E, Paul P. Lymphocyte activation via NKG2D: towards a new paradigm in immune recognition? Curr Opin Immunol. 2002;14(3):306-11.
Vivier E, Tomasello E, Baratin M, Walzer T, Ugolini S. Functions of natural killer cells. Nat Immunol. 2008;9:503-10.
Vivier E. Innate or adaptive immunity? The example of natural killer cells. Science. 2011;331(6013):44-9.
Wagtmann N, Rajagopalan S, Winter CC, Peruzzi M, Long EO. Killer cell inhibitory receptors specific for HLA-C and HLA-B identified by direct binding and by functional transfer. Immunity. 1995;3(6):801-9.
Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, Vitale M, Bottino C, Moretta L, Moretta A, Long EO. Molecular clones of the p58 NK cell receptor reveal immunoglobulin-related molecules with diversity in both the extra- and intracellular domains. Immunity. 1995;2(5):439-49.
Watzl C, Claus M. WhatSAP - 2B4 sends mixed messages in the absence of SAP. Eur J Immunol. 2014;44(5):1281-4.
Williams AP, Peh CA, Purcell AW, McCluskey J, Elliott T. Optimization of the MHC class I peptide cargo is dependent on tapasin. Immunity. 2002;16(4):509-20.
Williams F, Meenagh A, Sleator C, Cook D, Fernandez-Vina M, Bowcock AM, Middleton D. Activating killer cell immunoglobulin-like receptor gene KIR2DS1 is associated with psoriatic arthritis. Hum Immunol. 2005;66(7):836-41.
Wilson MJ, Torkar M, Haude A, Milne S, Jones T, Sheer D, Beck S, Trowsdale J. Plasticity in the organization and sequences of human KIR/ILT gene families. Proc Natl Acad Sci U S A. 2000;97(9):4778-83.
Winter CC, Gumperz JE, Parham P, Long EO, Wagtmann N. Direct binding and functional transfer of NK cell inhibitory receptors reveal novel patterns of HLA-C allotype recognition. J Immunol. 1998;161(2):571-7.
Winter C and Long E. A single amino acid in the p58 killer cell inhibitory receptor controls the ability of natural killer cells to discriminate between the two groups of HLA-C allotypes. J Immunol. 1997;158(9):4026-8.
Wong-Baeza I, Ridley A, Shaw J, Hatano H, Rysnik O, McHugh K, Piper C, Brackenbridge S, Fernandes R, Chan A, Bowness P, Kollnberger S. KIR3DL2 binds to HLA-B27 dimers and free H chains more strongly than other HLA class I and promotes the expansion of T cells in ankylosing spondylitis. J Immunol. 2013;190(7):3216-24.
Xiong S, Sharkey AM, Kennedy PR, Gardner L, Farrell LE, Chazara O, Bauer J, Hiby SE, Colucci F, Moffett A. Maternal uterine NK cell-activating receptor KIR2DS1 enhances placentation. J Clin Invest. 2013;123(10):4264-72.
240 Yen JH, Moore BE, Nakajima T, Scholl D, Schaid DJ, Weyand CM, Goronzy JJ. Major histocompatibility complex class I-recognizing receptors are disease risk genes in rheumatoid arthritis. J Exp Med. 2001;193(10):1159-67.
Yewdell JW. DRiPs solidify: progress in understanding endogenous MHC class I antigen processing. Trends Immunol. 2011;32(11):548-58.
Yewdell JW, Antón LC, Bennink JR. Defective ribosomal products (DRiPs): a major source of antigenic peptides for MHC class I molecules? J Immunol. 1996;157:1823-6.
Yewdell JW. Immunoproteasomes: regulating the regulator.Proc Natl Acad Sci U S A. 2005;102(26):9089-90.
Yewdell JW. Confronting complexity: real-world immunodominance in antiviral CD8+ T cell responses. Immunity. 2006;25(4):533-43.
Yewdell JW, Nicchitta CV. The DRiP hypothesis decennial: support, controversy, refinement and extension.Trends Immunol. 2006;27:368-73.
Yewdell JW, Reits E, Neefjes J. Making sense of mass destruction: quantitating MHC class I antigen presentation. Nat Rev Immunol. 2003;3(12):952-61.
Yewdell JW, Antón LC, Bennink JR. Defective ribosomal products (DRiPs): a major source of antigenic peptides for MHC class I molecules? J Immunol. 1996;157(5):1823-6.
Yewdell JW, Schubert U, Bennink JR. At the crossroads of cell biology and immunology: DRiPs and other sources of peptide ligands for MHC class I molecules. J Cell Sci. 2001;114(5):845-51.
Young NT, Mulder A, Cerundolo V, Claas FH, Welsh KI. Expression of HLA class I antigens in transporter associated with antigen processing (TAP)-deficient mutant cell lines. Tissue Antigens. 1998;52(4):368-73.
Zappacosta F, Borrego F, Brooks AG, Parker KC, Coligan JE. Peptides isolated from HLA-Cw*0304 confer different degrees of protection from natural killer cell-mediated lysis. Proc Natl Acad Sci U S A. 1997;94(12):6313-8.
241 Appendix I
Primary HCV isolate with peptides highlighted in red (Itakura et al, 2005) gi|18027684| Genbank Accession AF313916.1|Hepatitis C virus polyprotein mRNA, partial cds
>MSTNPKPQRKTKRNTNRRPQDVKFPGGGQIVGGVYLLPRRGPRLGVRATRKTSERSQPRGRRQPIPKARRPE GRTWAQPGYPWPLYGNEGMGWAGWLLSPRGSRPSWGPTDPRRRSRNLGKVIDTLTCGFADLMGYIPLVGAPL GGAARALAHGVRVLEDGVNYATGNLPGCSFSIFLLALLSCLTIPASAIEVRNVSGVYHVTNDCSNASIVYEAADMIM HTPGCVPCVRENNSSRCWVALTPTLAARNSSIPTTTIRRHVDLLVGTAAFCSAMYVGDLCGSVFLVSQLFTFSPR RHETVQDCNCSIYPGHVSGHRMAWDMMMNWSPTTALVVSQLLRIPQAVVDMVAGAHWGVLAGLAYYSMAGN WAKVLIVMLLFAGVDGETYTTGGAQAHTTRGLATLFSFGPSQNIQLINTNGSWHINRTALNCNDSLNTGFLAALFY AHRFNASGCPERMASCRPIDKFAQGWGPIAYAKPLSLDQKPYCWHYAPQPCGIVPASQVCGPVYCFTPSPVVV GTTDRFGVPTYSWGENETDVLLLNNTRPPRGNWFGCTWMNGTGFTKTCGGPPCDIGGVGNDTLICPTDCFRKH PEATYTKCGSGPWLTPRCMVDYPYRLWHYPCTVNFTIFKVRMYVGGVEHRLNAACNWTRGERCDLEDRDRSEL SPLLLSTTEWQVLPCSFTTLPALSTGLIHLHRNIVDVQYLYGIGSAVVSYAIKWEYVLLLFLFLADARVCACLWMML LIVQAEAALENLVVLNAASVAGTHGILSFLVFFCAAWYIKGRLVPGAAYAFYGVWPLLLLLLALPPRAYAMDREMA ASCGGAVFIGLALLTLSPHYKVFLARLIWWLQYFITRAEAHLQVWIPPLNVRGGRDAIILLMCAIHPELIFAITKILLAIL GPLMVLQAGITRVPYFVRAHGLIRACVLVREVAGGHYVQMALMRLAALTGTYVYDHLTPLRDWAHEGLRDLAVA VEPVVFSDMETKIITWGADTAACGDIILGLPVSARRGREILLGPADSLEGQGWRLLAPITAYAQQTRGLLGCIVTSL TGRDKNQVEGEVQVVSTATQSFLATCVNGVCWTVYHGAGTKTLAGQKGPITQMYTNVDQDLVGWLAPPGARSL TPCTCGSSDLYLVTRHADVIPVRRRGDSRGSLLSPRPVSYLKGSSGGPLLCPSGHAVGIFRAAVCTRGVAKAVDF VPVESMETTMRSPVFTDNSSPPAVPQTFQVAHLHAPTGSGKSTKVPAAYAAQGYKVLVLNPSVAATLGFGAYMS KAHGVDPNIRTGVRTITTGAPITYSTYGKFLADGGCSGGAYDIIICDECHSTDSTTILGIGTVLDQAETAGARLVVLA TATPPGSVTVPHPNIEEVALSNIGEIPFYGKAIPIETIKGGRHLIFCHSRKKCDELAAKLSGLGLNAVAYYRGLDVSVI PTSGDVVVVATDALMTGFTGDFDSVIDCNTCVTQTVDFSLDPTFTIETTTVPQDAVSRSQRRGRTGRGRTGIYRF VTPGERPSGMFDSSVLCECYDAGCAWYELTPAETSVRLRAYLNTPGLPVCQDHLEFWEGVFTGLTHIDAHFLSQ TKQAGDNFPYLVAYQATVCARAQAPPPSWDQMWKCLIRLKPTLHGPTPLLYRLGAVQNEVILTHPITKFIMACMS ADLEVATSTWVLVGGVLAALAAYCLTTGSVVIVGRIILSGKPAIVPDREVLYQQFDEMEECATHLPYIEQGMQLAE QFKQKALGLLQTATKQAEAAAPVVESKWRALESFWAKHMWNFISGIQYLAGLSTLPGNPAIASLMAFTASITSPLT TQHTLLFNILGGWVAAQLAPPSAASAFVGAGIAGAAVGSIGLGKVLVDILAGYGAGVAGALVAFKVMSGEMPSTE DLVNLLPAILSPGALVVGVVCAAILRRHVGPGEGAVQWMNRLIAFASRGNHVSPTHYVPESDAAARVTQILSSLTIT QLLKRLHQWINEDCSTPCSGSWLRDVWDWICTVLTDFKTWLQSKLLPRLPGVPFLSCQRGYKGVWRGDGIMQT TCPCGAQITGHVKNGSMRIVGPRTCSNTWHGTFPINAYTTGPCTPSPAPNYSRALWRVTAEEYLEVTQVGDFHY VTGMTTDNVKCPCQVPAPEFFTEVDGVRLHRYAPPCKPLLREEVTFQVGLNQYLVGSQLPCEPEPDVAVLTSML TDPSHITAEAAKRRLARGSPPSLASSSASQLSAPSLKATCTTRHDSPDADLIEANLLWRQEMGGNITRVESENKVV ILDSFEPLRAEEDEREVSVPAEILRRSRKFPRAMPIWARPDYNPPLVESWKDPDYVPPVVHGCPLPPTKAPPIPPP RRKRTVVLTESTVSSALAELATKTFGSSESSAVDSGTATAPPDQPSDDGDAGSDVESYSSMPPLEGEPGDPDLS DGSWSTVSEEASEDVVCCSMSYTWTGALITPCSAEESKLPINALSNSLLRHHNMVYATTSRSASQRQKKVTFDRL QVLDDHYRDVLKEMKAKASTVKAKLLPVEEACKLTPPHSAKSKFGYGAKDVRNLSSKAVNHILSVWKDLLEDTET PIDTTIMAKNEVFCVQPEKGGRKPARLIVFPDLGVRVCEKMALYDVVSTLPQAVMGSSYGFQYSPGQRVEFLVNA WKKKKNPMGFAYDTRCFDSTVTESDIRVEESIYQCCDLAPEARQAIRSLTERLYIGGPLTNSKGQNCGYRRCRAS GVLTTSCGNTLTCYLKASAACRAAKLQDCTMLVCGDDLVVICESAGTXEDAASLRVFTEAMTRYSAPPGDPPKPE YDLELITSCSSNVSVAHDASGKRVYYLTRDPTTPLARAAWETARHTPVNSWLGNIIMYAPTLWARMILMTHFFSILL AQEQLEKALDCQIYGACYSIEPLDLPQIIQRLHGLSAFSLHSYSPGEINRVAACLRKLGVPPLRVWRHRARSVRAKL LSQGGRAANCGKYLFNWAVRTKLKLTPIPAASQLDLSSWFVAGYSGGDIYHSLSRARPRWFMLCLLLLSVGVGIY LLPNR
242
List of HCV Isolate Genbank Accession numbers (Itakura et al, 2005)
gi|56342184|Genbank AB154177.1| isolate: No. 2 gi|56342186|Genbank AB154178.1| isolate: No. 3 gi|56342188|Genbank AB154179.1| isolate: No. 4 gi|56342190| Genbank AB154180.1| isolate: No. 5 gi|56342192| Genbank AB154181.1| isolate: No. 6 gi|56342194|Genbank AB154182.1| isolate: No. 7 gi|56342196|Genbank AB154183.1| isolate: No. 8 gi|56342198|Genbank AB154184.1| isolate: No. 9 gi|56342200|Genbank AB154185.1| isolate: No. 10 gi|56342202|Genbank AB154186.1| isolate: No. 11 gi|56342204|Genbank AB154187.1| isolate: No. 12 gi|56342206|Genbank AB154188.1| isolate: No. 13 gi|56342208|Genbank AB154189.1| isolate: No. 14 gi|56342210|Genbank AB154190.1| isolate: No. 15 gi|56342212|Genbank AB154191.1| isolate: No. 16 gi|56342214|Genbank AB154192.1| isolate: No. 17 gi|56342216|Genbank AB154193.1| isolate: No. 18 gi|56342218|Genbank AB154194.1| isolate: No. 19 gi|56342220|Genbank AB154195.1| isolate: No. 20 gi|56342222|Genbank AB154196.1| isolate: No. 21 gi|56342224|Genbank AB154197.1| isolate: No. 22 gi|56342226|Genbank AB154198.1| isolate: No. 23 gi|56342228|Genbank AB154199.1| isolate: No. 24 gi|56342230|Genbank AB154200.1| isolate: No. 25 gi|56342232|Genbank AB154201.1| isolate: No. 26 gi|56342234|Genbank AB154202.1| isolate: No. 27 gi|56342236|Genbank AB154203.1| isolate: No. 28 gi|56342238|Genbank AB154204.1| isolate: No. 29 gi|56342240|Genbank AB154205.1|isolate: No. 30
243 ClustalW2 Alignment of Irish HCV cohort sequences (LNPSVAATL HCV NS31253- 1262 peptide)
244 Appendix II
1. Sequence of target gene: Human KIR2DS2
KIR2DS2*00102, complete CDS with extracellular domain highlighted
>gi|38305393|gb|AY366243.1| Homo sapiens natural killer cell inhibitory receptor (KIR2DS2) mRNA, KIR2DS2*00102 allele, complete cds
AGACAGCACCATGTCGCTCATGGTCGTCAGCATGGCGTGTGTTGGGTTCTTCTTGCTGCAGGGGGCCTGGCCACA TGAGGGAGTCCACAGAAAACCTTCCCTCCTGGCCCACCCAGGTCCCCTGGTGAAATCAGAAGAGACAGTCATCCT GCAATGTTGGTCAGATGTCAGGTTTGAGCACTTCCTTCTGCACAGAGAGGGGAAGTATAAGGACACTTTGCACCT CATTGGAGAGCACCATGATGGGGTCTCCAAGGCCAACTTCTCCATCGGTCCCATGATGCAAGACCTTGCAGGGAC CTACAGATGCTACGGTTCTGTTACTCACTCCCCCTATCAGTTGTCAGCTCCCAGTGACCCTCTGGACATCGTCAT CACAGGTCTATATGAGAAACCTTCTCTCTCAGCCCAGCCGGGCCCCACGGTTTTGGCAGGAGAGAGCGTGACCTT GTCCTGCAGCTCCCGGAGCTCCTATGACATGTACCATCTATCCAGGGAGGGGGAGGCCCATGAACGTAGGTTCTC TGCAGGGCCCAAGGTCAACGGAACATTCCAGGCCGACTTTCCTCTGGGCCCTGCCACCCACGGAGGAACCTACAG ATGCTTCGGCTCTTTCCGTGACTCTCCCTATGAGTGGTCAAACTCGAGTGACCCACTGCTTGTTTCTGTCACAGG AAACCCTTCAAATAGTTGGCCTTCACCCACTGAACCAAGCTCCAAAACCGGTAACCCCAGACACCTACATGTTCT GATTGGGACCTCAGTGGTCAAAATCCCTTTCACCATCCTCCTCTTCTTTCTCCTTCATCGCTGGTGCTCCAACAA AAAAAATGCTGCTGTAATGGACCAAGAGCCTGCAGGGAACAGAACAGTGAACAGCGAGGATTCTGATGAACAAGA CCATCAGGAGGTGTCATACGCATAATTGGATCACTGTGTTTTCACACAGAGAGAAATCACTCGCCCTTCTGAGAG GCCCAAGACACCCCCAACAGATACCAGCATGTACATAGAACTTCCAAATGCTGAGCCCAGATCCAAAGTTGTCTT CTGTCCACGAGCACCACAGTCAGGCCTTGAGGGGATCTTCTAGGGAGACAACAGCCCTGTCTCAAAACCGGGTTG CCAGCTCCCATGTACCAGCAGCTGGAATCTGAAGGCATCAGTCTTCATCTTAGGGCATCGCTCTTCCTCACACCA CGAATCTGAACATGCCTCTCTCTTGCTTACAAATGTCTAAGGTCCCCACTGCCTGCTGGAGAGAAAACACACTCC TTTGCTTAGCCCACAATTCTCCATTTCACTTGACCCCTGCCCACCTCTCCAACCTAACTAGCTTACTTCCTAGTC TACCTGAGGCTGCAATCACACTGAGGAACTC
2. Restriction analysis of extracellular domain: 0 cutters include: BamHI, HindIII, NcoI, NdeI, NotI, BglII, EagI
(Note: 1 cutters include, amongst others: Xho1)
3. Translation of extracellular domain a) Compact (protein only): HEGVHRKPSLLAHPGPLVKSEETVILQCWSDVRFEHFLLHREGKYKDTLHLIGEHHDGVSKANFSIGPMMQDLAG TYRCYGSVTHSPYQLSAPSDPLDIVITGLYEKPSLSAQPGPTVLAGESVTLSCSSRSSYDMYHLSREGEAHERRF SAGPKVNGTFQADFPLGPATHGGTYRCFGSFRDSPYEWSNSSDPLLVSVTGNPSNSWPSPTEPSSKT
b) Including nucleotide sequence: catgagggagtccacagaaaaccttccctcctggcccacccaggtcccctggtgaaatca
H E G V H R K P S L L A H P G P L V K S gaagagacagtcatcctgcaatgttggtcagatgtcaggtttgagcacttccttctgcac
E E T V I L Q C W S D V R F E H F L L H agagaggggaagtataaggacactttgcacctcattggagagcaccatgatggggtctcc
R E G K Y K D T L H L I G E H H D G V S aaggccaacttctccatcggtcccatgatgcaagaccttgcagggacctacagatgctac
245 K A N F S I G P M M Q D L A G T Y R C Y ggttctgttactcactccccctatcagttgtcagctcccagtgaccctctggacatcgtc
G S V T H S P Y Q L S A P S D P L D I V atcacaggtctatatgagaaaccttctctctcagcccagccgggccccacggttttggca
I T G L Y E K P S L S A Q P G P T V L A ggagagagcgtgaccttgtcctgcagctcccggagctcctatgacatgtaccatctatcc
G E S V T L S C S S R S S Y D M Y H L S agggagggggaggcccatgaacgtaggttctctgcagggcccaaggtcaacggaacattc
R E G E A H E R R F S A G P K V N G T F caggccgactttcctctgggccctgccacccacggaggaacctacagatgcttcggctct
Q A D F P L G P A T H G G T Y R C F G S ttccgtgactctccctatgagtggtcaaactcgagtgacccactgcttgtttctgtcaca
F R D S P Y E W S N S S D P L L V S V T ggaaacccttcaaatagttggccttcacccactgaaccaagctccaaaacc
G N P S N S W P S P T E P S S K T
4. Primer Design a) FWD primer (Nco1, optimised for expression)
Original sequence 5’-3’: CATGAGGGAGTCCACAGAAAACCTTCCCT
Overhang NcoI/Start FWD PRIMER: GCAGCACCATGGGACATGAAGGAGTTCACAGAAAACCTTCCCT
M G H E G V H R K P b) REV primer (BamH1, with GP-linker to BirA-tag)
Original 5’-3’sequence: CCTTCACCCACTGAACCAAGCTCCAAAACC
GP BamHI Overhang CONSTRUCT 3’-5’: GAACCAAGCTCCAAAACCGGTCCGGGATCCAGTA E P S S K T G P G S
Overhang BamHI GP end of extracellular domain
REV PRIMER 5’-3’: TACTGGATCCCGGACCGGTTTTGGAGCTTGGTTC
246
5. Final construct (human KIR2DS2-BirA in pET23d)
TAAGAAGGAGATATAATGGGACATGAAGGAGTTCACAGAAAACCTTCCCTCCTGGCCCACCCAG GTCCCCTGGTGAAATCAGAAGAGACAGTCATCCTGCAATGTTGGTCAGATGTCAGGTTTGAGCACTTCCTTCTGC ACAGAGAGGGGAAGTATAAGGACACTTTGCACCTCATTGGAGAGCACCATGATGGGGTCTCCAAGGCCAACTTCT CCATCGGTCCCATGATGCAAGACCTTGCAGGGACCTACAGATGCTACGGTTCTGTTACTCACTCCCCCTATCAGT TGTCAGCTCCCAGTGACCCTCTGGACATCGTCATCACAGGTCTATATGAGAAACCTTCTCTCTCAGCCCAGCCGG GCCCCACGGTTTTGGCAGGAGAGAGCGTGACCTTGTCCTGCAGCTCCCGGAGCTCCTATGACATGTACCATCTAT CCAGGGAGGGGGAGGCCCATGAACGTAGGTTCTCTGCAGGGCCCAAGGTCAACGGAACATTCCAGGCCGACTTTC CTCTGGGCCCTGCCACCCACGGAGGAACCTACAGATGCTTCGGCTCTTTCCGTGACTCTCCCTATGAGTGGTCAA ACTCGAGTGACCCACTGCTTGTTTCTGTCACAGGAAACCCTTCAAATAGTTGGCCTTCACCCACTGAACCAA GCTCCAAAACCGGTCCGGGATCCGGTGGTGGTCTGAACGATATTTTTGAAGCTCAGAAAATCGAA TGGCATTAA
Translation
STOP_EGDIMGHEGVHRKPSLLAHPGPLVKSEETVILQCWSDVRFEHFLLHREGKYKDTLHLIGEHHDGVSKANF SIGPMMQDLAGTYRCYGSVTHSPYQLSAPSDPLDIVITGLYEKPSLSAQPGPTVLAGESVTLSCSSRSSYDMYHL SREGEAHERRFSAGPKVNGTFQADFPLGPATHGGTYRCFGSFRDSPYEWSNSSDPLLVSVTGNPSNSWPSPTEPS SKTGPGSGGGLNDIFEAQKIEWH-
6. KIR2DS2 Primers
FWD_KIR2DS2_tet: GCAGCACCATGGGACATGAAGGAGTTCACAGAAAACCTTCCCT
REV_KIR2DS2- tet: TACTGGATCCCGGACCGGTTTTGGAGCTTGGTTC
247