Natural Killer Cell Subsets and Multiple Sclerosis

Alison Jane Harris

Imperial College London Department of

Thesis submitted for the degree of

1

Declaration of Originality

I hereby declare that the work presented here is my own and that all references to the work of others are accompanied by appropriate citations.

2

ABSTRACT

This thesis addresses the hypothesis that multiple sclerosis (MS) patients have an imbalance in the frequencies of natural killer cell subsets, with fewer immunoregulatory and more proinflammatory NK cells compared to healthy controls. Ten NK cell subsets were defined according to expression of CD56, CD8, CD27 and CD57. Subset frequencies were compared by performing multiparameter flow cytometry on PBMCs from healthy controls and patients with clinically isolated syndrome (CIS) or relapsing-remitting MS. CD56dim CD8+ CD57+ CD27- NK cells had a higher frequency in untreated relapsing-remitting MS patients compared to healthy controls and their frequency was decreased in patients on interferon-β treatment. CD56bright NK cells, particularly the CD8- CD27- subset, had a significantly lower frequency in untreated MS and CIS patients. CD8- CD57- CD27- subsets of both CD56dim and CD56bright NK cells were expanded in interferon-β-treated patients in remission but not relapse. Characteristics of two CD56dim subsets displaying an apparent reciprocal relationship (CD8+ CD57+ CD27- and CD8- CD57- CD27-) were compared by qRT-PCR, flow cytometry and killing assays. CD56dim CD8+ CD57+ CD27- NK cells were more cytotoxic and expressed higher levels of transcripts encoding IFNγ, T-bet and other factors associated with IFNγ signalling. This is consistent with the hypothesis that there is a shift towards a more pro-inflammatory NK cell phenotype in MS patients, which is reversed by interferon-β treatment. In order to determine whether these NK cells exert their effects mainly in the blood or in the central nervous system (CNS), mononuclear cells were extracted from the CNS of Line 7 mice, which develop spontaneous autoimmune demyelination due to expression of a human MS-associated HLA/TCR combination. NK cells were detected in spleen, but not CNS, of these mice, suggesting that their role in this particular disease model is restricted to regulation of other leukocytes outside the CNS.

3

Acknowledgements

I would like to thank my supervisors Prof. Danny Altmann and Dr. Rosemary Boyton for their feedback on this thesis. I am grateful to Danny for giving me the opportunity to work on this project and giving me the freedom to pursue my own ideas and to Rosemary for helping me to focus them in the latter stages of my project.

A big thank you to everyone on the 8th floor who has helped me over the years, to all those who have taught me techniques, collected blood, kept the place running smoothly, discussed ideas and made the department a more fun place to be. In particular, I would like to thank Dr. Fatemah Kamel for her tireless work in collecting patient blood samples. Thank you also to Dr. Richard Nicholas, Dr. Omar Malik and Dr. Paolo Giannetti for providing access to MS and CIS pateints, the nurses at Charing Cross who helped to take the samples and to all the people that donated blood, both patients and controls, without whom this study would not have been possible.

I would also like to extend my gratitude to Anna Ettorre for looking after the FACS Aria II, which was essential for virtually all the experiments I performed, and for helping to set up the machine for cell sorting.

I am grateful to Dr. Dan Lowther who optimised and taught me the Line 7 CNS extraction procedure and to Marie-Laure Aknin, Dr. Debs Chong, Dr. Karen Chu and Kathryn Quigley who assisted in the scoring and processing of these mice.

Finally I would like to thank my friends, especially Nishkala Thiru and Richard Browning, for distracting me when experiments weren’t working and the Medical Research Council for funding me throughout my Masters and PhD.

4

Table of Contents

1. INTRODUCTION ...... 18

1.1 Natural Killer Cells ...... 18

1.1.1 NK Cell Development ...... 18

1.1.2 NK Cell Function is Determined by the Balance of Signals from Activating and Inhibitory Receptors ...... 21

1.1.3 NK Cell Education Determines Responsiveness to Activating Signals 27

1.1.4 NK Cells Form Synapses with Target Cells and Kill Susceptible Targets via Granzyme Release and Other Mechanisms ...... 31

1.1.5 NK Cells Produce a Range of Cytokines and Chemokines ...... 38

1.1.6 NK Cells Interact with Other Leukocytes and Regulate their Function 48

1.1.7 NK Cells can be Divided into Subsets with Distinct Functions ...... 55

1.2 The Role of T cells and NK Cells in Multiple Sclerosis ...... 66

1.2.1 Genetic and Environmental Risk Factors are Associated with Development of MS ...... 67

1.2.2 CD4+ T cells Infiltrate the CNS and Promote Recruitment and Activation of Other Leukocytes, leading to Myelin and Axonal Damage ...... 70

1.2.3 Disease Initiation Requires Activation of Peripheral T Cells – Potential Role of Viruses and Impaired Immunoregulatory Mechanisms ...... 78

1.2.4 Current Therapies for MS have Benefits and Limitations ...... 80

1.2.5 Evidence of a Role for NK Cells in Multiple Sclerosis ...... 86

1.3 Aims ...... 96

2. METHODS ...... 97

2.1 Human PBMC Isolation ...... 97

2.2 Flow Cytometry Evaluation of Human PBMC (Surface Staining) ...... 98

2.3 PBMC Stimulation and Intracellular Cytokine Staining ...... 98

2.4 Cell Sorting for RNA ...... 99

5

2.5 RNA Extraction ...... 100

2.6 Reverse Transcription ...... 101

2.7 Quantitative RT-PCR Arrays ...... 101

2.8 Individual qRT-PCR Assays ...... 103

2.9 K562 Killing Assays ...... 103

2.10 Preparation and Staining of Cells from Line 7 mice ...... 104

2.11 Buffers and Medium: ...... 107

3. RRMS PATIENTS EXHIBIT DIFFERENCES IN FREQUENCIES OF SPECIFIC NK CELL SUBSETS ...... 125

3.1 Expression and Co-expression of CD8, CD27 and CD57 on CD56dim and CD56bright NK Cells ...... 127

3.2 Confirmation of Differences in Frequency of NK Cell Subsets in RRMS Patients and the Effects of IFNβ Treatment ...... 129

3.3 Multiparameter Analysis Reveals Differences in NK Cell Subset Frequencies that are not Restricted to the CD56bright Population ...... 133

3.4 Subset Frequencies in Healthy Adults do not Differ Significantly in Relation to Age or Gender ...... 138

3.5 Discussion ...... 140

4. TRANSCRIPTIONAL AND FUNCTIONAL DIFFERENCES BETWEEN CD56DIM CD8- CD27- CD57- AND CD56DIM CD8+ CD27- CD57+ NK CELLS ..... 146

4.1 Identification of Transcriptional Differences between CD8- CD27- CD57- and CD8+ CD57+ CD56dim NK Cell Subsets ...... 147

4.2 Cell Surface Expression of Cytokine Receptors ...... 157

4.3 Differences in Expression of Activating and Inhibitory Receptors ...... 159

4.4 CD56dim CD8- CD27- CD57- NK Cells Produce Similar Levels of IFNγ to CD8+ CD57+ NK Cells and do not make IL-17 or IL-22 ...... 162

4.5 CD56dim CD8+ CD57+ NK cells are more cytotoxic than CD56dim CD8- CD27- CD57- NK Cells ...... 165

4.6 Discussion ...... 167

6

5. NK CELLS DO NOT INFILTRATE THE CNS IN THE LINE7 MURINE MODEL OF MULTIPLE SCLEROSIS ...... 173

5.1.1 NKp46+ NK Cells were Not Detected in the CNS of Line 7 Mice ...... 175

5.1.2 Lack of NKp46 Detection was not due to Antibody Sequestration ..... 178

5.1.3 NKG2D was also Not Detected on CNS-derived Lymphocytes ...... 179

5.1.4 Discussion ...... 181

6. DISCUSSION ...... 185

6.1 Summary ...... 195

7

List of Figures

Figure 1-i: NK cells develop from bone marrow HSCs...... 19 Figure 1-ii: Missing self and stressed self recognition...... 23 Figure 1-iii: NK cell education maintains tolerance of NK cells lacking self-specific inhibitory receptors...... 30 Figure 1-iv: NK cells stimulate T cell activation and Th1 polarisation both directly and indirectly...... 51 Figure 1-v: CD56bright CD16- NKG2A+ NK cells differentiate into CD56dim CD16+ NKG2A- NK cells...... 65 Figure 1-vi: CD4+ T cells initiate CNS inflammation and demyelination...... 73 Figure 2-i: Gating strategy for enumeration of NK cell subsets...... 108 Figure 2-ii: Classification of NK cells into ten subsets...... 109 Figure 2-iii: Sorted NK cell subsets have high purity...... 110 Figure 3-i: Co-expression pattern of NK cell surface markers...... 128 Figure 3-ii: Total NK cell frequency does not differ between MS patients and controls and is not significantly influenced by interferon-β or Tysabri therapy...... 129 Figure 3-iii: Untreated RRMS and CIS patients have a higher proportion of CD8+ compared to CD8- NK cells...... 130 Figure 3-iv: Untreated RRMS patients have a reduced frequency of CD56bright NK cells which is reversed by IFNβ treatment...... 132 Figure 3-v: CD56bright subset frequencies are altered in MS patients...... 134 Figure 3-vi: MS patients have a higher frequency of CD56dim CD8+ CD27- CD57+ NK cells...... 135 Figure 3-vii: Interferon-β alters frequencies of several CD56dim NK cell subsets..... 136 Figure 3-viii: Frequencies of CD56dim subsets are not significantly different in relapse compared to remission...... 137 Figure 3-ix: No significant gender- or age-related differences in the frequency of NK cells or NK cell subsets were detected...... 139 Figure 4-i: Clusters of co-regulated genes are associated with CD56dim CD8- CD27- CD57- and CD8+ CD27- CD57+ NK cell subsets...... 152 Figure 4-ii: Confirmation of transcriptional differences by individual qRT-PCR assays...... 156 Figure 4-v: Expression of cytokine receptors...... 158

8

Figure 4-iii: The CD8+ CD57+ subset contains significantly fewer NKG2A+ NK cells...... 160 Figure 4-iv: Differences in expression of activating receptors...... 161 Figure 4-vi: IFNγ production in the CD56dim CD8- CD27- CD57- subset in response to PMA/ionomycin stimulation is lower than that of CD56bright NK cells and not significantly different to that of the CD56dim CD8+ CD57+ subset...... 163 Figure 4-vii: CD56dim CD8- CD27- CD57- NK cells do not produce IL-22 or IL-17A...... 164 Figure 4-viii: CD8+ CD57+ NK cells express higher levels of cytotoxic molecules. 165 Figure 4-ix: CD8+ CD57+ NK cells exhibit higher cytotoxicity towards K562 target cells...... 166 Figure 5-i: NK cells expressing NKp46 are present in the spleen of Line 7 mice….175

Figure 5-ii: NKp46+ leukocytes are not found in the CNS of Line 7 mice………...176

Figure 5-iii: NKp46+ leukocytes are not found in the CNS of Line 7 mice………. 177

Figure 5-iv: Enrichment of lymphocyte population by FACS…………………….. 178

Figure 5-v: NKp46 was not detected in FACS-sorted lymphocytes isolated from CNS of a Line 7 mouse…………………………………………………………………...179

Figure 5-vi: NKG2D+ lymphocytes were not detected in the CNS of Line 7 mice. 180

Figure 6-i: Predicted signalling pathways and functions of CD8+ CD57+ and CD8- CD57- subsets of CD56dim CD27- NK cells...... 196

9

List of Tables Table 2-i: Reagents for reverse transcription……………………………………….101 Table 2-ii: Scoring of clinical disease in Line 7 mice……………………………....105 Table 2-iii: Antibodies for detection of human antigens……………………………111 Table 2-iv: Antibodies for detection of murine antigens……………………………113 Table 2-v: Excitation and detection wavelengths of fluorochromes and dead cell dyes measured on FACS Aria II………………………………………………………... 113 Table 2-vi: Primer list for Th17 for Autoimmunity & Inflammation RT2 Profiler PCR Array………………………………………………………………………………..114 Table 2-vii: Age and gender of healthy controls used for comparison of subset frequencies…………………………………………………………………………..119 Table 2-viii: Age and gender of CIS patients……………………………………….120 Table 2-ix: Age, gender and disease status of untreated MS patients………………121 Table 2-x: Age, gender and disease status of patients on interferon-β therapy.....…122 Table 2-xi: Age, gender and disease status of Tysabri-treated patients…………….123 Table 2-xiii: Summary of age and gender of patient and control groups…………..124

Table 4-i: Cytokine and chemokine expression determined by qRT-PCR array…...148 Table 4-ii: Expression of cell surface receptors and adaptor molecules determined by qRT-PCR…………………………………………………………………………. ..149 Table 4-iii: Expression of intracellular signalling molecules and transcription factors determined by qRT-PCR array……………………………………………………...150 Table 4-iv: Expression of other genes determined by qRT-PCR array……………..151 Table 4-v: Genes with significantly higher expression in the CD56dim CD8+ CD57+ NK cell subset………………………………………………………………………153 Table 4-vi: Genes with significantly higher expression in the CD56dim CD8- CD27- CD57- NK cell subset……………………………………………………………….153

10

ABBREVIATIONS

ADCC: antibody-dependent cellular cytotoxicity

AID: activation-induced cytidine deaminase

ANOVA: analysis of variance

AP-1: activator protein 1

Apaf-1: apoptotic peptidase activating factor 1

Ape-1: apurinic endonuclease-1

APC: antigen-presenting cell

ATF: activating transcription factor

ATP: adenosine triphosphate

Bak: bcl-2 homologous antagonist/killer

Bid: BH3 interacting-domain death agonist

Bim: Bcl-like protein 11

BBB: blood-brain barrier

BD: Becton Dickinson

BDNF: brain-derived neurotrophic factor bZIP: basic leucine zipper domain

CAD: caspase-activated DNase

CamKII: calmodulin-dependent protein kinase II

CapG: tropomysin, capping protein (actin filament), gelsolin-like

CBP: cAMP-responsive-element-binding protein

CD: cluster of differentiation cDNA: complementary deoxyribonucleic acid

CEBP: Ccaat-enhancer-binding protein

CI: confidence interval

CIITA: class II, major histocompatibility complex, transactivator

CIS: clinically isolated syndrome

CNS: central nervous system

11

CSF: cerebrospinal fluid cSMAC: central supramolecular activation cluster

Ct: cross threshold

DAP: DNAX-activating protein

DC: dendritic cell

DMEM: Dulbecco’s modified Eagle medium

DNA: deoxyribonucleic acid

DNAM-1: DNAX accessory molecule 1

E4BP4: E4-binding protein 4

EAE: experimental autoimmune encephalomyelitis

EBV: Epstein-Barr virus

ERK: extracellular signal-related kinase

ERM: Ezrin, radixin and moesin

F-actin: Filamentous actin

FACS: fluorescence-activated cell sorting

FADD: Fas-associated death domain

FITC: fluorescein isothyocyanate

FL: Flt3 ligand

FMO: fluorescence minus one control

FoxP3: Forkhead box P3

FSC: forward scatter

GA: Glatiramer acetate

GAPDH: Glyceraldehyde-3-phosphate dehydrogenase

GAS: Gamma-activated site

GATA3: GATA binding protein 3

G-CSF: granulocyte colony stimulating factor

GFP: green fluorescent protein

GM-CSF: granulocyte-macrophage colony stimulating factor

12

Grb2: growth factor receptor-bound protein 2

GWAS: genome-wide association study

HA: haemagglutinin

HCV: hepatitis C virus

HLA: human leukocyte antigen

HMG2: high mobility group protein 2

HMGB1: high mobility group box 1

HPRT1: hypoxanthine phosphoribosyltransferase 1

HSC: haematopoetic stem cell

ICAD: inhibitor of caspase-activated deoxyribonuclease

ICAM: intercellular adhesion molecule iDC: immature dendritic cell

IFN: interferon

Ig: immunoglobulin

IGF: insulin-like growth factor

IGSF4: immunoglobulin superfamily member 4

IL: interleukin iNKT cell: invariant natural killer T cell iNOS: inducible nitric oxide synthase

IP3: inositol 1,4,5-triphosphate

IP-10: IFNγ-inducible protein 10

IRF: interferon regulatory factor

ITAM: immunoreceptor tyrosine-based activation motif

ITIM: immunoreceptor tyrosine-based inhibitory motif

JAK: Janus kinase

JNK: cJun N-terminal kinase

KIR: killer immunoglobulin-like receptor

KLRG1: killer cell lectin-like receptor subfamily G member 1

13

LAT: linker of activated T cells

Lck: lymphocyte-specific protein tyrosine kinase

LFA-1: lymphocyte function-associated antigen 1

LIF: leukaemia inhibitory factor

LILR: leukocyte immunoglobulin-like receptor

LMP2: latent membrane protein 2

LPS: lipopolysaccharide

MAPK: mitogen-activated protein kinase

MAP3K: mitogen-activated protein kinase kinase kinase

MBP: myelin basic protein

MCP: monocyte chemotactic protein

MEK: mitogen-activated protein kinase kinase

MEKK1: MAP/ERK kinase kinase 1

MFI: median fluorescence intensity

MHC: major histocompatibility complex

MIC: MHC class I polypeptide-related sequence

Mig: monokine induced by IFN-gamma

MIP: macrophage inflammatory protein

MMP: matrix metalloproteinase

MOG: myelin oligodendrocyte protein mRNA: messenger ribonucleic acid

Mts1: metastasis-associated protein 1

MS: multiple sclerosis

Munc: mammalian uncoordinated

NADPH: nicotinamide adenine dinucleotide phosphate-oxidase

NCR: Natural cytotoxicity receptor

NFATC2: Nuclear factor of activated T cells, cytoplasmic 2

NF-κB: nuclear factor κB

14

NK cell: natural killer cell

NKG: natural killer group

OVA: ovalbumin

PAK: p21-activated kinase

PARP: poly (ADP-ribose) polymerase

PBMC: peripheral blood mononuclear cells

PBS: phosphate buffered saline

PCR: polymerase chain reaction pDC: plasmacytoid dendritic cell

PDGF: platelet-derived growth factor

PHA: phytohaemagglutinin

PI3K: phosphatidylinositol-3-OH kinase

PIP2: phosphatidylinositol 4,5-bisphosphate

PKB: protein kinase B

PKC: protein kinase C

PKR: protein kinase R

PLC-γ: phospholipase C-γ

PLP: proteolipid protein

PMA: phorbol myristate acetate

PML: progressive multifocal leukoencephalopathy

PPMS: primary progressive multiple sclerosis pSMAC: peripheral supramolecular activation cluster qPCR: quantitative polymerase chain reaction qRT-PCR: quantitative reverse transcription polymerase chain reaction

R: receptor

Rac1: ras-related C3 botulinum toxin substrate 1

RANTES: regulated on activation, normal T expressed and secreted

RIP: receptor interacting protein

15

RNA: ribonucleic acid

RNS: reactive nitrogen species

ROR: receptor tyrosine kinase-like orphan receptor

ROS: reactive oxygen species

RPMI: Roswell Park Memorial Institute medium

RRMS: relapsing-remitting multiple sclerosis

RT-PCR: reverse transcription polymerase chain reaction

S1P: sphingosine-1-phosphate

S1PR1: sphingosine-1-phosphate receptor 1

SCF: stem cell factor

SEM: standard error of the mean

SHP: Src homology domain-containing protein tyrosine phosphatase

SLAM: signalling lymphocyte activation molecule

SLP-76: SH2 domain-containing leukocyte protein of 76 kDa

SMAC: supramolecular activation cluster

SNARE: soluble NSF protein attachment receptor

SOCS3: suppressor of cytokine signalling 3

SPMS: secondary progressive multiple sclerosis

STAT: signal transducer and activator of transcription

SDF-1: stromal-derived factor 1

Syk: spleen tyrosine kinase

TAK1: TGFβ-activated kinase 1

TAP: transporter associated with antigen processing

T-bet (Tbx21): T-box expressed in T cells t-Bid: truncated BH3 interacting-domain death agonist

TCR: T cell receptor

TGF-β: transforming growth factor β

Th: Thelper

16

TIMP: tissue inhibitor of matrix metalloproteinases

TNF: tumour necrosis factor

TRADD: TNF receptor-asociated death domain

TRAF2: TNF-R associated factor 2

TRAIL: TNF-related apoptosis-inducing ligand

Treg: regulatory T cell

TRIF: TIR-domain-containing adapter-inducing interferon-β

UK: United Kingdom

ULBP: UL16 binding protein

VCAM: vascular cell adhesion molecule

VDR: vitamin D receptor

VDRE: vitamin D response element

VEGF: vascular endothelial growth factor

VLA-4: very late antigen 4

WASp: Wiscott-Aldrich Syndrome protein

WIP: WASp-interacting protein

ZAP-70: zeta-chain-associated protein kinase 70

17

1. INTRODUCTION

1.1 Natural Killer Cells

Natural killer (NK) cells are granular lymphocytes of the innate immune system, first noted for their ability to rapidly recognise and kill tumour cells (Kiessling et al. 1975, Glimcher et al. 1977). Early studies defined them by their ability to respond to cells lacking major histocompatibility complex class I (MHC class I), a molecule found on the surface of all healthy mammalian cells but often downregulated as a result of severe cellular stress, tumorigenic transformation or (Kärre et al. 1986, Ljunggren et al. 1985, Ljunggren et al. 1988). However, it is now understood that NK cell activation can occur in a number of ways, and is typically determined by the balance of many different activating and inhibitory signals. Furthermore, their function goes beyond direct killing of target cells: it is now known that NK cells interact with many other cell types via cytokines and cell-cell contact, to influence both innate and adaptive immune responses.

1.1.1 NK Cell Development

NK cells, like all leukocytes, are derived from haematopoetic stem cells (HSCs) found in cord blood and fetal and adult bone marrow and liver (Mrózek et al. 1996, Williams et al. 1997, Weissman and Shizuru 2008, Eissens et al. 2012). Development of both mouse and human HSCs into mature NK cells has been recapitulated in vitro and is summarised in figure 1-i. The earliest stages of development require the presence of stromal cells or stromal-derived cytokines such as Flt3-ligand (FL), interleukin (IL)-3 and stem cell factor (SCF) (Shibuya et al. 1995, Williams et al. 1997, Yu et al. 1998, Rosmaraki et al. 2001). This stimulates transition of CD34+ pluripotent HSCs into committed NK progenitors expressing CD122, a receptor chain shared by the IL-2 and IL-15 receptors (Rosmaraki et al. 2001, Yu et al. 1998, Ikawa et al. 1999, LeClerq et al. 1996). Upon reaching this stage, progenitors require either IL-2 or IL-15 to continue their development into mature CD56bright NK cells (Shibuya et al. 1995, Mrózek et al. 1996, Williams et al. 1997, Suzuki et al. 1997, Yu et al. 1998). CD56bright NK cells are then thought to give rise to the CD56dim population (discussed in chapter 1.1.7.1).

18

Figure 1-i: NK cells develop from bone marrow HSCs. Stem cells pass through multipotent and/or bi-potent stages before upregulating CD122 and committing to the NK cell lineage, after which they mature further into CD56bright NK cells. Receptors which have been acquired at each developmental stage are shown in green, while loss of a receptor is indicated in red.

19

During the transition from pluripotent HSCs to committed NK progenitors, stem cells are thought to pass through multipotent and/or bipotent stages, however these stages have not been fully elucidated. Common lymphoid progenitors have been described in adult mouse bone marrow, giving rise to T, B and NK cells, but not myeloid cells (Kondo et al. 1997). However, other studies have indicated that there are a number of different intermediates that can give rise to NK cells, including progenitors of the myeloid lineage (Perez et al. 2003, Vitale et al. 2008, Grzywacz et al. 2011). NK cells seem to have the closest developmental association with T cells. CD25- CD122- bipotent T/NK cell progenitors expressing GATA binding protein 3 (GATA3) and cytoplasmic CD3ε have been identified in mouse and human fetal thymus (Sanchez et al. 1994, Michie et al. 2000, Ikawa et al. 1999). These bipotent progenitors first become biased towards one lineage before fully committing; those which acquire CD25 mostly become T cells, but have not completely lost NK potential, whilst CD122+ progenitors are strongly biased towards the NK cell lineage (Ikawa et al. 1999). A recent study by Lewis Lanier’s team indicates that the transcriptional profile of mature NK cells shows a high level of similarity to T cells in comparison to other leukocytes, which further supports the notion that they are developmentally related (Bezman et al. 2012).

In vivo, development of HSCs into mature NK cells most likely takes place both in the bone marrow and at other sites. CD122+ NK progenitors are readily found in both lymph node and thymus (Ikawa et al. 1999, Freud et al. 2006, Eissens et al. 2012). However their CD34+ CD122- uncommitted precursors are reported only in very small numbers in lymph node (Freud et al. 2006, Eissens et al. 2012). In thymus, bipotent T/NK progenitors are present, but not pluripotent HSCs (Sanchez et al. 1994, Leclercq et al. 1996, Michie et al. 2000, Ikawa et al. 1999, Kawamoto et al. 1998). Though NK progenitors are found in the thymus, this does not appear to be a crucial site for NK cell development, since peripheral NK cells, unlike T cells, are present and functional in athymic mice and thymectomised mice and humans, but not in mice with depleted bone marrow (Colucci et al. 2003). This has led to a model in which NK cells begin the developmental process in the bone marrow, then migrate to other sites such as secondary lymphoid tissue, where they mature further into CD56bright NK cells (Freud et al. 2005, Eissens et al. 2012). The requirement for cytokines produced by bone marrow stromal cells is certainly consistent with the earliest developmental

20 stages occurring in the bone marrow. Progenitors of all stages, however, are found in bone marrow, from HSCs through committed NK progenitors to mature NK cells, so it is likely that many NK cells complete the entire developmental process in the bone marrow, while others migrate (Rosmaraki et al. 2001, Eissens et al. 2012). The relative importance of the different sites in the latter stages of NK cell development is unclear, although one can speculate on the roles that each might play. NK cell development in the bone marrow is likely to be driven by IL-15 constitutively produced by stromal cells (Mrózek et al. 1996). In lymph nodes however, development might be driven by IL-2 and IL-15 from activated T cells and dendritic cells (DCs) (Mrózek et al. 1996, Freud et al. 2005, Freud et al. 2006). Therefore one can hypothesise that the bone marrow might be a site for a more consistent, continuous generation of mature NK cells, whereas lymph node might provide bursts of NK cell maturation during the latter stages of infection, thus replenishing the mature NK cell pool. In addition, since these NK cells are initially of the immunoregulatory CD56bright subset and in close proximity to activated T cells and DCs they might have a role in dampening the immune response once an infection is cleared (Fehniger et al. 2003, Ferlazzo et al. 2004).

1.1.2 NK Cell Function is Determined by the Balance of Signals from Activating and Inhibitory Receptors

A characteristic of NK cells is the ability to recognise “missing self” (see figure 1-ii). This refers specifically to their ability to respond to a lack of MHC class I molecules on a target cell, often a characteristic of tumour and virus-infected cells (Kärre et al. 1986, Ljunggren et al. 1985, Ljunggren et al. 1988, Lanier et al. 2008). In this way, NK cells are able to kill pathogenic cells which have evaded recognition by CD8+ T cells. When an NK cell encounters a potential target cell, the outcome is determined by the recognition of ligands for a range of activating and/or inhibitory receptors (see figure 1-ii). Some activating ligands are constitutively expressed at low levels, which would not be enough to activate NK cells in the presence of inhibitory signals. However, when the inhibitory signal is absent, the weak activating signals may be sufficient to induce cytotoxicity (“missing self” recognition). Target cells expressing high levels of stress-induced activating ligands, such as MHC class I polypeptide- related sequence (MIC) A and MICB, however, may be killed even in the presence of an inhibitory ligand (“stressed self/ induced self”) (Bauer et al. 1999, Cosman et al.

21

2001). The balance of signals from activating and inhibitory receptors can also influence production of cytokines such as interferon (IFN)γ (Cassatella et al. 1989, Draghi et al. 2007, Fauriat et al. 2010a, De Maria et al. 2011). Meanwhile, cytokines produced by other cells can induce a more activated, cytotoxic phenotype amongst NK cells, which enhances their ability to kill susceptible targets and produce further proinflammatory cytokines (Nagler et al. 1990, Gerosa et al. 2005, Mailliard et al. 2005, Draghi et al. 2007).

Important receptors involved in regulation of NK cell activity include the natural cytotoxicity receptors (NCRs), CD16, the natural killer group (NKG)2 family and the killer immunoglobulin-like receptor (KIR) family in humans or Ly49 family in rodents (Lanier 2008). Many other receptors and co-receptors, such as DNAX accessory molecule-1 (DNAM-1), leukocyte immunoglobulin-like receptors (LILRs), killer cell lectin-like receptor subfamily G member 1 (KLRG1) and the signalling lymphocyte activation molecule (SLAM) family can also influence the decision whether or not to kill a target (Tahara-Hanaoka et al. 2004, Ito et al. 2006, Vaidya and Matthew 2006, Cruz-Munoz et al. 2009, Kaur et al. 2012, Stegmann et al. 2012).

Many of the activating receptors (including NCRs, CD16 and KIR) signal via immunoreceptor tyrosine-based activation motif (ITAM)-containing adaptor proteins such as DNAX-activating protein (DAP)12, CD3ζ and FcεRI-γ (Lanier 2008). ITAM phosphorylation upon receptor cross-linking induces binding of zeta-chain-associated protein kinase 70 (ZAP-70) and spleen tyrosine kinase (Syk) and activates a signalling cascade resulting in transcription of cytokine and chemokine genes as well as cytoskeletal reorganisation leading to polarisation and release of lytic granules (Brumbaugh et al. 1997, Lanier et al. 1998). Meanwhile, ligation of the inhibitory receptors NKG2A and KIR causes phosphorylation of the two immunoreceptor tyrosine-based inhibitory motifs (ITIMs) in the intracytoplasmic tail (Bruhns et al. 1999, Kabat et al. 2002). Phosphorylated ITIMs act as docking sites for the Src homology domain-containing protein tyrosine phosphatases (SHP)-1 and -2, which dephosphorylate components of the activation pathway, thereby neutralising the activating signal (Le Dréan et al. 1998, Bruhns et al. 1999, Yusa and Campbell 2003, Vyas et al. 2004). Signalling pathways associated with both activating and inhibitory receptors are transient and localised around the NK-target cell synapse, so as not to

22 affect the ability of an NK cell encountering inhibitory ligands to subsequently kill a different target cell which lacks such ligands (Lanier 2008, Kaplan et al. 2011).

Figure 1-ii: Missing self and stressed self recognition. (A) Missing self: When target cells do not express ligands specific for an NK cell’s inhibitory receptors, the threshold for activation is lowered, allowing the NK cell to kill target cells expressing normal levels of activating ligands. (B) Inhibitory signals antagonise components of the activation pathway, thereby sparing target cells expressing moderate levels of both activating and inhibitory ligands. (C) Stressed self: Cells expressing abnormally high levels of activating ligands (such as stressed or infected cells) induce cytotoxicity even in the presence of inhibitory ligands. Activating receptors and ligands are shown in green and yellow and inhibitory receptors and ligands in red and orange respectively.

23

One of the most potent activating receptors is NKG2D, which recognises the MHC class I chain-related molecules MICA, MICB and UL16-binding proteins (ULBPs) 1- 6 (Bauer et al. 1999, Cosman et al. 2001, Steinle et al. 2001). NKG2D is expressed on all NK cells, as well as γδ and CD8+ αβ T cells (Bauer et al. 1999). NKG2D homodimers associate with DAP10 (Wu et al. 1999). Phosphorylation of DAP10’s signalling motif allows it to bind phosphatidylinositol-3-OH kinase (PI3K) and growth factor receptor-bound protein 2 (Grb2), inducing a signalling pathway distinct from that of other activating receptors (Wu et al. 1999, Upshaw et al. 2006). NKG2D is particularly efficient at inducing cytotoxicity, but to a lesser extent also stimulates proliferation and production of cytokines including IFNγ (Bauer et al. 1999, Cosman et al. 2001, Kubin et al. 2001, Draghi et al. 2007, Lanier 2008). The NKG2D ligands ULBPs 1-3 are constitutively expressed at low levels on a variety of cell types, though they are further upregulated by viral infection (Cosman et al. 2001, Draghi et al. 2007). Conversely, MICA and MICB are virtually absent from most normal cells but are highly expressed on some tumours (Groh et al. 1999, Pende et al. 2002, Carbone et al. 2005). NKG2D ligands are induced/ upregulated by the DNA damage response pathway (Gasser et al. 2005, Tang et al. 2008). NKG2D therefore has a key role in recognition of “stressed self”. MICA is also expressed by activated effector T cells, regulatory T cells (Tregs) and macrophages, which may provide a mechanism for NK cell-mediated elimination of potentially pathogenic overstimulated leukocytes once an infection has been cleared (Molinero et al. 2002, Hamerman et al. 2004, Roy et al. 2008).

Other members of the NKG2 family found on NK cells include the activating receptor NKG2C and the inhibitory receptor NKG2A (Houchins et al. 1997). Both of these function as heterodimers with CD94 and recognise human leukocyte antigen (HLA)- E, a non-classical HLA molecule which presents signal peptides derived from other class I HLA molecules (Borrego et al. 1998, Braud et al. 1998, Lee et al. 1998, Brooks et al. 1999, Valés-Gómez et al. 1999). NKG2A is found on all CD56bright NK cells and around 50% of CD56dim NK cells (Vitale et al. 2004). It is more frequently expressed and has a higher binding affinity and more important role in regulating NK function than NKG2C (Valés-Gómez et al. 1999, Béziat et al. 2010). Binding affinity is, however, dependent on the sequence of the peptide bound to HLA-E, and this in

24 turn can affect the efficacy of NKG2A’s inhibitory function (Brooks et al. 1999, Valés-Gómez et al. 1999).

The more recently described NCRs (including NKp30, NKp44 and NKp46), another family of activating receptors common to humans and mice, recognise a wide range of ligands, some of which are likely still to be discovered. NKp44 and NKp46 recognise influenza virus haemagglutinin (HA) on the surface of infected cells (Arnon et al. 2001, Mandelboim et al. 2001). NKp44 is also able to bind directly to envelope proteins of Dengue and West Nile viruses and to the surface of mycobacteria, whilst NKp46 recognises the normally intracellular intermediate filament protein vimentin, which is upregulated on the surface of monocytes infected with Mycobacterium tuberculosis (Garg et al. 2006, Esin et al. 2008, Hershkovitz et al. 2009). NKp30 mediates activation by DCs and also recognises the tumour antigens B7-H6 and nuclear factor HLA-B-associated transcript 3 (Ferlazzo et al. 2002, Vitale et al. 2005, Pogge van Strandmann et al. 2007, Brandt et al. 2009). NKp46 and NKp30 are found on all NK cells, whereas NKp44 is predominantly restricted to CD56bright NK cells in secondary lymphoid tissue and gut, but may be induced on activated peripheral blood NK cells (Moretta et al. 2001). NCRs are considered to be specific to NK cells and are absent from other peripheral blood leukocytes, though data from our lab suggests that NKp46 may be found on astrocytes (Durrenberger et al. 2012).

Another important activating receptor is the Fc receptor CD16 (FcγRIII). This receptor mediates antibody-dependent cellular cytotoxicity (ADCC), a process which links NK cells to the adaptive immune response (Heijnen and van der Winkel 1997). Target cells coated with immunoglobulin (Ig)G cross-link CD16 on NK cells, resulting in tyrosine phosphorylation of the CD3ζ adaptor molecule, which triggers a cytotoxic response against the target as well as cytokine production (Ullberg and Jondal 1983, Cassatella et al. 1989, O’Shea et al. 1991, Vivier et al. 1991, Trinchieri and Valiente 1993). The efficacy of ADCC is dependent on both the density of antibody on the coated cell and the affinity of the antibody for its target antigen (Velders et al. 1998, Bowles and Weiner 2005). Unlike other NK cell activating receptors, triggering of CD16 alone is sufficient to elicit a cytolytic response amongst resting NK cells (Bryceson et al. 2005).

25

Missing self recognition in humans is largely mediated by a highly polymorphic family of receptors for classical human leukocyte antigen class I (HLA class I) molecules, the killer immunoglobulin-like receptors (KIRs), which are absent from lower mammals including mice (Parham 2005). The KIR family comprises both activating and inhibitory receptors. Inhibitory KIRs have long cytoplasmic tails containing two ITIMs (denoted “L”) (Biassoni et al. 1996, Bruhns et al. 1999). Activating KIRs usually have short cytoplasmic tails (“S”) containing a basic amino acid in the transmembrane region, allowing them to interact with a region containing an acidic residue in the ITAM-containing adaptor protein DAP-12 (Biassoni et al. 1996, Lanier et al. 1998, Lanier 2008). The exception to this is the long-tailed receptor KIR2DL4, which contains a single ITIM as well as a basic residue (Faure and Long 2002). It primarily activates NK cells via the FC receptor γ protein 1, but also possesses ITIM-dependent inhibitory function (Faure and Long 2002, Kikuchi- Maki et al. 2003, Kikuchi-Maki et al. 2005).

Each KIR has specificity for HLA molecules containing a particular motif. KIRs with two extracellular domains (KIR2D) mainly recognise HLA-C molecules and display a preference towards either “C1” alleles, with serine at position 77 and asparagine at position 80, or “C2” alleles, containing asparagine at position 77 and lysine at position 80 (Colonna et al. 1993). KIR3DL1 and KIR3DL2 recognise specific alleles of HLA- A and –B, whilst KIR2DL4 interacts with HLA-G (Cella et al. 1994, Gumperz et al. 1995, Rajagopalan and Long 1999, Hansasuta et al. 2004, Stern et al. 2008, Vivian et al. 2011). The affinity of the interaction and therefore the strength of the activating/ inhibitory signal, is influenced by KIR/HLA polymorphism and possibly by the peptide bound to the HLA molecule at the time (Hansasuta et al. 2004, Ianello et al. 2008, Sanjanwala et al. 2008, Sharma et al. 2009, Colantonio et al. 2011, Fadda et al. 2011, Vivian et al. 2011).

Not only are the individual genes encoding KIR and HLA highly polymorphic, there is also copy number and haplotype variation (Uhrberg et al. 1997, Hsu et al. 2002, Parham 2005, Jiang et al. 2012). Most KIR genes are not common to all individuals: some people have more activating KIRs whereas others have a more inhibitory repertoire (Uhrberg et al. 1997, Parham 2005). In addition to this, individuals vary in the proportion of NK cells expressing KIRs, and each KIR+ NK cell within an

26 individual expresses a different combination of KIRs on the cell surface (Yawata et al. 2008). KIR genotype is a major source of variation between individuals with regard to NK cell activation and influences susceptibility to infection, autoimmunity, cancer and reproductive disorders (Yen et al. 2001, Martin et al. 2002, van der Silk et al. 2003, Hiby et al. 2004, Khakoo et al. 2004, Suzuki et al. 2004, Verheyden et al. 2004, Jones et al. 2006, Martin et al. 2007, Hiby et al. 2008, Lorentzen et al. 2009, Hiby et al. 2010). Haplotypes associated with activation, especially those containing the activating KIR 2DS1 or 2DS2, have been associated with autoimmune diseases including multiple sclerosis, diabetes, psoriasis, rheumatoid arthritis and ulcerative colitis (Yen et al. 2001, van der Silk et al. 2003, Suzuki et al. 2004, Jones et al. 2006, Lorentzen et al. 2009). Inhibitory haplotypes may lead to increased susceptibility to infection, allowing haplotypes which predispose to autoimmunity to persist in the population (Martin et al. 2002). However, during some , certain inhibitory receptors may be protective, perhaps as a result of licensing conferring enhanced ability to recognise missing self (Khakoo et al. 2004, Martin et al. 2007) (see chapter 1.1.3).

Close to the KIR locus are genes encoding another family of both activating and inhibitory receptors, the LILRs, which are expressed on some NK cells, as well as T cells, B cells, monocytes and DCs (Kaur et al. 2012). In a similar manner to KIRs, activating LILRs (LIRA) associate with ITAM-containing adaptor proteins, while inhibitory LILRs (LIRB) transmit signals via their own ITIMs. Several of the LILRs bind classical and non-classical MHC class I ligands and LILRA4 binds tetherin (a membrane protein that inhibits release of viral particles), while the ligands for many of the LILR family are not yet known (Cosman et al. 1997, Colonna et al. 1998, Chapman et al. 1999, Allen et al. 2001, Cao et al. 2009, Ryu et al. 2011, Kaur et al. 2012).

1.1.3 NK Cell Education Determines Responsiveness to Activating Signals

NK cell education refers to a process thought to take place during development that prevents activation of NK cells lacking inhibitory receptors specific for self-ligands (see figure 1-iii). This is likely an important mechanism preventing autoreactivity. Olsson et al. first presented evidence to support this concept in C57BL/6 mice in 1995. Ly49+ (equivalent to human KIR+) NK cells were only able to kill tumour cells

27 lacking Ly49 ligands when they were derived from mice which expressed its ligand H-2D(d) on normal cells. Meanwhile, Ly49A+ NK cells are more effectively inhibited by their ligand, H-2b, when derived from mice which express this ligand (Dorfman and Raulet 1996). Therefore the immunological environment in which NK cells develop appears to tune their responsiveness to both presence and absence of inhibitory signals. The selective nature of education was demonstrated by Kim et al. (2005), who showed that adding a transgene encoding the ligand for Ly49C into MHC class I- deficient strains resulted in Ly49C+, but not Ly49A+, NK cells regaining the ability to mount an IFNγ response.

The exact mechanism by which education occurs is not yet understood. Some models suggest that NK cells expressing self-inhibitory receptors are “licensed” to become responsive to activating signals (Kim et al. 2005). Alternatively, education has been described as a process of “disarming”, whereby NK cells lacking such receptors are eliminated from the repertoire, inactivated, or perhaps become anergic due to persistent activating signals (Valiante et al. 1997, Fernandez et al. 2005, Raulet and Valance 2006, Yawata et al. 2008).

A study of human NK cell clones found that over 98% were inhibited by autologous HLA class I, leading to the “at least one” hypothesis, that NK cells not expressing an inhibitory receptor for self-HLA are eliminated from the repertoire (Valiante et al. 1997). However, this study examined NK cell clones from just two donors, who were most likely not representative of the population, since subsequent studies have described the presence of substantial numbers of NKG2A- KIR- NK cells in peripheral blood (Anfossi et al. 2006, Cooley et al. 2007, Yawata et al. 2008). Though present, such cells are hyporesponsive to stimulation by cytokines, tumour cells or antibody- mediated crosslinking of activating receptors (Anfossi et al. 2006, Cooley et al. 2007). In humans, NK cells which do not express NKG2A must not only express at least one inhibitory KIR in order to respond effectively, this KIR must also be specific for self-HLA (Anfossi et al. 2006, Yawata et al. 2008). However, NK cell education is not as simple as just an “on/off” switch; more recently it has been described as a “rheostat” (Joncker et al. 2009). The more self-specific inhibitory receptors an NK cell expresses, the greater the cytotoxic and IFNγ response to a variety of stimuli (Yu et al. 2007, Yawata et al. 2008, Brodin et al. 2009, Joncker et al. 2009). Just one HLA

28 ligand can abolish the response of NK cells expressing multiple self-specific inhibitory receptors, meaning that those cells with the highest potential response have the most stringent criteria for activation (they will only be activated when several self HLA molecules are missing) (Yu et al. 2007).

Since the NKG2A- KIR- NK cells described in peripheral blood resemble mature CD56dim NK cells in expression of CD16, CD162R and other surface receptors, and lack markers such as CD117 associated with progenitor cells, it seems most likely that these cells have differentiated from mature NKG2A+ KIR- NK cells (Anfossi et al. 2006). If this is indeed the case, it would argue against a model in which cells become permanently licensed at an early stage of maturity, since these cells were presumably functional and “licensed” prior to loss of NKG2A. Perhaps NK cells in fact require continual licensing throughout their lifetime in order to remain fully functional. Or perhaps both arming and disarming occur upon gain and loss of an inhibitory receptor respectively.

It appears that the hyporesponsiveness of unlicensed NK cells can be overcome under certain conditions. In vitro, pre-stimulation with IL-2 or polyinosinic-polycytidylic acid (poly I:C) partially abrogates hyporesponsiveness (Kim et al. 2005). In addition, strong activation of unlicensed NK cells has been reported during infection with Listeria monocytogenes or mouse cytomegalovirus and in leukaemia patients following HLA-matched HSC transplantation (Fernandez et al. 2005, Yu et al. 2009, Orr et al. 2010).

Although discussion of education has so far focussed largely on the role of inhibitory receptors, it seems that activating receptors may also be involved in the process. Noé mice, which have a defect in NKp46 expression, have hyperresponsive NK cells, which are more able to stimulate the T cell response, and are therefore able to clear infection more rapidly (Narni-Mancinelli et al. 2012). In humans, the responsiveness of KIR2DS2+ NK cells to target cells is tuned down in individuals homozygous for the KIR2DS2 ligand HLA-C2, though cytokine responsiveness is not affected (Fauriat et al. 2010b). Thus, it seems that education tunes an NK cell’s responsiveness according to its ability to recognise both activating and inhibitory signals.

29

Figure 1-iii: NK cell education maintains tolerance of NK cells lacking self-specific inhibitory receptors. NK cell responsiveness is determined by engagement of inhibitory and activating molecules during a process known as NK cell education. NK cells that lack inhibitory receptors (A) or that only express inhibitory receptors specific for non-self HLA molecules (B) are present in the repertoire, but are hyporesponsive to stimulation via their activating receptors. Conversely, inability to recognise common activation signals during education may result in hyperresponsiveness (D).

30

1.1.4 NK Cells Form Synapses with Target Cells and Kill Susceptible Targets via Granzyme Release and Other Mechanisms

NK cells induce death of targets primarily through directed secretion of cytotoxic granules, as well by engaging receptors of the tumour necrosis factor (TNF) superfamily such as Fas and TNF-related apoptosis-inducing ligand (TRAIL) receptors (Scott et al. 2008). Compared to resting CD8+ T cells, resting NK cells express higher levels of granzymes and molecules associated with vesicular transport and cytoskeletal rearrangement, allowing them to respond more rapidly (Bezman et al. 2012). NK cells must first contact the target cell, then form an immunological synapse through which cytolytic molecules and cytokines can be directly targeted towards the susceptible target cell (Davis et al. 1999, Vyas et al. 2001, McCann et al. 2003, Orange et al. 2003, Culley et al. 2009).

Initial interactions, possibly mediated by CD2/sialyl Lewis X and/or selectins, are followed by firm adhesion mediated by higher affinity receptor-ligand interactions involving integrins such as lymphocyte function-associated antigen 1 (LFA-1) and macrophage receptor 1, which rapidly cluster at the synapse during initiation and also participate in activation of cytolytic pathways (Davis et al. 1999, Riteau et al. 2003, Orange 2008, Mace et al. 2010). Signals from activating and inhibitory receptors will then determine progression to either an activating or inhibitory synapse (Vyas et al. 2001, Vyas et al. 2002, Culley et al. 2009). Once tight contacts have been formed between NK and susceptible target cells, around half of these interactions result in killing, which can be either rapid (within 2-5 minutes of conjugate formation) or delayed, with a prolonged period of contact during which calcium levels and actin accumulation fluctuate prior to polarisation and release of cytotoxic granules (Wulfing et al. 2003).

If activating signals are sufficient to override inhibitory ones, an activating (lytic) synapse can be formed. Filamentous (F)-actin accumulation at the synapse leads to symmetrical spreading which broadens the interface between NK cell and target cell and increases stability of the NK cell:target cell conjugate (McCann et al. 2003, Wulfing et al. 2003, Culley et al. 2009) . Ezrin, radixin and moesin (ERM) family proteins, which connect actin filaments to membrane structures, are present at the synapse and may be involved in recruitment of further receptors and associated

31 signalling molecules to synaptic lipid rafts, forming a supramolecular activation cluster (SMAC) (McCann et al. 2003, Watzl et al. 2003, Masilamani et al. 2006). A dense ring of F-actin forms at the peripheral (p) SMAC and integrins and talin also localise to this region in an actin-dependent manner (McCann et al. 2003, Orange et al. 2003, Vyas et al. 2001). Meanwhile, adaptor and signalling molecules such as Syk, ZAP-70, SH2 domain-containing leukocyte protein of 76 kDa (SLP-76) and SHP-1 accumulate in the centre of the synapse, forming the cSMAC (Vyas et al. 2001). Dynein drives migration of lytic granules along microtubules to the microtubule organising centre (MTOC), which then polarizes towards the centre of the synapse (Küpfer et al. 1983, Küpfer et al. 1985, Stinchcombe et al. 2011, Mentlik et al. 2010). A number of hypodense regions of actin form within the central part of the synapse, where the gaps between filaments are large enough to allow granules to pass through (Brown et al. 2011, Rak et al. 2011). Granules migrate along filaments towards the cell membrane in a myosin IIA-dependent manner before fusing with the cell membrane and being released in the direction of the target cell (Andzelm et al. 2007, Sanborn et al. 2009).

Following activating receptor ligation, the protein tyrosine kinase Syk is recruited to phosphorylated ITAMs and becomes activated (Brumbaugh et al. 1997). Activation of Syk has an important role in NK cell cytotoxicity, mediated in part by its phosphorylation of the nucleotide guanine exchange factors Vav 1-3 (Brumbaugh et al. 1997, Deckert et al. 1996, Galandrini et al. 1999). Vav 2 and 3 appear to be more important in signalling induced via DAP12 or FcRγ, whereas Vav1 is crucial for NKG2D/DAP10-induced cytotoxicity and may also be activated by integrin engagement (Riteau et al. 2003, Cella et al. 2004, Graham et al. 2006). Tyrosine phosphorylation of Vav and consequent activation of its target ras-related C3 botulinum toxin substrate 1 (Rac1), are important for triggering actin reorganisation, broadening of the immunological synapse and polarisation of lytic granules (Billadeau et al. 1998, Cella et al. 2004, Graham et al. 2006). Blocking of the p21-activated kinase (PAK1)/mitogen-activated protein kinase kinase (MEK)/extracellular signal- related kinase (ERK) pathway, which is activated by Rac1, severely impairs cytotoxic function (Cella et al. 2004). Syk also induces lipid raft polarisation and activates PI3K, another inducer of the Rac1/PAK/MEK/ERK pathway and phospholipase C-γ (PLC-γ), which initiates the inositol 1,4,5-triphosphate (IP3)/ Ca2+ signaling pathway

32 through its cleavage of membrane-associated phosphatidylinositol 4,5-bisphosphate (PIP2) (Ting et al. 1992, Cassatella et al. 1989, Azzoni et al. 1992, Windebank et al. 1988, Lou et al. 2000, Jiang et al. 2002). Protein kinase C is activated via this pathway and phosphorylates WASp-interacting protein (WIP), leading to formation of a complex containing WIP, actin, myosin IIA and the cytoskeletal regulator Wiscott- Aldrich Syndrome protein (WASp), which is crucial for F-actin accumulation and subsequent cytotoxic activity (Graves et al. 1986, Orange et al. 2002, Krzewski et al. 2006). Activation of talin in response to LFA-1 ligation by intercellular adhesion molecule 1 (ICAM-1) also induces the PIP2 pathway and recruits the actin nucleating protein complex Arp2/3, which mediates actin polymerisation promoted by WASp (Mace et al. 2010). Meanwhile, the increased calcium levels promote interaction of mammalian uncoordinated (Munc)- 13-4 with soluble NSF protein attachment receptor (SNARE) proteins and Rab27a, which is important for fusion of cytotoxic granules with the cell surface membrane (Feldmann et al. 2003, Johnson et al. 2011, Boswell et al. 2012).

If there is sufficient inhibitory input however, an inhibitory synapse will form between the NK cell and the target cell. Rings of inhibitory receptors such as long- tailed KIR and associated signalling molecules form at the synapse around a central cluster of adhesion molecules (Davis et al. 1999, Vyas et al. 2001, McCann et al. 2003, Treanor et al. 2006, Orange 2008). This so-called supramolecular inhibitory cluster (SMIC) blocks further recruitment of activating receptors and signalling molecules and antagonises their signalling function (Watzl et al. 2003,Vyas et al. 2004, Endt et al. 2007). Ligation of inhibitory receptors causes them to translocate to the centre of the synapse, where the phosphorylated ITIMs provide a docking site for the tyrosine phosphatases SHP-1 and -2, releasing them from autoinhibition (Yusa et al. 2003, Vyas et al. 2004). SHP-1 dephosphorylates components of the activation pathway such as Vav1, SLP-76, the ERM family, pp36 (an adaptor protein required for recruitment of PLC-γ) and possibly Syk, thereby preventing actin reorganisation, blocking recruitment of activating receptors and lipid raft polarisation, preventing symmetrical cell spreading and promoting detatchment from the target cell (Valiante et al. 1996, Brumbaugh et al. 1997, Binstadt et al. 1998, Burshtyn et al. 2000, Stebbins et al. 2003, Watzl et al. 2003, Masilamani et al. 2006, Endt et al. 2007, , Culley et al. 2009).

33

NK cell lytic granules consist of an inner electron-dense region containing cytotoxic mediators surrounded by a multivesicular lysosomal-like cortical layer (Burkhardt et al. 1990). Their contents include the membrane perturbing enzyme perforin, pro- apoptotic serine proteases known as granzymes and enzymes such as cathepsin B which protect the NK cell from apoptotic induction by components of its own granules (Lieberman 2003, Balaji et al. 2002).

Perforin is essential for induction of apoptosis by granzymes (Kägi et al. 1994, Walsh et al. 1994). Originally, it was proposed that perforin-mediated formation of pores in the cell surface membrane provided passages for granzymes to diffuse into the target cell. However there is limited evidence to support this hypothesis and in order to form large enough pores to facilitate diffusion of granzymes into the target cell, very high concentrations of perforin are required, much higher than the concentration needed to facilitate granzyme-induced apoptosis (Browne et al. 1999, Trapani and Smyth 2002). At lower concentrations of perforin, small pores are formed in the target cell membrane, triggering a Ca2+ flux– induced repair process, which prevents necrotic cell death and leads to endocytosis of perforin and granzymes in enlarged endosomes (Thiery et al. 2010). Several studies have shown that granzymes can enter the target cell independently of perforin, for instance by binding to the mannose 6-phosphate receptor, but that perforin is required for cytosolic and nuclear localisation of granzymes and apoptotic induction once inside the cell (Froelich et al. 1996, Shi et al. 1997, Pinkosi et al. 1998, Motyka et al. 2000). One important function of perforin is likely to be disruption of endosomal trafficking (Browne et al. 1999, Thiery et al. 2011). Perforin forms pores in the endocytic membrane, triggering endosome rupture and release of granzymes into the cytosol of the target cell (Thierry et al. 2011).

Human cytolytic granules contain granzymes A, B, H, K and M, with A and B being the most abundant (Lieberman 2003). Granzyme B can induce apoptosis via both caspase-dependent and –independent mechanisms. It is able to activate both apical caspases (8 and 10) and executioner caspases (3, 6 and 7) by cleaving after aspartic acid residues between the large and small subunits of the procaspase (Medema et al. 1997a, Talanian et al. 1997, Yang et al. 1998a). Caspases themselves also cleave at sites following aspartic acid residues, but vary in specificity. Apical caspases cleave downstream executioner caspases, which then cleave a variety of substrates, resulting

34 in activation of pro-apoptotic and degradation of anti-apoptotic molecules (Lieberman 2003). In cell-free assays, caspase 7 and 10 are the most efficiently processed by granzyme B, however in cell culture, full maturation of caspase 7 is dependent on the activation of caspase 3 and/or 10 (Talanian et al. 1997, Yang et al. 1998a). Cleavage of caspase 10 is likely to occur first (Talanian et al. 1997). Caspase 10 can in turn activate caspase 3, which can also be directly but inefficiently cleaved by granzyme B (Talanian et al. 1997). Activated caspase 3 removes the propeptide from the caspase 7 precursor, making it accessible to granzyme B which then cleaves it at a different site, resulting in production of active caspase 7 (Yang et al. 1998a).

Caspases 3 and 7 proteolytically inactivate inhibitor of caspase-activated deoxyribonuclease (ICAD) which releases caspase-activated DNase (CAD) from inhibition (Enari et al. 1998, Liu et al. 1999, Wolf et al. 1999). CAD oligomerizes into a functional complex and creates blunt ended double-stranded breaks in DNA, producing oligonucleosomal fragments of approximately 200 base pairs (Liu et al. 1999, Wolf et al. 1999, Enari et al. 1998 Lieberman 2003). Caspases also cleave poly (ADP-ribose) polymerase (PARP), preventing it from synthesising ADP-ribose polymers for DNA repair (Labeznik et al. 1994, Gu et al. 1995, Tewari et al. 1995, Germain et al. 1999). Chromatin condensation is induced by the cleaved form of Acinus, another target of caspase 3 (Sahara et al. 1999). Caspase 8 cleaves the pro- apoptotic protein BH3 interacting-domain death agonist (Bid), the truncated form of which (tBid) translocates to the mitochondrial outer membrane, where it induces oligomerization of bcl-2 homologous antagonist/killer (Bak), allowing it to form pores in the outer mitochondrial membrane, facilitating release of cytochrome C (Li et al. 1998). Another important mitochondrial target is NDUFS1, the 75 kDa subunit of respiratory complex I (Ricci et al. 2004). Caspase 3- mediated cleavage disrupts its function in the electron transport chain, causing loss of membrane potential, increased reactive oxygen species (ROS) production and reduced adenosine triphosphate (ATP) levels (Ricci et al. 2003, Ricci et al. 2004). Morphological changes, membrane blebbing and packaging of fragmented DNA into apoptotic bodies have also been attributed to enzymes activated by caspases, including Rho-associated protein kinase 1 (ROCK-1), which mediates myosin light phosphorylation and coupling of actin- myosin filaments to the plasma membrane, gelsolin which severs actin filaments, and PAK-2, a serine/threonine kinase which phosphorylates myosin light chain, MEK1

35

(inducer of p38 and cJun N-terminal kinase (JNK) pathways) and the p47phox nicotinamide adenine dinucleotide phosphate-oxidase (NADPH) subunit (Kothakota et al. 1997, Bokoch 1998, Rudel et al. 1998, Coleman et al. 2001).

Granzyme B is also able to directly cleave many caspase targets (Lieberman 2003). These include ICAD, as well as proteins involved in recognising and repairing DNA damage such as PARP and DNA-dependent protein kinases (Sharif-Askari et al. 2001, Thomas et al. 2000, Andrade et al. 1998). Granzyme B also induces microtubule polymerisation and cytoskeletal rearrangement by removing an autoinhibitory domain from alpha tubulin (Goping et al. 2006). In addition, it can directly cleave Bid, causing release of cytochrome C from the mitochondrianda (Barry et al. 2000, Sutton et al. 2000, Wang et al. 2001a). Granzyme B also cleaves myeloid cell leukaemia sequence 1L (Mcl-1L), preventing its sequestration of Bcl-like protein 11 (Bim), another inducer of cytochrome C release (Han et al. 2004). Cytochrome C and dATP promote binding of caspase 9 to apoptotic peptidase activating factor 1(Apaf-1), resulting in activation of caspase 9, which in turn cleaves the executioner caspases 3 and 7 (Li et al. 1997). Direct cleavage of Bid by granzyme B appears to have a more important role in granzyme B-induced apoptosis than caspase-mediated Bid cleavage (Trapani et al. 1998, Heibein et al. 1999, MacDonald et al. 1999, Sutton et al. 2000, Waterhouse et al. 2006). However this positive feedback loop may be important for preventing repolarisation of the mitochondrial membrane (MacDonald et al. 1999, Waterhouse et al. 2006).

Granzyme A does not activate caspase 3 or downstream effectors such as PARP, but induces DNA damage, nuclear condensation and membrane blebbing in a caspase- independent manner (Beresford et al. 1999). The mitochondrial membrane becomes depolarised and production of ROS is increased, however, in contrast to granzyme B, these effects are not dependent on the Bid/Bcl2 or caspase pathways and cytochrome C is not released (Lieberman 2003, Martinvalet et al. 2005). Granzyme A directly and efficiently cleaves lamins A, B and C (which are also cleaved by granzyme B- activated caspases), resulting in disruption of the nuclear envelope (Zhang et al. 2001a). It also inactivates several components of the endoplasmic reticulum- associated SET complex. These include the DNA bending protein high mobility group protein 2 (HMG2) and apurinic endonuclease-1 (Ape1), an enzyme which inhibits

36 generation of reactive oxygen species, repairs oxidised DNA and reduces oxidised transcription factors (Fan et al. 2002, Fan et al. 2003a, Guo et al. 2008). SET itself is also cleaved by granzyme A, releasing the tumour metastasis suppressor NM23H1 from inhibition, prompting its translocation to the nucleus, where it creates single- strand nicks in chromosomal DNA (Fan et al. 2003b). Therefore granzyme A both induces DNA damage and prevents its repair. Granzyme A also has functions beyond apoptosis, such as inducing monocyte proinflammtory cytokine secretion and retraction of neurites by mouse neuronal cells (Suidan et al. 1994, Metkar et al. 2008).

Both granzyme A and B can independently induce apoptosis in the presence of perforin, however they also act synergistically (Lieberman 2003). One explanation for this is that granzyme A cleaves several histone proteins, including histone H1, which facilitates chromatin hypercondensation (Zhang et al. 2001b). This results in unfolding, thereby increasing access to DNases including CAD, which mediates oligonucleosomal DNA degradation in response to granzyme B.

Granzyme K induces apoptosis in a similar manner to granzyme A. It induces loss of mitochondrial membrane potential and rapid generation of ROS via cleavage of Ape1 and possibly other mechanisms (Guo et al. 1998, MacDonald et al. 1999). Like granzyme A, it cleaves SET, activating NM23H1 to create single-stranded DNA nicks (Zhao et al. 2007a). It can also directly cleave Bid, leading to release of cytochrome C and endonuclease G (Zhao et al. 2007b).

The functions of granzymes H and M are only now being elucidated. Cell death- related substrates of granzyme H include Bid and ICAD (Hou et al. 2008). Granzyme M indirectly triggers caspase cascades by cleavage of survivin and directly cleaves PARP, ICAD, PAK2 and the nuclear phosphoprotein nucleophosmin which is essential for cell survival (Lu et al. 2006, Cullen et al. 2009, Hu et al. 2010). It inactivates the ROS antagonist heat shock protein 75 and cleaves the cytoskeletal components ezrin and alpha-tubulin, disrupting microtubule organisation (Bovenschen et al. 1999, Hua et al. 2007). Granzyme M also enhances granzyme B- mediated killing by inactivating the granzyme B inhibitor proteinase inhibitor 9 (Mahrus et al. 2004). In addition to their apoptotic function, granzymes H and M both cleave proteins essential for viral replication (Andrade et al. 2007, van Domselaar et al. 2010).

37

NK cells can also trigger cell death independently of granzymes, by ligation of the TNF superfamily receptors, Fas (CD95) and the TRAIL receptors DR4 and DR5. Binding of trimeric or hexameric Fas or TRAIL to receptor trimers causes dimerization of the trimers and organisation of receptor complexes into large clusters (Chan et al. 2000, Siegel et al. 2000, Holler et al. 2003, Valley et al. 2012). Fas- associated death domain (FADD) proteins are recruited to the cytoplasmic death domains of the activated Fas/TRAIL receptors (Boldin et al. 1995, Chinnaiyan et al. 1995, Kischkel et al. 2000, Sprick et al. 2000). FADD then recruits pro-caspase 8 to the receptor complex, triggering proximity-induced dimerisation (Muzio et al. 1996, Medema et al. 1997b, Muzio et al. 1998, Kischkel et al. 2000, Sprick et al. 2000, Boatright et al. 2003). Dimerisation is essential for inducing autoproteolytic activity of pro-caspase 8 and renders it susceptible to cleavage by other nearby pro-caspase 8 dimers (Muzio et al. 1998, Yang et al. 1998b, Chang et al. 2003, Donepudi et al. 2003). Pro-caspases are cleaved first between the large and small subunits, and then at a second site separating the large subunit from the pro-domain (Chang et al. 2003). Substrate specificity is altered upon processing: whereas pro-caspase 8 specifically cleaves other pro-caspase 8 molecules, mature caspase 8 instead cleaves pro-caspase 3, triggering the same apoptotic pathways as described for granzyme B (Stennicke et al. 1998, Chang et al. 2003). Caspase 8 also cleaves receptor interacting protein (RIP), which activates activator protein 1 (AP-1) and nuclear factor (NF)-κB signalling and initiates necrotic cell death when the apoptotic pathway is blocked (Lin et al. 1999, Holler et al. 2000). There is evidence that caspase 10, another apical caspase activated upon dimerization, can also be recruited to FADD following Fas or TRAIL stimulation, however whether or not this plays a significant role in Fas/TRAIL-mediated apoptosis is contraversial (Kischkel et al. 2001, Wang et al. 2001b, Sprick et al. 2002, LaFont et al. 2010). Caspase 10 also cleaves caspase 3 and therefore may provide and alternative apoptotic pathway in caspase 8-deficient cells (Stennicke et al. 1998, Wang et al. 2001b).

1.1.5 NK Cells Produce a Range of Cytokines and Chemokines

In addition to their direct role in killing infected or abnormal cells, NK cells are also able to exert influence on many other aspects of the immune response, via cytokine production, contact-dependent stimulation or inhibition and killing of other leukocytes such as T cells, B cells and DCs.

38

Activated NK cells release large quantities of IFNγ, a pleiotropic cytokine which is also produced by Thelper (Th)1, CD8+ T cells and invariant natural killer T (iNKT) cells (Schroder et al. 2004). IFNγ production can be induced by cytokines (particularly IL-12 with IL-18), ligation of activating receptors upon encounter with target cells, interactions with other leukocytes, or a combination of these (Cassatella et al. 1989, Cuturi et al. 1989, Fehniger et al. 1999, Fauriat et al. 2010a). It has effects on antigen presentation, antibody production, T cell polarisation, macrophage function, leukocyte trafficking, cell cycle and apoptosis (Schroder et al. 2004).

IFNγ upregulates expression of MHC class II and CIITA (class II, major histocompatibility complex, transactivator), a master regulator of the MHC class II antigen presentation pathway (Figueiredo et al. 1989, Chang et al. 1995, Barrerra et al. 1997, O’Keefe et al. 2001, Valledor et al. 2008). It also enhances expression of cathepsins and lysosomal proteases involved in peptide production for the MHC class II pathway and induces expression of new proteasome subunits that increase the quantity and diversity of peptides available for MHC class I (Schroder et al. 2004). In addition, IFNγ upregulates co-stimulatory molecules on antigen-presenting cells (APCs), leading to more efficient activation of antigen-specific CD4+ T cells (Menèndez Iglesias et al. 1997, Lee et al. 2006).

If present during T cell receptor (TCR) stimulation, IFNγ has both direct and indirect effects on T cell polarisation, promoting Th1 and Treg and inhibiting Th2 and Th17 differentiation (Harrington et al. 2005, Park et al. 2005, Murugaiyan et al. 2010). Effects on dendritic cells are important in this: IFNγ promotes production of the Th1- promoting cytokine IL-12 and the Treg-promoting cytokine IL-27, but inhibits expression of the Th17-polarising protein osteopontin by DCs (Murugaiyan et al. 2010). It also has effects on B cells, promoting isotype switching to IgG2a and antagonising IL-4-induced expression of the FcεR2 receptor for IgE (Rousset et al. 1988, Gao et al. 2001).

IFNγ enhances macrophage function by upregulating Fc receptors, promoting phagocytosis and inducing transcription of NADPH oxidase subunits and inducible nitric oxide synthase (iNOS), causing generation of ROS and reactive nitrogen species (RNS) (Nathan et al. 1983, Newburger et al. 1988, Cassatella et al. 1990a, Cassatella et al. 1990b, Gupta et al. 1992, Zhang et al. 2012). It also promotes leukocyte

39 trafficking by inducing dilation of blood vessels and expression of chemokines and adhesion molecules involved in adhesion of leukocytes to vascular endothelial cells (Schroder et al. 2004, Kebir et al. 2009). In contrast to its activating effects, IFNγ can also act to inhibit cell proliferation and increases sensitivity of activated T cells and myeloid cells to pro-apoptotic signals by promoting expression of death receptors, caspases 8 and 9, protein kinase R (PKR) and various components of the mitochondrial apoptotic pathway (Refaeli et al. 2002, Li et al. 2007).

IFNγ binds to the type II IFN receptor, composed of IFNGR1 and IFNGR2 subunits, which associate with Janus kinase (JAK)1 and JAK2 respectively (Platanias 2005). Binding of IFNγ causes the receptor subunits to dimerise, leading to autophosphorylation of JAKs, which then phosphorylate the signal transducer and activator of transcription 1 (STAT1) transcription factor at tyrosine701. Phosphorylated STAT1 dimerises and translocates to the nucleus, where it binds to Gamma-activated sites (GAS) in the DNA and modulates transcription. For optimal transcriptional induction, co-activators are required, such as the histone acetyltransferases, p300 and cAMP-responsive-element-binding protein (CBP), which are involved in chromatin remodelling (Zhang et al. 1998, Da Fonseca et al. 2001, Varinou et al. 2003, Li et al. 2010a). Interaction of STAT1 with these co-activators requires serine727 phosphorylation, mediated by protein kinase B (PKB), protein kinase C (PKC)δ, casein kinase 2, calmodulin-dependent protein kinase II (CamKII), ERK or JNK, several of which are activated by other IFNγ-induced signalling pathways such as the PI3K pathway (Nair et al. 2002, Uddin et al. 2002, Deb et al. 2003, Varinou et al. 2003, Harvey et al. 2007, Valledor 2008, Li et al. 2010a). In addition to its role in phosphorylation of STAT1, JNK1 also induces upregulation of MHC class II molecules and stabilises CITTA mRNA (Valledor et al. 2008).

One of STAT1’s direct target genes is the transporter associated with antigen processing (TAP) 1 (Min et al. 1996, Cramer et al. 2000, Rouyez et al. 2005). TAP transports peptides from the cytoplasm to the endoplasmic reticulum, where they bind to MHC class I, and is essential for presentation of intracellular antigen by MHC class I (Spies et al. 1991, Neefjes et al. 1993, de la Salle et al. 1994, Owen et al. 1999, Johnson et al. 2000). However, many of STAT1’s effects are mediated via expression of other transcriptional master regulators containg GAS sites in their promoter,

40 including interferon regulatory factor 1 (IRF1) and Ccaat-enhancer-binding protein (C/EBP)β (Poli et al. 1998, Taniguchi et al. 2001).

C/EBP binding motifs are found in promoters of pro-inflammatory genes including IL-6, IL-1β, TNFα, IL-8, IL-12, granulocye-colony stimulating factor (G-CSF), iNOS, lysozyme, myeloperoxidase and neutrophil elastase (Poli 1998). IRF1 is serine phosphorylated by PKC and CamKII and binds strongly to IRF-E sequences, found in many promoters, mostly those of pro-inflammatory genes (Taniguchi et al. 2001). IRF1 promotes antigen presentation by MHC class I and II by modulating expression of β2-microglobulin, B7-H1, CIITA, TAP1, the proteosomal subunit latent membrane protein 2 (LMP2) and proteosomal activators PA28α and -β, (Min et al. 1996, White et al. 1996, Hobart et al. 1997, O’Keefe et al. 2001, Taniguchi et al. 2001, Lee et al. 2006). It promotes anti-viral responses via induction of IFN-α and –β, and antibacterial responses via induction of iNOS in macrophages and microglia and the NADPH oxidase component gp91phox in myelomonocytic cells (Matsuyama et al. 1993, Kamijo et al. 1994, Eklund et al. 1998). By upregulating IL-12 expression in macrophages, it indirectly promotes Th1 differentiation and in the absence of IRF1 only a Th2 response can be mounted to infection (Lohoff et al. 1997, Taki et al. 1997). IRF-1 also induces IL-15 in bone marrow stromal cells, promoting differentiation and activation of NK cells (Ogasawara et al. 1998). The proinflammatory role of IRF1 is highlighted by its role in autoimmunity: mice deficient in IRF1 are protected against experimental autoimmune encephalomyelitis (EAE) and collagen-induced arthritis (Tada et al. 1997).

Strong p38 activation has also been observed following IFNγ stimulation of macrophages (Valledor et al. 2008, Matsuzawa et al. 2012). p38 stabilises transcripts for CCR5, CXCL10, CCL5 and TNFα, increases expression of CXCL9, IL-1β and iNOS2 and promotes autophagy (Valledor et al. 2008, Matsuzawa et al. 2012).

TNFα is also commonly produced by NK cells and regulates many of the same genes as IFNγ, since NF-κB, a key target of TNFα signalling, not only binds to the IRF-1 promoter, enhancing its expression, but also interacts with phosphorylated IRF-1 to synergistically induce expression of multiple target genes (Taniguchi et al. 2001). Most target genes can be induced by either IFNγ or TNFα alone, but are more effectively induced when both are present (Schroder et al. 2004).

41

Binding of TNFα to TNF-R1 induces receptor trimerisation and release of the Silencer of Death Domain inhibitory protein (Li and Lin 2008). This allows binding of TNF Receptor-Associated Death Domain (TRADD), which recruits FADD, TNF-R Associated Factor 2 (TRAF2) and RIP. TRAF2 oligomerisation induces binding and activation of mitogen-activated protein kinase kinase kinases (MAP3Ks) such as MAP/ERK kinase kinase 1 (MEKK1) and TGFβ-activated kinase 1 (TAK1), triggering a mitogen-activated protein kinase (MAPK) cascade leading to activation of JNK, p38 and IκB kinase (Baud et al. 1999, Li and Lin 2008, Fan et al. 2010).

JNK and p38 activate the AP-1 family of transcription factors, which consists of homodimers or heterodimers of basic leucine zipper domain (bZIP) proteins such as Jun, Fos and activating transcription factor (ATF) (Karin et al. 1997). C-Jun and Fos dimers promote cell proliferation by inducing expression of cyclins and both downregulating and inhibiting function of the tumour suppressor p53 (Shaulian and Karin 2001, Hommura et al. 2004). AP-1 transcription factors also target various genes involved in motility and invasion, thereby upregulating expression of cathepsin L and matrix metalloproteinases (MMPs) 1, 3 and 9 (involved in cellular matrix degradation), CapG (tropomysin, capping protein (actin filament), gelsolin-like) and metastasis-associated protein 1 (Mts1) (cytoskeletal regulation) and CD44 (adhesion) and downregulating the adhesion molecules immunoglobulin superfamily member 4 (IGSF4) and protocadherin gamma-C3 (Ozanne et al. 2007).

IκB kinase phosphorylates the NFκB inhibitor IκB, thereby targeting it for ubiquitination and degradation (Chen 2005). NF-κB therefore becomes activated and translocates to the nucleus, where it activates transcription of multiple target genes, including IRF1, TAP1 and LMP2 (Min et al. 1996, Taniguchi et al. 2001, Marqués et al. 2004, Li and Lin 2008). NF-κB has been described as ‘a central mediator of the human immune response’, directly regulating over 150 target genes (Pahl 1999). These include at least 27 cytokines and chemokines (mostly pro-inflammatory), cytokine receptors, adhesion molecules such as ICAM1 and vascular cell adhesion molecule 1 (VCAM1), immunoglobuins, matrix metalloproteinases, lysozyme, iNOS, and genes involved in antigen presentation, complement pathways and regulation of ROS production (Pahl 1999).

42

TNFα signalling can also lead to activation of caspases (Deng et al. 2003, Li and Lin 2008). Caspase 8 does not appear to be recruited to TNF receptors as it is to Fas or TRAIL, rather it is activated indirectly by mitochondrial Smac (second mitochondria- derived activator of caspases) and DIABLO (direct IAP binding protein with low pI) release following Jnk-mediated Bid cleavage (Deng et al. 2003, Harper et al. 2003). However, TNF-α does not usually induce apoptosis or necrosis due to the protective effects of NF-κB activation (Sakon et al. 2003, Morgan et al. 2008). NF-κB induces expression of cellular inhibitor of apoptosis proteins (cIAPs), the capase 8 inhibitor cFLIP (cellular FLICE-like inhibitory protein), antioxidant enzymes and the Bcl-2 family protein A1, which inhibits cytochrome C release and caspase activation (Mattson et al. 1997, Djavaheri-Mergny et al. 2004, Morgan et al. 2008). Sustained JNK activation is required for both apoptosis and necrosis and this is dependent on ROS, which inhibit JNK phosphatases by converting a catalytic cysteine residue to sulfenic acid and induce dissociation of the JNK inhibitor glutathione S-transferase π (Adler et al. 1999, Deng et al. 2003, Kamata et al. 2005). NF-κB promotes expression of antioxidant enzymes thioredoxin and manganese superoxide dismutase, thus preventing sustained JNK activation and protecting against both TNFα-induced and exogenous ROS-induced apoptosis (Mattson et al. 1997, Javelaud et al. 2001, Djavaheri-Mergny et al. 2004).

Another cytokine produced by NK cells is granulocyte-macrophage colony stimulating factor (GM-CSF), which is effectively induced by α-CD16 and IL-2 or by combinations of cytokines such as IL-15 and IL-18 or IL-1 and TNFα (Cuturi et al. 1989, Fehniger et al. 1999, Hamilton 2002). GM-CSF promotes differentiation of bone marrow progenitors into granulocytes, macrophages and eosinophils, stimulates differentiation of monocytes into macrophages and promotes survival of monocytes, macrophages and neutrophils (Metcalf et al. 1986, Young et al. 1990, Coxon et al. 1999, Hashimoto et al. 2001). It has also been used together with IL-4 to derive dendritic cells from monocytes in vitro (Agakawa et al. 1996). GM-CSF and TNF-α both enhance phagocytosis, ADCC and N-formyl-methionyl-leucyl-phenylalanine- induced oxidative burst in neutrophils (Klebanoff et al. 1986, Metcalf et al. 1986, Perussia et al. 1987, Weisbart et al. 1987, Lopez et al. 1988). Eosinophil ADCC, phagocytosis and ROS production is also promoted by GM-CSF, as is eosinophil and neutrophil adhesion to endothelial cells (Lopez et al. 1988, Walsh et al. 1990).

43

Stimulation of monocytes with GM-CSF again enhances adherence to vascular endothelial cells, as well as enhancing ADCC and lipopolysaccharide (LPS)-induced cytokine production and increasing activity of urokinase plasminogen activators, which convert plasminogen to plasmin, a serine protease which degrades fibrin and activates collagenases to promote degradation of the extracellular matrix (Gamble et al. 1989, Wing et al. 1989, Hart et al. 1991).

NK cells incubated with malignant melanoma cells and IL-2 also produce large quantities of IL-5, which are augmented by addition of IL-4 (Warren et al. 1995). Long-term culture with IL-4 in the presence of an IL-12- neutralising antibody polarises NK cells to an “NK2” phenotype, characterised by high production of IL-5 and IL-13, with impaired IFNγ secretion (Peritt et al. 1998). IL-5 was first identified as an eosinophil proliferation and differentiation factor (Lopez et al. 1986). It also activates mature eosinophils, promoting ADCC, phagocytosis, granule polarisation and ROS production, while inhibiting apoptosis (Lopez et al. 1986, Lopez et al. 1988, Walsh et al. 1990, Yamaguchi et al. 1991). In addition, IL-5 upregulates the complement receptor C3bi and increases eosinophil adhesion by stimulating CDllb expression (Lopez et al. 1986, Lopez et al. 1988, Walsh et al. 1990). IL-5-induced eosinophil infiltration during infection is important for induction of a protective antibody response and IL-5 production by NK cells specifically has been shown to have an important role in recruitment of eosinophils in vivo (Walker et al. 1998, Herbert et al. 2000). Activation of eosinophils is important for control of parasitic infections (Herbert et al. 2000, Vallance et al. 2000). Unlike GM-CSF, IL-5 does not mediate these effects in neutrophils (Lopez et al. 1988, Walsh et al. 1990). However, effects on B and T lymphocytes have been identified, including enhanced tumour- specific cytotoxic T cell responses and increased production of IgG or IgM in B cells stimulated with 2,4-dinitrophenyl or Staphylococcus aureus Cowan I (SAC) respectively (Harada et al. 1987, Nagasawa et al. 1991, Bertolini et al. 1993, Morikawa et al. 1993).

IL-13 induces isotype switching and IgE production and upregulates expression of the IgE receptor FcεRII and MHC class II on monocytes, macrophages and B cells (Cocks et al. 1993, de Waal Malefyt et al. 1993a, McKenzie et al. 1993, Punnonen et al. 1993, Defrance et al. 1994, Sinha et al. 2008). B cell proliferation and production

44 of IgM and IgG4 in response to CD40 stimulation are also enhanced by IL-13 (Cocks et al. 1993, Defrance et al. 1994, McKenzie et al. 1993). EAE induction in IL-13- deficient mice suggests that IL-13 enhances inflammation in this model, promoting T cell proliferation, macrophage central nervous system (CNS) infiltration and production of IL-12 and IFNγ (Sinha et al. 2008). However, IL-13 also has anti- inflammatory functions, protecting against endotoxic shock by antagonising effects of TNFα and LPS stimulation (Muchamuel et al. 1997). Activation of MEK and JNK in myeloid cells is inhibited by IL-13, thereby preventing activation of AP-1 and NF-κB in response to TNF-α or LPS and inhibiting expression of multiple cytokine and chemokine genes and production of ROS and RNS (de Waal Malefyt et al. 1993a, Minty et al. 1993, Sozzani et al. 1995, Bogdan et al. 1997, Manna and Agarwal 1998). IL-13 also protects against apoptosis by inhibiting caspase 3 activation (Manna and Agarwal 1998).

NK cells can also produce the regulatory cytokine IL-10. In the absence of polarising cytokines, only around 7% of freshly isolated human peripheral blood NK cells from healthy individuals make IL-10 protein in response to phytohaemagglutinin (PHA) stimulation, most of which are found within the CD56dim KIR+ subset (Deniz et al. 2008). However, NK cells from hepatitis C virus (HCV) -infected patients secrete substantial amounts of IL-10 in response to melanoma target cells (De Maria et al. 2007). Upregulation of NKp30 on these NK cells suggests interaction with DCs, perhaps forming a negative feedback loop whereby prolonged DC IL-12 production stimulates NK cell IL-10 production and suppressive function (De Maria et al. 2007). In vitro, strong IL-10 production can be induced in nearly all NK cells (including both CD56dim and CD56bright) by culture with IL-12 in combination with IL-2, IL-15 or IL- 4-neutralising antibody (Mehrotra et al. 1998, Peritt et al. 1998, Fehniger et al. 1999). IL-10 mRNA is also induced by stimulating antibodies directed against CD2 and CD16 and by vitamin D3 plus dexamethasone (Deniz et al. 2008). IL-10 transcription reaches its peak after 24 hours, but protein secretion is delayed (Mehrotra et al.1998, Maroof et al. 2008). For instance, during Leishmania donovani infection, NK cells begin secreting IL-10 and acquire regulatory function after around 21 days (Maroof et al. 2008). Its slow induction in comparison to other cytokines is consistent with a role in preventing chronic inflammation by suppressing activated leukocytes in the late stages of an immune response.

45

Incubation of APCs with IL-10 suppresses their ability to drive antigen-induced T cell + proliferation, CD8 T cell cytotoxic function and production of both Thelper (Th)-1 and Th2 cytokines by CD4+ T cells (Del Prete et al. 1993, Groux et al. 1998). This is due to downregulation of MHC class I and II and adhesion/ co-stimulatory molecules such as CD80, CD86 and ICAM1 (de Waal Malefyt et al. 1991, Groux et al. 1998). IL-10 also directly inhibits IL-2 production and proliferation of T cells stimulated via their TCR and blocks activation of the PI3K pathway by CD28 co-stimulation (de Waal Malefyt et al. 1993b, Taga et al. 1993, Akdis et al. 2000). In monocytes, IL-10 antagonises lipopolysaccharide (LPS)- or IFNγ-mediated induction of IL-1α and –β, IL-6, IL-8, TNF-α, G-CSF and GM-CSF transcription (Hashimoto et al. 2001, de Waal Malefyt et al. 1991). IL-12 production is also inhibited, which in turn leads to a reduction in IFNγ (D’Andrea et al. 1993). However, IL-10 appears to have positive effects on the humoral response. Though it promotes apoptosis of newly activated B cells, when added during the maturation stage it increases cell viability, promotes differentiation into plasma cells and enhances the proliferative and immunoglobulin responses to antigen or CD40 stimulation (Rousset et al. 1991, Itoh et al. 1994, Itoh and Hirohata 1995). It also upregulates expression of FcγRI, II and III on monocytes and promotes FcγR-mediated phagocytosis (te Welde et al. 1992, Hashimoto et al. 1997).

NK cells also produce chemokines, including IL-8, RANTES (regulated on activation, normal T expressed and secreted /CCL5) and macrophage inflammatory proteins (MIP) 1α and 1β. IL-8 production can be induced by a variety of stimuli, including poly(I:C), IL-15 or activation of Fc receptors in the presence of IL-2, IL-12 or platelet factor 4 (Marti et al. 2002, Chuang et al. 2006, Roda et al. 2006, El-Shazly et al. 2011). IL-8 is a chemotactic factor for neutrophils and T cells (Larsen et al.1989, Leonard et al. 1991, Douglass et al. 1996). It induces shape change in neutrophils, promoting development of the long lamellipodia required for cell migration, upregulates β2-integrins and induces lectin shedding, facilitating their transmigration through blood vessel walls (Thelen et al. 1988, Huber et al. 1991). It also activates neutrophils to produce ROS and release granules containing antimicrobial proteins such as elastase and chemotactic factors for T cells and monocytes (Peveri et al. 1988, Thelen et al. 1988, Taub et al. 1996). In addition, IL-8 regulates angiogenesis by

46 enhancing endothelial cell proliferation, inhibiting apoptosis, upregulating MMP2 and MMP9 and promoting capillary tube reorganisation (Li et al. 2003).

RANTES, MIP1α and MIP1β are related chemokines with overlapping functions and receptor specificities. CCR3 binds only RANTES (Daugherty et al. 1996, Ponath et al. 1996a, Ponath et al. 1996b), CCR1 and CCR4 bind RANTES and MIP1α but not MIP1β (Gao et al. 1993, Neote et al. 1993, Ben-Baruch et al. 1995, Power et al. 1995), and CCR5 binds all three (Raport et al. 1996). Secretion of these chemokines can be induced in NK cells by multiple mechanisms, including culture with IL-2, IL- 12 and/or IL-15, incubation with K562 or antibody-coated tumour cells, CD16 crosslinking, or stimulation of the activating receptors NKG2D or 2B4 (Oliva et al. 1998, Fehniger et al. 1999, Roda et al. 2006, Fauriat et al. 2010a). All three are chemotactic factors for monocytes, macrophages, DCs and CD4+ T cells, particularly of the Th1 type (Schall et al. 1990, Taub et al. 1993, Wang et al. 1993, Uguccioni et al. 1995, Raport et al. 1996, Sozzani et al. 1997, Roth et al. 1998, Siveke and Hamann 1998, Kawai et al. 1999, Lee et al. 2000). Monocytes and Th1 cells upregulate integrins in response to these chemokines, increasing their ability to adhere to endothelial cells and migrate across the vascular endothelium (Taub et al. 1993, Vaddi and Newton 1994, Roth et al. 1998, Kawai et al. 1999). RANTES, MIP1α and - MIP1β also enhance macrophage NO2 production and phagocytosis (Villalta et al. 1998). B cell antigen-specifc antibody production is also increased, with MIP1α and MIP1β preferentially promoting IgG1 and IgG2b, while RANTES stimulates IgG2a and IgG3 (Lillard et al. 2001, Lillard et al. 2003). RANTES, and to a lesser extent MIP1α, induce basophil and eosinophil chemotaxis and leukotriene release (Rot et al. 1992, Bischoff et al. 1993, Daugherty et al. 1996, Beck et al. 1997, Lee et al. 2000). MIP1α is also a chemotactic factor for neutrophils, B cells and CD8+ memory T cells, which do not appear to migrate in response to either RANTES or MIP1β (Schall et al.1990, Schall et al. 1993, Taub et al. 1993, Beck et al. 1997, Lee et al. 2000, Ramos et al. 2005). It also has effects on NK cells themselves, as a potent NK cell chemoattractant and enhancer of cytotoxicity (Taub et al. 1995, Bluman et al. 1996). Finally, MIP1α also has specific effects on stem cells: it is the only one of these chemokines to suppress proliferation of HSC and GM-CSF-stimulated early myeloid progenitors (Broxmeyer et al. 1993, Van Ranst et al. 1996).

47

1.1.6 NK Cells Interact with Other Leukocytes and Regulate their Function

NK cells engage in contact-dependent interactions with other leukocytes, in some cases promoting their activation and in other cases inducing cell death.

Interaction of NK cells with B cells results in contact-dependent reciprocal activation. Activated B cells interact with NK cells via LFA1-ICAM1 and CD40-CD40L, inducing IFNγ and TNFα production in both the NK cell and the B cell (Becker et al. 1990, Blanca et al. 2001, Gao et al. 2005). IFNγ then induces B cell proliferation and TNF-α promotes both proliferation and antigen-specific immunoglobulin production (Becker et al. 1990, Blanca et al. 2001). NK cells also induce transcripts for Iγ2a and activation-induced cytidine deaminase (AID), which are necessary, though not sufficient, for isotype switching to IgG2a (Gao et al. 2001, Gao et al. 2005). Recently it has been shown that NK cells can also enhance antigen processing and presentation in B cells, resulting in increased proliferation of antigen-specific CD4+ T cells (Jennings and Yuan 2009).

CD56bright NK cells activated with IL-12, IL-15 and IL-18 promote TNF-α production by monocytes in a contact-dependent manner (Dalbeth et al. 2004). In turn, contact with monocytes synergises with cytokines to induce NK cell IFNγ production. IL-15- activated NK cells can also induce contact-dependent differentiation of monocytes into DCs with a preference for Th1 priming, partly due to secretion of GM-CSF (Zhang et al. 2007).

Both resting and activated NK cells can induce maturation of immature (i) DCs (Carbone et al. 1999, Gerosa et al. 2002, Tosi et al. 2004). Maturation is induced most efficiently by NKG2Alow KIR- NK cells (Vitale et al. 2005). Formation of NK-DC conjugates is important in promoting upregulation of HLA-DR and co-stimulatory molecules CD83 and CD86 during DC maturation and activation (Gerosa et al. 2002, Piccioli et al. 2002, Tosi et al. 2004, Gerosa et al. 2005). NK cells in contact with iDCs are activated via NKp30 and appear to upregulate TNFα production in the iDC, which promotes DC maturation in an autocrine manner (Gerosa et al. 2002, Piccioli et al. 2002, Vitale et al. 2005). Neutralisation of IFNγ has relatively little effect on maturation, implying that other mechanisms are required (Gerosa et al. 2002, Piccioli et al. 2002, Vitale et al. 2005). These may include secretion of the leaderless cytokine high-mobility group box 1 (HMGB1), a powerful inducer of DC maturation, which is 48 produced spontaneously by activated NK cells and induced in resting NK cells upon co-culture with iDC (Semino et al. 2005). HMGB1 release is induced by IL-18, which iDCs secrete directly into the synapse via lysosomal exocytosis (Semino et al. 2005).

Activated NK cells can also selectively lyse iDCs, which express lower levels of HLA-E and HLA class I than mature DCs (Carbone et al. 1999, Wilson et al. 1999, Spaggiari et al. 2001, Della Chiesa et al. 2003). Only NKG2A+ NK cells lacking self- specific KIR are able to kill autologous iDCs, with NKG2Alow NK cells killing more effectively than NKG2Ahigh (Della Chiesa et al. 2003). Killing of iDCs is dependent on NKp30 (and to a lesser extent NKp46) and is most likely mediated by perforin and granzymes (Wilson et al. 1999, Spaggiari et al. 2001, Ferlazzo et al. 2002, Vitale et al. 2005).

The outcome of the interaction between NK cell and iDC is dependent on the ratio of the cell types present. If NK cells outnumber DCs, they are likely to kill them, with most efficient killing at ratios of 5:1 (NK:iDC) and above (Piccioli et al. 2002, Semino et al. 2005). However, at a 1:5 ratio, with iDCs outnumbering NK cells, iDCs are not efficiently lysed (in fact their survival is enhanced) and instead mature (Piccioli et al. 2002, Semino et al. 2005, Vitale et al. 2005). This stimulatory effect on DC maturation was observed at raios as low as 1:40 (NK:iDC) (Piccioli et al. 2002). Induction of iDC maturation or apoptosis during an immune response is likely to increase the chances of T cells interacting with mature, rather than immature DCs, thus promoting activation rather than anergy.

In addition to being responsible for most of the early IFNγ production during infection, NK cells are also important in stimulation of mature dendritic cell function and IL-12 production, which in turn induces IFNγ production in T cells, the major IFNγ producers at later stages of infection (Mocikat et al. 2003) (see figure 1-iv). CD56bright NK cells colocalise with DCs in human lymph nodes (Ferlazzo et al. 2004). LPS-activated DC in the lymph node can also release factors which recruit further NK cells from the blood (Martín-Fontecha et al. 2004). Meanwhile, IL-18 produced by macrophages and DCs in the periphery induces CCR7 expression on CD56dim NK cells (which do not usually express this receptor), allowing them to migrate towards CCL19 and CCL21 in the lymph node (Mailliard et al. 2005). Activated DCs express IL-15 on their cell surface, which is presented to NK cells by IL-15Rα and induces

49 their proliferation (Ferlazzo et al. 2004). Cytoskeletal rearrangement in the DC permits synaptic delivery of IL-12 directly to the NK cell (Borg et al. 2004). Together with NCR ligation, this induces secretion of IFNγ, which reciprocally upregulates IL- 12 expression in the DC (Ferlazzo et al. 2002, Borg et al. 2004, Ferlazzo et al. 2004, Gerosa et al. 2005). IFNγ production by NK cells is reported to be crucial for Th1 polarisation, due to its effects on DCs and its ability to induce the central Th1 transcriptional mediator, T-bet, in activated T cells (Afkarian et al. 2002, Martín- Fontecha et al. 2004). NK cells primed with IL-18 produce more IFNγ upon interaction with DCs and are particularly effective at inducing DC IL-12 production and promoting Th1 as opposed to Th2 differentiation (Mailliard et al. 2005).

Activated NK cells can also induce a stable type-1 polarised “effector/ memory” DC, termed “DC1” (Mailliard et al. 2003). DC1 continue to efficiently prime T cells and produce IL-12 when NK cells are removed from culture. Compared to DCs matured in the absence of NK cells, DC1 produce more IL-12 upon CD40L stimulation and strongly induce production of IFNγ but not IL-4 by antigen-specific CD8+ T cells and naïve CD4+ T cells. In order to polarise DCs, NK cells must be activated by at least two signals e.g. IL-2 and IL-18 or K562 cells and IFNα (Mailliard et al. 2003).

IFNα and –β produced by activated myeloid and plasmacytoid (p) DCs enhance NK cytolytic activity (Gerosa et al. 2005). This promotes subsequent lysis of susceptible targets including iDCs, however mature DCs, particularly pDCs, are themselves relatively resistant to NK-mediated cell death (Gerosa et al. 2005).

50

Figure 1-iv: NK cells stimulate T cell activation and Th1 polarisation both directly and indirectly. Stimulation via NKp30 causes the NK cell to release unidentified soluble factors that induce DC production of TNFα, which acts in an autocrine manner to drive DC maturation. Meanwhile, IL-18 from DCs drives NK cell production of HMGB1, which also induces DC maturation and upregulation of costimulation molecules. IL-12, IL-15 and IL-18 produced by mature DCs activate NK cells to upregulate costimulation molecules and produce IFNγ. IFNγ promotes expression of the Th1 transcription factor T-bet in CD4+ T cells and enhances production of IL-12 by DCs, which also promotes Th1 polarisation. Contact with co-stimulatory molecules such as OX40L and CD80 on NK cells enhances priming of antigen-specific T cells by mature DCs.

In addition to promoting DC priming, NK cells also have direct effects on T cells. Firstly, they can enhance T cell responses by engaging the co-stimulatory receptors CD28 and OX40 (Zingoni et al. 2004). Co-stimulation is required for optimal proliferation, cytokine production and long-term survival in response to TCR- mediated antigen recognition (Gramaglia et al. 2000, Zingoni et al. 2005). OX40L is induced on NK cells stimulated with cytokines (IL-2, -12 or -15) followed by ligation of activating receptors (NKG2D, CD16 or 2DS2) and binds to OX40, which is

51 expressed primarily on activated T cells (Ohshima et al. 1998, Gramaglia et al. 2000, Zingoni et al. 2004). Cytokine-activated NK cells also express CD80 and CD86, ligands for the CD28 co-stimulatory receptor (Hanna et al. 2004a, Zingoni et al. 2004). NK cells activated by both cytokine and receptor stimulation strongly promote CD4+ T cell proliferation and production of IL-2 and IFNγ, but not IL-4 (Hanna et al. 2004a, Zingoni et al. 2004). Blocking experiments indicate that both OX40 and CD28 are involved, though OX40 appears to be more important (Zingoni et al. 2004). These results are in contrast to a previous study where OX40 ligation enhanced Th2 polarisation of human umbilical cord-derived CD4+ T cells (Ohshima et al. 1998). This suggests that OX40L itself might be a more general co-activator of T cells, and that the Th1 bias observed in the former study was due to other mechanisms, such as NK cell IFNγ production. Interestingly, in vivo OX40 stimulation in mice induces a Th1 bias during the initial response to antigen, but enhances both Th1 and Th2 recall responses (De Smedt et al. 2002). The latter is likely due to enhanced expansion and survival of antigen-specific T cells during the latter phase of the primary response, which means that more memory cells remain to deal with a secondary challenge (Gramaglia et al. 2000).

NK cells themselves can also have an antigen presenting function. HLA-DR- expressing NK cells represent less than 10% of NK cells in healthy peripheral blood and around 50% of lymph node NK cells (Burt et al. 2008, Evans et al. 2011). HLA- DR+ NK cells proliferate preferentially in response to IL-2 or IL-15 stimulation and upregulate CXCR3, a receptor promoting migration to lymph nodes and sites of inflammation (Evans et al. 2011). Both CD56dim and CD56bright NK cells can process and present soluble protein antigens to CD4+ T cells via MHC class II (Roncarolo et al. 1991). This stimulates T cell proliferation and production of IFNγ and IL-2, which in turn enhances IFNγ production by NK cells (Hanna et al. 2004a, Evans et al. 2011). NK cells are able both to expand antigen-specifc memory T cells and to prime naïve T cells, albeit not as effectively as DCs (Hanna et al. 2004a). They also induce IFNγ production in CD8+ T cells, again less effectively than DCs (Burt et al. 2008). In the presence of plentiful antigen, NK cells induce a similar level of T cell proliferation to that induced by B cell lines, however at suboptimal antigen concentrations NK cells are less effective (Roncarolo et al. 1991).

52

NK cell antigen presenting function appears to be coupled to cytotoxicity. Activated NK cells that have killed influenza-infected target cells can specifically activate T cells recognising influenza HA peptides (Hanna et al. 2004a). This suggests that NK cells that have killed infected cells are presenting peptides derived from the cell that they have killed, thus bypassing the need for DCs. HA binds to NKp46, triggering internalisation of the receptor-HA complex to MHC class II-loading compartments, allowing the protein to be processed and presented to CD4+ T cells (Hanna et al. 2004a). B cell antibody production may be important for internalisation of proteins not directly recognised by NK cell receptors: incubation of HA protein with HA- specific antibodies leads to formation of immune complexes that bind to CD16 and these can also be internalised and processed for peptide presentation (Hanna et al. 2004a).

NK cells also appear to be important in inducing T cell memory. NK cells which have been activated by MHC class I-deficient RMA-S tumour cells promote subsequent T- cell mediated rejection of MHC class I-sufficient RMA cells, by a mechanism involving CD70-CD27 interaction (Kelly et al. 2001). Vaccination with ovalbumin (OVA)-expressing targets was found to be more effective at inducing a CD8+ recall response and protecting against subsequent infection with OVA-expressing Listeria monocytogenes when the original target cells lacked the MHC class I molecule Kb (Krebs et al. 2009). Kb-sufficient targets, which would be lysed by T cells rather than NK cells, were poor inducers of memory responses, suggesting that direct stimulation of CD8+ T cells in the absence of NK cell help is not effective at generating memory cells. NK cell enhancement of CD8 memory appears to be cell intrinsic and is dependent on MyD88/ TIR-domain-containing adapter-inducing interferon-β (Trif) signalling in the T cell (Krebs et al. 2009). NK cells also promoted OVA-specific IgG production and enhanced the CD4+ IFNγ response by a cell extrinsic mechanism, probably due to enhanced DC priming. Interaction of the murine HLA-E homolog, Qa1-Qdm, with NKG2A has also been reported to play an important role in expansion and memory function of CD4+ T cells (Lu et al. 2007).

NK cells also have an important role during the memory response itself. Though they do not themselves have antigen-specific memory, they are rapidly activated during a recall response. NK cells represent over 70% of IFNγ-secreting and degranulating

53 cells 12-18 hours after restimulation of peripheral blood mononuclear cells (PBMC) from individuals vaccinated against rabies (Horowitz et al. 2010). Activation of NK cells is dependent on IL-2 produced by CD4+ memory T cells, as well as IL-12 and IL-18 production by myeloid cells. Large amounts of IL-2 are produced within 6 hours of restimulation with inactivated rabies virus, whereas IFNγ production by T cells is initially low. Both pre- and post-vaccination NK cells respond to rabies virus when cultured with post-vaccination T cells, but neither degranulate when stimulated in the presence of pre-vaccination T cells, indicating that the memory function is mediated by T cells, not NK cells. Similar enhancement of NK cell responses in vaccinated individuals has been observed in PBMC stimulated with influenza virus or malarial antigens (Long et al. 2008, Horowitz et al. 2012).

Activated NK cells have also been shown to lyse Tregs expanded by Mycobacterium tuberculosis, which upregulates the NKG2D ligand ULBP1 on Tregs but not T effector cells (Roy et al. 2008). This might help NK cells to selectively target Tregs that would otherwise impair the immune response against the bacterium.

However, NK cells can also act to dampen the immune response. Following α-CD3 stimulation, T cells phosphorylate ataxia-telangiectasia mutated, a kinase involved in upregulation of NKG2D ligands following DNA damage (Cerboni et al. 2007). Both CD4+ and CD8+ T cells upregulate ligands for NKG2D in response to antigen stimulation or other stimuli and are lysed by IL-2-activated NK cells in an NKG2D/LFA-1-dependent, granzyme-dependent manner (Rabinovich et al. 2003, Cerboni et al. 2007, Nielsen et al. 2012). NK cells preferentially kill activated, rather than resting T cells, despite slight upregulation of MHC class I on activated T cells (Rabinovich et al. 2003, Molinero et al. 2006, Nielsen et al. 2012). Neutralisation of NKG2A further enhances killing of CD4+ T cells, indicating that expression of HLA- E on activated T cells does confer some degree of protection (Nielsen et al.2012). Killing of activated CD8+ T cells by NK cells occurs during lymphocytic choriomeningitis virus infection and has a profound effect on the anti-viral immune response, promoting viral persistence (Lang et al. 2012).

IL-2- activated NK cells can lyse monocytes (Spaggiari et al. 2001). Furthermore, macrophages stimulated with high doses of LPS are killed by both resting and activated NK cells (Nedvetzki et al. 2007). They upregulate the NKG2D ligands

54

ULBP1-3 and MICA, which induce NK cell proliferation, IFNγ production and formation of a supramolecular activating cluster, leading to release of cytotoxic molecules and death of the macrophage with which it is interacting (Nedvetzki et al. 2007). NK cells are also activated by NKp46 ligands on neutrophils and induce their apoptosis via the Fas/FasL pathway (Thorén et al. 2012).

1.1.7 NK Cells can be Divided into Subsets with Distinct Functions

Natural killer cells can be divided into subsets with different functional characteristics, which are typically defined by the presence/absence or expression level of one particular surface marker. The most well characterised human NK cell subsets are the CD56dim and CD56bright described by Lanier et al. in 1986. In peripheral blood, the majority of NK cells (around 90%) are CD56dim, whereas CD56bright NK cells predominate in lymph nodes and tonsil and are enriched at sites of inflammation (Cooper et al. 2001a, Fehniger et al. 2003, Dalbeth et al. 2004, Ferlazzo et al. 2004). CD56dim NK cells are considered to account for cytotoxicity, while CD56bright NK cells are thought to be the main producers of cytokines (Cooper et al. 2001b). The latter assumption is based mainly on the higher levels of both pro- and anti-inflammatory cytokine secretion following stimulation with phorbol myristate acetate (PMA) and ionomycin, LPS or cytokines (Cooper et al. 2001b, Jacobs et al. 2001). However, recent studies indicate that under certain circumstances CD56dim NK cells can also secrete large quantities of cytokines such as IFNγ and CD56bright NK cells can efficiently lyse certain kinds of target cell.

The nature of the activating signal is likely to be key in determining the relative response of each subset. Though CD56bright NK cells produce more IFNγ and other cytokines in response to stimulation with soluble mediators such as PMA/ionomycin, IL-12, IL-15 or IL-18, CD56dim NK cells are reported to produce more cytokines than CD56bright when stimulated with target cells or via activating receptors (Jacobs et al. 2001, Fauriat et al. 2010a, Juelke et al. 2010). An important study by Fauriat et al. showed that a much higher proportion of CD56dim compared to CD56bright NK cells produce IFNγ, TNFα, MIP1α and MIP1β following stimulation with K562 or Drosophila S2 cells transfected with ICAM-1, ULBP1 and CD48 (Fauriat et al. 2010a). Following stimulation with various combinations of IL-12, IL-15 and IL-18, CD56bright NK cells were more likely to produce IFNγ and TNFα. However the

55 frequency of MIP1α+ cells was similar between the two subsets, while MIP1β- producing cells were in fact more frequent within the CD56dim subset. When NK cells were stimulated simultaneously with both target cells and cytokines, the frequency of IFNγ-producing cells was similar, but a significantly higher proportion of CD56dim NK cells produced TNFα, MIP1α and MIP1β, demonstrating that CD56dim NK cells are also an important source of cytokines (Fauriat et al. 2010a). The level of cytokine and chemokine production by each subset, either per cell or overall secretion, was however not reported.

Differences in the kinetics of cytokine release between the two subsets have also been analysed. CD56dim NK cells produce IFNγ 2-4 hours after triggering of NKp46 or NKp30 (De Maria et al. 2011). Conversely CD56bright NK cells only begin to produce substantial amounts of IFNγ after 16 hours, by which time CD56dim IFNγ production has ceased. Resting CD56dim NK cells had 2.14-times more IFNγ transcript than CD56bright, which may explain their ability to respond so rapidly.

Though it is now clear that both subsets are capable of producing a range of cytokines and chemokines, there are differences in the secretory profiles of the two subsets. For example, CD56dim NK cells preferentially produce osteopontin, angiogenin and insulin-like growth factor (IGF)-1, whereas transforming growth factor (TGF)β-1/2/3, IL-15, IL-16, IL-18, GM-CSF and vascular endothelial growth factor (VEGF) are produced predominatly by CD56bright NK cells (Hanna et al. 2004b, Wendt et al. 2006). Amongst chemokines, higher expression of CCL1, CCL17, XCL1 and CXCL6 was observed in the CD56bright subset, while the CD56dim subset expressed higher levels of monocyte chemotactic protein (MCP)-1,2 and 3, MIP-1α and β, RANTES, IL-8, CXCL2 and eotaxins (Hanna et al. 2004b, Wendt et al. 2006). Upon activation, CD56bright NK cells also exhibited higher expression of TNF-superfamily members such as lymphotoxin-β, OX-40L, TRAIL and receptor activator of nuclear factor κB ligand (RANKL) (Wendt et al. 2006).

CD56dim NK cells are characterised by high expression of CD16, which is usually absent, or otherwise low in the CD56bright subset (Cooper et al. 2001b, Jacobs et al. 2001). This restricts ADCC function to the CD56dim subset (Nagler et al. 1989). CD56bright NK cells typically lack KIR, with the exception of KIR2DL4 (Jacobs et al. 2001, Wendt et al. 2006). Transcripts for CD58 (LFA-3), CD160, CD161, 2B4 and

56 lectin-like transcript 1 (LLT-1) costimulatory receptors and the CD3ζ adaptor (associated with CD16 and NCR signalling) are present in both subsets, but expressed at higher levels by CD56dim NK cells (Baume et al. 1992, Hanna et al. 2004b, Wendt et al. 2006). CD2, on the other hand, is upregulated in the CD56bright subset (Hanna et al. 2004b). Both subsets express similar levels of NKG2D and NKp30 on the cell surface, but NKp46 expression is slightly higher on CD56bright NK cells (Vitale et al. 2004). The inhibitory receptor NKG2A is expressed on all CD56bright NK cells and some CD56dim (Vitale et al. 2004).

CD56bright NK cells display poor cytotoxicity towards many of the targets that are lysed efficiently by CD56dim NK cells (Cooper et al. 2001b, Jacobs et al. 2001). This is likely to be due to lower expression of certain adhesion molecules, which may account for their reduced ability to form conjugates with K562 cells, and lower expression of cytotoxic molecules including perforin and granzymes A and B (Jacobs et al. 2001, Wendt et al. 2006, Chattopadhyay et al. 2009). Despite this, they are still capable of effectively lysing certain targets such as activated T cells (Bielekova et al. 2006, Jiang et al. 2011, Nielsen et al. 2012). Granzyme K is expressed almost exclusively in the CD56bright subset and is important for induction of T cell apoptosis by this population (Hanna et al. 2004b, Wendt et al. 2006, Jiang et al. 2011). IL-2 activated CD56bright NK cells also express higher levels of TRAIL than CD56dim NK cells and this pathway also contributes to killing of activated CD4+ T cells, which express DR4 and DR5 TRAIL receptors (Nielsen et al. 2012). The frequency of degranulating cells amongst the CD56bright subset is around double that of the CD56dim when incubated with autologous activated CD4+ T cells from healthy controls, however they lyse the T cells to a similar extent (Jiang et al. 2011, Nielsen et al. 2012). This may be due to higher levels of the more potent granzymes such as granzyme B in granules of CD56dim NK cells. However, during an inflammatory response, the greater IL-2 responsiveness of CD56bright NK cells is likely to mean a greater augmentation of their cytotoxic function compared to the CD56dim subset. Accordingly, CD56bright NK cells from daclizumab-treated MS patients, which are activated by the higher levels of available IL-2, had considerably higher cytotoxicity towards activated T cells than did CD56dim NK cells (Bielekova et al. 2006).

57

CD56bright NK cells are able to proliferate and have enhanced cytotoxic function in response to stimulation with very low doses of IL-2 that are insufficient to stimulate CD56dim NK cells (Nagler et al. 1989, Nagler et al. 1990, Baume et al. 1992). CD56bright NK cells constitutively express CD25, a component of the high affinity IL- 2 receptor which is only found on CD56dim NK cells following activation with target cells and IL-2 (Caligiuri et al. 1990, Nagler et al. 1990, Baume et al. 1992, Wendt et al. 2006). However, even when expression of CD25 is induced on CD56dim NK cells, their proliferation is still lower than that of CD56bright NK cells, suggesting that differences in signalling mechanisms may also be involved (Baume et al. 1992). CD56bright NK cells also exclusively express c-kit receptor, stimulation of which enhances IL-2-induced IFNγ production and proliferation (Matos et al. 1993, Carson et al. 1997). IL-7 receptor (IL7R) is also exclusively expressed on the CD56bright population, therefore only CD56bright NK cells proliferate and upregulate cytotoxic function in response to IL-7 (Hanna et al. 2004b, Wendt et al. 2006, Nielsen et al. 2012). In addition, CD56bright NK cells express considerably higher levels of IL12Rβ2 transcript, though this difference is abrogated following activation (Wendt et al. 2006). This may explain the enhanced IFNγ response and strong upregulation of cytotoxic function when this subset is stimulated with IL-12/IL-18. Although both subsets express similar levels of IL-21R, CD56bright NK cells also seem to be more responsive to this cytokine, which again is likely due to the differential expression of various signalling molecules (Hanna et al. 2004b, Wilk et al. 2008, Nielsen et al. 2012).

These subsets also have differing capacity for interaction with other leukocytes. Interaction with monocytes induces CD56bright NK cells to produce large amounts of IFNγ and in turn, NK cells promote TNFα production in the monocytes (Dalbeth et al. 2004). CD56dim NK cells do not promote TNFα production as strongly (Dalbeth et al. 2004). CD56bright NK cells are also more effective at inducing monocyte-derived DCs due to their higher level of GM-CSF secretion (Zhang et al. 2007). LPS-activated dendritic cells also preferentially induce proliferation and IFNγ production in the CD56bright subset (Vitale et al. 2004). In addition, CD56bright NK cells have higher levels of antioxidants such as cell surface thiols and are resistant to H2O2 produced by phagocytes, which induces apoptosis in CD56dim NK cells (Harlin et al. 2007, Thorén et al. 2007).

58

Differences in chemokine receptors and adhesion molecules are likely to influence the localisation of these two subsets. CD56bright NK cells constitutively express CD62L, CD44 and CCR7, which are associated with migration to lymph nodes (Frey et al. 1998, Campbell et al. 2001, Hanna et al. 2004b, Vitale et al. 2004, Wendt et al. 2006). CCR7 expression can be induced on activated CD56dim NK cells, but CD62L is only expressed on a small minority of CD56dim NK cells, both resting and activated (Wendt et al. 2006). CCR5, CCR6 and CXCR3 are also found predominantly on the CD56bright subset (Campbell et al. 2001, Vitale et al. 2004, Berahovich et al. 2006, Wendt et al. 2006). Conversely, CXCR1, CXCR2 and CX3CR1 are largely restricted to the CD56dim subset (Campbell et al. 2001). A number of other adhesion molecules appear to be differentially expressed, which may have implications for both migration and interaction with target cells: ICAM-1, ICAM-2, LFA-3, activated leukocyte cell adhesion molecule and integrin αE have all been associated with CD56dim NK cells, whereas ICAM3 and integrins α5, αM and αX are preferentially expressed by the CD56bright subset (Frey et al. 1998, Hanna et al. 2004b, Wendt et al. 2006).

A number of other markers have also been used to define distinct subsets of NK cells, including CD8, CD27, CD57, CD62L and CX3CR1 (Srour et al. 1990, Lowdell et al. 2002, Addison et al. 2005, Silva et al. 2008, Vossen et al. 2008, Chattopadhyay et al. 2009, Björkström et al. 2010, Juelke et al. 2010, Lopez-Vergès et al. 2010, Hamann et al. 2011).

CD8 is expressed on around 40% of NK cells, most of which express only α/α homodimers, as opposed to the α/β heterodimers found on most CD8+ T cells (Baume et al. 1990). The CD8+ NK cell subset is reported to be more cytotoxic towards K562 cells and primary leukaemic blasts than the CD8- subset (Srour et al. 1990, Lowdell et al. 2002, Addison et al. 2005). However, CD8+ and CD8- NK cells express similar levels of KIR, activating receptors (NKP30, NKp44, NKp46 and NKG2D) and intracellular perforin and granzyme A, the factors typically associated with differing cytotoxic function in other subsets (Addison et al. 2005). This led to speculation that the CD8 molecule itself might contribute directly to NK cell function. CD8α/α dimers bind to both classical and non-classical HLA class I and are predicted to modulate avidity of HLA class I/TCR interactions (Sanders et al. 1991, Gao et al. 1997, Gao et al. 2000). Conceivably, CD8 might also influence the interaction between HLA class I

59 and KIR or LILR, thereby modulating induction of positive or negative signalling by these receptors. However, target cells transfected with HLA-Cw*0304 were still protected from lysis when the NK cell effector population expressed CD8, indicating that CD8 does not modulate signals from inhibitory KIR (Addison et al. 2005). CD8 might also trigger signalling in its own right. In T cells, CD8 associates with lymphocyte-specific protein tyrosine kinase (Lck) and dephosphorylates it, causing Lck in turn to phosphorylate CD3ζ, triggering a signalling cascade resulting in activation of linker of activated T cells (LAT) and PLCγ, which modulate cytoskeletal reorganisation, calcium flux, proliferation and gene transcription (Gibbings and Dean Befus 2009). Meanwhile, CD8 ligation in myeloid cells enhances FcRγ-mediated responses (Gibbings and Dean Befus 2009). In support of a signalling role for CD8 on NK cells, its ligation causes a rapid and sustained influx of extracellular-derived calcium (Lowdell et al. 2002, Spaggiari et al. 2002, Addison et al. 2005). However, ligation of CD8 does not appear to induce lytic activity per se (Lowdell et al. 2002, Addison et al. 2005). According to Addison et al., CD8+ NK cells undergo a considerably lower rate of activation-induced apoptosis than CD8- NK cells, an effect that is abrogated by blocking MHC class I (Addison et al. 2005). CD8+ NK cells are protected from apoptosis even when killing MHC-deficient targets, due to simultaneous formation of both NK cell:target cell and NK cell:NK cell conjugates, whereby NK cells themselves provide the MHC class I ligand for CD8 (Addison et al. 2005). This study appears to contradict the previous findings of Spaggiari et al. (2002), who found that soluble HLA class I caused apoptosis of CD8+ NK cell clones by inducing expression of FasL, which then binds to the NK cell’s own Fas (CD95) receptors. However, this may reflect differential effects of soluble versus membrane- bound ligands, as has been observed for other receptor-ligand combinations, including NKG2D and its ligands (Groh et al. 2002, Song et al. 2006, Cao et al. 2007, Cerboni et al. 2009).

CD27 is a member of the TNF receptor family which distinguishes NK cell subsets in both humans and mice. Interaction of CD27 with its ligand CD70 promotes NK cell activation and cytokine production (Takeda et al. 2000, Kelley et al. 2001). CD27 is more frequently expressed on CD56bright NK cells and, like the CD56bright subset, CD27+ NK cells are enriched in tonsil and spleen (Silva et al. 2008, Vossen et al. 2008). CD27 is also found on a smaller percentage of CD56dim NK cells (Silva et al.

60

2008, Vossen et al. 2008). CD27+ NK cells express high levels of NKp46 and nearly all are CD94+ CD62L+, whereas the frequency of KIR-expressing cells is considerably lower than amongst CD27- NK cells (Silva et al. 2008, Vossen et al. 2008).Within CD56bright and CD56dim subsets, expression of CD27 is associated with a lower level of perforin and granzyme B expression and lower cytotoxicity towards K562 cells. However, NKG2D-mediated killing of P815-MICA targets is similar or even higher using CD27+ NK cells (Silva et al. 2008). In this case the slightly higher expression of NKG2D observed on CD27+ NK cells might compensate for the lower expression of perforin and granzyme B (Silva et al. 2008). CD27+ NK cells also make more IFNγ in response to stimulation with PMA/ionomycin or cytokines (IL-12 combined with IL- 15 or IL-18) (Silva et al. 2008, Vossen et al. 2008). CD56dim NK cells always produced less IFNγ than CD56bright, but within each of those subsets, there was a significant reduction in cytokine production when cells lacked CD27 (Vossen et al. 2008). In mice, CD27 expression is also reported to enhance activation by DCs (Hayakawa and Smyth 2006).

CD57 is a carbohydrate antigen expressed on some CD56dim, but not CD56bright NK cells (Mitsumoto et al. 2000, Chattopadhyay et al. 2009, Lopez-Vergès et al. 2010). The CD57+ NK cell population expands during viral infection and CD57 expresion can be induced on CD57- CD56dim NK cells in vitro by incubation with IL-2 (Lopez- Vergès et al. 2010, Lopez -Vergès et al. 2011). CD56dim CD57+ NK cells express lower levels of NKp30, NKp46, NKG2D, KLRG1, CCR5, CXCR3 and CXCR4 and higher levels of most KIRs than CD56dim CD57- NK cells and are typically LIR-1+, CD27- and CD62L- (Björkström et al. 2010, Lopez-Vergès et al. 2010). They express slightly higher levels of CD16 and accordingly produce more IFNγ in response to CD16 stimulation (Lopez-Vergès et al. 2010). However, CD56dim CD57- NK cells produce more IFNγ in response to IL-12+IL-18 stimulation, which may be related to their higher surface expression of IL-18Rα and higher levels of IL-12Rβ2 transcripts (Björkström et al. 2010, Lopez-Vergès et al. 2010). When stimulated with target cells or a combination of IL-2 and antibodies directed against NKG2D, NKp46 and 2B4, both subsets produce similar quantities of IFNγ (Lopez-Vergès et al. 2010). CD57+ NK cells have a reduced proliferative response to stimulation with target cells plus IL- 2 and to various combinations of cytokines, however, unlike CD57+ T cells, they are not more susceptible to activation-induced cell death than their CD57- counterparts

61

(Björkström et al. 2010, Lopez-Vergès et al. 2010). CD56dim CD57+ NK cells also express higher levels of granzyme B and perforin, however cytotoxicity towards K562 is not significantly different (Chattopadhyay et al. 2009, Lopez-Vergès et al. 2010). Antibody-coated target cells are killed more efficiently by CD57+ NK cells, but this is probably related to their higher expression of CD16 (Lopez-Vergès et al. 2010).

A number of functional differences between CX3CR1+ and CX3CR1- NK cells have recently been identified. CX3CR1 expression is upregulated by stimulation with IL-2, IL-12, IL-15 or TGFβ (Hamann et al. 2011). CX3CR1+ NK cells are more cytotoxic towards K562 target cells, whereas CX3CR1- NK cells produce 100-fold more IL-13, 30-fold more IL-5 and significantly more IL-10, GM-CSF and TNF-α in response to PMA/ionomycin stimulation (Infante-Duarte et al. 2005, Hamann et al. 2011). However these differences may in part be due to the fact that CX3CR1+ NK cells are all CD56dim, whereas the CX3CR1 popualtion contains both CD56bright and CD56dim NK cells (Hamann et al. 2011). CD56dim CX3CR1low/- NK cells appear to represent an intermediate population between CD56bright and CD56dim CX3CR1+, as they have intermediate expression of KIR (expressed highly on CX3CR1+ NK cells) and NKG2A, NKp30, NKp46, CD27 and CD62L, all of which are expressed preferentially on CD56bright NK cells (Hamann et al. 2011). CX3CR1high NK cells show virtually no proliferative response to IL-2, whereas CD56dim CX3CR1low/- cells exhibit proliferation only slightly lower than that of CD56bright NK cells. The percentage of CD56dim NK cells expressing CD57 however is the same in both CX3CR1+ and CX3CR1- subsets (Hamann et al. 2011).

All the above subsets are found within peripheral blood. However, other subsets have been described that are associated with specific tissues. For instance an NKp44+ subset, termed “NK-22,” has been identified in tonsils that expresses CCR6, receptor tyrosine kinase-like orphan receptor (ROR) α, RORC, IRF4 and aryl hydrocarbon receptor and produces IL-22, IL-26 and leaukaemia inhibitory factor (LIF) in response to IL-23 stimulation or interaction with activated monocytes (Cella et al. 2008). These cytokines stimulate epithelial cells to proliferate and release IL-10 (Cella et al. 2008). IL-17 production was not observed in NK-22 cells, however another RORγ+ subset has also been reported in tonsils which produces both IL-17 and IL-22 as well as TNF and lymphotoxin-α and -β when activated (Cella et al. 2008, Cupedo et al. 2009).

62

These cells appear to be derived from lymphoid tissue inducer cells and express IL7R, NKp44, NKp46 and NKp30, but lack NKG2D (Cupedo et al. 2009). It is possible that these two subsets are in fact the same population and that IL-17 was not detected in the former study due to the shorter stimulation time. Both subsets lack perforin, have low granzyme expression and produce little IFNγ in response to PMA/ionomycin stimulation (Cella et al. 2008, Cupedo et al. 2009).

During pregnancy, another distinctive subset of NK cells appears in the decidua, constituting around 70% of all decidual lymphocytes (Moffett-King 2002). Decidual NK cells are CD56superbright CD16- NKG2Ahigh CD9+, have low cytotoxicity and express many genes that are not transcribed in peripheral NK cells (King et al. 2000, Koopman et al. 2003, Kopcow et al. 2005, Keskin et al. 2007). Peripheral blood NK cells take on a similar phenotype when cultured with TGFβ, a cytokine produced by decidual stromal cells (Keskin et al. 2007).

1.1.7.1 Developmental Relationship of CD56dim and CD56bright NK Cells

Observations of an age-related decrease in the proportion of peripheral blood CD56bright NK cells led to the proposition that CD56dim NK cells might be derived from CD56bright, but for a long time there was no solid evidence of this (Borrego et al. 1999, Chidrawar et al. 2006). In fact, some speculated that the opposite developmental relationship might exist, given that CD56dim CD16+ NK cells have been shown to convert to CD56bright CD16- NK cells and acquire their functional characteristics in vitro when stimulated with certain combinations of cytokines (Loza and Perussia 2004, Mailliard et al. 2005, Takahashi et al. 2007). Conversion of CD56bright to CD56dim CD16+ NK cells was finally observed by Chan et al. in 2007, who cultured CD56bright NK cells from peripheral blood with synovial fibroblasts. The conversion was found to be dependent on NK cell-fibroblast contact mediated by interaction of NK cell CD56 with fibroblast growth factor receptor 1 on fibroblasts. The resulting NK cells had surface markers typical of the peripheral blood CD56dim subset, were highly cytotoxic and displayed a poor IFNγ response to PMA/ionomycin. Perhaps more importantly, the study went on to demonstrate that the natural direction of development in vivo is bright to dim, since CD56dim NK cells injected into NOD- SCID mice gave rise only to CD56dim NK cells, whereas CD56bright NK cells gave rise to both bright and dim subsets. CD56bright NK cells are also enriched shortly after

63

HSC transplant in humans (Cooley et al. 2007). In addition, CD56dim NK cells have a reduced telomere length compared to the CD56bright subset, indicating that they are at a later stage of differentiation (Chan et al. 2007, Romagnani et al. 2007, Björkström et al. 2010). Overall the balance of evidence now suggests that the CD56bright subset is indeed a developmental precursor to the CD56dim.

A number of studies have reported functional intermediates between CD56bright CD16- and CD56dim CD16+ NK cells. Changes in the repertoire of surface receptors have been linked to progressive functional changes, during what is apparently an ongoing developmental process occurring in “mature” NK cells (see figure 1-v). For instance, CD56bright CD16+ cells, induced by cytokine stimulation or incubation with T cells, are likely to represent NK cells about to transition to CD56dim (Romagnani et al. 2007, Béziat et al. 2011). Found in efferent lymph nodes, they exhibit both the cytokine responsesiveness of CD56bright CD16- NK cells and cytotoxicity resembling CD56dim NK cells (Romagnani et al. 2007, Takahashi et al. 2007, Béziat et al. 2011). Subsets with intermediate phenotypes have also been identified within the CD56dim population, characterised by expression of markers such as CD62L, CD94 and NKG2A, which are also highly expressed on the CD56bright population. CD62L is expressed only on a small proportion of CD56dim NK cells, which have a similar polyfunctional phenotype to that of CD56bright CD16+ NK cells (Juelke et al. 2010). CD56dim CD62L+ NK cells have high proliferative and cytotoxic capcity and produce large quantities of IFNγ in response to cytokines, interaction with DCs and ligation of activating receptors (Juelke et al. 2010). CD94 and NKG2A are found on around 50% of CD56dim NK cells and are typically expressed together, consistent with their functional interaction (Béziat et al. 2010, Yu et al. 2010). As CD56dim NK cells mature, they gain KIR and CD57 and lose NKG2A (Béziat et al. 2010, Björkstrom et al. 2010, Yu et al. 2010). Each of these steps is associated with a decrease in proliferative capacity and IFNγ production in response to cytokine stimulation (Béziat et al. 2010, Björkstrom et al. 2010, Yu et al. 2010). Cytotoxicity towards most susceptible targets, including K562, is already fully achieved in NKG2A+ CD56dim NK cells, though only NKG2A- KIR+ NK cells kill HLA-E positive targets (Béziat et al. 2010, Yu et al. 2010). Most NK cells gain KIR before losing NKG2A (Béziat et al. 2010). However, some NK cells lose NKG2A before gaining KIR and are

64 hyporesponsive as a result of NK cell education (Cooley et al. 2007, Béziat et al. 2010).

Figure 1-v: CD56bright CD16- NKG2A+ NK cells differentiate into CD56dim CD16+ NKG2A- NK cells. “Mature” NK cells are thought to pass through a series of developmental stages before reaching a terminally differentiated CD56dim NKG2A- KIR+ state. CD56bright NK cells gain CD16 then downregulate expression of CD56 and lose CD62L. This is followed by gain of KIRs and finally loss of NKG2A. Molecules acquired at a particular stage of differentiation are shown in green and those that are lost or downregulated are indicated in red.

65

1.2 The Role of T cells and NK Cells in Multiple Sclerosis

Multiple sclerosis (MS) is a presumed autoimmune disease of the central nervous system (CNS), involving myelin loss, axonal damage and brain atrophy (Hauser and Oksenberg 2006). Symptoms include limb weakness or loss of dexterity, visual disturbance, fatigue and cognitive deficits such as memory loss and impaired problem-solving ability. Initially, most patients experience relapsing-remitting disease (RRMS), i.e. episodes of acute worsening in symptoms followed by periods of partial or complete recovery (Hauser and Oksenberg 2006, Wakerley et al. 2008). However, approximately 10% of patients experience progressive degeneration without first going through a relapsing-remitting stage (Wakerley et al. 2008). This is known as primary progressive MS (PPMS).

Relapses are associated with breakdown of the blood-brain barrier (BBB) and formation of lesions in both white and grey matter, where infiltrating leukocytes accumulate causing inflammation and demyelination (Brück 2005, Hauser and Oksenberg 2006, Popescu and Lucchinetti 2012). The myelin sheath and oligodendrocytes which form it are subject to attack; this is presumed to involve infiltrating T cells and macrophages, although in some post-mortem neuropathological studies, T cells may not be abundant (Jurewicz et al. 1998, Barnett and Prineus 2004, Hill et al. 2004, Jurewicz et al. 2005, Zeis and Schaeren-Wiemers 2008, Goverman 2009). Oligodendrocytes have many processes, each of which wraps multiple times around a different axon, encasing it in a lipid-rich myelin sheath which provides electrical insulation (Baumann and Pham Dinh 2001). Demyelination reduces neuronal conductivity and leaves the axons themselves vulnerable to immune attack (Bitsch et al. 2000, Neumann et al. 2003, Hauser and Oksenberg 2006, Melzer et al. 2009).

Symptoms during relapse are caused mainly by inflammation and demyelination (Brück 2005). Once inflammation has subsided, axons can be remyelinated, explaining why function can be recovered during remission (Lassman et al. 2001, Brück 2005). However, axons which are not remyelinated are vulnerable to further degeneration even once inflammation has subsided (Bitsch et al. 2000, Kornek et al. 2000). Damage to neurons is irreversible and once substantial damage has accumulated, the CNS becomes unable to compensate and patients enter the

66 secondary progressive (SP) phase of disease, during which they become progressively more disabled (Neumann et al. 2003, Brück et al. 2005). The rate of disability progression correlates with cerebellar atrophy and the degree of axonal damage, both in lesions and in normal appearing white matter (Lossef et al. 1996, De Stefano et al. 1998, Fu et al. 1998).

1.2.1 Genetic and Environmental Risk Factors are Associated with Development of MS

Around 85,000 (approximately 1 in 900) people in the UK have MS, two thirds of whom are female (Multiple Sclerosis International Federation, Atlas of MS database [online], http://www.atlasofms.org/index.aspx, accessed August 25th 2012). Prevalence is similar in other European countries, North America, Australia and New Zealand, with considerably lower rates in Asia and South America. Mean age of onset is between 24 and 33 years.

This suggests that differences in susceptibility are partly caused by genetic factors. However, susceptibility to MS cannot be determined by the presence or absence of any one individual gene, allele or polymorphism. The strongest genetic association with MS occurs within the HLA gene complex on chromosome 6p21. In particular, one common Northern European haplotype, HLA-DQB1*0602, -DQA1*0102, - DRB1*1501, -DRB5*0101, has consistently been associated with susceptibility to MS (Spurkland et al. 1991, Allen et al. 1994, Amirzargar et al. 1998, Barcellos et al. 2003, Fernández et al. 2004). The risk is thought to be associated primarily with the HLA-DQB1*0602 and DRB1*1501 alleles and both of these may need to be present in order to confer susceptibility (Caballero et al. 1999, Oksenberg et al. 2004, Caillier et al. 2008, Lincoln et al. 2009). Possession of two copies of this haplotype confers an even greater susceptibility and increases the likelihood of more aggressive forms of disease (Barcellos et al. 2003). Other HLA haplotypes and specific alleles have also been proposed to positively or negatively influence risk of MS (Allen et al. 1994, Ligers et al. 2001, Cocco et al. 2012). However, due to a high level of linkage disequilibrium between HLA alleles, it remains difficult to determine just how many of these alleles are genuine functional determinants of disease (Etzensperger et al. 2008).

67

Genome-wide association studies (GWAS) have identified a number of other non- HLA alleles that are overrepresented in MS patients. Most of these have relatively modest effects on susceptibility and possession of several may be required in order to predispose to disease (Baranzini 2011). One of the most established associations outside of the HLA region is with certain alleles of IL2RA that favour production of the soluble rather than the membrane-bound form of the receptor, which promotes proliferation and activation of CD4+ effector T cells (International Multiple Sclerosis Genetics Consortium et al. 2007, ANZgene 2009, De Jager et al. 2009a, Maier et al. 2009a, Maier et al. 2009b). An allele of IL7RA associated with an increase in the proportion of soluble IL7R is also a risk factor for MS (Gregory et al. 2007, International Multiple Sclerosis Genetics Consortium et al. 2007, Lundmark et al. 2007, De Jager et al. 2009a). Expression of both IL7 and IL7R is increased in cerebrospinal fluid (CSF) of MS patients compared to other neurological disease and responses to myelin basic protein (MBP) and myelin oligodendrocyte protein (MOG) are enhanced by IL-7 in T cells from MS patients but not healthy controls (Traggiai et al. 2001, Lundmark et al. 2007).

Another genetic risk factor for MS is a glycine to serine single nucleotide polymorphism in the DNAM-1 (CD226) gene, a mutation which has also been associated with several other autoimmune diseases (Hafler et al. 2009, Wieczorek et al. 2009, Patsopoulos et al. 2011). DNAM-1 is expressed on most T cells, NK cells, monocytes and some B cells (Hafler et al. 2009). It binds to poliovirus receptor or nectin 2 and co-stimulates T and NK cell responses and also functions as an adhesion molecule promoting transendothelial migration of leukocytes (Shibuya et al.1996, Bottino et al. 2003, Reymond et al. 2004, Tahara-Hanaoka et al. 2004). Blocking of DNAM-1 delays EAE onset and reduces its severity (Dardalhon et al. 2005).

Other loci identified by GWAS include the costimulation molecules CD58 (LFA-3), CD40 and CD86, IL12B (a component of both IL-12 and IL-23 cytokines), and the transcription factors T-bet and STAT3 (International Multiple Sclerosis Genetics Consortium et al. 2007, ANZgene 2009, De Jager et al. 2009b, Patsopoulos et al. 2011). A general increase in MS prevalence with distance from the equator has led to the suggestion that environmental factors such as sunlight and pathogens may play a role. A role for sunlight is further supported by decreased MS susceptibility in those

68 with outdoor occupations (Freedman et al. 2000, Westberg et al. 2009). Migration studies suggest that exposure to sunlight in childhood is particularly important in protecting against MS and the association of birth month with likelihood of MS has led to speculation that vitamin D levels in utero may also have an effect (Alter et al. 1966, Dean and Elian 1997, Hammond et al. 2000, Willer et al. 2005). Sunlight is important for the synthesis of vitamin D. UVB radiation causes photolysis of 7- dehydrocholesterol to pre-vitamin D3, which spontaneously isomerises to vitamin D3 in the skin (Kamen and Tangpricha 2010). This is then hydroxylated twice to produce the active form of vitamin D, 1,25-dihydroxyvitamin D3 (calcitriol). Calcitriol mediates most of its effects via binding to the vitamin D receptor (VDR), a ligand- activated transcription factor that binds to vitamin D response elements (VDRE) in gene promoters (Ramagopalan et al. 2010). VDREs are found at many putative MS- related loci, including IRF8, STAT3, VCAM-1, CXCR4, CXCR5, NFκB1, CD40, CLEC16A and members of the TNF/TNFR superfamilies (Ramagopalan et al. 2010, DiSanto et al. 2012).

Higher vitamin D intake and serum levels of calcitriol have been associated with reduced risk of MS and with reduced relapse rate amongst RRMS patients (Munger et al. 2004, Munger et al. 2006, Pierrot-Deseilligny et al. 2008, Soilu-Hänninen et al. 2008 , Correale et al. 2009, Ascherio et al. 2010, Simpson et al. 2010). The serum level of calcitriol correlates with the ability of Tregs to suppress T effector proliferation in vitro, and higher calcitriol levels also favour skewing towards Th2 instead of Th1 (Smolders et al. 2009). VDR binds to and activates Forkhead box P3 (FoxP3) in T regs stimulated with calcitriol, enhancing their ability to suppress T cell proliferation (Kang et al. 2012). It also downregulates IL-12 and IL-23 expression in DCs, causing them to induce Th2 or Treg polarisation instead of Th1 or Th17 (Daniel et al. 2008).Vitamin D is essential for development of functional invariant (i)NKT cells and has a variety of other immunosuppressive effects, including upregulating expression of IL-10, TGFβ and IL-5, inhibiting DC maturation, downregulating the CNS-homing receptor CCR6 on T cells, inhibiting NK and T cell proliferation, promoting T cell apoptosis and inhibiting NK cell cytotoxic function (Leung 1989, Merino et al. 1989, Penna and Adorini 2000, Griffin et al. 2001, Spach et al. 2004, Spach et al. 2006, Pedersen et al. 2007, Yu and Cantorna 2008, Correale et al. 2009, Chang et al. 2010, Lysandropoulos et al. 2011).

69

Smoking is also a major risk factor for MS, particulary PPMS, though the mechanism is not known (Hernán et al. 2001, Riise et al. 2003, Hedström et al. 2009, Healy et al. 2009). It also increases the likelihood of progression to SPMS amongst RRMS patients (Hernán et al. 2005, Healy et al. 2009, Pittas et al. 2009).

In summary, the epidemiology of MS is complex, with both genetic and environmental factors likely to play a role.

1.2.2 CD4+ T cells Infiltrate the CNS and Promote Recruitment and Activation of Other Leukocytes, leading to Myelin and Axonal Damage

Due to the inaccessibility of brain tissue in MS, much of our understanding of disease mechanisms is based on animal models such as experimental autoimmune encephalomyelitis (EAE). EAE is induced by administration of immunogenic white matter, protein or myelin antigen peptides from myelin components such as myelin oligodendrocyte protein (MOG), myelin basic protein (MBP) or proteolipid protein (PLP) or by adoptive transfer of T cell clones specific for these epitopes combined with an adjuvant (Gold et al. 2006). This model showed in the 1980s that activated autoreactive CD4+ T cells are necessary and sufficient to initiate EAE (Ben-Nun et al. 1981). By extrapolation, it had been widely believed that CD4+ T cells are responsible for initiating human MS, although this has more recently been debated, as disease may occur in the presence of predominant CD8 infiltrates or in the virtual absence of lymphocytes (Babbe et al. 2000, Huseby et al. 2001, Barnett and Prineas 2004, Gold et al. 2006, Goverman 2009) (see figure 1-vi).

The type of CD4+ T cell that may induce MS is a matter of debate. MS was originally considered a stereotypical Th1-mediated autoimmune disease, based on preferential CNS infiltration of IFNγ- but not IL-4- expressing T cells following peptide immunisation and on the lack of EAE induction in IRF-1-deficient mice (Merrill et al. 1992, Renno et al. 1994, Tada et al. 1997, Juedes et al. 2000). However, since the discovery of Th17 cells, several studies have shown that these are also capable of initiating MS-like disease in mice and implied that they may be more crucial than Th1 (Cua et al. 2003, Langrish et al. 2005, Kebir et al. 2007, Reboldi et al. 2009). The relative importance of the two subsets is still a matter of debate, though it seems likely

70 that both might play a role either in initiation or perpetuation of CNS inflammation (O’Connor et al. 2008). Th1 and Th17 cells have been shown to initiate distinct types of EAE and, given the heterogenous nature of MS, it seems likely that different subsets may be responsible for initiating human MS in different individuals (Kroenke et al. 2008, Stromnes et al. 2008, Jäger et al. 2009).

However, T cell plasticity should also be taken into account. When Th17 polarisation is induced by IL-23 stimulation of human peripheral blood T cells in vitro, a subset of these cells also produces IFNγ and expresses both the Th1 transcription factor T-bet and the Th17 transcription factor RORγt (Kebir et al. 2009). The frequency of these IFNγ, IL-17 co-expressing cells is higher amongst T cells derived from MS patients, particularly those who are in relapse and large numbers of T cells co-expressing these cytokines have been detected in MS brain tissue. IFNγ induces strong upregulation of ICAM-1 on BBB endothelial cells, which facilitates adhesion and transmigration of both Th1 and Th17 cells (Kebir et al. 2009). The Th17 cytokines IL-17 and IL-22 also promote migration of human T cells through the BBB, by decreasing expression of the tight junction proteins occludin and zonula occludens-1 and promoting expression of chemokines including CCL2 and IL-8 (Kebir et al. 2007). Consequently, T cells co-expressing IFNγ and IL-17 migrate into the CNS more efficiently compared to those expressing either IFNγ or IL-17 alone and play an important role in the widespread recruitment of other leukocytes which follows (Kebir et al. 2009).

In order to cross the BBB, T cells must first be activated in the lymph nodes by dendritic cells presenting their cognate antigen (Goverman 2009, Inoue et al. 2012). Activated T cells upregulate chemokine receptors and integrins, allowing them to migrate towards the CNS and adhere to vascular endothelial cells of the BBB (Hauser and Oksenberg 2006, Bauer et al. 2009, Holmann et al. 2011, Inoue et al. 2012). Activated T cells roll along blood vessel walls until they encounter chemokines on the surface of the endothelial cell, which trigger signalling via the cognate chemokine receptors, leading to enhanced binding affinity and clustering of integrins (Carrithers et al. 2000, Kerfoot and Kubes 2002, Engelhart 2008, Holman et al. 2011). This facilitates stronger binding to the endothelial cell, causing the leukocyte to migrate across the vessel wall. Leukocytes are thought to enter the CNS at post-capillary venules in the meninges and via the choroid plexus epithelium, the barrier between

71 blood and CSF, both of which constitutively express P-selectins, one of the molecules that facilitates leukocyte rolling (Carrithers et al. 2000, Kivisäkk et al. 2003, Döring et al. 2007). The very late antigen-4 (VLA-4) integrin in particular is important for migration into the CNS parenchyma, since blocking antibodies effectively attenuate leukocyte infiltration in both EAE and MS (Yednock et al. 1992, Baron et al. 1993, Brocke et al. 1999, Theien et al. 2001, Kerfoot and Kubes 2002, Kivisäkk et al. 2009a).

During EAE, CD4+ T cells accumulate initially in the subarachnoid space, where they are reactivated by resident APCs, proliferate and release pro-inflammatory cytokines (Govermann 2009, Kivisäkk et al. 2009b). This causes upregulation of adhesion molecules on parenchymal vasculature, enabling them to migrate into the parenchyma. Infiltrating CD4+ T cells create an inflammatory environment within the brain, which promotes infiltration and activation of CD8+ T cells and macrophages (Kebir et al. 2007, Goverman 2009). Activated Th17 cells produce the neutrophil attractant CXCL8 and GM-CSF which induces maturation of microglia, CNS-specific APCs with similarity to macrophages (Aloisi et al. 2000, Codarri et al. 2011). Meanwhile, IFNγ activates macrophages and induces expression of MHC class II and co-stimulatory molecules on parenchymal micoglia allowing them to become competent antigen-presenting cells (Menèndez Iglesias et al. 1997, Goverman 2009). Th1 cells specifically enhance the ability of microglia and macrophages to induce Th1 rather than Th2 responses and stimulate them to produce TNFα (Renno et al. 1995, Aloisi et al. 2000, Juedes et al. 2000). IFNγ also induces IFNγ-inducible protein 10 (IP-10) and monokine induced by IFN-gamma (Mig) production by macrophages, microglia and astrocytes (Balashakov et al. 1999, Simpson et al. 2000). Together with the aforementioned disruption of the BBB by both Th1 and Th17 cytokines, accumulation of chemokines in lesions and in the CSF results in recruitment of much larger numbers of leukocytes to the CNS, including macrophages, neutrophils, memory and naïve B cells and CD4+ and CD8+ T cells (Glabinski et al. 1997, Balashakov et al. 1999, Sørensen et al. 1999, Simpson et al. 2000, Kebir et al. 2007, Carlson et al. 2008).

72

Figure 1-vi: CD4+ T cells initiate CNS inflammation and demyelination. T cells are activated in the periphery and upregulate integrins required to adhere to and cross the BBB or blood-CSF barrier. Once inside the CNS, myelin-specific CD4+ T cells are reactivated by APCs in the perivascular space and produce cytokines and chemokines which increase BBB permeability and promote migration and activation of other leukocytes such as macrophages, CD8+ T cells and B cells. Leukocytes migrate into the parenchyma and cause damage to myelin-forming oligodendrocytes and axons by releasing graznzymes, ROS, TNFα and MMPs.

73

Infiltrating leukocytes produce a number of tissue-degrading enzymes including ADAM-12 and MMPs (Bar-Or et al. 2003, Vos et al. 2003, Toft-Hansen et al. 2004). These degrade components of the subendothelial basal lamina, increasing BBB permeability and facilitating migration into the brain parenchyma (Rosenberg et al. 1992, Leppert et al. 1995, Stüve et al. 1996, Anthony et al. 1998, Lou et al. 1999, Zhang et al. 2011). MMPs, especially MMP9, also contribute to neuronal cell death and axonal damage (Vos et al. 2000, Newman et al. 2001). Several MMPs have been shown to cleave MBP, including macrophage-derived MMP12 (Gijbels et al. 1993, Chandler et al. 1995, Chandler et al. 1996, Vos et al. 2003, Toft-Hansen et al. 2004). MMP9 also has a protective role, cleaving the neuron-glial antigen 2 proteoglycan which inhibits oligodendrocyte maturation and therefore remyelination (Larsen et al. 2003).

Activated macrophages and microglia induce apoptotic pathways in oligodendrocytes and neurons by production of TNFα and FasL, to which these cell types are particularly sensitive; (Selmaj and Raine 1988, D’Souza et al. 1996, Becher et al. 1998, Li et al. 2002, Jurewicz et al. 2005). They also produce high levels of toxic ROS and RNS which mediate various forms of damage such as lipid peroxidation, tyrosine nitrosylation, and DNA strand breaks (Merrill et al. 1993, Zhang et al. 1994, De Groot et al. 1997, van der Veen and Roberts 1999, Willenborg et al. 1999, Hill et al. 2004, Li et al. 2005, Jack et al. 2007). MBP-specific T cells activate microglia to upregulate MHC class II and ICAM-1, secrete TNFα and phagocytose axons and antibody-coated oligodendrocytes (Gimsa et al. 2000, Zeis and Schaeren-Wiemers 2008). Macrophages and microglia present myelin products to T cells, further perpetuating the inflammatory response. Microglia secrete IL-12 and preferentially induce Th1 rather than Th2 cytokines (Krakowski and Owens 1997, Stalder et al. 1997). They also release IL-1β, IL-3, IL-6, TNF-α, VEGF, lymphotoxin, macrophage inhibitory protein (MIP)-1α and matrix metalloproteinases (MMPs) (Stoll et al. 2002). Microglia also selectively phagocytose apoptotic leukocytes, a function which is strongly enhanced in the presence of IFNγ (Chan et al. 2001). Phagocytosis of apoptotic cells by microglia causes them to downregulate expression of TNFα and IL- 12 and reduces their ability to activate MBP-specific T cells (Magnus et al. 2001).

74

IL-15 produced by B cells, monocytes, astrocytes and microglia upregulates NKG2D and granzymes in CD8+ T cells, enhancing their cytotoxicity towards glial cells (Saikali et al. 2010, Schneider et al. 2011). Cytotoxic CD8+ T cells attack neurons and the supporting oligodendrocytes via granzyme release, ligation of the death receptor Fas (CD95) and production of soluble factors such as TNF-α (Goverman 2009, Melzer et al. 2009, Meuth et al. 2009). Oligodendrocytes are lysed by antigen-specific CD8+ T cells via an MHC class I-dependent mechanism (Jurewicz et al. 1998). Activated CD4+ and CD8+ T cells also gather along axons and kill neurons (Giuliani et al. 2003, Nitsch et al. 2004). T cell degranulation induces calcium oscillations in the neuron, followed by a sustained increase in calcium concentration, triggering nitric oxide production leading to cell death (Malipiero et al. 1999, Nitsch et al. 2004). This process is dependent on activation of the glutamate recpeptor N-methyl- D-aspartate (Malipiero et al. 1999, Nitsch et al. 2004).

Once myelin, oligodendroglial and axon damage has begun, antigen presentation is enhanced, due to availability of soluble antigens and release of antigen presenting cells such as astroglial cells from suppression by neurons (Melzer et al. 2009). Activated macrophages, microglia and CD8+ T cells suppress neuronal excitability, which together with cytokine release, causes an upregulation of MHC class I expression, making them vulnerable to further attack by CD8+ T cells (Neumann et al. 2003, Wang et al. 2008, Melzer et al. 2009, Meuth et al. 2009).

B cells may also have a role in pathogenesis. Anti-MOG antibodies are increased in the CNS of MS patients and have been shown to cause myelin destruction in EAE (Genain et al. 1999, Berger et al. 2003, Hafler et al. 2005, Lalive et al. 2006, Menge et al. 2011). In addition, high concentrations of IgM in the CSF are associated with more rapid disease progression (Perini et al. 2006). Antibody deposited on oligodendrocytes is recognised by Fc receptors on macrophages, leading to phagocytosis of the oligodendrocyte (Scolding and Compston 1991). Antibodies can also direct formation of a membrane attack complex and subsequent complement- mediated lysis (Mead et al. 2002). It appears that B cells are able to mature within the CNS in response to specific antigen (Corcione et al. 2004). In some patients, memory B cells cluster into follicular structures in the meninges of the brain (Magliozzi et al. 2007). These follicles are typically found adjacent to large lesions and are associated

75 with more extensive cortical pathology and rapid disease progression. There has been considerable controversy around whether or not the B cells within these ectopic, sub- meningeal follicles transcribe reactivated Epstein-Barr virus (EBV) (Serafini et al. 2007, Willis et al. 2009, Serafini et al. 2010). This is of considerable interest, both with respect to the role of the anti-viral response in neuropathology and with respect to the consideration of infectious agents in the aetiology of the disease.

IFNγ, TNFα and IL-1β produced by microglia and infiltrating leukocytes activate astrocytes, which can have both protective and detrimental consequences (Liberto et al. 2004). Astrocytes are the most abundant cell type in the CNS, representing around 90% of cells in the human brain (Nair et al. 2008). They store glycogen, regulate local pH and ion homeostasis and modulate neurotransmission by release of neuromodulators into the synaptic cleft and uptake of neurotransmitters such as glutamate (Martin 1992, Kang et al. 1998, Brown and Ransom 2007, Nair et al. 2008). Despite being of neural, rather than haematopoetic origin, astrocytes also have some properties of immune cells (Nair et al. 2008). Upon stimulation with IFNγ, astrocytes upregulate MHC class II, CD40, CD80 and CD86 and acquire antigen presenting function (Nikcevich et al. 1997, Aloisi et al. 1998). Astrocytes are less efficient antigen presenting cells than microglia, however antigen presentation by astrocytes preferentially stimulates differentiation into Th2 cells, which may have a protective function in MS (Aloisi et al. 1998). Astrocytes secrete both pro- and anti- inflammatory cytokines, including IL-1β, IL-6, IL-10, IL-12, IL-17, IL-23, TNFα and TGFβ (Nair et al. 2008). Their endfeet surround blood vessels, allowing them to modulate BBB function. IL-1β, IL-6 and TNFα increase BBB permeability, while TGFβ and VEGF restore its integrity (Liberto et al. 2004, Nair et al. 2008). Astrocytes also secrete the chemokines MCP-1, RANTES, IL-8, IP-10, stromal derived factor 1 (SDF-1) and fractalkine (Glabinski et al. 1997). Astrocytes activated with IFNγ and TNFα upregulate ICAM-1 and VCAM-1, which bind LFA-1 and VLA-4 respectively on activated leukocytes, facilitating their extravasation and migration through the parenchyma (Sobel et al. 1990, Tan et al. 1998, Gimenez et al. 2004). However, astrocytes also inhibit microglial IL-12 secretion and upregulate cytotoxic T-lymphocyte-antigen 4 on activated CD4+ T cells, causing them to take on a more regulatory function (Aloisi et al. 1997, Gimsa et al. 2004, Trajkovic et al. 2004).

76

Severe damage to the CNS can cause “reactive astrogliosis” and glial scar formation (Smith and Eng 1987, Liberto et al. 2004). Astrocytes upregulate production of the intermediate filaments glial fibrillary acidic protein, vimentin and nestin, proliferate and extend processes, binding the cells tightly together to form a dense glial scar (Liberto et al. 2004). Glial scars block migration of oligodendrocytes and axons to repair lesions (Nair et al. 2008). Astrocyte-derived ephrin-B2 creates a basal lamina around the damaged area, thereby physically blocking axon growth (Bundesen et al. 2003). Astrocytes also express neurocan and brevican proteoglycans which inhibit axon growth and fibroblast growth factor 2, which promotes oligodendrocyte proliferation and survival but inhibits maturation (Yamada et al. 1997, Asher et al. 2000, Nair et al. 2008). Reactive astrocytes also produce ROS, NO and TNFα which may contribute to neuron and oligodendrocyte damage (Liberto et al. 2004).

However, astrocytes also perform important protective functions following neuronal damage, as evidenced by prolonged inflammation, continued influx of leukocytes and increased neuronal damage following neuronal injury when astrocytes are depleted (Bush et al. 1999). Activated astrocytes at the edge of lesions express tissue inhibitor of matrix metalloproteinases (TIMP)-1, which might prevent expansion of the lesion by the destructive action of MMPs (Pagenstecher et al. 1998). Following injury, increased sequestration of glutamate by astrocytes and conversion to glutamine protects neurons and oligodendrocytes from damage (Liberto et al. 2004). Astrocytes also produce glutathione-S transferase-µ and ceruloplasmin which protect against oxidative damage, and a number of factors promoting neuron and oligodendrocyte survival and neurogenesis, such as brain-derived neurotrophic factor (BDNF), nerve- derived growth factor, neurotrophin 3, platelet-derived growth factor (PDGF) and LIF, which are upregulated following damage (Levison et al. 1996, Kuhlow et al. 2003, Schwartz and Nishiyama 2004, Nair et al. 2008). Secretion of MCP-1 and SDF- 1 recruits neural progenitor cells to the site of injury and promotes maturation of neural and oligodendrocyte progenitors (Calderon et al. 2006, Maysami et al. 2006, Xu et al. 2007). CXCL1-expressing astrocytes interact with oligodendrocytes, activating CXCR2 signalling which is important for survival and differentiation of oligodendrocytes and promotes remyelination (Padovani-Claudio et al. 2006). Activated astrocytes also produce IL-6 and IL-11, which are thought to promote oligodendrocyte survival, maturation and myelin synthesis (Schonrock et al. 2000,

77

Valerio et al. 2002, Zhang et al. 2006). Therefore astrocytes may be an important determining factor in perpetuation or repair of damage to the CNS.

1.2.3 Disease Initiation Requires Activation of Peripheral T Cells – Potential Role of Viruses and Impaired Immunoregulatory Mechanisms

Autoreactive CD4+ and CD8+ cells are present at similar frequencies in peripheral blood of MS patients and healthy controls (Crawford et al. 2004). However, the activity of these cells is believed to be controlled in healthy individuals by Tregs. Furthermore, T cells specific for myelin epitopes would not normally be expected to encounter their cognate antigen in the lymph nodes or blood. Since only activated T cells cross the intact BBB, cells restricted to CNS-specific epitopes such as MOG should not be able to enter under normal conditions (Steinmann 2007). Given that the current paradigm involves an important role for antigen-specific cells in disease initiation, this raises the question of how they come to enter the CNS in the first place. Either they are somehow being activated in the periphery, allowing them to cross the intact BBB, or the BBB is first compromised by other factors. T cells do not need to be specific for CNS antigens in order to enter the CNS and it is thought that small numbers of lymphocytes perform an immunosurveillance function in the ventricular and subarachnoid CSF (Hickey et al. 1991, Holman et al. 2011). However, the CNS is a highly immunosuppressive environment and infiltrating leukocytes will only initiate an immune response if they are restimulated with their cognate antigen within the CNS (Govermann 2009, Holman et al. 2011). One possibility is that infection or physical damage to the brain tissue induces an inflammatory response which if not adequately controlled, might compromise the integrity of the BBB and permit infiltration of other cells, including those specific for CNS antigens. Damage to the CNS might also result in escape of myelin antigens into the periphery, where they could be presented to autoimmune T cells (Steinmann 2007). Another possibility is non-specific polyclonal activation of B and T cells by bacterial or viral superantigens (Suspedra and Martin 2005).

Alternatively, myelin-reactive T cells might be activated by molecular mimicry. Many viruses contain peptides with close similarity to CNS antigens. In particular the peptide between residues 82 and 99 of myelin basic protein, a major target of T cell responses during MS, resembles peptides from EBV, cytomegalovirus, influenza A,

78

Hepatitis A, adenovirus, herpes simplex virus 1 and various papilloma viruses (Steinman 2007). It has been postulated that activation of myelin-specific T cells by viruses in the periphery, causing them to upregulate chemokine receptors and integrins, might facilitate their migration into the CNS where they initiate inflammation and mount a response against the related self-antigen.

As discussed above, the notion of an aetiological role for EBV infection is being reappraised in the light of debate about possible EBV-positive cells in MS post- mortem brain tissue. Several studies have demonstrated EBV seropositivity in around 99% of MS patients (Bray et al.1983, Hahr et al. 2006, Levin et al. 2010). In addition, reactivation of EBV has been associated with increased clinical activity (Wandinger et al. 2000). It should be noted that latent EBV infection also occurs in 90-95% of the healthy population, most of whom will never develop MS (Haahr and Höllsberg 2006). Higher serum levels of EBV-specific antibodies or prior development of mononucleosis, however, are associated with increased risk of developing MS (Lindberg et al. 1991, Ascherio et al. 2001, Levin et al. 2005, De Lorenze et al. 2006, Thacker et al. 2006, Lünemann et al. 2010). Individuals with a history of mononucleosis who also carry HLA-DR15 are reported to have a ten-fold increased risk of MS compared to those with neither (Nielsen et al. 2009). HLA-DR15 conferred only a 2.4-fold increase when there was no prior mononucleosis, suggesting that infection plays a role in genetically susceptible people (Nielsen et al. 2009). Strikingly, antibodies recognising one particular epitope of Epstein-Barr nuclear antigen 1 increased the risk of MS 24-fold in HLA-DRB1*1501+ individuals (Sundström et al. 2009). Thus, it appears that MS may be associated with a loss of latency and/or an excessive immune response to EBV, though it remains to be conclusively proven whether this truly plays a role in MS induction and in how many people.

Regardless of whether or not viruses are important, the fact that they are not unique to MS patients suggests that disease initiation requires a perturbation in the balance between activation and regulation of the immune response. In accordance with this, several studies have reported reduced frequency and function of CD4+ Tregs, CD8+ CD28- suppressor cells and iNKT cells in peripheral blood of MS patients (Crucian et al. 1995, Viglietta et al. 2004, Haas et al. 2005, Korn et al. 2007, Venken et al. 2007,

79

O’Keeffe et al. 2008, Frisullo et al. 2009, Mikulkova et al. 2010). MS patients also have a defect in thymic export of newly generated T cells, resulting in a decreased ratio of naïve: memory T cells (Hug et al. 2003, Duszczysyn et al. 2010). Circulation of newly generated naïve Tregs is reported to be important for maintaining function of all Tregs in the blood, therefore this thymic defect is proposed to be one of the causes of impaired Treg-mediated suppression in MS patients (Haas et al. 2007). Since NK cells can also have immunosuppressive functions, it is possible that perturbation of NK cell function may also play a role in MS (as discussed in chapter 1.2.5). Modest genetic associations with so many immune-related loci suggest that there may be many factors contributing to this imbalance and that the processes leading to disease initiation may vary between patients.

1.2.4 Current Therapies for MS have Benefits and Limitations

For a number of years, the main first-line treatments for RRMS have been interferon- β (IFNβ) and glatiramer acetate (GA). IFNβ comes in two forms: IFNβ1a is essentially the same as endogenous human IFNβ, whereas IFNβ1b has a slightly altered sequence and lacks glycosylation, which may play a role in stabilising the molecule (Clerico et al. 2007). These treatments are usually self-administered intravenously once or twice weekly, though IFNβ1a can also be injected intramuscularly (Clerico et al. 2007, Racke et al. 2010). All three of these therapies reduce development of new lesions and reduce relapse rate by approximately one third on average, but show only a small reduction in progression of disability (Johnson et al. 1995, PRISMS 1998, Khan et al. 2001, Clanet et al. 2002, Fletcher et al. 2002, Martinelli Boneschi et al. 2003, Haas and Firzlaff 2005, Goldberg et al. 2009, Uitdehaag et al. 2011). Many patients show no improvement and effectiveness may deteriorate due to development of autoantibodies (PRISMS 2001, Traboulsee et al. 2008). Injection site reactions and flu-like symptoms are very common, particularly amongst pateints treated with IFNβ1a (Fletcher et al. 2002, Clerico et al. 2007, Schwid and Panitch 2007). Muscle weakness, depression and leuko- or thrombocytopenia have also been reported in around 10- 20% of IFNβ-treated patients, partcicularly those on higher dosage (Clanet et al. 2002, Fletcher et al. 2002, Clerico et al. 2007). In addition, adherence to these therapies is poor due to the need for regular (at least weekly) injections (Clerico et al. 2007, Fernández et al. 2012).

80

GA is a synthetic co-polymer of alanine, glutamic acid, lysine and tyrosine, which binds to MHC class I and II and induces suppressive function in GA-specific CD8+ cytotoxic regulatory T cells, causing them to kill activated CD4+ T cells presenting GA via HLA class I (Tennakoon et al. 2006, Racke et al. 2010). It upregulates FoxP3 expression and promotes IL-10 production by Tregs (Putheti et al. 2003, Hong et al. 2005). It also converts monocytes and microglia to a “type II” phenotype, producing IL-10 instead of IL-12 and TNFα and promoting Th2 polarisation of naïve CD4+ T cells (Kim et al. 2004, Weber et al. 2004). GA-specifc CD4+ T cells can penetrate the CNS and produce BDNF (Aharoni et al. 2003, Chen et al. 2003, Ziemssen et al. 2005). They also secrete IL-10 and TGF-β within the CNS and induce production of these anti-inflammatory cytokines by resident microglia and astrocytes (Aharoni et al. 2003).

PI3K signalling induces production of IL-1R antagonist in monocytes in response to both IFNβ and GA (Molnarfi et al. 2005, Carpintero et al. 2010). IFNβ also induces a variety of effects via activation of JAK/STAT pathways (Boxel-Dezaire et al. 2010). It inhibits breakdown of the BBB and migration of leukocytes into the CNS by suppressing MMP9, upregulating TIMPs and inhibiting expression of CCR7 on DCs (Leppert et al. 1996, Stüve et al. 1996, Karabudak et al. 2004, Boz et al. 2006, Comabella et al. 2009, Yen et al. 2010). It inhibits production of TNFα and lymphotoxin, both of which are toxic to oligodendrocytes (Selmaj and Raine 1988, Selmaj et al. 1991, Soliven et al. 1991, Abu Khabar et al. 1992, McRae et al. 1998). It also inhibits proliferation and production of IFNγ and TNFα and induces IL-10 and IL-4 production in MBP-reactive T cell lines (Kozovska et al. 1999, Weber et al. 1999). IFNβ1a-treated patients have higher levels of IL-10 in both serum and CSF, which correlate with clinical response (Rudick et al. 1998). IFNβ treatment also increases frequency and function of CD4+ Tregs, CD8+ CD28- Tsuppressor cells and CD56bright NK cells (Namdar et al. 2000, Javed and Reder 2006, de Andres et al. 2007, Saraste et al. 2007, Korporal et al. 2008, Aristimuno et al. 2010, Martín- Rodriguez et al. 2010, Martín-Rodriguez et al. 2011).

IFNβ modulates antigen presentation in several ways. Firstly, it antagonises IFNγ- induced MHC class II upregulation and induces apoptosis of mature (but not immature) DCs by activation of caspase 3 and induction of caspase 11 expression

81

(Ling et al. 1985, Inaba et al. 1986, Fertsch et al. 1987, Panitch et al. 1989, Yen and Ganea 2009). It downregualtes expression of IL-12, IL-23 and IL-1β and upregulates IL-27 in DCs, thereby promoting Treg induction and inhibiting Th1 and Th17 differentiation in naïve T cells (McRae et al. 1998, Ramgolam et al. 2009). IL-12 production in monocytes is also downregulated (Karp et al. 2000). IFNβ increases the number of monocytes expressing CD86, a co-stimulatory molecule favouring Th2 induction, whilst decreasing the number of B cells expressing CD80, which promotes Th1 induction (Genç et al. 1997). It also modulates the interaction between microglia and T cells, promoting IL-10 production and inhibiting expression of TNFα, IL-1β, IL-4, IL-12 and IL-13 (Chabot and Yong 2000).

Oral alternatives for first-line therapy have recently been trialled, the most promising of which is fingolimod. Fingolimod (FTY20) is a structural analog of sphingosine. It is rapidly phosphorylated in vivo to produce fingolimod-phosphate, which binds to four of the five sphingosine-1-phosphate receptors (S1PRs), S1PR1 and S1PR3-5 (Mandala et al. 2002, Chun and Hartung 2010). S1PR1 is upregulated during clonal expansion of activated T cells, prompting T cell egression from lymph nodes into peripheral blood, where the concentration of S1P is higher (Matloubian et al. 2004, Pappu et al. 2007). Although fingolimod-phosphate initially activates receptor- signalling, it also causes receptor internalisation and degradation, leading to a decrease in cell surface S1P receptor expression (Matloubian et al. 2004). By inhibiting responsiveness to S1P, fingolimod-phosphate causes retention of recently activated and central memory T cells and B cells within thymus and secondary lymphoid tissue, thereby preventing subsequent migration to the CNS (Mandala et al. 2002, Xie et al. 2003, Brinkmann et al. 2004, Matloubian et al. 2004). Accordingly, a reduction in Th17 cells has been observed in peripheral blood of fingolimod-treated patients (Brinkmann 2009). Fingolimod-phosphate is also likely to have direct effects within the central nervous system, since fingolimod is able to cross the BBB (Foster et al. 2007). S1PRs are expressed on various CNS cell types, including neurons, oligodendrocytes, astrocytes and microglia and S1P signalling has an important role in many neural functions (Chun and Hartung 2010). Fingolimod promotes survival and proliferation of oligodendrocytes and modulates extension and retraction of processes, effects which are likely to modulate remyelination (Coelho et al. 2007, Mirion et al. 2008). It also promotes neuronal survival by upregulation of BDNF and

82 induces ERK signalling and migration in astrocytes (Mullershausen et al. 2007, Osinde et al. 2007, Deogracias et al. 2012). Based on known effects of S1P signalling, it is likely that fingolimod also modulates other aspects of neural biology such as neural process extension, astrogliosis and migration of neural stem/progenitor cells to sites of damage (Chun and Hartung 2010). In addition, fingolimod has effects on vascular endothelial cells which may alter vascular permeability (Chun and Hartung 2010).

Fingolimod is generally well-tolerated in comparison to IFNβ and is reportedly more effective (Cohen et al. 2010). Hypertension and an increased susceptibility to respiratory infections were reported during phase III trials, but serious side-effects were rare (Kappos et al. 2010). Relapse rate was reduced by around 50% compared to placebo, with around two-thirds of patients remaining relapse-free after 48 months (Kappos et al. 2010). Appearance of new gadolinium-enhancing lesions was also reduced considerably. However, fingolimod has only recently been approved and the true efficacy and safety can be more effectively assessed over time once it becomes widely available.

Several monoclonal antibodies are also available for treatment of MS: natalizumab (Tysabri), alemtuzumab, dacluzimab and rituximab. Most of these have higher efficacy compared to IFNβ and GA and can be administered less frequently, but are more expensive and associated with a higher level of risk. For this reason, they are mainly used in patients with highly active disease. The most widely used antibody treatment is Tysabri, which is administered to RRMS patients by monthly intravenous infusions. Tysabri binds to the α4 subunit of the α4β1 integrin VLA-4 on lymphocytes, monocytes and DCs, thereby blocking its interaction with VCAM, inhibiting leukocyte adhesion to vascular endothelial cells and preventing migration across the BBB (Yednock et al. 1992, Baron et al. 1993, Carrithers et al. 2000, Putzki et al. 2010, Gan et al. 2012). It also causes downregulation of both alpha and beta integrin subunits on the surface of B, T, NK and NKT cells and DCs (Harrer et al. 2011, de Andrés et al. 2012). Downregulation of VLA-4 on DCs inhibits their ability to induce antigen-specific T cell responses (de Andrés et al. 2012). Cell counts in the CSF are reduced during Tysabri treatment and the cytokine profile is altered, with decreased expression of IL-1b, IL-6, IL-8, IFNγ, IL-23, osteopontin and

83 proinflammatory chemokines and increased IL-10 (Khademi et al. 2009, Mellergard et al. 2010). However, Treg function is not restored and the frequency of activated T cells and concentration of both pro- and anti-inflammatory cytokines in peripheral blood is increased (Khademi et al. 2008, Stenner et al. 2008, Kivisäk et al. 2009, Putzki et al. 2010, Ramos-Cejudo et al. 2011). In vitro stimulation with Tysabri does not affect T cell proliferation or cytokine profile, suggesting that the changes seen in the blood are due to sequestration of cells that would otherwise have migrated to the CNS, rather than immunomodulation (Ramos-Cejudo et al. 2011).

Tysabri is much more effective than IFNβ and GA, reducing the relapse rate by 68%, risk of progression over two years by 42% and lesions by 92% (Polman et al. 2006). The most common side-effects are fatigue and allergic reaction, though the increase in these over the placebo group was relatively modest (Polman et al. 2006). Overall, it is better tolerated than other antibody treatments and severe side-effects are rare. However, Tysabri causes progressive multifocal leukoencephalopathy (PML) in around 1 in 500 patients, a serious and often fatal condition caused by reactivation of JC virus, resulting in rapid and severe damage to the CNS (Bloomgren et al. 2012). Risk increases with treatment duration and is higher in patients with anti-JC virus antibodies and those who have undergone prior immunosuppressive therapy. Amongst patients with these risk factors who had undergone at least two years of Tysabri therapy, just over 1% developed PML (Bloomgren et al. 2012). Considering the serious and potentially fatal nature of PML, this level of risk may be unacceptable for many patients, so for those who test positive for JC antibodies this is often not a suitable long-term treatment option.

Rituximab (α-CD20) binds to the B cell surface antigen CD20, causing depletion of B cells by a combination of apoptosis induction, ADCC and complement-mediated cytotoxicity (Barun and Bar-Or 2012). This leads to reduced proliferation and cytokine production by CD8+ and CD4+ (Th1/Th17) cells and reduced T cell infiltration of the CNS (Cross et al. 2006, Barun and Bar-Or 2012). In a RRMS trial, it reduced new lesion formation and halved the number of patients experiencing relapses, however the mean annualised relapse rate was not significantly different (Hauser et al. 2008). Side-effects were mostly infusion-related rather than rituximab- specific and there was no increase in the frequency of severe adverse events (Hauser

84 et al. 2008). These results suggest that the drug is relatively safe, but may be ineffective in a substantial proportion of patients, though it has not been extensively tested in MS.

Dacluzimab binds to IL2Rα (CD25) and blocks binding of IL-2 to high affinity (CD25/CD122/CD132) IL2R complexes (Martin 2012). It downregulates the co- stimulation molecule CD40L and prevents transpresentation of IL-2 to antigen- specific T cells via CD25 during interaction with DCs, thereby inhibiting T cell activation (Snyder et al. 2007, Wuest et al. 2011). However, the most important effect of daclizumab treatment may be a 7-fold increase in the frequency of peripheral blood CD56bright NK cells, which are thought to mediate killing of activated T cells (Bielekova et al. 2006, Wynn et al. 2010). Daclizumab also upregulates expression of granzymes A and K and enhances cytotoxic function of the CD56bright subset (Jiang et al. 2011). Data so far indicates that use of dacluzimab as a combination treatment with IFNβ is safe and reduces lesion formation more effectively than IFNβ alone (Wynn et al. 2010). However, trials so far have been either very small or of limited duration, so the long-term efficacy and safety of dacluzimab in MS patients remains to be elucidated (Martin 2012).

Alemtuzumab binds to CD52 on the surface of mature leukocytes, inducing ADCC by NK cells and neutrophils (Klotz et al. 2012). This results in rapid and long-lasting depletion of T cells and a more transient depletion of B cells and monocytes. Alemtuzumab also enhances secretion of BDNF, PDGF and ciliary neurotrophic factor (Jones et al. 2010). In a clinical trial it was found to halve the relapse rate, significantly inhibit disease progression and in some cases to reverse it (Coles et al. 2011, Coles et al. 2012). However, around 20% of patients develop autoimmune disease. These are most commonly thyroid disorders, but there have also been incidences of idiopathic thrombocytopenic purpura, a potentially fatal autoimmune condition which inhibits blood clotting (Cossburn et al. 2011, Coles et al. 2012).

Most of these treatments are only effective during the relapsing-remitting phase of disease. Progressive MS and aggressive RRMS may be treated with mitoxantrone. Mitoxantrone intercalates into DNA, causing cross-linking and double-strand breaks and inhibits the DNA repair enzyme topoisomerase II (Fox 2006). It has a broad immunosuppressive effect by inhibiting proliferation and inducing cell death in B

85 cells, T cells, DCs and monocytes (Chan et al. 2005, Neuhaus et al. 2005). It also stimulates suppressive function in APCs, causing them to inhibit T cell proliferation and B cell antibody production and promote Treg rather than T effector function (Fidler et al. 1986a, Fidler et al. 1986b, Neuhaus et al. 2005). In RRMS patients, mitoxantrone reduces relapse rate by 65-80% and development of new lesions by 80% (Hartung et al. 2002, Krapf et al. 2005, Le Page et al. 2008). It also stabilises or reverses disability progression (Hartung et al. 2002, Le Page et al. 2008). Common side-effects of mitoxantrone therapy include nausea/vomiting, alopecia, menstrual disorders and susceptibility to urinary and respriratory tract infections (Hartung et al. 2002, Cohen and Mikel 2004). Rare cases of serious infection and leukaemia have also been reported and there is a dose-dependent risk of cardiotoxicity (Cohen and Mikel 2004, Fox 2006, Goffette et al. 2005, Le Page et al. 2008, Le Page et al. 2011). Due to its toxicity, duration of mitoxantrone therapy is usually limited to 2-3 years (Fox 2006).

In summary, though there are a number of therapies available for MS, most either have limited efficacy or high risks. Tysabri has proven effective and safe in the short term but the increasing risk over time means it is not an appropriate long-term solution for many patients. In order to develop more effective and safer therapies it is important to understand the immunological mechanisms involved in the disease. The majority of research into MS has so far focussed on CD4+ and CD8+ effector T cells. However, there is increasing evidence that other leukocytes including NK cells have a role in the disease.

1.2.5 Evidence of a Role for NK Cells in Multiple Sclerosis

1.2.5.1 NK Cells Influence Development of EAE

NK cell depletion studies with anti-NK1.1 antibody in the EAE model have indicated both pathogenic (Shi et al. 2000, Winkler-Pickett et al. 2008) and protective (Zhang et al. 1997, Xu et al. 2005) roles for NK cells. The exact reason for these discrepancies is unclear, though differences in the method of EAE induction and dose, timing and route of administration of NK-cell depleting antibodies could all have an impact (Winkler-Pickett et al. 2008). Notably, the form of EAE induced in the studies where NK cell depletion ameriolated disease was considerably more severe, with a much

86 higher clinical disease score in control mice compared to the other studies. Therefore it seems that the role of NK cells may vary depending on the exact nature of disease.

None of the depletion methods used in these studies were specific to NK cells. All three studies used an anti-NK1.1 antibody, which also depletes a subset of T cells. Winkler-Pickett et al. obtained similar results using anti-asialo GM, which does not deplete NK1.1+ T cells, but may have damaging effects on both T cells and macrophages, and using anti-Ly49H, which depletes only a subset of NK cells. The study by Zhang et al. used an anti-NK1.1 antibody in both WT and β2m-/- mice, which do not have NK1.1+ T cells. However, the latter mice also lack all CD8+ T cells and CD4- CD8- T cells, and the effect of depleting NK cells in this environment will not necessarily be the same as in an immunocompetent mouse. Nonetheless, the combined data do strongly suggest that NK cells influence development of EAE.

EAE studies have also provided some mechanistic insights into the role of NK cells. Firstly, NK cells must be depleted prior to disease induction in order to suppress EAE, suggesting that they are more important in initiation, rather than maintenance, of inflammation (Winkler-Pickett et al. 2008). Immunisation with MOG increases the number of NK cells in the draining lymph node, where they might influence antigen presentation by DCs. T cells isolated from draining lymph nodes of NK cell-depleted mice were severely impaired in their ability to proliferate and produce cytokines in response to MOG peptide and fewer T cells infiltrated the CNS of NK cell-depleted mice (Winkler-Pickett et al. 2008).

IL-18-/- mice are defective in mounting Th1 and antibody responses and are resistant to EAE (Shi et al. 2000). As described in chapter 1.1.6, IL-18 promotes co- stimulatory function and IFNγ production in NK cells and the presence of IL-18- stimulated NK cells during T cell priming promotes Th1 differentiation (Mailliard et al. 2005). In accordance with this, adding exogenous IL-18 to these mice restored Th1 responses and disease susceptibility only when NK cells were present (Shi et al. 2000). Transfer of NK cells from IL-18-sufficient mice also reversed the Th1 defect, but only when the donor mice expressed IFNγ (Shi et al. 2000).

Similar effects were seen in mice treated with IL-21. IL-21 injected prior to, but not after, EAE induction increased the disease severity, however this effect was abrogated

87 by depletion of NK cells, suggesting that activated NK cells can enhance disease initiation (Vollmer et al. 2005). IL-21 doubled NK cell IFNγ production on a per cell basis. EAE induction by transfer of autoreactive T cells was also inhibited when the donor mice had been treated with IL-21 prior to MOG immunisation. Though this could be due to direct effects of IL-21 on T cells, it may indicate that IL-21-activated NK cells are enhancing priming/ Th1 polarisation of encephalatogenic T cells. The studies by Vollmer and Shi both used anti-NK1.1 antibodies for NK cell depletion, however effects of NKT cells were excluded using CD1d-/- mice.

In the study by Zhang et al., where NK cells were protective, serum levels of TNFα and IFNγ were much higher in NK cell-depleted mice, implying a suppressive effect of NK cells on pro-inflammatory cytokine production (Zhang et al. 1997). Antigen- induced proliferation of MOG-specific T cells was higher when incubated with APCs from NK cell-depleted mice, indicating that NK cells were negatively regulating APC function, in contrast to the data from the Shi and Winkler-Pickett studies (Zhang et al. 1997). NK cells also selectively kill encephalitogenic PLP-specific T cells (Xu et al. 2005).

The role of NK cells within the CNS has also been investigated during EAE (Huang et al. 2006, Hao et al. 2010). In these studies, green fluorescent protein (GFP) was inserted into the CX3CR1 gene, creating CX3CR1+/GFP mice, in which CX3CR1+ cells could be tracked by fluorescent imaging, and CX3CR1GFP/GFP mice, which express GFP instead of CX3CR1. CX3CR1 is a receptor for fractalkine (CX3CL1), which is expressed in the inflamed CNS (Pan et al. 1997). It is important for recruitment of NK cells to the CNS, but is not required for infiltration of T cells, monocytes, macrophages or iNKT cells (Huang et al. 2006). CX3CR1GFP/GFP mice had much fewer NK cells in the CNS and developed more severe EAE, with extensive demyelination and an increased mortality rate (Huang et al. 2006, Hao et al. 2010). Anti-NK1.1-mediated NK cell depletion in CX3CR1+/GFP mice resulted in a similar exacerbation of disease (Huang et al. 2006). Both CX3CR1 knockout and NK cell depletion caused a marked increase in MOG-induced IL-17 production by CD4+ T cells in the CNS, but NK cell depletion did not significantly effect IL-17 production by lymph node CD4+ T cells (Hao et al. 2010). Conversely, NK cell-depleted mice had much lower IFNγ production in lymph node CD4+ T cells, but slightly higher

88 production in the CNS. NK cell depletion also led to enhanced proliferation of CD8+ T cells in lymph node but not CNS. This suggests that the role of NK cells in the CNS is likely to be different from that in the periphery.

The profound effect of NK cells on IL-17 production in the CNS is likely to be mediated via effects on microglia. Microglia are the most potent APCs for Th17 induction and this function is enhanced in NK cell-depleted mice (Hao et al. 2010). NK cell MIP1α production and microglial MCP-1 production result in reciprocal chemoattraction of the two cell types. NK cells are then able to inhibit microglial activation and production of cytokines involved in Th17 polarisation (IL-1, IL-6 and IL-23). Microglia are also susceptible to lysis by NK cells, owing in part to their low expression of the NKG2A ligand Qa1 (Hao et al. 2010). Blockade of NKG2A with an anti-NKG2A F(ab’)2 fragment led to marked ameriolation of MOG-induced EAE (Leavenworth et al. 2010). This was associated not only with inhibition of microglial activation, but also increased numbers of activated NK cells in the CNS, reduced T cell infiltration and impaired recall responses to MOG, though it was not clear whether the latter was due to reduced activation at the individual cell level or increased killing of MOG-specific T cells. NKG2A blockade also resulted in a shift in the cytokine profile of splenic, lymph node and CNS-derived CD4+ T cells, from pro- inflammatory IFN-γ and IL-17 towards anti-inflammatory IL-4 and IL-10 (Leavenworth et al. 2010). NK cells might also ameliorate EAE by promoting neural regeneration, since NK cells infiltrating the CNS of EAE rats produce high quantities of neurotrophic factors such as BDNF and neurotrophin-3 (Hammarberg et al. 2000).

1.2.5.2 Genetic Factors Affecting NK Cell Function have been Associated with Susceptibility to MS

Identification of HLA class I and KIR genes associated with increased or decreased susceptibility to MS indicates a potential role for NK cells in development of human MS. HLA alleles containing the Bw4 motif recognised by KIR3DL1, such as HLA- B*44, are underrepresented amongst MS patients, as is HLA-Cw5, a ligand for KIR2DL1 and KIR2DS2 (Yeo et al. 2007, International MHC and Autoimmunity Genetics Network et al. 2009, Lorentzen et al. 2009, Bergamaschi et al. 2010, Healey et al. 2010). Conversely, HLA-A3, a ligand for KIR3DL2, is associated with increased risk of MS (Fogdell-Hahn et al. 2000, Harbo et al. 2004). Whether these

89 associations are related to modulation of T cell function or of NK cells, or both, is unclear. Information on the contribution of KIR genotype to MS susceptibility is relatively sparse, however there do appear to be some associations. KIR2DS1 protects aginst MS in individuals who also possess alleles encoding its ligand HLA-C2 (Lorentzen et al. 2009, Fusco et al. 2010). Individuals homozygous for a haplotype containing the tightly linked 2DL2 and 2DS2 alleles (which lack 2DL3 since this segragates to the same locus) have increased susceptibility to MS, as do those carrying the KIR2DS4*001 or *002 allele (Fusco et al. 2010, Jelcić et al. 2012). These studies do not rule out a possible contribution of KIR+ T cells to disease modification, however KIRs are central to NK cell biology, whereas they are only expressed on a small percentage of T cells.

Other genetic associations have been identified which might impact upon NK cell function, including predispotion towards MS in those possessing a null allele of LILRA3, encoding a soluble receptor for MHC class I which is predicted to antagonise activating receptors for MHC class I (Koch et al. 2005, Ordóñez et al. 2009). The aforementioned association with DNAM-1 could also affect NK cell activation, though neither of these receptors is exclusive to NK cells. Given that blocking of IL2Rα causes expansion of protective CD56bright NK cells, the MS-associated IL2Rα alleles might also modulate NK cell proliferation and function. Variation at the IL7R locus might also have effects on CD56bright NK cells, all of which express this receptor (Hanna et al. 2004b, Nielsen et al. 2012).

1.2.5.3 The Role of NK Cells in Human MS Requires Further Investigation

As described in chapter 1.1.5, NK cells are known to produce many of the cytokines associated with MS pathogenesis, including IFNγ, TNFα and GM-CSF. In addition to their ability to influence T cell polarisation and kill activated T cells and monocytes, there is evidence that NK cells might also have effects on cells of the CNS, such as oligodendrocytes, neurons, astrocytes and microglia. Oligodendrocytes in MS white matter lesions express high levels of MICA and MICB (Saikali et al. 2007). Primary human oligodendrocytes expressing these ligands are susceptibe to NKG2D- dependent lysis by IL-2-activated autologous NK cells in vitro (Morse et al. 2001, Saikali et al. 2007). NK cells also kill autologous resting microglia, which constitutively express ligands for NKG2D and DNAM-1 (Lünemann et al. 2008).

90

LPS stimulation of microglia causes them to upregulate MHC class I and downregulate NKG2D ligands, making them less susceptible to lysis by NK cells, particularly at high doses of LPS (Lünemann et al. 2008). Whether activation by the pro-inflammatory cytokine mileu in the inflamed CNS would also protect microglia from NK cells is uncertain. NK cells have also been shown to kill human fetal astrocytes in culture, leading to subsequent displacement of neurons in mixed glial cultures (Morse et al. 2001, Darlington et al. 2008). However, the astrocytes used in these studies were heterologous, so it isn’t clear if NK cells would kill astrocytes in vivo. Direct killing (or lack of) of human neurons by NK cells has not been demonstrated, but murine data indicates that NK cells can kill peripheral, but not CNS-derived neurons, since only the former express ligands for NKG2D (Backström et al. 2003).

Presence of NK cells in demyelinating lesions has not been conclusively demonstrated in human MS. Although one study suggested that they were present, the antibody used to identify NK cells was directed against CD57, which is not in fact specific to NK cells (Traugott and Raine 1984). Our research group did not detect substantial infiltration of NK cells in MS brain, however since samples are only available post- mortem, and mostly from patients in a progressive stage of disease, this doesn’t exclude the possibility that they are present in the brain of relapsing-remitting patients (Durrenberger et al. 2012). Both CD56dim and CD56bright NK cells have been found in the CSF of RRMS patients, so it seems likely that they might also penetrate the parenchyma (Bielekova et al. 2011).

A number of studies have investigated changes in NK cell function in MS patients, with many of these suggesting that NK cell cytotoxic function might be important in preventing inflammation. Benczur et al. (1980) reported a decrease in cytotoxic function of NK cells towards K562 targets using peripheral blood lymphocytes as effectors, and lack of augmentation by interferon or poly(I:C), though this was contradicted by others (Santoli et al. 1981). Some of the patients in the Benczur study were on treatment and were not stratified accordingly. Decreased cytotoxicity and impaired augmentation by IL-2 was also reported in another study, in which over half the patients had the progressive form of MS (Braakman et al. 1986). None of the studies differentiated between patients with relapsing-remitting and progressive forms

91 of MS, though all included both in varying proportions. A subsequent study, where patients were stratified, indicated that abnormally low lytic activity was a feature of progressive, but not relapsing-remitting disease (Vraneš et al. 1989). Lysis of K562 cells in this study correlated with the frequency of CD56+ leukocytes in peripheral blood, which was also decreased in progressive patients. This suggests that apparent differences in previous studies may also have been due to differences in NK cell frequency rather than their individual cytotoxic capacity.

Impaired NK cell lytic capacity, as determined by in vitro lysis of K562 targets, has been observed immediately prior to onset of clinical relapse and appearance of new lesions in RRMS patients (Kastrukoff et al. 1998, Kastrukoff et al. 2003). However, these so-called valleys in lytic activity did not always result in a relapse, and given the apparent cyclical nature of lytic capacity in both patients and controls and the small number of patients studied, it is hard to be confident in whether this is really relevant to disease.

All of the studies mentioned so far were also complicated by the use of PBMC rather than sorted NK cells as the effector population, since CD33+ and CD56+ CD3+ cells are also capable of killing K562 (Kastrukoff et al. 2003), differences in frequencies of effector populations could influence results and because differences in NK cell cytotoxicity in this assay could be either cell intrinsic or the result of regulation by other cell types. Data from our own lab, using the degranulation marker CD107a, suggests that NK cells from patients in relapse actually degranulate more in response to K562 stimulation compared to patients in remission, though stable patients still had higher levels of degranulation than healthy controls (A. Al-Aneezi, unpublished data). Furthermore, the frequency of CD69+ NK cells is higher in MS patients, suggesting that they are in a more activated state (Lünemann et al. 2011).

1.2.5.4 NK Cell Subsets have been Associated with Protective and Pathogenic Roles in Multiple Sclerosis

Given that NK cells have both pro- and anti-inflammatory functions, some of which are associated with particular subsets, it seems likely that different subsets of NK cells could have opposing roles in multiple sclerosis. Several subsets have already been associated with regulatory roles in MS.

92

Munscahuer et al. (1995) reported a significantly lower frequency of CD8- NK cells (CD3- CD56+ CD8- ) in PBMCs of untreated RRMS patients compared to controls. The frequency of CD16+ CD56+ CD8- NK cells was also significantly lower, which likely indicates a decrease in the CD56dim CD8- NK cell population. Whether the CD56bright CD8- population was also reduced is unclear. There was also a slight, but non-significant increase in the frequency of CD8+ NK cells (again both CD3-CD56+ CD8+ and CD16+ CD56+ CD8+ populations were altered). In contradiction to this, De Jager et al. (2008) reported a decrease in CD8low CD4- lymphocytes in RRMS patients compared to healthy controls, and upon further investigation this difference was found to be attributable to CD8+ NK cells. No clear associations were found between disease course and frequency of CD8+ NK cells, though the sample size was fairly small given the large number of variables. Since the study was not originally designed to investigate NK cell subsets specifically, there is no mention of whether or not the frequency of CD8- NK cells was also altered. This makes it difficult to assess whether the data is due to a decrease in total NK numbers or whether the proportion of NK cells expressing CD8 is altered.

Two studies have reported similar frequencies of CD56dim and CD56bright NK cells in MS peripheral blood compared to healthy controls (Lünemann et al. 2011, Martin- Rodriguez et al. 2011). However, CD56bright NK cells from RRMS patients had impaired in vitro proliferative and IFNγ response to IL-12 stimulation, both when purified and in the presence of other leukocytes (Lünemann et al. 2011). CD56bright NK cells, unlike Tregs, are expanded during the third trimester of pregnancy when the risk of relapse is decreased, and might contribute to the increased ratio of Th2: Th1 observed during this time (Airas et al. 2008). They induce caspase-independent apoptosis of activated (but not resting) CD4+ T cells by releasing granzyme K, which causes depolarisation of the mitochondrial membrane and ROS production in the T cell (Jiang et al. 2011).

CD56bright NK cells expand in response to IFNβ therapy, and to an even greater extent following dacluzimab (αCD25) therapy (Bielekova et al. 2006, Saraste et al. 2007, Vandenbark et al. 2009, Martín-Rodriguez et al. 2010, Wynn et al. 2010, Martín- Rodriguez et al. 2011). IFNβ (but not daclizumab) also causes a decrease in absolute numbers of CD56dim NK cells, though the difference is less substantial (Bielekova et

93 al. 2006, Martín-Rodriguez et al. 2011). The proportion of CD56bright NK cells in the CSF is also increased following daclizumab treatment (Bielekova et al. 2011). CD25 combines with the main IL2R chains CD122 and CD132 to form a receptor with higher affinity than the other two chains alone. By blocking CD25, which facilitates high affinity binding of IL-2 to the T cells which produced it, there is more IL-2 available to bind to the intermediate affinity CD122/CD132 IL-2 receptor, which is constitutively expressed on NK cells (Martin et al. 2010, Martin 2012). CD56bright NK cells express higher levels of CD122 than other lymphocytes, therefore are well- placed to take advantage of increased IL-2 availability (Bielekova et al. 2006). CD56bright expansion is associated with positive outcomes during both daclizumab and IFNβ treatment, indicating that these cells may have an important role in suppression of autoimmunity (Wynn et al. 2010, Martín-Rodriguez et al. 2011).

Another study reported a decrease in CD16+ CD3- cells (corresponding to the CD56dim NK cell subset) and CD57+ CD3- cells following interferon-β treatment (Perini et al. 2000). This suggests that the therapy is not only increasing the frequency of regulatory NK cells, but also reducing the frequency of more mature, cytolytic NK cells. However, no comparison was made to healthy controls in this study.

NK cells from Japanese MS patients in remission express higher levels of CD95 and produce more IL-5 compared to patients in relapse or healthy controls (Takahashi et al. 2001). NK cells which are polarised towards this CD95high “NK2” phenotype by in vitro culture with IL-4 and α-IL-12 neutralising antibody inhibit IFN-γ and enhance IL-4 production by T cells stimulated with PMA and ionomycin. IL-5 is partially, but not completely responsible for mediating Th2 polarisation. This may be an important regulatory mechanism in some patients, as NK cell-depleted PBMC from patients with high CD95 expression contain a higher proportion of MBP-reactive CD4+ T cells than undepleted PBMC, whereas NK depletion has no effect on MBP response of T cells from “CD95low” patients (Takahashi et al. 2004). Low expression of CD11c is also associated with expression of IL-5 and GATA-3, although the CD11clow subset does not correspond to the CD95high subset (Aranami et al. 2006). Patients with a high proportion of CD11clow NK cells had a longer time until their next relapse compared to those with predominantly CD11chigh NK cells (Aranami et al. 2006). High

94 expression of CD11c correlates with the percentage of HLA-DR+ NK cells, which is also increased in MS patients (Aranami et al. 2006, Lünemann et al. 2011).

Stable RRMS patients also have a lower percentage of CX3CR1+ NK cells than healthy controls, whereas patients with active disease have a marginally higher frequency than controls (Infante-Duarte et al. 2005). This was measured as a percentage of NK cells rather than percentage of lymphocytes. Since stable MS patients also had lower levels of CX3CR1 mRNA in total PBMC, it is most likely that there was an overall decrease in frequency of CX3CR1+ NK cells, however it is unclear whether there was also a corresponding increase in the frequency of CX3CR1- NK cells. CX3CR1+ NK cells have higher cytotoxic activity towards K562 targets (Infante-Duarte et al. 2005). Since CX3CR1 is important in mediating migration to the CNS in mice, it is possible that these more cytotoxic NK cells are participating in destruction of oligodendrocytes and astrocytes in the CNS during relapse and that patients able to regulate their numbers might therefore be protected from relapse. Alternatively, CX3CR1+ NK cells might be promoting inflammation in the periphery, perhaps by killing regulatory cells, thereby increasing the likelihood of T cell activation and CNS infiltration. The lack of difference between healthy controls and patients in relapse suggests that the frequency of this subset is not however a primary cause of disease.

95

1.3 Aims

Given the evidence for both pathogenic and anti-inflammatory functions for NK cells in MS, we hypothesised that each of these functions might be performed by specific subsets of NK cells and that the frequency of such subsets might be altered in MS patients. Altered frequencies of several subsets have already been observed in MS patients compared to healthy controls and beneficial effects of therapies have been associated with specific types of NK cells. However, these subsets have each been identified based on expression of just one marker and it is not clear how the different observations are related. The aim of this study was to obtain a more comprehensive understanding of NK cell subsets in MS, in the following ways:

1. Classify NK cell subsets using a combination of markers and identify specific subsets which have higher or lower frequency in RRMS patients compared to healthy controls (chapter 3).

2. Investigate the effects of two widely-used MS therapies (IFNβ and Tysabri) on frequencies of NK cell subsets (chapter 3).

3. Investigate the phenotype and function of subsets of interest, including transcriptional profiles, cytokine production and cytotoxic function (chapter 4).

4. Determine whether pathogenic and/or regulatory NK cells are found in the CNS during spontaneous demyelinating autoimmune disease, using the Line 7 mouse model of MS (chapter 5).

96

2. METHODS

2.1 Human PBMC Isolation Blood samples were obtained from patients and healthy controls with full informed consent and ethical approval. MS and clinically isolated syndrome (CIS) patients were recruited from Charing Cross Hospital, West London, by Dr. Fatemah Kamel, Dr. Paolo Giannetti and Dr. Pascal Durrenberger. Patients were classified by neurology consultants Dr. Richard Nicholas and Dr. Omar Malik. RRMS patients experiencing an acute exacerbation in symptoms at the time of sampling were said to be in relapse, whereas those with no new visible symptoms were said to be in remission. CIS patients are defined in the clinic patient notes and database as those that have been referred to the MS Clinic due to a single, isolated attack of symptoms that may be MS related (such as blurred vision) but have not yet received a confirmed diagnosis of MS. A significant proportion of these will subsequently be confirmed as MS cases following further relapse and clinical confirmation according to McDonald criteria (www.nationalmssociety.org/download.aspx?id=29943). As such, analysis of CIS patients is considered to offer an invaluable insight into the early immunological events in a long-term, chronic disease. Sixteen healthy controls (see table 2-vii), twelve CIS patients (table 2-viii), twelve untreated RRMS patients (table 2-ix), twelve RRMS patients on interferon-β therapy (table 2-x) and eight on Tysabri therapy (table 2-xi) were analysed for peripheral blood frequencies of NK cell subsets. Only patients who had not been treated with steroids for a relapse within the last two months were included. Patients on IFNβ or Tysabri therapy had been on therapy for at least 2 months prior to sampling. Eleven of the twelve untreated patients had never been on disease-modifying treatment for MS and one (MSU2) had ceased treatment with GA one year prior to sample acquisition.

Blood samples were collected in sodium heparinised vacutainers (Becton Dickinson, UK) and PBMC isolated by Histopaque density gradient centrifugation as follows. Blood was mixed with an equal volume of sterile phosphate buffered saline (PBS) (Invitrogen, UK) and overlayed onto 15ml room temperature Histopaque (Sigma- Aldrich, UK) in a 50ml centrifuge tube, then centrifuged at 800g for 20 minutes (acceleration 5, brake 1). PBMC were removed from the Histopaque/plasma interface using sterile Pasteur pipettes and transferred into fresh 50ml tubes, which were then

97 filled to the top with PBS and centrifuged at 350g for 10 minutes (acceleration 9, brake 9). Following removal of supernatant, PBMC were washed twice more with PBS, first at 300g, then at 200g for platelet removal. Cells were resuspended in 10ml FACS buffer and mixed thoroughly prior to counting with trypan blue (Sigma- Aldrich, UK) for exclusion of dead cells.

2.2 Flow Cytometry Evaluation of Human PBMC (Surface Staining) Cells were centrifuged once more at 300g and supernatant removed completely, using a 1000μl pipette to remove residual liquid from the inverted tube. Cells were blocked using 1:10 diluted human FcR blocking reagent (Miltenyi Biotec, UK) in 100µl FACS buffer per 107 cells, for 15 minutes at 4ºC. Cells were resuspended in FACS buffer at a concentration of 1x 107 cells/ml and 50µl (5 x 105) cells were added to each well of a 96 well plate containing 100μl pre-mixed antibodies in FACS buffer. Antibody cocktails for experimental and fluorescence minus one controls (FMOs) were prepared on the morning of the experiment prior to adding cells, in order to maintain a consistent staining duration. For each FMO, one antibody was removed and replaced with an isotype control, in order to control for both non-specific binding and spectral overlap. (For full list of antibodies see table 2-iii.) PBMC plus antibody cocktails were mixed briefly by pulse vortexing and incubated for 30 minutes at 4ºC in the dark. Cells were then washed twice with FACS buffer, resuspended in 1% paraformaldehyde (Sigma-Aldrich, UK), stored at 4ºC and acquired on a BD FACS Aria II flow cytometer within 48 hours (50,000 lymphocytes were acquired per sample). Single antibody-stained cells were also acquired in order to perform compensation. Compensation was performed using BD FACSDiva software and further data analysis was carried out using FlowJo 7.5 (Treestar, USA).

Lymphocytes were gated according to forward and side-scatter and doublets were excluded using the forward scatter-width parameter. For multiparameter subset frequency analysis, cells were gated as shown in figure 2-i and categorised into ten subsets (summarised in figure 2-ii).

2.3 PBMC Stimulation and Intracellular Cytokine Staining PBMC were resuspended at a concentration of 1 x 106 cells/ml in culture medium and incubated for 6 hours (37ºC, 5% CO2) in the presence of 50ng/ml PMA, 1µg/ml

98 ionomycin (Sigma-Aldrich, UK) and 1µl/ml (1000x) monensin (Biolegend, UK). Cells were transferred to a 50ml centrifuge tube and washed twice with FACS buffer. Fc receptor-blocking was performed as described in section 2.2, before diluting cells to 2 x 107 cells/ml in FACS buffer. Cells were then transferred to FACS tubes (1 x 106 PBMC/ tube) and stained with antibodies against surface markers as described in section 2.2. Intracellular staining was performed using Cytofix/Cytoperm buffer and

Perm/Wash solution (diluted 1:10 in H2O) from the BD Cytofix/Cytoperm fixation/permeabilization kit (Becton Dickinson, UK). Following surface staining, cells were washed twice with FACS buffer, resuspended thoroughly in 250µl Cytofix/Cytoperm buffer and incubated at 4ºC for 20 minutes. After washing twice with 1ml per tube (1x) Perm/Wash solution, supernatant was removed thoroughly and cells were resuspended in 100µl (1x) Perm/Wash solution. Fluorochrome conjugated anti-cytokine antibodies and isotype controls were added and samples incubated at 4ºC in the dark for 30 minutes. Cells were washed twice with (1x) Perm/Wash solution and resuspended with 100µl 2% paraformaldehyde. Samples were stored at 4ºC in the dark and acquired the following day on BD FACS Aria II flow cytometer. (100,000 lymphocytes were acquired per sample). Data analysis was performed as described in section 2.2.

2.4 Cell Sorting for RNA PBMC were isolated from 40ml of blood obtained from healthy female volunteers and Fc-receptor blocking was performed as described in section 2.2. 5 x 105 PBMC were added to pre-prepared antibody mixes in FACS tubes for compensation (single stains) and gating (all five antibodies and FMOs for CD56, CD8, CD27 and CD57). The rest of the cells were stained in a 50ml centrifuge tube. 130µl FACS buffer, 10µl CD56, 10µl CD27, 20µl CD8 PerCP-Cy5.5, 20µl CD57 FITC and 10µl CD3 APC were added per million PBMC. Cells were incubated at 4ºC for 30 minutes in the dark, then washed twice and resuspended in FACS buffer. PBMC for sorting were diluted to approximately 5 x 106/ml and filtered through a 100µM filter into sterile polypropylene FACS tubes (both from Becton Dickinson, UK). Cells were sorted on the BD FACS Aria II fluorescence activated cell sorter using the 70 micron nozzle and “4-way purity” mode, with a sample incubation temperature of 4ºC. Voltages and compensations were set using single-stained cells prior to acquiring the FMO control tubes, which were used for setting gates. Lymphocytes were identified according to

99 forward and side scatter and doublets were excluded by plotting forward scatter-area vs. forward scatter-width. NK cells were defined as CD56+ CD3- singlets. CD56dim CD8+ CD57+ CD27- and CD56dim CD8- CD57- CD27- NK cell populations were defined using FMOs. In order to enhance purity, cells were sorted into far left and far right positions only, which are furthest from the waste stream and flow rate was set so as to sort no more than 2000 events per second. First, ≥ 2000 cells of each subset were sorted into polypropylene FACS tubes containing 300µl FACS buffer, then resuspended by repeatedly running buffer down the sides of the tube followed by vortexing. These samples were then acquired on the FACS Aria II to check their purity (see figure 2-iii). The remainder of the PBMC were then sorted into cold polypropylene tubes containing 300µl lysis buffer RLT Plus (Qiagen, UK) supplemented with 2-Mercaptoethanol (Sigma-Aldrich, UK). Once sorting was complete, samples were made up to a volume of 350µl (or a multiple of 350µl, if volume already over 350µl) with lysis buffer RLT Plus and resuspended by running buffer lysis buffer down side of tube. Suspensions were then transferred to 1.5ml Eppendorf tubes and vortexed thoroughly in order to homogenize samples. Lysates were stored at -80ºC until required for RNA extraction.

2.5 RNA Extraction RNA was extracted from sorted NK cells using the RNeasy Plus Micro Kit (Qiagen, UK), as follows. Buffer RPE was diluted with 4 volumes of 99% molecular grade ethanol (Sigma-Aldrich, UK). Lysates were thawed and briefly vortexed before transferring 350µl to gDNA Eliminator spin columns inside 2ml collection tubes. Following centrifugation for 30 seconds at 16,000g, columns were discarded and 350µl 70% ethanol was added to the flow-through and mixed by pipetting. Flow- through was then transferred to RNeasy MiniElute spin columns in 2ml collection tubes and centrifuged at 16,000g for 30 seconds. Flow-through was discarded using a pipette and column was then washed with 700µl buffer RW1, followed by 500µl buffer RPE by centrifuging for 30 seconds, 16,000g and discarding flow-through. 500µl 80% ethanol was then added to the spin column and centrifuged for 2 minutes at 16,000g. Collection tube and flow-through were discarded and column was transferred to a fresh 2ml collection tube. Columns were then centrifuged for 5 minutes at 16,000g with lids open in order to remove residual ethanol from spin column membrane. Columns were placed in 1.5ml collection tubes and 20µl RNase-

100 free water was added directly onto the spin column membrane, ensuring that the membrane was fully covered. Columns were then centrifuged at 16,000g for 2 minutes to elute RNA. RNA concentrations were determined using a NanoDrop spectrophotometer (Thermo Scientific, UK).

2.6 Reverse Transcription Reverse transcription was performed using the RT2 Nano PreAMP cDNA Synthesis Kit (Qiagen, UK). The following were added to PCR tubes: 2µl (5x) gDNA elimination buffer, 20ng RNA and RNase-free water up to a volume of 10µl. Contents were mixed by gentle pipetting, followed by brief centrifugation. Samples were incubated for 5 minutes at 42ºC, then immediately chilled on ice for at least one minute. RT cocktail was prepared as shown in table 2-i.

10μl RT cocktail was then added to each sample, mixed well by pipetting and briefly centrifuged. Samples were incubated at 42ºC for 30 minutes for reverse transcription, then immediately at 95ºC for 5 minutes to stop the reaction. cDNA was then pre- amplified for use in quantitative(q) RT-PCR arrays or individual qRT-PCR reactions.

Table 2-i: Reagents for reverse transcription reaction

Reagent Volume per reaction (μl)

BC3 (5x reverse transcription buffer 3) 4

P2 (Primer and external control mix) 1

RE (cDNA enzyme synthesis mix) 1

RI (RNase inhibitor) 1

RNase-free water 3

2.7 Quantitative RT-PCR Arrays Nano PreAMP cocktail was prepared by mixing 12.5μl RT2 PreAMP PCR master mix per reaction with 7.5μl RT2 PreAMP pathway primer mix Human Th17 for Autoimmunity and Inflammation. 5µl cDNA and 20µl Nano PreAMP cocktail were

101 added to 0.2ml PCR tubes, mixed gently by pipetting and centrifuged before performing PCR as follows: 10 minutes at 95ºC (to activate HotStart Taq DNA polymerase), followed by 12 PCR cycles (95ºC for 15 seconds, 60ºC for 2 minutes). Samples were placed on ice briefly, then 2μl side reaction reducer was added, mixed and centrifuged. Samples were incubated in a thermal cycler at 37ºC for 15 minutes to eliminate residual primers, followed by 95ºC for 5 minutes to stop the reaction. 84μl of RNAse/DNase-free water (Sigma-Aldrich, UK) was added to each sample and mixed. Samples were then stored at -20ºC until use.

For qPCR array, 102µl pre-amplified cDNA was mixed with 1350µl (2x) RT2 SYBR Green ROX qPCR Mastermix (Qiagen, UK) and 1248µl RNase/DNase-free water (Sigma-Aldrich, UK) in a 50ml centrifuge tube. 25µl of this mixture was added to each well of the Human Th17 for Autoimmunity & Inflammation RT² Profiler PCR Array (Qiagen, UK). Plate was sealed and centrifuged for 1 minute at 1000g to remove bubbles, before running qPCR on Mx3000P (Stratagene, UK), as follows:

1. Activation of HotStart DNA Taq polymerase (1 cycle, 10 minutes, 95ºC) 2. Amplification: 40 cycles of (95ºC for 15 seconds, 60ºC for 1 minute) 3. Automated dissociation curve

Threshold was set at 0.13 for all samples and Ct values exported to Excel for analysis. Values were uploaded to the SABiosciences PCR Array Data Analysis Web Portal (http://pcrdataanalysis.sabiosciences.com/pcr/arrayanalysis.php) for calculation of fold changes in gene expression (by ΔΔCt method) and clustergram generation. Gene expression was normalised to four of the five housekeeping genes provide in the array: HPRT1, RPL13A, GAPDH and ACTB. (β2-microglobulin was excluded as its expression was not consistent with that of the other housekeeping genes or between the two NK cell subsets.) Statistical significance of fold differences of ≥1.2 between gene expression in the CD56dim CD27- CD8+ CD57+ subset and the CD56dim CD27- CD8- CD57- subset was determined by paired Wilcoxon signed rank test (performed using GraphPad Prism version 5).

102

2.8 Individual qRT-PCR Assays Reverse transcription was performed as described in section 2.6. Primers for human GAPDH, HPRT1, granzyme A, granzyme B, perforin, interferon-γ, Tbx21, RORC, IL12RB1, IL12RB2, CD127, IL23R, IL-23A, IL-17D, IL-27 and S1PR1 were purchased from Qiagen, UK. Primer cocktail was prepared by mixing primers for 6 or 8 genes of interest with primers for GAPDH and HPRT1. PreAMP cocktail was prepared by mixing 12.5μl RT2 PreAMP PCR master mix (Qiagen, UK) per reaction with 7.5μl primer cocktail. cDNA and no-RT controls (RNA which had not undergone reverse transcription) were preamplified as described in section 2.7, with the following modification: only 8 PCR cycles were performed. For qRT-PCR, 1050µl RT2 SYBR Green Mastermix (Qiagen, UK) was mixed with 882µl RNase/DNase-free water and 84µl primer cocktail, then aliquoted into a 96-well PCR plate (24µl/well). 1µl/ well cDNA (in triplicate) or no-RT control (singlet) was added to the plate, which was then sealed, centrifuged and run on the Mx3000P as described in section 2.7, for 50 PCR cycles. Data was analysed using GraphPad Prism version 5 and statistical significance determined by Wilcoxon signed rank test.

2.9 K562 Killing Assays PBMC from six healthy female donors were stained with antibodies and sorted for CD56dim CD27- CD8+ CD57+ and CD56dim CD27- CD8+ CD57- NK cells as described in chapter 2.4, with the following modifications: samples were incubated at room temperature during sorting and sorted into polypropylene FACS tubes containing 200µl culture medium.

1 x 107 K562 cells were thawed from liquid nitrogen on day of experiment, washed and resuspended in 20ml culture medium and incubated at 37ºC, 5% CO2 for 2-4 hours prior to assay.

NK cells were diluted to a concentration of 60,000 cells/ml in culture medium and resuspended thoroughly by repeatedly running buffer down the sides of the tube, then vortexing. K562 cells were centrifuged, resuspended in 10ml culture medium and counted twice with trypan blue for exclusion of dead cells, then diluted to a concentration of 30,000 cells/ml. Experimental wells (100µl K562 + 100µl NK cells), K562 controls (100µl K562 + 100µl culture medium) and NK cell controls (100µl NK cells + 100µl culture medium) were set up in triplicate in a 96-well round-bottomed

103 tissue culture plate (total 1000 K562 and 2000 NK cells per well). Plates were centrifuged (250g, 5 minutes) in order to optimise cell contact, and then incubated for

2 hours at 37°C, 5% CO2. TO-PRO-3 (Invitrogen, UK) was diluted 1in 100 in PBS to produce a working solution, of which 1µl was added per well immediately after incubation. Countbright counting beads (Invitrogen, UK) were vortexed thoroughly and 10µl was added to each well. Contents of each well were mixed thoroughly, transferred to FACS tubes and acquired immediately on the BD FACS Aria II. Data was analysed using FlowJo verion 7.5 and Microsoft Excel. K562, NK cell and bead populations were identified by forward and side scatter and apoptotic cells identified according to uptake of TO-PRO-3 dye, which only enters cells when cell surface membrane integrity is compromised. Cell death was calculated as follows:

Expected number of viable cells = 3000 / (beads added/ beads acquired)

Actual number of viable cells = total K562 events – number of TO-PRO-3+ K562

Raw percentage cell death = 100 x (expected number of viable cells – actual number of viable cells) / expected number of viable cells

Normalised percentage cell death = 100 x (mean experimental cell death – mean control cell death) / (100 – mean control cell death)

2.10 Preparation and Staining of Cells from Line 7 mice

Line 7 mice were scored twice weekly according to their degree of paralysis (see table 2-i). Mice which had reached a clinical score of 2- 3 were selected for sacrifice.

Mice were injected intraperitoneally with 250μl euthatal (Mirion, UK). Following removal of the spleen, mice were perfused with sterile PBS into the left ventricle whilst the heart was still beating, until all blood was removed from circulation. This ensured that leukocytes isolated from the CNS included only those that had crossed the BBB. (This part of the procedure requires a Home Office license and was therefore carried out by Dr. Karen Chu, Dr. Daniel Lowther, Dr. Deborah Chong or Marie-Laure Aknin.) Splenocytes were flushed out into medium using syringe and needle and transferred to 15ml centrifuge tube, then centrifuged (300g, 10 minutes) and resuspended in fresh medium. Cells were kept at room temperature until ready to use, then washed once with PBS.

104

Table 2-ii: Scoring of clinical disease in Line 7 mice

Score Description

0 No clinical signs

1 Limp tail

2 Loss of righting reflex or waddling gait

2.5 Slight hind limb paralysis

3 Paralysis of one hind limb

4 Paralysis of both hind limbs

5 Total limb paralysis

6 Moribund

Brain and spinal cord were dissected out and placed in 1ml Dulbecco’s modified Eagle medium (DMEM) + Glutamax (Invitrogen, UK) with 25μl 100mg/ml collagenase and 10μl 100mg/ml DNase (Sigma-Aldrich, UK). Tissue was chopped up using scissors and further disrupted by repeatedly taking up and releasing with 1ml syringe, then incubated at 37ºC, 5% CO2 for 2 hours. CNS suspension was then passed through a 70μM cell strainer into petri dish. Remaining liquid was gently pushed through strainer using plunger from 1ml syringe. Filtered CNS was then transferred to 15ml tube and centrifuged (300g, 10mins) prior to careful aspiration of supernatant using a stripette. 30% Percoll (Sigma-Aldrich, UK) was made up in DMEM+glutamax and 70% Percoll was made up in PBS. Sample was resuspended in 4ml 30% Percoll and overlayed onto 4ml 70% Percoll, with 3ml Hanks balanced saline solution (Sigma-Aldrich, UK) overlayed on top. Sample was centrifuged at 400g for 30 minutes with no brake. Fatty layer from upper interface was removed using Pasteur pipette and discarded. Using a fresh Pasteur pipette, cell layer from lower interface was transferred to a fresh 15ml centrifuge tube then washed once with PBS.

105

Spleen and CNS samples were mixed with ACK buffer for erythrocyte lysis, centrifuged once (300g, 10 minutes) and washed once with PBS, then resuspended in 1ml PBS (CNS) or 10ml PBS (spleen) before counting with trypan blue for exclusion of dead cells. Cells were incubated with 1:1000 diluted live/dead yellow viability dye (Invitrogen, UK) for 30 minutes at 4ºC, then washed with FACS buffer. Cells were then blocked with 1:1000 diluted α-CD16/32 (e-bioscience, UK) in FACS buffer for 10 minutes at 4ºC, washed and resuspended in FACS buffer, then incubated with antibodies against surface antigens (see table 2-iv) for 30 minutes at 4ºC in V- bottomed plates. Cells were washed twice with FACS buffer then fixed with 1% paraformaldehyde. Samples were stored at 4ºC and acquired on FACS Aria II within 72 hours. Doublet exclusion was performed by gating lymphocyte population on forward scatter-width and live/dead yellow+ apoptotic cells were also excluded.

106

2.11 Buffers and Medium:

FACS buffer:

500ml sterile PBS (Invitrogen, UK) supplemented with 50ml fetal bovine serum (Biosera, UK).

Culture medium:

500ml RPMI 1640 + GlutaMax (Invitrogen, UK) supplemented with 50ml fetal bovine serum (Biosera, UK).

ACK (erythrocyte lysis) buffer:

8.29 g Ammonuim chloride NH4Cl (0.155 mM) Sigma-Aldrich UK

1 g Potassuim Bicarbonate KHCO3 (1 mM) Sigma-Aldrich UK

37.2 mg NA2EDTA (0.1 mM) Sigma-Aldrich UK

in 1 litre of ddH2O.

107

Figure 2-i: Gating strategy for enumeration of NK cell subsets. Lymphocytes were gated based on forward- and side scatter (A) and doublets were excluded by gating out cells with high forward scatter- width (FSC-W) (B). NK cells were defined as CD56+ CD3- cells within the singlet population (C). NK cells were then divided into four populations based on expression of CD56 and CD8 (D). These populations were subdivided according to expression of CD27 and CD57 (E and F). CD56dim CD8- and CD8+ NK cells were each divided into three subpopulations (E), whereas CD56bright CD8- and CD8+ NK cells lacked CD57 expression and were therefore divided into two subpopulations (F). CD27+ CD57+ NK cells were not present amongst any of the aforementioned populations. Gating is shown on PBMC from a healthy control and is representative of gating strategy for all controls, CIS and MS patients.

108

Figure 2-ii: Classification of NK cells into ten subsets. CD56bright NK cells from peripheral blood of both healthy controls and patients did not express CD57 and therefore were classified into just four subsets: CD27+ CD8-, CD27+ CD8+, CD27- CD8- and CD27- CD8+. CD8, CD27 and CD57 were all found on a subpopulation of CD56dim NK cells, however CD27 and CD57 expression were mutually exclusive, therefore CD56dim NK cells were divided into six subsets: CD27+ CD8+, CD27+ CD8-, CD57+ CD8+, CD57+ CD8-, CD27- CD57- CD8+ and CD27- CD57- CD8-. The frequencies of these ten subsets were compared in peripheral blood of healthy controls, CIS patients and MS patients (untreated, interferon-β-treated and Tysabri-treated).

109

Figure 2-iii: Sorted NK cell subsets have high purity. PBMC were stained with antibodies against CD3, CD56, CD8, CD27 and CD57. CD56dim CD27- CD8+ CD57+ and CD56dim CD27- CD8- CD57- NK cell subsets were sorted using a FACS Aria II cell sorter for extraction of RNA and use in killing assays. A minimum of 2000 cells per subset were sorted into FACS buffer and reacquired on the FACS Aria to check purity before commencing cell sorting for assay. Plots are representative of 13 cell sorts performed using PBMC from healthy female donors, with a purity of ≥98%.

110

Table 2-iii: Antibodies for detection of human antigens

Antigen Fluorochrome Clone Isotype Supplier Cat. Number

CD3 APC HIT3a Mouse BD 555342 IgG1, κ

CD3 V500 UCHT1 Mouse BD 561416 IgG1, κ

CD8 PerCP-Cy5.5 SK1 Mouse BD 341050 IgG1, κ

CD27 APC-H7 M-T271 Mouse BD 560222 IgG1, κ

CD56 V450 B159 Mouse BD 560360 IgG1, κ

CD57 FITC NK-1 Mouse BD 555619 IgM, κ

CD57 Alexa Fluor HCD57 Mouse Biolegend 322308 647 IgM, κ

CD95 PE DX2 Mouse BD 555674 IgG1, κ

CD127 PE hIL-7R-M21 Mouse BD 557938 IgG1, κ

IL12RB1 PE 69310 Mouse R&D FAB839P IgG1

IL12RB2 PE 305719 Mouse R&D FAB1959P IgG1

111

NKG2A PE Z199 Mouse BC IM3291U (CD159a) IgG2b

NKG2D PE 1D11 Mouse BD 557940 (CD314) IgG1, κ

NKp30 PE Z25 Mouse BC IM3709 IgG1

NKp44 Alexa Fluor p44-8.1 Mouse BD 558564 647 IgG1, κ

NKp44 PE p44-8.1 Mouse BD 558563 IgG1, κ

NKp46 PE 9E2/NKp46 Mouse BD 557991 IgG1, κ

IL-17 PE eBio64DEC17 Mouse e-bioscience 12-7179 IgG1, κ

IL-22 PE 22URTI Mouse e-bioscience 12-7229 IgG1, κ

IFNγ PE 4S.B3 Mouse e-bioscience 12-7319 IgG1, κ

BC = Beckmann Coulter, BD = Becton Dickinson, R&D = R&D Systems. FITC = fluorescein isothyocyanate, PE= phycoerythrin, APC = allophycocyanin

112

Table 2-iv: Antibodies for detection of murine antigens

Antigen Fluorochrome Clone Isotype Supplier Cat. Number

CD3 FITC 145-2C11 Armenian Beckton 553062 hamster IgG1, κ Dickinson

NKG2D APC CX5 Rat IgG1 e-bioscience 17-5882

NKp46 eFluor 450 29A1.4 Rat IgG2a e-bioscience 48-3351

Table 2-v: Excitation and detection wavelegths of fluorochromes and dead cell dyes measured on FACS Aria II

Fluorochrome or Excitation Detection Bandpass Filter Dye Wavelength (nm) Wavelength (nm) Tolerance

FITC 488 525 50

PE 488 575 25

PerCP-Cy5.5 488 710 50

APC/ Alexa Fluor 633 670 14 647

APC-H7 633 780 60

V450/ efluor450 405 450 40

V500 405 525 50

TO-PRO-3 633 670 14

Live/dead yellow 405 610 20

113

Table 2-vi: Primer list for Th17 for Autoimmunity & Inflammation RT² Profiler PCR Array

Unigene Refseq Symbol Description

Hs.508524 NM_014412 CACYBP Calcyclin binding protein

Hs.72918 NM_002981 CCL1 Chemokine (C-C motif) ligand 1

Hs.303649 NM_002982 CCL2 Chemokine (C-C motif) ligand 2

Hs.75498 NM_004591 CCL20 Chemokine (C-C motif) ligand 20

Hs.534347 NM_002990 CCL22 Chemokine (C-C motif) ligand 22

Hs.251526 NM_006273 CCL7 Chemokine (C-C motif) ligand 7

Hs.156445 NM_000734 CD247 CD247 molecule

Hs.591629 NM_006139 CD28 CD28 molecule

Hs.374990 NM_001773 CD34 CD34 molecule

Hs.504048 NM_000732 CD3D CD3δ molecule (CD3-TCR complex)

Hs.3003 NM_000733 CD3E CD3ε molecule (CD3-TCR complex)

Hs.2259 NM_000073 CD3G CD3γ molecule (CD3-TCR complex)

Hs.631659 NM_000616 CD4 CD4 molecule

Hs.592244 NM_000074 CD40LG CD40 ligand

Hs.85258 NM_001768 CD8A CD8α molecule

CCAAT/enhancer binding protein (C/EBP), Hs.517106 NM_005194 CEBPB beta

Hs.143929 NM_022570 CLEC7A C-type lectin domain family 7, member A

Hs.1349 NM_000758 CSF2 Colony stimulating factor 2 (GM-CSF)

Hs.2233 NM_000759 CSF3 Colony stimulating factor 3 (G-CSF)

Hs.531668 NM_002996 CX3CL1 Chemokine (C-X3-C motif) ligand 1

114

Hs.789 NM_001511 CXCL1 Chemokine (C-X-C motif) ligand 1

Hs.522891 NM_000609 CXCL12 Chemokine (C-X-C motif) ligand 12

Hs.590921 NM_002089 CXCL2 Chemokine (C-X-C motif) ligand 2

Hs.89714 NM_002994 CXCL5 Chemokine (C-X-C motif) ligand 5

Hs.164021 NM_002993 CXCL6 Chemokine (C-X-C motif) ligand 6

Hs.154210 NM_001400 S1PR1 Sphingosine-1-phosphate receptor 1

Hs.247700 NM_014009 FOXP3 Forkhead box P3

Hs.524134 NM_002051 GATA3 GATA binding protein 3

Hs.643447 NM_000201 ICAM1 Intercellular adhesion molecule 1

Hs.56247 NM_012092 ICOS Inducible T-cell co-stimulator

Hs.856 NM_000619 IFNG Interferon, gamma

Hs.193717 NM_000572 IL10 Interleukin 10

Hs.674 NM_002187 IL12B Interleukin 12B (p40)

Hs.567294 NM_005535 IL12RB1 Interleukin 12 receptor, beta 1

Hs.479347 NM_001559 IL12RB2 Interleukin 12 receptor, beta 2

Hs.845 NM_002188 IL13 Interleukin 13

Hs.654378 NM_000585 IL15 Interleukin 15

Hs.41724 NM_002190 IL17A Interleukin 17A

Hs.278911 NM_013278 IL17C Interleukin 17C

Hs.655142 NM_138284 IL17D Interleukin 17D

Hs.272295 NM_052872 IL17F Interleukin 17F

Hs.654970 NM_018725 IL17RB Interleukin 17 receptor B

Hs.129959 NM_032732 IL17RC Interleukin 17 receptor C

115

Hs.150725 NM_017563 IL17RD Interleukin 17 receptor D

Hs.390823 NM_153480 IL17RE Interleukin 17 receptor E

Interleukin 18 (interferon-gamma-inducing Hs.83077 NM_001562 IL18 factor)

Hs.126256 NM_000576 IL1B Interleukin 1, beta

Hs.89679 NM_000586 IL2 Interleukin 2

Hs.567559 NM_021803 IL21 Interleukin 21

Hs.287369 NM_020525 IL22 Interleukin 22

Hs.98309 NM_016584 IL23A Interleukin 23, alpha subunit p19

Hs.677426 NM_144701 IL23R Interleukin 23 receptor

Hs.302036 NM_022789 IL25 Interleukin 25

Hs.528111 NM_145659 IL27 Interleukin 27

Hs.694 NM_000588 IL3 Interleukin 3

Hs.73917 NM_000589 IL4 Interleukin 4

Hs.2247 NM_000879 IL5 Interleukin 5

Hs.654458 NM_000600 IL6 Interleukin 6

Hs.709210 NM_000565 IL6R Interleukin 6 receptor

Hs.591742 NM_002185 IL7R Interleukin 7 receptor

Hs.624 NM_000584 IL8 Interleukin 8

Interferon stimulated exonuclease gene Hs.459265 NM_002201 ISG20 20kDa

Hs.207538 NM_002227 JAK1 Janus kinase 1

Hs.656213 NM_004972 JAK2 Janus kinase 2

Hs.2936 NM_002427 MMP13 Matrix metallopeptidase 13 (collagenase 3)

116

Matrix metallopeptidase 3 (stromelysin 1, Hs.375129 NM_002422 MMP3 progelatinase)

Matrix metallopeptidase 9 (gelatinase B, 92kDa gelatinase, 92kDa type IV Hs.297413 NM_004994 MMP9 collagenase)

Nuclear factor of activated T-cells, Hs.713650 NM_012340 NFATC2 cytoplasmic, calcineurin-dependent 2

Nuclear factor of kappa light polypeptide Hs.654408 NM_003998 NFKB1 gene enhancer in B-cells 1

Hs.256022 NM_005060 RORC RAR-related orphan receptor C (RORγ)

Hs.50640 NM_003745 SOCS1 Suppressor of cytokine signaling 1

Hs.527973 NM_003955 SOCS3 Suppressor of cytokine signaling 3

Signal transducer and activator of transcription 3 (acute-phase response Hs.463059 NM_003150 STAT3 factor)

Signal transducer and activator of Hs.80642 NM_003151 STAT4 transcription 4

Signal transducer and activator of Hs.437058 NM_003152 STAT5A transcription 5A

Signal transducer and activator of Hs.524518 NM_003153 STAT6 transcription 6, interleukin-4 induced

Hs.371720 NM_003177 SYK Spleen tyrosine kinase

Hs.272409 NM_013351 TBX21 T-box 21 (T-bet)

Hs.645227 NM_000660 TGFB1 Transforming growth factor, beta 1

Toll-interleukin 1 receptor (TIR) domain Hs.537126 NM_001039661 TIRAP containing adaptor protein

Hs.174312 NM_138554 TLR4 Toll-like receptor 4

Hs.241570 NM_000594 TNF Tumor necrosis factor

117

Hs.591983 NM_004620 TRAF6 TNF receptor-associated factor 6

Hs.388927 NM_003403 YY1 YY1 transcription factor

Hs.534255 NM_004048 B2M Beta-2-microglobulin

Hs.412707 NM_000194 HPRT1 Hypoxanthine phosphoribosyltransferase 1

Hs.728776 NM_012423 RPL13A Ribosomal protein L13a

Glyceraldehyde-3-phosphate Hs.592355 NM_002046 GAPDH dehydrogenase

Hs.520640 NM_001101 ACTB Actin, beta

Table is adapted from version on SABiosciences website found at: http://www.sabiosciences.com/rt_pcr_product/HTML/PAHS-073A.html

118

Table 2-vii: Age and gender of healthy controls used for comparison of subset frequencies

Healthy Control Number Age Gender

HC1 37 F

HC2 52 M

HC3 52 M

HC4 43 F

HC5 25 F

HC6 40 F

HC7 46 M

HC8 40 M

HC9 31 F

HC10 53 F

HC11 29 F

HC12 60 F

HC13 37 F

HC14 26 F

HC15 29 F

HC16 29 F

119

Table 2-viii: Age and gender of CIS patients

Patient Number Age Gender

CIS1 47 M

CIS2 50 F

CIS3 48 M

CIS4 38 M

CIS5 41 F

CIS6 49 F

CIS7 34 F

CIS8 29 F

CIS9 32 F

CIS10 36 F

CIS11 38 F

CIS12 34 F

120

Table 2-ix: Age, gender and disease status of untreated MS patients

Patient Number Age Gender Relapse or Remision

MSU1 54 F Relapse

MSU2 45 F Relapse

MSU3 35 F Remission

MSU4 29 F Remission

MSU5 63 F Remission

MSU6 37 M Remission

MSU7 53 F Remission

MSU8 40 M Remission

MSU9 44 F Remission

MSU10 52 M Remission

MSU11 47 F Remission

MSU12 55 F Remission

121

Table 2-x: Age, gender and disease status of patients on interferon-β therapy

Patient Number Age Gender Relapse or Remision

MSI1 41 F Relapse

MSI2 36 M Relapse

MSI3 47 F Relapse

MSI4 39 F Relapse

MSI5 55 F Remission

MSI6 43 F Remission

MSI7 26 F Remission

MSI8 37 F Remission

MSI9 57 F Remission

MSI10 37 M Remission

MSI11 48 F Remission

MSI12 40 M Remission

122

Table 2-xi: Age, gender and disease status of Tysabri-treated patients

Patient Number Age Gender Relapse or Remission

MST1 30 F Remission

MST2 35 F Remission

MST3 51 F Remission

MST4 34 M Remission

MST5 50 F Remission

MST6 24 M Remission

MST7 46 F Remission

MST8 34 F Remission

123

Table 2-xii: Summary of age and gender of patient and control groups

Patient or Mean Age Age Range Number of Number of Control Group Males Females

Healthy 39.3 25 - 60 4 12 controls

CIS 39.7 29 - 50 3 9

Untreated 46.2 29 - 63 3 9 RRMS

IFNβ-treated 42.2 26 - 57 3 9 RRMS

Tysabri-treated 38.0 24 - 51 2 6 RRMS

Patient and control groups exhibited no significant difference in the mean or variance in age (determined by one-way ANOVA and Bartlett’s test for equal variance). All groups had a male: female ratio of 1:3.

124

3. RRMS PATIENTS EXHIBIT DIFFERENCES IN FREQUENCIES OF SPECIFIC NK CELL SUBSETS

Given the contradictory findings regarding the role of NK cells in MS and its animal models, I hypothesised that the overall influence of NK cells might be determined by relative frequencies of NK cell subsets with different functions. Several studies have already alluded to involvement of particular types of NK cell in downregulating inflammatory activity during MS. These include the CD56bright subset, which expands during pregnancy and in response to IFNβ and daclizumab therapies and CD95+ NK cells which are reported to have a higher frequency in the blood of MS patients in remission than either patients in relapse or healthy controls (Takahashi et al. 2001, Bielekova et al. 2006, Saraste et al. 2007, Airas et al. 2008, Vandenbark et al. 2009, Martín-Rodríguez et al. 2010, Wynn et al. 2010, Martín-Rodríguez et al. 2011). Killing of activated T cells has been proposed as a mechanism by which CD56bright NK cells regulate autoimmunity (Bielekova et al. 2006, Jiang et al. 2011), though they may also suppress disease by other mechanisms such as release of anti- inflammatory cytokines. NK cells from patients with high expression of CD95 are reported to specifically inhibit IFNγ production but not proliferation of MBP-specific T cells, though it is unclear whether this is by a direct mechanism or via effects on APCs (Takahashi et al. 2004). Though these two subsets appear to be capable of suppressing relapses amongst MS patients, there is no evidence that MS is related to a deficiency in these NK cell populations. It is expansion of these subsets to levels higher than those typical of healthy controls that has a protective role during MS (Takahashi et al. 2001, Vandenbark et al. 2009, Martín-Rodríguez et al. 2011).

One subset of NK cells in which MS patients may have a deficiency is the CD8+ subset. This subset has been reported to have a decreased frequency in both RRMS patients and patients with clinically isolated syndrome (CIS) (De Jager et al. 2008). However, an earlier study suggested that it was CD8- NK cells that were decreased in frequency in untreated MS patients compared to controls (Munschauer et al. 1995).

No-one has so far performed a comprehensive investigation of NK cell subsets in MS. Therefore my aim was to gain a better understanding of changes in NK cell subset frequencies in MS patients compared to healthy controls using a combination of surface markers (as opposed to previous studies which have identified subsets using

125 one marker at a time). In addition to comparing the frequencies of NK cell subsets in peripheral blood of RRMS patients to that of healthy controls, I also studied subset frequencies in CIS patients, individuals who have had a single demyelinating episode and are likely to develop MS in future, in order to give an insight as to whether any imbalance in subset frequencies was present at an early stage of disease. Given that IFNβ has previously been reported to have effects on the frequency of CD56bright NK cells, I was keen to find out more about the effect of this treatment on NK cell subsets. I also investigated whether Tysabri (natalizumab), one of the most successful treatments for highly active MS, affected the frequency of NK cell subsets; its function is normally evaluated only in terms of T cell ingress across the blood brain barrier and there has been relatively little consideration of effects on peripheral innate or adaptive immune phenotype.

The first step was to select suitable markers to include in the panel. In doing this, I sought to include all the most relevant markers, whilst balancing this with the need for adequate resolution of positive and negative (or bright and dim) populations, which may be compromised by spectral overlap when too many markers are used together.

The most widely-studied functional distinction amongst NK cells is that between CD56dim and CD56bright subsets, so this was a natural starting point for my classification, especially given the association of CD56bright NK cells with amelioration of MS in response to disease-modifying therapies (Wynn et al. 2010, Martín-Rodríguez et al. 2011). The contrasting data on changes in CD8+ and CD8- NK cell frequencies in MS patients made this marker my next choice for inclusion in the panel. The differences reported in CD95 expression in Japanese MS patients along with its functional association with IL-5 production and inhibition of T cell IFNγ production (Takahashi et al. 2001, Takahashi et al. 2004) led me also to consider including this marker. Next, I considered markers which have been associated with distinct functional properties in healthy subjects and chose CD27 and CD57 as markers of interest. A reduction in the percentage of CD57+ NK cells has previously been observed in MS patients treated with IFNβ, however it was not clear how these frequencies compared to that of healthy controls (Perini et al. 2000). The association of CD57 with increased maturity and cytotoxic function (Chattopadhyay et al. 2009, Björkström et al. 2010, Lopez-Vergѐs et al. 2010) made this marker an ideal choice.

126

Conversely, CD27 is associated with lower cytotoxic function and enhanced responsiveness to cytokines, properties that have also been associated with the CD56bright subset (Silva et al. 2008, Vossen et al. 2008). Unlike CD56 and most of the other markers used to distinguish human NK cell subsets, CD27 expression also distinguishes NK cell subsets in mice (Hayakawa and Smyth 2006) and might therefore provide an opportunity to link differences observed in MS patients to functional roles in disease models.

Staining for CD95, with two different phycoerythrin-conjugated antibody clones showed that there were not distinct positive and negative populations amongst peripheral blood NK cells, although there were amongst peripheral blood T cells. This rendered it unsuitable for inclusion in my classification. Indeed, one might suggest that close scrutiny of the FACS plots in the paper by Takahashi et al. (2001) does not allow delineation of distinct positive and negative populations, though the data have been interpreted and cited with this interpretation. Therefore I proceeded in designing a panel with the other four markers (CD56, CD8, CD27 and CD57), with CD3 for exclusion of iNKT and NK-like T cells.

3.1 Expression and Co-expression of CD8, CD27 and CD57 on CD56dim and CD56bright NK Cells

Various combinations of fluorochromes were tested in order to combine the markers in such a way that spectral overlap did not prevent identification of the positive and negative populations for CD8 and CD27, which were expressed only at low levels on NK cells in comparison to T cells. Once a suitable panel had been chosen, it was tested on healthy controls in order to evaluate which markers were or were not expressed together and to devise a classification system. Expression of each marker (CD8, CD27 and CD57) on CD56bright and CD56dim subsets and their frequency of co- expression with the other markers is summarised in figure 3-i. CD8 was typically expressed on 30-40% of NK cells, with a slightly higher frequency in the CD56dim compared to the CD56bright population. There did not appear to be any relationship between expression of CD8 and that of either CD27 or CD57. CD27 was expressed on around 10-20% of NK cells and was more frequent amongst CD56bright NK cells. Meanwhile, CD57 was expressed on around 50% of CD56dim NK cells and was absent from the CD56bright population. Expression of CD27 and CD57 was mutually

127 exclusive. Based on this information, a gating strategy was devised which divided NK cells into ten subsets as shown in chapter 2, figures 2-i and 2-ii.

Figure 3-i: Co-expression pattern of NK cell surface markers. Multiparameter flow cytometry was performed on PBMC from ten healthy donors using antibodies directed against CD3, CD56, CD8, CD27 and CD57. Lymphocytes were identified based on forward and side scatter and NK cells were defined as CD56+ CD3- lymphocytes. The mean percentage of cells expressing CD8, CD27 or CD57 (columns) was calculated for all NK cells, for CD56dim and CD56bright subsets and for NK cells expressing each of the other markers (rows). CD57 was expressed only on CD56dim NK cells and was never expressed in combination with CD27.

128

3.2 Confirmation of Differences in Frequency of NK Cell Subsets in RRMS Patients and the Effects of IFNβ Treatment

Before analysing the multiparameter data, I wanted to find out if there was any difference in the overall frequency of NK cells in MS or CIS patients compared to healthy controls.

No significant difference was observed in the frequency of NK cells in MS patients versus controls and neither IFNβ nor Tysabri treatment had a significant effect on NK cell frequency (see figure 3-ii).

Figure 3-ii: Total NK cell frequency does not differ between MS patients and controls and is not significantly influenced by interferon-β or Tysabri therapy. PBMC were isolated from 12 untreated, 12 IFNβ-treated and 8 Tysabri-treated MS patients (with mean ages of 46, 42 and 38 respectively), 12 CIS patients (mean age 40) and 16 healthy controls (mean age 39). All groups were gender matched with a 3:1 female: male ratio. Lymphocytes were gated according to forward and side scatter and doublets excluded. The frequency of NK cells (CD56+ CD3-) within the lymphocyte gate (with mean and SEM) is shown (A) in RRMS and CIS patients compared to healthy controls and (B) in IFNβ- and Tysabri-treated patients compared to untreated RRMS patients. Kruskal Wallis tests indicated no significant difference between MS patients, CIS patients and controls (A) or between treated and untreated MS patients (B).

Next, I wanted to see if I could replicate the data of De Jager et al., who reported fewer CD8+ NK cells in blood of MS patients and to see if IFNβ or Tysabri treatment affected the frequency of CD8+ NK cells.

129

I did not find any significant difference in the frequency of CD8+ NK cells within the whole lymphocyte population (figure 3-iii A), however, amongst the NK cell population, there was a significantly higher proportion of CD8+ cells in peripheral blood of both CIS and RRMS patients compared to healthy controls (figure 3-iii B). This difference was reversed in patients treated with either IFNβ or Tysabri.

Figure 3-iii: Untreated RRMS and CIS patients have a higher proportion of CD8+ compared to CD8- NK cells. PBMC were isolated from 12 untreated, 12 IFNβ-treated and 8 Tysabri-treated MS patients (with mean ages of 46, 42 and 38 respectively), 12 CIS patients (mean age 40) and 16 healthy controls (mean age 39). All groups were gender matched with a 3:1 female: male ratio. Lymphocytes were gated according to forward and side scatter with doublet exclusion and NK cells were identified as CD3- CD56+. The frequency of CD8+ NK cells (with mean and SEM) is shown (A) as a percentage of all lymphocytes and (B) as a percentage of NK cells. Significant differences were determined by Kruskal Wallis test followed by Dunn’s multiple comparison test of RRMS and CIS versus healthy (left) or IFNβ and Tysabri versus no treatment (right). Significant differences are indicated by *(p<0.05) or ** (p<0.01).

130

I then compared the overall frequencies of CD56bright and CD56dim NK cells amongst peripheral blood lymphocytes of untreated and treated MS patients and controls. There were no disease- or treatment-related differences in the overall frequency of CD56dim NK cells (figure 3-iv A), however there were differences in the frequency of CD56bright NK cells (figure 3-iv B). Untreated RRMS patients had a significantly lower frequency of CD56bright NK cells than healthy controls. CIS patients had a frequency between that of healthy controls and MS patients.

Consistent with previous studies, the frequency of CD56bright NK cells was higher in IFNβ-treated patients. However, there was a high level of variability amongst IFNβ- treated patients, with some having similar frequencies to untreated MS patients, while others had frequencies considerably higher than healthy controls. Therefore, I decided to compare the frequency of this subset in IFNβ-treated patients who were in relapse at the time the sample was taken to those who were in remission. Though the difference in frequency of CD56bright NK cells amongst all lymphocytes did not reach statistical significance, the percentage of NK cells which were CD56bright was significantly lower in patients who were in relapse compared to those in remission (figure 3-iv C).

131

Figure 3-iv: Untreated RRMS patients have a reduced frequency of CD56bright NK cells which is reversed by IFNβ treatment. (A) and (B): PBMC were isolated from 12 untreated, 12 IFNβ-treated and 8 Tysabri-treated MS patients (with mean ages of 46, 42 and 38 respectively), 12 CIS patients (mean age 40) and 16 healthy controls (mean age 39) and analysed by flow cytometry. All groups were gender matched with a 3:1 female: male ratio. Lymphocytes were gated according to forward and side scatter and doublets excluded. The frequency of (A) CD56dim NK cells and (B) CD56bright NK cells in MS and CIS patients versus controls (left) and in untreated versus IFNβ or Tysabri-treated pateints (right) is indicated as a percentage of lymphocytes (with mean and SEM). Significant differences were determined by Kruskal Wallis and Dunn’s multiple comparison tests. (C) IFNβ-related expansion of CD56bright NK cells is restricted to patients in remission. Blood samples were taken from age and gender-matched IFNβ-treated patients, four of whom were in relapse and eight of whom were in remission. Frequency of CD56bright NK cells was compared by Mann-Whitney U test. Significant differences ((A), (B) and (C)) are indicated by * (p <0.05), ** (p<0.01) or *** (p<0.001).

132

3.3 Multiparameter Analysis Reveals Differences in NK Cell Subset Frequencies that are not Restricted to the CD56bright Population

I then went on to analyse the frequencies of the ten subsets classified according to expression of CD56, CD8, CD27 and CD57. The CD56bright population was divided into four subsets: CD8- CD27-, CD8- CD27+, CD8+ CD27- and CD8+ CD27+. Three of these subsets had a significantly lower frequency in MS patients compared to controls and the other subset (CD8+ CD27+) had a low frequency in both (figure 3-v A). In both CIS and RRMS patients, the CD56bright CD8- CD27- population was approximately half the size of that in healthy controls. This subset was the only subset that was significantly increased in IFNβ-treated patients (figure 3-v B). However, all four CD56bright subsets had significantly higher frequency in IFNβ-treated patients who were in remission compared to those who were in relapse (figure 3-v C).

The CD56dim subset was divided into six subsets, according to expression of CD8, CD27 and CD57 as described in chapter 2 (see figure 2-ii). The CD56dim CD8+ CD27- CD57+ subset constituted a significantly higher proportion of NK cells in MS patients compared to healthy controls, with an intermediate frequency in CIS patients (figure 3-vi). The CD56dim CD8- CD27- CD57- subset had a slightly lower frequency in MS patients than controls. The Kruskall-Wallis test indicated a significant difference for this subset, however the post-test did not identify a difference between either patient group and controls once corrected for multiple comparisons. This indicates that there is likely to be some difference, but a larger sample size is needed to elucidate whether this subset is reduced in frequency in MS patients, CIS patients or both.

This subset is however significantly increased in IFNβ-treated compared to untreated MS patients (figure 3-vii). Conversely, IFNβ significantly decreased the frequency of CD56dim CD8+ CD57+, CD8+ CD27+ and CD8- CD27+ NK cell subsets. Tysabri- treated patients also had a significantly lower frequency of the CD56dim CD8+ CD57+ subset. Other subsets were not significantly altered in Tysabri-treated individuals, but this may have been due to the relatively small sample size. Unlike CD56bright NK cells, none of the CD56dim subsets had significantly different frequencies in IFNβ- treated patients in relapse compared to those in remission (figure 3-viii).

133

(A) 10 Healthy 8 CIS RRMS 6 ** ** 4 * *

(% of NK cells) NK of (% 2 Subset frequency Subset 0 8- 27+ 8- 27- 8+27+ 8+27-

(B) 15 Untreated * Interferon- 10 Tysabri

5

(% of NK cells) NK of (% Subset frequency Subset 0 8- 27+ 8- 27- 8+27+ 8+27-

(C) 15 * Relapse Remission 10

* *

5 **

(% of NK cells) NK of (% Subset frequency Subset 0 8- 27+ 8- 27- 8+27+ 8+27-

Figure 3-v: CD56bright subset frequencies are altered in MS patients. PBMC were isolated from 12 untreated, 12 IFNβ-treated and 8 Tysabri-treated MS patients (with mean ages of 46, 42 and 38 respectively), 12 CIS patients (mean age 40) and 16 healthy controls (mean age 39). All groups were gender matched with a 3:1 female: male ratio. CD56bright NK cells were classified into four subsets (CD8- CD27+, CD8- CD27-, CD8+ CD27+ and CD8+ CD27-) and flow cytometry was performed to compare frequencies of these subsets in peripheral blood. Mean and SEM were plotted and significant differences in the frequency of each subset in (A) RRMS and CIS patients compared to healthy controls and (B) IFN-β and Tysabri-treated compared to untreated RRMS patients, were determined by Kruskal-Wallis test followed by Dunn’s multiple comparison test. (C) Significant differences between age and gender-matched IFNβ-treated patients in relapse (n=4) and those in remission (n=8) were identified by Mann-Whitney U test. Significant differences are denoted as: * p<0.05 or ** p<0.01.

134

Figure 3-vi: MS patients have a higher frequency of CD56dim CD8+ CD27- CD57+ NK cells. PBMC were isolated from twelve untreated RRMS patients with a mean age of 46, twelve CIS patients with a mean age of 40 and sixteen healthy controls with a mean age of 39 and analysed by flow cytometry. All groups were gender matched with a 3:1 female: male ratio. CD56dim NK cells were classified into six subsets based on expression of CD8, CD27 and CD57, as described in chapter 2. Frequency of each of these subsets in patients and controls is shown with mean and SEM. Statistical significance was determined by Kruskal Wallis test followed by Dunn’s multiple comparison test. ** indicates a statistically significant difference (p<0.01) compared to healthy controls. Kruskal Wallis test also indicated a significant difference (p<0.05) in the frequency of the CD56dim CD8- CD27- CD57- subset, but this was not confirmed by Dunn’s multiple comparison test.

135

Figure 3-vii: Interferon-β alters frequencies of several CD56dim NK cell subsets. PBMC were isolated from 12 untreated RRMS patients with a mean age of 46, 12 IFNβ-treated patients with a mean age of 42 and 8 Tysabri-treated patients with a mean age of 38 and analysed by flow cytometry. All groups were gender matched with a 3:1 female: male ratio. CD56dim NK cells were divided into six subsets based on expression of CD8, CD27 and CD57, as described in chapter 2. Frequency of each of these subsets in untreated and IFNβ- or Tysabri-treated patients is shown with mean and SEM. Statistically significant differences were determined by Kruskal Wallis test followed by Dunn’s multiple comparison test and are indicated by * (p<0.05), ** (p<0.01) or *** (p<0.001).

136

Figure 3-viii: Frequencies of CD56dim subsets are not significantly different in relapse compared to remission. PBMC from interferon-β-treated patients were incubated with antibodies against CD3, CD56, CD8, CD27 and CD57 and analysed by flow cytometry. CD56dim NK cells were divided into six subsets based on expression of CD8, CD27 and CD57, as described in chapter 2. Frequency of these subsets (mean and SEM) is shown in age and gender-matched IFNβ-treated patients experiencing disease exacerbation (relapse) (n=4) and those in remission (n=8). Mann-Whitney U test detected no significant difference in frequency of any of the six subsets.

137

3.4 Subset Frequencies in Healthy Adults do not Differ Significantly in Relation to Age or Gender

Age-related changes in the frequency of NK cells and of particular NK cell subsets have been reported previously in peripheral blood of healthy donors. Both the absolute number and percentage of NK cells in peripheral blood increase with age (Borrego et al. 1999, Le Garff-Tavernier et al. 2010). The proportion of NK cell subsets also appears to change over time, with an accumulation of NK cell populations that are thought to be more developmentally mature in elderly subjects. For instance, the ratio of CD56dim to CD56bright NK cells is reported to increase with age, due to a decline in absolute numbers of CD56bright NK cells (Chidrawar et al. 2006, Le Garff-Tavernier et al. 2010) and an increase in absolute frequency of the CD56dim subset (Borrego et al. 1999). A negative correlation between CD56bright NK cell frequency and age has also been demonstrated in MS patients (Martín-Rodríguez et al. 2010). Older individuals also have fewer NK cells expressing CD94 and NKG2A and more expressing KIR (Lutz et al. 2005). CD57, which has also been associated with increased maturity of NK cells, was initially found to increase significantly with age amongst T cells but not NK cells (Borrego et al. 1999), however a subsequent study suggested that NK cells also contained a higher proportion of CD57+ cells in older subjects (Le Garff-Tavernier et al. 2010). These associations were typically based on comparisons of elderly subjects (60+ or 80+) with adults aged 18-60, and do not demonstrate age-related differences amongst the under 60s (Borrego et al. 1999, Chidrawar et al. 2006, Le Garff-Tavernier et al. 2010).

Given these previous associations of subset frequencies with age and the inferred relationship with developmental maturity, I was curious to know whether my subsets of interest were altered across the age range that was included in my study and whether any slight differences in the age distribution of patient groups were likely to affect my results. Though there did appear to be a slight decrease in CD56bright NK cell frequency with age in healthy subjects, this was not significant over the age range of 24-60 that was examined (figure 3-ix B).There was also no change in the overall frequency of NK cells in the lymphocyte population and no change in either the CD56dim CD8+ CD57+ subset or the CD56dim CD8- CD27- CD57- subset.

138

(A) (B)

Figure 3-ix: No significant gender- or age-related differences in the frequency of NK cells or NK cell subsets were detected. Overall NK cell frequency and frequency of CD56bright, CD56dim CD8- CD27- CD57- and CD56dim CD8+ CD27- CD57+ NK cell subsets in peripheral blood lymphocytes was compared (A) in male versus female age-matched healthy controls (n=8) and (B) in healthy individuals of different ages (age 24-60, n=16). The relationship between age and frequency of CD56bright NK cells was tested by linear regression, but the trend was not statistically significant (R2 = 0.1586, 95% CI of slope = -0.6025 to 0.08324, p = 0.1266).

139

Given the higher incidence of MS amongst females compared to males, I also wanted to know if there was any difference between males and females in the frequency of these subsets, which might potentially affect susceptibility to MS. However, no significant differences were observed between male and female healthy controls (figure 3-ix A).

3.5 Discussion

Though I did not observe differences in the overall frequency of NK cells in peripheral blood of MS patients in relation to either disease or therapy, a number of differences were observed in frequencies of particular subsets of NK cells, consistent with the hypothesis that an imbalance in the proportions of regulatory versus pro- inflammatory NK cell subsets might contribute to MS pathogenesis and that changes in this balance may be an important mechanism of action of MS therapies such as Tysabri and IFNβ, or biomarkers of their efficacy.

First, I have confirmed the observation that CD56bright NK cells are expanded in IFNβ- treated patients (Saraste et al. 2007, Vandenbark et al. 2009, Martín-Rodríguez et al. 2010, Martín-Rodríguez et al. 2011) and extended this to show that this expansion of CD56bright NK cells is only observed amongst patients who are in remission. I also noticed that those IFNβ-treated patients with stable MS (who had not had any relapses in the last year) had a particularly high frequency of CD56bright NK cells, however the number of patients was not sufficient to perform a gender-matched statistical comparison. This is consistent with a previous report that CD56bright NK cells are preferentially expanded in patients who show clinical responses to IFNβ therapy (Martín-Rodríguez et al. 2011). Contrary to previous reports, which implied that untreated MS patients had a similar frequency of CD56bright NK cells to healthy controls (Vandenbark et al. 2009, Lünemann et al. 2011, Martín-Rodríguez et al. 2011), I found that there was a statistically significant decrease in the frequency of the CD56bright subset both as a percentage of NK cells and as a percentage of lymphocytes in untreated RRMS patients. A decreased ratio of CD56bright: CD56dim NK cells was also observed in CIS patients, suggesting that changes in subset frequencies occur early in disease. Therefore, it is possible that alterations in NK cell subset frequencies might have a role in initiation of disease, though we do not have direct evidence of this. It would be interesting to follow up these CIS patients over time and see whether

140 there is any relationship between NK cell subset frequencies at this early stage of disease and how rapidly they progress to MS, if at all.

In contrast to De Jager et al., I did not find any significant decrease in the percentage of CD8+ NK cells in the peripheral blood lymphocyte population of either CIS or RRMS patients compared to healthy controls. Although my sample size was smaller, the discrepancy between the studies is not likely to be due to lack of statistical power, since the proportion of NK cells that expressed CD8 in my study was in fact higher in both CIS and MS patients than in controls. I also observed a reduction in the frequency of NK cells expressing CD8 in both IFNβ- and Tysabri-treated patients.

However, the aim of my study was to go beyond simply confirming or refuting prior observations regarding the frequency of single-marker-defined NK cell subsets such as CD56bright or CD8+. After elucidating the relationship in expression of key NK cell markers CD8, CD27 and CD57, I devised a panel that enabled me to dissect NK cells into ten different subsets, and compared the frequency of these ten subsets in untreated MS and CIS patients to that of healthy controls as well as patients on therapy. When the CD56bright population was subdivided into four subsets based on expression of CD8 and CD27, three of these subsets were found to constitute a significantly smaller proportion of NK cells in RRMS patients compared to healthy controls. The two CD56bright CD8- subsets both had a mean frequency in MS patients of approximately half that of healthy controls and the CD8- CD27- subset also had a significantly lower frequency in CIS patients. This subset is particularly interesting, since it was also substantially and significantly expanded in IFNβ-treated patients. Though treatment-related differences in frequency of the other CD56bright subsets did not reach statistical significance, the trend across all CD56bright subpopulations was broadly similar and all four were significantly higher in IFNβ-treated patients who were in remission compared to those in relapse. Therefore, all these subsets might play a role in regulating disease activity during IFNβ treatment and possibly also protect against development of MS.

The overall percentage of CD56dim NK cells within the lymphocyte population was not significantly different in MS patients compared to controls and was not altered in relation to treatment. However, when the CD56dim subset was subdivided according to expression of CD8, CD27 and CD57, some differences were observed. The most

141 interesting subset was the CD56dim CD8+ CD27- CD57+ subset, which represented a significantly higher proportion of NK cells in untreated RRMS patients than in healthy controls. The frequency of this subset was also significantly lower in both IFNβ- and Tysabri-treated patients. This suggests that therapies for MS may not only boost numbers of immunoregulatory CD56bright NK cells, but also reduce the frequency of NK cell subsets that are associated with inflammation. Naturally, this data cannot confirm whether or not this subset is actively contributing to pathogenesis or whether its expansion in MS patients is an inconsequential by-product of inflammation. However, it suggests that the function of this particular subset warrants further investigation. Since CD57 has been associated with increased cytotoxic function (Chattopadhyay et al. 2009, Lopez-Vergѐs et al. 2010) and CD8 with resistance to activation-induced cell death (Addison et al. 2005), one might hypothesise that the CD56dim CD8+ CD57+ subset directly participates in killing of neurons and oligodendrocytes in the CNS and/or contributes to inflammation by killing of Tregs.

IFNβ treatment has striking effects on the proportions of NK cells subsets. IFNβ- treated patients have a significantly lower percentage not only of CD56dim CD8+ CD57+ NK cells, but also of CD56dim CD8+ CD27+ and CD8- CD27+ NK cells. Meanwhile, the CD56dim subset lacking all three markers (CD8- CD27- CD57-) has a significantly higher frequency in IFNβ-treated patients. The latter subset had a slightly lower frequency in untreated patients compared to healthy controls, suggesting that its frequency is regulated in the opposite manner to the CD8+ CD57+ subset and I hypothesised that this subset might have functional characteristics more similar to that of the CD56bright subset and also play a role in suppressing inflammation. However, it is not clear if these changes in NK cell subset frequencies are a primary effect of IFNβ treatment that may be responsible for some of the changes observed in other leukocyte populations, or whether direct effects of IFNβ on other cell populations lead to secondary effects on NK cells. Therefore it would be interesting to study patients longitudinally from the point at which they commence IFNβ treatment, in order to determine whether changes in the phenotype and frequency of NK cells and other leukocytes, including T cells and APCs, occur in a particular order.

142

Analysis of differences between untreated and treated patients is difficult since patients with more active MS are more likely to be given therapy, therefore the IFNβ- treated group contained a higher number of subjects who were in relapse than the untreated group. However, the differences observed between IFNβ-treated and untreated patients do not appear to be related to disease activity, since IFNβ-treated patients in relapse had subset frequencies that were either closer to that of healthy controls (as was the case for CD56bright subsets) or the same as IFNβ-treated patients in remission (CD56dim subsets). Therefore, if anything, the relative lack of untreated patients in relapse may have made the effects of therapy appear less substantial than they would if patients were studied longitudinally. The difficulty in obtaining samples from untreated patients who were in relapse may also have hindered identification of differences between untreated MS patients and healthy controls. For instance, as mentioned, one of the CD56dim subsets appeared to be reduced in MS patients but this did not quite reach statistical significance in the post-hoc test. Ideally, the analysis would be extended to increase overall sample size as well as to include more patients in relapse and with highly active disease. This would also allow comparison of subset frequencies between untreated patients with active versus benign MS and relapse versus remission. It would be interesting to see if the same differences between relapse and remission observed in IFNβ-treated patients also apply to untreated patients.

It would also be interesting to investigate the mechanism by which IFNβ alters NK cell subset frequency. Expansion of the CD56bright subset, for instance, might be related to IL-2. IFNβ downregualtes IL2Rα expression on lymphocytes both in vitro and in vivo (Noronha et al. 1993, Leppert et al. 1996, Genç et al. 1997). Blocking of IL-2Rα by daclizumab has been shown to expand the CD56bright subset by reducing binding of IL-2 to receptors on the T cells that produced it, leaving more IL-2 available for binding to IL-2 receptors on CD56bright NK cells (Bielekova et al. 2006, Martin et al. 2010). Therefore IL2Rα downregulation by IFNβ might have a similar effect. Expansion of the CD56bright subset in daclizumab-treated patients is reportedly much more robust than in IFNβ-treated patients (Bielekova et al. 2006, Wynn et al. 2010), which might reflect more efficient blocking of IL-2/ IL-2Rα interactions than is mediated by IFNβ.

143

Tysabri treatment also signficantly reduced the frequency of CD56dim CD8+ CD57+ NK cells and may have effects on other subsets which were not statistically significant in this small sample. An increase in the total frequency of NK cells has previously been observed in peripheral blood of Tysabri-treated patients (Putzki et al. 2010). Alterations seen in the blood of Tysabri patients are more difficult to interpret than changes related to IFNβ therapy. Tysabri’s main mechanism of action is thought to be preventing lymphocytes from entering the CNS by blocking and downregulating the VLA-4 integrin required for adhesion to endothelial cells of the BBB (Yednock et al. 1992, Baron et al.1993, Carrithers et al. 2000, Kivisäkk et al. 2009, Putzki et al. 2010, Harrer et al. 2011, Gan et al. 2012). VLA-4 is expressed on NK cells and α-VLA4 has been found to suppress NK infiltration of the CNS during EAE (Gismondi et al. 1991, Gan et al. 2012). Therefore, changes in NK cell frequencies in peripheral blood could be caused either by an overall expansion or contraction of a particular subset or by maintenance of cell types in the blood that would otherwise have migrated to the CNS. The fact that the only subset for which there was a significant difference in Tysabri-treated patients (CD56dim CD8+ CD57+) was decreased in frequency in peripheral blood means that this difference is not likely to be due to effects on the migration of the CD56dim CD8+ CD57+ subset itself. However, the alteration in the relative frequency of this subset amongst NK cells and lymphocytes could in theory be due to an increase in peripheral blood numbers of other NK cells and leukocytes which are no longer able to migrate to the CNS. Therefore it would be useful to know if the absolute numbers of each NK cell subset (i.e. per μl of blood) were altered in peripheral blood, but we did not have the facility to measure this. It would also be of particular interest to see if the CD56dim CD8+ CD57+ subset and others are present in the CSF of RRMS patients, whether their frequency in CSF is related to active disease or relapse and whether disease-modifying therapies, especially Tysabri affect CSF frequencies of NK cell subsets.

In summary, I have identified significant changes in the frequencies of specific subsets of NK cells in MS patients compared to healthy controls, consistent with the hypothesis that MS patients have an imbalance in the relative frequencies of regulatory and pro-inflammatory NK cell subsets. These differences include an increased frequency of CD56dim CD8+ CD57+ NK cells and a reduced frequency of several CD56bright subsets, in particular the CD8- CD27- subset, which were reversed

144 by IFNβ therapy. Two subsets of CD56dim NK cells (CD56dim CD8+ CD27- CD57+ and CD56dim CD8- CD27- CD57-) were selected for follow-up in order to identify differences that would shed light on the mechanisms by which they might differentially modulate inflammation.

145

4. TRANSCRIPTIONAL AND FUNCTIONAL DIFFERENCES BETWEEN CD56DIM CD8- CD27- CD57- AND CD56DIM CD8+ CD27- CD57+ NK CELLS

Having identified subsets which had altered frequencies in MS patients compared to healthy controls, I sought to learn more about the function of these NK cell subsets. The CD56bright subset, which had reduced frequency in untreated MS patients and was increased amongst IFNβ-treated patients, has previously been well characterised and within the CD56bright population all four subsets showed the same trends in frequency, albeit to different extents. Therefore characterisation of CD56bright subsets was not a priority. However, within the CD56dim NK cell population, there were subsets that appeared to have a reciprocal relationship as biomarkers of disease state: the CD56dim CD8+ CD57+ subset was identified as a disease-associated subset, the frequency of which was reduced by IFNβ treatment. CD56dim CD27+ NK cell frequency was also reduced in IFNβ-treated patients, however there was no evidence that MS patients had a higher frequency than healthy controls and the CD27+ subsets represent a relatively small proportion of NK cells, making it difficult to isolate sufficient cells for functional analysis. Conversely, one subset of CD56dim NK cells, lacking CD8, CD27 and CD57, had a significantly higher frequency in IFNβ-treated patients. This subset also appeared to have a slightly lower frequency in MS patients than controls. Since the frequencies of CD56dim CD27- CD8+ CD57+ and CD56dim CD27- CD8- CD57- subsets appear to have a reciprocal relationship in disease, I decided to compare the function and transcriptional profile of these two subsets. Both these subsets constitute around 20% of NK cells in healthy controls, a sufficient proportion to enable isolation of these subsets for qRT-PCR and killing assays.

CD57+ NK cells are more likely to express KIR and have a decreased proliferative response to IL-2, therefore the CD8+ CD57+ subset is likely to represent a more mature effector population (Björkström et al. 2010, Lopez-Vergѐs et al. 2010). CD57+ NK cells are reported to be more effective mediators of ADCC and express higher levels of granzymes and perforin (Chattopadhyay et al. 2009, Lopez-Vergѐs et al. 2010). There is limited data on the functional differences between CD8+ and CD8- NK cells, however the CD8+ subset is reported to be more effective at killing target cells over time as it undergoes a lower rate of activation-induced apoptosis (Addison

146 et al. 2005). Therefore the CD56dim CD8+ CD57+ subset would be expected to have greater cytotoxic function.

The function of the CD8- CD27- CD57- subset is harder to predict, given that this subset lacks not only CD57, associated with cytotoxic function, but also CD27, associated with enhanced cytokine production. It is possible that this subset represents a population of NK cells with limited functional capacity and that an increased frequency of this subset might be protective purely in the sense that there is a lower proportion of pathogenic effector NK cells. Therefore I was keen to find out whether this subset expressed genes that might indicate a distinct function, for instance in immunosuppression. The frequency of the CD56dim CD27- CD8- CD57- subset showed a similar pattern to that of the CD56bright subset. Therefore I hypothesised that the function, transcriptional profile and expression of activating, inhibitory and cytokine receptors in the CD56dim CD8- CD27- CD57- subset might be closer to that of the CD56bright subset than the CD56dim CD8+ CD57+ subset. As discussed in chapter 1, a number of functional and developmental intermediates have been described between the CD56bright and CD56dim populations (Romagnani et al. 2007, Takahashi et al. 2007, Juelke et al. 2010, Béziat et al. 2011, Hamann et al. 2011) and it is possible that the CD56dim CD8- CD27- CD57- subset also falls into this category.

4.1 Identification of Transcriptional Differences between CD8- CD27- CD57- and CD8+ CD57+ CD56dim NK Cell Subsets

In order to identify transcriptional differences between the two subsets, I chose a qRT-PCR array (SABiosciences Human Th17 for Autoimmunity and Inflammation) that contained primers for a range of pro-inflammatory and anti-inflammatory cytokine and chemokine genes, cytokine receptors, signalling molecules and matrix metalloproteinases (expression of which might aid migration into the CNS parenchyma and contribute to neuronal damage). All the key cytokines known to be produced by NK cells were included in this array, including IFNγ, TNFα, GM-CSF, IL-5, IL-13 and IL-10. CD56dim CD8- CD27- CD57- and CD56dim CD8- CD27- CD57+ NK cells were isolated by FACS for RNA extraction and qRT-PCR analysis. Expression levels for each of these genes are summarised in tables 4-(i-iv).

147

Table 4-i: Cytokine and chemokine expression determined by qRT-PCR array

Gene Expression in CD8- CD57- Expression in CD8+ CD57+ CCL1 - - CCL2 +/- +/- CCL7 +/- - CCL20 +/- +/- CCL22 +/- +/- CSF2 (GM-CSF) +/- + CSF3 +/- +/- CX3CL1 - +/- CXCL1 + ++ CXCL2 +/- + CXCL5 + +/- CXCL6 +/- +/- CXCL12 - +/- IFNγ +++ +++ IL-1b +/- ++ IL-2 +/- - IL-3 - - IL-4 - +/- IL-5 +/- +/- IL-6 - - IL-8 +++ +++ IL-10 + + IL-12B - - IL-13 +/- +/- IL-15 ++ ++ IL-17A - - IL-17C ++ ++ IL-17F +/- - IL-18 +++ +++ IL-21 - - IL-22 - - IL-23A +/- +/- IL-25 +/- +/- IL-27 +/- +/- TGF-β +++ +++ TNF-α +++ +++ PBMC were isolated from seven healthy female donors. cDNA from FACS-sorted NK cell subsets was analysed by qRT-PCR array.Genes with a mean Ct value of <25 are defined as high expression (+++), those with mean Ct 25-30 as medium (++), Ct of 30-35 low (+) if detected in all 7 samples, (+/-) if detected in 2-6 samples or (-) if detected in less than 2 samples.

148

Table 4-ii: Expression of cell surface receptors and adaptor molecules determined by qRT-PCR array

Gene Expression in CD8- CD57- Expression in CD8+ CD57+

CD3D ++ +++ CD3E +++ +++ CD3G ++ +++ CD247 +++ +++ CD4 ++ ++ CD8A ++ +++ CD28 - - CD34 - - CD40L ++ ++ S1PR1 +++ +++ IL12RB1 +++ +++ IL12RB2 +++ +++ IL17RB ++ ++ IL17RC +/- +/- IL17RD - +/- IL17RE +/- +/- IL23R +/- - IL6R ++ ++ IL7R ++ ++ TLR4 ++ ++ ICAM1 +++ +++ ICOS + + PBMC were isolated from seven healthy female donors. cDNA from FACS-sorted NK cell subsets was analysed by qRT-PCR array.Genes with a mean Ct value of <25 are defined as high expression (+++), those with mean Ct 25-30 as medium (++), Ct of 30-35 low (+) if detected in all 7 samples, (+/-) if detected in 2-6 samples or (-) if detected in less than 2 samples.

149

Table 4-iii: Expression of intracellular signalling molecules and transcription factors determined by qRT-PCR array

Gene Expression in CD8- CD57- Expression in CD8+ CD57+

CACYBP +++ +++ CEBPB +++ +++ CLEC7A ++ ++ FOXP3 +/- +/- GATA3 ++ ++ JAK1 +++ +++ JAK2 +++ +++ NFATC2 +++ +++ NFKB1 +++ +++ RORC ++ + SOCS1 +++ +++ SOCS3 ++ ++ STAT3 +++ +++ STAT4 +++ +++ STAT5A +++ +++ STAT6 +++ +++ SYK ++ ++ TBX21 +++ +++ TIRAP ++ ++ TRAF6 +++ +++ YY1 +++ +++ PBMC were isolated from seven healthy female donors. cDNA from FACS-sorted NK cell subsets was analysed by qRT-PCR array. Genes with a mean Ct value of <25 are defined as high expression (+++), those with mean Ct 25-30 as medium (++), Ct of 30-35 low (+) if detected in all 7 samples, (+/-) if detected in 2-6 samples or (-) if detected in less than 2 samples.

150

Table 4-iv: Expression of other genes determined by qRT-PCR array

Gene Expression in CD8- CD57- Expression in CD8+ CD57+

ISG20 +++ +++ MMP3 - - MMP9 ++ ++ MMP13 - - PBMC were isolated from seven healthy female donors. cDNA from FACS-sorted NK cell subsets was analysed by qRT-PCR array.Genes with a mean Ct value of <25 are defined as high expression (+++), those with mean Ct 25-30 as medium (++), Ct of 30-35 low (+) if detected in all 7 samples, (+/-) if detected in 2-6 samples or (-) if detected in less than 2 samples.

Data was uploaded to the SABiosciences data analysis portal for calculation of fold changes in gene expression in the CD56dim CD8+ CD57+ subset compared to the CD56dim CD8- CD27- CD57- subset and for generation of a clustergram showing relationships in expression of different genes (figure 4-i). Genes that had a fold change of at least +/-1.2 in the CD8+ CD57+ subset compared to the CD8- CD57- subset were subjected to statistical analysis. Genes which had significantly higher expression in the CD8+ CD57+ subset are summarised in table 4-v. Genes which had significantly higher expression in the CD8- CD57- subset are summarised in table 4- vi.

151

Figure 4-i: Clusters of co-regulated genes are associated with CD56dim CD8- CD27- CD57- and CD8+ CD27- CD57+ NK cell subsets. PBMC were isolated from seven healthy female donors. cDNA derived from FACS-sorted NK cell subsets was analysed by qRT-PCR array. Clustergram shows all genes that were consistently expressed in at least one subset (see tables 4-(i-iv)). Genes are clustered according to similar relative expression levels in each sample. Upper box (purple) shows a cluster of co-expressed genes with higher expression in CD56dim CD8- CD27- CD57- NK cells. Lower box (blue) shows a cluster of co-expressed genes with higher expression in CD56dim CD8+ CD57+ NK cells.

152

Table 4-v: Genes with significantly higher expression in the CD56dim CD8+ CD57+ NK cell subset

Gene Fold change CD8A 49.08 CD3D 3.33 CD3G 2.70 SOCS3 1.54 IFNγ 1.54 IL12RB1 1.42 S1PR1 1.38 NFATC2 1.37 STAT3 1.36 TBX21 (T-bet) 1.34 CD247 (CD3ζ) 1.32 PBMC were isolated from seven healthy female donors. cDNA derived from FACS-sorted NK cell subsets was analysed by qRT-PCR array. All genes shown were significantly different (p<0.05) as determined by Wilcoxon matched pairs test, which was performed for all genes with a mean fold change of at least 1.2 as determined by ΔΔCt method with normalisation to four housekeeping genes.

Table 4-vi: Genes with significantly higher expression in the CD56dim CD8- CD27- CD57- NK cell subset

Gene Fold change RORC 5.60 IL7R 4.06 IL12RB2 2.53 IL-18 1.55 ICAM1 1.30 NFKB1 1.21 PBMC were isolated from seven healthy female donors. cDNA derived from FACS-sorted NK cell subsets was analysed by qRT-PCR array. All genes shown were significantly different (p<0.05) as determined by Wilcoxon matched pairs test, which was performed for all genes with a mean fold change of at least 1.2 as determined by ΔΔCt method with normalisation to four housekeeping genes.

153

The biggest fold change was in expression of RORC, the hallmark transcription factor associated with driving the expression of IL-17 and IL-22 (Manel et al. 2008, Cella et al. 2009, Cupedo et al. 2009), which had significantly higher expression in the CD8- CD57- subset. NFκB, a key transcriptional mediator in response to certain cytokines and TLR stimulation (Pahl 1999), was also modestly but significantly increased in the CD8- CD57- subset. Conversely STAT3, T-bet and NFATC2 (nuclear factor of activated T cells, cytoplasmic 2) transcription factors, all of which upregulate and/or are upregulated by IFNγ (Lighvani et al. 2001, Szabo et al. 2002, Harris et al. 2005, Macian 2005, Bluyssen et al. 2010), had higher expression in the CD8+ CD57+ subset. In line with this, IFNγ itself was also expressed at a higher level in this subset, as was suppressor of cytokine signalling 3 (SOCS3), which is activated by STAT3 and mediates a negative feedback loop in several cytokine-induced signalling pathways including IFNγ (Bluyssen et al. 2010). However, IL-18, a cytokine which co- stimulates IFNγ production (Fehniger et al. 1999) was expressed at a higher level in the CD8- CD57- subset. IL12RB1, the receptor chain shared by the IL-12 and IL-23 receptors was expressed at a higher level in the CD8+ CD57+ subset, whereas the IL12RB2 chain, unique to the IL-12 receptor, was expressed at a higher level in the CD8- CD57- subset, as was IL7R.

There were also significant differences in expression of adhesion molecules, with S1PR1 expressed preferentially in the CD8+ CD57+ subset and ICAM1 expressed at higher levels in the CD8- CD57- subset. Several TCR chains (CD3 δ, γ and ζ) were also upregulated in the CD8+ CD57+ subset compared to the CD8- CD57- subset.

In addition to the genes listed in table 4-v, IL-23A also had much higher expression in the CD8+ CD57+ NK cell subset, however the Ct value was above the cut-off value of 35 in one of the donors. Statistical comparison was performed for the remaining six donors but there was a high level of variability and overall the difference between the two subsets was not significant.

The PCR array contained a control with primers for sensitive detection of genomic DNA, which confirmed that in each sample genomic DNA contamination was either absent or sufficiently low not to affect the results when using a Ct of 35 as a cut-off point for a positive result. However, it is still preferable to confirm results with a no reverse transcription control and I also wanted to see if a significant difference in IL23A expression could be detected when Ct values above 35 were included in the

154 analysis. Therefore individual qRT-PCR assays for selected genes were carried out in triplicate to confirm differences in gene expression in the context of a separate no reverse transcription control for each gene.

Individual qRT-PCR assays confirmed significantly higher expression of IFNγ, IL- 23A, TBX21 (T-bet), IL12RB1 and S1PR1 in the CD8+ CD57+ subset (figure 4-ii). Significantly higher expression of RORC and IL12RB2 in the CD8- CD57- subset was also confirmed. Although expression of IL7R was again higher in the CD8- CD57- subset, this did not reach statistical significance in the individual qRT-PCR assay.

155

IFN IL23A 0.025 0.6 * * 0.020

0.4

0.015

Ct

Ct

-

- 2

2 0.010 0.2 0.005

0.0 0.000 8- 57- 8+ 57+ 8- 57- 8+ 57+

Tbx21 RORC 15 0.015 *

10 0.010

Ct

Ct

-

-

2 2 5 0.005 * *

0 0.000 8- 57- 8+ 57+ 8- 57- 8+ 57+

IL12RB1 IL12RB2 3 1.0 * 0.8 2

0.6

Ct

Ct

-

- 2 2 0.4 * 1 0.2

0 0.0 8- 57- 8+ 57+ 8- 57- 8+ 57+

S1PR1 IL7R 6 0.15 *

4 0.10

Ct

Ct

-

-

2 2 2 0.05

0 0.00 - - + + 8 57 8 57 8- 57- 8+ 57+

Figure 4-ii: Confirmation of transcriptional differences by individual qRT-PCR assays. The cDNA samples previously used for the array analysis (isolated from FACS-sorted CD56dim CD8- CD27- CD57- and CD56dim CD8+ CD57+ NK cell subsets of seven healthy donors) were reanalysed by individual qRT-PCR assays. Assays were performed in triplicate with no reverse transcription controls for each gene. ΔCt was calculated by normalisation to the mean of the Cts of GAPDH and HPRT1. Significant differences (p<0.05) as determined by Wilcoxon matched pairs test are indicated by *.

156

4.2 Cell Surface Expression of Cytokine Receptors

In order to determine whether the transcriptional differences identified in cytokine receptor expression corresponded to differences on the cell surface, I performed flow cytometry to examine surface expression of IL12Rβ1, IL12Rβ2 and IL7R (CD127) on each of the subsets (figure 4-v). IL-12Rβ1 was expressed at slightly higher levels on CD56dim compared to CD56bright NK cells, however there was no apparent difference between the CD56dim CD8- CD57- and CD8+ CD57+ subsets. IL-12Rβ2 expression was not detected on any of the subsets. IL7R was expressed on the majority of CD56bright NK cells, was virtually absent from the CD56dim CD8+ CD57+ subset and was expressed on only a small percentage of CD56dim CD8- CD57- NK cells.

157

Figure 4-iii: Expression of cytokine receptors. PBMC from four healthy donors were incubated with antibodies against CD3, CD56, CD8, CD27, CD57 and IL-12Rβ1, IL-12Rβ2 or IL7R (CD127) or isotype controls. (A) and (B) Representative histograms for each subset are shown with mean MFI ratio (median fluorescence for receptor of interest divided by median fluorescence FMO isotype control) of all four donors. Black line = FMO isotype control, blue line = receptor of interest. (C) Percentage of each subset expressing IL7R (representative of 3 donors).

158

4.3 Differences in Expression of Activating and Inhibitory Receptors

Next, I wanted to determine if CD8+ CD57+ and CD8- CD57- subsets expressed different levels of activating and inhibitory receptors on the cell surface. First I examined expression of the inhibitory receptor NKG2A. Loss of NKG2A has been associated with NK cell maturation and previous reports indicate that it is expressed on all CD56bright NK cells and around 50% of CD56dim NK cells (Béziat et al. 2010, Yu et al. 2010). The proportion of NK cells expressing NKG2A was significantly higher within the CD56dim CD8- CD27- CD57- subset (figure 4-iii). However, a substantial portion of the CD56dim CD8+ CD57+ population also expressed this receptor. As expected, I found that NKG2A was found on nearly all CD56bright NK cells.

I also compared expression of the activating receptors NKG2D, NKp30, NKp44 and NKp46 in the CD56dim CD8+ CD57+ and CD8- CD57- subsets as well as in CD56bright NK cells (figure 4-iv). NKp30 expression was similar across all three subsets. NKp44 was expressed at low levels on the CD56bright subset and was not detected on either CD56dim subset. NKG2D had slightly higher expression on the CD56bright subset, with no difference in expression between the two CD56dim subsets. NKp46 expression was highest on CD56bright NK cells and lowest on CD56dim CD8+ CD57+ NK cells, with an intermediate level of expression on CD56dim CD8- CD57- NK cells.

159

Figure 4-iv: The CD8+ CD57+ subset contains significantly fewer NKG2A+ NK cells. PBMC were isolated from nine healthy donors and incubated with antibodies against CD3, CD56, CD8, CD27, CD57 and NKG2A. (A) Representative flow cytometry plots showing expression of NKG2A in the CD56bright subset, the CD56dim CD8- CD27- CD57- subset and the CD56dim CD8+ CD27- CD57+ subset. (B) Percentage of each NK cell subset expressing NKG2A: each line represents one donor. Significant differences determined by repeated measures ANOVA followed by Tukey’s multiple comparison test are denoted *** (p<0.001).

160

Figure 4-v: Differences in expression of activating receptors. PBMC from four healthy donors were incubated with antibodies against CD3, CD56, CD8, CD27, CD57 and NKp30, NKp44, NKp46 or NKG2D. Representative histograms for each subset are shown with mean MFI ratio (median fluorescence for receptor of interest divided by median fluorescence of FMO isotype control) of all four donors. Black line = FMO isotype control, blue line = α-NKp30, NKp44, NKp46 or NKG2D.

161

4.4 CD56dim CD8- CD27- CD57- NK Cells Produce Similar Levels of IFNγ to CD8+ CD57+ NK Cells and do not Produce IL-17 or IL-22

CD56bright NK cells make more IFNγ than CD56dim NK cells when stimulated with cytokines or PMA and ionomycin (Fehniger et al. 1999, Cooper et al. 2001b, Jacobs et al. 2001). One of the features typical of previously identified functional and developmental intermediaries between the CD56bright and CD56dim subsets, such as CD56dim CD62L+, is the ability to produce levels of IFNγ commensurate with that of the CD56bright subset (Juelke et al. 2010). If CD56dim CD8- CD27- CD57- NK cells are another intermediate population, they might be expected to produce a similar level of IFNγ in response to cytokines or PMA/ionomycin as the CD56bright subset, whereas CD57+ NK cells, as a more “mature” population would be expected to produce less. However, the CD8+ CD57+ subset expressed higher levels of IFNγ transcript, therefore they might in fact produce more IFNγ. Therefore, I set out to determine whether there was any difference in the percentage of cells producing IFNγ or the amount of IFNγ that they were producing in response to stimulation with PMA and ionomycin.

There was no significant difference in the percentage of cells producing IFNγ (figure 4-vi-A). The intensity of IFNγ staining in the CD56bright population was significantly higher than that of the CD56dim CD8- CD57- subset (figure 4-vi-B). The amount produced by the CD56dim CD8+ CD57+ subset varied considerably between donors and on average was between that of the other two subsets. In this respect, the CD8- CD57- subset does not therefore have a closer functional association with the CD56bright subset.

162

Figure 4-vi: IFNγ production in the CD56dim CD8- CD27- CD57- subset in response to PMA/ionomycin stimulation is lower than that of CD56bright NK cells and not significantly different to that of the CD56dim CD8+ CD57+ subset. PBMC from five healthy donors were stimulated with PMA/ionomycin in the presence of monensin for 6 hours. (A) Percentage of IFNγ+ cells and (B) the level of IFNγ expression (MFI ratio = MFI IFNγ+ population of test sample divided by MFI of FMO isotype control). Statistical significance was determined by Friedman test followed by Dunn’s multiple comparison test.* denotes p<0.05. (C) Representative flow cytometry plots showing IFNγ expression in CD56bright, CD56dim CD8- CD27- CD57- and CD56dim CD8+ CD57+ NK cell subsets from the same donor.

163

As mentioned in chapter 4.1, the CD56dim CD8- CD27- CD57- subset expressed a significantly higher level of RORC, a transcription factor that has been associated with NK cell subsets found in tonsil that produce IL-22 and/or IL-17 but very little IFNγ (Cella et al. 2009, Cupedo et al. 2009). Transcripts for IL-22 and IL-17A were not detected in freshly isolated CD56dim CD8- CD57- NK cells (table 4-i). In order to determine if these cytokines were expressed in this subset following activation, I stimulated PBMC with PMA and ionomycin for 6 hours then performed intracellular staining for IL-22 and IL-17A. IL-22 production has previously been observed over this same time period in tonsil NK cells (Cella et al. 2009).

IL-22 and IL-17A were not expressed by the CD56dim CD8- CD27- CD57- subset or by any other peripheral blood NK cell population, though they were detected in a subpopulation of T cells (figure 4-vii).

Figure 4-vii: CD56dim CD8- CD27- CD57- NK cells do not produce IL-22 or IL-17A. PBMC from four healthy donors were stimulated with PMA/ionomycin for 6 hours. Cells were labelled with antibodies against CD3, CD56, CD8, CD27 and CD57, fixed and permeabilized, then incubated with α- IL-22, α-IL17A or isotype control. IL-17A- and IL-22-expressing cells were detected within the T cell, but not the NK cell, population.

164

4.5 CD56dim CD8+ CD57+ NK cells are More Cytotoxic than CD56dim CD8- CD27- CD57- NK Cells

Cytotoxicty towards susceptible targets is an important function of NK cells. Both CD57 and CD8 have previously been associated with increased cytotoxic function (Srour et al. 1990, Lowdell et al. 2002, Addison et al. 2005, Chattopadhyay et al. 2009, Lopez-Vergѐs et al. 2010), therefore the CD56dim CD8+ CD57+ would be expected to have higher cytotoxicity than the CD56dim CD8- CD57- subset. In order to test this hypothesis, I first performed qRT-PCR using RNA from FACS-sorted NK cell subsets to compare expression of key cytotoxic mediators granzymes A and B and perforin, then carried out killing assays to confirm this at a functional level.

As predicted, granzyme A, granzyme B and perforin all had significantly higher expression in the CD56dim CD8+ CD57+ subset (figure 4-viii).

Granzyme A Granzyme B Perforin 80 60 * 80 * * 60 60

40

Ct

Ct

Ct 

 40 

40 -

-

-

2 2 20 2 20 20

0 0 0 - - + + 8- 57- 8+ 57+ 8- 57- 8+ 57+ 8 57 8 57

Figure 4-viii: CD8+ CD57+ NK cells express higher levels of cytotoxic molecules. Expression of cytotoxicity genes in CD8- CD57- and CD8+ CD57+ subsets of CD56dim CD27- NK cells as determined by qRT-PCR. PBMC were isolated from seven healthy female donors. RNA was extracted from NK cell subsets isolated by FACS and subjected to qRT-PCR analysis for expression of granzyme A, granzyme B and perforin. Expression was measured in triplicate with no reverse transcriptase controls for each gene. ΔCt was calculated by normalisation to the mean Ct value for GAPDH and HPRT1 housekeeping genes. Significant differences (p<0.05) as determined by Wilcoxon matched pairs test are indicated by *.

Next I compared the ability of these two NK cell subsets to kill the NK cell- susceptible K562 erythroleukaemia cell line. FACS-sorted NK cell subsets were incubated with K562 cells for 2 hours and the percentage of NK cell-mediated K562

165 cell death and activation-induced NK cell death (normalised to controls for spontaneous cell death) was determined by a flow cytometric assay. This assay involved the use of a DNA-binding dye (TO-PRO-3), which only enters cells when the integrity of the cell membrane is perturbed, as is the case during apoptosis. Counting beads enabled determination of the number of dead cells that were lost from the target cell gate. This enabled calculation of the total percentage of dead and apoptotic cells, which was normalised to controls to account for spontaneous cell death, as described in chapter 2.9. The CD56dim CD8+ CD57+ NK cell subset killed significantly more K562 cells than the CD56dim CD8- CD27- CD57- subset (figure 4- ix-A). However, there was no significant difference in activation-induced cell death (figure 4-ix-B).

(A) (B)

30 ** 40

30 20

20

10 NK cell death NK cell

% killing K562 10 % activation- %induced activation- 0 0 8- 57- 8+ 57+ 8- 57- 8+ 57+

Figure 4-ix: CD8+ CD57+ NK cells exhibit higher cytotoxicity towards K562 target cells. PBMC were isolated from six healthy female donors. CD56dim CD8- CD27- CD57- and CD56dim CD8+ CD57+ NK cells were isolated by FACS and incubated for 2 hours with K562 at a 2:1 NK cell to K562 ratio. Countbright counting beads were used to calculate the number of dead cells that were lost from the K562 gate (A) or NK cell gate (B) and TO-PRO-3 DNA-binding dye was used to identify apoptotic cells. (A) K562 killing was determined by comparing the relative numbers of viable (TO-PRO-3- negative) K562 cells in triplicate experimental wells vs. K562-only controls. Paired t test indicated a statistically significant difference between CD8+ CD57+ and CD8- CD57- NK cells (p=0.0085 **). (B) The percentage of NK cells lost due to activation-induced cell death was estimated by comparing the number of viable NK cells following incubation with K562 to that in NK cell-only controls. Paired t test indicated no significant difference in activation-induced NK cell death.

166

4.6 Discussion

In this chapter I have compared transcriptional profiles, cell surface receptor expression, cytokine production and cytotoxic function of two subsets of CD56dim NK cells and identified both similarities and differences.

Many of the transcriptional differences identified by qRT-PCR array are related to IFNγ signalling. IFNγ-related genes that were upregulated in the CD8+ CD57+ subset include T-bet, STAT3, NFATC2, SOCS3 and IL12RB1. T-bet is crucial for IFNγ transcription in NK cells and CD4+ T cells and its expression is augmented by IFNγ (Lighvani et al. 2001, Szabo et al. 2002). STAT3 is activated by a number of cytokines including the IL-6 family, IL-10 family, interferons and IL-23 (Schindler and Plumlee 2008). It promotes survival and proliferation (Takeda et al. 1998, Akira 2000, Narimatsu et al. 2001). IFNγ activates both STAT3 and SOCS3, which inhibits STAT3-mediated signalling (Bluyssen et al. 2010). SOCS proteins are activated by cytokines and growth factors such as IL-2, IL-6, erythropoietin and PDGF and form part of a negative feedback loop, suppressing further signalling through cytokine receptors (Starr et al. 1997, Cacalano et al. 2001). SOCS3 binds to JAK2 associated with receptors that signal via STAT3 and inhibits its tyrosine kinase activity (Sasaki et al. 1999, Yoshimora et al. 2012). It also inhibits IL-12 signalling by binding to the STAT4 docking site on IL-12Rβ2 (Yamamoto et al. 2003).

Expression of IFNγ can be induced by a number of mechanisms, including stimulation with IL-12 and IL-18 or ligation of activating receptors (Anegón et al. 1988, Cassatella et al. 1989, Cuturi et al. 1989, Warren et al. 1995, Fehniger et al. 1999). Previous reports have indicated that CD56bright NK cells express higher levels of IL- 12RB2 transcript than CD56dim (Wendt et al. 2006, Lopez-Vergѐs et al. 2010) and that CD56dim CD57+ NK cells have a lower level than CD56dim CD57- NK cells, which is consistent with the data presented here showing that transcription is lower in the CD8+ CD57+ subset of CD56dim NK cells. IL-12Rβ2 expression has also been associated with “NK2” polarisation and production of IL-5 and IL-13 (Peritt et al. 1998, Takahashi et al. 2001), but I did not detect any significant difference in the expression of these two cytokines. Curiously, however, I found that IL-12RB1 transcript was more abundant in the CD8+ CD57+ subset. IL-12Rβ1can pair either with IL-12Rβ2 to form the high affinity IL-12 receptor or with IL-23R to form the IL-23 receptor

167

(Presky et al. 1996, Parham et al. 2002). However, expression of IL23R was low to undetectable in both subsets, therefore it seems unlikely that IL-12RB1 expression in the CD8+ CD57+ subset was linked to preferential formation of IL-23R rather than IL- 12R heterodimers.

The transcriptional differences in expression of these receptor subunits did not translate into differences on the cell surface of resting NK cells. IL-12Rβ2 was not detected on the surface of either CD56dim subset, or on the surface of CD56bright NK cells. This is consistent with a previous study demonstrating that IL-12Rβ2 is only expressed on the cell surface following activation with IL-18 (Mailliard et al. 2005). This may be part of the mechanism by which IL-18 synergises with IL-12 for induction of IFNγ production. The CD8- CD57- subset also had higher expression of IL-18 transcript, therefore it is possible that in response to an appropriate stimulus, these cells might produce IL-18 which acts in an autocrine manner to induce surface expression of IL-12Rβ2. However, no antibodies were available in order to assess whether this subset actually secretes IL-18 protein. IL-12Rβ1 was detected on the surface of freshly isolated PBMC, but there was no apparent difference in surface expression between the two subsets. This could be due to the lower sensitivity of flow cytometry compared to qRT-PCR, especially given that the fluorescence intensity for this antibody was low on all subsets. Activation might be required in order to upregulate cell surface IL-12Rβ1 levels. The presence of higher levels of transcript might enable a more rapid upregulation of cell surface levels on CD8+ CD57+ NK cells when translation or protein transport is increased. CD8+ CD57+ NK cells might also have a higher rate of IL-12Rβ1 protein turnover, resulting in similar levels of expression on the surface of resting NK cells, but allowing them to more rapidly replace internalised receptors following binding of IL-12 or IL-23.

Following PMA and ionomycin stimulation, CD8+ CD57+ NK cells did not produce significantly more IFNγ than CD8- CD57- NK cells, despite higher levels of IFNγ transcript. However, cytokine production was only measured at one time point and it would be interesting to know if there are differences in the kinetics of the response. Higher levels of IFNγ transcript in CD56dim compared to CD56bright NK cells have previously been associated with the ability to produce IFNγ more rapidly in response to activating signals (De Maria et al. 2011). Therefore it seems likely that the CD8+

168

CD57+ subset may be activated more rapidly than the CD8- CD57- subset. PMA and ionomycin stimulation demonstrated that both subsets have the capacity to produce substantial amounts of IFNγ, but given the findings of Fauriat et al. (2010a) regarding preferential responsiveness to different stimuli in CD56bright and CD56dim NK cell subsets, further investigation is clearly warranted. CD8+ CD57+ NK cells might produce more IFNγ in response to target cells or stimulation of activating receptors such as CD16, a characteristic associated with CD56dim NK cells in general (Anfossi et al. 2006, Fauriat et al. 2010a). I found that CD8+ CD57+ NK cells expressed higher levels of CD3ζ, an important ITAM-containing adaptor protein that mediates signalling via CD16, NKp30 and NKp46 and possibly through CD8 itself (Lanier et al. 1989, Pessino et al. 1998, Vitale et al. 1998, Pende et al. 1999, Gibbings and Dean Brefus 2009). They also exhibited higher expression of NFATC2, which is induced by CD16 stimulation (Arambu et al. 1995). In Th1 cells, NFAT proteins cooperate with STAT4, which is activated in response to IL-12R engagement, to induce IFNγ expression and with T-bet to promote and maintain Th1 polarisation (Macian 2005).

However, if the CD56dim CD8- CD27- CD57- subset does indeed resemble the CD56bright population, it might produce more IFNγ in response to cytokine stimulation. Given the opposing differences in transcripts for each of the IL-12 receptor components, it would be interesting to see how the two subsets respond to IL-12 stimulation and how this compares to the CD56bright subset. It would also be interesting to determine the effect of IL-18 co-stimulation on each subset. CD57- NK cells are reported to have higher surface expression of IL-18Rα (Lopez-Vergѐs et al. 2010), so a greater enhancement of IFNγ production by IL-18 might be expected in the CD8- CD57- subset. IL-18 might upregulate IL-12Rβ2 surface expression more strongly or more rapidly on the CD56dim CD8- CD27- CD57- subset since these cells already express high levels of transcript. Alternatively IL-18 might upregulate transcription in the CD8+ CD57+ subset, abolishing the difference at the mRNA level.

Though the CD56dim CD8- CD27- CD57- subset did appear to share some characteristics with the CD56bright subset, I also considered the possibility that it might be related to “NK-22” cells found in human tonsils. Two RORC-expressing subsets have been described in human tonsil, both of which produce IL-22 and one of which also produces IL-17A (Cella et al. 2008, Cupedo et al. 2009). A similar IL-22

169 producing RORC+ subset has also been described in the murine gut (Satoh-Takayama et al. 2008). Human NK-22 cells are NKp44+ CD127+ and have low expression of perforin and IFNγ (Cella et al. 2008, Cupedo et al. 2009). I did not detect expression of NKp44 on the CD56dim CD8- CD27- CD57- subset and though IFNγ production was significantly lower than CD56bright NK cells and slightly lower than CD56dim CD8+ CD57+ NK cells, the high percentage of cells expressing IFNγ was not consistent with RORC+ tonsil NK cells. More importantly, expression of IL-22 was not detected in this subset, or in any peripheral blood NK cell subset when stimulated with PMA and ionomycin for the same duration in which IL-22 production was observed in the study by Cella et al.

For the qRT-PCR and killing assays, I did not make comparisons to the CD56bright subset since I found that sorting of more than two populations compromised purity, which was of paramount importance for the PCR assays in particular. However, expression patterns of activating and inhibitory receptors were consistent with the hypothesis that CD56dim CD8- CD27- CD57- NK cells are a functional intermediate between the CD56bright and the CD56dim CD8+ CD57+ subset. NKG2A, which is present on virtually all CD56bright NK cells, was present on a significantly higher proportion of CD56dim CD8- CD27- CD57- NK cells compared to CD56dim CD8+ CD57+ NK cells. The CD8- CD57- subset also had a level of NKp46 expression in between that of CD56bright NK cells (which, consistent with a previous report (Vitale et al. 2004), had the highest expression of NKp46) and that of the CD8+ CD57+ subset. Lower expression of NKp46 on CD57+ NK cells is consistent with the findings of Lopez-Vergѐs et al. (2010), however they also reported lower expression of NKp30 and NKG2D on CD57+ compared to CD57- CD56dim NK cells, which I did not observe in the CD8+ CD57+ subset. CD8- CD57- NK cells had higher levels of both IL-12RB2 and IL7R receptor transcripts, which have also previously been reported to have higher expression in CD56bright compared to CD56dim NK cells (Hanna et al. 2004b, Wendt et al. 2006). However, surface staining revealed that actually only a small percentage of CD56dim CD8- CD27- CD57- NK cells expressed IL-7 receptor, whereas the majority of CD56bright NK cells expressed it. Another transcriptional similarity with the CD56bright subset was lower expression of CD3ζ. CD3ζ is expressed at ~3-fold higher levels in CD56dim compared to CD56bright NK cells (Hanna et al. 2004b). The fold-difference that I observed between the two

170

CD56dim subsets was less than this, indicating that the CD56dim CD8- CD27- CD57- subset probably has an intermediate level of CD3ζ expression, though it is impossible to be confident of this without having directly compared all three subsets in the same analysis.

In addition to CD3ζ, I also detected increased mRNA levels of CD3δ and CD3γ in the CD8+ CD57+ subset. CD3δ and CD3γ transcripts have previously been reported to be absent from adult peripheral blood NK cells, using less sensitive detection methods on NK cell clones or unsorted polyclonal NK cells (Biassoni et al. 1988, Anderson et al. 1989, Lanier et al. 1992). Since the NK cell populations isolated in my study had very high purity, it is unlikely that these transcripts were derived from contaminating T cells. CD3δ and/or γ also appear to be expressed in thymic bipotent T/NK cell precursors (Sanchez et al. 1994). In addition, both transcripts and protein have been detected in the cytoplasm of NK cell clones derived from human fetal liver, however the protein is not transported to the cell surface (Lanier et al. 1992, Phillips et al. 1992). Despite being derived from fetus, these NK cells display the cytotoxic function of mature NK cells (Phillips et al. 1992). Therefore expression of CD3δ and CD3γ is not restricted to immature progenitors with T cell potential and it is quite conceivable that they would also be expressed in adult mature NK cells. If these molecules are retained in the cytoplasm of CD56dim CD8- CD27- CD57- NK cells, as they are in fetal NK cells, the functional relevance of their expression is questionable. However they might contribute to signalling pathways in a similar way to CD3ζ.

CD56dim CD8+ CD57+ NK cells also expressed higher levels of IL23A transcript. However they did not express the other subunit that makes up functional IL-23 heterodimers, IL-12B (p40). IL-23 is not a cytokine normally associated with NK cells and it would be interesting to find out if there are any circumstances under which they also express p40 and actually secrete IL-23, or if IL-23A has some function that is independent of the p40 subunit.

The CD56dim CD8+ CD57+ NK cell subset expressed higher levels of perforin and granzymes A and B, as expected, since higher expression of these molecules has previously been reported in CD57+ compared to CD57- CD56dim NK cells (Chattopadhyay et al. 2009). Cytotoxicity towards K562 is apparently similar in CD57- and CD57+ subsets of CD56dim NK cells (Lopez-Vergѐs et al. 2010). However,

171

CD8+ NK cells are reported to kill more K562 due to their relative resistance to activation-induced cell death (Addison et al. 2005). Therefore I hypothesised that the CD8+ CD57+ population would kill more K562 cells than the CD8- CD57- subset. The CD8+ CD57+ subset did indeed kill a higher proportion of K562 cells, however there was no significant difference in the rate of activation-induced NK cell death, which occurred in both subsets during the 2 hour incubation period. Since increased granzyme expression is apparently not sufficient to confer enhanced cytotoxic function on CD57+ NK cells (Lopez-Vergѐs et al. 2010), the mechanism for this increased cytotoxicity remains unclear. However differential expression of signalling molecules such as CD3ζ could have a role.

In conclusion, a number of transcriptional and functional differences have been identified between the two CD56dim subsets that were examined. The data so far suggests that CD8+ CD57+ NK cells, as hypothesised, are the more cytotoxic population, whereas CD8- CD27- CD57- NK cells show more similarities to the CD56bright subset. These results highlight future avenues for investigation that would further improve understanding of the functional differences between these subsets.

172

5. NK CELLS DO NOT INFILTRATE THE CNS IN THE LINE 7 MURINE MODEL OF MULTIPLE SCLEROSIS

Since multiple sclerosis is the result of inflammation specifically within the central nervous system, an obvious question is whether natural killer cells are playing a direct role within the CNS or whether NK cells in the periphery are influencing the activation and therefore migration of other cells such as T cells prior to crossing the blood-brain barrier.

Studies of human MS are limited in their ability to determine what is happening inside the CNS itself, due to its inaccessibility. CSF is difficult to obtain and brain and spinal cord tissue can only be obtained post-mortem. As a result, MS tissue samples are usually from patients with advanced, progressive disease, and therefore may not be representative of pathology in the earlier relapsing-remitting phase. Furthermore, control samples can only be obtained from people who have died due to other causes and therefore it is debatable whether they can be said to be healthy.

Immunohistochemistry and confocal microscopy work performed by my colleague Pascal Durrenberger indicated very few NKp46+ NK cells present in post-mortem CNS samples from either MS patients or controls (Durrenberger et al. 2012). This would support the theory that the main role of NK cells in MS is in influencing inflammatory processes in the periphery that result in the abnormal migration of T cells and other leukocytes into the CNS. However, it does not exclude a possible role of NK cells within the CNS during relapsing-remitting MS. Indeed, human NK cells have been shown to kill oligodendrocytes, astrocytes and microglia in vitro (Morse et al. 2001, Saikali et al. 2007). NK cell cytotoxicity towards microglia is a proposed mechanism for inhibition of Th17 polarisation within the CNS during EAE (Hao et al. 2010). Transgenic CX3CR1-/- mice with a selective NK cell homing deficiency to the CNS develop EAE more rapidly and with higher mortality than wild type, which suggests that NK cells have an immunoregulatory function within the CNS during demyelinating disease (Huang et al. 2006). However, systemic depletion of NK cells using an antibody directed against NK1.1 not only alters the proliferation and IL-17 production of T cells in the CNS, it also increases T cell proliferation in the periphery

173 and causes an increase in the amount of leukocytes infiltrating the CNS (Hao et al. 2010) , suggesting that NK cells affect disease both inside and outside of the CNS.

In order to determine whether NK cells might have a role within the CNS at an earlier stage of disease, I sought to determine which types of NK cell were present in the CNS of Line 7 mice with a demyelinating autoimmune disease resembling MS. These humanised transgenic mice generated in the Altmann lab express high levels of HLA- DRB1*1501 (an HLA class II molecule associated with susceptibility to human MS) on an Aβ0 background and a TCR specific for the DR15-restricted epitope MBP 85- 99, cloned from the MS patient-derived Ob1A12 T cell clone (Ellmerich et al. 2005). Line 7 mice exhibit a high rate (around 60%) of spontaneous demyelinating disease leading to paralysis, without the need for disease induction by injection of myelin peptides or myelin-reactive T cells (Ellmerich et al. 2005). The spontaneous nature of demyelinating disease in the Line 7 model avoids bias caused by adjuvants required for EAE induction. Large numbers of T cells infiltrate the CNS of Line 7 mice and by the time disease is clinically evident, epitope spreading has occurred, with T cell responses directed at HLA-DR15-restricted epitopes of MBP, MOG and αβ-crystallin (Ellmerich et al. 2005).

Since these mice have a full repertoire of other immune cell-types including B cells, NK cells and  T cells in the context of their autoimmune T cell repertoire and the development of severe, spontaneous disease, they provide an opportunity for reappraisal of the role of CNS NK cells which might be considered to offer advantages over the adjuvant-induced EAE models.

174

5.1.1 NKp46+ NK Cells were Not Detected in the CNS of Line 7 Mice

Using the Line 7 model, I set out to determine which subsets of NK cells, if any, were present in the CNS of mice exhibiting symptoms of autoimmune demyelination. Cells were extracted from the CNS and spleen of Line 7 mice with moderate clinical disease (score 2-3) and stained with antibodies against NK cell markers. Score 2-3 mice in this model have previously been shown in the lab to exhibit profound demyelination and pronounced CNS mononuclear infiltrates.

Mouse NK cells are commonly defined either as NK1.1+ CD3- or as NKp46+. NKp46 was identified as the marker of choice, since one of the NK1.1 antibodies tested produced a very low signal in spleen, whilst the other exhibited an unacceptably high level of non-specific binding as determined using a matched concentration of appropriate isotype control.

For each of the six mice tested, staining of spleen and CNS were carried out simultaneously. In spleen samples, a clear population of NKp46+ cells could be identified, representing 1-3% of splenocytes (figure 5-i). In CNS samples from the same mice however, there was no clear NKp46+ population and no discernible difference between NKp46 staining and that of the isotype control (figure 5-ii and 5- iii).

(A) (B)

IgG2a efluor450 NKp46 efluor450

Figure 5-i: NK cells expressing NKp46 are present in the spleen of Line 7 mice. Data is representative of six Line 7 mice with moderate clinical disease (score 2-3). Splenocytes were extracted by syringe and treated with erythrocyte lysis buffer as described in methods. Cells were stained with live/dead yellow dye for exclusion of dead cells followed by (A) rat IgG2a eflour450 (isotype control) or (B) rat anti-mouse NKp46 eflour450 for identification of NK cells.

175

Isotype control Anti-NKp46

IgG2a efluor450 NKp46 efluor450

Figure 5-ii: NKp46+ leukocytes are not found in the CNS of Line 7 mice. CNS tissue was extracted from three Line 7 mice (two female and one male) with a clinical disease score of 2 (low/moderate). Tissue was physically and enzymatically disrupted prior to isolation of leukocytes by Percoll density gradient centrifugation. Cells in the leukocyte fraction were stained with live/dead yellow for exclusion of dead cells and either rat anti-mouse NKp46 (right) or IgG2a isotype control (left).

176

Isotype control Anti-NKp46

(A)

(B)

(C)

IgG2a efluor450 NKp46 efluor450

Figure 5-iii: NKp46+ leukocytes are not found in the CNS of Line 7 mice. CNS tissue was extracted from three female Line 7 mice with moderate clinical disease (score 2.5 - 3) and physically and enzymatically disrupted prior to isolation of leukocytes by Percoll density gradient centrifugation. Cells in the leukocyte fraction were stained with live/dead yellow for exclusion of dead cells and rat anti- mouse NKp46 (right) or IgG2a isotype control (left). (A) and (B) represent mice sacrificed at a score of 2.5 and sample (C) is from a mouse sacrificed at score 3.

177

5.1.2 Lack of NKp46 Detection was not due to Antibody Sequestration

To test the hypothesis that sequestration of NKp46 antibody by residual CNS debris or non-lymphocytic cells might be preventing the detection of NKp46 on CNS- derived NK cells, the lymphocyte population was enriched by fluorescence-activated cell sorting (FACS) of CNS-derived leukocytes prior to staining with NKp46. Forward/side scatter analysis of sorted and unsorted samples revealed that sorting had succeeded in enriching the lymphocyte population, although some larger cells were still present (figure 5-iv). NKp46 was still not detected in samples that were stained with anti-NKp46 after sorting (figure 5-v), suggesting that the failure to detect NKp46+ lymphocytes was not likely to be caused by antibody sequestration.

(A)

(B)

Figure 5-iv: Enrichment of lymphocyte population by FACS. Forward and side scatter plots showing CNS leukocyte fraction from a female Line 7 mouse with moderate clinical disease (score 2), (A) prior to sorting and (B) after sorting. Leukocytes were isolated from physically and enzymatically disrupted CNS tissue by Percoll density gradient centrifugation. Cells were incubated with live/dead yellow dye, then live lymphocytes were sorted on FACS Aria II according to forward- and side-scatter with doublet exclusion and negativity for live/dead yellow dye.

178

Figure 5-v: NKp46 was not detected in FACS-sorted lymphocytes isolated from the CNS of a Line 7 mouse. Leukocytes were isolated from physically and enzymatically disrupted CNS tissue from a female Line 7 mouse with a clinical disease score of 2 by Percoll density gradient centrifugation and incubated with live/dead yellow dye for exclusion of dead cells. Lymphocytes were identified based on forward and side scatter and sorted using the FACS Aria II with doublet and dead cell exclusion. 8 x 104 sorted lymphocytes were stained with α-CD3 FITC and either rat IgG2a efluor450 or α-NKp46 efluor450 for identification of NK cells.

5.1.3 NKG2D was also Not Detected on CNS-derived Lymphocytes

To address the possibility that NK cells are present in the CNS but have downregulated expression of NKp46, samples from spleen and CNS were stained with an antibody against NKG2D. NKG2D is typically found on all NK cells, however it is also found on activated CD8+ T cells (Diefenbach et al. 2000) therefore would not normally be the marker of choice. NK cells can however be identified by co-staining with CD3 for exclusion of NKG2D+ T cells.

In spleen samples, well-defined NKG2D+ populations were found amongst both CD3+ and CD3- leukocytes. However, in CNS samples neither CD3- nor CD3+ populations contained NKG2D+ cells (figure 5-vi). In one of the mice, scoring 2.5, there was a small proportion of cells that appeared to be weakly positive for NKG2D, but the difference from the isotype control was not sufficient to be confident that this was a genuine NKG2D+ population.

179

Figure 5-vi: NKG2D+ lymphocytes were not detected in the CNS of Line 7 mice. Leukocytes were isolated from three Line 7 mice with moderate clinical disease (two female and one male, score 2 - 2.5), by disruption of CNS tissue followed by Percoll density gradient centrifugation, then stained with α-CD3 and either α-mouse NKG2D APC (right) or rat IgG1 APC isotype control (left).

180

5.1.4 Discussion

Neither NKp46 nor NKG2D were detected on cells infiltrating the CNS of Line 7 mice exhibiting moderate clinical symptoms of demyelinating autoimmune disease.

The possibility of sequestration of antibody by impurities in the CNS sample was considered. NKp46 expression has been detected on human astrocytes (Durrenberger et al. 2012), so it is possible that CNS-resident cells or debris from such cells might specifically sequester NKp46 antibody. In addition, there is a substantial proportion of dead cells in these CNS samples, which have a tendency to bind antibody non- specifically. However, CD3 staining was still detected in CNS samples, so non- specific sequestration of antibody seems unlikely. Enriching lymphocytes by cell sorting did not make a difference to NKp46 staining, which also suggests that the antibody was not being sequestered.

There are several other possible reasons why these NK cell markers were detected in the spleen but not the CNS. NK cells might be absent from the CNS because they are not sufficiently activated and/or not expressing the molecules necessary for migration across the blood-brain barrier. Alternatively, they might be able to enter the CNS but then either undergo apoptosis or downregulate the activating receptors by which we would usually identify them.

Widespread downregulation of NK cell receptors, including NKp46 and NKG2D, is observed in certain types of cancer (Costello et al. 2002, Garcia-Iglesias et al. 2009). Downregulation of NKG2D has also been observed in transgenic mice expressing constitutively high levels of ligands such as MICA or Rae-1 (Oppenheim et al. 2005, Wiemann et al. 2005) and several studies have demonstrated downregulation of human NKG2D by soluble ligands including MICA, MICB and ULBPs released by tumour cells and activated CD4+ T cells (Groh et al. 2002, Song et al. 2006, Cao et al. 2007, Cerboni et al. 2009, Ashiru et al. 2010). NKG2D ligands are expressed on activated T cells and macrophages (Molinero et al. 2002, Hamerman et al. 2004) and reactive oxygen species produced by phagocytes have been shown to downregulate expression of both NKG2D and NKp46 on CD56dim NK cells (Romero et al. 2006). In addition, NKG2D is downregulated by the combination of IFN-β and IL-12 produced during viral infection (Muntasell et al. 2010). Therefore, receptor expression on CNS NK cells could conceivably be influenced by expression of stress-induced ligands in 181 the CNS tissue and on activated infiltrating leukocytes, as well as by the cytokine mileu within the inflamed CNS.

Work carried out in our group indicates a substantial increase in MICA and MICB mRNA in white matter lesions of MS patients compared to healthy white matter (Hyun-Woong Do, unpublished data). Increased levels of MICA are also found on intestinal epithelial cells of patients with Crohn’s disease and in both synovium and peripheral blood of rheumatoid arthritis patients (Groh et al. 2003, Allez et al. 2007). However, unlike cancer patients, these patients did not exhibit downregulation of NKG2D, most likely because of the counteracting influence of cytokines such as IL- 15 and TNFα, which upregulate NKG2D expression (Groh et al. 2003, Allez et al. 2007). These cytokines are also upregulated in MS and EAE (Hofman et al. 1989, Selmaj et al. 1991, Renno et al. 1995, Kivisäkk et al. 1998, Blanco-Jerez et al. 2002, Rentzos et al. 2006, Wu et al. 2010), which makes it seem less likely that downregulation of NKG2D would take place.

Meanwhile, NKp46 expression on astrocytes is increased in the CNS of MS patients (Durrenberger et al. 2012), which excludes the possibility of generic downregulation of this receptor within the CNS. If downregulation of NKp46 was taking place, it would have to be by a mechanism specific to NK cells.

Overall, the most likely possibility is that NK cells are not present in the CNS tissue of Line 7 mice at this stage of disease. Prior to extraction of the CNS, mice were perfused with PBS to flush out the blood vessels, therefore only NK cells which had crossed the BBB would be detected in the sample. It is possible therefore, that NK cells have migrated towards the CNS but not penetrated the BBB. Since the data presented here represents mice that already had a moderate degree of paralysis (score 2 – 3), it is also possible that NK cells might enter the CNS at a different stage of disease, possibly before onset of visible symptoms.

During EAE, NK cells have been reported to infiltrate the murine CNS in substantial numbers and to have a clinically significant role in regulating inflammation once there (Huang et al. 2006, Hao et al. 2010). The contradictory results between these studies and ours with regard to the presence of NK cells is most likely due to differences between the EAE and Line 7 disease models. For instance, there might be a difference

182 in the activation status of NK cells in the blood, since EAE induction involves use of an adjuvant, which is likely to activate NK cells independently of myelin-specific T cell responses. There could also be differences in chemokine expression and in the degree of selectivity of the BBB.

It is important to bear in mind that neither murine model is a completely faithful representation of human disease. The Line 7 model has more similarity to MS than most EAE models in so far as it is spontaneous and involves a humanised HLA/TCR combination associated with MS. However it goes without saying that there are significant differences in NK cell biology and other aspects of the immune system between mice and humans. Furthermore, human MS is a complex disease and although HLA-DRB15*01 is a strong genetic risk factor, in reality disease initiation and progression in humans is likely to be determined by a range of factors. Some of these factors, in particular KIR and HLA class I genotypes, might influence the activation state of NK cells, and therefore their ability to cross vascular endothelial barriers such as the BBB.

The data from the Line 7 mice is however consistent with staining of human post- mortem brain sections carried out by Pascal Durrenberger, which detected very few NKp46+ lymphocytes in white matter lesions of patients with progressive MS (Durrenberger et al. 2012). Of course, if NKp46 downregulation was occurring in the Line 7 model, it might also occur in the human CNS, so it is not completely certain that there are no NK cells there. Also, these samples are representative of very late disease and do not exclude the possibility of NK cell infiltration during the relapsing- remitting phase. In support of a role for NK cells within the CNS during human MS, presence of NK cells has been reported in the CSF of untreated RRMS patients (Bielekova et al. 2011). The ratio of NK cells to CD4+ and CD8+ T cells in the CSF is increased during daclizumab treatment due to an increase in the size of the CD56bright NK cell population, indicating a potential regulatory role for NK cells within the CNS. Whether they are able to penetrate other parts of the human CNS, in particular white matter lesions, during RRMS remains to be seen.

In summary, the data presented in this chapter does not support a role for NK cells within the CNS in either promoting or regulating demyelination in the Line 7 model. However, the absence of NK cell infiltration of the CNS in this particular model does

183 not exclude the possibility that they might regulate inflammation within the CNS of human RRMS patients. As a result of the studies initiated here and the problems of reconciling these with some of the published work on EAE, experiments were subsequently initiated in the lab to cross the Line 7 mice onto the E4BP4 -/- strain lacking all NK cells, as reported from Hugh Brady’s lab (Gascoyne et al. 2009).

184

6. DISCUSSION

In this study, I set out to identify subsets of NK cells with altered frequencies in MS patients that might be implicated in exacerbation or amelioration of disease, to characterise these subsets and determine whether they might have a role within the CNS itself.

I identified subsets of NK cells with a decreased frequency as well as one subset with a higher frequency in untreated MS patients compared to healthy controls. I confirmed the previously reported expansion of CD56bright NK cells in IFNβ-treated MS patients and showed that IFNβ also has pronounced effects on frequencies of CD56dim subpopulations. Three of the four CD56bright subsets had a reduced frequency in MS patients compared to controls, and all were reduced in patients in relapse compared to remission, suggesting that CD56bright subsets may share a protective immunoregulatory role in MS. However, these subsets were not all equally affected by IFNβ treatment. Three of them showed a slight, non-significant increase in IFNβ- treated patients, but the average frequency of the CD8- CD27- subset of CD56bright NK cells was doubled in IFNβ-treated patients and tripled in IFNβ-treated patients in remission. It is interesting that the only subset of CD56dim NK cells that was increased in IFNβ-treated patients also lacked expression of CD8, CD27 and CD57 (the latter of which is absent from all CD56bright NK cells). This suggests that CD56dim and CD56bright subsets lacking these markers share a common feature that increases their responsiveness to IFNβ treatment. They might for example have higher expression of signalling components associated with type I interferon signalling, allowing them to respond directly to IFNβ stimulation. However, there are many possible indirect mechanisms by which IFNβ might act on NK cells. These include modulation of IL-2 signalling as discussed in chapter 3 and induction of other cytokines and growth factors by IFNβ, which might in turn affect NK cell function. NK cells also interact with other leukocytes including monocytes, macrophages and DCs (Ferlazzo et al. 2002, Mocikat et al. 2003, Borg et al. 2004, Dalbeth et al. 2004, Ferlazzo et al. 2004, Gerosa et al. 2005, Mailliard et al. 2005, Nedvetzki et al. 2007). IFNβ has a number of effects on APCs and overall tends to inhibit their ability to induce Th1 polarisation (Genç et al. 1997, McRae et al. 1998, Karp et al. 2000, Ramgolam et al. 2009). Thus APCs exposed to IFNβ might also preferentially activate and induce proliferation of NK cells with a more immunoregulatory phenotype. It would also be interesting to

185 investigate whether the CD56bright CD8- CD27- NK cell subset has any functional differences, besides responsiveness to IFNβ, to other CD56bright NK cell subsets and whether the functional overlap of CD56dim CD8- CD27- CD57- NK cells with CD56bright CD8- CD27- NK cells is greater than that with the CD56bright population as a whole.

It is not clear from the data presented so far whether the changes in subset frequencies are the result of increased proliferation or survival rates of certain subsets, or if changes in the cytokine mileu and other factors are causing NK cells to alter their phenotype. Markers such as Ki67 could be employed to determine whether the proliferation of particular subsets is enhanced during disease or by treatments such as IFNβ. Cytokines such as IL-2, produced by activated T cells and IL-12, IL-15 and IL- 18 produced by DCs have previously been reported to induce changes in NK cell phenotype, in terms of both surface markers and functional attributes (Peritt et al. 1998, Loza et al. 2004, Mailliard et al. 2005, Romagnani et al. 2007, Takahashi et al. 2007, Lopez-Vergѐs et al. 2010, Béziat et al. 2011). Activation of other leukocytes and consequent release of proinflammatory cytokines during autoimmune disease could therefore induce changes in NK cell phenotype.

Another interesting question is whether the frequency of NK cell subsets is influenced by genetic factors. Differences in KIR/HLA genotype might influence the process of NK cell maturation. MS-associated polymorphisms of IL-2Rα and IL-7R (Hafler et al. 2007, De Jager et al. 2009a) might affect the survival, proliferation and maturation of the more immature subsets, such as CD56bright that typically express these receptors (Caligiuri et al. 1990, Nagler et al. 1990, Baume et al. 1992, Hanna et al. 2004b, Wendt et al. 2006, Nielsen et al. 2012). In addition, polymorphisms in the genes encoding T-bet and STAT3, of which I detected higher expression in the disease- associated CD56dim CD8+ CD57+ NK cell subset, were implicated in MS susceptibility by a recent GWAS study (Patsopoulos et al. 2011). NK cell-extrinsic genetic factors could also play a role in determining the way in which they develop.

In order to gain insight into the function of the CD56dim CD27- CD8+ CD57+ subset and its CD8- CD57- counterpart, I sorted these two subsets by FACS and identified differences in their transcriptional profiles and cytotoxic function. If technicalities permitted, CD56dim CD27- NK cells which are single-positive for either CD8 or CD57

186 could be sorted and characterised in a similar manner, in order to dissect whether particular functional attributes are related to expression of one molecule or the other, or if they are defined solely by the combination of both markers.

The CD56dim CD8+ CD57+ subset expressed a higher level of mRNA encoding S1PR1, a surface molecule that is upregulated upon activation and which is involved in egression from lymph nodes (Matloubian et al. 2004, Pappu et al. 2007). It would be interesting to see if CD56dim CD8- CD27- CD57- NK cells, like CD56bright NK cells, are enriched in lymph nodes and whether they express different levels of receptors associated with migration to lymph nodes, in particular CD62L, which has previously been identified as a marker of polyfunctional CD56dim NK cells with similarity to the CD56bright subset (Juelke et al. 2010). The phosphorylated form of fingolimod, a new therapy for MS, is an analog of sphingosine-1-phosphate, which binds to S1PR1, causing its internalisation and therefore preventing egression of autoreactive T cells from lymph nodes (Mandala et al. 2002, Xie et al. 2003, Brinkmann et al. 2004, Matloubian et al. 2004, Brinkmann et al. 2009). Therefore this therapy may also have implications for migration of NK cells and if it preferentially blocks migration of the CD8+ CD57+ subset this might enhance suppression of autoimmunity by this therapy. CNS infiltration of leukocytes is thought to be reduced by fingolimod, primarily because T cells which have been primed in lymph nodes cannot access the CNS without first migrating out of lymph node into the blood. However, this receptor might also influence other aspects of cell migration and activate other functions. Therefore effects of this therapy on migration and function of NK cell subsets would be worth studying, since CD8+ CD57+ NK cells might be affected more strongly than CD8- CD57- CD56dim NK cells. Another avenue for future investigation is comparing expression of chemokine receptors, including CX3CR1, which is crucial for migration of NK cells into the CNS during EAE (Huang et al. 2006, Hao et al. 2010). Differences in CX3CR1 expression between healthy controls and MS patients have been reported previously (Infante-Duarte et al. 2005) and it would be interesting to know whether the CX3CR1+ population corresponds to the CD56dim CD8+ CD57+ subset, since both have higher frequency in untreated MS patients. If so, this could indicate that this highly cytotoxic subset is likely to enter the human CNS and participate in killing of oligodendrocytes and neurons.

187

I did not detect CNS infiltration of NK cells in the humanised transgenic Line 7 mouse model of MS. However, given the abundant presence of NK cells in the CNS in other disease models (Hammarberg et al. 2000, Huang et al. 2006, Hao et al. 2010, Leavenworth et al. 2010), it is still plausible that they enter the human CNS. Infiltration of NK cells has been demonstrated in CSF (Bielekova et al. 2011), however this does not prove that they are able to migrate into the parenchyma and contribute to demyelination in MS lesions. Very few NK cells were observed in white or grey matter lesions of post-mortem samples from progressive MS patients and their localisation tended to be perivascular rather than parenchymal (Durrenberger et al. 2012). NK cells were only found in patients with pronounced focal demyelination and were more likely to be found in active lesions. Ideally, brain tissue samples would be obtained from a cohort of RRMS patients who have died prior to onset of progressive disease in order to determine if NK cells penetrate the parenchyma during the acute inflammatory stage of disease. However, even in RRMS patients, substantial NK cell infiltration may well only be observed during relapses and it is difficult to acquire such specific post-mortem samples. Given the limitations of studying tissue, it would be worth evaluating more precisely which subsets are present in the CSF of MS patients, both untreated and treated. If pathogenic NK cells are indeed entering the human CNS, then inhibition of NK cell migration might be another beneficial effect of treatments such as fingolimod and Tysabri. However, if such therapies prevented immunoregulatory subsets from entering the CNS this might prevent suppression of inflammation mediated by leukocytes that have already infiltrated the CNS prior to treatment.

However, even if NK cells do not infiltrate and participate in or regulate inflammatory processes and damage within the CNS, they could still have important roles in regulating onset or exacerbation of disease by modulating activation of T cells and other leukocytes in blood or lymph nodes. In order for T cells to cross the BBB and initiate CNS inflammation, they must first become activated and upregulate molecules involved in cellular adhesion and migration (Hauser and Oksenberg 2006, Bauer et al. 2009, Goverman et al. 2009, Holmann et al. 2011, Inoue et al. 2012). Interaction with other cells including NK cells could affect the activation status of T cells and therefore their ability to infiltrate the CNS, as well as their cytokine profile. In addition, NK cells have the ability to kill T cells that have been inappropriately

188 activated, or to kill Tregs and therefore prevent them from regulating effector cells (Rabinovich et al. 2003, Molinero et al. 2006, Cerboni et al. 2007, Roy et al. 2008, Lang et al. 2012, Nielsen et al. 2012). Therefore peripheral NK cells could be an important factor determining initiation of and relapses in MS, without ever entering the CNS.

An important area for future investigation is the regulation of IFNγ production in the CD56dim CD8+ CD57+ and CD56dim CD8- CD27- CD57- NK cell subsets. IFNγ upregulates expression of MHC class II, CIITA and co-stimulatory molecules, enhancing the ability of APCs to activate antigen-specific T cells (Figueiredo et al. 1989, Chang et al. 1995, Barrerra et al. 1997, Menèndez Iglesias et al. 1997, O’Keefe et al. 2001, Lee et al. 2006, Valledor et al. 2008). IFNγ produced by NK cells has a role in inducing Th1 polarisation, by upregulating T-bet in CD4+ T cells during priming and by enhancing DC production of IL-12 and IL-18, which preferentially induce IFNγ as opposed to Th2 cytokines (Afkarian et al .2002, Mocikat et al. 2003, Martín-Fontecha et al. 2004, Mailliard et al. 2005). Therefore production of IFNγ by NK cells could have important roles in regulating inflammation in the periphery and possibly the CNS. The CD56dim CD8+ CD57+ subset expressed significantly higher levels of mRNA encoding IFNγ as well as T-bet and NFATC2 which promote IFNγ transcription (Szabo et al. 2002, Macian 2005). Therefore it seems likely that this subset might participate in the inflammatory process by producing IFNγ, leading to activation of T cells, DCs, macrophages and microglia. IFNγ produced by NK cells might also contribute to migration of both NK cells and other leukocytes across the BBB by upregulating ICAM-1 on vascular endothelial cells, thereby promoting attachment of leukocytes expressing LFA-1 (Kebir et al. 2009).

However, in order to truly understand differences in IFNγ production, it will be important to compare the effects of different stimuli. IFNγ production is preferentially induced by cytokine stimulation in CD56bright NK cells and by ligation of activating receptors in CD56dim NK cells (Fauriat et al. 2010a). Antibody production in the CNS of many MS patients could trigger IFNγ production via CD16, therefore the response to ligation of this receptor in each subset should be tested in vitro. CD57+ NK cells have previously been reported to respond more strongly to stimulation via CD16 in terms of both IFNγ production and cytotoxic function (Lopez-Vergѐs et al. 2010). In

189 addition I found that the CD8+ CD57+ subset expressed higher levels of the CD3ζ adaptor protein that mediates signalling via CD16, NKp30 and NKp46 (Lanier et al. 1989, Pessino et al. 1998, Vitale et al. 1998, Pende et al. 1999). Therefore this subset would be expected to produce more IFNγ in response to stimulation of activating receptors, especially given the higher levels of IFNγ transcript in this subset. However, IFNγ production in response to cytokines may not show the same trend. Since CD56dim CD8- CD27- CD57- NK cells share features with the CD56bright subset, they might be expected to respond more effectively to stimulation with cytokines such as IL-12 and IL-18. IL-18R was not included in my PCR or flow cytometry analysis, but a previous study indicates that its expression is higher on CD57- NK cells (Lopez- Vergѐs et al. 2010). Given that NK cells might be exposed to IL-12 and IL-18 from activated APCs during MS, it would be worth investigating how each subset responds to stimulation with each of these cytokines both alone and in combination. Conversely, higher expression of STAT3, which is involved in the response to IFNγ, IL-6, IL-23 and other cytokines (Schindler and Plumlee 2008, Bluyssen et al. 2010), might enable the CD8+ CD57+ subset to be preferentially activated by certain MS- associated cytokines. STAT3 promotes survival and proliferation (Takeda et al. 1998, Akira 2000, Narimatsu et al. 2001), which might explain the expansion of this subset by pro-inflammatory cytokines in the blood of MS patients.

The ability to produce cytokines in response to direct interaction with other leukocytes may also be important in the context of MS. NK cell cytokines including IFNγ can be induced by interaction with APCs (Ferlazzo et al. 2002, Borg et al. 2004, Ferlazzo et al. 2004, Gerosa et al. 2005, Nedvetzki et al. 2007). NK cells which are activated in such a way form an immunological synapse with the other leukocyte and can direct secretion of cytokines and other factors towards the cell with which it is interacting, resulting in reciprocal activation (Borg et al. 2004, Ferlazzo et al. 2004, Semino et al. 2005). The ability to interact with DCs and to produce IFNγ in response to this interaction is likely to have a strong influence on the ability of each subset to promote T cell priming by APCs and Th1 polarisation. Therefore, it would be worth investigating the ability of each of these subsets to produce IFNγ and other cytokines such as TNFα and GM-CSF following in vitro incubation with APCs such as DCs, macrophages and microglia.

190

Another aspect of NK cell function that could be examined is responsiveness to viral and bacterial antigens via pattern recognition receptors such as TLRs. CD56dim CD8- CD57- NK cells exhibited higher expression of NF-κB, which is an important transcriptional mediator of TLR-induced signalling pathways (Kawai and Akira 2007). There might also be differences in other components of the signalling pathway or in expression of the receptors themselves. Stimulation with ligands for TLR and NOD receptors or with whole bacteria could be used to determine the ability of each subset to directly recognise infectious agents. NF-κB is also associated with induction of antioxidants that protect against apoptosis in response to ROS produced by phagocytes (Mattson et al. 1997, Javelaud et al. 2001, Djavaheri-Mergny et al. 2004). The CD56bright subset is reported to be more resistant to ROS (Harlin et al. 2007, Thorén et al. 2007) and it would be interesting to see if the higher level of NF-κB in the CD56dim CD8- CD27- CD57- subset confers upon it a similar resistance to that seen in CD56bright NK cells.

One of the most crucial functions of NK cells, which could have roles in both pathogenesis and suppression of inflammation, is cytotoxicity. I observed higher expression of perforin and granzymes and higher cytotoxicity towards K562 cells in the CD56dim CD8+ CD57+ subset. This more cytotoxic phenotype might indicate involvement of this subset in killing of oligodendrocytes and neurons, or of other leukocytes. However, NK cell subsets do not necessarily exhibit the same differences in cytotoxicity towards all types of target cell. For instance, the CD56bright subset is fully capable of killing activated T cells, despite its lower cytotoxicity towards classic NK cell-susceptible target cell lines (Bielekova et al. 2006, Jiang et al. 2011, Nielsen et al. 2012). It would be interesting to see if the CD56dim CD8- CD27- CD57- subset expresses granzyme K, a cytotoxic mediator associated primarily with the CD56bright subset, which is involved in killing of activated T cells (Hanna et al. 2004b, Wendt et al. 2006, Jiang et al. 2011). Comparing the ability of CD8+ CD57+ and CD8- CD57- CD56dim subsets to kill activated T cells would give more insight into their role in autoimmunity. In addition, there may be differences within the CD56bright population; perhaps, for instance, the CD56bright CD8- CD27- subset, which shows the greatest reduction in frequency in untreated MS patients compared to healthy controls and IFNβ-treated patients, is particularly important in mediating this function.

191

Human NK cells can also kill microglia in vitro (Lünemann et al. 2008) and killing of these CNS-resident APCs is thought to suppress disease in the EAE model (Hao et al. 2010, Leavenworth et al. 2010). Murine microglia are reported to be more susceptible to NK cell-mediated killing than other leukocytes due to their low expression of the ligand for the inhibitory receptor NKG2A (Hao et al. 2010). Induction of this ligand in microglia protects them from killing by NK cells (Hao et al. 2010), suggesting that killing of microglia, at least in mice, is mediated primarily by NKG2A+ NK cells. This is not necessarily the case in humans, since mouse and human NK cells differ in their repertoire of inhibitory receptors. It would be interesting to test whether NKG2A+ or KIR+ human NK cells are more effective at killing autologous microglia. I observed a higher percentage of NKG2A+ cells in the CD8- CD27- CD57- subset compared to the CD8+ CD57+ subset of CD56dim NK cells. Therefore if NKG2A expression is important for activation of the cytotoxic response to microglia, the CD8- CD27- CD57- subset might kill this subset more effectively despite lower expression of the main cytotoxic mediators.

The CD56dim CD8+ CD57+ subset may have a role in exacerbation of disease by killing neurons and oligodendrocytes. During MS, antibodies are deposited at sites of active demyelination and these antibodies are thought to induce macrophage phagocytosis of myelin and oligodendrocytes by triggering their Fc receptors (Scolding and Compston 1991, Storch et al. 1998, Lucchinetti et al. 2000, Breij et al. 2008). These antibodies might also trigger the Fc receptor CD16 on NK cells and if the myelin is coated with a sufficient density of immunoglobulin, this could trigger ADCC in the NK cell, leading to release of cytotoxic mediators that could kill the oligodendrocyte and possibly the underlying neuron. Human NK cells can also kill autologous oligodendrocytes by an NKG2D-dependent mechanism and oligodendrocytes in MS lesions express high levels of the NKG2D ligands MICA and MICB, meaning they are likely to be susceptible to NK cell lysis during MS if NK cells are present in the lesion (Morse et al. 2001, Saikali et al. 2007).

NK cells can also kill immature DCs, thereby promoting interaction of T cells with mature DCs instead, resulting in their activation (Carbone et al. 1999, Wilson et al. 1999, Spagiari et al. 2001, Ferlazzo et al. 2002, Della Chiesa et al. 2003, Vitale et al. 2005). Cytotoxic NK cells could also exacerbate disease by killing of Tregs, thereby

192 impairing an important immunosuppressive mechanism. Therefore the CD56dim CD8+ CD57+ and CD8- CD57- subsets would ideally be compared in killing assays using a range of target cells with relevance to MS, including CD4+ and CD8+ T effector cells, Tregs, macrophages, microglia, oligodendrocytes and neurons. Due to the large quantities of cells and reagents required to perform such experiments, this was beyond the scope of this study.

One of the more surprising results of the qRT-PCR array was the presence of CD3δ and CD3γ transcripts in CD56dim NK cells, particularly the CD8+ CD57+ subset. This observation could be investigated further by looking for protein expression. Clues as to the function of these receptor chains in NK cells might be gained by investigating their cellular localisation by fluorescence confocal microscopy and by searching for interactions with other proteins, including activating receptors and signalling molecules such as ZAP-70 and Syk.

Another curious observation from the PCR assays was the higher level of IL-12Rβ1 transcript in CD8+ CD57+ NK cells, whereas IL-12Rβ2 was expressed preferentially in the CD8- CD57- subset. Knockout of IL-12Rβ2 is reported to increase EAE severity, whereas IL-12Rβ1 deficiency ameliorates disease (Zhang et al. 2003a, Zhang et al. 2003b). This was presumed to indicate a more important role of Th17 compared to Th1 cells, since IL-23 signalling, which promotes Th17 differentiation, is dependent on IL-12Rβ1 and not IL-12Rβ2 (Parham et al. 2002, Hoeve et al. 2006). However, since both Th17 and Th1 cells are able to mediate EAE (Kroenke et al. 2008, Stromnes et al. 2008, Jager et al. 2009), it seems counterintuitive that preventing induction of T cell IFNγ production via IL-12Rβ2 would enhance EAE severity. Given the association of IL-12Rβ2 with immunoregulatory subsets of NK cells (Takahashi et al. 2001, Wendt et al. 2006, Perritt et al. 2008) and my observation that MS patients have a higher proportion of NK cell subsets that preferentially express IL-12Rβ1 rather than IL-12Rβ2, it is possible that some of the effects of IL-12Rβ2 and/or IL-12Rβ1 knockout are caused by loss of NK cell regulatory or effector function respectively.

It would also be interesting to investigate whether the frequency of NK cell subsets could be used as a predictor of future disease progression or responsiveness to therapy. The higher frequency of CD56bright NK cells in IFNβ-treated patients in

193 remission compared to those in relapse suggests that the frequency of this subset might correlate with disease activity. There is a pressing need for clinical application of biomarkers that will offer early indicators as to which patients will be responsive to IFN. The patients with the most striking expansion of CD56bright NKs were those with stable MS who had not had any relapses in the year prior to taking the sample. However, due to the relatively small number of samples, patients in remission with both stable and active disease were grouped together for analysis. Previous studies have reported that there is a relationship between expansion of CD56bright cells and response to therapy with either daclizumab or IFNβ (Wynn et al. 2010, Martín- Rodriguez et al. 2011). Extension of our study to compare stable and active patients could elucidate if there is a particular type of CD56bright NK cell that is particularly important in determining responsiveness to therapy. The CD8- CD27- population was the only CD56bright subset that was significantly expanded when the entire IFNβ- treated group was compared to the untreated patients. However, the result may have been different if only patients who were responding well to IFNβ therapy were used for comparison, since all four CD56bright subsets had higher frequencies in remission. Longitudinal studies could be performed in order to determine at what point expansion of CD8- CD27- CD57- subsets and contraction of the CD56dim CD8+ CD57+ subset occurs during therapy. If it occurs early on in therapy, it could be used as a predictor of responsiveness in order to determine if IFNβ is an appropriate choice of therapy for each individual patient. CIS patients could also be followed up to determine if NK cell subset frequencies are a predictor of future disease course. CIS patients with more CD56dim CD8+ CD57+ NK cells and fewer CD56bright NK cells might be more likely to progress rapidly to MS or to have more active disease. If a convenient predictor of disease severity was available it might help to determine the choice of therapy. Patients with an NK cell profile predisposing to severe disease might then be started on therapy earlier, which could be more effective at suppressing accumulation of disability.

Frequencies of these NK cell subsets could also be investigated in other autoimmune diseases such as arthritis, systemic lupus erythematosus (SLE) and type I diabetes, to see if there are similar changes to those observed in MS patients. However, evidence so far suggests that NK cell frequencies are altered in a different manner in different autoimmune diseases. CD16+ NK cell frequencies are reduced in the blood of SLE

194 patients but increased in arthritis patients (Li et al. 2010b). CD56bright NK cells actually have a higher frequency in the circulation of patients with SLE, which is thought to be due to increased serum levels of IFNα (Schepis et al. 2009). It would be interesting to see if these patients exhibit the preferential expansion of CD8- CD27- NK cells that I observed in IFNβ-treated MS patients. CD56bright NK cells are also found in the synovium of inflamed joints in arthritis patients and the phenotype of these cells could again be investigated further (Dalbeth and Callan 2002). Meanwhile, decreased numbers of circulating NK cells have been reported in psoriasis and Sjögren’s syndrome and my panel of antibodies could be used to investigate whether this is due to a selective deficiency in immunoregulatory NK cell subsets (Cameron et al. 2003, Izumi et al. 2006).

6.1 Summary

CD56bright NK cells have a significantly lower frequency in untreated MS patients compared to healthy controls, both as a percentage of NK cells and of all lymphocytes. I discovered that this decrease is particularly prominent in the CD8- CD27- subset of CD56bright NK cells. CIS patients have a similar frequency of this subset to MS patients, indicating that this is a change that occurs early on in the disease process. I have also identified a subset of NK cells that has a significantly higher frequency in MS patients: the CD56dim CD8+ CD57+ subset. This subset has an intermediate frequency in CIS patients, suggesting that its frequency is amplified by inflammatory disease processes. IFNβ has profound effects on the relative frequencies of different NK cell subsets. CD8- CD27- CD57- NK cell subsets, both CD56dim and CD56bright, are significantly expanded in IFNβ-treated patients, whereas the CD56dim CD8+ CD57+ subset is substantially reduced in frequency. CD56bright NK cells have previously been shown to kill activated T cells, therefore the decrease in CD56bright NK cells in MS patients is likely to have a role in promoting T cell-mediated inflammation. I have found that the CD56dim CD8- CD27- CD57- subset has some similarities with the CD56bright subset, such as higher levels of IL-12Rβ2 transcript, and hypothesise that it might have a similar function. In contrast, I have demonstrated that the CD56dim CD8+ CD57+ subset is highly cytotoxic and expresses higher levels of CD3ζ, which I hypothesise would make it more responsive to signals from activating receptors, enhancing its ability to release IFNγ and cytotoxic mediators upon encounter with target cells. It also expresses transcription factors associated with 195 induction of IFNγ and enhancement of survival and proliferation in response to IFNγ and other cytokines, which may allow this subset to be preferentially activated by elevated levels of pro-inflammatory cytokines in the blood of MS patients and to amplify inflammation by releasing further IFNγ that could promote Th1 polarisation and promote antigen presentation by upregulating components of the MHC class I and II pathways, APC co-stimulation molecules and the Th1 polarising cytokines IL-12 and IL-18. Actual and hypothetical differences between the CD8+ CD57+ and CD8- CD57- subsets of CD56dim CD27- NK cells are summarised in figure 6-i.

In conclusion, I have succeeded in identifying NK cell subsets with altered frequencies in MS patients and I have elucidated transcriptional and functional characteristics of the disease-associated CD56dim CD8+ CD57+ subset that could contribute to pathogenesis. Although the increased frequency of CD56dim CD8+ CD57+ and decreased frequency of CD56bright CD8- CD27- NK cells in MS patients does not directly prove that they have a role in promoting and suppressing disease respectively, the data presented here, combined with previous studies demonstrating the immunoregulatory function of CD56bright NK cells, are consistent with the hypothesis that the ratio of pro-inflammatory versus anti-inflammatory NK cell subsets is perturbed in MS patients.

196

Fig ure 6-i: Predicted signalling pathways and functions of CD8+ CD57+ and CD8- CD57- subsets of dim - CD56 CD27 NK cells. Signalling pathways likely to be associated with each subset were predicted based on differences observed in qRT-PCR comparisons. CD8+ CD57+ NK cells expressed higher levels of IFNγ and associated transcription factors (T-bet, NFATC2 and STAT3). During MS, IL-12 from APCs and IFNγ from T cells might induce IFNγ transcription via IL-12Rβ1 and type II IFN receptors respectively and IFNγ from NK cells could in turn activate APCs and T cells. CD8+ CD57+ NK cells express higher levels of CD3ζ which transduces signals from CD16 and the NCRs, NKp30 and NKp46, inducing IFNγ - - secretionKawai and T, release Akira of S. cytolytic Signaling granules to NF containing-kappaB granzyme by Toll A-like and B.receptors. CD8 CD57 Trends NK cells Mol express Med. IL - 18, 2007which Nov;13(11):460might act in an autocrine-9. manner together with IL-18 from DCs to induce surface expression of IL-12Rβ2, transcript of which is enriched in this subset. IL-18, TNF and TLR ligands could activate NF- Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman CG, Glimcher LH. A novel κB, which is also more abundant in this subset. NF-κB may be responsible for the upregulation of ICAM1 transcription factor, T-bet, directs Th1 lineage commitment. Cell. 2000 Mar observed17;100(6):655 in this subset-69. and could confer resistance to ROS. Expression of granzyme K and killing of activated T cells is hypothetical, based on similarity to the CD56bright subset.

197

7. REFERENCES

Abu-Khabar KS, Armstrong JA, Ho M. (1992) Type I interferons (IFN-alpha and - beta) suppress cytotoxin (tumor necrosis factor-alpha and lymphotoxin) production by mitogen-stimulated human peripheral blood mononuclear cell. J. Leukoc. Biol. 52:165-72.

Addison EG, North J, Bakhsh I, Marden C, Haq S, Al-Sarraj S, Malayeri R, Wickremasinghe RG, Davies JK, Lowdell MW. (2005) Ligation of CD8alpha on human natural killer cells prevents activation-induced apoptosis and enhances cytolytic activity. . 116:354-61.

Adler V, Yin Z, Fuchs SY, Benezra M, Rosario L, Tew KD, Pincus MR, Sardana M, Henderson CJ, Wolf CR, Davis RJ, Ronai Z. (1999) Regulation of JNK signalling by GSTp. EMBO J. 18:1321-34.

Afkarian M, Sedy JR, Yang J, Jacobson NG, Cereb N, Yang SY, Murphy TL, Murphy KM. (2002) T-bet is a STAT1-induced regulator of IL-12R expression in naïve CD4+ T cells. Nat. Immunol. 3:549-57.

Aharoni R, Kayhan B, Eilam R, Sela M, Arnon R. (2003) Glatiramer acetate-specific T cells in the brain express T helper 2/3 cytokines and brain-derived neurotrophic factor in situ. Proc. Natl. Acad. Sci. U.S.A. 100:14157-62.

Airas L, Saraste M, Rinta S, Elovaara I, Huang YH, Wiendl H; Finnish Multiple Sclerosis and Pregnancy Study Group. (2008) Immunoregulatory factors in multiple sclerosis patients during and after pregnancy: relevance of natural killer cells. Clin. Exp. Immunol. 151:235-43.

Akagawa KS, Takasuka N, Nozaki Y, Komuro I, Azuma M, Ueda M, Naito M, Takahashi K. (1996) Generation of CD1+RelB+ dendritic cells and tartrate-resistant acid phosphatase-positive osteoclast-like multinucleated giant cells form human monocytes. Blood. 88:4029-39.

Akdis CA, Joss A, Akdis M, Faith A, Blaser K. (2000) A molecular basis for T cell suppression by IL-10: CD28-associated IL-10 receptor inhibits CD28 tyrosine phosphorylation and phosphatidylinositol 3-kinase binding. FASEB J. 14:1666-8.

198

Akira S. (2000) Roles of STAT3 defined by tissue-specific gene targeting. Oncogene. 19:2607-11.

Allen M, Sandberg-Wollheim M, Sjögren K, Erlich HA, Petterson U, Gyllensten U. (1994) Association of susceptibility to multiple sclerosis in Sweden with HLA class II DRB1 and DQB1 alleles. Hum. Immunol. 39:41-8.

Allen RL, Raine T, Haude A, Trowsdale J, Wilson MJ. (2001) Leukocyte receptor complex-encoded immunomodulatory receptors show differing specificity for alternative HLA-B27 structures. J. Immunol. 167: 5543–7.

Allez M, Tieng V, Nakazawa A, Treton X, Pacault V, Dulphy N, Caillat-Zucman S, Paul P, Gornet JM, Douay C, Ravet S, Tamouza R, Charron D, Lémann M, Mayer L, Toubert A. (2007) CD4+NKG2D+ T cells in Crohn's disease mediate inflammatory and cytotoxic responses through MICA interactions. Gastroenterology. 132:2346-58.

Aloisi F, Penna G, Cerase J, Menendez IB, Adorini L. (1997) IL-12 production by central nervous system microglia is inhibited by astrocytes. J. Immunol. 159:1604–12.

Aloisi F, Ria F, Penna G, Adorini L. (1998) Microglia are more efficient than astrocytes in antigen processing and in Th1 but not Th2 cell activation. J. Immunol. 160:4671-80.

Aloisi F, De Simone R, Columba-Cabezas S, Penna G, Adorini L. (2000) Functional maturation of adult mouse resting microglia into an APC is promoted by granulocyte- macrophage colony-stimulating factor and interaction with Th1 cells. J. Immunol. 164:1705-12.

Alter M, Leibowitz U, Speer J. (1966) Risk of multiple sclerosis related to age at immigration to Israel. Arch. Neurol. 15:234–7.

Amirzargar A, Mytilineos J, Yousefipour A, Farjadian S, Scherer S, Opelz G, Ghaderi A. (1998) HLA class II (DRB1, DQA1 and DQB1) associated genetic susceptibility in Iranian multiple sclerosis (MS) patients. Eur. J. Immunogenet. 25:297-301.

Anderson P, Caligiuri M, Ritz J, Schlossman SF. (1989) CD3-negative natural killer cells express zeta TCR as part of a novel molecular complex. Nature. 341:159-62.

199

Andrade F, Roy S, Nicholson D, Thornberry N, Rosen A, Casciola-Rosen L. (1998) Granzyme B directly and efficiently cleaves several downstream caspase substrates: implications for CTL-induced apoptosis. Immunity. 8:451-60.

Andrade F, Fellows E, Jenner DE, Rosen A, Young CSH. (2007) Granzyme H destroys the function of critical adenoviral proteins required for viral DNA replication and granzyme B inhibition. EMBO J. 26:2148–2157.

Andzelm MM, Chen X, Krzewski K, Orange JS, Strominger JL. (2007) Myosin IIA is required for cytolytic granule exocytosis in human NK cells. J. Exp. Med. 204: 2285- 91.

Anegón I, Cuturi MC, Trinchieri G, Perussia B. (1988) Interaction of Fc receptor (CD16) ligands induces transcription of interleukin 2 receptor (CD25) and lymphokine genes and expression of their products in human natural killer cells. J. Exp. Med. 167:452-72.

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. (2006) Human NK cell education by inhibitory receptors for MHC class I. Immunity. 25:331-42.

Anthony DC, Miller KM, Fearn S, Townsend MJ, Opdenakker G, Wells GM, Clements JM, Chandler S, Gearing AJ, Perry VH. (1998) Matrix metalloproteinase expression in an experimentally-induced DTH model of multiple sclerosis in the rat CNS. J. Neuroimmunol. 87:62-72.

ANZgene (Australia and New Zealand Multiple Sclerosis Genetics Consortium). (2009) Genome-wide association study identifies new multiple sclerosis susceptibility loci on chromosomes 12 and 20. Nat. Genet. 41:824–8.

Arambu JA, Azzoni L, Rao A, Perussia B. (1995) Activation and expression of the nuclear factors of activated T cells, NFATp and NFATc, in human natural killer cells: regulation upon CD16 ligand binding. J. Exp. Med. 182: 801–810.

Aranami T, Miyake S, Yamamura T. (2006) Differential expression of CD11c by peripheral blood NK cells reflects temporal activity of multiple sclerosis. J. Immunol. 177: 5659-5667.

200

Aristimuño C, de Andrés C, Bartolomé M, de las Heras V, Martínez-Ginés ML, Arroyo R, Fernández-Cruz E, Sánchez-Ramón S. (2010) IFNbeta-1a therapy for multiple sclerosis expands regulatory CD8+ T cells and decreases memory CD8+ subset: a longitudinal 1-year study. Clin. Immunol. 134:148-57.

Arnon T, Lev M, Katz G, Chernobrov Y, Porgador A, Mandelboim O. (2001) Recognition of viral haemagglutinins by NKp44 but not by NKp30. Eur. J. Immunol. 31:2680-9.

Ascherio A, Munger KL, Lennette ET, Spiegelman D, Hernán MA, Olek MJ, Hankinson SE, Hunter DJ. (2001) Epstein-Barr virus antibodies and risk of multiple sclerosis: a prospective study. JAMA. 286:3083-8.

Ascherio A, Munger KL, Simon KC. (2010) Vitamin D and multiple sclerosis. Lancet Neurol. 9:599-612.

Asher RA, Morgenstern DA, Fidler PS, Adcock KH, Oohira A, Braistead JE, Levine JM, Margolis RU, Rogers JH, Fawcett JW. (2000) Neurocan is upregulated in injured brain and in cytokine-treated astrocytes. J. Neurosci. 20:2427–38.

Ashiru O, Boutet P, Fernández-Messina L, Agüera-González S, Skepper JN, Valés- Gómez M, Reyburn HT. (2010) Natural killer cell cytotoxicity is suppressed by exposure to the human NKG2D ligand MICA*008 that is shed by tumor cells in exosomes. Cancer Res. 70:481-9.

Azzoni L, Kamoun M, Salcedo TW, Kanakaraj P, Perussia B. (1992) Stimulation of Fc gamma RIIIA results in phospholipase C-gamma 1 tyrosine phosphorylation and p56lck activation. J. Exp. Med. 176:1745–50.

Babbe H, Roers A, Waisman A, Lassmann H, Goebels N, Hohlfeld R, Friese M, Schröder R, Deckert M, Schmidt S, Ravid R, Rajewsky K. (2000) Clonal Expansions of Cd8+ T Cells Dominate the T Cell Infiltrate in Active Multiple Sclerosis Lesions as Shown by Micromanipulation and Single Cell Polymerase Chain Reaction. J. Exp. Med. 192:393-404.

Backström E, Chambers BJ, Ho EL, Naiden OV, Mariotti R, Fremont DH, Yokoyama WM, Kristensson K, Ljunggren HG. (2003) Natural killer cell-mediated lysis of

201 dorsal root ganglia neurons via RAE1/NKG2D interactions. Eur. J. Immunol. 33: 92- 100.

Balaji KN, Schaschke N, Machleidt W, Catalfamo M, Henkart PA. (2002) Surface cathepsin B protects cytotoxic lymphocytes from self-destruction after degranulation. J. Exp. Med. 196:409-503.

Balashov KE, Rottman JB, Weiner HL, Hancock WW. (1999) CCR5(+) and CXCR3(+) T cells are increased in multiple sclerosis and their ligands MIP-1alpha and IP-10 are expressed in demyelinating brain lesions. Proc. Natl. Acad. Sci. U.S.A. 96:6873-8.

Baranzini SE. (2011) Revealing the genetic basis of multiple sclerosis: Are we there yet? Curr. Opin. Genet. Dev. 21: 317–24.

Barcellos LF, Oksenberg JR, Begovich AB, Martin ER, Schmidt S, Vittinghoff E, Goodin DS, Pelletier D, Lincoln RR, Bucher P, Swerdlin A, Pericak-Vance MA, Haines JL, Hauser SL, Multiple Sclerosis Genetics Group. (2003) HLA-DR2 dose effect on susceptibility to multiple sclerosis and influence on disease course. Am. J. Hum. Genet. 72:710-6.

Barnett MH, Prineas JW. (2004) Relapsing and remitting multiple sclerosis: Pathology of the newly forming lesion. Ann. Neurol. 55:458-68.

Baron JL, Madri JA, Ruddle NH, Hashim G, Janeway CA Jr. (1993) Surface expression of alpha 4 integrin by CD4 T cells is required for their entry into brain parenchyma. J. Exp. Med. 177:57-68.

Bar-Or A, Nuttal RK, Duddy M, Alter A, Kim HJ, Ifergan I, Pennington CJ, Bourgin P, Edwards DR, Yong, VW. (2003) Analyses of all matrix metalloproteinase members in leukocytes emphasize monocytes as major inflammatory mediators in multiple sclerosis. Brain. 126:2738-49.

Barrera LF, Kramnik I, Skamene E, Radzioch D (1997) I-A beta gene expression regulation in macrophages derived from mice susceptible or resistant to infection with M. bovis BCG. Mol. Immunol. 34:343-55.

202

Barry M, Heibein JA, Pinkoski MJ, Lee SF, Moyer RW, Green DR, Bleackley RC. (2000) Granzyme B short-circuits the need for caspase 8 activity during granule- mediated cytotoxic T-lymphocyte killing by directly cleaving Bid. Mol .Cell Biol. 20:3781-94.

Barun B, Bar-Or A. (2012) Treatment of multiple sclerosis with anti-CD20 antibodies. Clinical Immunology. 142: 31–37.

Baud V, Liu ZG, Bennett B, Suzuki N, Xia Y, Karin M. (1999) Signaling by proinflammatory cytokines: oligomerization of TRAF2 and TRAF6 is sufficient for JNK and IKK activation and target gene induction via an amino-terminal effector domain. Genes Dev. 13:1297-308.

Bauer S, Groh V, Wu J, Steinle A, Phillips JH, Lanier LL, Spies T. (1999) Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science. 285:727-9.

Bauer M, Brakebusch C, Coisne C, Sixt M, Wekerle H, Engelhardt B, Fässler R. (2009) Beta1 integrins differentially control extravasation of inflammatory cell subsets into the CNS during autoimmunity. Proc. Natl. Acad. Sci. U.S.A. 106:1920-5.

Baumann N and Pham Dinh D. (2001) Biology of Oligodendrocyte and Myelin in the Mammalian Central Nervous System. Physiol. Rev. 81: 871-927.

Baume DM, Caligiuri MA, Manley TJ, Daley JF, Ritz J. (1990) Differential expression of CD8 alpha and CD8 beta associated with MHC-restricted and non- MHC-restricted cytolytic effector cells. Cell. Immunol. 131:352-65.

Baume DM, Robertson MJ, Levine H, Manley TJ, Schow PW, Ritz J. (1992) Differential responses to interleukin 2 define functionally distinct subsets of human natural killer cells. Eur. J. Immunol. 22:1-6.

Becher B, Barker PA, Owens T, Antel JP. (1998) CD95-CD95L: Can the brain learn from the immune system? Trends Neurosci. 21:114–7.

203

Beck LA, Dalke S, Leiferman KM, Bickel CA, Hamilton R, Rosen H, Bochner BS, Schleimer RP. (1997) Cutaneous injection of RANTES causes eosinophil recruitment: comparison of nonallergic and allergic subjects. J. Immunol. 159: 296

Becker JC, Kolanus W, Lonnemann C, Schmidt RE. (1990) Human natural killer clones enhance in vitro antibody production by tumour necrosis factor alpha and gamma interferon. Scand. J. Immunol. 32:153-62.

Ben-Baruch A, Xu L, Young PR, Bengali K, Oppenheim JJ, Wang JM. (1995) Monocyte chemotactic protein-3 (MCP3) interacts with multiple leukocyte receptors. C-C CKR1, a receptor for macrophage inflammatory protein-1 alpha/Rantes, is also a functional receptor for MCP3. J. Biol. Chem. 270:22123-8.

Benczur M, Petrányl GG, Pálffy G, Varga M, Tálas M, Kotsy B, Földes I, Hollán SR. (1980) Dysfunction of natural killer cells in multiple sclerosis: a possible pathogenetic factor. Clin. Exp. Immunol. 39:657-62.

Ben-Nun A, Wekerle H, Cohen I. (1981) Vaccination against autoimmune encephalomyelitis with T-lymphocyte line cells reactive against myelin basic protein. Nature. 292:60-1.

Berahovich RD, Lai NL, Wei Z, Lanier LL, Schall TJ. (2006) Evidence for NK cell subsets based on chemokine receptor expression. J. Immunol. 177:7833-40.

Beresford PJ, Xia Z, Greenberg AH, Lieberman J. (1999) Granzyme A Loading Induces Rapid Cytolysis and a Novel Form of DNA Damage Independently of Caspase Activation. Immunity. 10: 585–595.

Bergamaschi L, Leone MA, Fasano ME, Guerini FR, Ferrante D, Bolognesi E, Barizzone N, Corrado L, Naldi P, Agliardi C, Dametto E, Salvetti M, Visconti A, Galimberti D, Scarpini E, Vercellino M, Bergamaschi R, Monaco F, Caputo D, Momigliano-Richiardi P, D'Alfonso S. (2010) HLA-class I markers and multiple sclerosis susceptibility in the Italian population. Genes Immun. 11: 173–80.

Berger T, Rubner P, Schautzer F, Egg R, Ulmer H, Mayringer I, Dilitz E, Deisenhammer F, Reindl M. (2003) Antimyelin Antibodies as a Predictor of

204

Clinically Definite Multiple Sclerosis after a First Demyelinating Event. N. Engl. J. Med. 349:139-45.

Bertolini JN, Sanderson CJ, Benson EM. (1993) Human interleukin-5 induces staphylococcal A Cowan 1 strain-activated human B cells to secrete IgM. Eur. J. Immunol. 23:398-402.

Béziat V, Descours B, Parizot C, Debré P, Vieillard V. (2010) NK cell terminal differentiation: correlated stepwise decrease of NKG2A and acquisition of KIRs. PLos One. 5:e11966.

Béziat V, Duffy D, Quoc SN, Le Garff-Tavernier M, Decocq J, Combadière B, Debré P, Vieillard V. (2011) CD56brightCD16+ NK cells: a functional intermediate stage of NK cell differentiation. J Immunol. 186:6753-61.

Bezman NA, Kim CC, Sun JC, Min-Oo G, Hendricks DW, Kamimura Y, Best JA, Goldrath AW, Lanier LL; The Immunological Genome Project Consortium, Gautier EL, Jakubzick C, Randolph GJ, Best AJ, Knell J, Goldrath A, Miller J, Brown B, Merad M, Jojic V, Koller D, Cohen N, Brennan P, Brenner M, Shay T, Regev A, Fletcher A, Elpek K, Bellemare-Pelletier A, Malhotra D, Turley S, Jianu R, Laidlaw D, Collins JJ, Narayan K, Sylvia K, Kang J, Gazit R, Rossi DJ, Kim F, Rao TN, Wagers A, Shinton SA, Hardy RR, Monach P, Bezman NA, Sun JC, Kim CC, Lanier LL, Heng T, Kreslavsky T, Painter M, Ericson J, Davis S, Mathis D, Benoist C. (2012) Molecular definition of the identity and activation of natural killer cells. Nat. Immunol. 13:1000-1009.

Biassoni R, Ferrini S, Prigione I, Moretta A, Long EO. (1988) CD3-negative lymphokine-activated cytotoxic cells express the CD3 epsilon gene. J. Immunol. 140:1685-9.

Biassoni R, Cantoni C, Falco M, Verdiani S, Bottino C, Vitale M, Conte R, Poggi A, Moretta A, Moretta L. (1996) 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. 183:645-50.

205

Bielekova B, Catalfamo M, Reichert-Scrivner S, Packer A, Cerna M, Waldmann TA, McFarland H, Henkart PA, Martin R. (2006) Regulatory CD56(bright) natural killer cells mediate immunomodulatory effects of IL-2Ralpha-targeted therapy (daclizumab) in multiple sclerosis. Proc. Natl. Acad. USA. 103:5941-6.

Bielekova B, Richert N, Herman ML, Ohayon J, Waldmann TA, McFarland H, Martin R, Blevins G. (2011) Intrathecal effects of daclizumab treatment of multiple sclerosis. Neurology. 77:1877-86.

Billadeau DD, Brumbaugh KM, Dick CJ, Schoon RA, Bustelo XR, Leibson PJ. (1998) The Vav-Rac1 pathway in cytotoxic lymphocytes regulates the generation of cell-mediated killing. J. Exp. Med. 188:549-59.

Binstadt BA, Billadeau DD, Jevremović D, Williams BL, Fang N, Yi T, Koretzky GA, Abraham RT, Leibson PJ. (1998) SLP-76 is a direct substrate of SHP-1 recruited to killer cell inhibitory receptors. J. Biol. Chem. 273: 27518.

Bischoff SC, Kriger M, Brunner T, Rot A, Vontscharner V, Baggliolini M, Dahinden CA. (1993) . RANTES and related chemokines activate human basophil granulocytes through different G-protein-coupled receptors. Eur. J. Immunol. 23: 761

Bitsch A, Schuchardt J, Bunkowski S, Kuhlmann T, Brück W. (2000) Acute axonal injury in multiple sclerosis. Correlation with demyelination and inflammation. Brain. 123:1174-83.

Björkström NK, Riese P, Heuts F, Andersson S, Fauriat C, Ivarsson MA, Björklund AT, Flodström-Tullberg M, Michaëlsson J, Rottenberg ME, Guzmán CA, Ljunggren HG, Malmberg KJ. (2010) Expression patterns of NKG2A, KIR, and CD57 define a process of CD56dim NK-cell differentiation uncoupled from NK-cell education. Blood. 116:3853-64.

Blanca IR, Bere EW, Young HA, Ortaldo JR. (2001) Human B cell activation by autologous NK cells is regulated by CD40-CD40 ligand interaction: role of memory B cells and CD5+ B cells. J. Immunol. 167:6132-9.

206

Blanco-Jerez C, Plaza JF, Masjuan J, Orensanz LM, Alvarez-Cermeño JC. (2002) Increased levels of IL-15 mRNA in relapsing--remitting multiple sclerosis attacks. J. Neuroimmunol. 128:90-4.

Bloomgren G, Richman S, Hotermans C, Subramanyam M, Goelz S, Natarajan A, Lee S, Plavina T, Scanlon JV, Sandrock A, Bozic C. (2012) Risk of natalizumab- associated progressive multifocal leukoencephalopathy. N. Engl. J. Med. 366:1870- 80.

Bluman EM, Bartynski KJ, Avalos BR, Caligiuri MA. (1996) Human natural killer cells produce abundant macrophage inflammatory protein-1 alpha in response to monocyte-derived cytokines. J Clin Invest. 97:2722-7.

Bluyssen HA, Rastmanesh MM, Tilburgs C, Jie K, Wesseling S, Goumans MJ, Boer P, Joles JA, Braam B. (2010) IFN gamma-dependent SOCS3 expression inhibits IL-6- induced STAT3 phosphorylation and differentially affects IL-6 mediated transcriptional responses in endothelial cells. Am. J. Physiol. Cell. Physiol. 299:C354- 62.

Boatright KM, Renatus M, Scott FL, Sperandio S, Shin H, Pedersen IM, Ricci JE, Edris WA, Sutherlin DP, Green DR, Salvesen GS. (2003) A unified model for apical caspase activation. Mol. Cell. 11:529-41.

Bodmer, J. L., Holler, N., Reynard, S., Vinciguerra, P., Schneider, P., Juo, P., Blenis, J., and Tschopp, J. TRAIL receptor-2 signals apoptosis through FADD and caspase-8 (2000) Nat. Cell Biol. 2, 241-243.

Bogdan C, Thuring H, Dlaska M, Rollinghoff M, Weiss G. (1997). Mechanism of suppression of macrophage nitric oxide release by IL-13 influence of the macrophage population. J. Immunol. 159: 4506

Bokoch GM. (1998) Caspase-mediated activation of PAK2 during apoptosis: proteolytic kinase activation as a general mechanism of apoptotic signal transduction? Cell Death & Differentiation. 5:637-645.

207

Boldin MP, Varfolomeev EE, Pancer Z, Mett IL, Camonis JH, Wallach D. (1995) A novel protein that interacts with the death domain of Fas/APO1 contains a sequence motif related to the death domain. J. Biol. Chem. 270:7795-8.

Borg C, Jalil A, Laderach D, Maruyama K, Wakasugi H, Charrier S, Ryffel B, Cambi A, Figdor C, Vainchenker W, Galy A, Caignard A, Zitvogel L. (2004) NK cell activation by dendritic cells (DCs) requires the formation of a synapse leading to IL- 12 polarization in DCs. Blood. 104:3267-75.

Borrego F, Ulbrecht M, Weiss EH, Coligan JE, Brooks AG. (1998) Recognition of Human Histocompatibility Leukocyte Antigen (HLA)-E Complexed with HLA Class I Signal Sequence-derived Peptides by CD94/NKG2 Confers Protection from Natural Killer Cell-mediated Lysis. J. Exp. Med. 187:813-8.

Borrego F, Alonso MC, Galiani MD, Carracedo J, Ramirez R, Ostos B, Peña J, Solana R. (1999) NK phenotypic markers and IL2 response in NK cells from elderly people. Exp. Gerontol. 34:253-65.

Boswell KL, James DJ, Esquibel JM, Bruinsma S, Shirakawa R, Horiuchi H, Martin TFJ. (2012) Munc13-4 reconstitutes calcium-dependent SNARE-mediated membrane fusion. J. Cell. Biol. 197:301-12.

Bottino C, Castriconi R, Pende D, Rivera P, Nanni M, Carnemolla B, Cantoni C, Grassi J, Marcenaro S, Reymond N, Vitale M, Moretta L, Lopez M, Moretta A. (2003) Identification of PVR (CD155) and Nectin-2 (CD112) as cell surface ligands for the human DNAM-1 (CD226) activating molecule. J. Exp. Med. 198:557-67.

Bovenschen N, de Koning PJ, Quadir R, Broekhuizen R, Damen JM, Froelich CJ, Slijper M, Kummer JA. (2008) NK cell protease granzyme M targets alpha-tubulin and disorganizes the microtubule network. J. Immunol. 180:8184-91.

Bowles JA, Weiner GJ. (2005) CD16 polymorphisms and NK activation induced by monoclonal antibody-coated target cells. J. Immunol. Methods. 304:88-99.

Boxel-Dezaire AHH, Zula JA, Xu Y, Ransohoff R, Jacobberger JW, Stark GR. (2010) Major differences in the response of primary human leukocyte subsets to IFN-β. J. Immunol. 185:5888-99.

208

Boz C, Ozmenoglu M, Velioglu S, Kilinc K, Orem A, Alioglu Z, Altunayoglu V. (2006) Matrix metalloproteinase-9 (MMP-9) and tissue inhibitor of matrix metalloproteinase (TIMP-1) in patients with relapsing-remitting multiple sclerosis treated with interferon beta. Clin. Neurol. Neurosurg. 108:124-8.

Braakman E, van Tunen A, Meager A, Lucas CJ. (1986) Natural cytotoxic activity in multiple sclerosis patients: defects in IL-2/interferon gamma-regulatory circuit. Clin. Exp. Immunol. 66:285-94.

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. (2009) The B7 family member B7-H6 is a tumor cell ligand for the activating natural killer cell receptor NKp30 in humans. J. Exp. Med. 206:1495-503.

Braud VM, Allan DS, O'Callaghan CA, Söderström K, D'Andrea A, Ogg GS, Lazetic S, Young NT, Bell JI, Phillips JH, Lanier LL, McMichael AJ. (1998) HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature. 391:795-9.

Bray PF, Bloomer LC, Salmon VC, Bagley MH, Larsen PD. (1983) Epstein-Barr virus infection and antibody synthesis in patients with multiple sclerosis. Arch. Neurol. 40:406-8.

Breij EC, Brink BP, Veerhuis R, van den Berg C, Vloet R, Yan R, Dijkstra CD, van der Valk P, Bö L. (2008) Homogeneity of active demyelinating lesions in established multiple sclerosis. Ann Neurol. 63:16-25.

Brinkmann V, Cyster JG, Hla T. (2004) FTY720: sphingosine 1-phosphate receptor-1 in the control of lymphocyte egress and endothelial barrier function. Am. J. Transplant. 4:1019–1025.

Brinkmann V. (2009) FTY720 (fingolimod) in multiple sclerosis: therapeutic effects in the immune and the central nervous system. Br. J. Pharmacol. 158:1173-1182.

Brocke S, Piercy C, Steinman L, Weissman IL, Veromaa T. (1999) Antibodies to CD44 and integrin alpha4, but not L-selectin, prevent central nervous system inflammation and experimental encephalomyelitis by blocking secondary leukocyte recruitment. Proc. Natl. Acad. Sci. U.S.A. 96:6896-901.

209

Brodin P, Lakshmikanth T, Johansson S, Kärre K, Höglund P. (2009) The strength of inhibitory input during education quantitatively tunes the functional responsiveness of individual natural killer cells. Blood. 113:2434-41.

Brooks AG, Borrego F, Posch PE, Patamawenu A, Scorzelli CJ, Ulbrecht M, Weiss EH, Coligan JE. (1999) Specific recognition of HLA-E, but not classical, HLA class I molecules by soluble CD94/NKG2A and NK cells. J. Immunol. 162:305-13.

Brown AM, Ransom BR. (2007) Astrocyte glycogen and brain energy metabolism. Glia. 55:1263–71.

Brown AC, Oddos S, Dobbie IM, Alakoskela JM, Parton RM, Eissmann P, Neil MA, Dunsby C, French PM, Davis I, Davis DM. (2011) Remodelling of cortical actin where lytic granules dock at natural killer cell immune synapses revealed by super- resolution microscopy. PLoS Biol. 9:e1001152.

Browne KA, Blink E, Sutton VR, Froelich CJ, Jans DA, Trapani JA. (1999) Cytosolic delivery of granzyme B by bacterial toxins: evidence that endosomal disruption, in addition to transmembrane pore formation, is an important function of perforin. Mol. Cell. Biol. 19:8604-15.

Broxmeyer HE, Sherry B, Cooper S, Lu L, Maze R, Beckmann MP, Cerami A, Ralph P. (1993) Comparative analysis of the human macrophage inflammatory protein family of cytokines (chemokines) on proliferation of human myeloid progenitor cells. Interacting effects involving suppression, synergistic suppression, and blocking of suppression. J. Immunol. 150:3448-58.

Brück W. (2005) The pathology of multiple sclerosis is the result of focal inflammatory demyelination with axonal damage. J. Neurol. 252:v3-9.

Bruhns P, Marchetti P, Fridman WH, Vivier E, Daëron M. (1999) Differential roles of N- and C-terminal immunoreceptor tyrosine-based inhibition motifs during inhibition of cell activation by killer cell inhibitory receptors. J. Immunol. 162:3168-75.

Brumbaugh KM, Binstadt BA, Billadeau DD, Schoon RA, Dick CJ, Ten RM, Leibson PJ. (1997) Functional Role for Syk Tyrosine Kinase in Natural Killer Cell–mediated Natural Cytotoxicity. J. Exp. Med. 186:1965-74.

210

Bryceson YT, March ME, Barber DF, Ljunggren HG, Long EO. (2005) Cytolytic granule polarization and degranulation controlled by different receptors in resting NK cells. J. Exp. Med. 202:1001-12.

Bundesen LQ, Scheel TA, Bregman BS, Kromer LF. (2003) Ephrin-B2 and EphB2 regulation of astrocyte-meningeal fibroblast interactions in response to spinal cord lesions in adult rats. J. Neurosci. 23:7789–7800.

Burkhardt JK, Hester S, Lapham CK, Argon Y. (1990) The lytic granules of natural killer cells are dual-function organelles combining secretory and pre-lysosomal compartments. J. Cell Biol. 111:2327-40.

Burshtyn DN, Shin J, Stebbins C, Long EO. (2000) Adhesion to target cells is disrupted by the killer cell inhibitory receptor. Current Biology. 10:777-80.

Burt BM, Plitas G, Nguyen HM, Stableford JA, Bamboat ZM, Dematteo RP. (2008) Circulating HLA-DR(+) natural killer cells have potent lytic ability and weak antigen- presenting cell function. Hum. Immunol. 69:469-74.

Bush TG, Puvanachandra N, Horner CH, Polito A, Ostenfeld T, Svendsen CN, Mucke L, Johnson MH, Sofroniew MV. (1999) Leukocyte infiltration, neuronal degeneration, and neurite outgrowth after ablation of scar-forming, reactive astrocytes in adult transgenic mice. Neuron. 23:297–308.

Caballero A, Alvés-León S, Papais-Alvarenga R, Fernández O, Navarro G, Alonso A. (1999) DQB1* 0602 confers genetic susceptibility to multiple sclerosis in Afro- Brazilians. Tissue Antigens. 54:524–526.

Cacalano NA, Sanden D, Johnston JA. (2001) Tyrosine-phosphorylated SOCS-3 inhibits STAT activation but binds to p120 RasGAP and activates Ras. Nat. Cell Biol. 3:460-5.

Caillier SJ, Briggs F, Cree BA, Baranzini SE, Fernandez-Viña M, Ramsay PP, Khan O, Royal W 3rd, Hauser SL, Barcellos LF, Oksenberg JR. (2008) Uncoupling the roles of HLA-DRB1 and HLA-DRB5 genes in multiple sclerosis. J. Immunol. 181:5473-80.

211

Calderon TM, Eugenin EA, Lopez L, Kumar SS, Hesselgesser J, Raine CS, Berman JW. (2006) A role for CXCL12 (SDF-1alpha) in the pathogenesis of multiple sclerosis: regulation of CXCL12 expression in astrocytes by soluble myelin basic protein. J. Neuroimmunol. 177:27–39.

Caligiuri MA, Zmuidzinas A, Manley TJ, Levine H, Smith KA, Ritz J. (1990) Functional consequences of interleukin 2 receptor expression on resting human lymphocytes. Identification of a novel natural killer cell subset with high affinity receptors. J. Exp. Med. 171:1509-26.

Cameron AL, Kirby B, Griffiths CE. (2003) Circulating natural killer cells in psoriasis. Br. J. Dermatol. 149:160-4.

Campbell JJ, Qin S, Unutmaz D, Soler D, Murphy KE, Hodge MR, Wu L, Butcher EC. (2001) Unique subpopulations of CD56+ NK and NK-T peripheral blood lymphocytes identified by chemokine receptor expression repertoire. J. Immunol. 166:6477-82.

Cao W, Xi X, Hao Z, Li W, Kong Y, Cui L, Ma C, Ba D, He W. (2007) RAET1E2, a Soluble Isoform of the UL16-binding Protein RAET1E Produced by Tumor Cells, Inhibits NKG2D-mediated NK Cytotoxicity. J. Biol. Chem. 282:18922-8.

Cao W, Bover L, Cho M, Wen X, Hanabuchi S, Bao M, Rosen DB, Wang YH, Shaw JL, Du Q, Li C, Arai N, Yao Z, Lanier LL, Liu YJ. (2009) Regulation of TLR7/9 responses in plasmacytoid dendritic cells by BST2 and ILT7 receptor interaction. J. Exp. Med. 206:1603–14.

Carbone E, Terrazzano G, Ruggiero G, Zanzi D, Ottaiano A, Manzo C, Kärre K, Zappacosta S. (1999) Recognition of autologous dendritic cells by human NK cells. Eur. J. Immunol. 29:4022-9.

Carbone E, Neri P, Mesuraca M, Fulciniti MT, Otsuki T, Pende D, Groh V, Spies T, Pollio G, Cosman D, Catalano L, Tassone P, Rotoli B, Venuta S. (2005) HLA class I, NKG2D, and natural cytotoxicity receptors regulate multiple myeloma cell recognition by natural killer cells. Blood. 105:251-8.

212

Carlson T, Kroenke M, Rao P, Lane TE, Segal B. (2008) The Th17-ELR+ CXC chemokine pathway is essential for the development of central nervous system autoimmune disease. J. Exp. Med. 205:811-23.

Carpintero R, Brandt KJ, Gruaz L, Molnarfi N, Lalive PH, Burger D. (2010) Glatiramer acetate triggers PI3Kdelta/Akt and MEK/ERK pathways to induce IL-1 receptor antagonist in human monocytes. Proc. Natl. Acad. Sci. U.S.A. 107:17692-97.

Carrithers MD, Visintin I, Kang SJ, Janeway CA Jr. (2000) Differential adhesion molecule requirements for immune surveillance and inflammatory recruitment. Brain. 123:1092-101.

Carson WE, Fehniger TA, Caligiuri MA. (1997) CD56bright natural killer cell subsets: characterization of distinct functional responses to interleukin-2 and the c-kit ligand. Eur. J. Immunol. 27:354-60.

Cassatella MA, Anegón I, Cuturi MC, Griskey P, Trinchieri G, Perussia B. (1989) Fc gamma R(CD16) interaction with ligand induces Ca2+ mobilization and phosphoinositide turnover in human natural killer cells. Role of Ca2+ in Fc gamma R(CD16)-induced transcription and expression of lymphokine genes. J. Exp. Med. 169:549-67.

Cassatella MA, Flynn RM, Amezaga MA, Bazzoni F, Vicentini F, Trinchieri G. (1990a) Interferon gamma induces in human neutrophils and macrophages expression of the mRNA for the high affinity receptor for monomeric IgG (Fc gamma R-I or CD64). Biochem. Biophys. Res. Commun. 170:582-8.

Cassatella MA, Bazzoni F, Flynn RM, Dusi S, Trinchieri G, Rossi F. (1990b) Molecular basis of interferon-gamma and lipopolysaccharide enhancement of phagocyte respiratory burst capability. Studies on the gene expression of several NADPH oxidase components. J. Biol. Chem. 265:20241-6.

Cella M, Longo A, Ferrara GB, Strominger JL, Colonna M. (1994) NK3-specific natural killer cells are selectively inhibited by Bw4- positive HLA alleles with isoleucine 80. J. Exp. Med. 180:1235-42.

213

Cella M, Fujikawa K, Tassi I, Kim S, Latinis K, Nishi S, Yokoyama W, Colonna M, Swat W. (2004) Differential requirements for Vav proteins in DAP10- and ITAM- mediated NK cell cytotoxicity. J. Exp. Med. 200:817-23.

Cella M, Fuchs A, Vermi W, Facchetti F, Otero K, Lennerz JK, Doherty JM, Mills JC, Colonna M. (2009) A human natural killer cell subset provides an innate source of IL-22 for mucosal immunity. Nature. 457:722-5.

Cerboni C, Zingoni A, Cippitelli M, Piccoli M, Frati L, Santoni A. (2007) Antigen- activated human T lymphocytes express cell-surface NKG2D ligands via an ATM/ATR-dependent mechanism and become susceptible to autologous NK- cell lysis. Blood. 110:606-15.

Cerboni C, Ardolino M, Santoni A, Zingoni A. (2009) Detuning CD8+ T lymphocytes by down-regulation of the activating receptor NKG2D: role of NKG2D ligands released by activated T cells. Blood. 113:2955-64.

Chabot S, Yong VW. (2000) Interferon beta-1b increases interleukin-10 in a model of T cell-microglia interaction: Relevance to MS. Neurology. 55:1497-1505.

Chan FK, Chun HJ, Zheng L, Siegel RM, Bui KL, Lenardo MJ. (2000) A domain in TNF receptors that mediates ligand-independent receptor assembly and signaling. Science. 288:2351-4.

Chan A, Magnus T, Gold R. (2001) Phagocytosis of apoptotic inflammatory cells by microglia and modulation by different cytokines: mechanism for removal of apoptotic cells in the inflamed nervous system. Glia. 33:87-95.

Chan A, Weilbach FX, Toyka KV, Gold R. (2005) Mitoxantrone induces cell death in peripheral blood leucocytes of multiple sclerosis patients. Clin. Exp. Immunol. 139:152-8.

Chan A, Hong DL, Atzberger A, Kollnberger S, Filer AD, Buckley CD, McMichael A, Enver T, Bowness P. (2007) CD56bright human NK cells differentiate into CD56dim cells: role of contact with peripheral fibroblasts. J. Immunol. 179:89-94.

214

Chandler S, Coates R, Gearing A, Lury J, Wells G, Bone E. (1995) Matrix metalloproteinases degrade myelin basic protein. Neurosci. Lett. 201:223-226.

Chandler S, Cossins J, Lury J, Wells G. (1996) Macrophage metalloelastase degrades matrix and myelin proteins and processes a tumour necrosis factor-alpha fusion protein. Biochem. Biophys. Res. Commun. 228:421–9.

Chang CH, Flavell RA. (1995) Class II transactivator regulates the expression of multiple genes involved in antigen presentation. J. Exp. Med. 181:765-767.

Chang DW, Xing Z, Capacio VL, Peter ME, Yang X. (2003) Interdimer processing mechanism of procaspase-8 activation. EMBO J. 22:4132-42.

Chang J-H, Cha H-R, Lee D-S, Seo KY, Kweon M-N (2010) 1,25-Dihydroxyvitamin D3 Inhibits the Differentiation and Migration of TH17 Cells to Protect against Experimental Autoimmune Encephalomyelitis. PLoS One. 5: e12925.

Chapman TL, Heikeman AP, Bjorkman PJ. (1999) The inhibitory receptor LIR-1 uses a common binding interaction to recognize class I MHC molecules and the viral homolog UL18. Immunity. 11: 603–13.

Chattopadhyay PK, Betts MR, Price DA, Gostick E, Horton H, Roederer M, De Rosa SC. (2009) The cytolytic enzymes granyzme A, granzyme B, and perforin: expression patterns, cell distribution, and their relationship to cell maturity and bright CD57 expression. J. Leukoc. Biol. 85:88-97.

Chen M, Valenzuela RM, Dhib-Jalbut S. (2003) Glatiramer acetate-reactive T cells produce brain-derived neurotrophic factor. J. Neurol. Sci. 215:37-44.

Chen ZJ. (2005) Ubiquitin signalling in the NF-kappaB pathway. Nat. Cell Biol. 7:758-65.

Chidrawar SM, Khan N, Chan YL, Nayak L, Moss PA. (2006) Ageing is associated with a decline in peripheral blood CD56bright NK cells. Immun. Ageing. 3:10.

Chinnaiyan AM, O'Rourke K, Tewari M, Dixit VM. (1995) FADD, a novel death domain-containing protein, interacts with the death domain of Fas and initiates apoptosis. Cell. 81:505-12.

215

Chuang YH, Lian ZX, Tsuneyama K, Chiang BL, Ansari AA, Coppel RL, Gershwin E. (2006) Increased killing activity and decreased cytokine production in NK cells in patients with primary biliary cirrhosis. J. Autoimmunity. 26:232-40.

Chun J, Hartung HP. (2010) Mechanism of action of oral fingolimod (FTY720) in multiple sclerosis. Clin. Neuropharmacol. 33:91-101.

Clanet M, Radue EW, Kappos L, Hartung HP, Hohlfeld R, Sandberg-Wollheim M, Kooijmans-Coutinho MF, Tsao EC, Sandrock AW; European IFNbeta-1a (Avonex) Dose-Comparison Study Investigators. (2002) A randomized, double-blind, dose- comparison study of weekly interferon beta-1a in relapsing MS. Neurology. 59:1507- 17.

Clerico M, Contessa G, Durelli L. (2007) Interferon-beta1a for the treatment of multiple sclerosis. Expert Opin. Biol. Ther. 7:535-42.

Cocco E, Sardu C, Pieroni E, Valentini M, Murru R, Costa G, Tranquilli S, Frau J, Coghe G, Carboni N, Floris M, Contu P, Marrosu MG. (2012) HLA-DRB1-DQB1 haplotypes confer susceptibility and resistance to multiple sclerosis in Sardinia. PLoS One. 7:e33972.

Cocks BG, de Waal Malefyt R, Galizzi JP, de Vries JE, Aversa G. (1993) IL-13 induces proliferation and differentiation of human B cells activated by the CD40 ligand. Int. Immunol. 5:657-63.

Codarri L, Gyülvészi G, Tosevski V, Hesske L, Fontana A, Magnenet L, Suter T, Becher B. (2011) RORgt drives production of the cytokine GM-CSF in helper T cells, which is essential for the effector phase of autoimmune neuroinflammation. Nat. Immunol. 12: 560-568.

Coelho RP, Payne SG, Bittman R, Spiegel S, Sato-Bigbee C. (2007) The immunomodulator FTY720 has a direct cytoprotective effect in oligodendrocyte progenitors. J. Pharmacol. Exp. Ther. 323:626-635.

Cohen BA, Mikol DD. (2004) Mitoxantrone treatment of multiple sclerosis: safety considerations. Neurology. 63:S28-32.

216

Cohen JA, Barkhof F, Comi G, Hartung HP, Khatri BO, Montalban X, Pelletier J, Capra R, Gallo P, Izquierdo G, Tiel-Wilck K, de Vera A, Jin J, Stites T, Wu S, Aradhye S, Kappos L; TRANSFORMS Study Group. (2010) Oral fingolimod or intramuscular interferon for relapsing multiple sclerosis. N. Engl. J. Med. 362:402-15.

Colantonio AD, Bimber BN, Neidermyer WJ Jr, Reeves RK, Alter G, Altfeld M, Johnson RP, Carrington M, O'Connor DH, Evans DT. (2011) KIR polymorphisms modulate peptide-dependent binding to an MHC class I ligand with a Bw6 motif. PLoS Pathog. 7:e1001316.

Coleman ML, Sahai EA, Yeo M, Bosch M, Dewar A, Olson MF. (2001) Membrane blebbing during apoptosis results from caspase-mediated activation of ROCK I. Nat. Cell Biol.3:339-45.

Coles AJ, Fox E, Vladic A, Gazda SK, Brinar V, Selmaj KW, Bass AD, Wynn DR, Margolin DH, Lake SL, Moran S, Palmer J, Smith MS, Compston DA. (2011) Alemtuzumab versus interferon β-1a in early relapsing-remitting multiple sclerosis: post-hoc and subset analyses of clinical efficacy outcomes. Lancet Neurol. 10:338-48.

Coles AJ, Fox E, Vladic A, Gazda SK, Brinar V, Selmaj KW, Skoromets A, Stolyarov I, Bass A, Sullivan H, Margolin DH, Lake SL, Moran S, Palmer J, Smith MS, Compston DA. (2012) Alemtuzumab more effective than interferon β-1a at 5- year follow-up of CAMMS223 clinical trial. Neurology. 78:1069-78.

Colonna M, Borsellino G, Falco M, Battista Ferrara G, Strominger J. (1993) HLA-C is the inhibitory ligand that determines dominant resistance to lysis by NK1- and NK2-specific natural killer cells. Proc. Natl. Acad. Sci. USA. 90:12000-4.

Colonna M, Samaridis J, Cella M, Angman L, Allen RL, O'Callaghan CA, Dunbar R, Ogg GS, Cerundolo V, Rolink A. (1998) Human myelomonocytic cells express an inhibitory receptor for classical and nonclassical MHC class I molecules. J. Immunol. 160:3096-100.

Colucci F, Caligiuri MA, Di Santo JP. (2003) What does it take to make a natural killer? Nat. Rev. Immunol. 3: 413-425.

217

Comabella M, Río J, Espejo C, Ruiz de Villa M, Al-Zayat H, Nos C, Deisenhammer F, Baranzini SE, Nonell L, López C, Julià E, Oksenberg JR, Montalban X. (2009) Changes in matrix metalloproteinases and their inhibitors during interferon-beta treatment in multiple sclerosis. Clin. Immunol. 130:145-50.

Cooley S, Xiao F, Pitt M, Gleason M, McCullar V, Bergemann TL, McQueen KL, Guethlein LA, Parham P, Miller JS. (2007) A subpopulation of human peripheral blood NK cells that lacks inhibitory receptors for self-MHC is developmentally immature. Blood. 110:578-86.

Cooper MA, Fehniger TA, Turner SC, Chen KS, Ghaheri BA, Ghayur T, Carson WE, Caligiuri MA. (2001a) Human natural killer cells: a unique innate immunoregulatory role for the CD56(bright) subset. Blood. 97:3146-51.

Cooper MA, Fehniger TA, Caligiuri MA. (2001b) The biology of human natural killer cell subsets. Trends Immunol. 22:633-40.

Corcione A, Casazza S, Ferretti E, Giunti D, Zappia E, Pistorio A, Gambini C, Mancardi GL, Uccelli A, Pistoia V. (2004) Recapitulation of B cell differentiation in the central nervous system of patients with multiple sclerosis. Proc. Natl. Acad. Sci. U.S.A. 101:11064-9.

Correale J, Ysrraelit MC, Gaitán MI. (2009) Immunomodulatory effects of Vitamin D in multiple sclerosis. Brain. 132:1146-60.

Cosman D, Fanger N, Borges L, Kubin M, Chin W, Peterson L, Hsu ML. (1997) A novel immunoglobulin superfamily receptor for cellular and viral MHC class I molecules. Immunity. 7:273-82

Cosman D, Müllberg J, Sutherland CL, Chin W, Armitage R, Fanslow W, Kubin M, Chalupny NJ. (2001) ULBPs, Novel MHC Class I-Related Molecules, Bind to CMV Glycoprotein UL16 and Stimulate NK Cytotoxicity through the NKG2D Receptor. Immunity 14:123-33.

Cossburn M, Pace AA, Jones J, Ali R, Ingram G, Baker K, Hirst C, Zajicek J, Scolding N, Boggild M, Pickersgill T, Ben-Shlomo Y, Coles A, Robertson NP. (2011)

218

Autoimmune disease after alemtuzumab treatment for multiple sclerosis in a multicenter cohort. Neurology. 77:573-9.

Costello RT, Sivori S, Marcenaro E, Lafage-Pochitaloff M, Mozziconacci MJ, Reviron D, Gastaut JA, Pende D, Olive D, Moretta A. (2002) Defective expression and function of natural killer cell- triggering receptors in patients with acute myeloid leukemia. Blood. 99:3661-7.

Coxon A, Tang T, Mayadas TN. (1999) Cytokine-activated Endothelial Cells Delay Neutrophil Apoptosis In Vitro and In Vivo: A Role for Granulocyte/Macrophage Colony-stimulating Factor. J. Exp. Med. 190:923-33.

Cramer LA, Nelson SL, Klemsz MJ. (2000) Synergistic induction of the Tap-1 gene by IFN-γ and lipopolysaccharide in macrophages is regulated by STAT1. J. Immunol. 165:3190.

Crawford MP, Yan SX, Ortega SB, Mehta RS, Hewitt RE, Price DA, Stastny P, Douek DC, Koup RA, Racke MK, Karandikar NJ. (2004) High prevalence of autoreactive, neuroantigen-specific CD8+ T cells in multiple sclerosis revealed by novel flow cytometric assay. Blood. 103:4222-31.

Cross AH, Stark JL, Lauber J, Ramsbottom MJ, Lyons J-A. (2006) Rituximab reduces B cells and T cells in cerebrospinal fluid of multiple sclerosis patients. J. Neuroimmunol. 180 :63–70.

Crucian B, Dunne P, Friedman H, Ragsdale R, Pross S, Widen R. (1995) Alterations in levels of CD28-/CD8+ suppressor cell precursor and CD45RO+/CD4+ memory T lymphocytes in the peripheral blood of multiple sclerosis patients. Clin. Diagn. Lab. Immunol. 2:249-52.

Cruz-Munoz ME, Dong Z, Shi X, Zhang S, Veillette A. (2009) Influence of CRACC, a SLAM family receptor coupled to the adaptor EAT-2, on natural killer cell function. Nat. Immunol . 10:297-305.

Cua DJ, Sherlock J, Chen Y, Murphy CA, Joyce B, Seymour B, Lucian L, To W, Kwan S, Churakova T, Zurawski S, Wiekowski M, Lira SA, Gorman D, Kastelein

219

RA, Sedgwick JD. (2003) Interleukin-23 rather than interleukin-12 is the critical cytokine for autoimmune inflammation of the brain. Nature. 421:744-8.

Cullen SP, Afonina IS, Donadini R, Lüthi AU, Medema JP, Bird PI, Martin SJ. (2009) Nucleophosmin is cleaved and inactivated by the cytotoxic granule protease granzyme M during natural killer cell-mediated killing. J. Biol. Chem. 284:5137-47.

Culley FJ, Johnson M, Evans JH, Kumar S, Crilly R, Casasbuenas J, Schnyder T, Mehrabi M, Deonarain MP, Ushakov DS, Braud V, Roth G, Brock R, Köhler K, Davis DM. (2009) Natural Killer Cell Signal Integration Balances Synapse Symmetry and Migration. PLoS Biol. 7:e1000159.

Cupedo T, Crellin NK, Papazian N, Rombouts EJ, Weijer K, Grogan JL, Fibbe WE, Cornelissen JJ, Spits H. (2009) Human fetal lymphoid tissue-inducer cells are interleukin 17-producing precursors to RORC+ CD127+ natural killer-like cells. Nat. Immunol. 10:66-74.

Cuturi MC, Anegón I, Sherman F, Loudon R, Clark SC, Perussia B, Trinchieri G. (1989) Production of hematopoietic colony-stimulating factors by human natural killer cells. J. Exp. Med. 169:569-83.

DaFonseca CJ, Shu F, Zhang JJ. (2001) Identification of two residues in MCM5 critical for the assembly of MCM complexes and Stat1-mediated transcription activation in response to IFN-γ. Proc. Natl Acad. Sci. USA. 98:3034–9.

Dalbeth N, Callan MF. (2002) A subset of natural killer cells is greatly expanded within inflamed joints. Arthritis Rheum. 46:1763-72.

Dalbeth N, Gundle R, Davies RJ, Lee YC, McMichael AJ, Callan MF. (2004) CD56bright NK cells are enriched at inflammatory sites and can engage with monocytes in a reciprocal program of activation. J. Immunol. 173:6418-26.

D’Andrea A, Aste-Amezaga M, Valiante NM, Ma X, Kubin M, Trinchieri G. (1993) Interleukin 10 (IL-10) Inhibits Human Lymphocyte Interferon 3,-Production by Suppressing Natural Killer Cell Stimulatory Factor/IL-12 Synthesis in Accessory Cells. J. Exp. Med. 178:1041-8.

220

Daniel C, Sartory NA, Zahn N, Radeke HH, Stein JM. (2008) Immune modulatory treatment of trinitrobenzene sulfonic acid colitis with calcitriol is associated with a change of a T helper (Th) 1/Th17 to a Th2 and regulatory T cell profile. J. Pharmacol. Exp. Ther. 324: 23–33.

Dardalhon V, Schubart AS, Reddy J, Meyers JH, Monney L, Sabatos CA, Ahuja R, Nguyen K, Freeman GJ, Greenfield EA, Sobel RA, Kuchroo VK. (2005) CD226 is specifically expressed on the surface of Th1 cells and regulates their expansion and effector functions. J. Immunol. 175:1558–65.

Darlington PJ, Podjaski C, Horn KE, Costantino S, Blain M, Saikali P, Chen Z, Baker KA, Newcombe J, Freedman M, Wiseman PW, Bar-Or A, Kennedy TE, Antel JP. (2008) Innate immune-mediated neuronal injury consequent to loss of astrocytes. J. Neuropathol. Exp. Neurol. 67:590-9.

Daugherty BL, Siciliano SJ, DeMartino JA, Malkowitz L, Sirotina A, Springer MS. (1996) Cloning, expression, and characterization of the human eosinophil eotaxin receptor. J. Exp. Med. 183:2349-54.

Davis DM, Chiu I, Fassett M, Cohen GB, Mandelboim O, Strominger JL. The human natural killer cell immune synapse. (1999) Proc. Natl. Acad. Sci. USA. 96:15062-7.

Dean G, Elian M. (1997) Age at immigration to England of Asian and Caribbean immigrants and the risk of developing multiple sclerosis. J. Neurol. Neurosurg. Psychiatry. 63: 565–8. de Andrés C, Aristimuño C, de Las Heras V, Martínez-Ginés ML, Bartolomé M, Arroyo R, Navarro J, Giménez-Roldán S, Fernández-Cruz E, Sánchez-Ramón S. (2007) Interferon beta-1a therapy enhances CD4+ regulatory T-cell function: an ex vivo and in vitro longitudinal study in relapsing-remitting multiple sclerosis. J. Neuroimmunol. 182:204-11. de Andrés C, Teijeiro R, Alonso B, Sánchez-Madrid F, Martínez ML, Guzmán de Villoria J, Fernández-Cruz E, Sánchez-Ramón S. (2012) Long-term decrease in VLA- 4 expression and functional impairment of dendritic cells during natalizumab therapy in patients with multiple sclerosis. PLoS One. 7:e34103.

221

Deb DK, Sassano A, Lekmine F, Majchrzak B, Verma A, Kambhampati S, Uddin S, Rahman A, Fish EN, Platanias LC. (2003) Activation of protein kinase C delta by IFN-gamma. J. Immunol. 171:267–273.

Deckert M, Tartare-Deckert S, Couture C, Mustelin T, Altman A. (1996) Functional and physical interactions of Syk family kinases with the Vav proto-oncogene product. Immunity. 5:591-604.

Defrance T, Carayon P, Billian G, Guillmot JC, Minty A, Caput D, Ferrara P. (1994) Interleukin 13 is a B cell stimulating factor. J. Exp. Med. 179: 135

De Groot CJ, Ruuls SR, Theeuwes JW, Dijkstra CD, Van der Valk P. (1997) Immunocytochemical characterization of the expression of inducible and constitutive isoforms of nitric oxide synthase in demyelinating multiple sclerosis lesions. J. Neuropath. Exp. Neurol. 56:10-20.

De Jager PL, Rossin E, Pyne S, Tamayo P, Ottoboni L, Viglietta V, Weiner M, Soler D, Izmailova E, Faron-Yowe L, O'Brien C, Freeman S, Granados S, Parker A, Roubenoff R, Mesirov JP, Khoury SJ, Hafler DA, Weiner HL. (2008) Cytometric profiling in multiple sclerosis uncovers patient population structure and a reduction of CD8low cells. Brain. 131:1701-11.

De Jager PL, Jia X, Wang J, de Bakker PI, Ottoboni L, Aggarwal NT, Piccio L, Raychaudhuri S, Tran D, Aubin C, Briskin R, Romano S; International MS Genetics Consortium, Baranzini SE, McCauley JL, Pericak-Vance MA, Haines JL, Gibson RA, Naeglin Y, Uitdehaag B, Matthews PM, Kappos L, Polman C, McArdle WL, Strachan DP, Evans D, Cross AH, Daly MJ, Compston A, Sawcer SJ, Weiner HL, Hauser SL, Hafler DA, Oksenberg JR. (2009a) Meta-analysis of genome scans and replication identify CD6, IRF8 and TNFRSF1A as new multiple sclerosis susceptibility loci. Nat. Genet. 41:776-82.

De Jager PL, Baecher-Allan C, Maier LM, Arthur AT, Ottoboni L, Barcellos L, McCauley JL, Sawcer S, Goris A, Saarela J, Yelensky R, Price A, Leppa V, Patterson N, de Bakker PI, Tran D, Aubin C, Pobywajlo S, Rossin E, Hu X, Ashley CW, Choy E, Rioux JD, Pericak-Vance MA, Ivinson A, Booth DR, Stewart GJ, Palotie A, Peltonen L, Dubois B, Haines JL, Weiner HL, Compston A, Hauser SL, Daly MJ,

222

Reich D, Oksenberg JR, Hafler DA. (2009b) The role of the CD58 locus in multiple sclerosis. Proc. Natl. Acad. Sci. U.S.A. 106:5264-9. de la Salle HD, Hanau D, Fricker A, Urlacher A, Kelly J, Salamero SH, Powis L, Donato H, Bausinger M. Laforet M, et al. (1994) Homozygous human TAP peptide transporter mutation in HLA class I deficiency. Science 265:237.

Della Chiesa M, Vitale M, Carlomagno S, Ferlazzo G, Moretta L, Moretta A. (2003) The natural killer cell-mediated killing of autologous dendritic cells is confined to a cell subset expressing CD94/NKG2A, but lacking inhibitory killer Ig-like receptors. Eur. J. Immunol. 33:1657-66.

DeLorenze GN, Munger KL, Lennette ET, Orentreich N, Vogelman JH, Ascherio A. (2005) Temporal relationship between elevation of Epstein-Barr virus antibody titers and initial onset of neurological symptoms in multiple sclerosis. JAMA. 293:2496 – 500.

Del Prete G, De Carli M, Almerigogna F, Giudizi MG, Biagiotti R, Romagnani S. (1993) Human IL-10 is produced by both type 1 helper (Th1) and type 2 helper (Th2) T cell clones and inhibits their antigen-specific proliferation and cytokine production. J. Immunol. 150:353-60.

De Maria A, Fogli M, Mazza S, Basso M, Picciotto A, Costa P, Congia S, Mingari MC, Moretta L. (2007) Increased natural cytotoxicity receptor expression and relevant IL-10 production in NK cells from chronically infected viremic HCV patients. Eur. J. Immunol. 37:445-55.

De Maria A, Bozzano F, Cantoni C, Moretta L. (2011) Revisiting human natural killer cell subset function revealed cytolytic CD56(dim)CD16+ NK cells as rapid producers of abundant IFN-gamma on activation. Proc. Natl. Acad. Sci. USA. 108:728-32.

Deng Y, Ren X, Yang L, Lin Y, Wu X. (2003) A JNK-dependent pathway is required for TNFalpha-induced apoptosis. Cell. 115:61-70.

Deniz G, Erten G, Kücüksezer UC, Kocacik D, Karagiannidis C, Aktas E, Akdis CA, Akdis M. (2008) Regulatory NK cells suppress antigen-specific T cell responses. J. Immunol. 180:850-7.

223

Deogracias R, Yazdani M, Dekkers MP, Guy J, Ionescu MC, Vogt KE, Barde YA. (2012) Fingolimod, a sphingosine-1 phosphate receptor modulator, increases BDNF levels and improves symptoms of a mouse model of Rett syndrome. Proc. Natl. Acad. Sci. U.S.A. 109:14230-5.

De Smedt T, Smith J, Baum P, Fanslow W, Butz E, Maliszewski C. (2002) Ox40 costimulation enhances the development of T cell responses induced by dendritic cells in vivo. J. Immunol. 168:661-70.

De Stefano N, Matthews PM, Fu L, Narayanan S, Stanley J, Francis GS, Antel JP, Arnold DL. (1998) Axonal damage correlates with disability in patients with relapsing-remitting multiple sclerosis. Results of a longitudinal magnetic resonance spectroscopy study. Brain. 121: 1469-77. de Waal Malefyt R, Haanen J, Spits H, Roncarolo MG, te Velde A, Figdor C, Johnson K, Kastelein R, Yssel H, de Vries JE. (1991) Interleukin 10 (IL-10) and viral IL-10 strongly reduce antigen-specific human T cell proliferation by diminishing the antigen-presenting capacity of monocytes via downregulation of class II major histocompatibility complex expression. J. Exp. Med. 174:915-24. de Waal Malefyt R, Figdor R, Huijbens R, Mohen-Peterson S, Bennet B, Culpepper J, Dang W, Zurawski G, de Vries JE. (1993a) Effects of IL-13 on phenotype, cytokine production, and cytotoxic function of human monocytes: comparison with IL-4 and modulation by IFN-γ or IL-10. J. Immunol. 151: 6370. de Waal Malefyt R, Yssel H, de Vries JE. (1993b) Direct effects of IL-10 on subsets of human CD4+ T cell clones and resting T cells. Specific inhibition of IL-2 production and proliferation. J. Immunol. 150:4754-65.

Diefenbach A, Jamieson AM, Liu SD, Shastri N, Raulet DH. (2000) Ligands for the murine NKG2D receptor: expression by tumor cells and activation of NK cells and macrophages. Nat. Immunol. 1:119-26.

Disanto G, Sandve GK, Berlanga-Taylor AJ, Ragnedda G, Morahan JM, Watson CT, Giovannoni G, Ebers GC, Ramagopalan SV. (2012) Vitamin D receptor binding,

224 chromatin states and association with multiple sclerosis. Hum. Mol. Genet. 21:3575- 86.

Djavaheri-Mergny M, Javelaud D, Wietzerbin J, Besançon F. (2004) NF-kappaB activation prevents apoptotic oxidative stress via an increase of both thioredoxin and MnSOD levels in TNFalpha-treated Ewing sarcoma cells. FEBS Lett. 578:111-5.

Donepudi M, Sweeney AM, Briand C, Grütter MG. (2003) Insights into the regulatory mechanism for caspase-8 activation. Mol. Cell. 11:543–549.

Dorfman JR, Raulet DH. (1996) Major histocompatibility complex genes determine natural killer cell tolerance. Eur. J. Immunol. 26:151-5.

Döring A, Wild M, Vestweber D, Deutsch U, Engelhardt B. (2007) E- and P-selectin are not required for the development of experimental autoimmune encephalomyelitis in C57BL/6 and SJL mice. J. Immunol. 179:8470-9.

Douglass J, Dhami D, Bulpitt M, Lindley IJ, Shute J, Church MK, Holgate ST. (1996) Intradermal challenge with interleukin-8 causes tissue oedema and neutrophil accumulation in atopic and non-atopic human subjects. Clin. Exp. Allergy. 26:1371-9.

Draghi M, Pashine A, Sanjanwala B, Gendzekhadze K, Cantoni C, Cosman D, Moretta A, Valiante NM, Parham P. (2007) NKp46 and NKG2D Recognition of Infected Dendritic Cells Is Necessary for NK Cell Activation in the Human Response to Influenza Infection. J. Immunol. 178:2688-98.

D’Souza SD, Alinauskas K A, Antel JP. (1996). Ciliary neurotrophic factor selectively protects human oligodendrocytes from tumor necrosis factor-mediated injury. J. Neurosci. Res. 43:289–98.

Durrenberger PF, Ettorre A, Kamel F, Webb LV, Sim M, Nicholas RS, Malik O, Reynolds R, Boyton RJ, Altmann DM. (2012) Innate immunity in multiple sclerosis white matter lesions: expression of natural cytotoxicity triggering receptor 1 (NCR1). J. Neuroinflammation. 2012 9:1.

225

Duszczyszyn DA, Williams JL, Mason H, Lapierre Y, Antel J, Haegert DG. (2010) Thymic involution and proliferative T-cell responses in multiple sclerosis. J. Neuroimmunol. 221:73-80.

Eissens DN, Spanholtz J, van der Meer A, van Cranenbroek B, Dolstra H, Kwekkeboom J, Preijers FW, Joosten I. (2012) Defining early human NK cell developmental stages in primary and secondary lymphoid tissues. PLoS One. 7:e30930.

Eklund EA, Jalava A, Kakar R. (1998) PU.1, interferon regulatory factor 1, and interferon consensus sequence-binding protein cooperate to increase gp91(phox) expression. J. Biol. Chem. 273:13957–65.

Ellmerich S, Mycko M, Takacs K, Waldner H, Wahid FN, Boyton RJ, King RHM, Smith PA, Amor S, Herlihy AH, Hewitt RE, Jutton M, Price DA, Hafler DA, Kuchroo VK, Altmann DM. (2005) High Incidence of Spontaneous Disease in an HLA-DR15 and TCR Transgenic Multiple Sclerosis Model. J. Immunol. 174:1938- 1946.

El-Shazly E, Lefebvre PP. (2011) Modulation of NK cell autocrine-induced eosinophil chemotaxis by interleukin-15 and vitamin D(3): a possible NK-eosinophil crosstalk via IL-8 in the pathophysiology of . Mediators Inflamm. 2011:373589.

Enari M, Sakahira H, Yokoyama H, Okawa K, Iwamatsu A, Nagata S. (1998) A caspase-activated DNase that degrades DNA during apoptosis, and its inhibitor ICAD. Nature. 391:43-50.

Endt J, McCann FE, Almeida CR, Urlaub D, Leung R, Pende D, Davis DM, Watzl C. (2007) Inhibitory Receptor Signals Suppress Ligation-Induced Recruitment of NKG2D to GM1-Rich Membrane Domains at the Human NK Cell Immune Synapse. J. Immunol. 178:5606-11.

Engelhardt B. (2008) Immune cell entry into the central nervous system: involvement of adhesion molecules and chemokines. J. Neurol. Sci. 274:23-6.

226

Esin S, Batoni G, Counoupas C, Stringaro A, Brancatisano FL, Colone M, Maisetta G, Florio W, Arancia G, Campa M. (2008) Direct Binding of Human NK Cell Natural Cytotoxicity Receptor NKp44 to the Surfaces of Mycobacteria and Other Bacteria. Infect. Immun. 76:1719-27.

Etzensperger R, McMahon RM, Jones EY, Fugger L. (2008) Dissection of the multiple sclerosis associated DR2 haplotype. J. Autoimmun. 31:201-7.

Evans JH, Horowitz A, Mehrabi M, Wise EL, Pease JE, Riley EM, Davis DM. (2011) A distinct subset of human NK cells expressing HLA-DR expand in response to IL-2 and can aid immune responses to BCG. Eur. J. Immunol. 41:1924-33.

Fadda L, O'Connor GM, Kumar S, Piechocka-Trocha A, Gardiner CM, Carrington M, McVicar DW, Altfeld M. (2011) Common HIV-1 peptide variants mediate differential binding of KIR3DL1 to HLA-Bw4 molecules. J. Virol. 85:5970-4.

Fan Z, Beresford PJ, Zhang D, and Lieberman J. (2002) HMG2 Interacts with the nucleosome assembly protein SET and is a target of the cytotoxic T-lymphocyte protease granzyme A. Mol. Cell. Biol. 22:2810–2820.

Fan Z, Beresford PJ, Zhang D, Xu Z, Novina CD, Yoshida A, Pommier Y, and Lieberman J. (2003a) Cleaving the oxidative repair protein Ape1 enhances cell death mediated by granzyme A. Nat. Immunol. 4:145–153.

Fan Z, Beresford PJ, Oh DY, Zhang D, Lieberman J. (2003b) Tumor Suppressor NM23-H1 Is a Granzyme A-Activated DNase during CTL-Mediated Apoptosis, and the Nucleosome Assembly Protein SET Is Its Inhibitor. Cell. 112:659-72.

Fan Y, Yu Y, Shi Y, Sun W, Xie M, Ge N, Mao R, Chang A, Xu G, Schneider MD, Zhang H, Fu S, Qin J, Yang J. (2010) Lysine 63-linked polyubiquitination of TAK1 at lysine 158 is required for tumor necrosis factor alpha- and interleukin-1beta-induced IKK/NF-kappaB and JNK/AP-1 activation. J. Biol. Chem. 285:5347-60.

Faure M, Long EO. (2002) KIR2DL4 (CD158d), an NK cell-activating receptor with inhibitory potential. J. Immunol. 168:6208-14.

227

Fauriat C, Long EO, Ljunggren HG, Bryceson YT. (2010a) Regulation of human NK- cell cytokine and chemokine production by target cell recognition. Blood. 115:2167- 76

Fauriat C, Ivarsson MA, Ljunggren HG, Malmberg KJ, Michaëlsson J. (2010b) Education of human natural killer cells by activating killer cell immunoglobulin-like receptors. Blood. 115:1166-74.

Fehniger TA, Shah MH, Turner MJ, VanDeusen JB, Whitman SP, Cooper MA, Suzuki K, Wechser M, Goodsaid F, Caligiuri MA. (1999) Differential cytokine and chemokine gene expression by human NK cells following activation with IL-18 or IL- 15 in combination with IL-12: implications for the innate immune response. J. Immunol. 162:4511-20.

Fehniger TA, Cooper MA, Nuovo GJ, Cella M, Facchetti F, Colonna M, Caligiuri MA. (2003) CD56bright natural killer cells are present in human lymph nodes and are activated by T cell-derived IL-2: a potential new link between adaptive and innate immunity. Blood. 101:3052–7.

Feldmann J, Callebaut I, Raposo G, Certain S, Bacq D, Dumont C, Lambert N, Ouachée-Chardin M, Chedeville G, Tamary H, Minard-Colin V, Vilmer E, Blanche S, Le Deist F, Fischer A, de Saint Basile G. (2003) Munc13–4 is essential for cytolytic granules fusion and is mutated in a form of familial hemophagocytic lymphohistiocytosis (FHL3). Cell. 115:461-473.

Ferlazzo G, Tsang ML, Moretta L, Melioli G, Steinman RM, Munz C. (2002) Human Dendritic Cells Activate Resting Natural Killer (NK) Cells and Are Recognized via the NKp30 Receptor by Activated NK Cells. J. Exp. Med. 195:343-51.

Ferlazzo G, Thomas D, Lin SL, Goodman K, Morandi B, Muller WA, Moretta A, Münz C. (2004) The abundant NK cells in human secondary lymphoid tissues require activation to express killer cell Ig-like receptors and become cytolytic. J. Immunol. 172:1455–62.

228

Fernández O, Fernández V, Alonso A, Caballero A, Luque G, Bravo M, León A, Mayorga C, Leyva L, de Ramón E. (2004) DQB1*0602 allele shows a strong association with multiple sclerosis in patients in Malaga, Spain. J. Neurol. 251:440-4.

Fernandez NC, Treiner E, Vance RE, Jamieson AM, Lemieux S, Raulet DH. (2005) A subset of natural killer cells achieves self-tolerance without expressing inhibitory receptors specific for self-MHC molecules. Blood. 105:4416-23.

Fernández O, Agüera J, Izquierdo G, Millán-Pascual J, Ramió I Torrentà L, Oliva P, Argente J, Berdei Y, Soler JM, Carmona O, Errea JM, Farrés J; Group on Adherence to IFNb-1b in Spain. (2012) Adherence to interferon β-1b treatment in patients with multiple sclerosis in Spain. PLoS One. 7:e35600.

Fertsch D, Schoenberg DR, Germain RN, Tou JY, Vogel SN. (1987) Induction of macrophage Ia antigen expression by rIFN-gamma and down-regulation by IFN- alpha/beta and dexamethasone are mediated by changes in steady-state levels of Ia mRNA. J. Immunol. 139:244–249.

Fidler JM, DeJOy SQ, Gibbons JJ Jr. (1986a) Selective immunomodulation by the antineoplastic agent mitoxantrone. I. Suppression of B lymphocyte function. J. Immunol. 137:727–32.

Fidler JM, DeJoy SQ, Smith FR III, Gibbons JJ Jr. (1986b) Selective immunomodulation by the antineoplastic agent mitoxantrone. II. Nonspecific adherent suppressor cells derived from mitoxantrone-treated mice. J. Immunol. 136:2747–54.

Figueiredo F, Koerner TJ, Adams DO. (1989) Molecular mechanisms regulating the expression of class II histocompatibility molecules on macrophages. Effects of inductive and suppressive signals on gene transcription. J. Immunol. 143:3781-6.

Flechter S, Vardi J, Pollak L, Rabey JM. (2002) Comparison of glatiramer acetate (Copaxone) and interferon beta-1b (Betaferon) in multiple sclerosis patients: an open- label 2-year follow-up. J. Neurol. Sci. 197:51-5.

Fogdell-Hahn A, Ligers A, Gronning M, Hillert J, Olerup O. (2000) Multiple sclerosis: a modifying influence of HLA class I genes in an HLA class II associated autoimmune disease. Tissue Antigens. 55: 140–8.

229

Foster CA, Howard LM, Schweitzer A, Persohn E, Hiestand PC, Balatoni B, Reuschel R, Beerli C, Schwartz M, Billich A. (2007) Brain penetration of the oral immunomodulatory drug FTY720 and its phosphorylation in the central nervous system during experimental autoimmune encephalomyelitis: consequences for mode of action in multiple sclerosis. J. Pharmacol. Exp. Ther. 323:469-475.

Fox EJ. (2006) Management of worsening multiple sclerosis with mitoxantrone: a review. Clin. Ther. 28:461-74.

Freedman DM, Dosemici M, Alavanja MC. (2000) Mortality from multiple sclerosis and exposure to residential and occupational solar radiation: a case-control study based on death certificates. Occup. Environ. Med. 57:418–21.

Freud AG, Becknell B, Roychowdhury S, Mao HC, Ferketich AK, Nuovo GJ, Hughes TL, Marburger TB, Sung J, Baiocchi RA, Guimond M, Caligiuri MA. (2005) A human CD34(+) subset resides in lymph nodes and differentiates into CD56bright natural killer cells. Immunity. 22:295-304.

Freud AG, Yokohama A, Becknell B, Lee MT, Mao HC, Ferketich AK, Caligiuri MA. (2006) Evidence for discrete stages of human natural killer cell differentiation in vivo. J. Exp. Med. 203:1033-43.

Frey M, Packianathan NB, Fehniger TA, Ross ME, Wang WC, Stewart CC, Caligiuri MA, Evans SS. (1998) Differential Expression and Function of L-Selectin on CD56bright and CD56dim Natural Killer Cell Subsets. J. Immunol. 161:400-408.

Frisullo G, Nociti V, Iorio R, Patanella AK, Caggiula M, Marti A, Sancricca C, Angelucci F, Mirabella M, Tonali PA, Batocchi AP. (2009) Regulatory T cells fail to suppress CD4T+-bet+ T cells in relapsing multiple sclerosis patients. Immunology. 127:418-28.

Froelich CJ, Orth K, Turbov J, Seth P, Gottlieb R, Babior B, Shah GM, Bleackley RC, Dixit VM, Hanna W. (1996) New paradigm for lymphocyte granule-mediated cytotoxicity. Target cells bind and internalize granzyme B, but an endosomolytic agent is necessary for cytosolic delivery and subsequent apoptosis. J. Biol. Chem. 271:29073-9.

230

Fu L, Matthews PM, De Stefano N, Worsley KJ, Narayanan S, Francis GS, Antel JP, Wolfson C, Arnold DL. (1998) Imaging axonal damage of normal-appearing white matter in multiple sclerosis. Brain. 121:103-13.

Fusco C, Guerini FR, Nocera G, Ventrella G, Caputo D, Valentino MA, Agliardi C, Gallotti J, Morra VB, Florio C, Clerici M, Lombardi ML. (2010) KIRs and their HLA ligands in remitting-relapsing multiple sclerosis. J. Neuroimmunol. 229:232-7.

Galandrini R, Palmieri G, Piccoli M, Frati L, Santoni A. (1999) Role for the Rac1 exchange factor Vav in the signaling pathways leading to NK cell cytotoxicity. J. Immunol. 162:3148-52.

Gamble JR, Elliott MJ, Jaipargas E, Lopez AF, Vadas MA. (1989) Regulation of human monocyte adherence by granulocyte-macrophage colony-stimulating factor. Proc. Natl. Acad. Sci. U. S. A. 86: 7169–7173.

Gan Y, Liu R, Bomprezzi R, Shi F-D. (2012) Antibody to α4 integrin suppresses natural killer cells infiltration in central nervous system in experimental autoimmune encephalomyelitis. J. Neuroimmunol. 247:9-15.

Gao JL, Kuhns DB, Tiffany HL, McDermott D, Li X, Francke U, Murphy PM. (1993) Structure and functional expression of the human macrophage inflammatory protein 1 alpha/RANTES receptor. J. Exp. Med. 177:1421-7.

Gao GF, Tormo J, Gerth UC, Wyer JR, McMichael AJ, Stuart DI, Bell JI, Jones EY, Jakobsen BK. (1997) Crystal structure of the complex between human CD8alpha(alpha) and HLA-A2. Nature. 387:630-634.

Gao GF, Willcox BE, Wyer JR, Boulter JM, O'Callaghan CA, Maenaka K, Stuart DI, Jones EY, Van Der Merwe PA, Bell JI, Jakobsen BK. (2000) Classical and nonclassical class I major histocompatibility complex molecules exhibit subtle conformational differences that affect binding to CD8alphaalpha. J. Biol. Chem. 275:15232-8.

Gao N, Dang T, Yuan D. (2001) IFN-gamma-dependent and -independent initiation of switch recombination by NK cells. J. Immunol. 167:2011-8.

231

Gao N, Dang T, Dunnick WA, Collins JT, Blazar BR, Yuan D. (2005) Receptors and counterreceptors involved in NK-B cell interactions. J. Immunol. 174:4113-9.

Garcia-Iglesias T, Del Toro-Arreola A, Albarran-Somoza B, Del Toro-Arreola S, Sanchez-Hernandez PE, Ramirez-Dueñas MG, Balderas-Peña LM, Bravo-Cuellar A, Ortiz-Lazareno PC, Daneri-Navarro A. (2009) Low NKp30, NKp46 and NKG2D expression and reduced cytotoxic activity on NK cells in cervical cancer and precursor lesions. BMC Cancer. 9:186.

Garg A, Barnes PF, Porgador A, Roy S, Wu S, Nanda JS, Griffith DE, Girard WM, Rawal N, Shetty S, Vankayalapati R. (2006) Vimentin Expressed on Mycobacterium tuberculosis-Infected Human Monocytes Is Involved in Binding to the NKp46 Receptor. J. Immunol. 177:6192-8.

Gascoyne DM, Long E, Veiga-Fernandes H, de Boer J, Williams O, Seddon B, Coles M, Kioussis D, Brady HJ. (2009) The basic leucine zipper transcription factor E4BP4 is essential for natural killer cell development. Nat. Immunol. 10:1118-24.

Gasser S, Orsulic S, Brown EJ, Raulet DH. (2005) The DNA damage pathway regulates innate immune system ligands of the NKG2D receptor. Nature. 436:1186- 90.

Genain CP, Cannella B, Hauser SL, Raine CS. (1999) Identification of autoantibodies associated with myelin damage in multiple sclerosis. Nat. Med. 5:170-5.

Genç K, Dona DL, Reder AT. (1997) Increased CD80+ B cells in active multiple sclerosis, and reversal by IFNβ-1b therapy. J. Clin. Invest. 99:2664–71.

Germain M, Affar EB, D'Amours D, Dixit VM, Salvesen GS, Poirier GG. (1999) Cleavage of automodified Poly(ADP-ribose) polymerase during apoptosis. Evidence for involvement of caspase-7. J. Biol. Chem. 274:28379–28384.

Gerosa F, Baldani-Guerra B, Nisii C, Marchesini V, Carra G, Trinchieri G. (2002) Reciprocal activating interaction between natural killer cells and dendritic cells. J. Exp. Med. 195:327-33.

232

Gerosa F, Gobbi A, Zorzi P, Burg S, Briere F, Carra G, Trinchieri G. (2005) The reciprocal interaction of NK cells with plasmacytoid or myeloid dendritic cells profoundly affects innate resistance functions. J. Immunol. 174:727-34.

Gibbings D and Dean Befus A. (2009) CD4 and CD8: an inside-out coreceptor model for innate immune cells. J. Leukoc. Biol. 86:251-259.

Gijbels K, Proost P, Masure S, Carton H, Billiau A, Opdenakker G. (1993) Gelatinase B is present in the cerebrospinal fluid during experimental autoimmune encephalomyelitis and cleaves myelin basic protein. J. Neurosci. Res. 36:432-40.

Gimenez MA, Sim JE, Russell JH. (2004) TNFR1-dependent VCAM-1 expression by astrocytes exposes the CNS to destructive inflammation. J. Neuroimmunol. 151:116– 25.

Gimsa U, Peter SV, Lehmann K, Bechmann I, Nitsch R. (2000) Axonal damage induced by invading T cells in organotypic central nervous system tissue in vitro: involvement of microglial cells. Brain Pathol. 10:365-77.

Gimsa U, Oren A, Pandiyan P, Teichmann D, Bechmann I, Nitsch R, Brunner- Weinzierl MC. (2004) Astrocytes protect the CNS: antigen-specific responses are inhibited by astrocyte-induced upregulation of CTLA-4 (CD152). J. Mol. Med. 82:364–72.

Gismondi A, Morrone S, Humphries MJ, Piccoli M, Frati L, Santoni A. (1991) Human natural killer cells express VLA-4 and VLA-5, which mediate their adhesion to fibronectin. J. Immunol. 146:384-92.

Giuliani F, Goodyer CG, Antel JP, Yong VW. (2003) Vulnerability of human neurons to T cell-mediated cytotoxicity. J. Immunol. 171: 368-379.

Glabinski AR, Tani M, Strieter RM, Tuohy VK, Ransohoff RM. (1997) Synchronous synthesis of alpha- and beta-chemokines by cells of diverse lineage in the central nervous system of mice with relapses of chronic experimental autoimmune encephalomyelitis. Am. J. Pathol. 150:617-30.

233

Glimcher L, Shen FW, Cantor H. (1977) Identification of a cell-surface antigen selectively expressed on the natural killer cell. J Exp Med. 145:1-9.

Goffette S, van Pesch V, Vanoverschelde JL, Morandini E, Sindic CJ. (2005) Severe delayed heart failure in three multiple sclerosis patients previously treated with mitoxantrone. J. Neurol. 252:1217-22.

Gold R, Linington C, Lassman H. (2006) Understanding pathogenesis and therapy of multiple sclerosis via animal models: 70 years of merits and culprits in experimental autoimmune encephalomyelitis research. Brain. 129:1953-71.

Goldberg LD, Edwards NC, Fincher C, Doan QV, Al-Sabbagh A, Meletiche DM. (2009) Comparing the cost-effectiveness of disease-modifying drugs for the first-line treatment of relapsing-remitting multiple sclerosis. J. Manag. Care Pharm. 15:543- 55.

Goping IS, Sawchuk T, Underhill DA, Bleackley RC. (2006) Identification of alpha- tubulin as a granzyme B substrate during CTL-mediated apoptosis. J Cell Sci. 119:858-65.

Goverman J. (2009) Autoimmune T cell responses in the central nervous system. Nat. Rev. Immunol. 9:393-407.

Graham DB, Cella M, Giurisato E, Fujikawa K, Miletic AV, Kloeppel T, Brim K, Takai T, Shaw AS, Colonna M, Swat W. (2006) Vav1 controls DAP10-mediated natural cytotoxicity by regulating actin and microtubule dynamics. J. Immunol. 177:2349-55.

Gramaglia I, Jember A, Pippig SD, Weinberg AD, Killeen N, Croft M. (2000) The OX40 costimulatory receptor determines the development of CD4 memory by regulating primary clonal expansion. J. Immunol. 165:3043-50.

Graves SS, Bramhall J, Bonavida B. (1986) Studies on the lethal hit stage of natural killer cell-mediated cytotoxicity. I. Both phorbol ester and ionophore are required for release of natural killer cytotoxic factors (NKCF), suggesting a role for protein kinase C activity. J. Immunol. 137:1977-84.

234

Gregory SG, Schmidt S, Seth P, Oksenberg JR, Hart J, Prokop A, Caillier SJ, Ban M, Goris A, Barcellos LF, Lincoln R, McCauley JL, Sawcer SJ, Compston DA, Dubois B, Hauser SL, Garcia-Blanco MA, Pericak-Vance MA, Haines JL; Multiple Sclerosis Genetics Group. (2007) Interleukin 7 receptor alpha chain (IL7R) shows allelic and functional association with multiple sclerosis. Nat. Genet. 39:1083-91.

Griffin MD, Lutz W, Phan VA, Bachman LA, McKean DJ, Kumar R. (2001) Dendritic cell modulation by 1alpha,25 dihydroxyvitamin D3 and its analogs: a vitamin D receptor-dependent pathway that promotes a persistent state of immaturity in vitro and in vivo. Proc. Natl. Acad. Sci. U.S.A. 98: 6800–5.

Groh V, Rhinehart R, Secrist H, Bauer S, Grabstein KH, Spies T. (1999) Broad tumor-associated expression and recognition by tumor-derived gamma,delta T cells of MICA and MICB. Proc. Natl. Acad. Sci. USA. 96:6879-84.

Groh V, Wu J, Yee C, Spies T. (2002) Tumour-derived soluble MIC ligands impair expression of NKG2D and T-cell activation. Nature. 419:734-8.

Groh V, Bruhl A, El-Gabalawy H, Nelson JL, Spies T. (2003) Stimulation of T cell autoreactivity by anomalous expression of NKG2D and its MIC ligands in rheumatoid arthritis. Proc. Natl. Acad. Sci. U.S.A. 100:9452-7.

Groux H, Bigler M, de Vries JE, Roncarolo MG. (1998) Inhibitory and Stimulatory Effects of IL-10 on Human CD8 + T Cells. J. Immunol. 160:3188-93.

Grzywacz B, Kataria N, Kataria N, Blazar BR, Miller JS, Verneris MR. (2011) Natural killer-cell differentiation by myeloid progenitors. Blood. 117:3548-58.

Gu Y, Sarnecki C, Aldape RA, Livingston DJ, Su MSS. Cleavage of poly(ADP- ribose) polymerase by interleukin-1-beta converting enzyme and its homologs TX and Nedd-2. (1995) J. Biol. Chem. 270: 18715–8.

Gumperz JE, Litwin V, Phillips JH, Lanier LL, Parham P. (1995) 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. 181: 1133–44.

235

Guo Y, Chen J, Zhao T, Fan Z. (2008) Granzyme K degrades the redox/DNA repair enzyme Ape1 to trigger oxidative stress of target cells leading to cytotoxicity. Mol. Immunol. 45:2225-35.

Gupta JW, Kubin M, Hartman L, Cassatella M, Trinchieri G. (1992) Induction of expression of genes encoding components of the respiratory burst oxidase during differentiation of human myeloid cell lines induced by tumor necrosis factor and gamma-interferon. Cancer Res. 52:2530-7.

Haahr S, Höllsberg P. (2006) Multiple sclerosis is linked to Epstein-Barr virus infection. Rev. Med. Virol. 16:297 – 310.

Haas J, Hug A, Viehöver A, Fritzsching B, Falk CS, Filser A, Vetter T, Milkova L, Korporal M, Fritz B, Storch-Hagenlocher B, Krammer PH, Suri-Payer E, Wildemann B. (2005) Reduced suppressive effect of CD4+ CD25high regulatory T cells on the T cell immune response against myelin oligodendrocyte glycoprotein in patients with multiple sclerosis. Eur. J. Immunol. 35:3343-52.

Haas J, Firzlaff M. (2005) Twenty-four-month comparison of immunomodulatory treatments - a retrospective open label study in 308 RRMS patients treated with beta interferons or glatiramer acetate (Copaxone). Eur. J. Neurol. 12:425-31.

Haas J, Fritzsching B, Trübswetter P, Korporal M, Milkova L, Fritz B, Vobis D, Krammer PH, Suri-Payer E, Wildemann B. (2007) Prevalence of newly generated naive regulatory T cells (Treg) is critical for Treg suppressive function and determines Treg dysfunction in multiple sclerosis. J. Immunol. 179:1322-30.

Hafler D, Slavik J, Anderson D, O'Connor K, De Jager P, Baecher-Allan C. (2005) Multiple sclerosis. Immunological reviews. 204:208-31.

Hafler JP, Maier LM, Cooper JD, Plagnol V, Hinks A, Simmonds MJ, Stevens HE, Walker NM, Healy B, Howson JM, Maisuria M, Duley S, Coleman G, Gough SC; International Multiple Sclerosis Genetics Consortium (IMSGC), Worthington J, Kuchroo VK, Wicker LS, Todd JA. (2009) CD226 Gly307Ser association with multiple autoimmune diseases. Genes Immun. 10:5-10.

236

Hamann I, Unterwalder N, Cardona AE, Meisel C, Zipp F, Ransohoff RM, Infante- Duarte C. (2011) Analyses of phenotypic and functional characteristics of CX3CR1- expressing natural killer cells. Immunology. 133:62-73.

Hamerman JA, Ogasawara K, Lanier LL. (2004) Cutting Edge: Toll-Like Receptor Signaling in Macrophages Induces Ligands for the NKG2D Receptor. J. Immunol. 172:2001-5.

Hamilton JA. (2002) GM-CSF in inflammation and autoimmunity. Trends Immunol. 23:403-8.

Hammarberg H, Lidman O, Lundberg C, Eltayeb SY, Gielen AW, Muhallab S, Svenningson A, Lindå H, van der Meide PH, Cullheim S, Olsson T, Piehl F. (2000) Neuroprotection by encephalomyelitis: Rescue of Mechanically Injured Neurons and Neurotrophin Production by CNS-Infiltrating T and Natural Killer Cells. J. Neurosci. 20: 5283-5291.

Hammond SR, English DR, McLeod JG. (2000) The age-range of risk of developing multiple sclerosis: evidence from a migrant population in Australia. Brain. 123: 968– 74.

Han J, Goldstein LA, Gastman BR, Froelich CJ, Yin XM, Rabinowich H. (2004) Degradation of Mcl-1 by granzyme B: implications for Bim-mediated mitochondrial apoptotic events. J. Biol. Chem. 279:22020-9.

Hanna J, Gonen-Gross T, Fitchett J, Rowe T, Daniels M, Arnon TI, Gazit R, Joseph A, Schjetne KW, Steinle A, Porgador A, Mevorach D, Goldman-Wohl D, Yagel S, LaBarre MJ, Buckner JH, Mandelboim O. (2004a) Novel APC-like properties of human NK cells directly regulate T cell activation. J. Clin. Invest. 114:1612-23.

Hanna J, Bechtel P, Zhai Y, Youssef F, McLachlan K, Mandelboim O. (2004b) Novel insights on human NK cells' immunological modalities revealed by gene expression profiling. J. immunol. 173:6547-63.

Hansasuta P, Dong T, Thananchai H, Weekes M, Willberg C, Aldemir H, Rowland- Jones S, Braud VM. (2004) Recognition of HLA-A3 and HLA-A11 by KIR3DL2 is peptide-specific. Eur. J. Immunol. 34:1673-9.

237

Hao J, Liu R, Piao W, Zhou Q, Vollmer TL, Campagnolo DI, Xiang R, La Cava A, Van Kaer L, Shi FD. (2010) Central nervous system (CNS)-resident natural killer cells suppress Th17 responses and CNS autoimmune pathology. J. Exp. Med. 207:1907-21.

Harada N, Matsumoto M, Koyama N, Shimizu A, Honjo T, Tominaga A, Takatsu K. (1987) T cell replacing factor/interleukin 5 induces not only B-cell growth and differentiation, but also increased expression of interleukin 2 receptor on activated B- cells. Immunol. Lett. 15:205-15.

Harbo HF, Lie BA, Sawcer S, Celius EG, Dai KZ, Oturai A, Hillert J, Lorentzen AR, Laaksonen M, Myhr KM, Ryder LP, Fredrikson S, Nyland H, Sørensen PS, Sandberg- Wollheim M, Andersen O, Svejgaard A, Edland A, Mellgren SI, Compston A, Vartdal F, Spurkland A. (2004) Genes in the HLA class I region may contribute to the HLA class II-associated genetic susceptibility to multiple sclerosis. Tissue Antigens. 63:237-47.

Harlin H, Hanson M, Johansson CC, Sakurai D, Poschke I, Norell H, Malmberg KJ, Kiessling R. (2007) The CD16- CD56(bright) NK cell subset is resistant to reactive oxygen species produced by activated granulocytes and has higher antioxidative capacity than the CD16+ CD56(dim) subset. J. Immunol. 179:4513-9.

Harper N, Hughes M, MacFarlane M, Cohen GM. (2003) Fas-associated death domain protein and caspase-8 are not recruited to the tumor necrosis factor receptor 1 signaling complex during tumor necrosis factor-induced apoptosis. J. Biol. Chem. 278:25534-41.

Harrer A, Wipfler P, Einhaeupl M, Pilz G, Oppermann K, Hitzl W, Afazel S, Haschke-Becher E, Strasser P, Trinka E, Kraus J. (2011) Natalizumab therapy decreases surface expression of both VLA-heterodimer subunits on peripheral blood mononuclear cells. J. Neuroimmunol. 234:148-54.

Harrington LE, Hatton RD, Mangan PR, Turner H, Murphy TL, Murphy KM, Weaver CT. (2005) Interleukin 17-producing CD4+ effector T cells develop via a lineage distinct from the T helper type 1 and 2 lineages. Nat. Immunol. 6:1123-32.

238

Harris DP, Goodrich S, Gerth AJ, Peng SL, Lund FE. (2005) Regulation of IFN- gamma production by B effector 1 cells: essential roles for T-bet and the IFN-gamma receptor. J. Immunol. 174:6781-90.

Hart PH, Vitti GF, Burgess DR, Whitty GA, Royston K, Hamilton JA. (1991) Activation of human monocytes by granulocyte-macrophage colony-stimulating factor: increased urokinase-type plasminogen activator activity. Blood. 77:841-8.

Hartung HP, Gonsette R, König N, Kwiecinski H, Guseo A, Morrissey SP, Krapf H, Zwingers T; Mitoxantrone in Multiple Sclerosis Study Group (MIMS). (2002) Mitoxantrone in progressive multiple sclerosis: a placebo-controlled, double-blind, randomised, multicentre trial. Lancet. 360:2018-25.

Harvey EJ, Li M, Ramji GP. (2007) Critical role for casein kinase 2 and phosphoinositide-3-kinase in the interferon-γ-induced expression of monocyte chemoattractant protein-1 and other key genes implicated in atherosclerosis. Arterioscler. Thromb. Vasc. Biol. 27:806–12.

Hashimoto S, Yamada M, Motoyoshi K, Akagawa KS. (1997) Enhancement of macrophage-colony stimulating factor-induced growth and differentiation of human monocytes by interleukin-10. Blood. 89: 315

Hashimoto S, Komuro I, Yamada M, Agakawa K. (2001) IL-10 Inhibits Granulocyte- Macrophage Colony-Stimulating Factor-Dependent Human Monocyte Survival at the Early Stage of the Culture and Inhibits the Generation of Macrophages. J. Immunol. 167:3619-3625.

Hauser SL, Oksenberg JR. (2006) The neurobiology of multiple sclerosis: genes, inflammation, and neurodegeneration. Neuron. 52:61-76.

Hauser SL, Waubant E, Arnold DL, Vollmer T, Antel J, Fox RJ, Bar-Or A, Panzara M, Sarkar N, Agarwal S, Langer-Gould A, Smith CH; HERMES Trial Group. (2008) B-cell depletion with rituximab in relapsing-remitting multiple sclerosis. N. Engl. J. Med. 358:676-88.

Hayakawa Y, Smyth MJ. (2006) CD27 dissects mature NK cells into two subsets with distinct responsiveness and migratory capacity. J. Immunol. 176:1517-24.

239

Healy BC, Ali EN, Guttmann CR, Chitnis T, Glanz BI, Buckle G, Houtchens M, Stazzone L, Moodie J, Berger AM, Duan Y, Bakshi R, Khoury S, Weiner H, Ascherio A. (2009) Smoking and disease progression in multiple sclerosis. Arch. Neurol. 66:858-64.

Healy BC, Liguori M, Tran D, Chitnis T, Glanz B, Wolfish C, Gauthier S, Buckle G, Houtchens M, Stazzone L, Khoury S, Hartzmann R, Fernandez-Vina M, Hafler DA, Weiner HL, Guttmann CR, De Jager PL. (2010) HLA B*44: protective effects in MS susceptibility and MRI outcome measures. Neurology. 75: 634–40.

Hedström A K, Bäärnhielm M, Olsson T, Alfredsson L. (2009) Tobacco smoking, but not Swedish snuff use, increases the risk of multiple sclerosis. Neurology. 73:696– 701.

Heibein JA, Barry M, Motyka B, Bleackley RC. (1999) Granzyme B-Induced Loss of Mitochondrial Inner Membrane Potential (ΔΨm) and Cytochrome c Release Are Caspase Independent. J. Immunol. 163:4683-4693.

Heijnen IA, van de Winkel JG. (1997) Human IgG Fc receptors. Int. Rev. Immunol. 16:29-55.

Herbert DR, Lee JJ, Lee NA, Nolan TJ, Schad GA, Abraham D. (2000) Role of IL-5 in innate and adaptive immunity to larval Strongyloides stercoralis in mice. J. Immunol. 165:4544–51.

Hernán MA, Olek MJ, Ascherio, A. (2001) Cigarette smoking and incidence of multiple sclerosis. Am. J. Epidemiol. 154: 69–74.

Hernán MA, Jick SS, Logroscino G, Olek MJ, Ascherio A, Jick H. (2005) Cigarette smoking and the progression of multiple sclerosis. Brain. 128:1461-5.

Hershkovitz O, Rosental B, Rosenberg LA, Navarro-Sanchez ME, Jivov S, Zilka A, Gershoni-Yahalom O, Brient-Litzler E, Bedouelle H, Ho JW, Campbell KS, Rager- Zisman B, Despres P, Porgador A. (2009) NKp44 Receptor Mediates Interaction of the Envelope Glycoproteins from the West Nile and Dengue Viruses with NK Cells. J. Immunol. 183:2610-21.

240

Hiby SE, Walker JJ, O'shaughnessy KM, Redman CW, Carrington M, Trowsdale J, Moffett A. (2004) Combinations of maternal KIR and fetal HLA-C genes influence the risk of preeclampsia and reproductive success. J. Exp. Med. 200:957-65.

Hiby SE, Regan L, Lo W, Farrell L, Carrington M, Moffett A. (2008) Association of maternal killer-cell immunoglobulin-like receptors and parental HLA-C genotypes with recurrent miscarriage. Hum. Reprod. 23:972-6.

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. (2010) Maternal activating KIRs protect against human reproductive failure mediated by fetal HLA-C2. J. Clin. Invest. 120:4102-10.

Hickey WF, Hsu BL, Kimura H. (1991) T-lymphocyte entry into the central nervous system. J. Neurosci. Res. 28:254–60.

Hill KE, Zollinger L V, Watt HE, Carlson NG, Rose JW. (2004) Inducible nitric oxide synthase in chronic active multiple sclerosis plaques: Distribution, cellular expression and association with myelin damage. J. Neuroimmunol. 151: 171–9.

Hobart M, Ramassar V, Goes N, Urmson J, Halloran PF. (1997) IFN regulatory factor-1 plays a central role in the regulation of the expression of class I and II MHC genes in vivo. J. Immunol. 158:4260-9.

Hoeve MA, Savage ND, de Boer T, Langenberg DM, de Waal Malefyt R, Ottenhoff TH, Verreck FA. (2006) Divergent effects of IL-12 and IL-23 on the production of IL-17 by human T cells. Eur. J. Immunol. 36:661-70.

Hofman FM, Hinton DR, Johnson K, Merrill JE. (1989) Tumour necrosis factor identified in multiple sclerosis brain. J. Exp. Med. 170:607-12.

Holler N, Zaru R, Micheau O, Thome M, Attinger A, Valitutti S, Bodmer JL, Schneider P, Seed B, Tschopp J. (2000) Fas triggers an alternative, caspase-8- independent cell death pathway using the kinase RIP as effector molecule. Nat. Immunol. 1:489-95.

241

Holler N, Tardivel A, Kovacsovics-Bankowski M, Hertig S, Gaide O, Martinon F, Tinel A, Deperthes D, Calderara S, Schulthess T, Engel J, Schneider P, Tschopp J. (2003) Two Adjacent Trimeric Fas Ligands Are Required for Fas Signaling and Formation of a Death-Inducing Signaling Complex. Mol. Cell. Biol. 23:1428-1440.

Holman DW, Klein RS, Ransohoff RM. (2011) The blood-brain barrier, chemokines and multiple sclerosis. Biochim. Biophys. Acta. 1812:220-30.

Hommura F, Katabami M, Leaner VD, Donninger H, Sumter TF, Resar LM, Birrer MJ. (2004) HMG-I/Y is a c-Jun/activator protein-1 target gene and is necessary for c- Jun-induced anchorage-independent growth in Rat1a cells. Mol. Cancer Res. 2:305- 14.

Hong J, Li N, Zhang X, Zheng J, Zhang Z. (2005) Induction of CD4+CD25+ regulatory T cells by copolymer-I through activation of transcription factor Foxp3. Proc. Natl. Acad. Sci. U.S.A. 102: 6449–6454.

Hori T, de Waal Malefyt R, Duncan BW, Harrison MR, Roncarolo MG, Spits H. (1991) Cloning of a novel cell type from human fetal liver expressing cytoplasmic CD3 delta and epsilon but not membrane CD3. Int. Immunol. 3:353-7.

Horowitz A, Behrens RH, Okell L, Fooks AR, Riley EM. (2010) NK cells as effectors of acquired immune responses: effector CD4+ T cell-dependent activation of NK cells following vaccination. J. Immunol. 185:2808-18.

Horowitz A, Hafalla JC, King E, Lusingu J, Dekker D, Leach A, Moris P, Cohen J, Vekemans J, Villafana T, Corran PH, Bejon P, Drakeley CJ, von Seidlein L, Riley EM. (2012) Antigen-specific IL-2 secretion correlates with NK cell responses after immunization of Tanzanian children with the RTS,S/AS01 malaria vaccine. J. Immunol. 188:5054-62.

Hou Q, Zhao T, Zhang H, Lu H, Zhang Q, Sun L, Fan Z. (2008) Granzyme H induces apoptosis of target tumor cells characterized by DNA fragmentation and Bid- dependent mitochondrial damage. Mol. Immunol. 45:1044-55.

242

Houchins, J. P., Lanier, L. L., Niemi, E., Phillips, J. H.& Ryan, J. C. (1997) Natural killer cell cytolytic activity is inhibited by NKG2-A and activated by NKG2-C. J. Immunol. 158:3603–3609.

Hsu KC, Liu XR, Selvakumar A, Mickelson E, O'Reilly RJ, Dupont B. (2002) Killer Ig-like receptor haplotype analysis by gene content: evidence for genomic diversity with a minimum of six basic framework haplotypes, each with multiple subsets. J. Immunol. 169:5118-29.

Hu D, Liu S, Shi L, Li C, Wu L, Fan Z. (2010) Cleavage of survivin by Granzyme M triggers degradation of the survivin-X-linked inhibitor of apoptosis protein (XIAP) complex to free caspase activity leading to cytolysis of target tumor cells. J. Biol. Chem. 285:18326-35.

Hua G, Zhang Q, Fan Z. (2007) Heat shock protein 75 (TRAP1) antagonizes reactive oxygen species generation and protects cells from granzyme M-mediated apoptosis. J. Biol. Chem. 282:20553-60.

Huang D, Shi FD, Jung S, Pien GC, Wang J, Salazar-Mather TP, He TT, Weaver JT, Ljunggren HG, Biron CA, Littman DR, Ransohoff RM. (2006) The neuronal chemokine CX3CL1/fractalkine selectively recruits NK cells that modify experimental autoimmune encephalomyelitis within the central nervous system. FASEB J. 20:896-905.

Huber AR, Kunkel SL, Todd RF, Weiss SJ. (1991) Regulation of transendothelial neutrophil migration by endogenous interleukin-8. Science. 254: 99-102.

Hug A, Korporal M, Schröder I, Haas J, Glatz K, Storch-Hagenlocher B, Wildemann B. (2003) Thymic export function and T cell homeostasis in patients with relapsing remitting multiple sclerosis. J. Immunol. 171:432-7.

Huseby ES, Liggitt D, Brabb T, Schnabel B, Ohlen C, Goverman J. (2001) A Pathogenic Role for Myelin-Specific Cd8+ T Cells in a Model for Multiple Sclerosis. J. Exp. Med. 194:669-76.

243

Iannello A, Debbeche O, Samarani S, Ahmad A. (2008) Antiviral NK cell responses in HIV infection: I. NK cell receptor genes as determinants of HIV resistance and progression to AIDS. J. Leukoc. Biol. 84:1-26.

Ikawa T, Kawamoto H, Fujimoto S, Katsura Y. (1999) Commitment of common T/Natural killer (NK) progenitors to unipotent T and NK progenitors in the murine fetal thymus revealed by a single progenitor assay. J. Exp. Med. 190:1617-26.

Inaba K, Kitaura M, Kato T, Watanabe Y, Kawade Y, Muramatsu S. (1986) Contrasting effect of alpha/beta- and gamma-interferons on expression of macrophage Ia antigens. J. Exp. Med. 163: 1030–5.

Infante-Duarte C, Weber A, Krätzschmar J, Prozorovski T, Pikol S, Hamann I, Bellmann-Strobl J, Aktas O, Dörr J, Wuerfel J, Stürzebecher CS, Zipp F. (2005) Frequency of blood CX3CR1-positive natural killer cells correlates with disease activity in multiple sclerosis patients. FASEB J. 19:1902-4.

Inoue M, Williams KL, Gunn MD, Shinohara ML. (2012) NLRP3 inflammasome induces chemotactic immune cell migration to the CNS in experimental autoimmune encephalomyelitis. Proc. Nat. Acad. USA. 109:10480-5.

International MHC and Autoimmunity Genetics Network, Rioux JD, Goyette P, Vyse TJ, Hammarström L, Fernando MM, Green T, De Jager PL, Foisy S, Wang J, de Bakker PI, Leslie S, McVean G, Padyukov L, Alfredsson L, Annese V, Hafler DA, Pan-Hammarström Q, Matell R, Sawcer SJ, Compston AD, Cree BA, Mirel DB, Daly MJ, Behrens TW, Klareskog L, Gregersen PK, Oksenberg JR, Hauser SL. (2009) Mapping of multiple susceptibility variants within the MHC region for 7 immune- mediated diseases. Proc. Natl. Acad. Sci. U.S.A. 106:18680-5.

International Multiple Sclerosis Genetics Consortium, Hafler DA, Compston A, Sawcer S, Lander ES, Daly MJ, De Jager PL, de Bakker PI, Gabriel SB, Mirel DB, Ivinson AJ, Pericak-Vance MA, Gregory SG, Rioux JD, McCauley JL, Haines JL, Barcellos LF, Cree B, Oksenberg JR, Hauser SL. (2007) Risk alleles for multiple sclerosis identified by a genomewide study. N. Engl. J. Med. 357:851–62.

244

Ito M, Maruyama T, Saito N, Koganei S, Yamamoto K, Matsumoto N. (2006) Killer cell lectin-like receptor G1 binds three members of the classical cadherin family to inhibit NK cell cytotoxicity. J. Exp. Med. 203:289-95.

Itoh K, Inoue T, Ito K, Hirohata S. (1994) The interplay of interleukin-10 (IL-10) and interleukin-2 (IL-2) in humoral immune responses: IL-10 synergizes with IL-2 to enhance responses of human B lymphocytes in a mechanism which is different from upregulation of CD25 expression. Cell. Immunol. 157:478-88.

Itoh K, Hirohata S. (1995) The role of IL-10 in human B cell activation, proliferation, and differentiation. J. Immunol. 154:4341-50.

Izumi Y, Ida H, Huang M, Iwanaga N, Tanaka F, Aratake K, Arima K, Tamai M, Kamachi M, Nakamura H, Origuchi T, Kawakami A, Anderson P, Eguchi K. (2006) Characterization of peripheral natural killer cells in primary Sjögren's syndrome: impaired NK cell activity and low NK cell number. J. Lab. Clin. Med. 147:242-9.

Jack C, Antel J, Bruck W, Kuhlmann T. (2007) Contrasting potential of nitric oxide and peroxynitrite to mediate oligodendrocyte injury in multiple sclerosis. Glia. 55:926–934.

Jacobs R, Hintzen G, Kemper A, Beul K, Kempf S, Behrens G, Sykora KW, Schmidt RE. (2001) CD56bright cells differ in their KIR repertoire and cytotoxic features from CD56dim NK cells. Eur. J. Immunol. 31:3121-7.

Jäger A, Dardalhon V, Sobel RA, Bettelli E, Kuchroo VK. (2009) Th1, Th17, and Th9 effector cells induce experimental autoimmune encephalomyelitis with different pathological phenotypes. J. Immunol. 183:7169-77.

Javed A, Reder A. (2006) Therapeutic role of beta-interferons in multiple sclerosis. Pharmacol. Therapeut. 110:35–56.

Javelaud D, Besancon F. (2001) NF-kappa B activation results in rapid inactivation of JNK in TNF alpha-treated Ewing sarcoma cells: a mechanism for the anti-apoptotic effect of NF-kappa B. Oncogene. 20:4365-72.

245

Jelcić I, Hsu KC, Kakalacheva K, Breiden P, Dupont B, Uhrberg M, Martin R, Münz C, Lünemann JD. (2012) Killer immunoglobulin-like receptor locus polymorphisms in multiple sclerosis. Mult. Scler. 18:951-8.

Jennings P, Yuan D. (2009) NK cell enhancement of antigen presentation by B lymphocytes. J. Immunol. 182:2879-87.

Jiang K, Zhong B, Gilvary DL, Corliss BC, Vivier E, Hong-Geller E, Wei S, Djeu JY. (2002) Syk Regulation of Phosphoinositide 3-Kinase-Dependent NK Cell Function. J. Immunol. 168:3155-64.

Jiang W, Chai NR, Maric D, Bielekova B. (2011) Unexpected role for granzyme K in CD56bright NK cell-mediated immunoregulation of multiple sclerosis. J. Immunol. 187:781-90.

Jiang W, Johnson C, Jayaraman J, Simecek N, Noble J, Moffatt MF, Cookson WO, Trowsdale J, Traherne JA. (2012) Copy number variation leads to considerable diversity for B but not A haplotypes of the human KIR genes encoding NK cell receptors. Genome Res. Advanced publication. September 4, 2012, doi: 10.1101/gr.137976.112

Johnson KP, Brooks BR, Cohen JA, Ford CC, Goldstein J, Lisak RP, Myers LW, Panitch HS, Rose JW, Schiffer RB. (1995) Copolymer 1 reduces relapse rate and improves disability in relapsing-remitting multiple sclerosis: results of a phase III multicenter, double-blind placebo-controlled trial. The Copolymer 1 Multiple Sclerosis Study Group. Neurology. 45:1268-76.

Johnson DR, Mook-Kanamori B. (2000) Dependence of elevated human leukocyte antigen class I molecule expression on increased heavy chain, light chain (2- microglobulin), transporter associated with antigen processing, tapasin, and peptide. J. Biol. Chem. 275:16643.

Johnson JL, Hong H, Monfregola J, Kiosses WB, Catz SD. (2011) Munc13-4 restricts motility of Rab27a-expressing vesicles to facilitate lipopolysaccharide-induced priming of exocytosis in neutrophils. J. Biol. Chem. 286:5647-56.

246

Joncker NT, Fernandez NC, Treiner E, Vivier E, Raulet DH. (2009) NK cell responsiveness is tuned commensurate with the number of inhibitory receptors for self-MHC class I: the rheostat model. J. Immunol. 182:4572-80.

Jones DC, Edgar RS, Ahmad T, Cummings JR, Jewell DP, Trowsdale J, Young NT. (2006) Killer Ig-like receptor (KIR) genotype and HLA ligand combinations in ulcerative colitis susceptibility. Genes Immun. 7:576-82.

Jones JL, Anderson JM, Phuah CL, Fox EJ, Selmaj K, Margolin D, Lake SL, Palmer J, Thompson SJ, Wilkins A, Webber DJ, Compston DA, Coles AJ. (2010) Improvement in disability after alemtuzumab treatment of multiple sclerosis is associated with neuroprotective autoimmunity. Brain. 133:2232-47.

Juedes AE, Hjelmström P, Bergman CM, Neild AL, Ruddle NH. (2000) Kinetics and cellular origin of cytokines in the central nervous system: insight into mechanisms of myelin oligodendrocyte glycoprotein-induced experimental autoimmune encephalomyelitis. J. Immunol. 164:419-26.

Juelke K, Killig M, Luetke-Eversloh M, Parente E, Gruen J, Morandi B, Ferlazzo G, Thiel A, Schmitt-Knosalla I, Romagnani C. (2010) CD62L expression identifies a unique subset of polyfunctional CD56dim NK cells. Blood. 116:1299-307.

Jurewicz A, Biddison WE, Antel JP. (1998) MHC class I-restricted lysis of human oligodendrocytes by myelin basic protein peptide-specific CD8 T lymphocytes. J. Immunol. 160:3056–9.

Jurewicz A, Matysiak M, Tybor K, Kilianek L, Raine CS, Selmaj K. (2005) Tumour necrosis factor-induced death of adult human oligodendrocytes is mediated by apoptosis inducing factor. Brain. 128: 2675–88.

Kabat J, Borrego F, Brooks A, Coligan JE. (2002) Role That Each NKG2A Immunoreceptor Tyrosine-Based Inhibitory Motif Plays in Mediating the Human CD94/NKG2A Inhibitory Signal. J. Immunol. 169:1948-58.

Kägi D, Ledermann B, Bürki K, Seiler P, Odermatt B, Olsen KJ, Podack ER, Zinkernagel RM, Hengartner H. (1994) Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice. Nature. 369:31–7.

247

Kamata H, Honda S, Maeda S, Chang L, Hirata H, Karin M. (2005) Reactive oxygen species promote TNFalpha-induced death and sustained JNK activation by inhibiting MAP kinase phosphatases. Cell. 120:649-61.

Kamen DL, Tangpricha V. (2010) Vitamin D and molecular actions on the immune system: modulation of innate and autoimmunity. J. Mol. Med. 88:441-50.

Kamijo R, Harada H, Matsuyama T, Bosland M, Gerecitano J, Shapiro D, Le J, Koh SI, Kimura T, Green SJ, Mak TW, Taniguchi T, Vlicek J. (1994) Requirement for transcription factor IRF-1 in NO synthase induction in macrophages. Science. 263:1612–15.

Kang J, Jiang L, Goldman SA, Nedergaard M. (1998) Astrocyte-mediated potentiation of inhibitory synaptic transmission. Nat. Neurosci. 1:683–692.

Kang SW, Kim SH, Lee N, Lee WW, Hwang KA, Shin MS, Lee SH, Kim WU, Kang I. (2012) 1,25-Dihyroxyvitamin D3 promotes FOXP3 expression via binding to vitamin D response elements in its conserved noncoding sequence region. J. Immunol. 188:5276-82.

Kaplan A, Kotzer S, Almeida CR, Kohen R, Halpert G, Salmon-Divon M, Köhler K, Höglund P, Davis DM, Mehr R. (2011) Simulations of the NK cell immune synapse reveal that activation thresholds can be established by inhibitory receptors acting locally. J. Immunol. 187:760-73.

Kappos L, Radue EW, O'Connor P, Polman C, Hohlfeld R, Calabresi P, Selmaj K, Agoropoulou C, Leyk M, Zhang-Auberson L, Burtin P; FREEDOMS Study Group. (2010) A placebo-controlled trial of oral fingolimod in relapsing multiple sclerosis. N. Engl. J. Med. 362:387-401.

Karabudak R, Kurne A, Guc D, Sengelen M, Canpinar H, Kansu E. (2004) Effect of interferon beta-1a on serum matrix metalloproteinase-9 (MMP-9) and tissue inhibitor of matrix metalloproteinase (TIMP-1) in relapsing remitting multiple sclerosis patients. One year follow-up results. J. Neurol. 251:279-83.

Karin M, Liu Z, Zandi E. (1997)AP-1 function and regulation. Curr. Opin. Cell Biol. 9:240-6.

248

Karp CL, Biron CA, Irani DN. (2000) Interferon beta in multiple sclerosis: is IL-12 suppression the key? Immunol. Today. 21:24-8.

Kärre K, Ljunggren HG, Piontek G, Kiessling R. (1986) Selective rejection of H-2- deficient lymphoma variants suggests alternative immune defence strategy. Nature. 319:675-8.

Kastrukoff LF, Morgan NG, Zecchini D, White R, Petkau AJ, Satoh J, Paty DW. (1998) A role for natural killer cells in the immunopathogenesis of multiple sclerosis. J. Neuroimmunol. 86:123-33.

Kastrukoff LF, Lau A, Wee R, Zecchini D, White R, Paty DW. (2003) Clinical relapses of multiple sclerosis are associated with ‘novel’ valleys in natural killer cell functional activity. J. Neuroimmunol. 145:103-14.

Kaur G, Troüsdale J, Fugger L. (2012) Natural killer cells and their receptors in multiple sclerosis. Brain. Advance Access published on 25th June 2012. doi: 10.1093/brain/aws159

Kawai T, Seki M, Hiromatsu K, Eastcott JW, Watts GF, Sugai M, Smith DJ, Porcelli SA, Taubman MA. (1999) Selective diapedesis of Th1 cells induced by endothelial cell RANTES. J. Immunol. 163:3269-78.

Kawamoto H, Ohmura K, Katsura Y. (1998) Presence of progenitors restricted to T, B, or myeloid lineage, but absence of multipotent stem cells, in the murine fetal thymus. J. Immunol. 161:3799-802.

Kebir H, Kreymborg K, Ifergan I, Dodelet-Devillers A, Cayrol R, Bernard M, Giuliani F, Arbour N, Becher B, Prat A. (2007) Human TH17 lymphocytes promote blood-brain barrier disruption and central nervous system inflammation. Nat. Med. 13:1173-5.

Kebir H, Ifergan I, Alvarez JI, Bernard M, Poirier J, Arbour N, Duquette P, Prat A. (2009) Preferential recruitment of interferon-gamma-expressing TH17 cells in multiple sclerosis. Ann. Neurol. 66:390-402.

249

Kelly JM, Darcy PK, Markby JL, Godfrey DI, Takeda K, Yagita H, Smyth MJ. (2002) Induction of tumor-specific T cell memory by NK cell-mediated tumor rejection. Nat. Immunol. 3:83-90.

Kerfoot SM, Kubes P. (2002) Overlapping roles of P-selectin and alpha 4 integrin to recruit leukocytes to the central nervous system in experimental autoimmune encephalomyelitis. J. Immunol. 169:1000-6.

Keskin DB, Allan DSJ, Rybalov B, Andzelm MM, Stern JNH, Kopcow HD, Koopman LA, Strominger JL. (2007) TGFβ promotes conversion of CD16+ peripheral blood NK cells into CD16- NK cells with similarities to decidual NK cells. Proc. Nat. Acad. USA. 104: 3378-83.

Khademi M, Stol D, Olsson T, Wallström E. (2008) Induction of systemic TNFalpha in natalizumab-treated multiple sclerosis. Eur. J. Neurol. 15:309-12.

Khademi M, Bornsen L, Rafatnia F, Andersson M, Brundin L, Piehl F, Sellebjerg F, Olsson T. (2009) The effects of natalizumab on inflammatory mediators in multiple sclerosis: prospects for treatment-sensitive biomarkers. Eur. J. Neurol. 16:528-36.

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. (2004) HLA and NK Cell Inhibitory Receptor Genes in Resolving Hepatitis C Virus Infection. Science. 305:872-4.

Khan OA, Tselis AC, Kamholz JA, Garbern JY, Lewis RA, Lisak RP. (2001) A prospective, open-label treatment trial to compare the effect of IFN beta-1a (Avonex), IFNbeta-1b (Betaseron), and glatiramer acetate (Copaxone) on the relapse rate in relapsing-remitting multiple sclerosis. Eur. J. Neurol. 8:141-8.

Kiessling R, Klein E, Pross H, Wigzell H. (1975) "Natural" killer cells in the mouse. II. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Characteristics of the killer cell. Eur. .J Immunol. 5:117-21.

250

Kikuchi-Maki A, Yusa Si, Catina TL, Campbell KS. (2003) KIR2DL4 Is an IL-2- Regulated NK Cell Receptor That Exhibits Limited Expression in Humans but Triggers Strong IFN-gamma Production. J. Immunol. 171:3415-25.

Kikuchi-Maki A, Catina TL, Campbell KS. (2005) Cutting Edge: KIR2DL4 Transduces Signals into Human NK Cells through Association with the Fc Receptor gamma Protein. J. Immunol. 174:3859-63.

Kim HJ, Ifergan I, Antel JP, Seguin, Duddy M, Lapierre Y, Jalili F, Bar-Or A. (2004) Type 2 monocyte and microglia differentiation mediated by glatiramer acetate therapy in patients with multiple sclerosis. J. Immunol. 172: 7144–53.

Kim S, Poursine-Laurent J, Truscott SM, Lybarger L, Song YJ, Yang L, French AR, Sunwoo JB, Lemieux S, Hansen TH, Yokoyama WM. (2005) Licensing of natural killer cells by host major histocompatibility complex class I molecules. Nature. 436:709-13.

King A, Allan DS, Bowen M, Powis SJ, Joseph S, Verma S, Hiby SE, McMichael AJ, Loke YW, Braud VM. (2000) HLA-E is expressed on trophoblast and interacts with CD94/NKG2 receptors on decidual NK cells. Eur. J. Immunol. 30:1623-31.

Kischkel FC, Lawrence DA, Chuntharapai A, Schow P, Kim KJ, Ashkenazi A. (2000) Apo2L/TRAIL-dependent recruitment of endogenous FADD and caspase-8 to death receptors 4 and 5. Immunity. 12:611-20.

Kischkel FC, Lawrence DA, Tinel A, LeBlanc H, Virmani A, Schow P, Gazdar A, Blenis J, Arnott D, Ashkenazi A. (2001) Death receptor recruitment of endogenous caspase-10 and apoptosis initiation in the absence of caspase-8. J. Biol. Chem. 276:46639-46.

Kivisäkk P, Mahad DJ, Callahan MK, Trebst C, Tucky B, Wei T, Wu L, Baekkevold ES, Lassmann H, Staugaitis SM, Campbell JJ, Ransohoff RM. (2003) Human cerebrospinal fluid central memory CD4+ T cells: evidence for trafficking through choroid plexus and meninges via Pselectin. Proc. Natl. Acad. Sci. U.S.A. 100:8389– 94.

251

Kivisäkk P, Healy BC, Viglietta V, Quintana FJ, Hootstein MA, Weiner HL, Khoury SJ. (2009a) Natalizumab treatment is associated with peripheral sequestration of proinflammatory T cells. Neurology. 72:1922-30.

Kivisäkk P, Imitola J, Rasmussen S, Elyaman W, Zhu B, Ransohoff RM, Khoury SJ. (2009b) Localizing central nervous system immune surveillance: meningeal antigen- presenting cells activate T cells during experimental autoimmune encephalomyelitis. Ann. Neurol. 65:457–69.

Klebanoff SJ, Vadas MA, Harlan JM, Sparks LH, Gamble JR, Agosti JM, Waltersdorph AM. (1986) Stimulation of neutrophils by tumor necrosis factor. J. Immunol. 136:4220-5.

Klotz L, Meuth SG, Wiendl H. (2012) Immune mechanisms of new therapeutic strategies in multiple sclerosis—A focus on alemtuzumab. Cinical Immunol. 142: 25- 30.

Koch S, Goedde R, Nigmatova V, Epplen JT, Müller N, de Seze J, Vermersch P, Momot T, Schmidt RE, Witte T. (2005) Association of multiple sclerosis with ILT6 deficiency. Genes Immunit. 6:445-7.

Kondo M, Weissman IL, Akashi K. (1997) Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell. 91:661-72.

Koopman LA, Kopcow HD, Rybalov B, Boyson JE, Orange JS, Schatz F, Masch R, Lockwood CJ, Schachter AD, Park PJ, Strominger JL. (2003) Human decidual natural killer cells are a unique NK cell subset with immunomodulatory potential. J. Exp. Med. 198:1201-12.

Kopcow HD, Allan DS, Chen X, Rybalov B, Andzelm MM, Ge B, Strominger JL. (2005) Human decidual NK cells form immature activating synapses and are not cytotoxic. Proc. Natl. Acad. Sci. USA. 102:15563-8.

Korn T, Reddy J, Gao W, Bettelli E, Awasthi A, Petersen TR, Bäckström BT, Sobel RA, Wucherpfennig KW, Strom TB, Oukka M, Kuchroo VK. (2007) Myelin-specific regulatory T cells accumulate in the CNS but fail to control autoimmune inflammation. Nat. Med. 13:423-31.

252

Kornek B, Storch MK, Weissert R, Wallstroem E, Stefferl A, Olsson T, Linington C, Schmidbauer M, Lassmann H. (2000) Multiple sclerosis and chronic autoimmune encephalomyelitis: a comparative quantitative study of axonal injury in active, inactive, and remyelinated lesions. Am. J. Pathol. 157:267-76.

Kothakota S, Azuma T, Reinhard C, Klippel A, Tang J, Chu K, McGarry TJ, Kirschner MW, Koths K, Kwiatkowski DJ, Williams LT. (1997) Caspase-3-generated fragment of gelsolineffector of morphological change in apoptosis. Science. 278:294- 8.

Kozovska ME, Hong J, Zang YCQ, Li S, Rivera VM, Killian JM, Zhang JZ. (1999) Interferon beta induces T-helper 2 immune deviation in MS. Neurology. 53:1692

Krakowski ML, Owens T. (1997) The central nervous system environment controls effector CD4+ T cell cytokine profile in experimental allergic encephalomyelitis. Eur. J. Immunol. 27:2840-7.

Krapf H, Morrissey SP, Zenker O, Zwingers T, Gonsette R, Hartung HP; MIMS Study Group. (2005) Effect of mitoxantrone on MRI in progressive MS: results of the MIMS trial. Neurology. 65:690-5.

Krebs P, Barnes MJ, Lampe K, Whitley K, Bahjat KS, Beutler B, Janssen E, Hoebe K. (2009) NK-cell-mediated killing of target cells triggers robust antigen-specific T- cell-mediated and humoral responses. Blood. 113:6593-602.

Kroenke MA, Carlson TJ, Andjelkovic AV, Segal BM. (2008) IL-12- and IL-23- modulated T cells induce distinct types of EAE based on histology, CNS chemokine profile, and response to cytokine inhibition. J. Exp. Med. 205:1535-41.

Krzewski K, Chen X, Orange JS, Strominger JL. (2006) Formation of a WIP-, WASp- , actin-, and myosin IIA–containing multiprotein complex in activated NK cells and its alteration by KIR inhibitory signaling. J. Cell Biol. 173:121–132.

Kubin M, Cassiano L, Chalupny J, Chin W, Cosman D, Fanslow W, Müllberg J, Rousseau AM, Ulrich D, Armitage R. (2001) ULBP1, 2, 3: novel MHC class I-related molecules that bind to human cytomegalovirus glycoprotein UL16, activate NK cells. Eur. J. Immunol. 31:1428-37.

253

Kuhlow CJ, Krady JK, Basu A, Levison SW. (2003) Astrocytic ceruloplasmin expression, which is induced by IL-1beta and by traumatic brain injury, increases in the absence of the IL-1 type 1 receptor. Glia. 44:76–84.

Kupfer A, Dennert G, Singer SJ. (1983) Polarization of the Golgi apparatus and the microtubule-organizing center within cloned natural killer cells bound to their targets. Proc. Natl. Acad. Sci. USA. 80:7224-8.

Kupfer, A., G. Dennert, S. J. Singer. (1985) The orientation of the golgi apparatus and the microtubule-organizing center in the cytotoxic effector cell is a prerequisite in the lysis of bound target cells. J. Mol. Cell Immunol. 2: 3.

Lafont E, Milhas D, Teissié J, Therville N, Andrieu-Abadie N, Levade T, Benoist H, Ségui B. (2010) Caspase-10-dependent cell death in Fas/CD95 signalling is not abrogated by caspase inhibitor zVAD-fmk. PLoS One. 5:e13638.

Lalive PH, Menge T, Delarasse C, Della Gaspera B, Pham-Dinh D, Villoslada P, von Büdingen HC, Genain CP. (2006) Antibodies to native myelin oligodendrocyte glycoprotein are serologic markers of early inflammation in multiple sclerosis. Proc. Natl. Acad. Sci. U.S.A. 103:2280-5.

Lanier LL, Yu G, Phillips JH. (1989) Co-association of CD3 zeta with a receptor (CD16) for IgG Fc on human natural killer cells. Nature. 342:803-5.

Lanier LL, Chang C, Spits H, Phillips JH. (1992) Expression of cytoplasmic CD3 epsilon proteins in activated human adult natural killer (NK) cells and CD3 gamma, delta, epsilon complexes in fetal NK cells. Implications for the relationship of NK and T lymphocytes. J. Immunol. 149:1876-80.

Lanier LL, Corliss BC, Wu J, Leong C, Phillips JH. (1998) Immunoreceptor DAP12 bearing a tyrosine-based activation motif is involved in activating NK cells. Nature. 391:703-7.

Lanier LL. (2008) Up on the tightrope: natural killer cell activation and inhibition. Nat. Immunol. 9:495-502.

254

Lang PA, Lang KS, Xu HC, Grusdat M, Parish IA, Recher M, Elford AR, Dhanji S, Shaabani N, Tran CW, Dissanayake D, Rahbar R, Ghazarian M, Brüstle A, Fine J, Chen P, Weaver CT, Klose C, Diefenbach A, Häussinger D, Carlyle JR, Kaech SM, Mak TW, Ohashi PS. (2012) Natural killer cell activation enhances immune pathology and promotes chronic infection by limiting CD8+ T-cell immunity. Proc. Natl. Acad. USA. 109:1210-5.

Larsen CG, Anderson AO, Appella E, Oppenheim JJ, Matsushima K. (1989) The neutrophil-activating protein (NAP-1) is also chemotactic for T lymphocytes. Science. 243:1464-6.

Larsen PH, Wells JE, Stallcup WB, Opdenakker G, Yong VW. (2003) Matrix Metalloproteinase-9 Facilitates Remyelination in Part by Processing the Inhibitory NG2 Proteoglycan. J. Neurosci. 23: 11127-35.

Lassmann H, Brück W, Lucchinetti C. (2001) Heterogeneity of multiple sclerosis pathogenesis: implications for diagnosis and therapy. Trends Mol. Med. 7:115-21.

Lazebnik YA, Kaufmann SH, Desnoyers S, Poirier GG, Earnshaw WC. (1994) Cleavage of poly(ADP-ribose) polymerase by a proteinase with properties like ICE. Nature. 371:346–7.

Leavenworth JW, Schellack C, Kim HJ, Lu L, Spee P, Cantor H. (2010) Analysis of the cellular mechanism underlying inhibition of EAE after treatment with anti- NKG2A F(ab’)2. Proc. Natl. Acad. Sci. 107:2562-7.

Leclercq G, Debacker V, de Smedt M, Plum J. (1996) Differential effects of interleukin-15 and interleukin-2 on differentiation of bipotential T/natural killer progenitor cells. J. Exp. Med. 184:325-36.

Le Dréan E, Vély F, Olcese L, Cambiaggi A, Guia S, Krystal G, Gervois N, Moretta A, Jotereau F, Vivier E. (1998) Inhibition of antigen-induced T cell response and antibody-induced NK cell cytotoxicity by NKG2A: association of NKG2A with SHP- 1 and SHP-2 protein-tyrosine phosphatases. Eur. J. Immunol. 28:264-76.

255

Lee N, Llano M, Carretero M, Ishitani A, Navarro F, López-Botet M, Geraghty DE. (1998) HLA-E is a major ligand for the natural killer inhibitory receptor CD94/NKG2A. Proc. Natl. Acad. Sci. USA. 95:5199-204.

Lee SC, Brummet ME, Shahabuddin S, Woodworth TG, Georas SN, Leiferman KM, Gilman SC, Stellato C, Gladue RP, Schleimer RP, Beck LA. (2000) Cutaneous injection of human subjects with macrophage inflammatory protein-1 alpha induces significant recruitment of neutrophils and monocytes. J. Immunol. 164:3392-401.

Lee SJ, Jang BC, Lee SW, Yang YI, Suh SI, Park YM, Oh S, Shin JG, Yao S, Chen L, Choi IH. (2006) Interferon regulatory factor-1 is prerequisite to the constitutive expression and IFN-gamma-induced upregulation of B7-H1 (CD274). FEBS Lett. 580:755-62.

Le Garff-Tavernier M, Béziat V, Decocq J, Siguret V, Gandjbakhch F, Pautas E, Debré P, Merle-Beral H, Vieillard V. (2010) Human NK cells display major phenotypic and functional changes over the life span. Aging Cell. 9:527-35.

Leonard EJ, Yoshimura T, Tanaka S, Raffled M. (1991) Neutrophil recruitment by intradermal injected neutrophil attractant/activation protein-1. J. Invest. Dermatol. 96:690–694.

Le Page E, Leray E, Taurin G, Coustans M, Chaperon J, Morrissey SP, Edan G. (2008) Mitoxantrone as induction treatment in aggressive relapsing remitting multiple sclerosis: treatment response factors in a 5 year follow-up observational study of 100 consecutive patients. J. Neurol. Neurosurg. Psychiatry. 79:52-6.

Le Page E, Leray E, Edan G; French Mitoxantrone Safety Group. Long-term safety profile of mitoxantrone in a French cohort of 802 multiple sclerosis patients: a 5-year prospective study. (2011) Mult. Scler. 17:867-75.

Leppert D, Waubant E, Galardy R, Bunnett NW, Hauser SL. (1995) T cell gelatinases mediate basement membrane transmigration in vitro. J. Immunol. 154:4379-89.

Leppert D, Waubant E, Bürk MR, Oksenberg JR, Hauser SL. (1996) Interferon beta- 1b inhibits gelatinase secretion and in vitro migration of human T cells: a possible mechanism for treatment efficacy in multiple sclerosis. Ann. Neurol. 40:846-52.

256

Leung K.H. (1989) Inhibition of Human Natural Killer Cell and Lymphokine- Activated Killer Cell Cytotoxicity and Differentiation by Vitamin D3. Scand. J. Immunol. 30:199–208.

Levin LI, Munger KL, Rubertone MV, Peck CA, Lennette ET, Spiegelmann D, Ascherio A. (2005) Temporal relationship between elevation of Epstein-Barr virus antibody titers and initial onset of neurological symptoms in multiple sclerosis. JAMA. 293:2496 – 500 .

Levin LI, Munger KL, O'Reilly EJ, Falk KI, Ascherio A. (2010) Primary infection with the Epstein-Barr virus and risk of multiple sclerosis. Ann. Neurol. 67:824-30.

Levison SW, Ducceschi MH, Young GM, Wood TL. (1996) Acute exposure to CNTF in vivo induces multiple components of reactive gliosis. Exp. Neurol. 141:256–68.

Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M, Alnemri ES, Wang X. (1997) Cytochrome c and dATP-dependent formation of Apaf-1/caspase-9 complex initiates an apoptotic protease cascade. Cell. 91:479-89.

Li H, Zhu H, Xu CJ, Yuan J. (1998) Cleavage of BID by caspase 8 mediates the mitochondrial damage in the Fas pathway of apoptosis. Cell. 94:491-501.

Li W, Maeda Y, Ming X, Cook S, Chapin J, Husar W, Dowling F. (2002) Apoptotic death following Fas activation in human oligodendrocyte hybrid cultures. J. Neurosci. Res. 69:189-96.

Li A, Dubey S, Varney ML, Dave BJ, Singh RK. (2003) IL-8 directly enhanced endothelial cell survival, proliferation, and matrix metalloproteinases production and regulated angiogenesis. J. Immunol. 170:3369-76.

Li J, Baud O, Vartanian T, Volpe JJ, Rosenberg PA. (2005) Peroxynitrite generated by inducible nitric oxide synthase and NADPH oxidase mediates microglial toxicity to oligodendrocytes. Proc. Natl. Acad. Sci. U.S.A. 102: 9936–41.

Li X, McKinstry KK, Swain SL, Dalton DK. (2007) IFN-gamma acts directly on activated CD4+ T cells during mycobacterial infection to promote apoptosis by

257 inducing components of the intracellular apoptosis machinery and by inducing extracellular proapoptotic signals. J. Immunol. 179:939-49.

Li H, Lin X. (2008) Positive and negative signaling components involved in TNFalpha-induced NF-kappaB activation. Cytokine. 41:1-8.

Li N, McLaren JE, Michael DR, Clement M, Fielding CA, Ramji DP. (2010a) ERK is integral to the IFN-γ-mediated activation of STAT1, the expression of key genes implicated in atherosclerosis, and the uptake of modified lipoproteins by human macrophages. J. Immunol. 185:3041-8.

Li WX, Pan HF, Hu JL, Wang CZ, Zhang N, Li J, Li XP, Xu JH, Ye DQ. (2010b) Assay of T- and NK-cell subsets and the expression of NKG2A and NKG2D in patients with new-onset systemic lupus erythematosus. Clin. Rheumatol. 29:315-23.

Liberto CM, Albrecht PJ, Herx LM, Yong VW, Levison SW. (2004) Pro-regenerative properties of cytokine-activated astrocytes. J. Neurochem. 89:1092–1100.

Lieberman J. (2003) The ABCs of granule-mediated cytotoxicity: New weapons in the arsenal. Nat. Rev. Immunol. 3: 361-70.

Ligers A, Dyment DA, Willer CJ, Sadovnick AD, Ebers G, Risch N, Hillert J; Canadian Collaborative Study Groups. (2001) Evidence of linkage with HLA-DR in DRB1*15-negative families with multiple sclerosis. Am. J. Hum. Genet. 69:900-3.

Lighvani AA, Frucht DM, Jankovic D, Yamane H, Aliberti J, Hissong BD, Nguyen BV, Gadina M, Sher A, Paul WE, O'Shea JJ. (2001) T-bet is rapidly induced by interferon-gamma in lymphoid and myeloid cells. Proc. Natl. Acad. Sci. U.S.A. 98:15137-42.

Lillard JW Jr, Boyaka PN, Taub DD, McGhee JR. (2001) RANTES potentiates antigen-specific mucosal immune responses. J Immunol. 166:162-9.

Lillard JW Jr, Singh UP, Boyaka PN, Singh S, Taub DD, McGhee JR. (2003) MIP- 1alpha and MIP-1beta differentially mediate mucosal and systemic adaptive immunity. Blood. 101:807-14.

258

Lin Y, Devin A, Rodriguez Y, Liu ZG. (1999) Cleavage of the death domain kinase RIP by caspase-8 prompts TNF-induced apoptosis. Genes Dev. 13:2514-26.

Lincoln MR, Ramagopalan SV, Chao MJ, Herrera BM, Deluca GC, Orton SM, Dyment DA, Sadovnick AD, Ebers GC. (2009) Epistasis among HLA-DRB1, HLA- DQA1, and HLA-DQB1 loci determines multiple sclerosis susceptibility. Proc. Natl. Acad. Sci. U.S.A. 106:7542-7.

Lindberg C, Andersen O, Vahlne A, Dalton M, Runmarker B. (1991) Epidemiological investigation of the association between infectious mononucleosis and multiple sclerosis. Neuroepidemiology. 10:62-5.

Ling PD, Warren MK, Vogel SN. (1985) Antagonistic effect of interferon-beta on the interferon-gamma-induced expression of Ia antigen in murine macrophages. J. Immunol. 135:1857-63.

Liu X, Zou H, Widlak P, Garrard W, Wang X. (1999) Activation of the apoptotic endonuclease DFF40 (caspase-activated DNase or nuclease). Oligomerization and direct interaction with histone H1. J Biol. Chem. 274:13836–13840.

Ljunggren HG, Ohlen C, Hoglund P, Yamasaki T, Klein G, Karre K. (1988) Afferent and efferent cellular interactions in natural resistance directed against MHC class I deficient tumor grafts. J. Immunol. 140:671-8.

Ljunggren H, Karre K. (1985) Host resistance directed selectively against H-2- deficient lymphoma variants. Analysis of the mechanism. J. Exp. Med. 162:1745-59.

Lohoff M, Ferrick D, Mittrucker HW, Duncan GS, Bischof S, Rollinghoff M, Mak TW. (1997) Interferon regulatory factor-1 is required for a T helper 1 immune response in vivo. Immunity. 6:681–89.

Long BR, Michaelsson J, Loo CP, Ballan WM, Vu BA, Hecht FM, Lanier LL, Chapman JM, Nixon DF. (2008) Elevated frequency of gamma interferon-producing NK cells in healthy adults vaccinated against influenza virus. Clin. Vaccine Immunol. 15:120-30.

259

Lopez AF, Begley CG, Williamson DJ, Warren DJ, Vadas MA, Sanderson CJ. (1986) Murine eosinophil differentiation factor. An eosinophil-specific colony-stimulating factor with activity for human cells. J. Exp. Med. 163:1085-99.

Lopez AF, Sanderson CJ, Gamble JR, Campbell HD, Young IG, Vadas MA. (1988) Recombinant human interleukin 5 is a selective activator of human eosinophil function. J. Exp. Med. 167:219-224.

Lopez-Vergès S, Milush JM, Pandey S, York VA, Arakawa-Hoyt J, Pircher H, Norris PJ, Nixon DF, Lanier LL. (2010) CD57 defines a functionally distinct population of mature NK cells in the human CD56dimCD16+ NK-cell subset. Blood. 116:3865-74.

Lopez-Vergès S, Milush JM, Schwartz BS, Pando MJ, Jarjoura J, York VA, Houchins JP, Miller S, Kang SM, Norris PJ, Nixon DF, Lanier LL. (2011) Expansion of a unique CD57⁺NKG2Chi natural killer cell subset during acute human cytomegalovirus infection. Proc. Natl. Acad. Sci. USA. 108:14725-32.

Lorentzen AR, Karlsen TH, Olsson M, Smestad C, Mero IL, Woldseth B, Sun JY, Senitzer D, Celius EG, Thorsby E, Spurkland A, Lie BA, Harbo HF. (2009) Killer immunoglobulin-like receptor ligand HLA-Bw4 protects against multiple sclerosis. Ann. Neurol. 65:658-66.

Losseff NA, Wang L, Lai HM, Yoo DS, Gawne-Cain ML, McDonald WI, Miller DH, Thompson AJ. (1996) Progressive cerebral atrophy in multiple sclerosis. A serial MRI study. Brain. 119:2009-19.

Lou J, Gasche Y, Zheng L, Giroud C, Morel P, Clements J, Ythier A, Grau GE. (1999) Interferon-beta inhibits activated leukocyte migration through human brain microvascular endothelial cell monolayer. Lab. Invest. 79:1015-25.

Lou Z, Jevremovic D, Billadeau DD, Leibson PJ. (2000) A Balance between Positive and Negative Signals in Cytotoxic Lymphocytes Regulates the Polarization of Lipid Rafts during the Development of Cell-Mediated Killing. J, Exp. Med. 191:347-54.

Lowdell MW, Craston R, Samuel D, Wood ME, O'Neill E, Saha V, Prentice HG. (2002) Evidence that continued remission in patients treated for acute leukaemia is dependent upon autologous natural killer cells. Br. J. Haematol. 117:821-7.

260

Loza MJ, Perussia B. (2004) The IL-12 signature: NK cell terminal CD56+high stage and effector functions. J. Immunol. 172:88-96.

Lu H, Hou Q, Zhao T, Zhang H, Zhang Q, Wu L, Fan Z. (2006) Granzyme M directly cleaves inhibitor of caspase-activated DNase (CAD) to unleash CAD leading to DNA fragmentation. J. Immunol. 177:1171-8.

Lu L, Ikizawa K, Hu D, Werneck MB, Wucherpfennig KW, Cantor H. (2007) Regulation of activated CD4+ T cells by NK cells via the Qa-1-NKG2A inhibitory pathway. Immunity. 26:593-604.

Lucchinetti C, Bruck W, Parisi J, Scheithauer B, Rodriguez M, Lassmann H. (2000) Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of demyelination. Ann. Neurol. 47:707–717.

Lundmark F, Duvefelt K, Iacobaeus E, Kockum I, Wallström E, Khademi M, Oturai A, Ryder LP, Saarela J, Harbo HF, Celius EG, Salter H, Olsson T, Hillert J. (2007) Variation in interleukin 7 receptor alpha chain (IL7R) influences risk of multiple sclerosis. Nat. Genet. 39:1108-13.

Lünemann A, Lünemann JD, Roberts S, Messmer B, Barreira da Silva R, Raine CS, Münz C. (2008) Human NK Cells Kill Resting but Not Activated Microglia via NKG2D- and NKp46-Mediated Recognition. J. Immunol. 181:6170-7.

Lünemann JD, Tintoré M, Messmer B, Strowig T, Rovira A, Perkal H, Caballero E, Münz C, Montalban X, Comabella M. (2010) Elevated Epstein-Barr virus-encoded nuclear antigen-1 immune responses predict conversion to multiple sclerosis. Ann. Neurol. 67:159-69.

Lünemann A, Tackenberg B, DeAngelis T, da Silva RB, Messmer B, Vanoaica LD, Miller A, Apatoff B, Lublin FD, Lünemann JD, Münz C. (2011) Impaired IFN-γ production and proliferation of NK cells in multiple sclerosis. Int. Immunol. 23:139- 48.

Lutz CT, Moore MB, Bradley S, Shelton BJ, Lutgendorf SK. (2005) Reciprocal age related changes in natural killer cell receptors for MHC class I. Mech. Ageing Dev. 126:722-31.

261

Lysandropoulos AP, Jaquiéry E, Jilek S, Pantaleo G, Schluep M, Du Pasquier RA. (2011) Vitamin D has a direct immunomodulatory effect on CD8+ T cells of patients with early multiple sclerosis and healthy control subjects. J. Neuroimmunol. 233:240- 4.

MacDonald G, Shi L, Vande Velde C, Lieberman J, Greenberg AH. (1999) Mitochondria-dependent and -independent regulation of Granzyme B-induced apoptosis. J. Exp. Med. 189:131-44.

Mace EM, Zhang J, Siminovitch KA, Takei F. (2010) Elucidation of the integrin LFA-1-mediated signaling pathway of actin polarization in natural killer cells. Blood. 116:1272-1279.

Macian F. (2005) NFAT proteins: key regulators of T-cell development and function. Nat. Rev. Immunol. 5:472-84.

Magliozzi R, Howell O, Vora A, Serafini B, Nicholas R, Puopolo M, Reynolds R, Aloisi F. (2007) Meningeal B-cell follicles in secondary progressive multiple sclerosis associate with early onset of disease and severe cortical pathology. Brain. 130:1089- 104.

Magnus T, Chan A, Grauer O, Toyka KV, Gold R. (2001) Microglial phagocytosis of apoptotic inflammatory T cells leads to down-regulation of microglial immune activation. J. Immunol. 167:5004-10.

Mahrus S, Kisiel W, Craik CS. (2004) Granzyme M is a regulatory protease that inactivates proteinase inhibitor 9, an endogenous inhibitor of granzyme B. J. Biol. Chem. 279:54275-85.

Maier LM, Lowe CE, Cooper J, Downes K, Anderson DE, Severson C, Clark PM, Healy B, Walker N, Aubin C, Oksenberg JR, Hauser SL, Compston A, Sawcer S; International Multiple Sclerosis Genetics Consortium, De Jager PL, Wicker LS, Todd JA, Hafler DA. (2009a) IL2RA genetic heterogeneity in multiple sclerosis and type 1 diabetes susceptibility and soluble interleukin-2 receptor production. PLoS Genet. 5:e1000322.

262

Maier LM, Anderson DE, Severson CA, Baecher-Allan C, Healy B, Liu DV, Wittrup KD, De Jager PL, Hafler DA. (2009b) Soluble IL-2RA levels in multiple sclerosis subjects and the effect of soluble IL-2RA on immune responses. J. Immunol. 182:1541-7.

Mailliard RB, Son YI, Redlinger R, Coates PT, Giermasz A, Morel PA, Storkus WJ, Kalinski P. (2003) Dendritic cells mediate NK cell help for Th1 and CTL responses: two-signal requirement for the induction of NK cell helper function. J. Immunol. 171:2366-73.

Mailliard RB, Alber SM, Shen H, Watkins SC, Kirkwood JM, Herberman RB, Kalinski P. (2005) IL-18-induced CD83+CCR7+ NK helper cells. J. Exp. Med. 202:941-53.

Malipiero U, Heuss C, Schlapbach R, Tschopp J, Gerber U, Fontana A. (1999) Involvement of the N-methyl-D-aspartate receptor in neuronal cell death induced by cytotoxic T cell-derived secretory granules. Eur. J. Immunol. 29:3053-62.

Mandala S, Hajdu R, Bergstrom J, Quackenbush E, Xie J, Milligan J, Thornton R, Shei GJ, Card D, Keohane C, Rosenbach M, Hale J, Lynch CL, Rupprecht K, Parsons W, Rosen H. (2002) Alteration of lymphocyte trafficking by sphingosine-1-phosphate receptor agonists. Science. 296:346-9.

Mandelboim O, Lieberman N, Lev M, Paul L, Arnon TI, Bushkin Y, Davis DM, Strominger JL, Yewdell JW, Porgador A. (2001) Recognition of haemagglutinins on virus-infected cells by NKp46 activates lysis by human NK cells. Nature. 409:1055- 60.

Manel N, Unutmaz D, Littman DR. (2008) The differentiation of human T(H)-17 cells requires transforming growth factor-beta and induction of the nuclear receptor RORgammat. Nat. Immunol. 9:641-9.

Manna SK, Aggarwal BB. (1998) IL-13 suppresses TNF-induced activation of nuclear factor-kappa B, activation protein-1, and apoptosis. J. Immunol. 161:2863-72.

263

Maroof A, Beattie L, Zubairi S, Svensson M, Stager S, Kaye PM. (2008) Posttranscriptional regulation of II10 gene expression allows natural killer cells to express immunoregulatory function. Immunity. 29:295-305.

Marqués L, Brucet M, Lloberas J, Celada A. (2004) STAT1 regulates lipopolysaccharide- and TNF-alpha-dependent expression of transporter associated with antigen processing 1 and low molecular mass polypeptide 2 genes in macrophages by distinct mechanisms. J. Immunol. 173:1103-10.

Martí F, Bertran E, Llucià M, Villén E, Peiró M, Garcia J, Rueda F. (2002) Platelet factor 4 induces human natural killer cells to synthesize and release interleukin-8. J. Leukoc. Biol. 72:590-7.

Martin DL. (1992) Synthesis and release of neuroactive substances by glial cells. Glia. 5:81–94.

Martin MP, Gao X, Lee JH, Nelson GW, Detels R, Goedert JJ, Buchbinder S, Hoots K, Vlahov D, Trowsdale J, Wilson M, O'Brien SJ, Carrington M. (2002) Epistatic interaction between KIR3DS1 and HLA-B delays the progression to AIDS. Nat. Genet. 31:429-34.

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. (2007) Innate partnership of HLA-B and KIR3DL1 subtypes against HIV-1. Nat. Genet. 39:733-40.

Martin JF, Perry JS, Jakhete NR, Wang X, Bielekova B. (2010) An IL-2 paradox: blocking CD25 on T cells induces IL-2-driven activation of CD56(bright) NK cells. J. Immunol. 185:1311-20.

Martin R. (2012) Anti-CD25 (daclizumab) monoclonal antibody therapy in relapsing- remitting multiple sclerosis. Clin. Immunol. 142:9-14.

Martinelli Boneschi F, Rovaris M, Johnson KP, Miller A, Wolinsky JS, Ladkani D, Shifroni G, Comi G, Filippi M. (2003) Effects of glatiramer acetate on relapse rate

264 and accumulated disability in multiple sclerosis: meta-analysis of three double-blind, randomized, placebo-controlled clinical trials. Mult. Scler. 9:349-55.

Martínez-Rodríguez JE, Saez-Borderías A, Munteis E, Romo N, Roquer J, López- Botet M. (2010) Natural killer receptors distribution in multiple sclerosis: Relation to clinical course and interferon-beta therapy. Clin. Immunol. 137:41-50.

Martínez-Rodríguez JE, López-Botet M, Munteis E, Rio J, Roquer J, Montalban X, Comabella M. (2011) Natural killer cell phenotype and clinical response to interferon- beta therapy in multiple sclerosis. Clin. Immunol. 141:348-56.

Martín-Fontecha A, Thomsen LL, Brett S, Gerard C, Lipp M, Lanzavecchia A, Sallusto F. (2004) Induced recruitment of NK cells to lymph nodes provides IFN- gamma for T(H)1 priming. Nat. Immunol. 5:1260-5.

Martinvalet D, Zhu P, Lieberman J. (2005) Granzyme A induces caspase-independent mitochondrial damage, a required first step for apoptosis. Immunity. 22:355-70.

Masilamani M, Nguyen C, Kabat J, Borrego F, Coligan JE. (2006) CD94/NKG2A Inhibits NK Cell Activation by Disrupting the Actin Network at the Immunological Synapse. J. Immunol. 177:3590-6.

Matloubian M, Lo CG, Cinamon G, Lesneski MJ, Xu Y, Brinkmann V, Allende ML, Proia RL, Cyster JG. (2004) Lymphocyte egress from thymus and peripheral lymphoid organs is dependent on S1P receptor 1. Nature. 427:355-360.

Matos ME, Schnier GS, Beecher MS, Ashman LK, William DE, Caligiuri MA. (1993) Expression of a functional c-kit receptor on a subset of natural killer cells. J. Exp. Med. 178:1079-84.

Matsuyama T, Kimura T, Kitagawa M, Pfeffer K, Kawakami T, Watanabe N, Kündig TM, Amakawa R, Kishihara K, Wakeham A, Potter J, Furlonger CL, Narendran A, Suzuki H, Ohashi PS, Paige CJ, Taniguchi T, Mak TW. (1993) Targeted disruption of IRF-1 or IRF-2 results in abnormal type I IFN gene induction and aberrant lymphocyte development. Cell. 75:83–97.

265

Matsuzawa T, Kim BH, Shenoy AR, Kamitani S, Miyake M, Macmicking JD. (2012) IFN-γ elicits macrophage autophagy via the p38 MAPK signaling pathway. J. Immunol. 189:813-8.

Mattson MP, Goodman Y, Luo H, Fu W, Furukawa K. (1997) Activation of NF- kappaB protects hippocampal neurons against oxidative stress-induced apoptosis: evidence for induction of manganese superoxide dismutase and suppression of peroxynitrite production and protein tyrosine nitration. J. Neurosci. Res. 49:681-97.

Maysami S, Nguyen D, Zobel F, Pitz C, Heine S, Hopfner M, Stangel M. (2006) Modulation of rat oligodendrocyte precursor cells by the chemokine CXCL12. Neuroreport. 17:1187–90.

McCann FE, Vanherberghen B, Eleme K, Carlin LM, Newsam RJ, Goulding D, Davis DM. (2003) The size of the synaptic cleft and distinct distributions of filamentous actin, ezrin, CD43, and CD45 at activating and inhibitory human NK cell immune synapses. J. Immunol. 170:2862-70.

McKenzie GJ, Fallon PG, Emson CL, Grencis RK, McKenzie ANJ. (1999) Simultaneous Disruption of Interleukin (IL)-4 and IL-13 Defines Individual Roles in T Helper Cell Type 2-mediated Responses. J. Exp. Med. 189: 1565-72.

McRae BL, Semnani RT, Hayes MP, van Seventer GA. (1998) Type I IFNs inhibit human dendritic cell IL-12 production and Th1 cell development. J. Immunol. 160:4298–4304.

Mead RJ, Singhrao SK, Neal JW, Lassmann H, Morgan BP. (2002) The membrane attack complex of complement causes severe demyelination associated with acute axonal injury. J. Immunol. 168:458–65.

Medema JP, Toes RE, Scaffidi C, Zheng TS, Flavell RA, Melief CJ, Peter ME, Offringa R, Krammer PH. (1997a) Cleavage of FLICE (caspase-8) by granzyme B during cytotoxic T lymphocyte-induced apoptosis. Eur. J. Immunol. 27:3492-8.

Medema JP, Scaffidi C, Kischkel FC, Shevchenko A, Mann M, Krammer PH, Peter ME. (1997b) FLICE is activated by association with the CD95 death-inducing signaling complex (DISC). EMBO J. 16:2794-804.

266

Mehrotra PT, Donnelly RP, Wong S, Kanegane H, Geremew A, Mostowski HS, Furuke K, Siegel JP, Bloom ET. (1998) Production of IL-10 by human natural killer cells stimulated with IL-2 and/or IL-12. J. Immunol. 160:2637-44.

Mellergård J, Edström M, Vrethem M, Ernerudh J, Dahle C. (2010) Natalizumab treatment in multiple sclerosis: marked decline of chemokines and cytokines in cerebrospinal fluid. Mult. Scler. 16:208-17.

Melzer N, Meuth SG, Wiendl H. (2009) CD8+ T cells and neuronal damage: direct and collateral mechanisms of cytotoxicity and impaired electrical excitability. FASEB J. 23:3659-73.

Menèndez Iglesias B, Cerase J, Ceracchini C, Levi G, Aloisi F. (1997) Analysis of B7-1 and B7-2 costimulatory ligands in cultured mouse microglia: upregulation by interferon-gamma and lipopolysaccharide and downregulation by interleukin-10, prostaglandin E2 and cyclic AMP-elevating agents. J. Neuroimmunol. 72:83-93.

Menge T, Lalive PH, von Büdingen HC, Genain CP. (2011) Conformational epitopes of myelin oligodendrocyte glycoprotein are targets of potentially pathogenic antibody responses in multiple sclerosis. J. Neuroinflammation. 8:161.

Mentlik AN, Sanborn KB, Holzbaur EL, Orange JS. (2010) Rapid lytic granule convergence to the MTOC in natural killer cells is dependent on dynein but not cytolytic commitment. Mol. Biol. Cell. 21:2241-56.

Merino F, Alvarez-Mon M, de la Hera A, Ales JE, Bonilla F, Durantez A. (1989) Regulation of natural killer cytotoxicity by 1,25-dihydroxyvitamin D3. Cell. Immunol. 118:328–36.

Merrill JE, Kono DH, Clayton J, Ando DG, Hinton DR, Hofman FM. (1992) Inflammatory leukocytes and cytokines in the peptide-induced disease of experimental allergic encephalomyelitis in SJL and B10.PL mice. Proc. Natl. Acad. Sci. U.S.A. 89:574-8.

Merrill JE, Ignarro LJ, Sherman MP, Melinek J, Lane TE. (1993) Microglial cell cytotoxicity of oligodendrocytes is mediated through nitric oxide. J. Immunol. 151:2132-41.

267

Metcalf D, Begley CG, Johnson GR, Nicola NA, Vadas MA, Lopez AF, Williamson DJ, Wong GG, Clark SC, Wang EA. (1986) Biologic properties in vitro of a recombinant human granulocyte-macrophage colony-stimulating factor. Blood. 67:37- 45.

Metkar SS, Menaa C, Pardo J, Wang B, Wallich R, Freudenberg M, Kim S, Raja SM, Shi L, Simon MM, Froelich CJ. (2008) Human and mouse granzyme A induce a proinflammatory cytokine response. Immunity. 29:720-33.

Meuth SG, Herrmann AM, Simon OJ, Siffrin V, Melzer N, Bittner S, Meuth P, Langer HF, Hallermann S, Boldakowa N, Herz J, Munsch T, Landgraf P, Aktas O, Heckmann M, Lessmann V, Budde T, Kieseier BC, Zipp F, Wiendl H. (2009) Cytotoxic CD8+ T cell-neuron interactions: perforin-dependent electrical silencing precedes but is not causally linked to neuronal cell death. J. Neurosci. 29:15397-409.

Michie AM, Carlyle JR, Schmitt TM, Ljutic B, Cho SK, Fong Q, Zúñiga-Pflücker JC. (2000) Clonal characterization of a bipotent T cell and NK cell progenitor in the mouse fetal thymus. J. Immunol. 164:1730-3.

Mikulkova Z, Praksova P, Stourac P, Bednarik J, Strajtova L, Pacasova R, Belobradkova J, Dite P, Michalek J. (2010) Numerical defects in CD8+CD28- T- suppressor lymphocyte population in patients with type 1 diabetes mellitus and multiple sclerosis. Cell. Immunol. 262:75-9.

Min, W, Pober JS, Johnson DR. (1996) Kinetically coordinated induction of TAP1 and HLA class I by IFN-γ: the rapid induction of TAP1 by IFN-γ is mediated by Stat1α. J. Immunol. 156:3174.

Miron VE, Jung CG, Kim HJ, Kennedy TE, Soliven B, Antel JP. (2008) FTY720 modulates human oligodendrocyte progenitor process extension and survival. Ann. Neurol. 63:61-71.

Mitsumoto Y, Oka S, Sakuma H, et al. (2000). Cloning and chromosomal mapping of human glucuronyltransferase involved in biosynthesis of the HNK-1 carbohydrate epitope. Genomics. 65: 166–73.

268

Mocikat R, Braumüller H, Gumy A, Egeter O, Ziegler H, Reusch U, Bubeck A, Louis J, Mailhammer R, Riethmüller G, Koszinowski U, Röcken M. (2003) Natural killer cells activated by MHC class I(low) targets prime dendritic cells to induce protective CD8 T cell responses. Immunity. 19:561-9.

Moffett-King A. (2002) Natural killer cells and pregnancy. Nat. Rev. Immunol. 2:656- 63.

Molinero LL, Fuertes MB, Rabinovich GA, Fainboim L, Zwirner NW. (2002) Activation-induced expression of MICA on T lymphocytes involves engagement of CD3 and CD28. J. Leukoc. Biol. 71:791-7.

Molinero LL, Domaica CI, Fuertes MB, Girart MV, Rossi LE, Zwirner NW. (2006) Intracellular expression of MICA in activated CD4 T lymphocytes and protection from NK cell-mediated MICA-dependent cytotoxicity. Hum. Immunol. 67:170-82.

Molnarfi N, Hyka-Nouspikel N, Gruaz L, Dayer J-M, Burger D. (2005) The Production of IL-1 Receptor Antagonist in IFN-beta-Stimulated Human Monocytes Depends on the Activation of Phosphatidylinositol 3-Kinase but Not of STAT1. J. Immunol. 174:2974-80.

Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC, Biassoni R, Moretta L. (2001) Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu. Rev. Immunol. 19: 197-223

Morgan MJ, Kim YS, Liu ZG. (2008) TNFalpha and reactive oxygen species in necrotic cell death. Cell Res. 18:343-9.

Morikawa K, Oseko F, Morikawa S, Imai K, Sawada M. (1993) Recombinant human IL-5 augments immunoglobulin generation by human B lymphocytes in the presence of IL-2. Cell. Immunol. 149:390-401.

Morse RHA, Seguin R, McCrea EL, Antel JP. (2001) NK cell-mediated lysis of autologous human oligodendrocytes. J. Neuroimmunol. 116:107-15.

Motyka B, Korbutt G, Pinkoski MJ, Heibein JA, Caputo A, Hobman M, Barry M, Shostak I, Sawchuk T, Holmes CF, Gauldie J, Bleackley RC. (2000) Mannose 6-

269 phosphate/insulin-like growth factor II receptor is a death receptor for granzyme B during cytotoxic T cell-induced apoptosis. Cell. 103:491-500.

Mrózek E, Anderson P, Caligiuri MA. (1996) Role of interleukin-15 in the development of human CD56+ natural killer cells from CD34+ hematopoietic progenitor cells. Blood. 87:2632-40.

Muchamuel T, Menon S, Pisacane P, Howard MC, Cockayne DA. (1997) IL-13 protects mice from lipopolysaccharide-induced lethal endotoxemia: correlation with down-modulation of TNF-α, IFN-γ, and IL-12 production. J. Immunol. 158: 2898

Mullershausen F, Craveiro LM, Shin Y, Cortes-Cros M, Bassilana F, Osinde M, Wishart WL, Guerini D, Thallmair M, Schwab ME, Sivasankaran R, Seuwen K, Dev KK. (2007) Phosphorylated FTY720 promotes astrocyte migration through sphingosine-1-phosphate receptors. J. Neurochem. 102:1151-61.

Munger KL, Zhang SM, O'Reilly E, Hernan MA, Olek MJ, Willett WC, Ascherio A. (2004) Vitamin D intake and incidence of multiple sclerosis. Neurology. 62:60-65.

Munger KL, Levin LI, Hollis BW, Howard NS, Ascherio A. (2006) Serum 25- hydroxyvitamin D levels and risk of multiple sclerosis. JAMA. 296:2832-8.

Munschauer FE, Hartrich LA, Stewart CC, Jacobs L. (1995) Circulating natural killer cells but not cytotoxic T lymphocytes are reduced in patients with active relapsing multiple sclerosis and little clinical disability as compared to controls. J. Neuroimmunol. 62:177-81.

Muntasell A, Magri G, Pende D, Angulo A, López-Botet M. (2010) Inhibition of NKG2D expression in NK cells by cytokines secreted in response to human cytomegalovirus infection. Blood. 115:5170-9.

Murugaiyan G, Mittal A, Weiner HL. (2010) Identification of an IL-27/osteopontin axis in dendritic cells and its modulation by IFN-gamma limits IL-17-mediated autoimmune inflammation. Proc. Natl. Acad. USA. 107:11495-500.

Muzio M, Chinnaiyan AM, Kischkel FC, O'Rourke K, Shevchenko A, Ni J, Scaffidi C, Bretz JD, Zhang M, Gentz R, Mann M, Krammer PH, Peter ME, Dixit VM.

270

(1996) FLICE, a novel FADD homologous ICE/CED-3-like protease, is recruited to the CD95 (Fas/Apo-1) death-inducing signaling complex. Cell. 85:817–27.

Muzio M, Stockwell BR, Stennicke HR, Salvesen GS, Dixit VM. (1998) An induced proximity model for caspase-8 activation. J. Biol. Chem. 273:2926–2930.

Nagasawa M, Ohshiba A, Yata J. (1991) Effect of recombinant interleukin 5 on the generation of cytotoxic T cells (CTL). Cell. Immunol. 133:317–326.

Nagler A, Lanier LL, Cwirla S, Phillips JH. (1989) Comparative studies of human FcRIII-positive and negative natural killer cells. J. Immunol. 143:3183-91.

Nagler A, Lanier LL, Phillips JH. (1990) Constitutive expression of high affinity interleukin 2 receptors on human CD16-natural killer cells in vivo. J. Exp. Med. 171:1527-33.

Nair JS, DaFonseca CJ, Tjernberg A, Sun W, Darnell JE Jr, Chait BT, Zhang JJ. (2002) Requirement of Ca2+ and CaMKII for Stat1 Ser-727 phosphorylation in response to IFN-gamma. Proc. Natl. Acad. Sci. USA. 99:5971-6.

Nair A, Frederick TJ, Miller SD. (2008) Astrocytes in multiple sclerosis: a product of their environment. Cell. Mol. Life Sci. 65:2702-20.

Narimatsu M, Maeda H, Itoh S, Atsumi T, Ohtani T, Nishida K, Itoh M, Kamimura D, Park SJ, Mizuno K, Miyazaki J, Hibi M, Ishihara K, Nakajima K, Hirano T. (2001) Tissue-specific autoregulation of the stat3 gene and its role in interleukin-6-induced survival signals in T cells. Mol. Cell. Biol. 21:6615-25.

Narni-Mancinelli E, Jaeger BN, Bernat C, Fenis A, Kung S, De Gassart A, Mahmood S, Gut M, Heath SC, Estellé J, Bertosio E, Vely F, Gastinel LN, Beutler B, Malissen B, Malissen M, Gut IG, Vivier E, Ugolini S. (2012) Tuning of natural killer cell reactivity by NKp46 and Helios calibrates T cell responses. Science. 335:344-8.

Nathan CF, Murray HW, Wiebe ME, Rubin BY. (1983) Identification of interferon- gamma as the lymphokine that activates human macrophage oxidative metabolism and antimicrobial activity. J. Exp. Med. 158:670-89.

271

Nedvetzki S, Sowinski S, Eagle RA, Harris J, Vély F, Pende D, Trowsdale J, Vivier E, Gordon S, Davis DM. (2007) Reciprocal regulation of human natural killer cells and macrophages associated with distinct immune synapses. Blood. 109:3776-85.

Neefjes JJ, Momburg F, Hammerling GJ. (1993) Selective and ATP-dependent translocation of peptides by the MHC-encoded transporter. Science. 261:769.

Neote K, DiGregorio D, Mak JY, Horuk R, Schall TJ. (1993) Molecular cloning, functional expression, and signaling characteristics of a C-C chemokine receptor. Cell. 72:415-25.

Neuhaus O, Wiendl H, Kieseier BC, Archelos JJ, Hemmer B, Stüve O, Hartung HP. (2005) Multiple sclerosis: Mitoxantrone promotes differential effects on immunocompetent cells in vitro. J. Neuroimmunol. 168:128-37.

Neumann H. (2003) Molecular mechanisms of axonal damage in inflammatory central nervous system diseases. Curr. Opin. Neurol. 16:267-73.

Newburger PE, Ezekowitz RA, Whitney C, Wright J, Orkin SH. (1988) Induction of phagocyte cytochrome b heavy chain gene expression by interferon gamma. Proc. Natl. Acad. Sci. USA. 85:5215-9.

Newman TA, Woolley ST, Hughes PM, Sibson NR, Anthony DC, Perry VH. (2001) T-cell- and macrophage-mediated axon damage in the absence of a CNS-specific immune response: involvement of metalloproteinases. Brain. 124:2203-2214.

Nielsen TR, Rostgaard K, Askling J, Steffensen R, Oturai A, Jersild C, Koch- Henriksen N, Sorensen PS, Hjalgrim H. (2009) Effects of infectious mononucleosis and HLA-DRB1*15 in multiple sclerosis. Mult. Scler. 15 :431-436.

Nielsen N, Ødum N, Ursø B, Lanier LL, Spee P. (2012) Cytotoxicity of CD56(bright) NK cells towards autologous activated CD4+ T cells is mediated through NKG2D, LFA-1 and TRAIL and dampened via CD94/NKG2A. PLoS One. 7:e31959.

Nikcevich KM, Gordon KB, Tan L, Hurst SD, Kroepfl JF, Gardinier M, Barrett TA, Miller SD. (1997) Interferon-gamma activated primary murine astrocytes express B7

272 costimulatory molecules and prime naive antigen-specific T cells. J. Immunol. 158:614–621.

Nitsch R, Pohl, EE, Smorodchenko A, Infante-Duarte C, Aktas O, Zipp F. (2004) Direct Impact of T Cells on Neurons Revealed by Two-Photon Microscopy in Living Brain Tissue. J. Neurosci. 24: 2458-64.

Noronha A, Toscas A, Jensen MA. (1993) Interferon beta decreases T cell activation and interferon gamma production in multiple sclerosis. J. Neuroimmunol. 46:145-53.

O'Connor RA, Prendergast CT, Sabatos CA, Lau CW, Leech MD, Wraith DC, Anderton SM. (2008) Cutting edge: Th1 cells facilitate the entry of Th17 cells to the central nervous system during experimental autoimmune encephalomyelitis. J. Immunol. 181:3750-4.

Ogasawara K, Hida S, Azimi N, Tagaya Y, Sato T, Yokochi-Fukuda T, Waldmann TA, Taniguchi T, Taki S. (1998) Requirement for IRF-1 in the microenvironment supporting development of natural killer cells. Nature. 391:700–3.

Ohshima Y, Yang LP, Uchiyama T, Tanaka Y, Baum P, Sergerie M, Hermann P, Delespesse G. (1998) OX40 costimulation enhances interleukin-4 (IL-4) expression at priming and promotes the differentiation of naive human CD4(+) T cells into high IL- 4-producing effectors. Blood. 92:3338-45.

O'Keefe GM, Nguyen VT, Ping Tang LL, Benveniste EN. (2001) IFN-gamma regulation of class II transactivator promoter IV in macrophages and microglia: involvement of the suppressors of cytokine signaling-1 protein. J. Immunol. 166:2260-9.

O'Keeffe J, Gately CM, Counihan T, Hennessy M, Leahy T, Moran AP, Hogan EL. (2008) T-cells expressing natural killer (NK) receptors are altered in multiple sclerosis and responses to alpha-galactosylceramide are impaired. J. Neurol. Sci. 275:22-8.

Oksenberg JR, Barcellos LF, Cree BA, Baranzini SE, Bugawan TL, Khan O, Lincoln RR, Swerdlin A, Mignot E, Lin L, Goodin D, Erlich HA, Schmidt S, Thomson G, Reich DE, Pericak-Vance MA, Haines JL, Hauser SL. (2004) Mapping multiple

273 sclerosis susceptibility to the HLA-DR Locus in African Americans. Am. J. Hum. Genet. 74:160–7.

Oliva A, Kinter AL, Vaccarezza M, Rubbert A, Catanzaro A, Moir S, Monaco J, Ehler L, Mizell S, Jackson R, Li Y, Romano JW, Fauci AS. (1998) Natural killer cells from human immunodeficiency virus (HIV)-infected individuals are an important source of CC-chemokines and suppress HIV-1 entry and replication in vitro. J. Clin. Invest. 102:223-31.

Olsson MY, Kärre K, Sentman CL. (1995) Altered phenotype and function of natural killer cells expressing the major histocompatibility complex receptor Ly-49 in mice transgenic for its ligand. Proc. Natl. Acad. Sci. USA. 92:1649-53.

Oppenheim DE, Roberts SJ, Clarke SL, Filler R, Lewis JM, Tigelaar RE, Girardi M, Hayday AC. (2005) Sustained localized expression of ligand for the activating NKG2D receptor impairs natural cytotoxicity in vivo and reduces tumor immunosurveillance. Nat. Immunol. 6:928-37.

Orange J.S., Ramesh N, Remold-O'Donnell E, Sasahara Y, Koopman L, Byrne M, Bonilla FA, Rosen FS, Geha RS, Strominger JL. (2002) Wiskott-Aldrich syndrome protein is required for NK cell cytotoxicity and colocalizes with actin to NK cell- activating immunologic synapses. Proc. Natl. Acad. Sci. USA. 99:11351–56.

Orange JS, Harris KE, Andzelm MM, Valter MM, Geha RS, Strominger JL. (2003) The mature activating natural killer cell immunologic synapse is formed in distinct stages. Proc. Natl. Acad. Sci. USA. 100:14151-6.

Orange JS. (2008) Formation and function of the lytic NK-cell immunological synapse. Nat. Rev. Immunol. 8:713-25.

Ordóñez D, Sánchez AJ, Martínez-Rodríguez JE, Cisneros E, Ramil E, Romo N, Moraru M, Munteis E, López-Botet M, Roquer J, García-Merino A, Vilches C. (2009) Multiple sclerosis associates with LILRA3 deletion in Spanish patients. Genes Immun. 10:579-85.

Orr MT, Murphy WJ, Lanier LL. (2010) 'Unlicensed' natural killer cells dominate the response to cytomegalovirus infection. Nat. Immunol. 11:321-7.

274

O'Shea JJ, Weissman AM, Kennedy IC, Ortaldo JR. (1991) Engagement of the natural killer cell IgG Fc receptor results in tyrosine phosphorylation of the zeta chain. Proc. Natl. Acad. Sci. U.S.A. 88:350-4.

Osinde M, Mullershausen F, Dev KK. (2007) Phosphorylated FTY720 stimulates ERK phosphorylation in astrocytes via S1P receptors. Neuropharmacology. 52:1210- 8.

Owen BA, Pease LR. (1999) TAP association influences the conformation of nascent MHC class I molecules. J. Immunol. 162:4677.

Ozanne BW, Spence HJ, McGarry LC, Hennigan RF. (2007) Transcription factors control invasion: AP-1 the first among equals. Oncogene. 26:1-10.

Padovani-Claudio DA, Liu L, Ransohoff RM, Miller RH. (2006) Alterations in the oligodendrocyte lineage, myelin, and white matter in adult mice lacking the chemokine receptor CXCR2. Glia. 54:471–83.

Pagenstecher A, Stalder AK, Kincaid CL, Shapiro SD, Campbell IL. (1998) Differential expression of matrix metalloproteinase and tissue inhibitor of matrix metalloproteinase genes in the mouse central nervous system in normal and inflammatory states. Am. J. Pathol. 152:729-41.

Pahl HL. (1999) Activators and target genes of Rel/NF-kappaB transcription factors. Oncogene. 18:6853-66.

Pan Y, Lloyd C, Zhou H, Dolich S, Deeds J, Gonzalo JA, Vath J, Gosselin M, Ma J, Dussault B, Woolf E, Alperin G, Culpepper J, Gutierrez-Ramos JC, Gearing D. (1997) Neurotactin, a membrane-anchored chemokine upregulated in brain inflammation. Nature. 387:611-7.

Panitch HS, Folus JS, Johnson KP. (1990) Interferon beta augments suppressor cell function in multiple sclerosis. Ann. Neurol. 27:207–10.

Pappu R, Schwab SR, Cornelissen I, Pereira JP, Regard JB, Xu Y, Camerer E, Zheng YW, Huang Y, Cyster JG, Coughlin SR. (2007) Promotion of lymphocyte egress into blood and lymph by distinct sources of sphimgosine-1-phosphate. Science. 316:295-8.

275

Parham C, Chirica M, Timans J, Vaisberg E, Travis M, Cheung J, Pflanz S, Zhang R, Singh KP, Vega F, To W, Wagner J, O'Farrell AM, McClanahan T, Zurawski S, Hannum C, Gorman D, Rennick DM, Kastelein RA, de Waal Malefyt R, Moore KW. (2002) A receptor for the heterodimeric cytokine IL-23 is composed of IL-12Rbeta1 and a novel cytokine receptor subunit, IL-23R. J. Immunol. 168:5699-708.

Parham P. MHC class I molecules and kirs in human history, health and survival. (2005) Nat. Rev. Immunol. 5:201-14.

Park H, Li Z, Yang XO, Chang SH, Nurieva R, Wang YH, Wang Y, Hood L, Zhu Z, Tian Q, Dong C.(2005) A distinct lineage of CD4 T cells regulates tissue inflammation by producing interleukin 17. Nat. Immunol. 6:1133-41.

Patsopoulos NA; Bayer Pharma MS Genetics Working Group; Steering Committees of Studies Evaluating IFNβ-1b and a CCR1-Antagonist; ANZgene Consortium; GeneMSA; International Multiple Sclerosis Genetics Consortium, Esposito F, Reischl J, Lehr S, Bauer D, Heubach J, Sandbrink R, Pohl C, Edan G, Kappos L, Miller D, Montalbán J, Polman CH, Freedman MS, Hartung HP, Arnason BG, Comi G, Cook S, Filippi M, Goodin DS, Jeffery D, O'Connor P, Ebers GC, Langdon D, Reder AT, Traboulsee A, Zipp F, Schimrigk S, Hillert J, Bahlo M, Booth DR, Broadley S, Brown MA, Browning BL, Browning SR, Butzkueven H, Carroll WM, Chapman C, Foote SJ, Griffiths L, Kermode AG, Kilpatrick TJ, Lechner-Scott J, Marriott M, Mason D, Moscato P, Heard RN, Pender MP, Perreau VM, Perera D, Rubio JP, Scott RJ, Slee M, Stankovich J, Stewart GJ, Taylor BV, Tubridy N, Willoughby E, Wiley J, Matthews P, Boneschi FM, Compston A, Haines J, Hauser SL, McCauley J, Ivinson A, Oksenberg JR, Pericak-Vance M, Sawcer SJ, De Jager PL, Hafler DA, de Bakker PI. (2011) Genome-wide meta-analysis identifies novel multiple sclerosis susceptibility loci. Ann. Neurol. 70:897-912.

Pedersen L.B., Nashold F.E., Spach K.M., Hayes C.E. (2007) 1,25-dihydroxyvitamin D3 reverses experimental autoimmune encephalomyelitis by inhibiting chemokine synthesis and monocyte trafficking. J. Neurosci. Res. 85:2480–90.

Pende D, Parolini S, Pessino A, Sivori S, Augugliaro R, Morelli L, Marcenaro E, Accame L, Malaspina A, Biassoni R, Bottino C, Moretta L, Moretta A. (1999) Identification and molecular characterization of NKp30, a novel triggering receptor

276 involved in natural cytotoxicity mediated by human natural killer cells. J. Exp. Med. 190:1505-16.

Pende D, Rivera P, Marcenaro S, Chang CC, Biassoni R, Conte R, Kubin M, Cosman D, Ferrone S, Moretta L, Moretta A. (2002) Major histocompatibility complex class I- related chain A and UL16-binding protein expression on tumor cell lines of different histotypes: analysis of tumor susceptibility to NKG2D-dependent natural killer cell cytotoxicity. Cancer Res. 62:6178-86.

Penna G, Adorini L. (2000) 1 Alpha,25-dihydroxyvitamin D3 inhibits differentiation, maturation, activation, and survival of dendritic cells leading to impaired alloreactive T cell activation. J. Immunol. 164: 2405–11.

Perez SA, Sotiropoulou PA, Gkika DG, Mahaira LG, Niarchos DK, Gritzapis AD, Kavalakis YG, Antsaklis AI, Baxevanis CN, Papamichail M. (2003) A novel myeloid- like NK cell progenitor in human umbilical cord blood. Blood. 101:3444-50.

Perini P, Wadhwa M, Buttarello M, Meager A, Facchinetti A, Thorpe R, Biasi G, Gallo P. (2000) Effect of IFNbeta and anti-IFNbeta antibodies on NK cells in multiple sclerosis patients. J. Neuroimmunol. 105:91-5.

Perini P, Ranzato F, Calabrese M, Battistin L, Gallo P. (2006) Intrathecal IgM production at clinical onset correlates with a more severe disease course in multiple sclerosis. J. Neurol. Neurosurg. Psychiatry. 77:953-5.

Peritt D, Robertson S, Gri G, Showe L, Aste-Amezaga M, Trinchieri G. (1998) Differentiation of human NK cells into NK1 and NK2 subsets. J. Immunol. 161:5821- 4.

Perussia B, Kobayashi M, Rossi ME, Anegon I, Trinchieri G. (1987) Immune interferon enhances functional properties of human granulocytes: role of Fc receptors and effect of lymphotoxin, tumor necrosis factor, and granulocyte-macrophage colony-stimulating factor. J. Immunol. 138:765-74.

Pessino A, Sivori S, Bottino C, Malaspina A, Morelli L, Moretta L, Biassoni R, Moretta A. (1998) Molecular cloning of NKp46: a novel member of the

277 immunoglobulin superfamily involved in triggering of natural cytotoxicity. J. Exp. Med. 188:953-60.

Peveri P, Walz A, Dewald B, Baggiolini M. (1988) A novel neutrophil-activating factor produced by human mononuclear phagocytes. J. Exp. Med. 167:1547-1559.

Phillips JH, Hori T, Nagler A, Bhat N, Spits H, Lanier LL. (1992) Ontogeny of human natural killer (NK) cells: fetal NK cells mediate cytolytic function and express cytoplasmic CD3 epsilon,delta proteins. J. Exp. Med. 175:1055-66.

Piccioli D, Sbrana S, Melandri E, Valiante NM. (2002) Contact-dependent stimulation and inhibition of dendritic cells by natural killer cells. J. Exp. Med. 195:335-41.

Pierrot-Deseilligny C, Rivaud-Péchoux S, Clerson P, de Paz R, Souberbielle JC. (2012) Relationship between 25-OH-D serum level and relapse rate in multiple sclerosis patients before and after vitamin D supplementation. Ther. Adv. Neurol. Disord. 5:187-98.

Pinkoski MJ, Hobman M, Heibein JA, Tomaselli K, Li F, Seth P, Froelich CJ, Bleackley RC. (1998) Entry and trafficking of granzyme B in target cells during granzyme B-perforin-mediated apoptosis. Blood. 92:1044-54.

Pittas F, Ponsonby AL, van der Mei IA, Taylor BV, Blizzard L, Groom P, Ukoumunne OC, Dwyer T. (2009) Smoking is associated with progressive disease course and increased progression in clinical disability in a prospective cohort of people with multiple sclerosis. J. Neurol. 256:577-85.

Platanias LC. (2005) Mechanisms of type-I- and type-II-interferon–mediated signalling. Nat. Rev. Immunol. 5:375–386.

Pogge von Strandmann E, Simhadri VR, von Tresckow B, Sasse S, Reiners KS, Hansen HP, Rothe A, Böll B, Simhadri VL, Borchmann P, McKinnon PJ, Hallek M, Engert A. (2007) Human leukocyte antigen-B-associated transcript 3 is released from tumor cells and engages the NKp30 receptor on natural killer cells. Immunity. 27:965- 74

278

Poli V. (1998) The Role of C/EBP Isoforms in the Control of Inflammatory and Native Immunity Functions. J. Biol. Chem. 273:29279-82.

Polman CH, O'Connor PW, Havrdova E, Hutchinson M, Kappos L, Miller DH, Phillips JT, Lublin FD, Giovannoni G, Wajgt A, Toal M, Lynn F, Panzara MA, Sandrock AW; AFFIRM Investigators. (2006) A randomized, placebo-controlled trial of natalizumab for relapsing multiple sclerosis. N. Engl. J. Med. 354:899-910.

Ponath PD, Qin S, Post TW, Wang J, Wu L, Gerard NP, Newman W, Gerard C, Mackay CR. (1996a) Molecular cloning and characterization of a human eotaxin receptor expressed selectively on eosinophils. J. Exp. Med. 183:2437-48.

Ponath PD, Qin S, Ringler DJ, Clark-Lewis I, Wang J, Kassam N, Smith H, Shi X, Gonzalo JA, Newman W, Gutierrez-Ramos JC, Mackay CR. (1996b) Cloning of the human eosinophil chemoattractant, eotaxin. Expression, receptor binding, and functional properties suggest a mechanism for the selective recruitment of eosinophils. J. Clin. Invest. 97:604–12.

Popescu BF, Lucchinetti CF. (2012) Meningeal and cortical grey matter pathology in multiple sclerosis. BMC Neurol. 12:11.

Power CA, Meyer A, Nemeth K, Bacon KB, Hoogewerf AJ, Proudfoot AEI, Wells TNC. (1995) Molecular Cloning and Functional Expression of a Novel CC Chemokine Receptor cDNA from a Human Basophilic Cell Line. J. Biol. Chem. 270:19495-19500.

Presky DH, Yang H, Minetti LJ, Chua AO, Nabavi N, Wu CY, Gately MK, Gubler U. (1996) A functional interleukin 12 receptor complex is composed of two beta-type cytokine receptor subunits. Proc. Natl. Acad. Sci. U.S.A. 93:14002-7.

PRISMS Study Group. (1998) Randomised double-blind placebo-controlled study of interferon β-1a in relapsing/remitting multiple sclerosis. Lancet. 352:1498-1504.

PRISMS Study Group and the University of British Columbia MS/MRI Analysis Group. (2001) PRISMS-4: Long-term efficacy of interferon-beta-1a in relapsing MS. Neurology. 56:1628-36.

279

Punnonen J, Avers G, Cocks BG, McKenzie AN, Menon S, Bennet B, Culpepper J, Dang W, Zurawski G, de Vries JE. (1993) Interleukin 13 induces interleukin 4- independent IgG4 and IgE synthesis and CD23 expression by human B cells. Proc. Natl. Acad. Sci. USA. 90: 3730

Putheti P, Soderstrom M, Link H, Huang YM. (2003) Effect of glatiramer acetate (Copaxone) on CD4+CD25high T regulatory cells and their IL-10 production in multiple sclerosis. J. Neuroimmunol. 144:125-31.

Putzki N, Baranwal MK, Tettenborn B, Limmroth V, Kreuzfelder E. (2010) Effects of natalizumab on circulating B cells, T regulatory cells and natural killer cells. Eur. Neurol. 63:311-7.

Rabinovich BA, Li J, Shannon J, Hurren R, Chalupny J, Cosman D, Miller RG. (2003) Activated, but not resting, T cells can be recognized and killed by syngeneic NK cells. J. Immunol. 170:3572-6.

Racke MK, Lovett-Racke AE, Karandikar NJ. (2010) The mechanism of action of glatiramer acetate treatment in multiple sclerosis. Neurology. 74:S25-30.

Rajagopalan S, Long EO. (1999) A human histocompatibility leukocyte antigen (HLA)-G-specific receptor expressed on all natural killer cells. J. Exp. Med. 189:1093-100.

Rak GD, Mace EM, Banerjee PP, Svitkina T, Orange JS. (2011) Natural killer cell lytic granule secretion occurs through a pervasive actin network at the immune synapse. PLoS Biol. 9:e1001151.

Ramagopalan SV, Maugeri NJ, Handunnetthi L, Lincoln MR, Orton SM, Dyment DA, Deluca GC, Herrera BM, Chao MJ, Sadovnick AD, Ebers GC, Knight JC. (2009) Expression of the multiple sclerosis-associated MHC class II Allele HLA- DRB1*1501 is regulated by vitamin D. PLoS Genet. 5:e1000369.

Ramgolam VS, Sha Y, Jin J, Zhang X, Markovic-Plese S. (2009) IFN-beta inhibits human Th17 cell differentiation. J. Immunol. 183:5418-27.

280

Ramos CD, Canetti C, Souto JT, Silva JS, Hogaboam CM, Ferreira SH, Cunha FQ. (2005) MIP-1alpha[CCL3] acting on the CCR1 receptor mediates neutrophil migration in immune inflammation via sequential release of TNF-alpha and LTB4. J. Leukoc. Biol. 78:167-77.

Ramos-Cejudo J, Oreja-Guevara C, Stark Aroeira L, Rodriguez de Antonio L, Chamorro B, Diez-Tejedor E. (2011) Treatment with natalizumab in relapsing- remitting multiple sclerosis patients induces changes in inflammatory mechanism. J. Clin. Immunol. 31:623-31.

Raport CJ, Gosling J, Schweickart VL, Gray PW, Charo IF. (1996) Molecular cloning and functional characterization of a novel human CC chemokine receptor (CCR5) for RANTES, MIP-1β, and MIP-1α. J. Biol. Chem. 271: 17161

Raulet DH, Vance RE. (2006) Self-tolerance of natural killer cells. Nat. Rev. Immunol. 6:520-531.

Reboldi A, Coisne C, Baumjohann D, Benvenuto F, Bottinelli D, Lira S, Uccelli A, Lanzavecchia A, Engelhardt B, Sallusto F. (2009) C-C chemokine receptor 6- regulated entry of TH-17 cells into the CNS through the choroid plexus is required for the initiation of EAE. Nat. Immunol. 10:514-23.

Refaeli Y, Van Parijs L, Alexander SI, Abbas AK. (2002) Interferon gamma is required for activation-induced death of T lymphocytes. J. Exp. Med. 196:999-1005.

Renno T, Zeine R, Girard JM, Gillani S, Dodelet V, Owens T. (1994) Selective enrichment of Th1 CD45RBlow CD4+ T cells in autoimmune infiltrates in experimental allergic encephalomyelitis. Int. Immunol. 6:347-54.

Renno T, Krakowski M, Piccirillo C, Lin JY, Owens T. (1995) TNF-alpha expression by resident microglia and infiltrating leukocytes in the central nervous system of mice with experimental allergic encephalomyelitis. Regulation by Th1 cytokines. J. Immunol. 154:944

Rentzos M, Cambouri C, Rombos A, Nikolaou C, Anagnostouli M, Tsoutsou A, Dimitrakopoulos A, Triantafyllou N, Vassilopoulos D. (2006) J. Neurol. Sci. 241:25- 9.

281

Reymond N, Imbert AM, Devilard E, Fabre S, Chabannon C, Xerri L, Farnarier C, Cantoni C, Bottino C, Moretta A, Dubreuil P, Lopez M. (2004) DNAM-1 and PVR regulate monocyte migration through endothelial junctions. J. Exp. Med. 199:1331- 41.

Ricci JE, Gottlieb RA, Green DR. (2003) Caspase-mediated loss of mitochondrial function and generation of reactive oxygen species during apoptosis. J. Cell Biol. 160:65-75.

Ricci JE, Muñoz-Pinedo C, Fitzgerald P, Bailly-Maitre B, Perkins GA, Yadava N, Scheffler IE, Ellisman MH, Green DR. (2004) Disruption of mitochondrial function during apoptosis is mediated by caspase cleavage of the p75 subunit of complex I of the electron transport chain. Cell. 117:773–786.

Riise T, Nortvedt MW, Ascherio A. (2003) Smoking is a risk factor for multiple sclerosis. Neurology. 61: 1122–4.

Riteau B, Barber DF, Long EO. (2003) Vav1 Phosphorylation Is Induced by beta-2 Integrin Engagement on Natural Killer Cells Upstream of Actin Cytoskeleton and Lipid Raft Reorganization. J. Exp. Med. 198:469-74.

Roda JM, Parihar R, Magro C, Nuovo GJ, Tridandapani S, Carson WE. (2006) Natural Killer Cells Produce T Cell–Recruiting Chemokines in Response to Antibody-Coated Tumor Cells. Cancer Res. 66:517

Romagnani C, Juelke K, Falco M, Morandi B, D'Agostino A, Costa R, Ratto G, Forte G, Carrega P, Lui G, Conte R, Strowig T, Moretta A, Münz C, Thiel A, Moretta L, Ferlazzo G. (2007) CD56brightCD16- killer Ig-like receptor- NK cells display longer telomeres and acquire features of CD56dim NK cells upon activation. J. Immunol. 178:4947-55.

Romero AI, Thorén FB, Brune M, Hellstrand K. (2006) NKp46 and NKG2D receptor expression in NK cells with CD56dim and CD56bright phenotype: regulation by histamine and reactive oxygen species. Br. J. Haematol. 132:91-8.

282

Roncarolo MG, Bigler M, Haanen JB, Yssel H, Bacchetta R, de Vries JE, Spits H. (1991) Natural killer cell clones can efficiently process and present protein antigens. J. Immunol. 147:781-7.

Rosenberg GA, Kornfeld M, Estrada E, Kelley RO, Liotta LA, Stetler-Stevenson WG. (1992) TIMP-2 reduces proteolytic opening of blood-brain barrier by type IV collagenase. Brain Res. 576:203-7.

Rosmaraki EE, Douagi I, Roth C, Colucci F, Cumano A, Di Santo JP. (2001) Identification of committed NK cell progenitors in adult murine bone marrow. Eur. J. Immunol. 31:1900-9.

Rot A, Krieger M, Brunner T, Bischoff SC, Schall TJ, Dahinden A. (1992) RANTES and macrophage inflammatory protein 1α induce the migration and activation of normal human eosinophil granulocytes. J. Exp. Med. 176: 1489

Roth SJ, Diacovo TG, Brenner MB, Rosat JP, Buccola J, Morita CT, Springer TA. (1998) Transendothelial chemotaxis of human alpha/beta and gamma/delta T lymphocytes to chemokines. Eur. J. Immunol. 28:104-13.

Rousset F, Malefijt RW, Slierendregt B, Aubry JP, Bonnefoy JY, Defrance T, Banchereau J, de Vries JE. (1988) Regulation of Fc receptor for IgE (CD23) and class II MHC antigen expression on Burkitt’s lymphoma cell lines by human IL-4 and IFN- gamma. J. Immunol. 140:2625-2632.

Rousset F, Garcia E, Defrance T, Péronne C, Vezzio N, Hsu DH, Kastelein R, Moore KW, Banchereau J. (1992) Interleukin 10 is a potent growth and differentiation factor for activated human B lymphocytes. Proc. Natl. Acad. USA. 89:1890-3.

Rouyez MC, Lestingi M, Charon M, Fichelson S, Buzyn A, Dusanter-Fourt I. (2005) IFN regulatory factor-2 cooperates with STAT1 to regulate transporter associated with antigen processing-1 promoter activity. J. Immunol. 174:3948-58.

Roy S, Barnes PF, Garg A, Wu S, Cosman D, Vankayalapati R. (2008) NK Cells Lyse T Regulatory Cells That Expand in Response to an Intracellular Pathogen. J. Immunol. 180:1729-36.

283

Rudel T, Zenke FT, Chuang TH, Bokoch GM. (1998) p21-activated kinase (PAK) is required for Fas-induced JNK activation in Jurkat cells. J. Immunol. 160:7-11.

Rudick RA, Ransohoff RM, Lee JC, Peppler R, Yu M, Mathisen PM, Tuohy VK. (1998) In vivo effects of interferon beta-1a on immunosuppressive cytokines in multiple sclerosis. Neurology. 50:1294-1300.

Ryu M, Chen Y, Qi J, Liu J, Fan Z, Nam G, Shi Y, Cheng H, Gao GF. (2011) LILRA3 binds both classical and non-classical HLA class I molecules but with reduced affinities compared to LILRB1/LILRB2: structural evidence. PLoS One. 6:e19245.

Sahara S, Aoto M, Eguchi Y, Imamoto M, Yoneda Y, Tjuimoto Y. (1999) Acinus is a caspase-3-activated protein required for apoptotic chromatin condensation. Nature. 401:168–73.

Saikali P, Antel JP, Newcombe J, Chen Z, Freedman M, Blain M, Cayrol R, Prat A, Hall JA, Arbour N. (2007) NKG2D-mediated cytotoxicity toward oligodendrocytes suggests a mechanism for tissue injury in multiple sclerosis. J. Neurosci. 27:1220-8.

Saikali P, Antel JP, Pittet CL, Newcombe J, Arbour N. (2010) Contribution of astrocyte-derived IL-15 to CD8 T cell effector functions in multiple sclerosis. J. Immunol. 185:5693-703.

Sakon S, Xue X, Takekawa M, Sasazuki T, Okazaki T, Kojima Y, Piao JH, Yagita H, Okumura K, Doi T, Nakano H. (2003) NF-kappaB inhibits TNF-induced accumulation of ROS that mediate prolonged MAPK activation and necrotic cell death. EMBO J. 22:3898-909.

Sanborn KB, Rak GD, Maru SY, Demers K, Difeo A, Martignetti JA, Betts MR, Favier R, Banerjee PP, Orange JS. (2009) Myosin IIA Associates with NK Cell Lytic Granules to Enable Their Interaction with F-Actin and Function at the Immunological Synapse. J. Immunol. 182:6969-84.

Sánchez MJ, Muench MO, Roncarolo MG, Lanier LL, Phillips JH. (1994) Identification of a common T/natural killer cell progenitor in human fetal thymus. J. Exp. Med. 180:569-76.

284

Sanders SK, Giblin PA, Kavathas P. (1991) Cell-cell adhesion mediated by CD8 and human histocompatibility leukocyte antigen G, a nonclassical major histocompatibility complex class 1 molecule on cytotrophoblasts. J. Exp. Med. 174:737-40.

Sanjanwala B, Draghi M, Norman PJ, Guethlein LA, Parham P. (2008) Polymorphic sites away from the Bw4 epitope that affect interaction of Bw4+ HLA-B with KIR3DL1. J. Immunol. 181:6293-300.

Santoli D, Hall W, Kastrukoff L, Lisak RP, Perussia B, Trinchieri G, Koprowski H. (1981) Cytotoxic activity and interferon production by lymphocytes from patients with multiple sclerosis. J. Immunol. 126:1274-8.

Saraste M, Irjala H, Airas L. (2007) Expansion of CD56Bright natural killer cells in the peripheral blood of multiple sclerosis patients treated with interferon-beta. Neurol. Sci. 28:121–126.

Sasaki A, Yasukawa H, Suzuki A, Kamizono S, Syoda T, Kinjyo I, Sasaki M, Johnston JA, Yoshimura A. (1999) Cytokine-inducible SH2 protein-3 (CIS3/SOCS3) inhibits Janus tyrosine kinase by binding through the N-terminal kinase inhibitory region as well as SH2 domain. Genes Cells. 4:339-51.

Satoh-Takayama N, Vosshenrich CA, Lesjean-Pottier S, Sawa S, Lochner M, Rattis F, Mention JJ, Thiam K, Cerf-Bensussan N, Mandelboim O, Eberl G, Di Santo JP. (2008) Microbial flora drives interleukin 22 production in intestinal NKp46+ cells that provide innate mucosal immune defense. Immunity. 29:958-70.

Schall TJ, Bacon K, Toy KJ, Goeddel DV. (1990) Selective attraction of monocytes and T lymphocytes of the memory phenotype by cytokine RANTES. Nature. 347:669-71.

Schall TJ, Bacon K, Camp RD, Kaspari JW, Goeddel DV. (1993) Human macrophage inflammatory protein alpha (MIP-1 alpha) and MIP-1 beta chemokines attract distinct populations of lymphocytes. J. Exp. Med. 177:1821-1826.

285

Schepis D, Gunnarsson I, Eloranta ML, Lampa J, Jacobson SH, Kärre K, Berg L. (2009) Increased proportion of CD56bright natural killer cells in active and inactive systemic lupus erythematosus. Immunology. 126:140-6.

Schindler C, Plumlee C. (2008) Inteferons pen the JAK-STAT pathway. Semin. Cell. Dev. Biol. 19: 311–318.

Schneider R, Mohebiany AN, Ifergan I, Beauseigle D, Duquette P, Prat A, Arbour N. (2011) B cell-derived IL-15 enhances CD8 T cell cytotoxicity and is increased in multiple sclerosis patients. J. Immunol. 187:4119-28.

Schonrock LM, Gawlowski G, Bruck W. (2000) Interleukin-6 expression in human multiple sclerosis lesions. Neurosci. Lett. 294:45–48.

Schroder K, Hertzog PJ, Ravasi T, Hume DA. (2004) Interferon-gamma: an overview of signals, mechanisms and functions. J. Leukoc. Biol. 75:163-89.

Schwartz JP, Nishiyama N. (1994) Neurotrophic factor gene expression in astrocytes during development and following injury. Brain Res. Bull. 35:403–7.

Schwid SR, Panitch HS. (2007) Full results of the Evidence of Interferon Dose- Response-European North American Comparative Efficacy (EVIDENCE) study: a multicenter, randomized, assessor-blinded comparison of low-dose weekly versus high-dose, high-frequency interferon beta-1a for relapsing multiple sclerosis. Clin. Ther. 29:2031-48.

Scolding NJ, Compston DA. (1991) Oligodendrocyte-macrophage interactions in vitro triggered by specific antibodies. Immunolgy. 72:127-132.

Scott GB, Meade JL, Cook GP. (2008) Profiling killers; unravelling the pathways of human natural killer cell function. Briefings in Functional Genomics and Proteomics. 7:8-16.

Selmaj KW, Raine CS. (1988) Tumor necrosis factor mediates myelin and oligodendrocyte damage in vitro. Annals of Neurology. 23: 339–346.

286

Selmaj K, Raine CS, Farooq M, Norton WT, Brosnan CF. (1991) Cytokine cytotoxicity against oligodendrocytes: apoptosis induced by lymphotoxin. J. Immunol. 147:1522–9.

Semino C, Angelini G, Poggi A, Rubartelli A. (2005) NK/iDC interaction results in IL-18 secretion by DCs at the synaptic cleft followed by NK cell activation and release of the DC maturation factor HMGB1. Blood. 106:609-16.

Serafini B, Rosicarelli B, Franciotta D, Magliozzi R, Reynolds R, Cinque P, Andreoni L, Trivedi P, Salvetti M, Faggioni A, Aloisi F. (2007) Dysregulated Epstein-Barr virus infection in the multiple sclerosis brain. J. Exp. Med. 204:2899-912.

Serafini B, Severa M, Columba-Cabezas S, Rosicarelli B, Veroni C, Chiappetta G, Magliozzi R, Reynolds R, Coccia EM, Aloisi F. (2010) Epstein-Barr virus latent infection and BAFF expression in B cells in the multiple sclerosis brain: implications for viral persistence and intrathecal B-cell activation. J. Neuropatho. Exp. Neurol. 69:677-93.

Sharif-Askari E, Alam A, Rhéaume E, Beresford PJ, Scotto C, Sharma K, Lee D, DeWolf WE, Nuttall ME, Lieberman J, Sékaly RP. (2001) Direct cleavage of the human DNA fragmentation factor-45 by granzyme B induces caspase-activated DNase release and DNA fragmentation. EMBO J. 20:3101-13.

Sharma D, Bastard K, Guethlein LA, Norman PJ, Yawata N, Yawata M, Pando M, Thananchai H, Dong T, Rowland-Jones S, Brodsky FM, Parham P. (2009) Dimorphic motifs in D0 and D1+D2 domains of killer cell Ig-like receptor 3DL1 combine to form receptors with high, moderate, and no avidity for the complex of a peptide derived from HIV and HLA-A*2402. J. Immunol. 183:4569-82.

Shaulian E, Karin M. (2001) AP-1 in cell proliferation and survival. Oncogene. 20:2390-400.

Shi L, Mai S, Israels S, Browne K, Trapani JA, Greenberg AH. (1997) Granzyme B (GraB) autonomously crosses the cell membrane and perforin initiates apoptosis and GraB nuclear localization. J. Exp. Med. 185:855-66.

287

Shi FD, Takeda K, Akira S, Sarvetnick N, Ljunggren HG. (2000) IL-18 Directs Autoreactive T Cells and Promotes Autodestruction in the Central Nervous System Via Induction of IFN-gamma by NK Cells. J. Immunol. 165:3099-104.

Shibuya A, Nagayoshi K, Nakamura K, Nakauchi H. (1995) Lymphokine requirement for the generation of natural killer cells from CD34+ hematopoietic progenitor cells. Blood. 85:3538-46.

Shibuya A, Campbell D, Hannum C, Yssel H, Franz-Bacon K, McClanahan T, Kitamura T, Nicholl J, Sutherland GR, Lanier LL, Phillips JH. (1996) DNAM-1, a novel adhesion molecule involved in the cytolytic function of T lymphocytes. Immunity. 4:573–81.

Siegel RM, Frederiksen JK, Zacharias DA, Chan FK, Johnson M, Lynch D, Tsien RY, Lenardo MJ. (2000) Fas preassociation required for apoptosis signaling and dominant inhibition by pathogenic mutations. Science. 288:2354-7.

Silva A, Andrews DM, Brooks AG, Smyth MJ, Hayakawa Y. (2008) Application of CD27 as a marker for distinguishing human NK cell subsets. Int. Immunol. 20:625-30.

Simpson JE, Newcombe J, Cuzner ML, Woodroofe MN. (2000) Expression of the interferon-gamma-inducible chemokines IP-10 and Mig and their receptor, CXCR3, in multiple sclerosis lesions. Neuropathol. Appl. Neurobiol. 26:133-42.

Simpson S. Jr., Taylor B., Blizzard L., Ponsonby A.L., Pittas F., Tremlett H., Dwyer T., Gies P., van der Mei I. (2010) Higher 25-hydroxyvitamin D is associated with lower relapse risk in multiple sclerosis. Ann. Neurol. 68:193-203.

Sinha S, Kaler LJ, Proctor TM, Teuscher C, Vandenbark AA, Offner H. (2008) IL-13- Mediated Gender Difference in Susceptibility to Autoimmune Encephalomyelitis. J. Immunol. 180:2679-85.

Siveke JT, Hamann A. (1998) Cutting Edge: T Helper 1 and T Helper 2 Cells Respond Differentially to Chemokines. J. Immunol. 160:550-4.

Smith ME, Eng LF. (1987) Glial fibrillary acidic protein in chronic relapsing experimental allergic encephalomyelitis in SJL/J mice. J. Neurosci. Res. 18:203–208.

288

Smolders J, Thewissen M, Peelen E, Menheere P, Tervaert JW, Damoiseaux J, Hupperts R. (2009) Vitamin D Status Is Positively Correlated with Regulatory T Cell Function in Patients with Multiple Sclerosis. PLoS One. 4: e6635.

Snyder JT, Shen J, Amzi H, Hou J, Fowler DH, Ragheb JA. (2007) Direct inhibition of CD40L expression can contribute to the clinical efficacy of daclizumab independently of its effects on cell division and Th1/Th2 cytokine production. Blood. 109:5399–406.

Sobel RA, Mitchell ME, Fondren G. (1990) Intercellular adhesion molecule-1 (ICAM-1) in cellular immune reactions in the human central nervous system. Am. J. Pathol. 136:1309–16.

Soilu-Hänninen M, Laaksonen M, Laitinen I, Erälinna JP, Lilius EM, Mononen I. (2008) A longitudinal study of serum 25-hydroxyvitamin D and intact parathyroid hormone levels indicate the importance of vitamin D and calcium homeostasis regulation in multiple sclerosis. J. Neurol. Neurosurg. Psychiatry. 79:152-7.

Soliven B, Szuchet S, Nelson DJ. (1991) Tumor necrosis factor inhibits K+ current expression in cultured oligodendrocytes. J. Membr. Biol. 124:127–137.

Song H, Kim J, Cosman D, Choi I. (2006) Soluble ULBP suppresses natural killer cell activity via down-regulating NKG2D expression. Cell. Immunol. 239:22-30.

Sørensen TL, Tani M, Jensen J, Pierce V, Lucchinetti C, Folcik VA, Qin S, Rottman J, Sellebjerg F, Strieter RM, Frederiksen JL, Ransohoff RM. (1999) Expression of specific chemokines and chemokine receptors in the central nervous system of multiple sclerosis patients. J. Clin. Invest. 103:807-15.

Sospedra M, Martin R. (2005) Immunology of multiple sclerosis. Annu. Rev. Immunol. 23:683-747.

Sozzani S, Luini W, Borsatti A, Polentarutti N, Zhou D, Piemonti L, D'Amico G, Power CA, Wells TN, Gobbi M, Allavena P, Mantovani A. (1993) Receptor expression and responsiveness of human dendritic cells to a defined set of CC and CXC chemokines. J. Immunol. 159:1993-2000.

289

Sozzani P, Cambon C, Vita N, Seguelas MH, Caput D, Ferrara P, Pipy B. (1995) Interleukin-13 inhibits protein kinase C-triggered respiratory burst in human monocytes: role of calcium and cyclic AMP. J. Biol. Chem. 270: 5084

Spach KM, Pedersen LB, Nashold FE, Kayo T, Yandell BS, Prolla TA, Hayes CE. (2004) Gene expression analysis suggests that 1,25-dihydroxyvitamin D3 reverses experimental autoimmune encephalomyelitis by stimulating inflammatory cell apoptosis. Physiol. Genomics. 18:141-51.

Spach K.M., Nashold F.E., Dittel B.N., Hayes C.E. (2006) IL-10 Signaling Is Essential for 1,25-Dihydroxyvitamin D3-Mediated Inhibition of Experimental Autoimmune Encephalomyelitis. J. Immunol. 177: 6030–6037.

Spaggiari GM, Carosio R, Pende D, Marcenaro S, Rivera P, Zocchi MR, Moretta L, Poggi A. (2001) NK cell-mediated lysis of autologous antigen-presenting cells is triggered by the engagement of the phosphatidylinositol 3-kinase upon ligation of the natural cytotoxicity receptors NKp30 and NKp46. Eur. J. Immunol. 31:1656-65.

Spaggiari GM, Contini P, Carosio R, Arvigo M, Ghio M, Oddone D, Dondero A, Zocchi MR, Puppo F, Indiveri F, Poggi A. (2002) Soluble HLA class I molecules induce natural killer cell apoptosis through the engagement of CD8: evidence for a negative regulation exerted by members of the inhibitory receptor superfamily. Blood. 99:1706-14.

Spies T, DeMars R. (1991) Restored expression of major histocompatibility class I molecules by gene transfer of a putative peptide transporter. Nature. 351:323.

Sprick MR, Weigand MA, Rieser E, Rauch CT, Juo P, Blenis J, Krammer PH, Walczak H. (2000) FADD/MORT1 and caspase-8 are recruited to TRAIL receptors 1 and 2 and are essential for apoptosis mediated by TRAIL receptor 2. Immunity. 12:599-609.

Sprick MR, Rieser E, Stahl H, Grosse-Wilde A, Weigand MA, Walczak H. (2002) Caspase-10 is recruited to and activated at the native TRAIL and CD95 death- inducing signalling complexes in a FADD-dependent manner but can not functionally substitute caspase-8. EMBO J. 21:4520-30.

290

Spurkland A, Rønningen KS, Vandvik B, Thorsby E, Vartdal F. (1991) HLA-DQA1 and HLA-DQB1 genes may jointly determine susceptibility to develop multiple sclerosis. Hum. Immunol. 30:69-75.

Srour EF, Leemhuis T, Jenski L, Redmond R, Jansen J. (1990) Cytolytic activity of human natural killer cell subpopulations isolated by four-color immunofluorescence flow cytometric cell sorting. Cytometry. 11:442-6.

Stalder AK, Pagenstecher A, Yu NC, Kincaid C, Chiang CS, Hobbs MV, Bloom FE, Campbell IL. (1997) Lipopolysaccharide-induced IL-12 expression in the central nervous system and cultured astrocytes and microglia. J. Immunol. 159:1344-51.

Starr R, Willson TA, Viney EM, Murray LJ, Rayner JR, Jenkins BJ, Gonda TJ, Alexander WS, Metcalf D, Nicola NA, Hilton DJ. (1997) A family of cytokine- inducible inhibitors of signalling. Nature. 387:917-21.

Stebbins CC, Watzl C, Billadeau DD, Leibson PJ, Burshtyn DN, Long EO. (2003) Vav1 dephosphorylation by the tyrosine phosphatase SHP-1 as a mechanism for inhibition of cellular cytotoxicity. Mol. Cell. Biol. 23:6291-9.

Stegmann KA, Björkström NK, Ciesek S, Lunemann S, Jaroszewicz J, Wiegand J, Malinski P, Dustin LB, Rice CM, Manns MP, Pietschmann T, Cornberg M, Ljunggren HG, Wedemeyer H. (2012) Interferon α-stimulated natural killer cells from patients with acute hepatitis C virus (HCV) infection recognize HCV-infected and uninfected hepatoma cells via DNAX accessory molecule-1. J. Infect. Dis. 205:1351- 62

Steinle A, Li P, Morris DL, Groh V, Lanier LL, Strong RK, Spies T. (2001) Interactions of human NKG2D with its ligands MICA, MICB, and homologs of the mouse RAE-1 protein family. Immunogenetics. 53:279-87.

Steinman L. (2007) Antigen-Specific Therapy of Multiple Sclerosis: The Long- Sought Magic Bullet. Neurotherapeutics. 4:661-5.

Stenner MP, Waschbisch A, Buck D, Doerck S, Einsele H, Toyka KV, Wiendl H. (2008) Effects of natalizumab treatment on Foxp3+ T regulatory cells. PLoS One. 3:e3319.

291

Stennicke HR, Jürgensmeier JM, Shin H, Deveraux Q, Wolf BB, Yang X, Zhou Q, Ellerby HM, Ellerby LM, Bredesen D, Green DR, Reed JC, Froelich CJ, Salvesen GS. (1998) Pro-caspase-3 is a major physiologic target of caspase-8. J. Biol. Chem. 273:27084-90.

Stern M, Ruggeri L, Capanni M, Mancusi A, Velardi A. (2008) Human leukocyte antigens A23, A24, and A32 but not A25 are ligands for KIR3DL1. Blood. 112:708– 10.

Stinchcombe JC, Salio M, Cerundolo V, Pende D, Arico M, Griffiths GM. (2011) Centriole polarisation to the immunological synapse directs secretion from cytolytic cells of both the innate and adaptive immune systems. BMC Biol. 9:45.

Stoll G, Jander S, Schroeter M. (2002) Detrimental and beneficial effects of injury- induced inflammation and cytokine expression in the nervous system. Adv. Exp. Med. Biol. 513:87-113.

Storch MK, Piddlesden S, Haltia M, Iivanainen M, Morgan P, Lassmann H. (1998) Multiple sclerosis: in situ evidence for antibody- and complement-mediated demyelination. Ann. Neurol. 43:465-71.

Stromnes IM, Cerretti LM, Liggitt D, Harris RA, Goverman JM. (2008) Differential regulation of central nervous system autoimmunity by T(H)1 and T(H)17 cells. Nat. Med. 14:337-42.

Stüve O, Dooley NP, Uhm JH, Antel JP, Francis GS, Williams G, Yong VW. (1996) Interferon beta-1b decreases the migration of T lymphocytes in vitro: effects on matrix metalloproteinase-9. Ann. Neurol. 40:853-63.

Suidan HS, Bouvier J, Schaerer E, Stones SR, Monard D, Tschopp J. (1994) Granzyme A released upon stimulation of cytotoxic T lymphocytes activates the thrombin receptor on neuronal cells and astrocytes. Proc. Natl. Acad. Sci. USA. 91:8112-8116.

Sundström P, Nyström M, Ruuth K, Lundgren E. (2009) Antibodies to specific EBNA-1 domains and HLA DRB1*1501 interact as risk factors for multiple sclerosis. J. Neuroimmunol. 215:102-7.

292

Sutton VR, Davis JE, Cancilla M, Johnstone RW, Ruefli AA, Sedelies K, Browne KA, Trapani JA. (2000) Initiation of apoptosis by granzyme B requires direct cleavage of bid, but not direct granzyme B-mediated caspase activation. J. Exp. Med. 192:1403-14.

Suzuki H, Duncan GS, Takimoto H, Mak TW. (1997) Abnormal development of intestinal intraepithelial lymphocytes and peripheral natural killer cells in mice lacking the IL-2 receptor beta chain. J. Exp. Med. 185:499-505.

Suzuki Y, Hamamoto Y, Ogasawara Y, Ishikawa K, Yoshikawa Y, Sasazuki T, Muto M. (2004) Genetic polymorphisms of killer cell immunoglobulin-like receptors are associated with susceptibility to psoriasis vulgaris. J. Invest. Dermatol. 122:1133-6.

Szabo SJ, Sullivan BM, Stemmann C, Satoskar AR, Sleckman BP, Glimcher LH. (2002) Distinct effects of T-bet in TH1 lineage commitment and IFN-gamma production in CD4 and CD8 T cells. Science. 295:338-42.

Tada Y, Ho A, Matsuyama T, Mak TW. (1997) Reduced incidence and severity of antigen-induced autoimmune diseases in mice lacking interferon regulatory factor-1. J. Exp. Med. 185:231-8.

Taga K, Mostowski H, Tosato G. Human interleukin-10 can directly inhibit T-cell growth. Blood. 81: 2964-71.

Tahara-Hanaoka S, Shibuya K, Onoda Y, Zhang H, Yamazaki S, Miyamoto A, Honda S, Lanier LL, Shibuya A. (2004) Functional characterization of DNAM-1 (CD226) interaction with its ligands PVR (CD155) and nectin-2 (PRR-2/CD112). Int. Immunol. 16:533-8.

Takahashi K, Miyake S, Kondo T, Terao K, Hatakenaka M, Hashimoto S, Yamamura T. (2001) Natural killer type 2 bias in remission of multiple sclerosis. J. Clin. Invest. 107:R23-9.

Takahashi K, Aranami T, Endoh M, Miyake S, Yamamura T. (2004) The regulatory role of natural killer cells in multiple sclerosis. Brain. 127:1917-27.

293

Takahashi E, Kuranaga N, Satoh K, Habu Y, Shinomiya N, Asano T, Seki S, Hayakawa M. (2007) Induction of CD16+ CD56bright NK cells with antitumour cytotoxicity not only from CD16- CD56bright NK Cells but also from CD16- CD56dim NK cells. Scand. J. Immunol. 65:126-38.

Takeda K, Kaisho T, Yoshida N, Takeda J, Kishimoto T, Akira S. (1998) Stat3 activation is responsible for IL-6-dependent T cell proliferation through preventing apoptosis: generation and characterization of T cell-specific Stat3-deficient mice. J. Immunol. 161:4652-60.

Takeda K, Oshima H, Hayakawa Y, Akiba H, Atsuta M, Kobata T, Kobayashi K, Ito M, Yagita H, Okamura K. (2000) CD27-mediated activation of murine NK cells. J. Immunol. 164: 1741-1745.

Taki S, Sato T, Ogasawara K, Fukuda T, Sato M, Hida S, Suzuki G, Mitsuyama M, Shin EH, Kojima S, Taniguchi T, Asano Y. (1997) Multistage regulation of Th1-type immune responses by the transcription factor IRF-1. Immunity. 6:673–9.

Talanian RV, Yang X, Turbov J, Seth P, Ghayur T, Casiano CA, Orth K, Froelich CJ. (1997) Granule-mediated killing: pathways for granzyme B-initiated apoptosis. J. Exp. Med. 186:1323-31.

Tan LJ, Gordon KB, Mueller JP, Matis LA, Miller SD. (1998) Presentation of proteolipid protein epitopes and B7-1-dependent activation of encephalitogenic T cells by IFN-gamma-activated SJL/J astrocytes. J. Immunol. 160:4271–9.

Tang KF, Ren H, Cao J, Zeng GL, Xie J, Chen M, Wang L, He CX. (2008) Decreased Dicer expression elicits DNA damage and up-regulation of MICA and MICB. J. Cell Biol. 182:233-9.

Taniguchi T, Ogasawara K, Takaoka A, Tanaka N. (2001) IRF family of transcription factors as regulators of host defense. Annu. Rev. Immunol. 19:623-55.

Taub DD, Conlon K, Lloyd AR, Oppenheim JJ, Kelvin DJ. (1993) Preferential migration of activated CD4+ and CD8+ T cells in response to MIP-1α and MIP-1β. Science. 260:355

294

Taub DD, Sayers TJ, Carter CR, Ortaldo JR. (1995) Alpha and beta chemokines induce NK cell migration and enhance NK-mediated cytolysis. J. Immunol. 155:3877- 88.

Taub DD, Anver M, Oppenheim JJ, Longo DL, Murphy WJ. (1996) T lymphocyte recruitment by interleukin-8 (IL-8). IL-8-induced degranulation of neutrophils releases potent chemoattractants for human T lymphocytes both in vitro and in vivo. J. Clin. Invest. 97:1931-41.

Tennakoon DK, Mehta RS, Ortega SB, Bhoj V, Racke MK, Karandikar NJ. (2006) Therapeutic induction of regulatory, cytotoxic CD8+ T cells in multiple sclerosis. J. Immunol. 176:7119-29. te Velde AA, de Waal Malefijt R, Huijbens RJ, de Vries JE, Figdor CG. (1992) IL-10 stimulates monocyte Fc gamma R surface expression and cytotoxic activity. Distinct regulation of antibody-dependent cellular cytotoxicity by IFN-gamma, IL-4, and IL- 10. J. Immunol. 149:4048-52.

Tewari M, Quan LT, O'Rourke K, Desnoyers S, Zeng Z, Beidler DR, Poirier GG, Salvesen GS, Dixit VM. (1995) Yama/CPP32-beta, a mammalian homolog of CED-3, is a CrmA-inhibitable protease that cleaves the death substrate poly(ADP-ribose) polymerase. Cell. 81:801–9.

Thacker EL, Mirzaei F, Ascherio A. (2006) Infectious mononucleosis and risk for multiple sclerosis: a meta-analysis. Ann. Neurol. 59:499 – 503 .

Theien BE, Vanderlugt CL, Eagar TN, Nickerson-Nutter C, Nazareno R, Kuchroo VK, Miller SD. (2001) Discordant effects of anti-VLA-4 treatment before and after onset of relapsing experimental autoimmune encephalomyelitis. J. Clin. Invest. 107:995-1006.

Thelen M, Peveri P, Kernen P, von Tscharaner V, Walz A, Baggiolini M. (1988) Mechanism of neutrophil activation by NAF, a novel monocyte-derived peptide agonist. FASEB J. 2: 2702–6.

Thiery J, Keefe D, Saffarian S, Martinvalet D, Walch M, Boucrot E, Kirchhausen T, Lieberman J. (2010) Perforin activates clathrin- and dynamin-dependent endocytosis,

295 which is required for plasma membrane repair and delivery of granzyme B for granzyme-mediated apoptosis. Blood. 115:1582-93.

Thiery J, Keefe D, Boulant S, Boucrot E, Walch M, Martinvalet D, Goping IS, Bleackley RC, Kirchhausen T, Lieberman J. (2011) Perforin pores in the endosomal membrane trigger the release of endocytosed granzyme B into the cytosol of target cells. Nat. Immunol. 12:770-7.

Thomas DA, Du CY, Xu M, Wang XD, Ley TJ. (2000) DFF45/ICAD can be directly processed by granzyme B during the induction of apoptosis. Immunity. 12:621–632.

Thorén FB, Romero AI, Hermodsson S, Hellstrand K. (2007) The CD16-/CD56bright subset of NK cells is resistant to oxidant-induced cell death. J. Immunol. 179:781-5.

Thorén FB, Riise RE, Ousbäck J, Della Chiesa M, Alsterholm M, Marcenaro E, Pesce S, Prato C, Cantoni C, Bylund J, Moretta L, Moretta A. (2012) Human NK Cells induce neutrophil apoptosis via an NKp46- and Fas-dependent mechanism. J. Immunol. 188:1668-74.

Ting AT, Karnitz LM, Schoon RA, Abraham RT, Leibson PJ. (1992) Fc gamma receptor activation induces the tyrosine phosphorylation of both phospholipase C (PLC)- gamma 1 and PLC-gamma 2 in natural killer cells. J. Exp. Med. 176:1751–55.

Toft-Hansen H, Nuttal RK, Edwards DR, Owens T. (2004) Key Metalloproteinases Are Expressed by Specific Cell Types in Experimental Autoimmune Encephalomyelitis. J. Immunol. 173: 5209-18.

Tosi D, Valenti R, Cova A, Sovena G, Huber V, Pilla L, Arienti F, Belardelli F, Parmiani G, Rivoltini L. (2004) Role of cross-talk between IFN-alpha-induced monocyte-derived dendritic cells and NK cells in priming CD8+ T cell responses against human tumor antigens. J. Immunol. 172:5363-70.

Traboulsee A, Al-Sabbagh A, Bennett R, Chang P, Li DK; EVIDENCE Study Group; UBC MS/MRI Research Group. (2008) Reduction in magnetic resonance imaging T2 burden of disease in patients with relapsing-remitting multiple sclerosis: analysis of 48-week data from the EVIDENCE (EVidence of Interferon Dose-response: European North American Comparative Efficacy) study. BMC Neurol. 8:11.

296

Traggiai E, Biagioli T, Rosati E, Ballerini C, Mazzanti B, Ben Nun A, Massacesi L, Vergelli M. (2001) IL-7-enhanced T-cell response to myelin proteins in multiple sclerosis. J. Neuroimmunol. 121: 111–9.

Trajkovic V, Vuckovic O, Stosic-Grujicic S, Miljkovic D, Popadic D, Markovic M, Bumbasirevic V, Backovic A, Cvetkovic I, Harhaji L, Ramic Z, Mostarica SM. (2004) Astrocyte-induced regulatory T cells mitigate CNS autoimmunity. Glia. 47:168–79.

Trapani JA, Jans DA, Jans PJ, Smyth MJ, Browne KA, Sutton VR. (1998) Efficient nuclear targeting of granzyme B and the nuclear consequences of apoptosis induced by granzyme B and perforin are caspase-dependent, but cell death is caspase- independent. J. Biol. Chem. 273:27934-8.

Trapani JA, Smyth MJ. (2002) Functional significance of the perforin/granzyme cell death pathway. Nat. Rev. Immunol. 2:735-47.

Traugott U, Raine CS. (1984) Further lymphocyte characterization in the central nervous system in multiple sclerosis. Ann. N.Y. Acad. Sci. 436:163-80.

Treanor B, Lanigan PM, Kumar S, Dunsby C, Munro I, Auksorius E, Culley FJ, Purbhoo MA, Phillips D, Neil MA, Burshtyn DN, French PM, Davis DM. (2006) Microclusters of inhibitory killer immunoglobulin-like receptor signaling at natural killer cell immunological synapses. J.Cell Biol. 174:153-61.

Trinchieri G, Valiante N. (1993) Receptors for the Fc fragment of IgG on natural killer cells. Nat. Immun. 12:218-34.

Uddin S, Sassano A, Deb DK, Verma A, Majchrzak B, Rahman A, Malik AB, Fish EN, Platanias LC. (2002) Protein kinase C-delta (PKC-delta) is activated by type I interferons and mediates phosphorylation of Stat1 on serine 727. J. Biol. Chem. 277:14408–16.

Uguccioni M, D'Apuzzo M, Loetscher M, Dewald B, Baggiolini M. (1995) Actions of the chemotactic cytokines MCP-1, MCP-2, MCP-3, RANTES, MIP-1 alpha and MIP- 1 beta on human monocytes. Eur. J. Immunol. 25:64-8.

297

Uhrberg M, Valiante NM, Shum BP, Shilling HG, Lienert-Weidenbach K, Corliss B, Tyan D, Lanier LL, Parham P. (1997) Human diversity in killer cell inhibitory receptor genes. Immunity. 7:753-63.

Uitdehaag B, Constantinescu C, Cornelisse P, Jeffery D, Kappos L, Li D, Sandberg- Wollheim M, Traboulsee A, Verdun E, Rivera V. (2011) Impact of exposure to interferon beta-1a on outcomes in patients with relapsing-remitting multiple sclerosis: exploratory analyses from the PRISMS long-term follow-up study. Ther. Adv. Neurol. Disord. 4:3-14.

Ullberg M, Jondal M. (1983) Antibody- and interferon-dependent killer cells are part of the NK cell receptor positive subpopulation of human peripheral blood cells. Clin. Exp. Immunol. 53:101-8.

Upshaw JL, Arneson LN, Schoon RA, Dick CJ, Billadeau DD, Leibson PJ. (2006) NKG2D-mediated signaling requires a DAP10-bound Grb2-Vav1 intermediate and phosphatidylinositol-3-kinase in human natural killer cells. Nat. Immunol. 7:524-32

Vaddi K, Newton RC. (1994) Regulation of monocyte integrin expression by beta- family chemokines. J. Immunol. 153:4721-32.

Vaidya SV, Mathew PA. (2006) Of mice and men: Different functions of the murine and human 2B4 (CD244) receptor on NK cells. Immunology Letters. 105(2):180-4.

Valerio A, Ferrario M, Dreano M, Garotta G, Spano P, Pizzi M. (2002) Soluble interleukin-6 (IL-6) receptor/IL-6 fusion protein enhances in vitro differentiation of purified rat oligodendroglial lineage cells. Mol. Cell. Neurosci. 21:602–15.

Valés-Gómez M, Reyburn HT, Erskine RA, López-Botet M, Strominger JL. (1999) Kinetics and peptide dependency of the binding of the inhibitory NK receptor CD94/NKG2-A and the activating receptor CD94/NKG2-C to HLA-E. EMBO J. 18:4250-60.

Valiante NM, Phillips JH, Lanier LL, Parham P. (1996) Killer cell inhibitory receptor recognition of human leukocyte antigen (HLA) class I blocks formation of a pp36/PLC-gamma signaling complex in human natural killer (NK) cells. J. Exp. Med. 184:2243-50.

298

Valiante NM, Uhrberg M, Shilling HG, Lienert-Weidenbach K, Arnett KL, D'Andrea A, Phillips JH, Lanier LL, Parham P. (1997) Functionally and structurally distinct NK cell receptor repertoires in the peripheral blood of two human donors. Immunity. 7:739-51.

Vallance BA, Matthaei KI, Sanovic S, Young IG, Collins SM. (2000) Interleukin-5 deficient mice exhibit impaired host defence against challenge Trichinella spiralis infections. Parasite Immunol. 22:487–92.

Valledor AF, Sánchez-Tilló E, Arpa L, Park JM, Caelles C, Lloberas J, Celada A. (2008) Selective roles of MAPKs during the macrophage response to IFN-gamma. J. Immunol. 180:4523-9.

Valley CC, Lewis AK, Mudaliar DJ, Perlmutter JD, Braun AR, Karim CB, Thomas DD, Brody JR, Sachs JN. (2012) Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) induces death receptor 5 networks that are highly organized. J. Biol. Chem. 287:21265-78.

Vandenbark AA, Huan J, Agotsch M, La Tocha D, Goelz S, Offner H, Lanker S, Bourdette D. (2009) Interferon-beta-1a treatment increases CD56bright natural killer cells and CD4+CD25+ Foxp3 expression in subjects with multiple sclerosis. J. Neuroimmunol. 215:125-8. van der Slik AR, Koeleman BP, Verduijn W, Bruining GJ, Roep BO, Giphart MJ. (2003) KIR in type 1 diabetes: disparate distribution of activating and inhibitory natural killer cell receptors in patients versus HLA-matched control subjects. Diabetes. 52:2639-42.

Van der Veen RC, Roberts LJ. (1999) Contrasting roles for nitric oxide and peroxynitrite in the peroxidation of myelin lipids. J. Neuroimmunol. 95:1-7. van Domselaar R, Philippen LE, Quadir R, Wiertz EJ, Kummer JA, Bovenschen N. (2010) Noncytotoxic inhibition of cytomegalovirus replication through NK cell protease granzyme M-mediated cleavage of viral phosphoprotein 71. J. Immunol. 185:7605-13.

299

Van Ranst PC, Snoeck HW, Lardon F, Lenjou M, Nijs G, Weekx SF, Rodrigus I, Berneman ZN, Van Bockstaele DR. (1996) TGF-beta and MIP-1 alpha exert their main inhibitory activity on very primitive CD34+2CD38- cells but show opposite effects on more mature CD34+CD38+ human hematopoietic progenitors. Exp. Hematol. 24:1509-15.

Varinou L, Ramsauer K, Karaghiosoff M, Kolbe T, Pfeffer K, Müller M, Decker T. (2003) Phosphorylation of the Stat1 transactivation domain is required for full-fledged IFN-γ-dependent innate immunity. Immunity. 19: 793–802.

Velders MP, van Rhijn CM, Oskam E, Fleuren GJ, Warnaar SO, Litvinov SV. (1998) The impact of antigen density and antibody affinity on antibody-dependent cellular cytotoxicity: relevance for immunotherapy of carcinomas. Br. J. Cancer. 78:478-83.

Venken K, Hellings N, Thewissen M, Somers V, Hensen K, Rummens JL, Medaer R, Hupperts R, Stinissen P. (2007) Compromised CD4+ and CD25high regulatory T-cell function in patients with relapsing-remitting multiple sclerosis is correlated with a reduced frequency of FOXP3-positive cells and reduced FOXP3 expression at the single-cell level. Immunology. 123:79-89.

Verheyden S, Bernier M, Demanet C. (2004) Identification of natural killer cell receptor phenotypes associated with leukemia. Leukemia. 18:2002-7.

Viglietta V, Baecher-Allan C, Weiner HL, Hafler DA. (2004) Loss of Functional Suppression by CD4+CD25+ Regulatory T Cells in Patients with Multiple Sclerosis. J. Exp. Med. 199:971-9.

Villalta F, Zhang Y, Bibb KE, Kappes JC, Lima MF. (1998) The Cysteine-Cysteine Family of Chemokines RANTES, MIP-1a, and MIP-1b Induce Trypanocidal Activity in Human Macrophages via Nitric Oxide. Infect. Immun. 66:4690-5.

Vitale M, Bottino C, Sivori S, Sanseverino L, Castriconi R, Marcenaro E, Augugliaro R, Moretta L, Moretta A. (1998) 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. 187:2065-72.

300

Vitale M, Della Chiesa M, Carlomagno S, Romagnani C, Thiel A, Moretta L, Moretta A. (2004) The small subset of CD56brightCD16- natural killer cells is selectively responsible for both cell proliferation and interferon-gamma production upon interaction with dendritic cells. Eur. J. Immunol. 34:1715-22.

Vitale M, Della Chiesa M, Carlomagno S, Pende D, Aricò M, Moretta L, Moretta A. (2005) NK-dependent DC maturation is mediated by TNFalpha and IFNgamma released upon engagement of the NKp30 triggering receptor. Blood. 106:566-71.

Vitale C, Cottalasso F, Montaldo E, Moretta L, Mingari MC. (2008) Methylprednisolone induces preferential and rapid differentiation of CD34+ cord blood precursors toward NK cells. Int. Immunol. 20:565-75.

Vivian JP, Duncan RC, Berry R, O'Connor GM, Reid HH, Beddoe T, Gras S, Saunders PM, Olshina MA, Widjaja JM, Harpur CM, Lin J, Maloveste SM, Price DA, Lafont BA, McVicar DW, Clements CS, Brooks AG, Rossjohn J. (2011) Killer cell immunoglobulin-like receptor 3DL1-mediated recognition of human leukocyte antigen B. Nature. 479:401–5.

Vivier E, Morin P, O'Brien C, Druker B, Schlossman SF, Anderson P. (1991) Tyrosine phosphorylation of the Fc gamma RIII(CD16): zeta complex in human natural killer cells. Induction by antibody-dependent cytotoxicity but not by natural killing. J. Immunol. 146:206-10.

Vollmer TL, Liu R, Price M, Rhodes S, La Cava A, Shi FD. (2005) Differential effects of IL-21 during initiation and progression of autoimmunity against neuroantigen. J. Immunol. 174:2696-701.

Vos CM, Sjulson L, Nath A, McArthur JC, Pardo CA, Rothstein J, Conant K. (2000) Cytotoxicity by matrix metalloprotease-1 in organotypic spinal cord and dissociated neuronal cultures. Exp. Neurol. 163:324-30.

Vos CM, van Haastert ES, de Groot CJ, van der Valk P, de Vries HE. (2003) Matrix metalloproteinase-12 is expressed in phagocytotic macrophages in active multiple sclerosis lesions. J. Neuroimmunol. 138:106-14.

301

Vossen MT, Matmati M, Hertoghs KM, Baars PA, Gent MR, Leclercq G, Hamann J, Kuijpers TW, van Lier RA. (2008) CD27 defines phenotypically and functionally different human NK cell subsets. J. Immunol. 180:3739-45.

Vranes Z, Poljaković Z, Marusić M. (1989) Natural killer cell number and activity in multiple sclerosis. J. Neurol. Sci. 94:115-23.

Vyas YM, Mehta KM, Morgan M, Maniar H, Butros L, Jung S, Burkhardt JK, Dupont B. (2001) Spatial Organization of Signal Transduction Molecules in the NK Cell Immune Synapses During MHC Class I-Regulated Noncytolytic and Cytolytic Interactions. J. Immunol. 167:4358-67.

Vyas YM, Maniar H, Dupont B. (2002) Cutting edge: differential segregation of the SRC homology 2-containing protein tyrosine phosphatase-1 within the early NK cell immune synapse distinguishes noncytolytic from cytolytic interactions. J. Immunol. 168:3150-4.

Vyas YM, Maniar H, Lyddane CE, Sadelain M, Dupont B. (2004) Ligand Binding to Inhibitory Killer Cell Ig-Like Receptors Induce Colocalization with Src Homology Domain 2-Containing Protein Tyrosine Phosphatase 1 and Interruption of Ongoing Activation Signals. J. Immunol. 173:1571-8.

Wakerley B, Nicholas R, Malik O. (2008) Multiple sclerosis. Medicine. 36: 625-9.

Walker C, Checkel J, Cammisuli S, Leibson PJ, Gleich GJ. (1998) IL-5 Production by NK Cells Contributes to Eosinophil Infiltration in a Mouse Model of Allergic Inflammation. J. Immunol. 161:1962-9.

Walsh GM, Hartnell A, Wardlaw AJ, Kurihara K, Sanderson CJ, Kay AB. (1990) IL- 5 enhances the in vitro adhesion of human eosinophils, but not neutrophils, in a leucocyte integrin (CD11/18)-dependent manner. Immunology. 71:258-265.

Walsh CM, Matloubian M, Liu CC, Ueda R, Kurahara CG, Christensen JL, Huang MT, Young JD, Ahmed R, Clark WR. (1994) Immune function in mice lacking the perforin gene. Proc. Natl. Acad. Sci. USA. 91:10854–58.

302

Wandinger K, Jabs W, Siekhaus A, Bubel S, Trillenberg P, Wagner H, Wessel K, Kirchner H, Hennig H. (2000) Association between clinical disease activity and Epstein-Barr virus reactivation in MS. Neurology. 25:178 – 84.

Wang JM, Sherry B, Fivash MJ, Kelvin DJ, Oppenheim JJ. (1993) Human recombinant macrophage inflammatory protein-1 alpha and -beta and monocyte chemotactic and activating factor utilize common and unique receptors on human monocytes. J. Immunol. 150:3022-9.

Wang GQ, Wieckowski E, Goldstein LA, Gastman BR, Rabinovitz A, Gambotto A, Li S, Fang B, Yin XM, Rabinowich H. (2001a) Resistance to granzyme B-mediated cytochrome c release in Bak-deficient cells. J. Exp. Med. 194:1325-37.

Wang J, Chun HJ, Wong W,Spencer DM, Lenardo MJ. (2001b) Caspase-10 is an initiator caspase in death receptor signalling. Proc. Natl. Acad. USA. 98:13884-8.

Wang W, Hu D, Xiong H. (2008) Macrophage attenuation of neuronal excitability: implications for pathogenesis of neurodegenerative disorders. Glia. 56:241-6.

Warren HS, Kinnear BF, Phillips JH, Lanier LL. (1995) Production of IL-5 by human NK cells and regulation of IL-5 secretion by IL-4, IL-10, and IL-12. J. Immunol. 154:5144-52.

Waterhouse NJ, Sedelies KA, Sutton VR, Pinkoski MJ, Thia KY, Johnstone R, Bird PI, Green DR, Trapani JA. (2006) Functional dissociation of DeltaPsim and cytochrome c release defines the contribution of mitochondria upstream of caspase activation during granzyme B-induced apoptosis. Cell Death Differ. 13:607-18.

Watzl C, Long EO. (2003) Natural Killer Cell Inhibitory Receptors Block Actin Cytoskeleton-dependent Recruitment of 2B4 (CD244) to Lipid Rafts. J. Exp. Med. 197:77-85.

Weber F, Janovskaja J, Polak T, Poser S, Rieckmann P. (1999) Effect of interferon beta on human myelin basic protein-specific T-cell lines: comparison of IFNbeta-1a and IFNbeta-1b. Neurology. 52:1069-71.

303

Weber MS, Starck M, Wagenpfeil S, Meinl E, Hohlfeld R, Farina C. (2004) Multiple sclerosis: glatiramer acetate inhibits monocyte reactivity in vitro and in vivo. Brain. 127: 1370–1378.

Weisbart RH, Kwan L, Golde DW, Gasson JC. (1987) Human GM-CSF primes neutrophils for enhanced oxidative metabolism in response to the major physiological chemoattractants. Blood. 69:18-21.

Weissman IL, Shizuru JA. (2008) The origins of the identification and isolation of hematopoietic stem cells, and their capability to induce donor-specific transplantation tolerance and treat autoimmune diseases. Blood. 112:3543-53.

Wendt K, Wilk E, Buyny S, Buer J, Schmidt RE, Jacobs R. (2006) Gene and protein characteristics reflect functional diversity of CD56dim and CD56bright NK cells. J. Leukoc. Biol. 80:1529-41.

Westberg M, Feychting M, Jonsson F, Nise G, Gustavsson P. (2009) Occupational exposure to Uv light and mortality from multiple sclerosis. Am. J. Ind. Med. 52:353– 7.

White LC, Wright KL, Felix NJ, Ruffner H, Reis LFL, Pine R, Ting JP-Y. (1996) Regulation of LMP2 and TAP1 genes by IRF-1 explains the paucity of CD8 T cells in IRF-1 mice. Immunity. 5:247–53.

Wieczorek S, Hoffjan S, Chan A, Rey L, Harper L, Fricke H, Holle JU, Gross WL, Epplen JT, Lamprecht P. (2009) Novel association of the CD226 (DNAM-1) Gly307Ser polymorphism in Wegener's granulomatosis and confirmation for multiple sclerosis in German patients. Genes Immun. 10:591-5.

Wiemann K, Mittrücker HW, Feger U, Welte SA, Yokoyama WM, Spies T, Rammensee HG, Steinle A. (2005) Systemic NKG2D down-regulation impairs NK and CD8 T cell responses in vivo. J. Immunol. 175:720-9.

Wilk E, Kalippke K, Buyny S, Schmidt RE, Jacobs R. (2008) New aspects of NK cell subset identification and inference of NK cells' regulatory capacity by assessing functional and genomic profiles. Immunobiology. 213:271-83.

304

Willenborg DO, Staykova MA, Cowden WB. (1999). Our shifting understanding of the role of nitric oxide in autoimmune encephalomyelitis: A review. J. Neuroimmunol. 100: 21–35.

Willer CJ, Dyment DA, Sadovnick AD, Rothwell PM, Murray TJ, Ebers GC; Canadian Collaborative Study Group. (2005) Timing of birth and risk of multiple sclerosis: population based study. BMJ. 330: 120.

Williams NS, Moore TA, Schatzle JD, Puzanov IJ, Sivakumar PV, Zlotnik A, Bennett M, Kumar V. (1997) Generation of lytic natural killer 1.1+, Ly-49- cells from multipotential murine bone marrow progenitors in a stroma-free culture: definition of cytokine requirements and developmental intermediates. .J Exp. Med. 186:1609-14.

Willis SN, Stadelmann C, Rodig SJ, Caron T, Gattenloehner S, Mallozzi SS, Roughan JE, Almendinger SE, Blewett MM, Brück W, Hafler DA, O'Connor KC. (2009) Epstein-Barr virus infection is not a characteristic feature of multiple sclerosis brain. Brain. 132:3318-28.

Wilson JL, Heffler LC, Charo J, Scheynius A, Bejarano MT, Ljunggren HG. (1999) Targeting of human dendritic cells by autologous NK cells. J. Immunol. 163:6365-70.

Windebank KP, Abraham RT, Powis G, Olsen RA, Barna TJ, Leibson PJ. (1988) Signal transduction during human natural killer cell activation: inositol phosphate generation and regulation by cyclic AMP. J. Immunol. 141:3951–57.

Wing EJ, Magee DM, Whiteside TL, Kaplan SS, Shadduck RK. (1989) Recombinant human granulocyte/macrophage colony-stimulating factor enhances monocyte cytotoxicity and secretion of tumor necrosis factor α and interferon in cancer patients. Blood. 73:643-646.

Winkler-Pickett R, Young HA, Cherry JM, Diehl J, Wine J, Back T, Bere WE, Mason AT, Ortaldo JR. (2008) In vivo regulation of experimental autoimmune encephalomyelitis by NK cells: Alteration of primary adaptive responses. J. Immunol. 180: 4495-506.

305

Wolf BB, Schuler M, Echeverri F, Green DR. (1999) Caspase-3 is the primary activator of apoptotic DNA fragmentation via DNA fragmentation factor-45/inhibitor of caspase-activated DNase inactivation. J. Biol. Chem. 274:30651-6.

Wu J, Song Y, Bakker AB, Bauer S, Spies T, Lanier LL, Phillips JH. (1999) An activating immunoreceptor complex formed by NKG2D and DAP10. Science. 285:730-2.

Wu X, Pan W, He Y, Hsuchou H, Kastin AJ. (2010) Cerebral interleukin-15 shows upregulation and beneficial effects in experimental autoimmune encephalomyelitis. J. Neuroimmunol. 223:65-72.

Wuest SC, Edwan JH, Martin JF, Han S, Perry JS, Cartagena CM, Matsuura E, Maric D, Waldmann TA, Bielekova B. (2011) A role for interleukin-2 trans-presentation in dendritic cell-mediated T cell activation in humans, as revealed by daclizumab therapy. Nat. Med. 17:604–9.

Wulfing C, Purtic B, Klem J, Schatzle JD. (2003) Stepwise cytoskeletal polarization as a series of checkpoints in innate but not adaptive cytolytic killing. Proc. Natl. Acad. Sci. USA. 100:7767-72.

Wynn D, Kaufman M, Montalban X, Vollmer T, Simon J, Elkins J, O'Neill G, Neyer L, Sheridan J, Wang C, Fong A, Rose JW; CHOICE investigators. (2010) Daclizumab in active relapsing multiple sclerosis (CHOICE study): a phase 2, randomised, double- blind, placebo-controlled, add-on trial with interferon beta. Lancet Neurol. 9:381-90.

Xie JH, Nomura N, Koprak SL, Quackenbush EJ, Forrest MJ, Rosen H. (2003) Sphingosine-1-phosphate receptor agonism impairs the efficiency of the local immune response by altering trafficking of naive and antigen-activated CD4+ T cells. J. Immunol. 170:3662-70.

Xu W, Fazekas G, Hara H, Tabira T. (2005) Mechanism of natural killer (NK) cell regulatory role in experimental autoimmune encephalomyelitis. J. Neuroimmunol. 163:24-30.

Xu Q, Wang S, Jiang X, Zhao Y, Gao M, Zhang Y, Wang X, Tano K, Kanehara M, Zhang W, Ishida T. (2007) Hypoxia-induced astrocytes promote the migration of

306 neural progenitor cells via vascular endothelial factor, stem cell factor, stromal- derived factor-1alpha and monocyte chemoattractant protein-1 upregulation in vitro. Clin. Exp. Pharmacol. Physiol. 34:624–31.

Yamada H, Fredette B, Shitara K, Hagihara K, Miura R, Ranscht B, Stallcup WB, Yamaguchi Y. (1997) The brain chondroitin sulfate proteoglycan brevican associates with astrocytes ensheathing cerebellar glomeruli and inhibits neurite outgrowth from granule neurons. J. Neurosci. 17:7784–95.

Yamaguchi Y, Suda T, Ohta S, Tominaga K, Miura Y, Kasahara T. (1991) Analysis of the survival of mature human eosinophils: interleukin-5 prevents apoptosis in mature human eosinophils. Blood. 78:2542–2547.

Yamamoto K, Yamaguchi M, Miyasaka N, Miura O. (2003) SOCS-3 inhibits IL-12- induced STAT4 activation by binding through its SH2 domain to the STAT4 docking site in the IL-12 receptor beta2 subunit. Biochem. Biophys. Res. Commun. 310:1188- 93.

Yang X, Stennicke HR, Wang B, Green DR, Jänicke RU, Srinivasan A, Seth P, Salvesen GS, Froelich CJ. (1998a) Granzyme B mimics apical caspases. Description of a unified pathway for trans-activation of executioner caspase-3 and -7. J. Biol. Chem. 273:34278-83.

Yang X, Chang HY, Baltimore D. (1998b) Autoproteolytic Activation of Pro- Caspases by Oligomerization. Molecular Cell. 1:319–325, January, 1998

Yawata M, Yawata N, Draghi M, Partheniou F, Little AM, Parham P. (2008) MHC class I-specific inhibitory receptors and their ligands structure diverse human NK-cell repertoires toward a balance of missing self-response. Blood. 112:2369-80.

Yednock TA, Cannon C, Fritz LC, Sanchez-Madrid F, Steinman L, Karin N. (1992) Prevention of experimental autoimmune encephalomyelitis by antibodies against alpha 4 beta 1 integrin. Nature. 356:63-6.

Yen JH, Moore BE, Nakajima T, Scholl D, Schaid DJ, Weyand CM, Goronzy JJ. (2001) Major Histocompatibility Complex Class I-Recognizing Receptors Are Disease Risk Genes in Rheumatoid Arthritis. J. Exp. Med. 193:1159-68.

307

Yen JH, Ganea D. (2009) Interferon beta induces mature dendritic cell apoptosis through caspase-11/caspase-3 activation. Blood. 114:1344-54.

Yen JH, Kong W, Ganea D. (2010) IFN-beta inhibits dendritic cell migration through STAT-1-mediated transcriptional suppression of CCR7 and matrix metalloproteinase 9. J. Immunol. 184:3478-86.

Yeo TW, De Jager PL, Gregory SG, Barcellos LF, Walton A, Goris A, Fenoglio C, Ban M, Taylor CJ, Goodman RS, Walsh E, Wolfish CS, Horton R, Traherne J, Beck S, Trowsdale J, Caillier SJ, Ivinson AJ, Green T, Pobywajlo S, Lander ES, Pericak- Vance MA, Haines JL, Daly MJ, Oksenberg JR, Hauser SL, Compston A, Hafler DA, Rioux JD, Sawcer S. (2007) A second major histocompatibility complex susceptibility locus for multiple sclerosis. Ann. Neurol. 61:228–36.

Yoshimura A, Suzuki M, Sakaguchi R, Hanada T, Yasukawa H. (2012) SOCS, Inflammation, and Autoimmunity. Front Immunol. 3:20.

Young DA, Lowe LD, Clark SC. (1990) Comparison of the effects of IL-3, granulocyte-macrophage colony-stimulating factor, and macrophage colony- stimulating factor in supporting monocyte differentiation in culture: analysis of macrophage antibody-dependent cellular cytotoxicity. J. Immunol. 145: 607.

Yu H, Fehniger TA, Fuchshuber P, Thiel KS, Vivier E, Carson WE, Caligiuri MA. (1998) Flt3 ligand promotes the generation of a distinct CD34(+) human natural killer cell progenitor that responds to interleukin-15. Blood. 92:3647-57.

Yu J, Heller G, Chewning J, Kim S, Yokoyama WM, Hsu KC. (2007) Hierarchy of the human natural killer cell response is determined by class and quantity of inhibitory receptors for self-HLA-B and HLA-C ligands. J. Immunol. 179:5977-89.

Yu S, Cantorna MT. (2008) The vitamin D receptor is required for iNKT cell development. Proc. Natl. Acad. Sci. U.S.A. 105: 5207-12.

Yu J, Venstrom JM, Liu XR, Pring J, Hasan RS, O'Reilly RJ, Hsu KC. (2009) Breaking tolerance to self, circulating natural killer cells expressing inhibitory KIR for non-self HLA exhibit effector function after T cell-depleted allogeneic hematopoietic cell transplantation. Blood. 113:3875-84.

308

Yu J, Mao HC, Wei M, Hughes T, Zhang J, Park IK, Liu S, McClory S, Marcucci G, Trotta R, Caligiuri MA. (2010) CD94 surface density identifies a functional intermediary between the CD56bright and CD56dim human NK-cell subsets. Blood. 115:274-81.

Yusa S, Campbell KS. (2003) Src homology region 2-containing protein tyrosine phosphatase-2 (SHP-2) can play a direct role in the inhibitory function of killer cell Ig-like receptors in human NK cells. J. Immunol. 170:4539-47.

Zeis T, Schaeren-Wiemers N. (2008) Lame ducks or fierce creatures? The role of oligodendrocytes in multiple sclerosis. J. Mol. Neurosci. 35:91-100.

Zhang J, Dawson VL, Dawson TM, Snyder SH. (1994) Nitric oxide activation of poly(ADP-ribose) synthetase in neurotoxicity. Science. 263:687–9.

Zhang B, Yamamura T, Kondo T, Fujiwara M, Tabira T. (1997) Regulation of Experimental Autoimmune Encephalomyelitis by Natural Killer (NK) Cells. J. Exp. Med. 186:1677-87.

Zhang JJ, Zhao Y, Chait BT, Lathem WW, Ritzi M, Knippers R, Darnell JE Jr. (1998) Ser727-dependent recruitment of MCM5 by Stat1α in IFN-γ-induced transcriptional activation. EMBO J. 17:6963–71.

Zhang D, Beresford PJ, Greenberg AH, Lieberman J. (2001a) Granzymes A and B directly cleave lamins and disrupt the nuclear lamina during granule-mediated cytolysis. Proc. Natl. Acad. Sci. USA. 98:5746–51.

Zhang D, Pasternack MS, Beresford PJ, Wagner L, Greenberg AH, Lieberman J. (2002b) Induction of rapid histone degradation by the cytotoxic T lymphocyte protease Granzyme A. J. Biol. Chem. 276:3683-90.

Zhang GX, Gran B, Yu S, Li J, Siglienti I, Chen X, Kamoun M, Rostami A. (2003a) Induction of experimental autoimmune encephalomyelitis in IL-12 receptor-beta 2- deficient mice: IL-12 responsiveness is not required in the pathogenesis of inflammatory demyelination in the central nervous system. J. Immunol. 170:2153-60.

309

Zhang GX, Yu S, Gran B, Li J, Siglienti I, Chen X, Calida D, Ventura E, Kamoun M, Rostami A. (2003b) Role of IL-12 receptor beta 1 in regulation of T cell response by APC in experimental autoimmune encephalomyelitis. J. Immunol. 171:4485-92.

Zhang Y, Taveggia C, Melendez-Vasquez C, Einheber S, Raine CS, Salzer JL, Brosnan CF, John GR. (2006) Interleukin-11 potentiates oligodendrocyte survival and maturation, and myelin formation. J. Neurosci. 26:12174–85.

Zhang AL, Colmenero P, Purath U, Teixeira de Matos C, Hueber W, Klareskog L, Tarner IH, Engleman EG, Söderström K. (2007) Natural killer cells trigger differentiation of monocytes into dendritic cells. Blood. 110:2484-93.

Zhang H, Trivedi A, Lee J-U, Lohela M, Lee S M, Fandel TM, Werb Z, Noble- Haeusslein LJ. (2011) Matrix Metalloproteinase-9 and Stromal Cell-Derived Factor-1 Act Synergistically to Support Migration of Blood-Borne Monocytes into the Injured Spinal Cord. J. Neurosci. 31:15894-903.

Zhang Y, Wang H, Ren J, Tang X, Jing Y, Xing D, Zhao G, Yao Z, Yang X, Bai H. (2012) IL-17A synergizes with IFN-γ to upregulate iNOS and NO production and inhibit chlamydial growth. PLoS One. 7:e39214.

Zhao T, Zhang H, Guo Y, Zhang Q, Hua G, Lu H, Hou Q, Liu H, Fan Z. (2007a) Granzyme K cleaves the nucleosome assembly protein SET to induce single-stranded DNA nicks of target cells. Cell Death and Differentiation. 14:489–499.

Zhao T, Zhang H, Guo Y, Fan Z. (2007b) Granzyme K directly processes bid to release cytochrome c and endonuclease G leading to mitochondria-dependent cell death. J. Biol. Chem. 282:12104-11.

Ziemssen T, Kümpfel T, Schneider H, Klinkert WE, Neuhaus O, Hohlfeld R. (2005) Secretion of brain-derived neurotrophic factor by glatiramer acetate-reactive T-helper cell lines: Implications for multiple sclerosis therapy. J. Neurol. Sci. 233:109-12.

Zingoni A, Sornasse T, Cocks BG, Tanaka Y, Santoni A, Lanier LL. (2004) Cross- talk between activated human NK cells and CD4+ T cells via OX40-OX40 ligand interactions. J. Immunol. 173:3716-24.

310

Zingoni A, Sornasse T, Cocks BG, Tanaka Y, Santoni A, Lanier LL. (2005) NK cell regulation of T cell-mediated responses. Mol. Immunol. 42:451-4.

311