Immunological Mechanisms and Natalizumab Treatment in Multiple

Linköping University Medical Dissertations No. 1332 Linköping University Medical Dissertations No. 1332 Johan Mellergård

Immunological Mechanisms and Immunological Mechanisms and Natalizumab Treatment in Multiple

Natalizumab Treatment in Multiple Sclerosis and Mechanisms Immunological Sclerosis Studies on lymphocytes, inflammatory markers and magnetic resonance spectroscopy Studies on lymphocytes, inflammatory markers and magnetic resonance spectroscopy

Johan Mellergård Johan Mellergård

Natalizumab

Treatment in Multiple Sclerosis

Linköping 2012

Divisions of Neurology and Clinical Immunology Neurology and Clinical Immunology Department of Clinical and Experimental Medicine Department of Clinical and Experimental Medicine Faculty of Health Sciences, Linköping University, Faculty of Health Sciences, Linköping University, Sweden Sweden

 Johan Mellergård, 2012

Paper I, III and IV have been reprinted with permission of the respective copyright holders.

Cover: Cerebral MRI sagittal section showing MS white matter lesions and atrophy (FLAIR protocol). From Center for Medical Image Science and Visualization (CMIV), Linköping.

Printed by LiU-Tryck, Linköping, Sweden, 2012.

ISBN: 978-91-7519-787-6 ISSN: 0345-0082

To Sara, Selma, Morris & Bea

For now we see only a reflection as in a mirror; then we shall see face to face. Now I know in part; then I shall know fully, even as I am fully known. 1 Corinthians 13:12

Contents

Contents ______5 Original publications ______8 Abstract ______10 Abbreviations ______11 1. Introduction ______14 2. Background ______16 2.1 Historical aspects of MS – the treatment revolution ______16 2.2 Epidemiology and symptoms ______16 2.3 Diagnosing multiple sclerosis ______18 2.4 The immune system – introductory remarks ______19 2.5 Etiology and pathogenesis of multiple sclerosis ______19 2.6 Mechanisms of demyelination and neuroaxonal damage ______21 2.7 Components of the immune response ______22 2.8 Leukocyte adhesion and migration ______28 2.9 Inflammation versus neurodegeneration ______29 2.10 Measuring diffuse pathology in MS ______30 2.11 Disease-modifying treatments ______31 2.12 Progressive multifocal leukoencephalopathy (PML) ______33 3. Aims ______34 4. Methods ______36 4.1 Study subjects ______36 4.2 Procedures ______38 4.2.1 Disease scoring systems (paper I, II, III, IV) ______38 4.2.2 Routine CSF analyses (paper I, II, III) ______39 4.2.3 Multiplex bead assay (paper I, III) ______40 4.2.4 Flow cytometry (paper II, IV) ______41 4.2.5 Lymphocyte activation assay (paper II) ______42 4.2.6 Markers of neurodegeneration in CSF (paper III) ______42 4.2.7 Magnetic resonance spectroscopy – MRS (paper III) ______43 4.2.8 Absolute quantification of MRS data (paper III) ______45 4.2.9 Real-time RT-PCR (paper IV) ______45 4.3 Statistical methods ______46

Contents

4.4 Ethical considerations ______46 5. Results and Discussion ______48 5.1 Clinical outcome (paper I, II, III) ______48 5.2 CSF and plasma variables (paper I, II, III) ______49 5.2.1 Routine CSF analyses (paper I, II, III) ______49 5.2.2 Cytokines and chemokines in CSF and plasma – comparing multiple sclerosis with non-inflammatory neurological diseases (paper I) ______49 5.2.3 Cytokines and chemokines in CSF (paper I, III) ______51 5.2.4 Cytokines and chemokines in plasma (paper I) ______52 5.3 Blood lymphocyte composition (paper II) ______53 Main lymphocyte populations ______53 Lymphocyte subpopulations ______55 5.4 Lymphocyte activation assay (paper II) ______58 5.5 Markers of neurodegeneration in CSF (paper III) ______60 5.6 1H-MRS metabolite concentrations in NAWM (paper III) ______61 5.7 Transcriptional characteristics of CD4+ T cells (paper IV) ______63 6. Concluding remarks and future perspectives ______66 7. Conclusions ______68 8. Sammanfattning på svenska ______70 9. Acknowledgements ______72 10. References ______74

Original publications

This thesis is based on the following four papers, in the text referred to by their Roman numerals.

I Natalizumab treatment in multiple sclerosis: marked decline of chemokines and cytokines in cerebrospinal fluid. Johan Mellergård, Måns Edström, Magnus Vrethem, Jan Ernerudh and Charlotte Dahle Mult Scler. 2010 Feb;16(2):208-17

II An increase in B cell and cytotoxic NK cell proportions and increased T cell responsiveness in blood of natalizumab-treated multiple sclerosis patients. Johan Mellergård, Måns Edström, Maria C. Jenmalm, Charlotte Dahle, Magnus Vrethem, Jan Ernerudh Submitted

III Association between change in normal appearing white matter metabolites and intrathecal inflammation in natalizumab-treated multiple sclerosis. Johan Mellergård, Anders Tisell, Olof Dahlqvist Leinhard, Ida Blystad, Anne-Marie Landtblom, Kaj Blennow, Bob Olsson, Charlotte Dahle, Jan Ernerudh, Peter Lundberg, Magnus Vrethem PLoS One. 2012 Sept 17;7(9):e44739

IV Transcriptional characteristics of CD4+ T cells in multiple sclerosis: relative lack of suppressive populations in blood. Måns Edström, Johan Mellergård, Jenny Mjösberg, Maria C. Jenmalm, Magnus Vrethem, Rayomand Press, Charlotte Dahle, Jan Ernerudh Mult Scler. 2011 Jan;17(1):57-66

Abstract

Background: Multiple sclerosis (MS) is a chronic inflammatory, demyelinating disease of the central nervous system (CNS), and a frequent cause of neurological disability among young adults. In addition to focal inflammatory demyelinated lesions, diffuse white matter pathology as well as a neurodegenerative component with accumulating axonal damage and gliosis have been demonstrated and contribute to MS disease characteristics. The inflammatory component is considered autoimmune and mediated by auto-reactive T lymphocytes together with other cell populations of the immune system and their respective products like cytokines and chemokines. Treatment with natalizumab, a monoclonal antibody directed against the α4β1-integrin (VLA-4), reduces migration of potential disease-promoting cells to the CNS. The efficacy of natalizumab in reducing relapses and MRI activity is evident, however associated effects on the immune response and the neurodegenerative component in MS are not clear.

Methods: In total 72 MS patients were included, distributed among paper I-IV. We investigated effects associated with one-year natalizumab treatment in 31 MS patients regarding cytokine and chemokine levels in CSF and blood using multiplex bead assay analyses (paper I), as well as treatment effects on blood lymphocyte composition in 40 patients using flow cytometry, including functional assays of lymphocyte activation (paper II). Normal appearing white matter (NAWM) metabolite concentrations were assessed with proton magnetic resonance spectroscopy (1H-MRS) in 27 MS patients before and after one year of treatment (paper III). We also evaluated the balance between circulating T helper (Th) subsets in 33 MS patients using gene expression analyses of the CD4+ T cell related transcription factors in whole blood (paper IV).

Results: One-year natalizumab treatment was associated with a marked decline in pro-inflammatory cytokines (IL-1β and IL-6) and chemokines (CXCL8, CXCL9, CXCL10 and CXCL11) intrathecally. Circulating plasma levels of some cytokines (GM-CSF, TNF, IL-6 and IL-10) also decreased after treatment. Natalizumab treatment was further associated with an increase in lymphocyte numbers of major populations in blood (total lymphocytes, T cells, T helper cells, cytotoxic T cells, NK cells and B cells). In addition, T cell responsiveness to recall antigens and mitogens was restored after treatment. As to 1H-MRS metabolite concentrations in NAWM, no change in levels were detected post-to pretreatment on a group level. However, correlation analyses between one-year change in metabolite levels (total creatine and total choline) and levels of pro-inflammatory IL-1β and CXCL8 showed a pattern of high magnitude correlation coefficients (r=0.43-0.67). Gene expression analyses demonstrated a systemically reduced expression of transcription factors related to immunoregulatory T cell populations (regulatory T cells and Th2) in relapsing MS compared with controls.

Conclusions: Our findings support that an important mode of action of natalizumab is reducing lymphocyte extravasation, although cell-signalling effects through VLA-4 also may be operative. Correlation analyses between changes 1H-MRS metabolite concentrations and inflammatory markers possibly point towards an association between intrathecal inflammation and gliosis development in NAWM. Finally, gene expression analyses indicate a systemic defect at the mRNA level in relapsing MS, involving downregulation of beneficial CD4+ phenotypes.

Abbreviations

ANOVA analysis of variance APC antigen presenting cell BBB blood brain barrier Breg regulatory B cell CD cluster of differentiation cDNA complementary deoxyribonucleic acid (DNA) cMRI conventional magnetic resonance imaging Cr creatine CNS central nervous system CCL CC-chemokine ligand CSF cerebrospinal fluid CXCL CXC-chemokine ligand DC dendritic cell EAE experimental allergic (autoimmune) encephalomyelitis EDSS expanded disability status scale FOXP3 forkhead box P3 GA glatiramer acetate GFAP glial fibrillary acidic protein Gln glutamine Glu glutamate GM-CSF granulocyte-monocyte colony-stimulating factor HLA human leukocyte antigen 1H-MRS proton magnetic resonance spectroscopy IFN interferon Ig immunoglobulin IL interleukin Lac lactate LFA-1 leukocyte function-associated antigen 1 MAdCAM-1 mucosal addressin cell adhesion molecule 1 MBP myelin basic protein MHC major histocompatibility complex MHC I major histocompatibility complex class I MHC II major histocompatibility complex class II mIns myo-Inositol MRI magnetic resonance imaging mRNA messenger ribonucleic acid MS multiple sclerosis MSIS-29 multiple sclerosis impact scale 29 MSSS multiple sclerosis severity score NAA N-acetylaspartate NAAG N-acetylaspartylglutamate NAGM normal appearing grey matter NAWM normal appearing white matter NFL neurofilament light protein NK cell natural killer cell NKT cell natural killer cell with T cell properties OND other neurological diseases PBMC peripheral blood mononuclear cell 11

Abbreviations

PCR polymerase chain reaction PML progressive multifocal leukoencephalopathy PNS peripheral nervous system PPMS primary progressive multiple sclerosis P-tau phosphorylated tauprotein qMRI quantitative magnetic resonance imaging RRMS relapsing remitting multiple sclerosis S1P sphingosine-1-phosphate SDMT symbol digit modalities test SPMS secondary progressive multiple sclerosis rRNA ribosomal ribonucleic acid tCho total choline tCr total creatine TCR T cell receptor TGF tumor growth factor Th T helper tGlx total glutamate + glutamine tNA total N-acetylaspartate TNF tumor necrosis factor Treg cell regulatory T cell T-tau total tauprotein VCAM-1 vascular cell adhesion molecule 1 VLA-4 very late activation antigen 4

1. Introduction

The human nervous system contains about 100 billion nerve cells, neurons, and principally constitutes groups of neurons connected to each other, creating a network of nerve circuits with astounding complexity. Apart from neurons, specialized to receive and transmit electric impulses, the nervous system also consists of cells with supportive functions, the neuroglia cells. The brain and spinal cord define the central nervous system (CNS) and the cranial and spinal nerves with their ganglia the peripheral nervous system (PNS). A neuron consists of a cell body with two types of extensions; the receiving dendrites with many branches in the immediate neighborhood of the cell body, and one axon that conducts the nerve impulse away from the cell body. Axons may be myelinated or unmyelinated depending on whether they are surrounded by a sheath of isolating lipoproteins, the myelin, or not. Myelin is required for the rapid propagation of action potentials along the axon. Parts in the CNS containing many myelinated axons designate the white matter, whereas parts containing aggregations of neuron cell bodies designate the gray matter (Heimer 1995).

Multiple sclerosis (MS) is a chronic inflammatory demyelinating disease of the CNS. In MS, the inflammatory response results in myelin lesions (demyelination) causing deteriorated nerve impulse propagation (Bruck 2005). Although the hallmark of MS is focal demyelinated white matter, the disease is also denoted by axonal damage and gliosis (Dutta and Trapp 2007). Depending on the size and anatomical localization of the lesion, the clinical result is reduction or loss of the corresponding neurological function. The onset of symptoms is often referred to as a relapse. Within weeks the local inflammation attenuates and eventually the myelin may be partially restored (remyelination) resulting in fully or partial regained neurological function (Kotter, Stadelmann et al. 2011). With time there might be no fully recovery from relapse- caused neurological deficits and there is an accumulation of axonal damage causing increasing neurological disability (Dutta and Trapp 2011). The MS disease now clinically changes character into a progressive course, where inflammation as underlying pathomechanism seems replaced by neurodegeneration.

This thesis discusses the immunological mechanisms and the inflammatory response in MS and deals with the interplay between inflammation and neurodegeneration. The approach is by mainly focusing on the immunomodulating treatment with natalizumab, highlighted from different perspectives.

2. Background

2.1 Historical aspects of MS – the treatment revolution

The first description of MS pathology dates back to 1868, when Charcot documented the focal lesions in the brain of a young woman who had suffered from nystagmus, intention tremor and scanning speech. In his report, “Histologie de la sclérose en plaques”, he described the characteristic plaques that still constitute the hallmark of MS (Compston 1988). Up to the last decade of the 20th century there were no disease-modulating treatments available, only symptomatic treatment could be offered. Life with MS was an evitable journey towards increasing disability in various extensions. In the mid-90th interferon beta-1b was approved as the first disease-modifying therapy for relapsing-remitting MS (RRMS). The approval was a result of a randomized, double-blind, placebo-controlled trial in 1993 (Paty and Li 1993). That event started a new era of MS history with the possibility to modify the disease course for patients with relapsing MS. Ever since, the number of disease-modifying treatments has grown and today (august 2012) there are five approved immunomodulators for the treatment of relapsing MS (interferon beta-1a, interferon beta-1b, glatiramer acetate, natalizumab and fingolimod), and the number will further increase. By using disease-modifying treatment early in the disease course the accumulation of irreversible brain damage and subsequent neurological disability may decrease. In the foreseeable future there is no cure for MS, but with the increasing pallet of disease-modifying drugs, hopefully MS may become a disease with a far better reputation than in the past.

2.2 Epidemiology and symptoms

The MS prevalence worldwide is unevenly distributed, in general varying from about 10/100 000 to 100/100 000 what is believed to depend on geographical and ethnical/genetic factors (Rosati 2001). The highest prevalence is seen in northwestern Europe and in northern North America and the lowest in Asia and Africa. Women are affected roughly twice as often as men. Sweden has among the highest nationwide prevalence estimates of MS in the world with 189/100 000 (Ahlgren, Oden et al. 2012), which correspond to 600 new cases in Sweden each year. Most patients are diagnosed with MS between the ages of 20-50 (incidence peak at 30 year) and MS is, after acquired disability from accidents, the most frequent cause of neurological disability among young adults. The usual clinical presentation, occurring in approximately 85% of patients, is a relapsing-remitting disease course with varying neurological deficits fluctuating over time (Figure 1). Early features that predict a benign disease course are complete remission of the first relapse, low or moderate initial relapse frequency and dominance of afferent symptoms (Skoog, Runmarker et al. 2012). With time, however, the inflammatory activity in the CNS seems to decline and the relapsing disease turns into a progressive state distinguished by increasing axonal loss and deterioration of neurological function without clinical relapses (secondary progressive MS, SPMS). For 16

2. Background approximately 15% of patients the disease course is progressive from the start (primary progressive MS, PPMS) (Miller and Leary 2007).

Figure 1. The usual disease course in MS. Subclinical disease activity within the CNS is followed by a relapsing- remitting disease course. With time there is an accumulation of disability and a gradual transition to the progressive phase. In parallel with inflammation, neuroaxonal damage causes increasing cerebral atrophy. Magnetic resonance imaging (MRI) is a valuable tool to assess disease activity and progression.

The clinical symptoms related to a relapse vary depending on where the demyelinated lesion is located in the CNS and thus highly diverse signs and symptoms may arise. There is however a high propensity for focal lesions to occur in the periventricular white matter, cerebellum and brain stem (Fazekas, Barkhof et al. 1999). The first clinical presentation of MS is usually a clinically isolated syndrome (CIS) (Miller, Chard et al. 2012). Principally there are so called negative and positive symptoms. Negative symptoms are characterized by a reduction or loss of function like motor weakness, reduced sensation ability, reduction or loss of vision and reduced coordination. It is easy to comprehend that affected nerve conduction in demyelinated axons may cause these symptoms. Positive symptoms on the other hand arise as a consequence of ectopic and emphatic nerve transmission with symptoms like paresthesias, spasms and paroxysmal pain (Raine 1997). In general optic neuritis, sensory disturbances and mild to moderate paresis are common in the initial disease course, but with time ataxia, increased muscle tonus with spasticity and genito-urinary as well as bowel dysfunction develop. Apart from the described focal neurological deficits, cognitive dysfunction is common (Chiaravalloti

2. Background and DeLuca 2008). Loss of energy, fatigue, is one common symptom associated with cognitive dysfunction, although the mechanisms behind are elusive.

2.3 Diagnosing multiple sclerosis

The principal when diagnosing MS is to demonstrate dissemination of CNS lesions in time and space as well as excluding alternative diagnosis (Miller, Weinshenker et al. 2008) (Miller, Chard et al. 2012). A careful medical history and a thorough neurological examination form the basis for further diagnostic considerations. Since 2001 the McDonald criteria are the foundation for diagnosing MS (McDonald, Compston et al. 2001). Although MS may be diagnosed on clinical grounds alone, the use of magnetic resonance imaging (MRI) is fundamental for an accurate MS diagnosis. According to the latest revision of the McDonald criteria from 2010, the use of MRI for demonstration of dissemination of lesions in time and space has been emphasized, in some circumstances facilitating an MS diagnosis by a single MRI scan (Polman, Reingold et al. 2011). The investigation of cerebrospinal fluid (CSF) with the assessment of intrathecal production of immunoglobulins and oligoclonal bands (reflecting intrathecal Ig production) can support the diagnosis but is not mandatory. However, when to exclude an alternative diagnosis to MS, CSF analyses are crucial, although intrathecal Ig production occurs in other inflammatory conditions. In the future, with the identification of diagnostic and prognostic markers, the importance of CSF evaluations may increase.

2. Background

2.4 The immune system – introductory remarks

The immune system constitutes a complex network of different cells and molecules necessary for our protection from various microbes, but also to fight cancer cells and promote healing. Furthermore, the components of the immune system also participate in a wide range of other biological processes from embryology to behavior, in close interplay with the endocrine and the nervous system. Traditionally the immune system is divided into innate and adaptive responses (Abbas A.K. 2012). Innate immunity is the first, constantly mobilized, line of defense against infections, and involves epithelial barriers, phagocytic cells (neutrophils and macrophages), natural killer (NK) cells, mast cells, dendritic cells, the complement system and many other soluble mediators. Innate immunity uses “standardized” procedures to fight invading microbes, utilizing structures that are common to groups of microbes, but cannot evolve highly specific immune mechanisms and, importantly, has no memory function. On the contrary, adaptive immune responses evolve slower (within days), use highly specific immune mechanisms to combat various infections and have a memory function. The adaptive immune system consists of lymphocytes, with a total estimated number of 4.6 x 1011 (Ganusov and De Boer 2007). Lymphocytes are divided into T and B lymphocytes, commonly and here referred to as T and B cells. During T cell maturation in the thymus, the T cell receptor (TCR) evolves and there is an expression of either CD4 or CD8 co-receptor molecules, denoted as CD4+ and CD8+ T cells respectively. CD4+ and CD8+ T cells have different capabilities that complement each other. The development of antigen specific immunity is dependent on the TCR recognition of antigens presented by antigen presenting cells (APC). The most efficient APC is the dendritic cell, thus a member of innate immunity with a major role in initiating adaptive immunity. Also macrophages and B cells act as APCs. The antigen is presented to the T cells by major histocompatibility complex (MHC) molecules (in humans often referred to as HLA, human leukocyte antigen) on the APC. MHC molecules are polymorphic glycoproteins that transport antigens and display them on the cell-membrane. There are two main types of MHC molecules, class I and class II. MHC class I molecules (MHC I) are constitutively expressed on all nucleated cells and platelets, and present peptides from endogenously synthesized antigens to CD8+ T cells. MHC class II molecules (MHC II) are expressed mainly on dendritic cells, macrophages and B cells and present peptides from processed exogenous antigens to CD4+ T cells thereby initiating CD4+ T cell immune responses (Abbas A.K. 2012).

2.5 Etiology and pathogenesis of multiple sclerosis

The precise etiology of MS is still unknown but a complex interplay between genetic susceptibility and environmental factors is most likely (Sospedra and Martin 2005). The strongest disease association is linked to human leukocyte antigen (HLA) class II genes (in particular within the HLA-DRB1*15:01 haplotype), which increase the risk for disease about 3- 4 times (International Multiple Sclerosis Genetics, Wellcome Trust Case Control et al. 2011). MS is considered an autoimmune disease; i. e. the basis for the inflammatory lesions is an immune response against self-antigens. Generally, this conclusion is based on what we learned

2. Background from the animal model of MS, experimental allergic (or autoimmune) encephalomyelitis (EAE). The general procedure of inducing EAE is to inject a myelin-derived protein in conjunction with Freund´s adjuvant into a rodent. The rodent will then develop an immune response involving myelin-specific T cells that migrate into the brain initiating an inflammatory attack on myelin structures. This results in a disability and histopathology that share many similarities with MS (Raine 1997). Since EAE could be transferred to naïve animals by in vitro reactivated myelin-specific CD4+ T cells (adoptive transfer), it was assumed that MS too is a T cell-mediated autoimmune disease (Zamvil and Steinman 1990). Still, substantial evidence accumulated over many years confirms that the histopathological hallmark of MS is inflammatory demyelination with axonal damage (Hemmer, Archelos et al. 2002). However, the initial processes that trigger this inflammatory response are yet to be defined. In addition to inflammation, a process of degeneration is crucial in MS pathology and it has even been suggested that MS primarily is a neurodegenerative disease instead of an autoimmune disease. Thus there are data suggesting that the first event in the pathogenesis of MS lesions is the loss of oligodendrocytes independent of adaptive immunity mechanisms (Henderson, Barnett et al. 2009). On the contrary, among genes associated with MS, there is a significant overrepresentation of genes with immunological relevance, in particular genes related to T cell differentiation (International Multiple Sclerosis Genetics, Wellcome Trust Case Control et al. 2011), which supports the inflammatory theory.

Furthermore, although T and B cell populations are considered major contributors to inflammation and tissue damage (Goverman 2009, Fletcher, Lalor et al. 2010, Disanto, Morahan et al. 2012), there is increasing evidence that other mechanisms and factors may have a major impact on disease development and progression (Sriram 2011).

Principally, the pathogenesis of MS involves (1) the activation and clonal expansion of auto- reactive lymphocyte clones (T and B cells) in the peripheral lymph nodes and spleen or the gastrointestinal (Berer, Mues et al. 2011) or respiratory tract (Odoardi, Sie et al. 2012), (2) the migration and infiltration of these clones into the CNS and (3) the reactivation of these cells within the CNS by resident immune cells presenting self-antigens (Figure 2). (1) Although still uncertainties about what specific antigens constituting the target for auto-reactive cells, many studies indicate the importance of myelin proteins (Goverman 2011). CNS antigens could either be released to the periphery or there may be a cross-reaction to foreign antigens (Ochoa- Reparaz, Mielcarz et al. 2011). (2) The subsequent homing of activated T and B cells to the inflammatory region in CNS and migration through the blood-brain barrier (BBB) is dependent on chemokine signalling and the interaction between cell surface molecules (as integrins) and corresponding endothelial adhesion molecules (Rice, Hartung et al. 2005). (3) Once inside the CNS, T and B cells are reactivated by neurons and glial cells as well as APCs like microglia, through recognition of antigens presented on MHC class I and II molecules respectively (Hemmer, Archelos et al. 2002). This interaction is also largely dependent on signalling molecules like cytokines and chemokines. Furthermore, the magnitude of the inflammatory response depends on the balance between auto-reactive T cells and immunosuppressive T cells. With time, the inflammation may be self-generating and compartmentalized within the CNS (Meinl, Krumbholz et al. 2008).

2. Background

Figure 2. Schematic view of mechanisms of lymphocyte migration and effector responses. Briefly, after activation of CD4+ T cells in the periphery by antigen presenting cells (APC),migration through the blood-brain barrier (BBB) into the CNS are dependent on the interaction between integrin VLA-4 and VCAM-1 and chemokine gradient guidance. Within the CNS, reactivation of auto-reactive T cell clones by macrophages and microglia initiates an inflammatory cascade, which includes the secretion of different cytokines, eventually leading to demyelination and axonal damage. Pro-inflammatory (red) and immunoregulatory (green) cytokines and Th cell subsets are also shown.

2.6 Mechanisms of demyelination and neuroaxonal damage

The formation of a demyelinated lesion in CNS involves the interplay between mechanisms of both adaptive and innate immunity (Weiner 2009). As to adaptive immunity, activated CD4+ T cells facilitate effector mechanisms of other immune cells by cell-to cell interactions and the secretion of pro-inflammatory cytokines. A pro-inflammatory cytokine milieu activates residence cells like microglia and astrocytes as well as recruits additional immune cells like monocytes, mast cells, CD4+ and CD8+ T cells and B cells to the site of inflammation. CD8+ T cells, can mediate direct cytotoxic effects to neurons and glial cells by the release of toxic granules inducing apoptosis of the target cell. B cells, apart from acting as APCs and regulatory cells can secrete cytokines and exert humoral responses, by producing autoantibodies that specifically may target neuronal and glial cells. Turning to innate immunity responses, these are main players in inflammation and demyelination. As to cell populations, monocytes/macrophages, dendritic cells, NK cells, mast 21

2. Background cells, NKT cells and γδ-T cells have all been assigned involvement in inflammation and tissue damage (Gandhi, Laroni et al. 2010). As one prominent example, monocytes are differentiated into tissue macrophages releasing pro-inflammatory cytokines and injurious reactive oxygen intermediates and nitric oxide. Additionally, macrophages may also exert direct phagocytic attacks on myelin sheaths. Also microglia, being resident cells of the macrophage lineage, are involved in the process (Sriram 2011).

Soluble factors as matrix metalloproteinases and histamine further add to the breakdown of the BBB and damage to neurons and glial cells (Larochelle, Alvarez et al. 2011). Disturbances in glutamate metabolism among glia cells may also contribute to white matter damage (Matute, Domercq et al. 2006).

2.7 Components of the immune response

This thesis investigates some of the immunological mechanisms underlying MS. Below is a description of the main components assessed. As mentioned, MS pathogenesis has been linked to the adaptive immune system by T cell-mediated immune regulation, involving both CD4+ T helper cells and CD8+ T cytotoxic cells (McFarland and Martin 2007). However, the pathogenetic picture has become far more diverse also including B lymphocytes and lymphocytes of the innate immune system. Clearly, the distribution of different cell populations is an important facet of MS pathology and regulation.

CD4+ T cells

CD4+ T cell directed immune responses can roughly be mediated by effector T cells as T helper 1 (Th1), Th2, Th17 and regulatory T cells (Treg cells). As to MS, Th1 and Th17 cells are thought to represent auto-reactive disease-promoting T cell populations, whereas Th2 and Treg cells seem to have protective abilities (Fletcher, Lalor et al. 2010, Kasper and Shoemaker 2010). Although attempts have been done, mainly in experimental models (Racke, Bonomo et al. 1994, Ekerfelt, Dahle et al. 2001, Hallin, Mellergard et al. 2006), utilizing this concept therapeutically in autoimmunity has been proven difficult. CD4+ Th1 cells have traditionally been assigned a pivotal disease-promoting role in autoimmunity by the production of pro- inflammatory cytokine interferon γ (IFN-γ) (Petermann and Korn 2011), indeed underscored by the worsening of MS during IFN-γ therapy (Panitch, Hirsch et al. 1987). The discovery of the Th17 subset has changed this view since Th17 cells were shown absolutely essential for initiating EAE (Langrish, Chen et al. 2005). Th17 cells produce the pro-inflammatory cytokine IL-17 and elevated intracellular expression of IL-17 in peripheral blood mononuclear cells (PBMC) has been demonstrated in patients with active MS (Durelli, Conti et al. 2009) and also in CSF (Matusevicius, Kivisakk et al. 1999). Moreover, MS endothelial cells express high levels of IL-17 receptors and IL-17 is shown to facilitate BBB permeability, altogether further promoting CNS inflammation (Kebir, Kreymborg et al. 2007). Interestingly, it was recently proposed that granulocyte-monocyte colony-stimulating factor (GM-CSF) was required to

2. Background induce CNS pathology in EAE irrespective of a Th1- or a Th17-mediated disease mechanism (Codarri, Gyulveszi et al. 2011).

Treg cells are distinguished by their high expression of the IL-2 receptor alpha chain (CD25) and the transcriptor factor forkhead box p3 (FOXP3). Although FOXP3 is the signature marker of Treg cells, CD4+ cells with a slightly lower CD4 expression (CD4dim) and high CD25 expression (CD25bright) are also shown to correspond well to FOXP3-high expressing cells (Mjosberg, Svensson et al. 2009). Treg cells have been designated a central role for maintenance of tolerance and for the protection against autoimmune diseases (Zozulya and Wiendl 2008). Treg cells may prevent the development of EAE (Kohm, Carpentier et al. 2002) and immunosuppressive function of Treg cells has been shown impaired in MS patients compared to controls (Viglietta, Baecher-Allan et al. 2004, Haas, Hug et al. 2005).

The differentiation of a naive CD4+ T cell into an effector T cell is dependent on its recognition of antigens presented by MHC II molecules on APCs. However, for the proper activation of the T cell, co-stimulatory cell-to-cell interactions as well as soluble signals are also required. The co-stimulatory signals from APCs are referred to as the second signal. A well-known example is the binding of T cell surface molecule CD28 to B7-1 (CD80) and B7-2 (CD86) on APCs. The soluble signals required refers to the cytokine milieu in which the activation takes place. Furthermore, part of the T cell effector mechanisms involves secretion of specific cytokines that orchestrate the subsequent immune response (Figure 3). For example, Th1 requires IL-12 for its differentiation and is characterized by secretion of IFN-γ, whereas Th2 requires IL-4 and secretes IL-4, IL-5 and IL-13 (Farrar, Asnagli et al. 2002, Murphy and Reiner 2002). As to Th17, differentiation is dependent on IL-23, transforming growth factor β (TGF-β) and IL-6 and effector responses are characterized by secretion of IL-17A, IL-17F, IL-21 and IL-22 (Miossec, Korn et al. 2009). Treg differentiation requires TGF-β and is denoted by the secretion of IL-10 and TGF-β (Sakaguchi, Miyara et al. 2010). Intracellularly the development of Th subpopulations is denoted by expression of lineage specific transcription factors as Tbet/TBX21 (Th1) (Szabo, Kim et al. 2000), GATA3 (Th2) (Zheng and Flavell 1997), RORC (Th17) (Ivanov, McKenzie et al. 2006) and FOXP3 (Treg) (Fontenot, Gavin et al. 2003) (Figure 3).

CD8+ T cells

CD8+ T cells are major effector cells of the adaptive immune system. Consistent with this, it is reported that MS lesions show a predominance of CD8+ T cells and that axonal damage within lesions correlate best with the number of CD8+ T cells (Kuhlmann, Lingfeld et al. 2002, Friese and Fugger 2009). Thus, even though CD4+ T cells are considered to initiate or direct the immune attack on myelin, CD8+ T cells are crucial for continuous demyelination and subsequent axonal damage. Effector mechanisms of activated CD8+ T cells include the secretion of IFN-γ and tumor necrosis factor (TNF) and direct toxic properties to myelin and axons by perforin and granzymes (Kaech and Wherry 2007).

2. Background

Figure 3. Schematic view of CD4+ T cell lineages and corresponding cytokines. Different cytokines drive T helper (Th) cells into different subsets that express different lineage-specific transcription factors. Different Th subsets secrete different sets of cytokines. Note that the inducing cytokines are mainly secreted by other cells, in particular antigen presenting cells (APC). Also note that the Treg cells depicted here are induced Treg cells, while natural Treg cells are produced in thymus.

B cells

The importance of B cells in the pathophysiology of MS has become indisputable in recent years. The background for this include the findings of clonally expanded B cells producing immunoglobulins with overlapping repertoire within CSF and brain lesions (Obermeier, Lovato et al. 2011), large numbers of plasma cells in subacute and chronic MS plaques (Henderson, Barnett et al. 2009) and the therapeutic effect of B cell depletion (Bar-Or, Calabresi et al. 2008). In addition, some EAE models are mainly dependent on antibodies (Pollinger, Krishnamoorthy et al. 2009). The presence of antibody production in terms of CSF oligoclonal bands is demonstrated in more than 95% of MS patients and thus constitutes a typical finding in MS (Freedman, Thompson et al. 2005). Furthermore, CSF levels of B cell marker cytokine CXCL13, is shown to correlate both with clinical disease activity and progression as well as with paraclinical measures as CSF total leukocyte and B cell counts, intrathecal IgG synthesis, markers of demyelination and MRI activity (Khademi, Kockum et al. 2011) (Disanto, Morahan et al. 2012). Apart from antibody-dependent effector mechanisms of B cells, this lymphocyte population can also be divided into subsets with divergent functions including pro- inflammatory or regulatory subsets. Memory B cells (CD19+CD27+) are reported to be

2. Background distinguished from naïve B cells (CD19+ CD27-) by the secretion of pro-inflammatory cytokines tumor necrosis factor (TNF) and lymphotoxin (LT) upon stimulation, whereas naïve B cells secrete immunoregulatory IL-10 (Duddy, Niino et al. 2007). Reduction in IL-10 producing B cells (Breg) in MS was accompanied by a reduced naïve/memory Breg ratio during relapse but not in remission (Knippenberg, Peelen et al. 2011). Breg cells have, despite their name, no association with Treg cells with regard to differentiation, action and expression of FOXP3. Still, Breg cells constitute a heterogeneous population with different definitions, one being CD25 expression. CD19+CD25high Breg cells have been shown to suppress CD4+ T cell proliferation and enhance Treg properties (Kessel, Haj et al. 2012), as well as suggested a regulatory role in vasculitis (Eriksson, Sandell et al. 2010).

Natural killer cells

Natural killer (NK) cells are lymphocytes of the innate immune system, having both cytotoxic and regulatory functions (Gandhi, Laroni et al. 2010).(Berzins, Smyth et al. 2011, Kaur, Trowsdale et al. 2012). They constitute a unique cell population considering their ability to direct lyse tumor cells and virus-infected cells without a preceding sensitization. The regulatory functions are primarily executed by the CD3- CD56bright subset secreting cytokines as IL-10 (Poli, Michel et al. 2009). Based on different expression of chemokine receptors and adhesion molecules, cytotoxic NK cells (CD56dim) and regulatory NK cells (CD56bright) have different migration preferences with CD56dim migrating to inflammatory sites while CD56bright preferentially home to secondary lymphoid organs (Campbell, Qin et al. 2001). Interestingly, NK cells are shown to mediate differentiation of dendritic cells, thereby shaping the subsequent immune response (Moretta, Marcenaro et al. 2008). In that respect, NK cells are a striking example of the close interplay between innate and adaptive immunity. NK regulatory subset (CD56bright) is proposed having a beneficial effect in MS since there is an increase in their numbers on immunomodulatory treatment (Bielekova, Catalfamo et al. 2006, Saraste, Irjala et al. 2007) and in ameliorated MS during pregnancy (Airas, Saraste et al. 2008). Additionally, a reduction in NK cell function in the periphery was demonstrated to correlate with the onset of clinical relapse in MS patients (Kastrukoff, Morgan et al. 1998, Kastrukoff, Lau et al. 2003).

Cell surface markers of activation and co-stimulation

Activation of a T cell (upon ligand recognition of the TCR) triggers intracellular signalling pathways that cause activation of gene transcripts and eventually the production of corresponding proteins. Proteins synthesized may then be expressed on the cell-surface and used as a measure of T cell activation. Based on the time elapse from TCR stimulation to expression on the cell surface, activation markers may be referred to as early or late markers of activation.

CD69 is a membrane bound receptor that acts as a co-stimulatory molecule. It is a marker of very early activation of T cells since it is detectable within 1-2 hours of antigen ligation of the TCR (Ziegler, Ramsdell et al. 1994). It facilitates the activation and proliferation of other T cells and is also suggested to have a downregulatory role in immune responses (Sancho, Gomez et al. 2005). CD69 expression could be transient but in an environment of pro-inflammatory

2. Background cytokines, i. e. at the inflammatory site, CD69 expression is prolonged (Sancho, Gomez et al. 2005). CD25 is part of the IL-2 receptor (α-chain) and is expressed within 24 hours after TCR stimulation (Caruso, Licenziati et al. 1997). Signalling through the IL-2 receptor is crucial for the activation, differentiation and expansion of effector T cells (Liao, Lin et al. 2011). HLA- DR, a MHC class II cell-membrane receptor molecule, is expressed after about 48 hours and thus constitutes a marker of late T cell activation (Caruso, Licenziati et al. 1997).

OX40 (CD134) and its ligand OX40L (CD252) represent a co-stimulatory signalling system that promotes effector T cell expansion and survival (Croft, So et al. 2009). OX40L is not constitutively expressed but induced within hours of T cell activation. OX40/OX40L interactions also modulate cytokine production from NK cells, NKT cells and APCs and have also been shown important for the development of EAE (Weinberg, Bourdette et al. 1996). Moreover, co-stimulation by T cell OX40L was suggested to deviate a Th1 response towards a Th2 response under certain circumstances (Mendel and Shevach 2006).

Cytokines

Cytokines comprise a large group of small signalling proteins secreted by all kinds of cells of the immune system as well as by several non-immune cells. They constitute the key mediators of communication between immune cells and thus regulate and shape all immunological pathways. Cytokine effects are characterized by pleiotrophism and redundancy, the former refers to the ability of a specific cytokine to exert different effects on different cell types, the latter implicates that multiple cytokines may have the same or similar functional effects (Akdis, Burgler et al. 2011) (Table 1). However, as already described, different CD4+ T cells are denoted by their secretion of specific cytokines. Accordingly, the balance between pro- inflammatory and regulatory cytokines has a major influence on the initiation and development of MS (Sospedra and Martin 2005). As to IL-1 (IL-1α and IL-1β), it is an important regulator of homeostasis in CNS under physiologic conditions, but levels are rapidly increased in response to injury (Basu, Krady et al. 2004). It is suggested that IL-1 is at the top of the cytokine signalling cascade by acting as an upstream signal for other pro-inflammatory cytokines and mediators (Basu, Krady et al. 2004).

2. Background

Table 1. Description of cytokines and chemokine CXCL8 included in paper I and III. (Commins, Borish et al. 2010, Akdis, Burgler et al. 2011, Ernerudh 2012).

Cytokine Main producers Important functions Important cell targets IL-1β Monocytes/macrophages, Pro-inflammatory, induction of T cells, endothelial cells neutrophils, glia cells, adaptive immunity and epithelial cells. keratinocytes and numerous other Facilitates proliferation, cells. differentiation, and function of both innate and adaptive immune system cells, in particular T cells. Upregulation of adhesion molecules like integrins and selectins. Endogenous pyrogen. IL-2 Activated CD4+ and CD8+ T Regulation of adaptive CD4+ and CD8+ cells. immunity T cells, NK cells and B Proliferation of effector T and B cells. cells, development of Treg cells, differentiation and proliferation of NK cells and growth factor for B cells. IL-4 Th2 cells, basophils, eosinophils, Induction and regulation of T cells and B cells. mast cells and NKT cells. adaptive immunity Induction of Th2 differentiation, IgE class switch and upregulation of MHC II expression on B cells. IL-5 Th2 cells, eosinophils, mast cells, Regulation of adaptive Eosinophils, basophils, and NK cells and NKT cells. immunity Facilitates chemotactic mast cells. activity and adhesion capacity of eosinophils. IL-6 Monocytes/macrophages, Pro-inflammatory, induction of Hepatocytes, leukocytes, T endothelial cells, T cells, B cells adaptive immunity cells, B cells and and numerous other cells. T cell activation and hematopoietic cells. differentiation, induction of Th17 differentiation, B cell differentiation and inducer of acute phase proteins. IL-10 Monocytes/macrophages, Immunosuppressive Monocytes/macrophages T dendritic cells, T cells and B Downregulation of MHC II and cells, B cells, NK cells, cells. co-stimulatory molecules on mast cells, dendritic cells APCs, inhibition of production of and granulocytes. pro-inflammatory cytokines. TNF Mononcytes/macrophages, T Pro-inflammatory Endothelial cells and cells, B cells, NK cells, Induces anti-tumor immunity, granulocytes. endothelial cells, mast cells and facilitates granulocyte function neutrophils. and chemotaxis and promotes vascular leakage. IFN-γ Th1 cells, cytotoxic T cells and Induction and regulation of Monocytes/macrophages, NK cells. adaptive immunity dendritic cells, NK cells, Induces cell-mediated immunity granulocytes, endothelial by stimulating antigen cells, T cells and B cells. presentation and cytokine production, promotes macrophage mobilization and function. GM-CSF T cells, fibroblasts, epithelial Growth and differentiation Eosinophils, neutrophils cells, endothelial cells, Supports maturation of dendritic dendritic cells, monocytes/macrophages, mast cells, neutrophils and monocytes/macrophages, cells, eosinophils and neutrophils. macrophages, stimulates platelet platelets and erythrocytes. and erythrocyte production, activation of granulocytes and monocytes. CXCL8 (IL-8) Monocytes/macrophages, Chemoattractant Neutrophils, T cells, NK endothelial and epithelial cells, T Potent chemoattractant for cells, basophils, eosinophils cells, granulocytes and numerous neutrophils, NK cells, T cells and and endothelial cells. other cells. mobilization of hematopoietic stem cells.

2. Background

Chemokines

Chemokines are described as “chemotactic cytokines” and constitue a large protein family responsible for the trafficking of leukocytes to sites of inflammation in a concentration- dependent fashion (Commins, Borish et al. 2010). Chemokines are also involved in several other processes including the regulation of leukocyte maturation (Baggiolini 1998, Ubogu, Cossoy et al. 2006, Commins, Borish et al. 2010). Under normal conditions the CNS is believed to be a zone of relative immune privilege, which means a relative lack of immune surveillance. This implicates that although there are many immune competent cells in the CNS like microglia, there are few patrolling immune cells recruited from the circulation. To establish a local immune response, the recruitment of T cells to the site is therefore crucial and likely a cornerstone in MS pahogenesis. Accordingly, CXCL8 (previously denoted IL-8) as well as CXCL9 and CXCL10, the latter two denoted as Th1-associated chemokines, have all shown to be elevated in CSF during relapse (Sorensen, Tani et al. 1999, Bartosik-Psujek and Stelmasiak 2005). In contrast, CCL17 and CCL22 are associated with Th2 immune responses, and are highly expressed in Th2-related diseases like airway hypersensitivity and atopic dermatitis (Fujisawa, Fujisawa et al. 2002, Yamashita and Kuroda 2002, Hijnen, De Bruin- Weller et al. 2004).

2.8 Leukocyte adhesion and migration

The recruitment of pro-inflammatory cells to the CNS is dependent on the sequential interactions of different adhesion and signalling molecules on leukocytes and endothelial cells lining the BBB. The process can principally be divided into tethering, rolling, activation and arrest followed by diapedesis (Luster, Alon et al. 2005). Initial tethering and subsequent rolling of cells along the endothelium is mediated by E-, P-, and L-selectins or α4-integrins. E- and P- selectins are expressed by the endothelium and interacts with glycoproteins, glycolipids, L- selectins and α4-integrins expressed on the leukocyte. If the rolling leukocyte at this step encounters additional signals from chemoattractive molecules like chemokines, there is an activation of intracellular signalling pathways through G-protein–coupled receptors causing the rapid activation of integrins. Integrins are heterodimeric leukocyte surface proteins composed of an α- and β-popypeptide chain. The most important integrins for leukocyte trafficking belong to the β2 subfamily (leukocyte function-associated antigen 1, LFA-1) and the α4- integrins α4β1 (very late activation antigen 4, VLA-4) and α4β7 (Luster, Alon et al. 2005). The integrins α4β1 and α4β7 bind to their corresponding molecules on endothelial cells, vascular cell adhesion molecule 1 (VCAM-1) and mucosal addressin cell adhesion molecule 1 (MAdCAM-1), respectively. The activation of a leukocyte by chemoattractants thus upregulates its expression of these integrins, promoting the arrest and firm adhesion of the leukocyte to the endothelium. In the context of CNS inflammation, the ensuing diapedesis into CNS tissue is facilitated by inflammation-caused local impairment of the BBB. This impairment of the BBB implicates an increase in permeability with alteration of tight junction structures altogether promoting the influx of pro-inflammatory leukocytes (Alvarez, Cayrol et al. 2011). 28

2. Background

Inflammatory signals like cytokines and chemokines have a major role in cell-adhesion and migration processes since they induce expression of selectins and integrins on leukocytes and endothelial cells. To exemplify, VCAM-1 is normally not detectable in the brain parenchyma, however upon stimulation with IFN-γ, IL-1β and TNF, astrocytes express VCAM-1 (Rosenman, Shrikant et al. 1995) and VCAM-1 can also be expressed by glial cells near sites of MS lesions (Cannella and Raine 1995). Furthermore, peripheral lymphocyte expression of adhesion molecules like VLA-4 and LFA-1 was associated with the development of MS lesions as measured by T2 lesion load on MRI (Eikelenboom, Killestein et al. 2005).

2.9 Inflammation versus neurodegeneration

MS is traditionally considered an inflammatory disease with attacks on the myelin sheath by auto-reactive T cells causing demyelination and axonal damage. Evidence for the association between demyelination and axonal damage has come from histopathological studies showing that transected axons are common in MS lesions and that the density of transected axons correlated with the inflammatory activity in the lesion (Trapp, Peterson et al. 1998). Although patients use to recover clinically from demyelinating exacerbations, there is shown to be irreversible subclinical axonal damage also at early stages of the disease (Dutta and Trapp 2007). It is presumed that remyelination in combination with CNS plasticity accounts for the ability to mask this progressive axonal damage early in the disease course. However, when the accumulating axonal loss eventually reaches a threshold for the ability of the brain to compensate, the result is irreversible neurologic disability. Clinically this point is characterized by the transition to SPMS. MS treatment strategy has been focusing on the decrease of inflammation and thereby indirectly also possibly interfering with the subsequent neurodegenerative process. All available disease-modulating therapies for MS today are immunomodulators and there is no specific treatment once the patient has reached the secondary progressive phase of the disease. Consequently, in a recent systematic review of treatment with IFNβ in SPMS, the conclusion was that treatment did not delay permanent disability, although relapse risk was reduced (La Mantia, Vacchi et al. 2012).

However, the view that focal demyelinating inflammation drives the neurodegenerative process in MS has been challenged by evidence that axonal damage, reflected in brain atrophy progression, may occur independently of focal inflammatory lesions (Miller, Barkhof et al. 2002). In particular by means of new magnetic resonance imaging (MRI) techniques, several studies have shown that the disease process in MS is ongoing rather than episodic, and have emphasized the importance of neuronal and axonal pathology (De Stefano, Bartolozzi et al. 2005) (Filippi, Rocca et al. 2012). Wallerian degeneration has been suggested to partly explain the discrepancies between axonal damage and focal inflammation by providing a mechanism for axonal loss in non-lesional white matter distal to the lesion. This notion is also consistent with studies on axonal projections in corpus callosum (Evangelou, Konz et al. 2000) and the corticospinal tract (Simon, Kinkel et al. 2000). However, the relation between inflammation and neurodegeneration in MS is unclear and yet to be defined.

2. Background

Figure 4. Cerebral MRI (FLAIR protocol) showing a sagittal (left) and axial (right) section with periventricular white matter lesions and cortical atrophy in MS (from CMIV).

2.10 Measuring diffuse pathology in MS

MRI is indisputable one of the most important paraclinical tools not only when diagnosing MS but also in assessing disease progression and evaluating treatment effects (Figure 4). MRI in combination with pathological studies has indeed expanded the way we understand this disase, i. e. from a focal demyelinating disease of white matter to a disease with both focal and diffuse pathology and involving both white and grey matter (Filippi, Rocca et al. 2012). Still, conventional MRI (cMRI) only provides limited information about the degree of inflammation and remyelination as well as data about diffuse white and grey matter pathology (Zivadinov, Stosic et al. 2008). In this perspective novel MRI-techniques like proton magnetic resonance spectroscopy (1H-MRS) and quantitative MRI (qMRI) can give valuable additional information. 1H-MRS is a technique that measures signals originating from protons in organic molecules in contrast to signals derived from protons in water (cMRI technique), thereby allowing for non-invasive quantification of specific metabolites in cerebral tissues (Matthews, Francis et al. 1991, Sajja, Wolinsky et al. 2009). 1H-MRS has provided data about pathologic findings in what appears to be normal white (NAWM) and gray (NAGM) matter on cMRI (He, Inglese et al. 2005). In 1H-MRS specific neurometabolites are represented by different resonance peaks like N-acetylaspartate (NAA), creatine (Cr), myo-Inositol (mIns), choline (Cho) and glutamate (Glu) (Sajja, Wolinsky et al. 2009). NAA, an aminoacid produced and localized primarily in neurons, is considered to be a marker of axonal integrity and neuronal viability. The resonance from Cr derives from creatine and phosphocreatine, which are important in the energy metabolism. Since Cr is more abundant in glia cells (astrocytes and oligodendrocytes) compared to neurons, elevated levels are an indicator of gliosis. mIns is also 30

2. Background considered a glia cell marker whereas Cho indicates myelin turnover thus reflecting the process of de- and remyelination. Glutamate is the principal excitatory neurotransmitter of the CNS but has also been ascribed neurotoxic properties when at excessive amounts (Matute, Domercq et al. 2006).

As a complement to 1H-MRS data on neurodegeneration in MS, levels of neurofilament light protein (NFL) and glial fibrillary protein (GFAP) in CSF might be used (Giovannoni 2006). NFL is, together with neurofilament intermediate chain and neurofilament heavy chain, the major component of axonal cytoskeleton proteins. Axonal damage causes the release of neurofilament proteins in the extracellular fluid and may accordingly be assessed in CSF (Teunissen and Khalil 2012). CSF levels of NFL have been shown elevated in MS patients, in particular during acute relapses (Malmestrom, Haghighi et al. 2003, Norgren, Sundstrom et al. 2004, Teunissen, Iacobaeus et al. 2009), and in patients with CIS (Teunissen, Iacobaeus et al. 2009). NFL levels in CSF were significantly reduced in MS patients by treatment with natalizumab (Gunnarsson, Malmestrom et al. 2010). GFAP constitutes the major intermediate filament cytoskeletal protein of astrocytes and is considered a marker of astrocytes and astrogliosis (Middeldorp and Hol 2011). GFAP was found in a large MS plaque with fibrous astrocytes and demyelinated axons (Eng, Vanderhaeghen et al. 1971), and is increasingly expressed during CNS degeneration and aging. GFAP levels in CSF have been shown to correlate with disability and disease progression in MS (Malmestrom, Haghighi et al. 2003, Axelsson, Malmestrom et al. 2011).

2.11 Disease-modifying treatments

All approved disease-modifying therapies for relapsing MS are immunomodulators. The first- line treatment consists of interferon-β (IFNβ-1a or IFNβ-1b) and glatiramer acetate.

IFNβ is given as a subcutaneous injection every two or three days or as an intramuscular injection once a week. IFNβ treatment has been shown to reduce relapse rate with about 30% and reduce disease activity as measured by a reduction in new MRI lesions compared to placebo (1995, Jacobs, Cookfair et al. 1996, 1998). The precise effector mechanism is not fully defined, although inhibition of auto-reactive T cells, induction of Treg cells and interference with leukocyte migration and cytokine modulation has been suggested (Dhib-Jalbut and Marks 2010).

Glatiramer acetate (GA) is given once a day as a subcutaneous injection. GA treatment was shown to reduce relapse rate by 29% (Johnson, Brooks et al. 1995). As for IFNβ, the effector mechanisms of GA are uncertain, however studies on EAE suggest that GA may generate suppressor cells, inducing tolerance, expanding Treg populations and alter APC functions (Racke, Lovett-Racke et al. 2010).

The second-line treatment for MS includes natalizumab and fingolimod. Fingolimod (FTY720) is the first oral disease-modifying treatment available for relapsing MS and is a sphingosine-1- 31

2. Background phosphate (S1P) receptor agonist. S1P signalling mediate leukocyte trafficking in lymph nodes and treatment with S1P has been shown to reversibly sequester naïve and activated central memory T cells in the lymph nodes, whereas effector memory T cells remain in the peripheral circulation (Mehling, Brinkmann et al. 2008). Treatment with fingolimod has been shown to reduce relapse rate with 54% compared to placebo, and a significant decline in MRI lesions has also been demonstrated (Kappos, Radue et al. 2010).

Natalizumab, which is of particular interest in this thesis, is a humanized monoclonal antibody directed against the α4-chain (CD49d) of VLA-4 (α4β1) and α4β7. Humanized antibodies have major advantages, since they are less immunogenic, which increases the in vivo life time and allows for repeated administrations. Natalizumab antibodies are of the IgG4 subclass, which facilitates their persistence in the circulation and eliminates the risk of complement activation (Rice, Hartung et al. 2005). The α4-integrin is expressed on different types of leukocytes as T and B cells, monocytes, eosinophilic and basophilic granulocytes, albeit neutrophil granulocytes are excepted. The α4-chain heterodimerizes with β1 and β7 subunits and thereby form a functional molecule (Abbas A.K. 2012). VLA-4 (α4β1) adheres toVCAM-1 on vascular endothelium and α4β7 interacts with MAdCAM present on gut endothelium. The effect of natalizumab in Crohn´s disease is considered mediated by the blocking effect on α4β7 (Guagnozzi and Caprilli 2008).

Evidence for the crucial role of α4-integrins in leukocute activation, migration and differentiation originate from both in vitro and animal studies (Lobb and Hemler 1994). The first study to demonstrate prevention of EAE by treatment with α4β1-antibodies came 1992 (Yednock, Cannon et al. 1992). Natalizumab-binding to VLA-4 blocks the attachment of leukocytes to brain endothelial cells thereby preventing their entrance into the CNS. In a randomized placebo-controlled trial, natalizumab reduced the relapse rate with about 68% over one year and simultaneously reduced the accumulation of new or enlarging hyperintense lesions as measured by T2-weighted MRI (Polman, O'Connor et al. 2006). Thus, natalizumab is proven more efficient than the first-line treatments with IFNβ and GA. Apart from reduced entrance of inflammatory cells to the CNS, other immunological effects are likely to occur that may contribute to treatment effects. Principally, effects of α4-integrin chain antibody blockade could be referred either to the mere blocking of VLA-4 adhesion to its ligands or to natalizumab acting agonistically as a ligand on the VLA-4 complex (Gonzalez-Amaro, Mittelbrunn et al. 2005). The latter could generate intracellular signals that may affect the future behavior of the target cell. It is assumed that the major part of in vivo treatment effects of α4-integrin chain blockade are related to the former, i. e. reduced extravasation of pro- inflammatory leukocytes. This is also indirectly supported by studies showing an increase in the expression of pro-inflammatory lymphocytes in the peripheral circulation (Kivisakk, Healy et al. 2009, Frisullo, Iorio et al. 2011) and data showing reduced leukocyte counts in CSF (Stuve, Marra et al. 2006, Khademi, Bornsen et al. 2009) and cerebral perivascular spaces (del Pilar Martin, Cravens et al. 2008) after natalizumab treatment. However, there are many reasons to suspect additional effects of α4-integrin chain blockade apart from interference with cell trafficking. To exemplify, VLA-4 is involved in the development and differentiation of many cell types and participates in the induction of the immune response by facilitating APC and T cell adhesion during antigen presentation (Gonzalez-Amaro, Mittelbrunn et al. 2005). 32

2. Background

Accordingly, ligand-binding to VLA-4 is shown to be a co-stimulatory signal for T cells (Sato, Tachibana et al. 1995, Niino, Bodner et al. 2006) and VLA-4 expression is up-regulated during maturation of dendritic cells (Puig-Kroger, Sanz-Rodriguez et al. 2000). Further evidence of additional effects of VLA-4 blockade is alteration in gene expressions of peripheral blood cells (Lindberg, Achtnichts et al. 2008) and changes in PBMC population composition and activation (Skarica, Eckstein et al. 2011) in natalizumab treated MS patients, as well as the intriguing finding that treatment with a small α4-integrin inhibitor (CDP323) in a phase II study of relapsing MS failed to show efficacy (Harrer, Wipfler et al. 2011).

2.12 Progressive multifocal leukoencephalopathy (PML)

Exploring the full range of natalizumab treatment effects has become particularly important considering the association to development of progressive multifocal leukoencephalopathy (PML). PML is an opportunistic brain infection caused by JC polyomavirus (White and Khalili 2011). PML causes diverse neurological symptoms depending on the localization and extent of the infection. Visual deficits, cognitive impairment and motor weakness have been reported, and the disease course may be fatal (Tan and Koralnik 2010). In a survey study of cases of PML associated to natalizumab treatment and reported up to February 2012, there were 212 confirmed cases of PML among 99,571 patients (2.1 cases per 1000 patients) (Bloomgren, Richman et al. 2012). The risk of developing PML was related to positive status with respect to anti-JC virus antibodies, prior immunosuppressive treatment and increased duration of natalizumab treatment (Bloomgren, Richman et al. 2012). Considering the proposed action of natalizumab on cell trafficking, it is suggested that impaired immune surveillance in the CNS may explain the increased risk of PML (Stuve, Marra et al. 2006). Still, additional effects of VLA-4 blockade may contribute to the risk of PML, emphasizing the importance for further studies on these mechanisms.

3. Aims

The general aim of this thesis was to elucidate immunological mechanisms present in multiple sclerosis and the immunomodulatory effects of natalizumab treatment, focusing on lymphocytes, inflammatory markers and 1H-MRS.

The specific aims of each paper were:

I To assess the effects of one-year natalizumab treatment on cytokine and chemokine protein levels in plasma and cerebrospinal fluid of MS patients.

II To explore changes in blood lymphocyte composition and functional T cell responses in MS patients after one-year natalizumab treatment.

III To detect changes in NAWM metabolite concentrations in MS patients after one-year natalizumab treatment, and to assess whether changes were associated with intrathecal inflammation or neurodegeneration as measured by biomarkers in the CSF.

IV To evaluate the balance between circulating Th cell subsets in MS patients using the expression of the CD4+ T cell related transcription factors in whole blood.

4. Methods

4.1 Study subjects

All patients in this thesis were included consecutively from the Department of Neurology at the University Hospital of Linköping, Sweden. All were diagnosed with MS according to the McDonald criteria from 2001 (McDonald, Compston et al. 2001). In total 72 patients were included. Table 2 shows the distribution of patients in paper I-IV. The study design in paper I, II and III was observational and not interventional, since there was no randomization to a prospectively followed placebo-treated group. When treated with natalizumab, patients were given intravenously infusions (300 mg) once a month.

Paper I Natalizumab treatment was initiated in 31 patients, 18 men and 13 women (median age 35 years, range 23-50), because of unsatisfactory effect or intolerable adverse effects on treatment with first-line treatments. Patients were diagnosed as either RRMS (n=23) or RRMS with secondary progression (RRMS/SP) (n=8). RRMS/SP was defined as a clinically progressive MS though with presumed inflammatory activity because of superimposed relapses and/or disease progression measured by MRI. Median EDSS at inclusion was 3.5 (range 0-8). Within the last three months before inclusion 26 patients were on immunomodulatory treatment (17 IFN-β, 6 GA, 2 mitoxantrone, 1 IvIg) and 5 patients were not on any treatment. CSF from 10 patients with non-inflammatory neurological diseases was used as a control for baseline CSF analyses, and blood from 15 healthy donors was used as control for baseline blood analyses.

Paper II Natalizumab treatment was initiated in 40 patients, 22 men and 18 women (median age 37 years, range 22-62), with active MS. Initiation of treatment was based on clinical and MRI parameters, implicating an active relapsing disease. EDSS at inclusion was 2.5 (range 0-7). Within the last three months before inclusion 31 patients were on immunomodulatory treatment (26 IFN-β, 4 GA, 1 IvIg) and 9 patients were not.

Paper III Natalizumab treatment was initiated in 27 patients, 14 men and 13 women (median age 40 years, range 22-62), with active MS. Initiation of treatment was based on clinical and MRI parameters, implicating an active relapsing disease. EDSS at inclusion was 2.5 (range 0-7). Within the last three months before inclusion 24 patients were on immunomodulatory treatment (20 IFN-β, 4 GA) and 3 patients were not. For baseline comparisons of 1H-MRS variables, 20 healthy volunteers (median age 48 years, range 27-72) were used as controls.

4. Methods

Paper IV 33 patients, 18 men and 15 women (median age 35 years, 25th-75th percentile 30-43), with RRMS were included. Median EDSS at inclusion was 2.0 and within the last three months before inclusion 12 patients were on treatment with IFN-β and 23 patients were untreated. 20 age-and sex-matched healthy volunteers (median age 46 years, range 21-62) were used as controls. All participants (patients and controls) were checked for ongoing infection and/or other inflammatory conditions at the time of inclusion in order to minimize potential confounding factors. None of the participants were excluded on these grounds.

Table 2. Distribution of patients in paper I-IV. Total no. of patients=72.

Year of birth Sex I II III IV Year of birth Sex I II III IV 1987 f x 1970 f x x 1986 f x x 1970 f x 1983 f x 1970 f x x x 1983 f x 1968 f x 1982 m x x 1968 m x x x 1982 m x x x x 1967 f x x 1982 m x x x 1967 f x 1981 m x 1967 f x 1980 f x 1967 f x 1980 f x 1967 m x 1979 f x 1967 m x x x x 1979 f x 1966 f x 1979 m x x 1966 m x x 1978 m x 1965 f x 1978 m x 1965 m x x x x 1978 m x 1965 f x 1976 m x x x x 1965 f x 1976 m x x x 1965 f x x 1976 f x 1965 f x 1975 m x x x 1964 m x x x 1975 m x x x 1964 m x 1975 f x x x 1963 m x x x x 1975 f x 1963 f x 1974 m x 1963 f x x 1974 m x 1963 m x x 1974 f x 1962 m x x 1974 f x 1962 f x x x 1973 m x 1961 m x x x 1973 f x x x 1960 f x x 1973 f x 1959 m x 1973 m x x x 1959 f x 1973 f x x 1958 f x x 1972 m x x x x 1957 f x x 1971 m x x 1955 f x 1971 f x 1945 m x x 1970 f x x 1937 f x

Year of birth and sex are given for each patient. m=male, f=female, x mark participation in respective study.

4. Methods

4.2 Procedures

4.2.1 Disease scoring systems (paper I, II, III, IV)

Medical records and the Swedish MS registry were used in order to evaluate disease course and earlier immunomodulatory treatments. In paper I-III, all patients were examined by a neurologist (CD or MV) or resident in neurology (JM) at inclusion and after one year of treatment for the definition of Expanded Disability Status Scale (EDSS) score and subsequent Multiple Sclerosis Severity Score (MSSS). In paper IV, EDSS was defined once (CD, MV or JM). For paired comparisons the examination was carried out by the same clinician pre and post one-year treatment (paper I, II and III).

EDSS is widely used as a measure of disability in MS and dates back to 1955, but was revised to present form 1983 (Kurtzke 1983). It is composed of eight Functional Systems (FS) and allows neurologists to assign a Functional System Score (FSS) in each of these. The eight FS are pyramidal function, cerebellar function, brainstem function, sensory function, bowel and bladder function, visual function, higher cerebral function and other deficits. Other deficits can include paroxysmal symptoms. The sum of all FSS is then referred to the EDSS scale to define an EDSS score (Table 3). EDSS 0-4.5 refers to fully ambulatory patients and requires a thorough neurological examination, whereas EDSS 5.0-9.0 are defined as the impairment to ambulation.

MSSS was introduced in 2005 (Roxburgh, Seaman et al. 2005). It combines disability, as defined by EDSS, with disease duration and thus is a measure of the disease progression rate.

The Multiple Sclerosis Impact Scale 29 (MSIS-29) was introduced in 2001 and consists of a questionnaire about the patients’ view on the physical and psychological impact of MS within the last two weeks (Hobart, Lamping et al. 2001). There are 12 questions related to physical impairment and 17 related to psychological impairment. Each question constitutes a statement and can be answered with four alternatives rendering different scores (1=not at all, 2=a little, 3=moderately, 4=quite a bit, 5=extremely). The sum of the questionnaire-score constitutes the MSIS-29 score. MSIS-29 questionnaires were distributed to patients in connection to natalizumab infusions.

Symbol Digit Modalities Test (SDMT) is a quick screening test for cognitive dysfunction (Smith 1991). The procedure is that the patient, by using a reference key, has 90 seconds to pair specific numbers with a given geometric figure. Answers can be oral or written. SDMT was conducted by a nurse in connection to natalizumab infusions.

4. Methods

Table 3. The Kurtzke Expanded Disability Status Scale with description of EDSS scores.

0.0 Normal neurological examination. 1.0 No disability, minimal signs in one FS. 1.5 No disability, minimal signs in more than one FS. 2.0 Minimal disability in one FS. 2.5 Mild disability in one FS or minimal disability in two FS.

3.0 Moderate disability in one FS, or mild disability in three or four FS. Fully ambulatory. 3.5 Fully ambulatory but with moderate disability in one FS and more than minimal disability in several others. 4.0 Fully ambulatory without aid, self-sufficient, up and about some 12 hours a day despite relatively severe disability; able to walk without aid or rest some 500 meters. 4.5 Fully ambulatory without aid, up and about much of the day, able to work a full day, may otherwise have some limitation of full activity or require minimal assistance; characterized by relatively severe disability; able to walk without aid or rest some 300 meters. 5.0 Ambulatory without aid or rest for about 200 meters; disability severe enough to impair full daily activities (work a full day without special provisions). 5.5 Ambulatory without aid or rest for about 100 meters; disability severe enough to preclude full daily activities. 6.0 Intermittent or unilateral constant assistance (cane, crutch, brace) required to walk about 100 meters with or without resting. 6.5 Constant bilateral assistance (canes, crutches, braces) required to walk about 20 meters without resting. 7.0 Unable to walk beyond approximately five meters even with aid, essentially restricted to wheelchair; wheels self in standard wheelchair and transfers alone; up and about in wheelchair some 12 hours a day. 7.5 Unable to take more than a few steps; restricted to wheelchair; may need aid in transfer; wheels self but cannot carry on in standard wheelchair a full day; may require motorized wheelchair. 8.0 Essentially restricted to bed or chair or perambulated in wheelchair, but may be out of bed itself much of the day; retains many self-care functions; generally has effective use of arms. 8.5 Essentially restricted to bed much of day; has some effective use of arms retains some self-care functions. 9.0 Confined to bed; can still communicate and eat. 9.5 Totally helpless bed patient; unable to communicate effectively or eat/swallow. 10.0 Death due to MS.

4.2.2 Routine CSF analyses (paper I, II, III) Analyses of CSF total leukocyte counts, IgG index, albumin ratio and the investigation of oligoclonal bands were performed as routine analyses at the Department of Clinical Chemistry at the University Hospital of Linköping. IgG index is a measure of intrathecal immunoglobulin synthesis and constitutes, together with the presence of oligoclonal bands, the CSF hallmarks of an intrathecal inflammatory process as in MS. IgG index compensates for levels of IgG in serum and effects on the BBB (Figure 5 a). Albumin ratio is a measure of BBB integrity (Figure 5 b). An increase in albumin ratio is a non-specific sign of damage to the BBB and

4. Methods could thus be caused by inflammatory processes as well as by infections, tumors, stroke and trauma.

a: IgG index = [CSF-IgG (mg/l) / serum-IgG (g/l)] / [CSF-albumin (mg/l) / serum-albumin (g/l)]

b: Albumin ratio = [CSF-albumin (mg/l) / serum-albumin (g/l)]

Figure 5 a-b. How to calculate IgG index and albumin ratio.

4.2.3 Multiplex bead assay (paper I, III) For assessment of cytokine (IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8 (CXCL8), IL-10, IFN-γ, TNF and GM-CSF) concentrations in plasma and CSF, an ultra-sensitive 10-plex kit was used in accordance with the instructions provided (Invitrogen Cytokine Ultrasensitive Human 10- plex Panel). For assessment of chemokine (CXCL9, CKCL10, CXCL11, CCLL17 and CCL22) concentrations in plasma and CSF, an in-house assay was used as described in paper I. Both the assessment of cytokines and chemokines included the use of a multiplex bead assay (Figure 6). Briefly, polystyrene microspheres (beads) are internally dyed with two different fluorochromes in combinations, creating up to 100 different microspheres, all with a unique spectral signature (Vignali 2000). A capture antibody to the analyte of interest is then coupled to the microsphere, thereby obtaining a link between a specific analyte (in this thesis a cytokine or chemokine) and a spectral-specific bead. Several bead sets may be mixed, enabling the assessment of many analytes in a single sample. The beadmix with different captured analytes is then incubated with detection antibodies labeled with a reporter fluorochrome. Using flow cytometry, each bead is then characterized one at a time with respect to bead type and fluorescence intensity. The mean fluorescence intensity (MFI) is proportional to the amount of bound analyte, and by using a standard curve of defined concentrations of the analyte, the concentration of each analyte in the sample could be determined.

4. Methods

Figure 6. Principle of a multiplex bead assay.

4.2.4 Flow cytometry (paper II, IV) Flow cytometry makes it possible to detect, sort and count different cell populations on the basis of their properties related to size, granularity and characteristic cell-surface proteins. The cells are first incubated with fluorescent-labeled monoclonal antibodies against the cell-surface (or intra-cellular) molecules of interest. In the flow cytometer, labeled cells then passes one at a time through a laser beam causing scattering of the light and excitation of the attached fluorochrome, if any, to a higher energy state. The excitation of the fluorochrome causes emission of a certain wavelength that then can be detected. According to the specific scattering profile of each cell, its volume or size (forward scatter, FSC) and inner complexity including cytoplasmatic granules (side scatter, SSC), can be determined. In addition, based on the fluorescence emission characteristics of each cell the expressions of cell-surface molecules of interest are assessed.

Briefly, whole blood was drawn from EDTA tubes. The cells were labeled with antibodies for 15 min at room temperature in the dark. Antibodies used were (paper II) anti-CD3, anti-CD45, anti-CD4, anti-CD8, anti-CD16/56, anti-CD19, anti-CD28, anti-CD56, anti-CD57, anti-HLA- DR, anti-CD69, anti-CD25, anti-OX40L, anti-CD5 and anti-CD27 and (paper IV) anti-CD4, anti-CD3, anti-CD25. All antibodies were purchased from BD Biosciences (San José, CA, USA). For analysis of absolute cell numbers (cells/μl) in lymphocyte populations, TruecountTM

4. Methods tubes (BD Biosciences) were used. Removal of erythrocytes was done using FACS Lysing Solution (BD Biosciences) followed by another 15 min incubation in the dark. Cells were collected and analyzed using the FACS Canto II system (BD Biosciences). The initial lymphocyte gate was based on FSC and SSC properties. The principle for gating was then to define discrete populations or to use negative populations to define cut-off for continuously expressed markers as described in paper II.

4.2.5 Lymphocyte activation assay (paper II) The Flow-cytometric Assay for Specific Cell-mediated Immune-response in Activated whole blood (FASCIA) (Svahn, Linde et al. 2003) was used with some modifications to evaluate lymphocyte function. This method is based on determination of induced lymphoblast numbers after in vitro stimulation. Briefly, diluted peripheral whole blood cultures were stimulated with influenza antigen 1:1000 (Vaxigrip; Sanofi Pasteur, Solna, Sweden), PPD 10 µg/mL (SSI, Copenhagen, Denmark), CMV antigen 0.125 µg (BD Biosciences), tetanus toxin 5.7 Lf/mL (SSI), phytohaemagglutinin (PHA) 5 µg/mL (Sigma Aldrich), pokeweed mitogen (PWM) 10 µg/mL (Sigma Aldrich) or myelin basic protein (MBP) 100 µg/mL (Sigma Aldrich) diluted in cell-culturing media. Culturing ensued for seven days at 37°C with 5% CO2, after which cells were harvested and labeled with anti-CD3, anti-CD4, anti-CD8, and anti-CD108. After labeling, erythrocytes were lysed by incubating cells with 0.8% NH4Cl solution for 15 min at room temperature in the dark. Collection and analysis of data were performed using the FACS Canto II system. TruecountTM tubes were used separately for assessment of absolute cell numbers. Negative controls, cultured without antigen, were set up separately.

Initially, a lymphoblast gate was set based on FSC and SSC properties. A negative gate, gating non-activated lymphocytes, was set using the unstimulated cultures, and this gate was used to distinguish activated from resting lymphocytes. The cells were then further gated to CD3+CD4+ and CD3+CD8+ gates, respectively. From these populations, CD108+ cells were selected as a marker of cytotoxic T cells. The number of cells in the lymphoblast gate was compared between patients for the respective population and stimulus. For each stimulus, the mean number of lymphoblasts in unstimulated cultures was subtracted, thereby compensating for baseline activation of cells.

4.2.6 Markers of neurodegeneration in CSF (paper III) Assessment of CSF myelin basic protein (MBP), total tau (T-tau), tau phosphorylated at threonine 181 (P-tau), neurofilament light protein (NFL) and glial fibrillary acidic protein (GFAP) was performed at the Clinical Neurochemistry Laboratory at The Sahlgrenska Academy, Gothenburg, Sweden. Briefly, MBP was analyzed using a sandwich ELISA (Active MBP ELISA, Diagnostic Systems Laboratories Inc., Webster, TX, USA). T-tau and P-tau were determined using the Luminex xMAP technology and the INNOBIA AlzBio3 kit (Innogenetics Zwijndrecht, Belgium) as described previously (Olsson, Vanderstichele et al. 2005). NFL and

4. Methods

GFAP were analyzed using ELISA methods described previously (Rosengren, Wikkelso et al. 1994, Rosengren, Karlsson et al. 1996).

4.2.7 Magnetic resonance spectroscopy – MRS (paper III) All body tissues have a high water content and thus also protons (H+ ions). If a magnetic field is applied to a tissue, the containing protons will align to the direction of this field. In the static magnetic field of a MRI scanner, radio frequency currencies are systematically applied to create electromagnetic fields that alter the alignment of the protons. When the electromagnetic field is changed, the protons will return to the equilibrium state (relaxation), causing the emission of a radio frequency signal that can be detected by the scanner. Since protons in different tissues and under different pathophysiological conditions have different relaxation properties, these variations are used to create high-resolution images of the investigated tissue.

Proton MRS (1H-MRS) is based on MRI methodology and measures signals originating from protons in organic molecules in contrast to signals derived from protons in water as in conventional MRI (cMRI). It is an important complement to cMRI because of the ability to assess the chemical situation of tissues in vivo, as measured by the concentration of specific metabolites. Accordingly, the use of MRS for the investigation of cerebral tissues in MS is feasible (Sajja, Wolinsky et al. 2009). In clinic routine MRS is mainly used to differentiate between inflammatory and malignant focal pathology. Metabolite data is collected from a volume (voxel) defined by cMRI (Figure 7 A). The results are presented as a MRS-spectrum where each resonance peak represents a specific metabolite (Figure 7 B).

All MRS examinations as well as the calculation and evaluation of spectral data were performed by Anders Tisell at the Center for Medical Image Science and Visualization (CMIV), Linköping. Briefly, examinations were performed using an Achieva 1.5 T MR scanner (Philips, Best, The Netherlands), and an eight channels SENSE head coil. A T1 weighted axial, and a FLAIR sagittal, and a T2 weighted coronal image volumes were acquired for placing the

MRS voxels. For quantification of R2 (1/T2) an axial ‘Multi Echo GRAdient echo and Spin Echo’ (ME-GRASE) pulse sequence was used to acquire data of the tissue water signal. MRS data was acquired using ‘Point RESolved Spectroscopy’ (PRESS), with an echo time of 30 ms and a repetition time of 3.0 s, 128 water suppressed spectra were averaged. In addition, eight spectra without water suppression were obtained as an internal reference standard. The MRS voxel was placed in periventricular NAWM (Figure 7 A). Lesions were avoided by freely rotating the voxel and adjusting the size (ranging 2-3 ml). The healthy controls (HC) were similarly examined using two MRS voxels (PRESS TR 30 ms, TR 3.0 s), which were placed bilaterally in parietal white matter.

4. Methods

Figure 7. 1H-MRS examination with description of metabolites (from Mellergård, Tisell et al, 2012).

A. Typical placement of a spectral voxel in normal appearing white matter.

B. Typical 1H-MR spectrum with LCModel fit and residual of a natalizumab-treated MS patient (male, 30 years old). Spectral assignments: 1, Cr-

CH2-; 2, mIns; 3, Cho (CH3)3;

4, Cr (CH3); 5, NAA; 6, Gln γ/β; 7, Glu γ/β; 8, NAAG- methyl; 9, NAA-methyl; 10,

Lac CH3.

C. Description of 1H-MRS metabolites, for reference see Dahlqvist O. (Dahlqvist 2010).

4. Methods

4.2.8 Absolute quantification of MRS data (paper III) MRS results are often expressed as ratios, with the most frequent denominator creatine (Cr), assuming that Cr levels are stable over time. However, this is not a recommended procedure as it has been shown that Cr is neither unaffected nor stable (Mader, Roser et al. 2000, Li, Wang et al. 2003). In paper III we therefore utilized absolute quantitative 1H-MRS to measure metabolite concentrations by using the internal water signal as an internal reference. The metabolite concentrations are thus estimated as absolute ‘aqueous fraction concentrations’ (mM aq.). The MRS spectra were analysed using LCModel (S. Provencher, Canada) (Provencher 1993), and the metabolites total creatine (tCr), total choline (tCho), myo-Inositol (mIns), total N-acetylaspartate (tNA), the sum of glutamate and glutamine (tGlx) and lactate (Lac) (Figure 7 C) were analyzed.

Since we only intended to measure NAWM, evaluation of possible plaque contamination in each voxel was done. The volume of plaque contamination in each MRS voxel was estimated by segmenting all plaques in the proximity of the MRS voxel. The segmentations were performed on T2w image by a neuroradiologist using MEVISLAB (MeVis Medical Solutions AG, Bremen, Germany). Of all investigated MRS voxels, only one contained more than 2 % of plaque. Statistical tests were therefore redone excluding this potential outlier, however, the observed significant correlations presented were then still valid.

4.2.9 Real-time RT-PCR (paper IV) Gene expression analyses were performed using real-time Taqman reverse transcriptase (RT)- PCR. Real-time Taqman RT-PCR enables the amplification and quantification of mRNA in the same procedure. Principally, a reporter dye (fluorescent) and a quencher dye are attached to a target probe that is complementary to the specific mRNA sequence of interest. During the PCR procedure the probe hybridizes with the cDNA template. The subsequent extension of template by Taqman polymerase results in the cleavage and release of both the reporter and quencher dye causing an increase in fluorescent intensity. Hence, the amplification plot reflects the increasing amount of reporter dye during amplification and is directly related to the formation of the PCR product of interest.

For the purpose of acquiring a situation with as much resemblance to the in vivo situation as possible, whole blood was used to investigate gene expression. Briefly, total RNA was extracted from whole blood PAXgene tubes (PreAnalytix, GmbH, Hombrechtikon, Switzerland) and then converted to complementary DNA (cDNA) using a random hexamer method (High Capacity cDNA Synthesis Kit, Applied Biosystems, Foster City, CA, USA). Before cDNA synthesis, all samples were diluted to a fixed concentration to standardize the performance of the real-time RT-PCR. mRNA expression was quantified using the Applied Biosystems 7500 Fast Real-Time PCR System (Applied Biosystems). Amplification of cDNA was performed with the TaqMan Universal PCR Master Mix (2X), No AmpErase UNG (Applied Biosystems). For quantification of cDNA a five-point serially four-fold diluted

4. Methods standard curve was developed from PBMC cultures stimulated with phytohaemagglutinin (PHA). The mRNA expression of T helper cell transcription factor TBX21 (Th1), GATA3 (Th2), RORC (Th17), FOXP3 (Treg) and EBI3 was standardized to 18S (human rRNA), and all results expressed as ratio in relation to 18S-expression. Primers and probes for TBX21, RORC and Ebi3 expression were analyzed with Taqman® Gene Expression assays (Applied Biosystems), while GATA3, FOXP3 and 18S probes were constructed in-house using Primer Express 3.0 (Applied Biosystems).

4.3 Statistical methods Statistical analyses on variables with assumed non-normal distribution were performed using non-parametric methods. Such variables included clinical data (clinical scoring systems), CSF routine analyses (cell count, IgG index and albumin ratio), levels of cytokines, chemokines, markers of neurodegeneration and gene expression of transcription factors. In contrast, flow cytometry and 1H-MRS data were analyzed using paired and independent samples t test. Comparisons between several groups (lymphocyte activation assay data in MS patients and controls and different treatment at baseline 1H-MRS) were performed with ANOVA and Tukey´s post-hoc test (paper II and III). Parametric bi-variate correlation analyses (Pearson) were used to evaluate flow cytometry data in relation to clinical data (paper II) and non- parametric bi-variate correlation analyses (Spearman) were used to evaluate 1H-MRS data in relation to CSF biomarkers (paper III). A p-value of <0.05 was considered statistically significant, except for flow cytometry data in paper II, where p<0.01 was considered statistically significant and p<0.05 was considered a tendency.

4.4 Ethical considerations All studies included in this thesis were approved by The Regional Ethics Committee in Linköping. Written (paper I and III) or informed (paper II and IV) consent was obtained from all patients.

5. Results and Discussion

5.1 Clinical outcome (paper I, II, III) The studies were observational, thus the purpose was not to evaluate clinical effects of treatment. The studies were open and treatment was initiated on clinical indications. Nevertheless, it was mandatory to carefully record the clinical outcome including tests as described in the Methods section. One-year treatment with natalizumab was associated with a clinical improvement as measured both by a reduction in relapse rate and improvement of clinical scoring systems. Annualized relapse rates at follow up were 0.35 (paper I), 0.10 (paper II) and 0.15 (paper III) which is in line with earlier results from a randomized placebo- controlled trial with a reduction of relapse to 0.26 after one year of natalizumab treatment (Polman, O'Connor et al. 2006). EDSS was stable at follow-up or improved though not significantly in all studies. MSSS was assessed in paper II and III and declined significantly upon treatment as did MSIS-29 (Table 4 a-b).

Table 4 a-b. Clinical scoring system outcome in paper II and III. a: Paper II, n=40

Baseline Follow-up p EDSS 2.5 (0-7.0) 2.5 (0-8.0) 0.08 MSSS 3.82 (0.19-8.55) 3.20 (0.17-9.20) <0.0005 MSIS-29 physical 2.18 (1.00-4.75) 1.40 (1.00-4.20) a <0.0005 psychological 2.11 (1.00-4.56) 1.44 (1.00-4.56) a <0.0005 SDMT 48 (5-66) 50 (11-65) b 0.03

b: Paper III, n=27

Baseline Follow-up p EDSS 2.5 (0-7) 2.0 (0-6.5) 0.007 MSSS 3.9 (0.2-8.9) 3.1 (0.2-7.9) <0.0005 MSIS-29 physical 2.2 (1.0-3.7) 1.5 (1.0-3.6) <0.0005 psychological 2.1 (1.0-4.2) 1.6 (1.0-3.8) <0.0005 SDMT 49 (5-69) 53 (3-74) 0.03

Median values are given and range within brackets. p refers to Wilcoxon signed rank test comparing baseline and follow-up. a n=37 and b n=38 because of lack of follow-up data. EDSS=Expanded Disability Status Scale, MSSS=Multiple Sclerosis Severity Score, MSIS-29=Multiple Sclerosis Impact Scale 29, SDMT=Symbol Digit Modalities Test.

5. Results and Discussion

5.2 CSF and plasma variables (paper I, II, III)

5.2.1 Routine CSF analyses (paper I, II, III) There were significant declines in CSF total white blood cell counts and IgG index after one year of natalizumab treatment (Table 5). This result is in accordance with the proposed main mode of action of natalizumab; the reduced migration of leukocytes into the CNS, and have been demonstrated earlier (Stuve, Marra et al. 2006, del Pilar Martin, Cravens et al. 2008, Khademi, Bornsen et al. 2009). In contrast, albumin ratios did not change significantly post- to pre-treatment, which may reflect that damage to the BBB is irreversible or take long time to recover from.

Table 5. CSF total white blood cell (wbc) count and IgG index changes after one year of natalizumab treatment.

CSF Paper Baseline Follow-up p wbc count (x106 cells/l) I 2.5 (0.0-31.0) 0.9 (0.1-3.6) <0.0005 II 2.6 (0.2-28.0)a 1.1 (0.0-4.0)b <0.0005 III 2.2 (0.2-31.0) 1.0 (0.0-3.6) 0.002 IgG index I 0.97 (0.56-4.1) 0.83 (0.49-3.8) <0.0005 II 0.92 (0.48-3.0)a 0.77 (0.45-2.4)b <0.0005 III 0.83 (0.48-4.06) 0.75 (0.45-3.84) 0.001

Median values are given and range within brackets. p refers to Wilcoxon signed rank test comparing baseline and follow-up. n=31 (paper I), n=25 (paper III), a n=38, b n=36.

5.2.2 Cytokines and chemokines in CSF and plasma – comparing multiple sclerosis with non-inflammatory neurological diseases (paper I) CSF levels of IL-1β, IL-8 (hereafter referred to as CXCL8), CXCL10, CXCL11 and CCL22, were shown to be significantly higher in natalizumab-naive MS patients compared with patients with non-inflammatory other neurological diseases (OND), whereas levels of GM-CSF, IFN-γ, TNF, IL-2, IL-4, IL-5, IL-10 and CCL17 were too low to be detected (Table 6). In plasma, levels of IFN-γ, IL-2, IL-5 and CXCL8 were significantly higher in natalizumb-naive MS patients compared with healthy blood donors (Table 6).

Although levels of several cytokines proved to be low and therefore difficult to detect, still indications of an intrathecal inflammation in MS was present in terms of consistent and significant higher levels of chemokines (CXCL8, CXCL10, CXCL11, CCL22) and IL-1β in CSF from MS patients compared with OND patients. Of note, levels of IL-1β were higher intrathecally than in plasma, possibly reflecting its important role in hypothalamic activity.

5. Results and Discussion

Table 6. Cytokine and chemokine concentrations (pg/ml) in plasma (top) and CSF (bottom) in healthy blood donors and patients with other neurological diseases (OND) compared with natalizumab-naïve MS patients (paper I).

Plasma No. of MS MS (n=31) No. of healthy Healthy blood patients with blood donors donors (n=15) measurable with levels measurable levels Cytokines IL-1β 31/31 0.2 (0.2-0.5) 15/15 0.2 (0.2-0.4) IL-2 31/31 172.5 (117.6-194.4) * 15/15 123.4 (72.9-150.1) IL-4 31/31 77.4 (51.4-136.5) 15/15 95.4 (43.0-112.0) IL-5 31/31 18.1 (11.8-23.1) * 15/15 12.8 (6.3-19.0) IL-6 31/31 15.9 (7.4-27.3) 15/15 22.8 (7.7-27.5) IL-8 (CXCL8) 31/31 39.7 (23.3-46.3) * 15/15 30.3 (17.2-35.0) IL-10 31/31 19.2 (9.1-27.4) 15/15 15.2 (8.3-27.1) TNF-α 31/31 67.6 (48.8-81.8) 15/15 56.9 (31.4.70.0) IFN-γ 31/31 88.3 (58.1-98.4) * 15/15 51.3 (37.7-78.0) GM-CSF 31/31 57.4 (30.6-90.4) 15/15 62.1 (26.1-108.5) Chemokines CXCL9 31/31 79.5 (58.5-135.0) 15/15 84.6 (41.3-144.1) CXCL10 31/31 49.6 (35.0-80.3) 15/15 44.8 (33.0-67.5) CXCL11 31/31 48.3 (34.6-92.4) 15/15 48.7 (35.0-96.1) CCL17 31/31 22.3 (15.9-41.5) 15/15 36.6 (21.2-74.6) CCL22 31/31 196.8 (149.5-317.2) 15/15 192.2 (153.8-281.8)

CSF No.of MS MS (n=31) No. of OND OND (n=10) patients with patients with measurable measurable levels levels Cytokines IL-1β 29/30 a 4.4 (3.7-5.3) ** 9/9 b 2.7 (2.6-3.8) IL-2 0/30 - 0/9 - IL-4 0/30 - 0/9 - IL-5 29/30 0.6 (0.6-0.6) 9/9 0.6 (0.6-0.6) IL-6 28/30 1.1 (0.7-1.9) 7/9 1.4 (0.3-2.8) IL-8 (CXCL8) 29/30 83.8 (70.2-111.3) ** 9/9 51.1 (43.3-71.3) IL-10 24/30 0.2 (0.2-0.2) 0/9 - TNF-α 26/30 0.3 (0.3-0.3) 5/9 0.3 (0.3-0.3) IFN-γ 0/30 - 0/9 - GM-CSF 8/30 0.4 (0.4-0.4) 0/9 - Chemokines CXCL9 31/31 50.1 (37.3-73.6) 10/10 38.3 (28.8-55.6) CXCL10 31/31 188.3 (146.4-294.1) *** 10/10 87.1 (74.2-118.9) CXCL11 31/31 5.4 (4.7-7.2) * 10/10 4.4 (3.2-4.9) CCL17 31/31 1.0 (1.0-1.0) 10/10 1.0 (1.0-1.0) CCL22 31/31 21.4 (12.8-35.3) *** 10/10 8.8 (6.1-9.7)

Median values are given and interquartile range within brackets. Values in bold represent significant differences compared to healthy blood donors (plasma) or patients with OND (CSF). a n=30, while analyse was not done in one patient, b n= 9 while analyse was not done in one patient, * p<0.05 compared with healthy blood donors, ** p<0.01 compared with OND patients, *** p<0.001 compared with OND patients. 50

5. Results and Discussion

5.2.3 Cytokines and chemokines in CSF (paper I, III) There was a significant decline in cytokine (IL-1β, IL6) and chemokine (CXCL8, CXCL9, CXCL10, CXCL11, CCL22) levels in CSF after one-year natalizumab treatment as reported in paper I (Figure 8). Considering that CSF levels of IL-1β, CXCL8, CXCL10, CXCL11 and CCL22 were significantly higher in MS patients pre-treatment compared to OND patients, it is notable that after treatment with natalizumab, levels of CXCL11 and CCL22 were reduced to levels similar to OND patients (data not shown). In paper III the significant decline in CSF cytokine (IL-1β) and chemokines (CXCL10, CXCL11, CCL22) after one year of natalizumab treatment was confirmed, also when 12 patients were excluded from the cytokine and chemokine analyses because of their former inclusion in paper I.

Figure 8. CSF levels of cytokines and chemokines pre-and post-treatment (paper I). Median values are given and whiskers mark the interquartile range. * p<0.05, ** p<0.01, *** p<0.001. Please note the broken y-axis.

We interpret the decline in CSF cytokine and chemokine levels as a reflection of the decrease in intrathecal inflammation. It seems logic that a decline in CSF leukocyte numbers would lead to a subsequent decrease in cell-derived mediators like cytokines and chemokines intrathecally. However, the biological and clinical significance of these changes are uncertain.

5. Results and Discussion

5.2.4 Cytokines and chemokines in plasma (paper I) Interestingly, the decline in intrathecal levels of pro-inflammatory cytokines and chemokines was not accompanied by an increase in plasma levels of these pro-inflammatory mediators, as would have been expected. Instead, levels of IL-6, IL-10, GM-CSF, IL-6 and TNF decreased significantly after one-year natalizumab treatment. This result is partially in contrast to previous gene expression data, where mRNA expression of IFN-γ and TNF in PBMC were shown to increase after natalizumab treatment. However, the expression of IL-23, IL-17A, IL- 17F, IL-13, IL-10, IL-4, and TGF-β was unchanged (Khademi, Bornsen et al. 2009). There is however limited data on natalizumab treatment effects on cytokine and chemokine protein levels in plasma. In one report on protein levels in plasma assessed by flow cytometry, IL-1β, IL-2 and IL-5 were significant elevated after treatment for 9-12 months, whereas IL-10 and IFN-γ did not change (Ramos-Cejudo, Oreja-Guevara et al. 2011). IL-6, GM-CSF and TNF were not assessed in this study. To conclude, our findings of a decrease in plasma levels of IL- 6, IL-10, GM-CSF, IL-6 and TNF are difficult to interpret. However, the fact that there is no consistent finding in the literature of elevated plasma protein levels of pro-inflammatory cytokines and chemokines in natalizumab treated patients, as would have been expected, indicate that natalizumab has other effects apart from reducing leukocyte extravasation.

5. Results and Discussion

5.3 Blood lymphocyte composition (paper II)

Main lymphocyte populations

There was a significant increase in absolute numbers of all investigated main lymphocyte populations in peripheral blood after one year of natalizumab treatment (Table 7). This increase was most marked for NK cells (195% increase compared with baseline) and B cells (105% increase compared with baseline). As to changes in proportions within the lymphocyte population, the fractions of NK cells and CD8+ T cells increased significantly, whereas CD3+ T cells and CD4+ T cells decreased significantly (Table 7).

Table 7. Changes in main lymphocyte populations in blood before (baseline) and after one-year (follow-up) treatment with natalizumab.

Number (cells/μl) Percentage of parent population (%)

baseline follow-up change p baseline follow-up change p

Lymphocytes 1989 ± 630 3889 ±1163 +96 % <0.0005 27.7 ± 6.8 40.3 ± 7.0 +45 % <0.0005

T cells 1501 ± 554 2591 ± 915 +73 % <0.0005 75.0 ± 7.0 66.1 ± 6.7 -12 % <0.0005

CD4+ 896 ± 285 1435 ± 397 +60 % <0.0005 61.3 ± 9.7 56.6 ± 7.7 -8 % <0.0005

CD8+ 502 ± 316 975 ± 599 +94 % <0.0005 31.8 ± 8.4 35.9 ± 8.6 +13 % <0.0005

NK cells 277 ± 145 816 ± 248 +195 % <0.0005 14.3 ± 6.6 21.4 ± 5.2 +50 % <0.0005

B cells 258 ± 182 528 ± 296 +105 % <0.0005 12.5 ± 6.4 13.2 ± 5.1 +6 % 0.3

Mean ± SD, n=40 except for CD4+ where n=38, p values refers to paired samples t test comparing number of cells and percentage of parent population, respectively, at baseline and follow-up. Change refers to difference between baseline and follow-up mean given in % of baseline values.

5. Results and Discussion

An overview of blood lymphocyte population compositions before and after one-year natalizumab treatment is given in figure 9 a-d.

Figure 9 a-d. Overview of lymphocyte populations in patients before and after one year of natalizumab treatment. a-b: Distribution of lymphocytes (% of total leukocytes) and CD3+ T cells (% of total lymphocytes). Comparisons are pairwise. Bars denote mean values. c-d: Relative distribution of discrete lymphocyte subpopulations before (c) and after (d) natalizumab treatment.

The increase in circulating numbers of lymphocyte populations associated with natalizumab treatment is in accordance with the proposed mode of action of this therapy. However, since VLA-4 is expressed on many different lymphocyte populations and its main ligand VCAM-1 is ubiquitously expressed on activated endothelium throughout the body, the overall effect of VLA-4 antagonism on leukocyte populations measured in peripheral blood is not only a result of reduced leukocyte migration across the BBB but also across endothelium in many peripheral tissues. Furthermore, considering the low numbers of leukocytes intrathecally compared with numbers in the peripheral compartment, it is unlikely that reduced migration to the CNS may account for the total increase in circulating leukocytes during natalizumab treatment. Accordingly, it is also found that natalizumab mobilizes hematopoietic progenitor cells out of the bone marrow (Bonig, Wundes et al. 2008, Zohren, Toutzaris et al. 2008).

5. Results and Discussion

It is further reported that the magnitude of VLA-4 expression varies between different lymphocyte populations, e. g. being higher in B cells than T cells, and higher in CD8+ T cells than CD4+ T cells (Niino, Bodner et al. 2006). In addition, different lymphocyte subsets seem to respond to natalizumab treatment in various magnitudes. It has recently been reported that the amount of natalizumab binding to CD3- NK cells is higher than the amount binding to CD19+ B cells and CD3+ T cells in a descendant scale (Harrer, Pilz et al. 2012). This diversity in binding preference of natalizumab among different lymphocyte subsets is well in line with our observation of the highest increase in the number of NK cells after treatment (195 %) followed by B cells (105 %) and T cells (CD3+) (73 %). The overall effect of these differences (expression of VLA-4 and binding preference of natalizumab), regarding migration capacity and function for different lymphocyte subsets, is however uncertain.

Lymphocyte subpopulations

NK cells

Among CD3-CD56+ NK cells, a significant decrease in percentage of CD3-CD56bright NK cells, was accompanied by a significant increase in CD3-CD56dim NK cells (Figure 10 a-b). There was a tendency to a decrease in early activated CD3-CD56+CD69+ NK cells (Figure 10 c).

Based on different expression of chemokine receptors and adhesion molecules, cytotoxic NK cells (CD56dim) and regulatory NK cells (CD56bright) have different migration preferences with CD56dim migrating to inflammatory sites while CD56bright preferentially home to secondary lymphoid organs (Campbell, Qin et al. 2001), where they can exert an immunoregulatory function by interacting with dendritic cells (Gandhi, Laroni et al. 2010). Expansion of the CD56bright NK cell subset in blood has been reported in MS patients treated with interferon-beta or daclizumab, in contrast to what we observed during natalizumab treatment (monoclonal antibody against CD25, a part of the IL-2 receptor) (Bielekova, Catalfamo et al. 2006, Saraste, Irjala et al. 2007). Our finding of an increase in the fraction of circulating cytotoxic NK cells, may suggest that these cells are relatively more affected by a decrease in extravasation caused by natalizumab than regulatory NK cells. This interpretation seems logic considering the cytotoxic NK cell preference for homing to inflammatory sites with endothelial upregulation of VCAM-1, compared to regulatory NK cells that preferentially home for secondary lymphoid organs, with a possible less expression of VCAM-1.

T cells

Regarding activation of T cells, there was a significant decrease in the proportion of activated CD4+ cells when using CD4+CD25+ T cells among CD4+ cells as marker of activation (Figure 10 d). ). However, there was a tendency to an increase in CD4+CD69+ cells, representing early activated Th cells (Figure 10 e), with similar findings for CD8+HLA-DR+ cells, representing late activated cytotoxic T cells (Figure 10 f). Fractions of both CD4+OX40L+ cells and CD8+OX40L+ cells, representing activated immunomodulatory T cells, decreased significantly on treatment (Figure 10 g-h). There was a tendency to a decrease in CD4dimCD25bright Treg

5. Results and Discussion cells after treatment (Figure 10 i). The population of CD8+CD28-CD57+ T cells, representing senescent cytotoxic T cells, decreased significantly (Figure 10 j).

No significant change in the Treg (CD4dimCD25bright) population was demonstrated, a result in line with earlier observations and the demonstrated low expression of VLA-4 on Treg cells (Stenner, Waschbisch et al. 2008). The changes in expression status of activation markers and co-stimulatory molecules were to some extent inconsistent; both a decrease (significant in fractions of CD4+CD25+, CD4+OX40L+, CD8+OX40L+, CD8+CD28-CD57+ T cells and a tendency in fractions of CD3-CD56+CD69+ NK cells) as well as a tendency to an increase (in fractions of CD4+CD69+ and CD8+HLA-DR+ T cells) were observed. The interpretation of these findings is challenging. The OX40-OX40L interaction has been ascribed an important role in facilitating survival and clonal expansion of effector and memory T cells, as well as regulating T cell-mediated cytokine production (Croft, So et al. 2009) and possibly promoting Th2 immune responses (Ohshima, Yang et al. 1998). Our finding of a significant decline in proportions of both CD4+OX40L+ and CD8+OX40L+ cells after natalizumab treatment, possibly suggests an attenuation of effector T cell responses in the periphery. Since OX40- OX40L interaction also favors a Th2 immune response, our result may also have implications on the peripheral Th1 –Th2 balance. On the contrary, the observed tendency of an increase in fractions of early activated Th cells (CD4+CD69+) and late activated cytotoxic T cells (CD8+HLA-DR+) and the restored T cell responsiveness in the functional assays after treatment (as described below) may indicate a regained responsiveness of T cells in the periphery. A recovered T cell responsiveness after treatment is also in line with the observed significant decrease in CD8+CD28-CD57+ T cells, since the loss of CD28 and gain of CD57 expression is considered a sign of chronic activation and a marker for immune senescence (Strioga, Pasukoniene et al. 2011).

Whether these observations implicate direct effects of natalizuamb on cell-surface markers of activation and co-stimulation, or constitute indirect effects, demands further investigation.

B cells

The proportion of CD19+CD27+ cells, representing memory B cells, increased significantly on treatment as well as the fraction of CD19+CD25+ cells, possibly representing regulatory B cells (Figure 10 k-l). The increase in the fraction of memory B cells was significantly higher than the increase in regulatory B cells (p=0.005).

The increase in absolute numbers of B cells after treatment are in line with earlier studies showing an increase in circulating B cells during natalizumab treatment (Krumbholz, Meinl et al. 2008, Haas, Bekeredjian-Ding et al. 2011, Planas, Jelcic et al. 2012). Within the B cell population, the fractions of both memory CD19+CD27+ B cells and possibly regulatory CD19+CD25+ B cells increased significantly after treatment. This is in accordance to recent reports showing an increase in memory B cells while the population of naïve B cells decreased (Haas, Bekeredjian-Ding et al. 2011, Planas, Jelcic et al. 2012). The latter finding was suggested to depend on differences in α4-integrin expression between these two B cell subsets. Nevertheless, the marked increase in circulating B cells during natalizumab treatment is a

5. Results and Discussion consistent finding throughout many studies and further yields this lymphocyte population a probably essential role in treatment effects and side effects.

Figure 10 a-l. Phenotypic characteristics of blood lymphocyte subpopulations in MS patients before and after one year of natalizumab treatment (pre and post, respectively). p<0.01 is considered statistically significant. All comparisons are pairwise. Bars show mean value, whiskers denote SD.

5. Results and Discussion

5.4 Lymphocyte activation assay (paper II) The number of activated lymphocytes (lymphoblast-transformed cells) after antigen- or mitogen stimulation was assessed at baseline and compared with the response after one year of natalizumab treatment. The results from these assays are given in figure 11 a-f. As controls, we also analyzed blood cells of healthy individuals, showing a significantly stronger response compared with pre-treatment levels of patients regarding influenza antigen-induced CD4+ and CD4+CD108+ Th cell responses (Figure 12). However, post-treatment levels of patients were in the same range as those for controls.

Figure 11 a-f. Response towards antigens in different T cell populations. Pairwise comparison between patients before and after one year of natalizumab treatment. PPD=purified protein derivative, PWM=pokeweed mitogen, CMV=cytomegalovirus, n=37 in both groups.

5. Results and Discussion

Figure 12. Response towards antigens in healthy controls and patients before and after one year of natalizumab treatment. For visualization purposes, data is normalized to the average of the healthy controls for the respective antigen. Analyses performed with one-way ANOVA with Tukey’s post hoc test. *: p<0.05, comparison between controls and pre-treatment patients. ¶: p<0.05, comparison between pre- and post-treatment patients. n=23 for controls, n=37 for both pre- and post-treatment groups. Bars represent mean values, whiskers mark SEM.

The overall pattern of the lymphocyte activation assays was a consistently higher response post-treatment compared with pre-treatment, although not always statistically significant. Since the assay determines the number of responding cells, this observation is in accordance with a reduction in immune cell migration into tissues by VLA-4 blockade and thus expectedly an increase in the number of reactive memory T cells in the circulation. In addition, when comparing healthy individuals to MS patients post-treatment, comparable levels of reactivity were observed, although pre-treatment MS patients overall showed lower reactivity than healthy controls. Speculatively this may be explained by the notion that in a neuroinflammatory setting, the immune-competent T cells are residing at the inflammatory focus, i.e. the CNS. Normally, T cells are constantly migrating between peripheral lymphoid tissues, blood and other tissues. In MS, this migratory pattern could be disturbed by trapping of effector T cells to the focus of inflammation, however, during natalizumab treatment these cells can remain in the circulation.

5. Results and Discussion

5.5 Markers of neurodegeneration in CSF (paper III) There was a significant decrease in CSF levels of MBP (from 1.08 to 0.97 ng/ml, median values, p<0.0005) and NFL (from 509 to 289 ng/l, median values, p<0.0005) after one year of natalizumab treatment. GFAP levels increased (from 314 to 382 ng/l, median values, p=0.004), whereas T-tau and P-tau were unchanged. Reduced CSF levels of MBP are in accordance with a decline in demyelination activity (Cohen, Brooks et al. 1980) because of reduced intrathecal inflammation, and the decrease in levels of NFL is interpreted as the consequence of a general decline in neuroaxonal damage associated with natalizumab treatment (Norgren, Sundstrom et al. 2004). The increase in levels of GFAP may indicate a process of increasing astrogliosis (Malmestrom, Haghighi et al. 2003, Middeldorp and Hol 2011). However, in a multicenter Swedish cohort of 99 MS patients, including nine patients from the present study (paper III), there was no increase in GFAP levels during natalizumab treatment (Gunnarsson, Malmestrom et al. 2010). Thus, the increase in GFAP should be interpreted with caution. No change in T-tau and P-tau levels was observed. The accumulation of Tau-proteins in neurons and glia cells and elevated levels in CSF are typical signs of the neurodegenerative process in Alzheimer´s disease. Although elevated levels of tau-proteins in CSF among MS patients have been reported, this is not a consistent finding (Teunissen, Dijkstra et al. 2005).

5. Results and Discussion

5.6 1H-MRS metabolite concentrations in NAWM (paper III) Pre-treatment 1H-MRS metabolite concentration levels in NAWM of MS patients differed significantly from levels in healthy controls regarding tNA, tCho, mIns and tGlx, but no significant group level changes in metabolites were observed after one year of natalizumab treatment (Table 8).

Table 8. 1H-MRS metabolite concentrations at baseline and at follow-up after one year of natalizumab treatment (presented as units of mM aq).

HC MS MS baseline p a baseline (n=27) follow-up (n=27) change during p b (n=20) treatment mean SD mean SD mean SD mean CI difference tNA 13.35 0.82 0.001 12.20 1.40 11.95 1.25 -0.25 -0.72, 0.22 0.3 tCr 6.82 0.38 0.5 6.91 0.55 6.93 0.46 0.02 -0.25, 0.29 0.9 tCho 2.59 0.31 0.02 2.83 0.34 2.83 0.26 0.00 -0.11, 0.10 0.9 mIns 5.95 0.94 0.004 7.08 1.60 6.91 1.16 -0.17 -0.70, 0.36 0.5 tGlx 10.56 1.60 0.003 12.42 2.47 12.33 2.11 -0.09 -1.26, 1.09 0.9 Lac 0.29 0.27 0.2 0.45 0.52 0.55 0.62 0.10 -0.21, 0.41 0.5

To the left the mean and standard deviation (SD) of metabolite concentrations among healthy controls (HC). p a refers to significance level with independent samples t test comparing HC with MS patients at baseline. Significant differences were still valid when adjusting for age using logistic regression analyses. To the right the mean and the standard deviation (SD) of metabolite concentrations among MS patients are presented at baseline and at follow-up. At the far right the mean difference in metabolite concentrations post-to pretreatment are presented with a 95% confidence interval (CI). p b refers to significance level with paired samples t test comparing levels of metabolites in MS patients post-to pretreatment.

Since we aimed to explore possible associations between on one hand changes in 1H-MRS metabolite levels, reflecting the development of disease pathology, and on the other hand and the intrathecal status of inflammation and neurodegeneration, we performed correlation analyses between one-year change in H-MRS metabolite levels versus baseline and follow-up levels of markers for inflammation and neurodegeneration. The rationale for this procedure was that although there were no observed overall changes in 1H-MRS metabolite levels at one-year follow-up, this finding does not imply that there was a zero change in metabolite concentrations in all patients. We thus wanted to explore if individual variations in 1H-MRS metabolite change could be associated with individual levels of CSF markers of inflammation and neurodegeneration. By using correlation analyses we also could make use of the broad investigational approach (examinig both markers of inflammation and neurodegeneration as well as 1H-MRI data) and the longitudinal design of the study.

Interestingly, we found a consistent pattern of high magnitude correlation coefficients indicating an association between change in specific 1H-MRS metabolite concentrations in NAWM and the inflammatory markers IL-1β and CXCL8 (Table 9 a). In specific, intrathecal levels of IL-1β and CXCL8, the latter previously known as IL-8, correlated positively at 61

5. Results and Discussion

baseline and at follow-up to change in levels of tCr (IL-1β and CXCL8) and tCho (IL-1β and for CXCL8 at follow-up only) over one year (Table 9 b). These results suggest that patients, despite treatment, being characterized by high levels of inflammatory markers at follow-up, experienced an increase in density of glia cells and also increased membrane turnover in the MRS voxel during the one-year follow-up time. Inversely, elevated levels of inflammatory markers at baseline could possibly be a prognostic factor for the development of increased glia cell density and membrane turnover. This finding may be speculatively interpreted as an involvement of subclinical persistent inflammation in MS diffuse white matter pathology.

Table 9 a-b. Correlation analyses map (bivariate non-parametric) between change in 1H-MRS metabolite concentrations during one-year versus levels of CSF markers at baseline and follow-up.

a: Map of all correlation coefficients.

1H-MRS change Inflammation Th1 Th2 Neurodegeneration

CXCL8 IL-1β IL-6 TNF CXCL10 CXCL11 CCL22 NFL GFAP T-tau P-tau MBP tNA tCr * ** * ** * tCho ** * ** mIns tGlx Lac The magnitude of all correlation coefficients (r) is shown. ** r > 0.5, * r = 0.4-0.5, no asterisk r < 0.4. CSF markers are subdivided into markers of inflammation, chemokines representing T helper cells 1 (Th1) and 2 (Th2) and neurodegeneration.

b: Detail of correlation analyses map showing one-year change in metabolite concentrations versus baseline (left column for each analyte) and follow-up (right column for each analyte) levels of CXCL8 and IL-1β.

1H-MRS CXCL8 (n=24) IL-1β (n=24) change baseline follow-up baseline follow-up tNA r 0.01 -0.04 -0.02 -0.15 p 1.0 0.9 0.9 0.5 tCr r 0.48 0.67 0.49 0.52 p 0.02 <0.0005 0.02 0.01 tCho r 0.38 0.56 0.43 0.52 p 0.1 <0.01 0.03 0.01 mIns r 0.21 0.09 0.30 0.10 p 0.3 0.7 0.2 0.6 tGlx r -0.26 -0.14 -0.25 -0.25 p 0.2 0.5 0.2 0.2 Lac r 0.14 -0.004 0.09 -0.04 p 0.5 1.0 0.7 0.8

High magnitude correlation coefficients in bold. r=correlation coefficient, p=significance level.

5. Results and Discussion

The correlation analyses included several variables and when considering multiple comparisons, each with a significance test, the problem with mass significance is true. Our solution to this problem was not focusing on repeated significance testing of isolated correlations one by one. Instead, the magnitude of the correlation coefficients and their pattern was the main focus. The magnitude of the correlation coefficient reflects the strength of the association between two variables, irrespective of whether this coefficient happens to be “significant” or not in a null hypothesis testing. Furthermore, and most important, data revealed a pattern of high magnitude correlation coefficients in a consistent direction that was relevant and interesting.

5.7 Transcriptional characteristics of CD4+ T cells (paper IV) Paper IV is a cross-sectional, descriptive study on the balance of circulating CD4+ T cell populations in blood of 33 patients with relapsing MS. The different T cell populations were defined by expression of their respective associated transcription factor. The main finding was a significantly diminished expression of Treg-associated FOXP3, Th2-associated GATA3 and immunoregulatory EBI3 in RRMS compared with healthy blood donors (Figure 13 A-E). Similar findings were present also when recognizing the possible confounding factor of IFN-β treatment. In contrast, transcription factors associated with proposed pro-inflammatory and disease-promoting T cell subsets Th1 (TBX21) and Th17 (RORC) did not differ compared with controls. On the basis of the reciprocal relation between TBX21 and GATA3 (Ouyang, Ranganath et al. 1998) as well as RORC and FOXP3 (Ichiyama, Yoshida et al. 2008), ratios of TBX21/GATA3 and RORC/FOXP3 expression were also calculated. An increase in RORC/FOXP3 ratios among MS patients was then found (Figure 14).

5. Results and Discussion

Figure 13 A-E. Comparison of transcription factor expression fold change in patients with MS and healthy controls (HC). All values represent ratios between transcription factor and 18S expression. Bars represent median values and interquartile range and whiskers mark the 10th and 90th percentiles. rrMS=relapsing-remitting MS (n=33), HC=healthy controls (n=20), * p<0.05, ** p<0.01.

5. Results and Discussion

Figure 14. Transcription factor ratios in MS. Bars represent median values and whiskers mark the interquartile range (25th and 75th percentiles). rrMS=relapsing-remitting MS (n=33), HC=healthy controls (n=20), ***p<0.005.

We interpret these findings as an indication of defects at the mRNA level systemically, involving downregulation of beneficial CD4+ phenotypes in RRMS. These results stress the importance of assessing the balance between pro-inflammatory and regulatory subsets when exploring disease mechanisms. Notably, data was based on whole blood analyses and accordingly we cannot argue that myelin-specific cells were measured. Nevertheless, our findings may suggest that a defect in the relative immune suppression capacity of CD4+ T cells in the periphery could contribute to Th-mediated pathogenic mechanisms.

6. Concluding remarks and future perspectives

This thesis discusses different aspects of the immune response and its consequences in the context of neuroinflammation and neurodegeneration in MS. We assessed both soluble mediators as cytokines and chemokines in plasma and CSF, different lymphocyte populations in blood, diffuse white matter pathology using 1H-MRS as well as expression of intracellular transcription factors. Although one main aim of the thesis was to elucidate immunological mechanisms and consequences of natalizumab treatment, we did not compare natalizumab- induced changes with what would have been the natural course, that is without natalizumab treatment. For obvious reasons, a placebo group cannot be used and therefore the study design was to compare baseline status with status after one year of treatment. Consequently we could not, and did not aim at evaluating clinical effects of treatment. However, we still wanted to describe the clinical setting being a basic description of the material.

Our results regarding clinical outcome after one-year natalizumab treatment are well in line with previous reports and confirm the efficacy of VLA-4 blockade. As to the effects on lymphocyte populations in the circulation, changes in the major populations (T cells, B cells, NK cells) may be explained by their expression levels of VLA-4, thus reflecting the effect of natalizumab on cell trafficking. In addition, the reduction of cytokine and chemokine levels intrathecally, also tallies with a decline in leukocyte extravasation due to VLA-4 blockade. However, the differential effects on lymphocyte subpopulations, including cell-surface molecules of co-stimulation and immunomodulation, as well as the intriguing divergent findings of plasma levels of cytokines and chemokines, seem hard to explain solely by effects on cell trafficking. Hence, additional effects of natalizumab involving cell signalling are likely. Knowing these effects may help to understand the mechanisms of adverse reactions as the development of PML.

Although there is a reduction in intrathecal inflammation and immune surveillance during natalizumab treatment, the overall systemic increase in lymphocyte populations (especially considering cytotoxic NK cells and memory B cells) indicate a preserved or even an increase in the ability for immune responses systemically. However, since natalizumab also reduces leukocyte migration into peripheral tissues, immune surveillance and immune responses at other sites like the gut and lungs may be insufficient. This may have to be accounted for in treatment considerations.

Apart from the changes in soluble mediators and lymphocyte populations associated with natalizumab treatment, we also addressed diffuse white matter pathology in MS. The distinct reduction in inflammatory markers (total leukocyte cell counts, IgG index, cytokines and chemokines) and markers for demyelination and neuroaxonal damage (MBP and NFL) intrathecally after one year, constituted the rationale to explore if these variables were associated with NAWM changes as measured by 1H-MRS metabolite concentrations. The observation of a possible relation between intrathecal subclinical inflammation and development of presumably degenerative processes in NAWM, as measured by 1H-MRS, may 66

6. Concluding remarks and future perspectives be speculative and needs further confirmation, yet it may have important implications for the understanding of diffuse pathology in MS and how to prevent it.

In the future, it would be interesting to evaluate the effects of natalizumab treatment on the balance of circulating CD4+ T cell populations as defined by the expression of transcription factors (in analogy to paper IV). This would hint at possible mechanisms of natalizumab treatment that goes beyond the effect on cell trafficking. Such mechanisms could also be successfully studied in an in vitro setting. Furthermore, additional functional studies on lymphocytes, comparing pre-to post-treatment characteristics, should be done. One relevant study would be to find ways to identify responders and non-responders (as to clinical effects of natalizumab) by means of in vitro assessment of lymphocyte responsiveness. Another study of high relevance is to evaluate 1H-MRS metabolite concentrations after three years of natalizumab treatment.

7. Conclusions

Paper I Natalizumab treatment was associated with a global decline in cytokine and chemokine levels at a protein level. This finding was most pronounced in CSF, in line with the reduced migration of cells into the CNS, whereas reduction also in plasma levels indicates other possible mechanisms of natalizumab treatment.

Paper II Natalizumab treatment was associated with a pronounced effect on the lymphocyte composition in blood with an increase in numbers of all investigated major populations, most marked for NK cells and B cells. The overall pattern of the functional T cell responses was a consistently higher response post-treatment compared to pre-treatment, although not always statistically significant. Changes in the major populations (T cells, B cells and NK cells) may be explained by their expression levels of VLA-4, thus reflecting the effect of natalizumab on cell trafficking, but the differential effects on subsets of these populations, indicate additional effects of natalizumab involving cell-signalling.

Paper III The levels of 1H-MRS metabolite concentrations were unchanged after one year of natalizumab treatment on a group level. However, despite a clinical improvement and a global decrease in levels of inflammatory markers in CSF during treatment, high levels of pro-inflammatory CXCL8 and IL-1β were associated with an increase in 1H-MRS metabolites indicative of continued gliosis development and membrane turnover in NAWM.

Paper IV Gene expression analyses of transcriptional factors indicated systemic defects at the mRNA level in RRMS, involving downregulation of beneficial CD4+ phenotypes. This relative lack of suppression might play a role in disease development by permitting the activation of harmful T cell populations.

8. Sammanfattning på svenska

Bakgrund: Multipel skleros (MS) är en kronisk sjukdom i det centrala nervsystemet (CNS) där nervtrådarnas skyddshölje, myelin, angrips av en inflammatorisk process. MS utgör, näst efter olyckor, den vanligaste orsaken till neurologisk funktionsnedsättning hos unga vuxna. Förutom lokala skador på grund av inflammation har MS också visat sig medföra en diffust utbredd skadeprocess i hjärnan som på sikt innebär en tilltagande funktionsnedsättning och minskad hjärnvolym (atrofi). Huruvida denna process är en självständig del av MS-sjukdomen eller beror på inflammation är oklart. Den inflammatoriska komponenten vid MS anses orsakad av immunsystemets T-lymfocyter, även kallade T- celler (en typ av vita blodkroppar), i samarbete med andra immunceller och deras producerade substanser som cytokiner och kemokiner. Natalizumab, den hittills mest effektiva behandlingen mot MS, utgörs av så kallade antikroppar (sökmolekyler) som binder och blockerar molekylen α4β1-integrin (VLA-4) på cellytan. Genom denna blockering kan bland annat T-celler inte ta sig in till hjärnan eftersom antikropparna hindrar deras bindning till kärlväggen. Natalizumab minskar på så sätt inflammationen i hjärnan och därmed minskar också skovfrekvensen hos patienten. Däremot är det okänt om natalizumab också påverkar andra processer i hjärnan och likaså om blockering av VLA-4 molekylen leder till andra förändringar utöver effekten på celltrafik till hjärnan. Metoder: Totalt inkluderades 72 stycken MS-patienter, fördelade på 4 olika delarbeten. Vi undersökte effekterna av ett års natalizumab-behandling hos 31 MS-patienter vad gäller koncentrationer av inflammationsmediatorer, cytokiner och kemokiner, i ryggvätska (cerebrospinalvätska, CSV) och blod med hjälp av multiplexteknik (arbete I). Vidare undersöktes sammansättningen av immunceller i blod hos 40 patienter med hjälp av flödescytometri före och efter ett års behandling (arbete II). Diffusa sjukdomsförändringar i till synes normal vit hjärnvävnad (NAWM) bedömdes genom att mäta koncentrationen av olika ämnen i NAWM med hjälp av proton-magnetresonansspektroskopi (1H-MRS) hos 27 patienter före och efter behandling (arbete III). Slutligen analyserades balansen mellan olika hjälpar-T-celler i blod hos 33 MS-patienter med hjälp av molekylärbiologisk metod (arbete IV). Resultat: Ett års behandling med natalizumab ledde till en tydlig minskning av inflammationsmediatorer i ryggvätska, det vill säga både av cytokiner (IL-1β och IL-6) och kemokiner (CXCL8, CXCL9, CXCL10 och CXCL11). Nivåer av några cytokiner i blod minskade också (GM- CSF, TNF, IL-6 och IL-10). Behandlingen ledde dessutom till en ökning av det totala antalet lymfocyter i blod inklusive undergrupper av lymfocyter (hjälpar-T-celler, mördar-T-celler, NK-celler och B-celler). En normalisering av T-cellernas förmåga att reagera mot virus och på ämnen för celltillväxt påvisades också. Vi kunde däremot inte med 1H-MRS på gruppnivå påvisa någon skillnad före jämfört med efter ett års behandling i sammansättning av ämnen i till synes normal vit hjärnvävnad. Vi fann vidare ett samband mellan förändring i de 1H-MRS-påvisade ämnena kreatin och kolin under ett års behandling och nivåer av inflammationsmediatorerna IL-1β och CXCL8. Analys av balansen mellan olika hjälpar- T-celler visade tecken på en minskad andel gynnsamma hjälpar-T-celler i blod hos MS-patienter jämfört med hos friska kontrollpersoner. Slutsatser: Våra fynd stöder att behandling med natalizumab påverkar celltrafiken mellan blodbanan och annan vävnad. Samtidigt tycks behandlingen också påverka undergrupper av immunceller, vilket också kan bidra till behandlingsresultatet. Undersökningen av inflammationsmediatorer i ryggvätska i kombination med 1H-MRS påvisade ett möjligt samband mellan inflammationsprocesser och tilltagande diffus hjärnskada. Avslutningsvis tycks det finnas en ofördelaktig sammansättning mellan olika hjälpar- T-celler i blod hos MS-patienter jämfört med friska kontrollpersoner. Denna obalans skulle kunna bidra till utveckling av MS och en korrigering av obalansen kan möjligen utgöra en framtida behandlingsstrategi.

9. Acknowledgements

I wish to express my sincere gratitude to all who made this thesis possible, especially I would like to thank:

Magnus Vrethem, chief tutor and co-author, for great support and encouragement during these years. As head of the MS-team you have been invaluable for the progress of this thesis. A special thanks also for always generously sharing your knowledge and vast experience in MS clinics.

Jan Ernerudh, co-tutor and co-author, for sharing your impressive knowledge in the field of immunology and for invaluable support and guidance. Thanks for always taking the time for discussions and giving essential comments on manuscripts.

Charlotte Dahle, co-tutor and co-author, for your never-ending enthusiasm and encouragement throughout this project, as well as for inspiring discussions on topics beyond immunology.

Måns Edström, co-author, for truly appreciated collaboration both in the laboratory and during the preparation of manuscripts. Thanks for always answering my labquestions and sharing my interest for syncopated beats!

Anders Tisell, co-author at CMIV, for great collaboration during the work with the MRS study.

Peter Lundberg, Olof Dahlqvist Leinhard, Anne-Marie Landtblom, Ida Blystad, Kaj Blennow and Bob Olsson, co-authors, for stimulating collaboration during the MRS study.

Maria Jenmalm and Jenny Mjösberg, co-authors, for sharing your knowledge and skills as to immunological methods and for valuable comments on manuscripts.

Rayomand Press, co-author, for nice collaboration during the work with the transcription study.

Petra Cassel, co-worker at AIR, for helping me out with the laboratory work and for your generous attitude.

Jan-Edvin Olsson, professor emeritus, for all support and encouragement.

Mats Andersson, Anne-Marie Landtblom, Nil Dizdar and Patrick Vigren for providing me with the opportunity to combine clinical work and research during their time as Heads of the Department of Neurology.

To members of the MS-team, Helene Lundberg, Gunn Johansson, Britt Morberg, Anne Wickström, Marielle Kagre, Annette Sverker, Katja Asplund, Eva Nilsson, Liz Myhrinder, Annica Insulander, Elisabeth Fransson, Martin Hemmingsson, Daniel Ulrici, Iréne Håkansson and Yumin Link for all practical help and for sharing everyday work with our MS patients.

9. Acknowledgements

Per Hultman, my first research supervisor at Medical School, for introducing me to research and autoimmunity, in this case mercury-induced autoimmunity in rodents. I will never forget the shape of green-fluorescent antinucleolar antibodies in the dark!

Christina Ekerfelt, tutor during my year at the Biomedical Research School, for encouragement and introduction to MS immunology including “immune deviation in vitro”.

Olof Danielsson, former clinical tutor, for introducing me to the clinical skills of neurology and for your good example of devotion to the subject of neurology.

Mats Fredrikson, for statistical advice on paper III.

The staffs at AIR and Clinical Immunology, for handling of samples and performing flow cytometry analyses, respectively.

All highly appreciated colleagues and co-workers at the Department of Neurology for their support and understanding. A special thanks to the staff at neurology ward 73, for taking care of patients during natalizumab infusions.

To all patients who generously participated in all studies, without you this thesis would never have been written.

All dear friends, without you, life would be truly empty! A special help and encouragement when working on this thesis I got from Sara and Johan and from Charlotte and Mattias.

My parents-in-law, Margaretha and Rodney, for great generosity and practical help whenever needed.

My parents, Aina and Beo, for endless love and support from the very start and throughout the years. My sisters, Camilla, Martina and Hanna and my brother Joseph and their families for love and friendship.

My beloved Selma, Morris and Bea; you fill me with tremendous joy and gratitude!

My beloved Sara for all your love, immense support and for being a true life companion. Words fail me!

This work was supported by grants from The National Research Council (VR/NT), The Swedish Society of Medicine, The Swedish Association of Persons with Neurological Disabilities, Linköping Society of Medicine, The University Hospital of Linköping, The County Council of Östergötland and Linköping University. The project was further supported by an unrestricted grant and award from Biogen Idec Sweden.

10. References

(1995). "Interferon beta-1b in the treatment of Bartosik-Psujek, H. and Z. Stelmasiak (2005). "The multiple sclerosis: final outcome of the randomized levels of chemokines CXCL8, CCL2 and CCL5 in controlled trial. The IFNB Multiple Sclerosis Study multiple sclerosis patients are linked to the activity Group and The University of British Columbia of the disease." Eur J Neurol 12(1): 49-54. MS/MRI Analysis Group." Neurology 45(7): 1277- Basu, A., J. K. Krady and S. W. Levison (2004). 1285. "Interleukin-1: a master regulator of (1998). "Randomised double-blind placebo- neuroinflammation." J Neurosci Res 78(2): 151- controlled study of interferon beta-1a in 156. relapsing/remitting multiple sclerosis. PRISMS Berer, K., M. Mues, M. Koutrolos, Z. A. Rasbi, M. (Prevention of Relapses and Disability by Boziki, C. Johner, H. Wekerle and G. Interferon beta-1a Subcutaneously in Multiple Krishnamoorthy (2011). "Commensal microbiota Sclerosis) Study Group." Lancet 352(9139): 1498- and myelin autoantigen cooperate to trigger 1504. autoimmune demyelination." Nature 479(7374): Abbas A.K., L. A. H., Ed. (2012). Cellular and 538-541. Molecular Immunology. Philadelphia, Elsevier Berzins, S. P., M. J. Smyth and A. G. Baxter Saunders. (2011). "Presumed guilty: natural killer T cell Ahlgren, C., A. Oden and J. Lycke (2012). "A defects and human disease." Nat Rev Immunol nationwide survey of the prevalence of multiple 11(2): 131-142. sclerosis in immigrant populations of Sweden." Bielekova, B., M. Catalfamo, S. Reichert-Scrivner, Mult Scler 18(8): 1099-1107. A. Packer, M. Cerna, T. A. Waldmann, H. Airas, L., M. Saraste, S. Rinta, I. Elovaara, Y. H. McFarland, P. A. Henkart and R. Martin (2006). Huang and H. Wiendl (2008). "Immunoregulatory "Regulatory CD56(bright) natural killer cells factors in multiple sclerosis patients during and mediate immunomodulatory effects of IL-2Ralpha- after pregnancy: relevance of natural killer cells." targeted therapy (daclizumab) in multiple Clin Exp Immunol 151(2): 235-243. sclerosis." Proc Natl Acad Sci U S A 103(15): 5941-5946. Akdis, M., S. Burgler, R. Crameri, T. Eiwegger, H. Fujita, E. Gomez, S. Klunker, N. Meyer, L. Bloomgren, G., S. Richman, C. Hotermans, M. O'Mahony, O. Palomares, C. Rhyner, N. Ouaked, Subramanyam, S. Goelz, A. Natarajan, S. Lee, T. A. Schaffartzik, W. Van De Veen, S. Zeller, M. Plavina, J. V. Scanlon, A. Sandrock and C. Bozic Zimmermann and C. A. Akdis (2011). (2012). "Risk of natalizumab-associated "Interleukins, from 1 to 37, and interferon-gamma: progressive multifocal leukoencephalopathy." N receptors, functions, and roles in diseases." J Engl J Med 366(20): 1870-1880. Allergy Clin Immunol 127(3): 701-721 e701-770. Bonig, H., A. Wundes, K. H. Chang, S. Lucas and Alvarez, J. I., R. Cayrol and A. Prat (2011). T. Papayannopoulou (2008). "Increased numbers of "Disruption of central nervous system barriers in circulating hematopoietic stem/progenitor cells are multiple sclerosis." Biochim Biophys Acta 1812(2): chronically maintained in patients treated with the 252-264. CD49d blocking antibody natalizumab." Blood 111(7): 3439-3441. Axelsson, M., C. Malmestrom, S. Nilsson, S. Haghighi, L. Rosengren and J. Lycke (2011). "Glial Bruck, W. (2005). "The pathology of multiple fibrillary acidic protein: a potential biomarker for sclerosis is the result of focal inflammatory progression in multiple sclerosis." J Neurol 258(5): demyelination with axonal damage." J Neurol 252 882-888. Suppl 5: v3-9. Baggiolini, M. (1998). "Chemokines and leukocyte Campbell, J. J., S. Qin, D. Unutmaz, D. Soler, K. E. traffic." Nature 392(6676): 565-568. Murphy, M. R. Hodge, L. Wu and E. C. Butcher (2001). "Unique subpopulations of CD56+ NK and Bar-Or, A., P. A. Calabresi, D. Arnold, C. NK-T peripheral blood lymphocytes identified by Markowitz, S. Shafer, L. H. Kasper, E. Waubant, S. chemokine receptor expression repertoire." J Gazda, R. J. Fox, M. Panzara, N. Sarkar, S. Immunol 166(11): 6477-6482. Agarwal and C. H. Smith (2008). "Rituximab in relapsing-remitting multiple sclerosis: a 72-week, Cannella, B. and C. S. Raine (1995). "The adhesion open-label, phase I trial." Ann Neurol 63(3): 395- molecule and cytokine profile of multiple sclerosis 400. lesions." Ann Neurol 37(4): 424-435.

10. References

Caruso, A., S. Licenziati, M. Corulli, A. D. Canaris, Duddy, M., M. Niino, F. Adatia, S. Hebert, M. M. A. De Francesco, S. Fiorentini, L. Peroni, F. Freedman, H. Atkins, H. J. Kim and A. Bar-Or Fallacara, F. Dima, A. Balsari and A. Turano (2007). "Distinct effector cytokine profiles of (1997). "Flow cytometric analysis of activation memory and naive human B cell subsets and markers on stimulated T cells and their correlation implication in multiple sclerosis." J Immunol with cell proliferation." Cytometry 27(1): 71-76. 178(10): 6092-6099. Chiaravalloti, N. D. and J. DeLuca (2008). Durelli, L., L. Conti, M. Clerico, D. Boselli, G. "Cognitive impairment in multiple sclerosis." Contessa, P. Ripellino, B. Ferrero, P. Eid and F. Lancet Neurol 7(12): 1139-1151. Novelli (2009). "T-helper 17 cells expand in multiple sclerosis and are inhibited by interferon- Codarri, L., G. Gyulveszi, V. Tosevski, L. Hesske, beta." Ann Neurol 65(5): 499-509. A. Fontana, L. Magnenat, T. Suter and B. Becher (2011). "RORgammat drives production of the Dutta, R. and B. D. Trapp (2007). "Pathogenesis of cytokine GM-CSF in helper T cells, which is axonal and neuronal damage in multiple sclerosis." essential for the effector phase of autoimmune Neurology 68(22 Suppl 3): S22-31; discussion S43- neuroinflammation." Nat Immunol 12(6): 560-567. 54. Cohen, S. R., B. R. Brooks, R. M. Herndon and G. Dutta, R. and B. D. Trapp (2011). "Mechanisms of M. McKhann (1980). "A diagnostic index of active neuronal dysfunction and degeneration in multiple demyelination: myelin basic protein in sclerosis." Prog Neurobiol 93(1): 1-12. cerebrospinal fluid." Ann Neurol 8(1): 25-31. Eikelenboom, M. J., J. Killestein, T. Izeboud, N. F. Commins, S. P., L. Borish and J. W. Steinke Kalkers, P. A. Baars, R. A. van Lier, F. Barkhof, B. (2010). "Immunologic messenger molecules: M. Uitdehaag and C. H. Polman (2005). cytokines, interferons, and chemokines." J Allergy "Expression of adhesion molecules on peripheral Clin Immunol 125(2 Suppl 2): S53-72. lymphocytes predicts future lesion development in MS." J Neuroimmunol 158(1-2): 222-230. Compston, A. (1988). "The 150th anniversary of the first depiction of the lesions of multiple Ekerfelt, C., C. Dahle, R. Weissert, M. Kvarnstrom, sclerosis." J. Neurol. Neurosurg. Psychiatry 51: T. Olsson and J. Ernerudh (2001). "Transfer of 1249-1252. myelin-specific cells deviated in vitro towards IL-4 production ameliorates ongoing experimental Croft, M., T. So, W. Duan and P. Soroosh (2009). allergic neuritis." Clin Exp Immunol 123(1): 112- "The significance of OX40 and OX40L to T-cell 118. biology and immune disease." Immunol Rev 229(1): 173-191. Eng, L. F., J. J. Vanderhaeghen, A. Bignami and B. Gerstl (1971). "An acidic protein isolated from Dahlqvist, O. (2010). Quantitative Magnetic fibrous astrocytes." Brain Res 28(2): 351-354. Resonance in Diffuse Neurological and Liver Disease, Linköping University. Eriksson, P., C. Sandell, K. Backteman and J. Ernerudh (2010). "B cell abnormalities in De Stefano, N., M. L. Bartolozzi, L. Guidi, M. L. Wegener's granulomatosis and microscopic Stromillo and A. Federico (2005). "Magnetic polyangiitis: role of CD25+-expressing B cells." J resonance spectroscopy as a measure of brain Rheumatol 37(10): 2086-2095. damage in multiple sclerosis." J Neurol Sci 233(1- 2): 203-208. Ernerudh, J. (2012). Klinisk Immunologi. Lund, Studentlitteratur AB. del Pilar Martin, M., P. D. Cravens, R. Winger, E. M. Frohman, M. K. Racke, T. N. Eagar, S. S. Evangelou, N., D. Konz, M. M. Esiri, S. Smith, J. Zamvil, M. S. Weber, B. Hemmer, N. J. Palace and P. M. Matthews (2000). "Regional Karandikar, B. K. Kleinschmidt-DeMasters and O. axonal loss in the corpus callosum correlates with Stuve (2008). "Decrease in the numbers of dendritic cerebral white matter lesion volume and cells and CD4+ T cells in cerebral perivascular distribution in multiple sclerosis." Brain 123 ( Pt 9): spaces due to natalizumab." Arch Neurol 65(12): 1845-1849. 1596-1603. Farrar, J. D., H. Asnagli and K. M. Murphy (2002). Dhib-Jalbut, S. and S. Marks (2010). "Interferon- "T helper subset development: roles of instruction, beta mechanisms of action in multiple sclerosis." selection, and transcription." J Clin Invest 109(4): Neurology 74 Suppl 1: S17-24. 431-435. Disanto, G., J. M. Morahan, M. H. Barnett, G. Fazekas, F., F. Barkhof, M. Filippi, R. I. Grossman, Giovannoni and S. V. Ramagopalan (2012). "The D. K. Li, W. I. McDonald, H. F. McFarland, D. W. evidence for a role of B cells in multiple sclerosis." Paty, J. H. Simon, J. S. Wolinsky and D. H. Miller Neurology 78(11): 823-832. (1999). "The contribution of magnetic resonance

10. References imaging to the diagnosis of multiple sclerosis." (alpha4 and alphaL) monoclonal antibodies: two- Neurology 53(3): 448-456. edged swords?" Immunology 116(3): 289-296. Filippi, M., M. A. Rocca, F. Barkhof, W. Bruck, J. Goverman, J. (2009). "Autoimmune T cell T. Chen, G. Comi, G. DeLuca, N. De Stefano, B. J. responses in the central nervous system." Nat Rev Erickson, N. Evangelou, F. Fazekas, J. J. Geurts, C. Immunol 9(6): 393-407. Lucchinetti, D. H. Miller, D. Pelletier, B. F. Goverman, J. M. (2011). "Immune tolerance in Popescu and H. Lassmann (2012). "Association multiple sclerosis." Immunol Rev 241(1): 228-240. between pathological and MRI findings in multiple sclerosis." Lancet Neurol 11(4): 349-360. Guagnozzi, D. and R. Caprilli (2008). "Natalizumab in the treatment of Crohn's disease." Biologics 2(2): Fletcher, J. M., S. J. Lalor, C. M. Sweeney, N. 275-284. Tubridy and K. H. Mills (2010). "T cells in multiple sclerosis and experimental autoimmune Gunnarsson, M., C. Malmestrom, M. Axelsson, P. encephalomyelitis." Clin Exp Immunol 162(1): 1- Sundstrom, C. Dahle, M. Vrethem, T. Olsson, F. 11. Piehl, N. Norgren, L. Rosengren, A. Svenningsson and J. Lycke (2010). "Axonal damage in relapsing Fontenot, J. D., M. A. Gavin and A. Y. Rudensky multiple sclerosis is markedly reduced by (2003). "Foxp3 programs the development and natalizumab." Ann Neurol. function of CD4+CD25+ regulatory T cells." Nat Immunol 4(4): 330-336. Haas, J., I. Bekeredjian-Ding, M. Milkova, B. Balint, A. Schwarz, M. Korporal, S. Jarius, B. Fritz, Freedman, M. S., E. J. Thompson, F. H. M. Lorenz and B. Wildemann (2011). "B cells Deisenhammer, G. Giovannoni, G. Grimsley, G. undergo unique compartmentalized redistribution in Keir, S. Ohman, M. K. Racke, M. Sharief, C. J. multiple sclerosis." J Autoimmun 37(4): 289-299. Sindic, F. Sellebjerg and W. W. Tourtellotte (2005). "Recommended standard of cerebrospinal fluid Haas, J., A. Hug, A. Viehover, B. Fritzsching, C. S. analysis in the diagnosis of multiple sclerosis: a Falk, A. Filser, T. Vetter, L. Milkova, M. Korporal, consensus statement." Arch Neurol 62(6): 865-870. B. Fritz, B. Storch-Hagenlocher, P. H. Krammer, E. Suri-Payer and B. Wildemann (2005). "Reduced Friese, M. A. and L. Fugger (2009). "Pathogenic suppressive effect of CD4+CD25high regulatory T CD8(+) T cells in multiple sclerosis." Ann Neurol cells on the T cell immune response against myelin 66(2): 132-141. oligodendrocyte glycoprotein in patients with Frisullo, G., R. Iorio, D. Plantone, A. Marti, V. multiple sclerosis." Eur J Immunol 35(11): 3343- Nociti, A. K. Patanella and A. P. Batocchi (2011). 3352. "CD4+T-bet+, CD4+pSTAT3+ and CD8+T-bet+ T Hallin, E., J. Mellergard, M. Vrethem, J. Ernerudh cells accumulate in peripheral blood during NZB and C. Ekerfelt (2006). "In vitro Th2 deviation of treatment." Mult Scler 17(5): 556-566. myelin-specific peripheral blood lymphocytes from Fujisawa, T., R. Fujisawa, Y. Kato, T. Nakayama, patients with multiple sclerosis." J Neuroimmunol A. Morita, H. Katsumata, H. Nishimori, K. Iguchi, 171(1-2): 156-162. H. Kamiya, P. W. Gray, D. Chantry, R. Suzuki and Harrer, A., G. Pilz, M. Einhaeupl, K. Oppermann, O. Yoshie (2002). "Presence of high contents of W. Hitzl, P. Wipfler, J. Sellner, S. Golaszewski, S. thymus and activation-regulated chemokine in Afazel, E. Haschke-Becher, E. Trinka and J. Kraus platelets and elevated plasma levels of thymus and (2012). "Lymphocyte subsets show different activation-regulated chemokine and macrophage- response patterns to in vivo bound natalizumab--a derived chemokine in patients with atopic flow cytometric study on patients with multiple dermatitis." J Allergy Clin Immunol 110(1): 139- sclerosis." PLoS One 7(2): e31784. 146. Harrer, A., P. Wipfler, M. Einhaeupl, G. Pilz, K. Gandhi, R., A. Laroni and H. L. Weiner (2010). Oppermann, W. Hitzl, S. Afazel, E. Haschke- "Role of the innate immune system in the Becher, P. Strasser, E. Trinka and J. Kraus (2011). pathogenesis of multiple sclerosis." J "Natalizumab therapy decreases surface expression Neuroimmunol 221(1-2): 7-14. of both VLA-heterodimer subunits on peripheral Ganusov, V. V. and R. J. De Boer (2007). "Do most blood mononuclear cells." J Neuroimmunol 234(1- lymphocytes in humans really reside in the gut?" 2): 148-154. Trends Immunol 28(12): 514-518. He, J., M. Inglese, B. S. Li, J. S. Babb, R. I. Giovannoni, G. (2006). "Multiple sclerosis Grossman and O. Gonen (2005). "Relapsing- cerebrospinal fluid biomarkers." Dis Markers 22(4): remitting multiple sclerosis: metabolic abnormality 187-196. in nonenhancing lesions and normal-appearing white matter at MR imaging: initial experience." Gonzalez-Amaro, R., M. Mittelbrunn and F. Radiology 234(1): 211-217. Sanchez-Madrid (2005). "Therapeutic anti-integrin

10. References

Heimer, L. (1995). The Human Brain and Spinal Ellinghaus, I. Elovaara, F. Esposito, C. Fontenille, Cord. Functional Neuroanatomy and Dissection S. Foote, A. Franke, D. Galimberti, A. Ghezzi, J. Guide. New York, Springer-Verlag. Glessner, R. Gomez, O. Gout, C. Graham, S. F. Grant, F. R. Guerini, H. Hakonarson, P. Hall, A. Hemmer, B., J. J. Archelos and H. P. Hartung Hamsten, H. P. Hartung, R. N. Heard, S. Heath, J. (2002). "New concepts in the immunopathogenesis Hobart, M. Hoshi, C. Infante-Duarte, G. Ingram, of multiple sclerosis." Nat Rev Neurosci 3(4): 291- W. Ingram, T. Islam, M. Jagodic, M. Kabesch, A. 301. G. Kermode, T. J. Kilpatrick, C. Kim, N. Klopp, K. Henderson, A. P., M. H. Barnett, J. D. Parratt and J. Koivisto, M. Larsson, M. Lathrop, J. S. Lechner- W. Prineas (2009). "Multiple sclerosis: distribution Scott, M. A. Leone, V. Leppa, U. Liljedahl, I. L. of inflammatory cells in newly forming lesions." Bomfim, R. R. Lincoln, J. Link, J. Liu, A. R. Ann Neurol 66(6): 739-753. Lorentzen, S. Lupoli, F. Macciardi, T. Mack, M. Marriott, V. Martinelli, D. Mason, J. L. McCauley, Hijnen, D., M. De Bruin-Weller, B. Oosting, C. F. Mentch, I. L. Mero, T. Mihalova, X. Montalban, Lebre, E. De Jong, C. Bruijnzeel-Koomen and E. J. Mottershead, K. M. Myhr, P. Naldi, W. Ollier, A. Knol (2004). "Serum thymus and activation- Page, A. Palotie, J. Pelletier, L. Piccio, T. regulated chemokine (TARC) and cutaneous T cell- Pickersgill, F. Piehl, S. Pobywajlo, H. L. Quach, P. attracting chemokine (CTACK) levels in allergic P. Ramsay, M. Reunanen, R. Reynolds, J. D. diseases: TARC and CTACK are disease-specific Rioux, M. Rodegher, S. Roesner, J. P. Rubio, I. M. markers for atopic dermatitis." J Allergy Clin Ruckert, M. Salvetti, E. Salvi, A. Santaniello, C. A. Immunol 113(2): 334-340. Schaefer, S. Schreiber, C. Schulze, R. J. Scott, F. Hobart, J., D. Lamping, R. Fitzpatrick, A. Riazi and Sellebjerg, K. W. Selmaj, D. Sexton, L. Shen, B. A. Thompson (2001). "The Multiple Sclerosis Simms-Acuna, S. Skidmore, P. M. Sleiman, C. Impact Scale (MSIS-29): a new patient-based Smestad, P. S. Sorensen, H. B. Sondergaard, J. outcome measure." Brain 124(Pt 5): 962-973. Stankovich, R. C. Strange, A. M. Sulonen, E. Sundqvist, A. C. Syvanen, F. Taddeo, B. Taylor, J. Ichiyama, K., H. Yoshida, Y. Wakabayashi, T. M. Blackwell, P. Tienari, E. Bramon, A. Tourbah, Chinen, K. Saeki, M. Nakaya, G. Takaesu, S. Hori, M. A. Brown, E. Tronczynska, J. P. Casas, N. A. Yoshimura and T. Kobayashi (2008). "Foxp3 Tubridy, A. Corvin, J. Vickery, J. Jankowski, P. inhibits RORgammat-mediated IL-17A mRNA Villoslada, H. S. Markus, K. Wang, C. G. Mathew, transcription through direct interaction with J. Wason, C. N. Palmer, H. E. Wichmann, R. RORgammat." J Biol Chem 283(25): 17003-17008. Plomin, E. Willoughby, A. Rautanen, J. International Multiple Sclerosis Genetics, C., C. Winkelmann, M. Wittig, R. C. Trembath, J. Wellcome Trust Case Control, S. Sawcer, G. Yaouanq, A. C. Viswanathan, H. Zhang, N. W. Hellenthal, M. Pirinen, C. C. Spencer, N. A. Wood, R. Zuvich, P. Deloukas, C. Langford, A. Patsopoulos, L. Moutsianas, A. Dilthey, Z. Su, C. Duncanson, J. R. Oksenberg, M. A. Pericak-Vance, Freeman, S. E. Hunt, S. Edkins, E. Gray, D. R. J. L. Haines, T. Olsson, J. Hillert, A. J. Ivinson, P. Booth, S. C. Potter, A. Goris, G. Band, A. B. L. De Jager, L. Peltonen, G. J. Stewart, D. A. Oturai, A. Strange, J. Saarela, C. Bellenguez, B. Hafler, S. L. Hauser, G. McVean, P. Donnelly and Fontaine, M. Gillman, B. Hemmer, R. Gwilliam, F. A. Compston (2011). "Genetic risk and a primary Zipp, A. Jayakumar, R. Martin, S. Leslie, S. role for cell-mediated immune mechanisms in Hawkins, E. Giannoulatou, S. D'Alfonso, H. multiple sclerosis." Nature 476(7359): 214-219. Blackburn, F. Martinelli Boneschi, J. Liddle, H. F. Ivanov, II, B. S. McKenzie, L. Zhou, C. E. Harbo, M. L. Perez, A. Spurkland, M. J. Waller, M. Tadokoro, A. Lepelley, J. J. Lafaille, D. J. Cua and P. Mycko, M. Ricketts, M. Comabella, N. D. R. Littman (2006). "The orphan nuclear receptor Hammond, I. Kockum, O. T. McCann, M. Ban, P. RORgammat directs the differentiation program of Whittaker, A. Kemppinen, P. Weston, C. Hawkins, proinflammatory IL-17+ T helper cells." Cell S. Widaa, J. Zajicek, S. Dronov, N. Robertson, S. J. 126(6): 1121-1133. Bumpstead, L. F. Barcellos, R. Ravindrarajah, R. Abraham, L. Alfredsson, K. Ardlie, C. Aubin, A. Jacobs, L. D., D. L. Cookfair, R. A. Rudick, R. M. Baker, K. Baker, S. E. Baranzini, L. Bergamaschi, Herndon, J. R. Richert, A. M. Salazar, J. S. Fischer, R. Bergamaschi, A. Bernstein, A. Berthele, M. D. E. Goodkin, C. V. Granger, J. H. Simon, J. J. Boggild, J. P. Bradfield, D. Brassat, S. A. Broadley, Alam, D. M. Bartoszak, D. N. Bourdette, J. D. Buck, H. Butzkueven, R. Capra, W. M. Carroll, Braiman, C. M. Brownscheidle, M. E. Coats, S. L. P. Cavalla, E. G. Celius, S. Cepok, R. Chiavacci, F. Cohan, D. S. Dougherty, R. P. Kinkel, M. K. Mass, Clerget-Darpoux, K. Clysters, G. Comi, M. F. E. Munschauer, 3rd, R. L. Priore, P. M. Cossburn, I. Cournu-Rebeix, M. B. Cox, W. Cozen, Pullicino, B. J. Scherokman, R. H. Whitham and et B. A. Cree, A. H. Cross, D. Cusi, M. J. Daly, E. al. (1996). "Intramuscular interferon beta-1a for Davis, P. I. de Bakker, M. Debouverie, B. disease progression in relapsing multiple sclerosis. D'Hooghe M, K. Dixon, R. Dobosi, B. Dubois, D.

10. References

The Multiple Sclerosis Collaborative Research prospects for treatment-sensitive biomarkers." Eur J Group (MSCRG)." Ann Neurol 39(3): 285-294. Neurol 16(4): 528-536. Johnson, K. P., B. R. Brooks, J. A. Cohen, C. C. Khademi, M., I. Kockum, M. L. Andersson, E. Ford, J. Goldstein, R. P. Lisak, L. W. Myers, H. S. Iacobaeus, L. Brundin, F. Sellebjerg, J. Hillert, F. Panitch, J. W. Rose and R. B. Schiffer (1995). Piehl and T. Olsson (2011). "Cerebrospinal fluid "Copolymer 1 reduces relapse rate and improves CXCL13 in multiple sclerosis: a suggestive disability in relapsing-remitting multiple sclerosis: prognostic marker for the disease course." Mult results of a phase III multicenter, double-blind Scler 17(3): 335-343. placebo-controlled trial. The Copolymer 1 Multiple Kivisakk, P., B. C. Healy, V. Viglietta, F. J. Sclerosis Study Group." Neurology 45(7): 1268- Quintana, M. A. Hootstein, H. L. Weiner and S. J. 1276. Khoury (2009). "Natalizumab treatment is Kaech, S. M. and E. J. Wherry (2007). associated with peripheral sequestration of "Heterogeneity and cell-fate decisions in effector proinflammatory T cells." Neurology 72(22): 1922- and memory CD8+ T cell differentiation during 1930. viral infection." Immunity 27(3): 393-405. Knippenberg, S., E. Peelen, J. Smolders, M. Kappos, L., E. W. Radue, P. O'Connor, C. Polman, Thewissen, P. Menheere, J. W. Cohen Tervaert, R. R. Hohlfeld, P. Calabresi, K. Selmaj, C. Hupperts and J. Damoiseaux (2011). "Reduction in Agoropoulou, M. Leyk, L. Zhang-Auberson and P. IL-10 producing B cells (Breg) in multiple sclerosis Burtin (2010). "A placebo-controlled trial of oral is accompanied by a reduced naive/memory Breg fingolimod in relapsing multiple sclerosis." N Engl ratio during a relapse but not in remission." J J Med 362(5): 387-401. Neuroimmunol 239(1-2): 80-86. Kasper, L. H. and J. Shoemaker (2010). "Multiple Kohm, A. P., P. A. Carpentier, H. A. Anger and S. sclerosis immunology: The healthy immune system D. Miller (2002). "Cutting edge: CD4+CD25+ vs the MS immune system." Neurology 74 Suppl 1: regulatory T cells suppress antigen-specific S2-8. autoreactive immune responses and central nervous system inflammation during active experimental Kastrukoff, L. F., A. Lau, R. Wee, D. Zecchini, R. autoimmune encephalomyelitis." J Immunol White and D. W. Paty (2003). "Clinical relapses of 169(9): 4712-4716. multiple sclerosis are associated with 'novel' valleys in natural killer cell functional activity." J Kotter, M. R., C. Stadelmann and H. P. Hartung Neuroimmunol 145(1-2): 103-114. (2011). "Enhancing remyelination in disease--can we wrap it up?" Brain 134(Pt 7): 1882-1900. Kastrukoff, L. F., N. G. Morgan, D. Zecchini, R. White, A. J. Petkau, J. Satoh and D. W. Paty Krumbholz, M., I. Meinl, T. Kumpfel, R. Hohlfeld (1998). "A role for natural killer cells in the and E. Meinl (2008). "Natalizumab immunopathogenesis of multiple sclerosis." J disproportionately increases circulating pre-B and Neuroimmunol 86(2): 123-133. B cells in multiple sclerosis." Neurology 71(17): 1350-1354. Kaur, G., J. Trowsdale and L. Fugger (2012). "Natural killer cells and their receptors in multiple Kuhlmann, T., G. Lingfeld, A. Bitsch, J. sclerosis." Brain. Schuchardt and W. Bruck (2002). "Acute axonal damage in multiple sclerosis is most extensive in Kebir, H., K. Kreymborg, I. Ifergan, A. Dodelet- early disease stages and decreases over time." Brain Devillers, R. Cayrol, M. Bernard, F. Giuliani, N. 125(Pt 10): 2202-2212. Arbour, B. Becher and A. Prat (2007). "Human TH17 lymphocytes promote blood-brain barrier Kurtzke, J. F. (1983). "Rating neurologic disruption and central nervous system impairment in multiple sclerosis: an expanded inflammation." Nat Med 13(10): 1173-1175. disability status scale (EDSS)." Neurology 33(11): 1444-1452. Kessel, A., T. Haj, R. Peri, A. Snir, D. Melamed, E. Sabo and E. Toubi (2012). "Human La Mantia, L., L. Vacchi, M. Rovaris, C. Di CD19(+)CD25(high) B regulatory cells suppress Pietrantonj, G. Ebers, S. Fredrikson and G. Filippini proliferation of CD4(+) T cells and enhance Foxp3 (2012). "Interferon beta for secondary progressive and CTLA-4 expression in T-regulatory cells." multiple sclerosis: a systematic review." J Neurol Autoimmun Rev 11(9): 670-677. Neurosurg Psychiatry. Khademi, M., L. Bornsen, F. Rafatnia, M. Langrish, C. L., Y. Chen, W. M. Blumenschein, J. Andersson, L. Brundin, F. Piehl, F. Sellebjerg and Mattson, B. Basham, J. D. Sedgwick, T. T. Olsson (2009). "The effects of natalizumab on McClanahan, R. A. Kastelein and D. J. Cua (2005). inflammatory mediators in multiple sclerosis: "IL-23 drives a pathogenic T cell population that

10. References induces autoimmune inflammation." J Exp Med McDonald, W. I., A. Compston, G. Edan, D. 201(2): 233-240. Goodkin, H. P. Hartung, F. D. Lublin, H. F. McFarland, D. W. Paty, C. H. Polman, S. C. Larochelle, C., J. I. Alvarez and A. Prat (2011). Reingold, M. Sandberg-Wollheim, W. Sibley, A. "How do immune cells overcome the blood-brain Thompson, S. van den Noort, B. Y. Weinshenker barrier in multiple sclerosis?" FEBS Lett 585(23): and J. S. Wolinsky (2001). "Recommended 3770-3780. diagnostic criteria for multiple sclerosis: guidelines Li, B. S., H. Wang and O. Gonen (2003). from the International Panel on the diagnosis of "Metabolite ratios to assumed stable creatine level multiple sclerosis." Ann Neurol 50(1): 121-127. may confound the quantification of proton brain McFarland, H. F. and R. Martin (2007). "Multiple MR spectroscopy." Magn Reson Imaging 21(8): sclerosis: a complicated picture of autoimmunity." 923-928. Nat Immunol 8(9): 913-919. Liao, W., J. X. Lin and W. J. Leonard (2011). "IL-2 Mehling, M., V. Brinkmann, J. Antel, A. Bar-Or, N. family cytokines: new insights into the complex Goebels, C. Vedrine, C. Kristofic, J. Kuhle, R. L. roles of IL-2 as a broad regulator of T helper cell Lindberg and L. Kappos (2008). "FTY720 therapy differentiation." Curr Opin Immunol 23(5): 598- exerts differential effects on T cell subsets in 604. multiple sclerosis." Neurology 71(16): 1261-1267. Lindberg, R. L., L. Achtnichts, F. Hoffmann, J. Meinl, E., M. Krumbholz, T. Derfuss, A. Junker Kuhle and L. Kappos (2008). "Natalizumab alters and R. Hohlfeld (2008). "Compartmentalization of transcriptional expression profiles of blood cell inflammation in the CNS: a major mechanism subpopulations of multiple sclerosis patients." J driving progressive multiple sclerosis." J Neurol Sci Neuroimmunol 194(1-2): 153-164. 274(1-2): 42-44. Lobb, R. R. and M. E. Hemler (1994). "The Mendel, I. and E. M. Shevach (2006). "Activated T pathophysiologic role of alpha 4 integrins in vivo." cells express the OX40 ligand: requirements for J Clin Invest 94(5): 1722-1728. induction and costimulatory function." Immunology Luster, A. D., R. Alon and U. H. von Andrian 117(2): 196-204. (2005). "Immune cell migration in inflammation: Middeldorp, J. and E. M. Hol (2011). "GFAP in present and future therapeutic targets." Nat health and disease." Prog Neurobiol 93(3): 421-443. Immunol 6(12): 1182-1190. Miller, D. H., F. Barkhof, J. A. Frank, G. J. Parker Mader, I., W. Roser, L. Kappos, G. Hagberg, J. and A. J. Thompson (2002). "Measurement of Seelig, E. W. Radue and W. Steinbrich (2000). atrophy in multiple sclerosis: pathological basis, "Serial proton MR spectroscopy of contrast- methodological aspects and clinical relevance." enhancing multiple sclerosis plaques: absolute Brain 125(Pt 8): 1676-1695. metabolic values over 2 years during a clinical pharmacological study." AJNR Am J Neuroradiol Miller, D. H., D. T. Chard and O. Ciccarelli (2012). 21(7): 1220-1227. "Clinically isolated syndromes." Lancet Neurol 11(2): 157-169. Malmestrom, C., S. Haghighi, L. Rosengren, O. Andersen and J. Lycke (2003). "Neurofilament light Miller, D. H. and S. M. Leary (2007). "Primary- protein and glial fibrillary acidic protein as progressive multiple sclerosis." Lancet Neurol biological markers in MS." Neurology 61(12): 6(10): 903-912. 1720-1725. Miller, D. H., B. G. Weinshenker, M. Filippi, B. L. Matthews, P. M., G. Francis, J. Antel and D. L. Banwell, J. A. Cohen, M. S. Freedman, S. L. Arnold (1991). "Proton magnetic resonance Galetta, M. Hutchinson, R. T. Johnson, L. Kappos, spectroscopy for metabolic characterization of J. Kira, F. D. Lublin, H. F. McFarland, X. plaques in multiple sclerosis." Neurology 41(8): Montalban, H. Panitch, J. R. Richert, S. C. 1251-1256. Reingold and C. H. Polman (2008). "Differential diagnosis of suspected multiple sclerosis: a Matusevicius, D., P. Kivisakk, B. He, N. Kostulas, consensus approach." Mult Scler 14(9): 1157-1174. V. Ozenci, S. Fredrikson and H. Link (1999). "Interleukin-17 mRNA expression in blood and Miossec, P., T. Korn and V. K. Kuchroo (2009). CSF mononuclear cells is augmented in multiple "Interleukin-17 and type 17 helper T cells." N Engl sclerosis." Mult Scler 5(2): 101-104. J Med 361(9): 888-898. Matute, C., M. Domercq and M. V. Sanchez- Mjosberg, J., J. Svensson, E. Johansson, L. Gomez (2006). "Glutamate-mediated glial injury: Hellstrom, R. Casas, M. C. Jenmalm, R. Boij, L. mechanisms and clinical importance." Glia 53(2): Matthiesen, J. I. Jonsson, G. Berg and J. Ernerudh 212-224. (2009). "Systemic reduction of functionally suppressive CD4dimCD25highFoxp3+ Tregs in

10. References human second trimester pregnancy is induced by Ouyang, W., S. H. Ranganath, K. Weindel, D. progesterone and 17beta-estradiol." J Immunol Bhattacharya, T. L. Murphy, W. C. Sha and K. M. 183(1): 759-769. Murphy (1998). "Inhibition of Th1 development mediated by GATA-3 through an IL-4-independent Moretta, A., E. Marcenaro, S. Parolini, G. Ferlazzo mechanism." Immunity 9(5): 745-755. and L. Moretta (2008). "NK cells at the interface between innate and adaptive immunity." Cell Death Panitch, H. S., R. L. Hirsch, A. S. Haley and K. P. Differ 15(2): 226-233. Johnson (1987). "Exacerbations of multiple sclerosis in patients treated with gamma interferon." Murphy, K. M. and S. L. Reiner (2002). "The Lancet 1(8538): 893-895. lineage decisions of helper T cells." Nat Rev Immunol 2(12): 933-944. Paty, D. W. and D. K. Li (1993). "Interferon beta- 1b is effective in relapsing-remitting multiple Niino, M., C. Bodner, M. L. Simard, S. Alatab, D. sclerosis. II. MRI analysis results of a multicenter, Gano, H. J. Kim, M. Trigueiro, D. Racicot, C. randomized, double-blind, placebo-controlled trial. Guerette, J. P. Antel, A. Fournier, F. Grand'Maison UBC MS/MRI Study Group and the IFNB Multiple and A. Bar-Or (2006). "Natalizumab effects on Sclerosis Study Group." Neurology 43(4): 662-667. immune cell responses in multiple sclerosis." Ann Neurol 59(5): 748-754. Petermann, F. and T. Korn (2011). "Cytokines and effector T cell subsets causing autoimmune CNS Norgren, N., P. Sundstrom, A. Svenningsson, L. disease." FEBS Lett 585(23): 3747-3757. Rosengren, T. Stigbrand and M. Gunnarsson (2004). "Neurofilament and glial fibrillary acidic Planas, R., I. Jelcic, S. Schippling, R. Martin and protein in multiple sclerosis." Neurology 63(9): M. Sospedra (2012). "Natalizumab treatment 1586-1590. perturbs memory- and marginal zone-like B-cell homing in secondary lymphoid organs in multiple Obermeier, B., L. Lovato, R. Mentele, W. Bruck, I. sclerosis." Eur J Immunol 42(3): 790-798. Forne, A. Imhof, F. Lottspeich, K. W. Turk, S. N. Willis, H. Wekerle, R. Hohlfeld, D. A. Hafler, K. Poli, A., T. Michel, M. Theresine, E. Andres, F. C. O'Connor and K. Dornmair (2011). "Related B Hentges and J. Zimmer (2009). "CD56bright cell clones that populate the CSF and CNS of natural killer (NK) cells: an important NK cell patients with multiple sclerosis produce CSF subset." Immunology 126(4): 458-465. immunoglobulin." J Neuroimmunol 233(1-2): 245- Pollinger, B., G. Krishnamoorthy, K. Berer, H. 248. Lassmann, M. R. Bosl, R. Dunn, H. S. Domingues, Ochoa-Reparaz, J., D. W. Mielcarz, S. Begum- A. Holz, F. C. Kurschus and H. Wekerle (2009). Haque and L. H. Kasper (2011). "Gut, bugs, and "Spontaneous relapsing-remitting EAE in the SJL/J brain: role of commensal bacteria in the control of mouse: MOG-reactive transgenic T cells recruit central nervous system disease." Ann Neurol 69(2): endogenous MOG-specific B cells." J Exp Med 240-247. 206(6): 1303-1316. Odoardi, F., C. Sie, K. Streyl, V. K. Ulaganathan, Polman, C. H., P. W. O'Connor, E. Havrdova, M. C. Schlager, D. Lodygin, K. Heckelsmiller, W. Hutchinson, L. Kappos, D. H. Miller, J. T. Phillips, Nietfeld, J. Ellwart, W. E. Klinkert, C. Lottaz, M. F. D. Lublin, G. Giovannoni, A. Wajgt, M. Toal, F. Nosov, V. Brinkmann, R. Spang, H. Lehrach, M. Lynn, M. A. Panzara and A. W. Sandrock (2006). Vingron, H. Wekerle, C. Flugel-Koch and A. Flugel "A randomized, placebo-controlled trial of (2012). "T cells become licensed in the lung to natalizumab for relapsing multiple sclerosis." N enter the central nervous system." Nature Engl J Med 354(9): 899-910. 488(7413): 675-679. Polman, C. H., S. C. Reingold, B. Banwell, M. Ohshima, Y., L. P. Yang, T. Uchiyama, Y. Tanaka, Clanet, J. A. Cohen, M. Filippi, K. Fujihara, E. P. Baum, M. Sergerie, P. Hermann and G. Havrdova, M. Hutchinson, L. Kappos, F. D. Lublin, Delespesse (1998). "OX40 costimulation enhances X. Montalban, P. O'Connor, M. Sandberg- interleukin-4 (IL-4) expression at priming and Wollheim, A. J. Thompson, E. Waubant, B. promotes the differentiation of naive human Weinshenker and J. S. Wolinsky (2011). CD4(+) T cells into high IL-4-producing effectors." "Diagnostic criteria for multiple sclerosis: 2010 Blood 92(9): 3338-3345. revisions to the McDonald criteria." Ann Neurol 69(2): 292-302. Olsson, A., H. Vanderstichele, N. Andreasen, G. De Meyer, A. Wallin, B. Holmberg, L. Rosengren, E. Provencher, S. W. (1993). "Estimation of Vanmechelen and K. Blennow (2005). metabolite concentrations from localized in vivo "Simultaneous measurement of beta-amyloid(1-42), proton NMR spectra." Magn Reson Med 30(6): total tau, and phosphorylated tau (Thr181) in 672-679. cerebrospinal fluid by the xMAP technology." Clin Chem 51(2): 336-345.

10. References

Puig-Kroger, A., F. Sanz-Rodriguez, N. Longo, P. Liguori, E. Cocco, M. G. Marrosu, F. Tesser, M. A. Sanchez-Mateos, L. Botella, J. Teixido, C. Leone, A. Weber, F. Zipp, B. Miterski, J. T. Bernabeu and A. L. Corbi (2000). "Maturation- Epplen, A. Oturai, P. S. Sorensen, E. G. Celius, N. dependent expression and function of the CD49d T. Lara, X. Montalban, P. Villoslada, A. M. Silva, integrin on monocyte-derived human dendritic M. Marta, I. Leite, B. Dubois, J. Rubio, H. cells." J Immunol 165(8): 4338-4345. Butzkueven, T. Kilpatrick, M. P. Mycko, K. W. Selmaj, M. E. Rio, M. Sa, G. Salemi, G. Savettieri, Racke, M. K., A. Bonomo, D. E. Scott, B. Cannella, J. Hillert and D. A. Compston (2005). "Multiple A. Levine, C. S. Raine, E. M. Shevach and M. Sclerosis Severity Score: using disability and Rocken (1994). "Cytokine-induced immune disease duration to rate disease severity." deviation as a therapy for inflammatory Neurology 64(7): 1144-1151. autoimmune disease." J Exp Med 180(5): 1961- 1966. Sajja, B. R., J. S. Wolinsky and P. A. Narayana (2009). "Proton magnetic resonance spectroscopy in Racke, M. K., A. E. Lovett-Racke and N. J. multiple sclerosis." Neuroimaging Clin N Am Karandikar (2010). "The mechanism of action of 19(1): 45-58. glatiramer acetate treatment in multiple sclerosis." Neurology 74 Suppl 1: S25-30. Sakaguchi, S., M. Miyara, C. M. Costantino and D. A. Hafler (2010). "FOXP3+ regulatory T cells in Raine, M., Tourtellotte, Ed. (1997). Multiple the human immune system." Nat Rev Immunol Sclerosis. Clinical and pathogenetic basis. London, 10(7): 490-500. Chapman and Hall Medical. Sancho, D., M. Gomez and F. Sanchez-Madrid Raine, M. F., Tourtellotte, Ed. (1997). Multiple (2005). "CD69 is an immunoregulatory molecule Sclerosis. Clinical and pathogenetic basis. London, induced following activation." Trends Immunol Chapman and Hall Medical. 26(3): 136-140. Ramos-Cejudo, J., C. Oreja-Guevara, L. Stark Saraste, M., H. Irjala and L. Airas (2007). Aroeira, L. Rodriguez de Antonio, B. Chamorro "Expansion of CD56Bright natural killer cells in the and E. Diez-Tejedor (2011). "Treatment with peripheral blood of multiple sclerosis patients natalizumab in relapsing-remitting multiple treated with interferon-beta." Neurol Sci 28(3): sclerosis patients induces changes in inflammatory 121-126. mechanism." J Clin Immunol 31(4): 623-631. Sato, T., K. Tachibana, Y. Nojima, N. D'Avirro and Rice, G. P., H. P. Hartung and P. A. Calabresi C. Morimoto (1995). "Role of the VLA-4 molecule (2005). "Anti-alpha4 integrin therapy for multiple in T cell costimulation. Identification of the sclerosis: mechanisms and rationale." Neurology tyrosine phosphorylation pattern induced by the 64(8): 1336-1342. ligation of VLA-4." J Immunol 155(6): 2938-2947. Rosati, G. (2001). "The prevalence of multiple Simon, J. H., R. P. Kinkel, L. Jacobs, L. Bub and N. sclerosis in the world: an update." Neurol Sci 22(2): Simonian (2000). "A Wallerian degeneration 117-139. pattern in patients at risk for MS." Neurology 54(5): Rosengren, L. E., J. E. Karlsson, J. O. Karlsson, L. 1155-1160. I. Persson and C. Wikkelso (1996). "Patients with Skarica, M., C. Eckstein, K. A. Whartenby and P. amyotrophic lateral sclerosis and other A. Calabresi (2011). "Novel mechanisms of neurodegenerative diseases have increased levels of immune modulation of natalizumab in multiple neurofilament protein in CSF." J Neurochem 67(5): sclerosis patients." J Neuroimmunol 235(1-2): 70- 2013-2018. 76. Rosengren, L. E., C. Wikkelso and L. Hagberg Skoog, B., B. Runmarker, S. Winblad, S. Ekholm (1994). "A sensitive ELISA for glial fibrillary and O. Andersen (2012). "A representative cohort acidic protein: application in CSF of adults." J of patients with non-progressive multiple sclerosis Neurosci Methods 51(2): 197-204. at the age of normal life expectancy." Brain 135(Pt Rosenman, S. J., P. Shrikant, L. Dubb, E. N. 3): 900-911. Benveniste and R. M. Ransohoff (1995). Smith, A. (1991). Los Angeles, Western "Cytokine-induced expression of vascular cell Psychological services. adhesion molecule-1 (VCAM-1) by astrocytes and astrocytoma cell lines." J Immunol 154(4): 1888- Sorensen, T. L., M. Tani, J. Jensen, V. Pierce, C. 1899. Lucchinetti, V. A. Folcik, S. Qin, J. Rottman, F. Sellebjerg, R. M. Strieter, J. L. Frederiksen and R. Roxburgh, R. H., S. R. Seaman, T. Masterman, A. M. Ransohoff (1999). "Expression of specific E. Hensiek, S. J. Sawcer, S. Vukusic, I. Achiti, C. chemokines and chemokine receptors in the central Confavreux, M. Coustans, E. le Page, G. Edan, G. V. McDonnell, S. Hawkins, M. Trojano, M.

10. References nervous system of multiple sclerosis patients." J transection in the lesions of multiple sclerosis." N Clin Invest 103(6): 807-815. Engl J Med 338(5): 278-285. Sospedra, M. and R. Martin (2005). "Immunology Ubogu, E. E., M. B. Cossoy and R. M. Ransohoff of multiple sclerosis." Annu Rev Immunol 23: 683- (2006). "The expression and function of 747. chemokines involved in CNS inflammation." Trends Pharmacol Sci 27(1): 48-55. Sriram, S. (2011). "Role of glial cells in innate immunity and their role in CNS demyelination." J Weinberg, A. D., D. N. Bourdette, T. J. Sullivan, Neuroimmunol 239(1-2): 13-20. M. Lemon, J. J. Wallin, R. Maziarz, M. Davey, F. Palida, W. Godfrey, E. Engleman, R. J. Fulton, H. Stenner, M. P., A. Waschbisch, D. Buck, S. Doerck, Offner and A. A. Vandenbark (1996). "Selective H. Einsele, K. V. Toyka and H. Wiendl (2008). depletion of myelin-reactive T cells with the anti- "Effects of natalizumab treatment on Foxp3+ T OX-40 antibody ameliorates autoimmune regulatory cells." PLoS One 3(10): e3319. encephalomyelitis." Nat Med 2(2): 183-189. Strioga, M., V. Pasukoniene and D. Characiejus Weiner, H. L. (2009). "The challenge of multiple (2011). "CD8+ CD28- and CD8+ CD57+ T cells sclerosis: how do we cure a chronic heterogeneous and their role in health and disease." Immunology disease?" Ann Neurol 65(3): 239-248. 134(1): 17-32. White, M. K. and K. Khalili (2011). "Pathogenesis Stuve, O., C. M. Marra, K. R. Jerome, L. Cook, P. of progressive multifocal leukoencephalopathy-- D. Cravens, S. Cepok, E. M. Frohman, J. T. revisited." J Infect Dis 203(5): 578-586. Phillips, G. Arendt, B. Hemmer, N. L. Monson and M. K. Racke (2006). "Immune surveillance in Viglietta, V., C. Baecher-Allan, H. L. Weiner and multiple sclerosis patients treated with D. A. Hafler (2004). "Loss of functional natalizumab." Ann Neurol 59(5): 743-747. suppression by CD4+CD25+ regulatory T cells in patients with multiple sclerosis." J Exp Med 199(7): Svahn, A., A. Linde, R. Thorstensson, K. Karlen, L. 971-979. Andersson and H. Gaines (2003). "Development and evaluation of a flow-cytometric assay of Vignali, D. A. (2000). "Multiplexed particle-based specific cell-mediated immune response in flow cytometric assays." J Immunol Methods activated whole blood for the detection of cell- 243(1-2): 243-255. mediated immunity against varicella-zoster virus." J Yamashita, U. and E. Kuroda (2002). "Regulation Immunol Methods 277(1-2): 17-25. of macrophage-derived chemokine (MDC, CCL22) Szabo, S. J., S. T. Kim, G. L. Costa, X. Zhang, C. production." Crit Rev Immunol 22(2): 105-114. G. Fathman and L. H. Glimcher (2000). "A novel Yednock, T. A., C. Cannon, L. C. Fritz, F. Sanchez- transcription factor, T-bet, directs Th1 lineage Madrid, L. Steinman and N. Karin (1992). commitment." Cell 100(6): 655-669. "Prevention of experimental autoimmune Tan, C. S. and I. J. Koralnik (2010). "Progressive encephalomyelitis by antibodies against alpha 4 multifocal leukoencephalopathy and other disorders beta 1 integrin." Nature 356(6364): 63-66. caused by JC virus: clinical features and Zamvil, S. S. and L. Steinman (1990). "The T pathogenesis." Lancet Neurol 9(4): 425-437. lymphocyte in experimental allergic Teunissen, C. E., C. Dijkstra and C. Polman (2005). encephalomyelitis." Annu Rev Immunol 8: 579- "Biological markers in CSF and blood for axonal 621. degeneration in multiple sclerosis." Lancet Neurol Zheng, W. and R. A. Flavell (1997). "The 4(1): 32-41. transcription factor GATA-3 is necessary and Teunissen, C. E., E. Iacobaeus, M. Khademi, L. sufficient for Th2 cytokine gene expression in CD4 Brundin, N. Norgren, M. J. Koel-Simmelink, M. T cells." Cell 89(4): 587-596. Schepens, F. Bouwman, H. A. Twaalfhoven, H. J. Ziegler, S. F., F. Ramsdell and M. R. Alderson Blom, C. Jakobs and C. D. Dijkstra (2009). (1994). "The activation antigen CD69." Stem Cells "Combination of CSF N-acetylaspartate and 12(5): 456-465. neurofilaments in multiple sclerosis." Neurology 72(15): 1322-1329. Zivadinov, R., M. Stosic, J. L. Cox, D. P. Ramasamy and M. G. Dwyer (2008). "The place of Teunissen, C. E. and M. Khalil (2012). conventional MRI and newly emerging MRI "Neurofilaments as biomarkers in multiple techniques in monitoring different aspects of sclerosis." Mult Scler 18(5): 552-556. treatment outcome." J Neurol 255 Suppl 1: 61-74. Trapp, B. D., J. Peterson, R. M. Ransohoff, R. Zohren, F., D. Toutzaris, V. Klarner, H. P. Hartung, Rudick, S. Mork and L. Bo (1998). "Axonal B. Kieseier and R. Haas (2008). "The monoclonal anti-VLA-4 antibody natalizumab mobilizes

10. References

CD34+ hematopoietic progenitor cells in humans." Blood 111(7): 3893-3895. Zozulya, A. L. and H. Wiendl (2008). "The role of regulatory T cells in multiple sclerosis." Nat Clin Pract Neurol 4(7): 384-398.